Prudential Assurance Company Ltd v Commissioners for Her Majesty’s Revenue and Customs [2018] UKSC 39

Trinity Term[2018] UKSC 39On appeal from: [2016] EWCA Civ 376JUDGMENT
Between
Prudential Assurance Company LtdRespondentCommissioners for Her Majesty’s Revenue and CustomsAppellant
Before
Lord ManceLord SumptionLord ReedLord CarnwathLord HodgeDate 25 July 2018Hearing Heard on 20 and 21 February 2018David Ewart QC, Rupert Baldry QC, Andrew Burrows QC (Hon), Barbara Belgrano (Instructed by HMRC Solicitors Office) for AppellantGraham Aaronson QC, Tom Beazley QC, Jonathan Bremner (Instructed by Joseph Hage Aaronson LLP) for Respondent
[1]This is a test case brought against the Commissioners for Her Majesty’s Revenue and Customs (“HMRC”) by the Prudential Assurance Co Ltd (“PAC”). PAC is a typical United Kingdom-resident recipient of dividends on “portfolio” investments overseas, representing less (usually much less) than 10% of the relevant overseas companies’ share capital. The issues originate from two features of the UK tax position in the period 1990 to 1 July 2009. First, throughout that period dividend income received from overseas investments was in principle taxable, subject (as will appear) to certain reliefs. Second, until 6 April 1999 Advance Corporation Tax (“ACT”) was levied on dividends distributed to UK companies’ shareholders. The scope of the issues arising from these features and open on this appeal is, as will appear, itself in some dispute, but the appeal on any view involves a number of conceptually difficult points.[2]The principal issues on this appeal can be summarised as follows: I. Does EU law require a tax credit in respect of overseas dividends to be set by reference to the overseas tax actually paid, or by reference to the foreign nominal tax rate (“FNR”)? II. Is PAC entitled to compound interest in respect of tax which was levied in breach of EU law, on the basis that HMRC were unjustly enriched by the opportunity to use the money in question? III. Subject to HMRC’s being granted permission to argue the point, does a claim in restitution lie to recover lawful ACT which was set against unlawful mainstream corporation tax (“MCT”)? IV. If the answer to (I) is that EU law requires a tax credit to be set by reference to the overseas tax actually paid, PAC seeks permission to crossappeal on the following question: should the charge to corporation tax on the foreign dividend income under Case V of Schedule D (Income and Corporation Taxes Act 1988 (“ICTA”), section 18) (“DV tax”) be disapplied, or should PAC be allowed to rely on FNRs, or on consolidated effective tax rates, as a simplification or proxy for tax actually paid? V. If HMRC are granted permission to argue Issue III, PAC seek permission to cross-appeal on the following questions: (a) where ACT from a pool which includes unlawful and lawful ACT is utilised against an unlawful MCT liability, should the unlawful ACT be treated as a pre-payment of the unlawful MCT liability, or is the ACT so utilised to be treated as partly lawful and partly unlawful; and (b) where domestic franked investment income (“FII”) was carried back to an earlier quarter, is it to be treated as having been applied to relieve the lawful and unlawful ACT pro rata, or only lawful ACT?

Issue I

[3]The first issue - Issue I - arises from the approach adopted by UK law in order to avoid or mitigate double taxation of dividends. It is now clear that this was inconsistent with EU law, but in what precise respects and what is due by way of restitution or compensation are live issues. The inconsistency with EU law arose as follows. Domestically, dividends received by one UK-resident company, the source of which was a distribution made by another UK-resident company, were exempt from tax under section 208 of ICTA. The effect is that corporation tax was only levied once, on the latter company which made the profit out of which it distributed the dividend to the former company.[4]In contrast, dividends received by a UK-resident company, the source of which was an overseas company, were in principle subject to DV tax. But where the UK-resident company controlled a certain percentage of the voting power of the relevant overseas company (typically 10%), certain relief was given for foreign tax paid on the underlying profits out of which such dividends were paid. This was done either pursuant to a double taxation treaty or unilaterally under ICTA, section 790. No relief against DV tax was however afforded in respect of “portfolio” investments, that is investments involving lesser percentage holdings.[5]In Metallgesellschaft Ltd v Inland Revenue Comrs; Hoechst v Inland Revenue Comrs (Joined Cases C-397/98 and C-410/98) EU:C:2001:134; [2001] ECR I-1727; [2001] Ch 620, the European Court of Justice (“CJEU”) held that the unharmonized domestic tax regime fell under the EC Treaty, and could therefore be challenged if inconsistent with a Treaty provision. Pursuant to a group litigation order dated 30 July 2003, PAC was on 13 November 2003 appointed to conduct the present test case, in which PAC’s primary contention has been that the UK tax position is inconsistent with article 63 of the FEU Treaty.[6]Article 63FEU (ex article 56 of the EC Treaty) provides:
“1. Within the framework of the provisions set out in this Chapter, all restrictions on the movement of capital between member states and between member states and third countries shall be prohibited. 2. Within the framework of the provisions set out in this Chapter, all restrictions on payments between member states and between member states and third countries shall be prohibited.”
[7]At an early stage in the present case, a reference to the CJEU was found necessary. But, before that reference was heard, the CJEU determined a separate UK reference, in Test Claimants in the FII Group Litigation v Inland Revenue Comrs (Case C-446/04) EU:C:2006:774; [2006] ECR I-11753; [2012] 2 AC 436 (“FII ECJ I” - “FII” standing for franked investment income). In it, the CJEU held, at paras 1 and 2 of the operative part:
“1. … where a member state has a system for preventing or mitigating the imposition of a series of charges to tax or economic double taxation as regards dividends paid to residents by resident companies, it must treat dividends paid to residents by non-resident companies in the same way. [The Treaty provisions] do not preclude legislation of a member state which exempts from corporation tax dividends which a resident company receives from another resident company, when that state imposes corporation tax on dividends which a resident company receives from a non-resident company in which the resident company holds at least 10% of the voting rights, while at the same time granting a tax credit in the latter case for the tax actually paid by the company making the distribution in the member state in which it is resident, provided that the rate of tax applied to foreign-sourced dividends is no higher than the rate of tax applied to nationallysourced dividends and that the tax credit is at least equal to the amount paid in the member state of the company making the distribution, up to the limit of the amount of the tax charged in the member state of the company receiving the distribution. Article [63FEU] precludes legislation of a member state which exempts from corporation tax dividends which a resident company receives from another resident company, where that state levies corporation tax on dividends which a resident company receives from a non-resident company in which it holds less than 10% of the voting rights, without granting the company receiving the dividends a tax credit for the tax actually paid by the company making the distribution in the state in which the latter is resident. 2. [The Treaty provisions] preclude legislation of a member state which allows a resident company receiving dividends from another resident company to deduct from the amount which the former company is liable to pay by way of advance corporation tax the amount of that tax paid by the latter company, whereas no such deduction is permitted in the case of a resident company receiving dividends from a non-resident company as regards the corresponding tax on distributed profits paid by the latter company in the state in which it is resident.”
[8]This ruling was re-affirmed in the Reasoned Order by which the CJEU disposed of the reference made by the High Court in the present case: Test Claimants in the CFC and Dividend Group Litigation v Inland Revenue Comrs (Case C201/05) EU:C:2008:239; [2008] ECR I-2875; [2008] STC 1513. The issue of a Reasoned Order, without a formal Advocate General’s opinion and with the same juge rapporteur involved as in FII ECJ I, indicates that the CJEU saw the position as relatively straightforward.[9]In the light of these two decisions of the CJEU, it is common ground that the UK’s treatment of overseas dividends was incompatible with EU law. In a judgment in the present case, Prudential Assurance Co Ltd v Revenue and Customs Comrs [2013] EWHC 3249 (Ch); [2014] STC 1236, Henderson J held (para 148) that the appropriate means of rectifying this was for PAC to be accorded an appropriate tax credit. (This was on the basis that a complete exemption from UK corporation tax would go further than the CJEU had stated that EU law required.) HMRC also accept that PAC is entitled to repayment or restitution of any corporation tax unlawfully charged as a result of the incompatibility: Amministrazione delle Finanze dello Stato v SpA San Giorgio (Case C-199/82) [1983] ECR 3595 (“San Giorgio”). However, the amount to be awarded depends significantly on issues of EU law and domestic law which are either open or which HMRC seek to raise on this appeal.[10]Issue I is whether the credit in respect of overseas dividends should under EU law be set by reference to the overseas tax actually paid, as HMRC submit, or by reference to the foreign nominal tax rate (“FNR”), as PAC submits. HMRC rely in this connection upon the CJEU’s judgments in FII ECJ I and on its Reasoned Order in the present case, as well as upon a further judgment of the CJEU in Haribo Lakritzen Hans Riegel BetriebsgmbH v Finanzamt Linz and Österreichische Salinen AG v Finanzamt Linz (Joined Cases C-436/08 and C-437/08) EU:C:2011:61; [2011] ECR I-355; [2011] STC 917. In all three cases, the juge rapporteur was Judge Lenaerts, now the President of the CJEU. In HMRC’s submission, these cases demonstrate, first, a difference in principle between portfolio investments, such as PAC held, and non-portfolio investments, conferring a significant measure of control, and, secondly, that at any rate in relation to portfolio investments the credit to be imputed to PAC is in respect of the actual tax incurred overseas.[11]In response, PAC relies upon a later CJEU decision in the FII litigation, Test Claimants in the FII Group Litigation v Revenue and Customs Comrs (formerly Inland Revenue Comrs) (Case C-35/11) EU:C:2012:707; [2013] Ch 431 (“FII ECJ II”). Judge Lenaerts was once again the juge rapporteur. In this judgment, PAC submits, the CJEU refined its jurisprudence to require the use of the FNR in respect of all dividends received by PAC from overseas. HMRC in reply point out that FII ECJ II was concerned essentially with non-portfolio dividends, and criticise some aspects of its reasoning, particularly its treatment of Haribo. Finally, HMRC submit that the European legal position is unclear, and requires a further reference to the CJEU.[12]There are further issues which HMRC seek to attach to Issue I. The first, identified before us as “issue 4 CA”, is whether, when considering the relevant overseas tax position, attention should focus on the overseas company directly responsible for the remission of the dividend to the UK (the overseas “water’s edge” company) or on the overseas company (or companies) responsible for generating the profits out of which such dividend was paid and on which it (or they) paid tax overseas. The second issue, which HMRC submit that the Supreme Court should “take into account”, was identified as “issue 6 CA”, and is whether any difference has been shown to exist between the effective rate incurred by domestic companies declaring dividends to PAC and the nominal rate payable by UK companies. This is relevant, HMRC submit, because the existence of such a difference was a reason why the CJEU indicated in FII ECJ II that it was appropriate to give a credit for the FNR, rather than the actual tax, incurred on an overseas dividend. PAC submits that neither of issues 4 CA and 6 CA is open in this court. The Court of Appeal refused permission for either issue to be raised before it, and neither issue is properly part of or essential to the resolution of Issue I.[13]The CJEU in FII ECJ I and in its Reasoned Order in the present case clearly established that the discrimination involved in the UK’s arrangements for taxation Page 7 of dividends sourced domestically and from overseas could be resolved by a mixed system, whereby dividends with a domestic source remained exempt, while credit was given against DV tax for tax actually incurred overseas on dividends received from overseas.[14]HMRC point out that the CJEU in FII ECJ I addressed separately the position of dividends received from non-portfolio and from portfolio companies. In relation to the former, the question arose whether a mixed system of exemption in respect of domestically sourced dividends coupled with a credit in respect of dividends received from overseas was compatible with EU law. The CJEU dealt with this at paras 46 to 57. The claimants drew attention to the situation arising if, under the relevant UK legislation, such an exemption was granted in respect of a nationallysourced dividend received from a company which for some reason had no corporation tax liability or paid corporation tax at a lower rate than the normal UK rate (para 54). The CJEU understood the UK Government to explain that this arose only exceptionally (para 55), and on that basis contented itself with saying (para 56):
“In that respect, it is for the national court to determine whether the tax rates are indeed the same and whether different levels of taxation occur only in certain cases by reason of a change to the tax base as a result of certain exceptional reliefs.”
[15]The inference seems to be that, were a significant difference to exist between the effective rate of tax paid by the UK source of the dividend (eg because of some relief or allowance available to the company which was the source of the dividend) and the nominal rate of tax to which the exemption under section 208 of ICTA applied, then a system of credit in respect of overseas-sourced dividends which limited the tax credit to tax actually paid overseas (ie the effective rate of tax) would not be consistent with EU law - because the overseas-sourced dividend would remain liable for any DV tax chargeable after the credit had been taken into account. In other words, the overseas-sourced dividend would not be enjoying, under the tax credit system, any relief or allowance which had reduced the tax actually paid on it, whereas the UK-sourced dividend would enjoy any such relief or allowance.[16]In respect of portfolio dividends, the CJEU faced a more fundamental objection. The UK system was inherently discriminatory, because it failed to give any credit at all for overseas tax paid (paras 61 to 72). The CJEU gave short shrift to the UK Government’s argument that practical difficulties in ascertaining the tax actually paid justified a different system for portfolio dividends.[17]It does not however follow from the separate treatment of non-portfolio and portfolio holdings in FII ECJ I that the CJEU saw any significant difference between them regarding the manner in which the deficiencies in the UK tax system needed to be addressed. It is true that the judgment in Haribo in February 2011 concerned portfolio dividends; following FII ECJ I and the Reasoned Order in the present case, it spoke of the need to credit tax actually paid. But it was only in FII ECJ II, where the focus was on non-portfolio holdings, that the CJEU identified the FNR as a more relevant criterion in any context. That therefore in no way indicates that the FNR is not also relevant to portfolio investments.[18]As a matter of logic and principle, there seems no basis in this connection for any distinction between portfolio and non-portfolio holdings, when applying the mixed system of domestic exemption coupled with a credit in respect of overseassourced dividends to each. Rather than concentrating on the practical difficulties advanced before the CJEU in FII ECJ II, HMRC now suggest that there are important differences in the approaches and expectations which investors would have with regard to portfolio investments, when compared with non-portfolio investments. There are of course differences between holdings giving a degree of control and smaller holdings, but it is not obvious what relevance they have to the question of central interest on this appeal: that is, the proper treatment of domestically-sourced and overseas-sourced dividends so as to avoid unfair discrimination between them.[19]The CJEU’s change of approach in FII ECJ II arose from correction of the misunderstanding evidenced in paragraph 54 in FII ECJ I. Far from being exceptional, it had been established conclusively that it was commonly the position in the UK that a company’s effective rate of tax was (due for example to group relief, or the carry forward of trading losses or other reasons) less than the nominal tax rate, and on that basis it was held that the UK tax system infringed what is now article 63: Test Claimants in the FII Group Litigation v Revenue and Customs Comrs [2008] EWHC 2893 (Ch); [2009] STC 254 (“FII High Court I”), affirmed [2010] EWCA Civ 103; [2010] STC 1251 (“FII CA”).[20]Pursuant to an order for a further reference made by Henderson J on 20 December 2010, it became necessary for the CJEU to address the implications. The European Commission in written submissions in FII ECJ II said (para 28) that in circumstances where the effective rate borne by the company making the profits from which the dividend came was lower than the nominal rate:
“28. … to exempt domestic dividends (which in effect amounts to giving credit for the full amount of tax at the statutory rate even where this full amount has not been paid) while giving credit only for the actual amount of tax paid in respect of the profits giving rise to foreign dividends results in more favourable treatment for domestic dividends.”
[21]The Commission drew from this (para 29) that:
“29. … It can no longer be said that the credit method is equivalent to exemption, because foreign dividends receive less favourable treatment than domestic dividends. To understand why, let us imagine identical resident and nonresident companies which each have revenues of 100 and have, say, a loss carry-forward of 50. The tax rate is 30% in both the source and residence states. A company which is a shareholder in the resident company and receives a dividend from it will have no further tax obligation, even though that company has paid only 15 in tax (that is, has an effective rate of 15%). A shareholder in the non-resident company will receive a credit equivalent to only 15% and will have to pay an additional 15%. The same result will ensue where both states grant, for example, an identical research and development incentive. That is not equal treatment, and it constitutes a serious obstacle to outward investment.”
[22]The Commission then discussed how the problem might be addressed (paras 31-34):
“31. In such circumstances there seem to the Commission to be two ways of ensuring equal treatment. One is to exempt both domestic and foreign dividends. That solution has the drawback, as outlined above, that it may permit excessively favourable treatment of foreign dividends where the tax rate in the source state is lower than in the United Kingdom. The other, which is wholly consistent with the court’s reasoning in Case C-446/04 [‘FII ECJ I’], is to have regard solely to the nominal rate of tax in calculating the tax credit on foreign dividends. 32. That is to say, recipients of such dividends should receive a tax credit representing the amount which would flow from the application of the nominal rate of tax in the source state to the accounting profits of the distributing company. Such a measure would correspond more truly to the exemption of domestic dividends, since the latter amounts in effect to the grant of a credit for tax at the nominal rate. The court has seen and approved a measure of this kind in Joined Cases C-436/08 and C-437/08 Haribo, judgment of 10 February 2011 … see point 99 of the judgment. It would no doubt be desirable for a member state applying such a measure to insert a safeguard clause limiting its scope to dividends distributed by a company which is subject to the normal system of taxation in the source state. 33. It should be noted that such a measure would also alleviate to a very large extent the administrative burden faced by taxpayers in relation to foreign dividends, especially taxpayers with small shareholdings. 34. Such a solution does not ensure substantive equal treatment in all cases. In particular, where the tax system in the source state is a simple one in which the effective rate is systematically the same as the nominal rate (because the tax base is constituted by accounting profits, with no modifications), foreign dividends will treated less favourably than domestic dividends, since the latter will benefit from any tax advantages enjoyed by the distributing company. However, to ensure full substantive equal treatment would require systematic re-calculation of the tax position of the foreign company - essentially a simulation of the tax which it would have paid were it resident in the United Kingdom. Such an approach seems impractical. The solution advocated by the Commission ensures formal equality of treatment, is easy to apply and achieves a fair result.”
It is worth noting in passing para 33. The Commission evidently had no doubt about the relevance of its proposed solution to overseas portfolio holdings.[23]It was a neat solution which evidently appealed to the juge rapporteur, who (according to the informal transcript with which the Supreme Court has been supplied) put to counsel for the UK a series of points, starting with this “very simple” question:
“… does such an exemption system not in fact come down to as Mr Lyal said, ‘tantamount to’ a credit system applied at the normal rate of tax applicable to the taxing of those dividends with the shareholding company?”
Judge Lenaerts went on to put that an exemption system does more than a system crediting tax actually paid, because “It gives more relief than the tax actually paid by the distributing company and … when you say that you have an exemption system, in fact, you exonerate from any tax liability, you exempt from any tax liability, the shareholding companies at the rate applicable to the taxing of dividends with that shareholding company.” The reasoning was the same as the Commission’s.[24]Subsequently various questions were put by Advocate General Jääskinen to Mr Lyal for the Commission. The Advocate General expressed some doubt whether the Commission’s proposed solution was really consistent with the CJEU’s previous judgment. He suggested that it would seem to work if there was equivalence of both the domestic and the overseas nominal and effective tax rates, but pointed to a risk of distortion if the nominal and effective tax rates were similar in one state, but diverged significantly in the other. Mr Lyal’s response was:
“Yes, my Lord, that’s quite right. That’s a danger. It is a danger that can be minimised, if what I said a moment ago about recognizing only the type of tax benefits or discounts or manners of calculation that are recognized in the state of residence of the parent. The practical likelihood of the problem is to some extent limited to the extent that there is something of a correlation between higher taxes, higher company tax rates and lots of discounts, and equally a correlation between lower company tax rates and broad tax bases. So, if there is a source company in Slovakia, say, which has set their tax at, whatever their rate now is, 17%, perhaps, one can be pretty sure that it’s 17% on accounting profits. And thirdly, this is after all said to be a rough equivalent because the only other practical option that I can see for a rough equivalent of the domestic exemption is exemption for foreign-source dividends as well which would compound the problem that your Lordship refers to. That is to say you would not only have the freedom from taxation represented by the difference between the lower rate, the lower effective rate and [?in] the source state, but also the difference between the statutory rate in the UK and the lower statutory rate. So, again, the problem that your Lordship advances is certainly a correct one, but in circumstances in which the resident state applies an exemption system, we need to find something that gives substantive equivalent taxation for inbound dividends, a measure of which focuses on the actual tax, the effective rate of tax borne by inbound dividends simply is not equivalent.”
[25]Against this background, the CJEU in FII ECJ II addressed the problem as follows. First, it pointed out (paras 43 to 48) that the tax rate applied to foreignsourced dividends would be higher than that applied to domestic-sourced dividends, and their equivalence compromised, if the resident company generating the dividends was subject to either a nominal or an effective rate of tax below that to which the resident company receiving the dividends was subject, since in a case where an overseas company generating dividends was subject to an effective tax rate lower than the UK nominal rate, the difference would be chargeable to DV tax. On this basis, it explained that in its judgment in FII ECJ I, para 56, the reference to the “tax rates” related to the nominal rate of tax while the reference to the “different levels of taxation … by reason of a change to the tax base” related to the effective levels of taxation.[26]The CJEU in FII ECJ II, after stating why such discrimination could not be justified under EU law as necessary to preserve the cohesion of the domestic tax system, then adopted use of the FNR as an acceptable solution in the following paragraphs:
“61. The tax exemption to which a resident company receiving nationally-sourced dividends is entitled is granted irrespective of the effective level of taxation to which the profits out of which the dividends have been paid were subject. That exemption, in so far as it is intended to avoid economic double taxation of distributed profits, is thus based on the assumption that those profits were taxed at the nominal rate of tax in the hands of the company paying dividends. It thus resembles grant of a tax credit calculated by reference to that nominal rate of tax. 62. For the purpose of ensuring the cohesion of the tax system in question, national rules which took account in particular, also under the imputation method, of the nominal rate of tax to which the profits underlying the dividends paid have been subject would be appropriate for preventing the economic double taxation of the distributed profits and for ensuring the internal cohesion of the tax system while being less prejudicial to freedom of establishment and the free movement of capital. 63. It is to be observed in this connection that in the Haribo case [2011] ECR I-305, para 99, the court, after pointing out that the member states are, in principle, allowed to prevent the imposition of a series of charges to tax on dividends received by a resident company by applying the exemption method to nationally-sourced dividends and the imputation method to foreign-sourced dividends, noted that the national rules in question took account, for the purpose of calculating the amount of the tax credit under the imputation method, of the nominal rate of tax applicable in the state where the company paying dividends was established. 64. It is true that calculation, when applying the imputation method, of a tax credit on the basis of the nominal rate of tax to which the profits underlying the dividends paid have been subject may still lead to a less favourable tax treatment of foreign-sourced dividends, as a result in particular of the existence in the member states of different rules relating to determination of the basis of assessment for corporation tax. However, it must be held that, when unfavourable treatment of that kind arises, it results from the exercise in parallel by different member states of their fiscal sovereignty, which is compatible with the Treaty … 65. In light of the foregoing, the answer to the first question is that articles 49FEU and 63FEU must be interpreted as precluding legislation of a member state which applies the exemption method to nationally-sourced dividends and the imputation method to foreign-sourced dividends if it is established, first, that the tax credit to which the company receiving the dividends is entitled under the imputation method is equivalent to the amount of tax actually paid on the profits underlying the distributed dividends and, second, that the effective level of taxation of company profits in the member state concerned is generally lower than the prescribed nominal rate of tax.”
[27]These paragraphs are generally stated, and there is no reason why they should not be as applicable to portfolio holdings as they are to non-portfolio holdings. There is no hint of any distinction between portfolio and non-portfolio holdings. If any had been intended, one would have expected it to be mentioned, particularly when the same juge rapporteur was active in all four of the key decisions, covering between them both types of holding. Instead, in para 63 of FII ECJ II, quoted above, reliance is actually placed on Haribo, a case of portfolio holdings, in support of the use of the FNR. The reliance may, as Mr David Ewart QC for HMRC submitted, be misplaced, achieving a coherence in the development of the jurisprudence that is more apparent than real - because the reference to FNR in Haribo was in the context of a simplified method permitted by the Austrian tax authorities to show the tax actually paid. But that is presently irrelevant. What matters is that the CJEU in FII ECJ II presented its development of the law in the context of non-portfolio holdings as being in line with, and supported by, its previous jurisprudence in the context of portfolio holdings. It is inconceivable that it contemplated any material distinction in the principles applicable to both. It follows that, subject to one reservation, the CJEU’s jurisprudence establishes clearly that the credit for foreign dividends in the present case should be by reference to the FNR, rather than by reference to the actual or effective tax incurred overseas.[28]The one reservation arises from the assumption made throughout the discussion in FII ECJ II, that, in the Commission’s words (para 20 above): “to exempt domestic dividends … in effect amounts to giving credit for the full amount of tax at the statutory rate even where this full amount has not been paid;” or in Judge Lenaerts’ words (para 23 above) that “an exemption system does more than a system crediting tax actually paid by the distributing company”.[29]This assumption is readily understandable, if one also assumes that the domestic company which is the source of and distributes the dividend has an effective tax rate less than the nominal tax rate, while the receiving company which is exempted from tax would, but for the exemption, pay tax at the full nominal rate. There is then a benefit from the exemption, which would have no parallel if the credit in respect of overseas-sourced dividends was by reference only to the actual tax incurred overseas. However, in the UK domestic context, there appears to be no reason to think that companies receiving domestically-sourced dividends are any less able to reduce the effective tax rate they would have borne on such dividends than are the companies from which the dividend is sourced. In other words, the evidence appears to have been that all UK companies are generally taxed at an effective level below the nominal rate. That being so, the domestic exemption does not confer a benefit at the nominal rate, but at their effective rate.[30]It follows that to give a credit for overseas dividends at the FNR may confer a benefit on overseas dividends, compared with domestic dividends. By way of example, given by HMRC, one can suppose an overseas water’s edge company with a nominal tax rate of 20% and an effective rate of 10%, which makes a profit after tax of 100 and distributes a dividend of 100. The UK recipient company has a nominal rate of 30%, but an effective tax rate of 20%. One would expect the UK company to bear a 10% charge on the dividend to reflect the higher tax rate charged in the UK (and that is so, whether one is looking at and comparing the nominal or the effective tax rate in this connection). But a tax credit can necessarily only reduce the tax which would actually otherwise be charged. Here, since the UK company’s effective tax rate is only 20%, the effect of giving credit to the UK company for the FNR of 20% is in fact to eliminate any UK tax charge.[31]Does this reservation about the rationale and solution adopted by the CJEU mean that we should once again refer the case back to the CJEU, for it to reconsider once again whether its approach is appropriate? In our opinion, it does not. It is clear that the CJEU was well aware that the adoption of the FNR would not eliminate all inequities or incongruities: see the Commission’s written observations, para 34, cited in para 22 above, the Advocate General’s question put to Mr Lyal and Mr Lyal’s answer cited in para 24 above and the CJEU’s own judgment, para 64, cited in para 26 above. There could, depending on the incidence of nominal and effective tax rates, be swings and roundabouts in the equivalence achieved by a mixed system of domestic exemption combined with overseas credits. But the “ideal” alternative of a comparison between two tax systems to ensure equivalence (subject only to each state’s right to set its own nominal tax levels) was consciously rejected as wholly impractical. In these circumstances, such inequity as may arise from the reservation discussed in the previous three paragraphs is not in our view a reason for referring the matter yet again to the CJEU. The prospect that the CJEU would, at this stage in history, contemplate revising yet again its jurisprudence appears to us negligible to the point where it can be discarded.[32]We turn finally to the two further issues which HMRC suggest should be taken into account, and should lead to or encourage the making of a reference. The short answer in relation to each is that permission to raise it before the Court of Appeal was specifically refused by that court: [2016] EWCA Civ 376; [2016] STC 1798. In these circumstances, there is no constitutional basis for consideration of either before the Supreme Court: Access to Justice Act 1999, section 54(4), Supreme Court Practice Direction No 1 para 1.2.5. Issue I at first instance might have been wide enough to embrace one or both of issues 4 CA and 6 CA, since it asked inter alia what the appropriate amount of any tax credit required was, but in fact neither issue was raised at the trial before Henderson J. Issue 6 CA was first raised six weeks after the trial by letter to the judge, who refused permission to appeal on it. Issue 4 CA only emerged as ground 2 in HMRC’s Grounds of Appeal to the Court of Appeal. The Court of Appeal expressly refused permission to appeal to it on Issues 4 and 6 CA, and the declaration it made read simply that (para 99):
“… the effect of the rulings of the CJEU is that the foreign dividend should be afforded equivalent treatment, taking the form of the imputation method according to which credit should be given for the relevant foreign tax at the effective rate [ie the actual tax paid] or the nominal rate (whichever is the higher), subject to a cap at the rate of the UK’s nominal rate of ACT [ie the corporation tax rate or the rate of ACT as applicable].”
This declaration does not address or require the Supreme Court to address either of Issues 4 and 6 CA. (In the event, the Supreme Court has not even been asked to address the question whether the Court of Appeal was right to declare that credit required to be given for the higher of the effective and nominal foreign rate, that issue being we were told of no present relevance, but reserved for a further instalment of the FII litigation.)

Issue IV

[33]It is however appropriate at this point to deal with Issue IV which is before the Supreme Court. That is whether, if PAC had no entitlement under EU law to a credit by reference to the FNR, effect should, in the light of what is said to be the impossibility of showing the tax actually borne by its portfolio holdings, be given to the CJEU’s judgments either:(a) by disapplying the DV charge; or(b) by allowing PAC to rely on FNRs (or consolidated effective tax rates) as a simplification or proxy for tax actually paid. The short answer to this issue is that it does not arise or need answering, having regard to our conclusion that PAC is entitled to credit in respect of overseas-sourced dividends by reference to the FNR.

Issue II: Introduction

[34]PAC seeks restitution, on the ground of unjust enrichment, of an amount calculated on the basis of compound interest, in respect of each category of claim which has succeeded. The amounts on which interest is sought, and the periods over which it is submitted that interest should be compounded, are as follows:(a) unlawfully levied ACT which was subsequently set off against lawfully levied MCT, from the date of payment by PAC to the date of setoff;(b) all other unlawfully levied tax (including unlawfully levied ACT which was never set off against lawful MCT, and unlawfully levied ACT which was set off against unlawfully levied MCT), from the date of payment by PAC to the date of repayment by HMRC; and(c) the time value of utilised ACT (resulting from (a) above), from the date of set-off to the date of payment by HMRC. PAC submits that the interest should be compounded at conventional rates calculated by reference to the rates of interest, and rests, applicable to borrowings by the Government in the market during the relevant period, that being the approach favoured by a majority of the House of Lords in Sempra Metals Ltd v Inland Revenue Comrs (formerly Metallgesellschaft Ltd) [2007] UKHL 34; [2008] 1 AC 561.[35]HMRC have accepted that compound interest is payable in respect of the utilised ACT falling within category (a) above, since that is what the House of Lords decided in Sempra Metals. PAC submits that the principles set out in Sempra Metals entail that the same approach should also apply to the amounts falling within categories (b) and (c) above. HMRC, on the other hand, submit that only simple interest should be awarded, in accordance with section 35A of the Senior Courts Act 1981 (“the 1981 Act”), inserted by the Administration of Justice Act 1982, section 15(1) and Schedule 1, Part I. An award of interest on that basis would, they argue, be compatible with the requirement under EU law that PAC should receive an “adequate indemnity”, in accordance with the decision of this court in Littlewoods Ltd v Revenue and Customs Comrs [2017] UKSC 70; [2017] 3 WLR 1401.[36]Although the difference between simple and compound interest is modest in the present case, the point also arises in other cases which are pending against HMRC, and the total amount at stake, on HMRC’s estimate, is of the order of £4-5 billion. The point is also one of considerable importance from a legal perspective, since it raises some fundamental issues in the law of unjust enrichment.

The approach of the courts below

[37]In a careful judgment, Henderson J held that compound interest should be awarded in respect of all three categories of claim, on the basis that, applying the reasoning of the majority in Sempra Metals, PAC was entitled on the ground of unjust enrichment to compound interest on all its claims: [2013] EWHC 3249 (Ch); [2014] STC 1236, paras 242-246. His Lordship subsequently followed that decision in other proceedings, concerned with the recovery of VAT paid under a mistake: Littlewoods Retail Ltd v Revenue and Customs Comrs [2014] EWHC 868 (Ch); [2014] STC 1761, para 417. That paragraph was then expressly approved by the Court of Appeal in the Littlewoods proceedings: Littlewoods Ltd v Revenue and Customs Comrs [2015] EWCA Civ 515; [2016] Ch 373; [2015] STC 2014, paras 203-204. When the present case reached the Court of Appeal, it dismissed the appeal on this issue on the basis that it was bound by its previous decision in the Littlewoods case. So the only detailed consideration of this issue, in the context of the present case, has been that of Henderson J.[38]It is only necessary to add that, following the decision of this court in the Littlewoods proceedings (Littlewoods Ltd v Revenue and Customs Comrs [2017] UKSC 70; [2017] 3 WLR 1401), which reversed the decision of the Court of Appeal, it is no longer argued that there is a right to compound interest under EU law.

A preliminary point

[39]As a preliminary point, PAC submits that HMRC should not be permitted to advance an argument to the effect that the opportunity to use money mistakenly paid is not a benefit obtained at the expense of the person who made the mistaken payment. They point out that HMRC did not advance any argument along these lines at the trial before Henderson J in 2013. On the contrary, it was conceded in advance of the trial that PAC was entitled to compound interest in respect of the claims in category (a), following Sempra Metals, and that position was maintained in the agreed Statement of Facts and Issues. Although HMRC do not seek to withdraw that concession, PAC submits that the proposed argument challenges the reasoning in Sempra Metals, and therefore the basis on which the concession was made. Against this background, it is submitted that HMRC should not be permitted to advance this argument for the first time in this court.[40]In so far as HMRC wish to advance submissions questioning the soundness of the reasoning in Sempra Metals, the court is not inclined to prevent them from doing so, in the particular circumstances of this case. As will be explained, there have been some significant developments in the law of unjust enrichment since the trial before Henderson J, and indeed since the present appeal was brought. In particular, the concept of a benefit being obtained at the expense of the claimant, and the related concept of a transfer of value, were considered by this court only relatively recently. In this appeal, PAC invites the court to extend the reasoning in Sempra Metals beyond the scope of that decision itself, albeit PAC submits that the extension is the logical consequence of that decision. The appeal therefore involves analysing the reasoning in Sempra Metals, and unavoidably requires the court to consider whether that reasoning is consistent with the approach which it has more recently adopted, so as to form part of a coherent body of law. As we explain later, there is indeed a difficulty involved in reconciling Sempra Metals with this court’s more recent case law. Accordingly, even if HMRC had not wished to subject the decision in Sempra Metals to critical analysis, that is an exercise which this court could not have avoided. In addition, it is important to bear in mind that this appeal has to be decided in the context of a group litigation order, and also that the point of law which HMRC wish to argue is undoubtedly one of general public importance.[41]We do not consider that allowing these matters of law to be argued involves unfairness to PAC. The essence of HMRC’s argument was set out in the written case which they submitted in advance of the hearing of the appeal, although, as often happens, the argument was refined during the hearing.

What did Sempra Metals decide?

[42]The issue in Sempra Metals was how effect should be given in domestic law to the judgment of the CJEU in the Metallgesellschaft case (Metallgesellschaft Ltd v Inland Revenue Comrs (Joined Cases C-397/98 and C-410/98) EU:C:2001:134; [2001] Ch 620; [2001] ECR I-1727) (Sempra Metals Ltd being the same company, under a new name, as Metallgesellschaft Ltd). The case concerned claims for compound interest in respect of unlawfully levied ACT which had been set off against lawful MCT: in other words, claims falling within category (a) above. The CJEU made it clear that it was for domestic law to determine the juridical basis of the claims: in particular, whether they lay in restitution or in damages. On the hypothesis that, under domestic law, the appropriate basis was restitution, the CJEU stated:
“87 … In such circumstances, where the breach of Community law arises, not from the payment of the tax itself but from its being levied prematurely, the award of interest represents the ‘reimbursement’ of that which was improperly paid and would appear to be essential in restoring the equal treatment guaranteed by article 52 of the [EC] Treaty. 88. The national court has said that it is in dispute whether English law provides for restitution in respect of damage arising from loss of the use of sums of money where no principal sum is due. It must be stressed that in an action for restitution the principal sum due is none other than the amount of interest which would have been generated by the sum, use of which was lost as a result of the premature levy of the tax.”
Equally, on the hypothesis that the claim properly lay in damages, the argument that the claimants could not be awarded interest could not be accepted.[43]Giving effect to the judgment of the CJEU, the lower courts held that the claimants were entitled to recover compound interest on the ACT in respect of the period between the date of payment and the date of set-off. They also expressed the view, obiter, that the same principles should apply to claims in respect of unutilised ACT, falling within category (b) above.[44]HMRC’s appeal to the House of Lords was dismissed ([2008] AC 561). For varying reasons, the House held, by a majority, that a claim would lie in unjust enrichment for restitution of compound interest on money which had been paid prematurely as the consequence of a mistake, and that the appropriate measure of restitution in the instant case was compound interest calculated on a conventional basis applicable to government borrowing. The House also held that compound interest was available as damages, where it was the measure of the loss foreseeably suffered by the claimant from the loss of the use of his funds. That aspect of the decision is not in issue in the present case and need not be considered.[45]Lord Nicholls of Birkenhead and Lord Hope of Craighead, who were in the majority on the question of unjust enrichment, emphasised that the interest was not ancillary to a claim for the recovery of a principal sum: rather, the interest was itself the principal sum, claimed as restitution of the time value of money. They interpreted the CJEU’s judgment in Metallgesellschaft as meaning that EU law required, as Lord Hope put it at para 9, that “the companies must be provided with a remedy in domestic law which will enable them to recover a sum equal to the interest which would have been generated by the advance payments from the date of the payment of the ACT until the date on which the MCT became chargeable”: see also, to the same effect, the speech of Lord Nicholls at para 60. In that regard, both Lord Hope and Lord Nicholls referred to para 88 of the judgment of the CJEU, cited above. Lord Nicholls identified the crux of the dispute, at paras 71-73, as being whether the provision English law made for the payment of interest satisfied the EU principle of effectiveness.[46]Lord Hope and Lord Nicholls adopted similar analyses of the basis of the claim in unjust enrichment, at paras 33 and 102 respectively. Lord Hope described the Revenue’s enrichment at para 33 as “the opportunity to turn the money to account during the period of the enrichment”. Lord Nicholls analysed the issue in terms of Professor Birks’s theory of unjust enrichment by subtraction (that is, at the expense of the claimant), and stated at para 102:
“The benefits transferred by Sempra to the Inland Revenue comprised, in short, (1) the amounts of tax paid to the Inland Revenue and, consequentially, (2) the opportunity for the Inland Revenue, or the Government of which the Inland Revenue is a department, to use this money for the period of prematurity. The Inland Revenue was enriched by the latter head in addition to the former. The payment of ACT was the equivalent of a massive interest free loan. Restitution, if it is to be complete, must encompass both heads. Restitution by the Revenue requires (1) repayment of the amounts of tax paid prematurely (this claim became spent once set off occurred) and (2) payment for having the use of the money for the period of prematurity.”
[47]Since the enrichment which had to be undone was the opportunity to turn the money to account during the period before it was lawfully due, it followed that the measure of the enrichment did not depend on what HMRC actually did with the money during that period (Lord Hope at para 33, Lord Nicholls at para 117). In that connection, Lord Nicholls drew an analogy at para 116 with the award of “user damages”, although such awards are based on wrongdoing and are designed to compensate for loss: One Step (Support) Ltd v Morris-Garner [2018] UKSC 20; [2018] 2 WLR 1353, para 30. In the ordinary course, the market value of the benefit arising from having the use of money was said to be the cost the defendant would have incurred in borrowing the amount in question for the relevant period: a sum which, like all borrowings, would inevitably be calculated in terms of compound interest (Lord Nicholls at para 103). The court could however depart from the market value approach if it were established that it would produce an unjust outcome (Lord Hope at para 48, Lord Nicholls at para 119).[48]Lord Hope and Lord Nicholls proceeded on the basis of a presumption that the innocent recipient of a mistaken payment has benefited from the use of the money, the value of the benefit being the market cost of borrowing the money over the relevant period. The onus is on the defendant to displace that presumption. The innocent recipient, rather than the mistaken payer, is thus exposed to the risks of litigation.[49]Lord Nicholls acknowledged at para 125 that the decision might have serious consequences for public finances, because of the extended limitation period available in cases of mistake, but considered that the issue had been addressed by legislation:
“The seriously untoward consequences this may have for the Inland Revenue flow from the open-ended character of the extended limitation period prescribed by section 32(1)(c) of the Limitation Act 1980. Parliament has now recognised this extended period should not apply to payments of tax made by mistake: see section 320 of the Finance Act 2004.”
[50]Lord Walker of Gestingthorpe stated that he was essentially in agreement with Lord Hope and Lord Nicholls (para 154), and that he too would dismiss the appeal, “largely for the reasons which they give”. He also observed that the “crucial insight” in their speeches was the recognition that income benefits were more accurately characterised as an integral part of the overall benefit obtained by a defendant who is unjustly enriched (para 178). He went on, however, to state that he must confess that his own inclination would be to extend the equitable jurisdiction to award compound interest, rather than to recognise a restitutionary remedy available as of right at common law (para 184). He added that he “felt some apprehension” about the suggested conclusion that compound interest should be available as of right, subject only to an exception for “subjective devaluation”.[51]The other members of the Appellate Committee disagreed with the majority. Lord Scott of Foscote rejected the view that “the mere possession of mistakenly paid money - and accordingly the ability to use it if minded to do so - is sufficient to justify not simply a restitutionary remedy for recovery of the money, but a remedy also for recovery of the wholly conceptual benefit of an ability to use the money” (para 145). A restitutionary remedy could not in his view encompass the recovery of anything other than the money which the defendant had actually received. In reality, in his view, Sempra was asserting a claim for compensation for its loss of the use of the money, dressed up as a claim in restitution in order to take advantage of the more generous limitation period allowed by section 32(1)(c) of the Limitation Act 1980 (“the 1980 Act”).[52]Lord Mance also noted the practical context of the issue. The basis on which Sempra principally put their claim was that they had paid the ACT under a mistake of law. On that basis, section 32(1)(c) of the 1980 Act would postpone the commencement of the limitation period until the time when Sempra discovered or could with reasonable diligence have discovered that the ACT was not due: a time which they identified with the date in 2001 when the CJEU issued its judgment in the Metallgesellschaft case. Lord Mance commented (para 200) that the appropriateness of an extended time limit in this context was questionable. As he noted, Lord Hoffmann had recognised in the Kleinwort Benson case (Kleinwort Benson Ltd v Lincoln City Council [1999] 2 AC 349, 401) that “‘allowing recovery for mistake of law without qualification, even taking into account the defence of change of position, may be thought to tilt the balance too far against the public interest in the security of transactions’, adding that ‘the most obvious problem is the Limitation Act, which as presently drafted is inadequate to deal with the problem of retrospective changes in law by judicial decision’”. Like Lord Nicholls, Lord Mance noted that, as regards the future (although not as regards the instant case), section 320 of the Finance Act 2004 meant that section 32(1)(c) of the 1980 Act would no longer apply to mistakes of law relating to a taxation matter under the care and management of HMRC.[53]Like Lord Walker, Lord Mance cautioned against a radical reshaping of the law, observing at para 205 that “we must navigate using the reference points of precedent, Parliamentary intervention and analogy, and we should bear in mind the limitations of judicial knowledge and the assistance offered by a series of Law Commission reports”. European law left it, in his view, to national law to provide an effective remedy and did not prescribe that this should be by way of compound, rather than simple, interest (paras 201-204). The common law had recognised a claim for money had and received, but not a claim for the use of money had and received. A claim of the latter kind faced a long line of authority over a period of nearly 200 years, including the recent decision of the House in Westdeutsche Landesbank Girozentrale v Islington London Borough Council [1996] AC 669 (paras 203-220). The common law rule had been recognised and effectively endorsed by the Law Revision Committee, whose recommendations on interest in their Second Interim Report, 1934 (Cmd 4546) were implemented by provisions of the Law Reform (Miscellaneous Provisions) Act 1934 (later replaced by section 35A of the 1981 Act) (paras 211-212), by the Law Commission in all its reports on the subject (paras 222-224), and most importantly by Parliament, which had legislated in recent times for the payment of interest, but invariably on a simple basis (paras 212 and 221). There were in addition policy reasons making it unwise to introduce an absolute right to compound interest in restitution. As the Law Commission had noted, compound interest evoked deep-seated fears, because it increased in an exponential rather than a linear way, especially during periods of high inflation (para 222). In the light of such concerns, the Law Commission had made a number of recommendations relating to the introduction of a right to compound interest on a restricted basis. Those recommendations had not been acted on (para 224).[54]The decision of the House on the issues relevant to the present appeal can therefore be summarised as follows:(1) By a majority consisting of Lord Hope, Lord Nicholls and Lord Walker, the House held that the court had jurisdiction at common law (Lord Hope and Lord Nicholls) or at least in equity (Lord Walker) to make an award on the ground of unjust enrichment in respect of the time value of money which was paid prematurely as the consequence of a mistake. The basis of the award was that the benefit by which the recipient of the money was enriched was the time value of the money. The benefit was presumptively quantified as the market value of the use of the money during the period before it was lawfully due, that is, the cost of borrowing an equivalent amount in the market.(2) The same majority held that: (a) in the instant case, the presumption that the Government had benefited from the premature payment of the tax had not been displaced; but (b) the Government was in a different position from ordinary commercial borrowers, in that it could borrow at more favourable rates; and accordingly (c) the claims should be quantified on a conventional basis applicable to Government borrowing.

Legal developments since Sempra Metals

[55]A number of relevant developments in the law have occurred since Sempra Metals. First, the jurisprudence of the CJEU has developed since its Metallgesellschaft judgment. As was noted above, that judgment described the sum due under EU law, where tax was paid prematurely, and on the hypothesis that the appropriate remedy in domestic law lay in restitution, as “the amount of interest which would have been generated by the sum, use of which was lost as a result of the premature levy of the tax” (para 88). More recent judgments have provided greater clarity.[56]For example, in Littlewoods Retail Ltd v Revenue and Customs Comrs (Case C-591/10) EU:C:2012:478; [2012] STC 1714, the CJEU stated at para 27 that “it is for the internal legal order of each member state to lay down the conditions in which such interest [that is, interest on amounts levied in breach of EU law] must be paid, particularly the rate of that interest and its method of calculation (simple or ‘compound’ interest)”. The CJEU also made it clear, in relation to the principle of effectiveness, that national rules in relation to the calculation of interest “should not lead to depriving the taxpayer of an adequate indemnity for the loss occasioned” (para 29). In Littlewoods Ltd v Revenue and Customs Comrs [2017] 3 WLR 1401, this court held that an award of simple interest was sufficient to comply with that requirement, and that an award of compound interest on overpaid tax was therefore not required by the EU law principle of effectiveness. Recognition that an award of compound interest is not necessary in order to comply with the EU principle of effectiveness affects the context in which these issues have to be considered.[57]Secondly, the Littlewoods case also revealed a conflict between the decision in Sempra Metals and prior legislation. Long before Sempra Metals was decided, Parliament had created a scheme for the repayment of overpaid VAT, currently set out in section 80 of the Value Added Tax Act 1994 (“the 1994 Act”), with provision for the payment of simple interest in section 78. That section requires HMRC to pay interest on the repaid tax “if and to the extent that they would not be liable to do so apart from this section”. Entitlement to interest under section 78 is subject to limitations which would be defeated if it were possible for taxpayers to bring a common law claim for interest on mistaken payments.[58]Until Sempra Metals, it had been settled law for about 200 years that no such claim could be brought. In enacting section 78, Parliament legislated on that basis. In deciding Sempra Metals as it did, however, the House of Lords failed to have regard to the scheme which Parliament had established. Nor did it take account of section 826 of ICTA, which also provides for the payment of simple interest on overpaid tax, and covers a range of direct taxes, including ACT and MCT. These provisions are matched by corresponding provisions limiting the liability of taxpayers towards HMRC to simple interest on underpaid tax: see section 74 of the 1994 Act and section 826 of ICTA.[59]The persuasiveness of the majority’s approach in Sempra Metals is diminished by their failure to have regard to these provisions. As Lord Hoffmann observed in Johnson v Unisys Ltd [2001] UKHL 13; [2003] 1 AC 518, para 37:
“… judges, in developing the law, must have regard to the policies expressed by Parliament in legislation ... The development of the common law by the judges plays a subsidiary role. Their traditional function is to adapt and modernise the common law. But such developments must be consistent with legislative policy as expressed in statutes. The courts may proceed in harmony with Parliament but there should be no discord.”
[60]Against the background of the 1994 Act, in particular, the effect of Sempra Metals, was to create discord of a serious character: it rendered section 78 a dead letter, if that provision were given its natural construction. This court therefore decided in Littlewoods that, in order for section 78 to have the effect which Parliament had intended, it was necessary to depart from its natural construction. Thus the approach of the majority in Sempra Metals led, as Lord Mance had predicted, to a dislocation in a related area of the law which the Appellate Committee had not considered.[61]Thirdly, in Kleinwort Benson [1999] 2 AC 349 it was realised that allowing recovery of payments made under a mistake of law could create problems as the law of limitation then stood, since section 32(1)(c) of the 1980 Act would enable claims to be brought within six years of the mistake being discovered, no matter how long in the past the payment had been made. For that reason, Lord Browne-Wilkinson considered that “the correct course would be for the House to indicate that an alteration in the law is desirable but leave it to the Law Commission and Parliament to produce a satisfactory statutory change in the law which, at one and the same time, both introduces the new cause of action and also properly regulates the limitation period” (p 364). The majority, however, were unpersuaded that reform of the law of restitution should be delayed, and assumed that legislation could be enacted if Parliament considered it desirable to address the limitation question (see, for example, Lord Hoffmann at p 401). Parliament duly enacted such legislation. By the time of the decision in Sempra Metals, the majority therefore considered that the “seriously untoward consequences” for HMRC (as Lord Nicholls described them at para 125) of claims arising from mistaken payments of tax in the distant past were guarded against by section 320 of the Finance Act 2004, which provided that section 32(1)(c) of the 1980 Act should not apply in relation to a mistake of law relating to a taxation matter under the care and management of the Commissioners of Inland Revenue.[62]What has become apparent since Sempra Metals, however, is that the problems in relation to limitation which arise from the retrospective effect of that decision, and the decision in Kleinwort Benson, are incapable of being fully addressed by legislation. Repeated attempts by Parliament to address the retrospective impact of those decisions by introducing a limitation period with retrospective effect have been held to be incompatible with EU law: section 80 of the Value Added Tax Act 1994, as originally enacted, in Fleming (trading as Bodycraft) v Revenue and Customs Comrs [2008] UKHL 2; [2008] 1 WLR 195; section 320 of the Finance Act 2004 in Test Claimants in the FII Group Litigation v Revenue and Customs Comrs (Case C-362/12) EU:C:2013:834; [2014] AC 1161; and section 107 of the Finance Act 2007 in Test Claimants in the FII Group Litigation v Revenue and Customs Comrs [2012] UKSC 19; [2012] 2 AC 337.[63]This problem, of which the House of Lords was unaware at the time when Kleinwort Benson and Sempra Metals were decided, illustrates the risks of effecting major changes to the law of restitution by judicial decision. By applying the declaratory theory of adjudication, the law as altered by the decisions was deemed always to have applied, and the previously settled understanding of the law was treated as a mistake for the purposes of limitation. Consistently with that theory, in Kleinwort Benson the House of Lords held that a right of action had arisen when payments were made under a mistake of law, notwithstanding that no such right of action was recognised by the courts at that time. Similarly in Sempra Metals, the right of action in unjust enrichment arose when the defendant obtained the opportunity to use the money mistakenly paid, notwithstanding that no such right was understood to exist at that time. The tension inherent in the decisions is that the House adhered to the declaratory theory for the purpose of finding that a cause of action based on unjust enrichment had accrued in the past, based on a mistake of law capable of invoking section 32(1)(c) of the 1980 Act, while straining the premise of the theory, namely the need for judicial development of the law to be justifiable by reference to existing legal principles.[64]The consequence was that the rights established by those decisions were deemed to have vested in the claimants before the decisions were reached, with the result that, under EU law, they could not be taken away by retrospective legislation excluding or restricting the operation of section 32(1)(c) without a reasonable transitional period during which claims could be made. The position would have been different if the changes had been effected by legislation, since legislation can, and normally does, take effect prospectively.[65]Fourthly, decisions subsequent to Sempra Metals have demonstrated the degree of disruption to public finances which the decision in that case, taken together with Kleinwort Benson, is capable of causing. The decision in Kleinwort Benson enabled claims to be brought for the repayment of tax which had been paid in ignorance of the fact that the UK law under which it was levied was incompatible with EU law. Since the limitation period did not begin to run until the mistake was or could reasonably have been discovered, such claims could in principle be backdated to the UK’s entry into the EU in 1973. Not only could the principal amounts go back, in principle, for a period of several decades, but they had earned interest over that period. If, following Sempra Metals, the interest was compounded over that period, the resultant claims were potentially enormous.[66]The Littlewoods case, for example, concerned overpaid VAT on goods supplied to agents employed to make catalogue sales, as a form of commission paid in kind. Like the present appeal, it was a test case. The amount turning on the outcome of that appeal was estimated by HMRC at £17 billion. That was not the amount of the overpaid tax, or even the amount of the interest on the overpaid tax. It was the difference between compound interest and simple interest. In the present case, as we have explained, the total amount turning on the outcome is estimated by HMRC at £4-5 billion. Even in the context of public finances, these are very large sums.[67]Fifthly, the law of unjust enrichment has developed since Sempra Metals in ways which cannot easily be reconciled with the reasoning of the majority in that case. The development of greatest significance has been a more detailed analysis of the “at the expense of” question, in Investment Trust Companies v Revenue and Customs Comrs [2017] UKSC 29; [2018] AC 275. It is necessary next to consider that issue. “At the expense of”[68]Assuming for the present that an enrichment arises from having the opportunity to use money mistakenly paid, the question whether it is obtained “at the expense of” the claimant can best be answered by reference to the analysis of that question in the Investment Trust case. Lord Reed explained at para 42 that “the law of unjust enrichment … is designed to correct normatively defective transfers of value, usually by restoring the parties to their pre-transfer positions.” He went on at para 44 to endorse Lord Clyde’s dictum in Banque Financière de la Cité v Parc (Battersea) Ltd [1999] 1 AC 221, 237 that the principle of unjust enrichment:
“… requires at least that the plaintiff should have sustained a loss through the provision of something for the benefit of some other person with no intention of making a gift, that the defendant should have received some form of enrichment, and that the enrichment has come about because of the loss.”
Lord Reed also explained at para 50 that, as a general rule, a cause of action based on unjust enrichment is only available in respect of a benefit which the claimant has provided directly to the defendant (the only true exception identified being subrogation following the discharge of a debt, which is arguably based on a different principle). A causal connection between the claimant’s incurring a loss (in the relevant sense) and the defendant’s receiving a benefit was not enough to establish a transfer of value.[69]When money is paid by mistake, the claimant normally provides a benefit directly to the defendant: he pays him the money. He normally does so at his own expense: he is less wealthy by virtue of the payment. The transaction is normatively defective: the benefit is provided as the result of a mistake. In those circumstances, an obligation arises immediately under the law of unjust enrichment to reverse the enrichment by repaying the money (or an equivalent amount). The cause of action accrues when the money is mistakenly paid.[70]The majority in Sempra Metals considered that there was also an additional and simultaneous transfer of value, comprising the opportunity to use the money, which also gave rise to a cause of action based on unjust enrichment. That enrichment had to be reversed by the payment of compound interest.[71]This analysis has a number of questionable features, which can be illustrated by an example. If on 1 April the claimant mistakenly pays the defendant £1,000, with the result that the defendant is on that date obliged to repay the claimant £1,000, the defendant’s repayment of £1,000 on that date will effect complete restitution. Restitution of the amount mistakenly paid in itself restores to the claimant the opportunity to use the money: there is no additional amount due in restitution. That is because there has been only one direct transfer of value, namely the payment of the £1,000. The opportunity to use the money mistakenly paid can arise as a consequence of that transfer, but a causal link is not sufficient to constitute a further, independent, transfer of value. Contrary to the analysis of Lord Nicholls in Sempra Metals (at para 102), the recipient’s possession of the money mistakenly paid to him, and his consequent opportunity to use it, is not a distinct and additional transfer of value.[72]The position is essentially the same if the £1,000 is repaid not on 1 April but on 1 May. There has been no transfer of value subsequent to 1 April, when the mistaken payment was made. The only transfer of value needing to be reversed remains the payment of the £1,000. The claimant can however be awarded, in addition to the £1,000, simple interest on that amount under section 35A of the 1981 Act. That is because the obligation which arose under the law of unjust enrichment on 1 April, upon the making of the mistaken payment, created a debt. Interest can normally be awarded on a debt under section 35A of the 1981 Act.[73]That interest is intended to compensate the claimant for the loss of the use of the money to which he became entitled to restitution on 1 April. There is no right to interest on the basis of unjust enrichment: failure to pay a sum which is legally due is not a transfer of value, and does not give rise to an additional cause of action based on unjust enrichment. If there was no distinct cause of action for restitution of the opportunity to use the money on the date of the mistaken payment (as explained above), a cause of action based on unjust enrichment cannot have subsequently accrued, since no further defective transfer of value has taken place.[74]The point can also be illustrated by an example used by HMRC. If D owes C £1,000 under a contract, a claim also lies against D for interest under section 35A, from the date when the contractual payment became due. There is no claim against D for interest on the ground of unjust enrichment (even if an unjust factor is present). That is because any benefit obtained by D from his failure to pay the debt on time is not obtained at the expense of C in the relevant sense. There has been no transfer of value from C to D. The latter’s opportunity to use the money which remains in his possession is the result of his failure to pay the contractual debt. The same analysis applies where the debt is imposed by the law of unjust enrichment, for example as the result of a mistaken payment of £1,000. Any benefit obtained by D as a consequence of his possession of the £1,000 is derived from his failure to pay that debt. It cannot be said to have been transferred from C to D.[75]All this is consistent with a long-established understanding of, first, the nature of the cause of action based on a mistaken payment, and secondly, the basis on which interest is payable. As to the first of these, Lord Mansfield stated in the classic case of Moses v Macferlan (1760) 2 Burr 1005, 1010, that the defendant in an action for money had and received “can be liable no further than the money he has received”. That approach was followed in many later authorities, until Sempra Metals: see, to give only a few examples, Walker v Constable (1798) 1 Bos & P 306, 307 (“The court were of opinion, on the authority of Moses v Macferlan, 2 Burr 1005, that in an action for money had and received the plaintiff could recover nothing but the net sum received without interest”), Depcke v Munn (1828) 3 C & P 111 per Lord Tenterden CJ (“… the courts have held again and again that interest cannot be recovered in an action for money had and received … This has been decided so often, that I cannot now venture to allow the question to be agitated.”), Johnson v The King [1904] AC 817 and the Westdeutsche case [1996] AC 669.[76]As to the basis on which interest is payable, a clear explanation was provided by Lord Wright, a judge who was well aware of unjust enrichment (see, for example, Fibrosa Spolka Akcyjna v Fairbairn Lawson Combe Barbour Ltd [1943] AC 32), and had also had to consider interest as a member of the Law Revision Committee which reported in 1934, mentioned earlier. In Riches v Westminster Bank Ltd [1947] AC 390, 400, he stated:
“… the essence of interest is that it is a payment which becomes due because the creditor has not had his money at the due date. It may be regarded either as representing the profit he might have made if he had had the use of the money, or conversely the loss he suffered because he had not that use. The general idea is that he is entitled to compensation for the deprivation.”
[77]Once it is understood that the claim to interest is not truly based on unjust enrichment but on the failure to pay a debt on the due date, the conclusion inevitably follows that interest can be awarded on the claims within categories (b) and (c) under section 35A of the 1981 Act: see BP Exploration Co (Libya) Ltd v Hunt (No 2) [1979] 1 WLR 783, and Sempra Metals at paras 104 (Lord Nicholls) and 175 (Lord Walker).[78]On a literal reading of section 35A, no such interest could have been awarded on the claims under category (a). That is because section 35A applies only where there are proceedings for the recovery of a debt (or damages), and therefore does not apply where the defendant has repaid the debt (or has set it off) before the creditor has commenced proceedings for its recovery. An award of interest is nevertheless required in such circumstances by EU law, if an effective restitutionary remedy is to be available under English law in respect of San Giorgio claims: that was the point decided in Metallgesellschaft. It is unnecessary to decide in this appeal how an award of interest should be made available in those circumstances (and the court has heard no argument on the point). But there are a number of potential solutions.[79]For the foregoing reasons, we therefore depart from the reasoning in Sempra Metals so far as it concerns the award of interest in the exercise of the court’s jurisdiction to reverse unjust enrichment. As mentioned earlier, it is unnecessary for us to consider the reasoning in that case so far as it concerns the award of interest as damages, and nothing in this judgment is intended to question that aspect of the decision. Since the award of compound interest to PAC by the courts below was based on the application of the reasoning in Sempra Metals which we have disapproved, it follows that HMRC succeed on Issue II, and PAC’s claims to compound interest under categories (b) and (c) must be rejected. PAC’s claim to compound interest under category (a) would also have been rejected, if it had not been accepted by HMRC.[80]Finally, in relation to this aspect of the appeal, it is worth adding that the view of the majority in Sempra Metals that the opportunity to use money mistakenly paid should be regarded as an enrichment also raises a number of questions, particularly in relation to the method by which, and the date or dates as at which, the enrichment is to be measured. In addition, if one stands back and considers the realism, and also the fairness, of the approach to enrichment adopted by the majority in Sempra Metals, the results which it produces are concerning. As Professor Burrows has written, in relation to the decision of Henderson J in the Littlewoods case, in his contribution to Commercial Remedies: Resolving Controversies, eds Virgo and Worthington (2017), p 266:
“… if one were to step back from the complex detail, the result of Henderson J applying compound interest on all the mistaken payments from the date of receipt appears to be tantamount to saying that, had the Revenue not been paid those sums, it would have borrowed the same sums at a compound interest rate for five decades. Surely that cannot be right.”
These issues were not, however, discussed in argument in the present appeal, and in the circumstances it is inappropriate to consider them further.

The remaining issues

[81]There remain two issues which are relevant to the computation of the amounts which EU law requires HMRC to repay. In order to understand those issues it is necessary to recollect how, under the legislation which was in force in the relevant years, ACT was charged on distributions by a UK-resident company and was set against the paying company’s subsequent liability to MCT. Under section 14 of ICTA, ACT was charged when a UK-resident company paid a “qualifying distribution”, which included the payment of a dividend (section 209).[82]Under section 231 of ICTA, when a UK-resident company made a qualifying distribution, the recipient of the distribution was entitled to a tax credit equal to the proportion of the amount of the distribution which corresponded to the rate of ACT in force when the distribution was made. Under section 238, when a UK-resident company made a qualifying distribution, the sum of the amount of that distribution together with the proportion of that amount which corresponded to the rate of ACT in force when the distribution was made was known as a “franked payment” (“FP”). Section 238 also provided that when a UK-resident company received a distribution, in respect of which it was entitled to receive a tax credit, the aggregate of the distribution and the amount of the tax credit was “franked investment income” (“FII”). When the recipient company itself made a qualifying distribution in the same accounting period as it received FII, it paid ACT only on the excess of FP over FII (section 241).[83]The UK-resident company (“company A”) had to make a return and account to HMRC for the ACT quarterly by disclosing the FPs which it made and the FII which it received, foreign income dividends paid and received, and the ACT payable on the FPs and the foreign income dividends: paragraphs 1 and 3(1) of Schedule 13 to ICTA. The ACT which company A paid during an accounting period was then set against its liability (if any) to MCT on its profits in the accounting period by the operation of section 239 of ICTA, which we discuss below.[84]Thus, if company A received a distribution from another UK-resident company (company B) it would not be liable to pay MCT on that distribution (section 208). If company A, having received a distribution from company B, itself made a distribution, it would be liable to pay ACT on the excess of its FP over its FII. When company A came to pay MCT on its profits in the same accounting period it would have been entitled to set off the ACT which it had paid (section 239).[85]The illegality under EU law, which was caused by the UK’s failure to match the exemption conferred on dividends received from UK-resident companies by an equivalent credit in respect of overseas-sourced dividends, is to be remedied by a credit for foreign dividends by reference to the FNR, as we have held under Issue I above. Applying the example above but with the receipt by company A of overseassourced dividends in place of UK-sourced dividends, under EU law the FNR credit would fall to be applied to its payment of MCT. Thus the MCT which company A paid was unlawful to the extent that the credit had not been given. If company A had itself paid a dividend, the FNR credit should have been applied to reduce the ACT which it had paid.[86]With that introduction we turn to Issue III.

Issue III

[87]Issue III, on which HMRC seek permission to appeal, is whether a claim for restitution lies to recover “lawful ACT”, which has been set against “unlawful MCT”. By the expression “lawful ACT” we refer to the element within an undifferentiated ACT charge which did not represent unduly levied tax on overseassourced dividends. Lawful ACT on the (UK-resident) Company A’s distribution refers to such ACT as is due after giving effect to(i) the exemption given to income from dividends of UK-resident companies and(ii) (in light of our answer to Issue I) the tax credit which EU law requires to be given to income from dividends from overseas companies. “Unlawful MCT” in this context refers to such part of the charge to MCT as is attributable to the failure to give the overseas-sourced dividends, which company A received, a tax credit at the FNR to achieve equivalence to the exemption which section 208 gave to dividends received from UK-resident companies.[88]This issue was not argued in the courts below because Henderson J and the Court of Appeal, when considering this case, were bound by the Court of Appeal’s earlier decision in Test Claimants in the FII Group Litigation v Revenue and Customs Comrs [2010] EWCA Civ 103; [2010] STC 1251 (“FII CA”). In that judgment the Court of Appeal held that a taxpayer was entitled to reimbursement of lawful ACT, which HMRC had retained because it had been set against an unlawful MCT charge. The court held that such ACT related directly to the unlawful MCT because the CJEU treated ACT as an advance payment of MCT: FII CA paras 148- 151.[89]The statutory provisions governing the set-off of ACT against MCT were as follows.[90]Section 239(1) of ICTA provided for the automatic set-off of ACT against MCT, thereby reducing company A’s liability to pay MCT. It provided:
“… advance corporation tax paid by a company (and not repaid) in respect of any distribution made by it in an accounting period shall be set against its liability to corporation tax on any profits charged to tax for that accounting period and shall accordingly discharge a corresponding amount of that liability.”
[91]Where the tax-paying company did not have a sufficient liability to MCT on its profits to use up the ACT by way of set-off in the same accounting period, the unused ACT could be carried back under section 239(3) which provided:
“Where in the case of any accounting period of a company there is an amount of surplus advance corporation tax, the company may, within two years after the end of that period, claim to have the whole or any part of that amount treated for the purposes of this section (but not of any further application of this subsection) as if it were advance corporation tax paid in respect of distributions made by the company in any of its accounting periods beginning in the six years preceding that period … and corporation tax shall, so far as may be required, be repaid accordingly. In this subsection ‘surplus advance corporation tax’ in relation to any accounting period of a company, means advance corporation tax which cannot be set against the company’s liability to corporation tax for that period because the company has no profits charged to corporation tax for that period …”
[92]The surplus ACT could also be carried forward automatically under section 239(4) which provided:
“Where in the case of any accounting period of a company there is an amount of surplus advance corporation tax which has not been dealt with under subsection (3) above, that amount shall be treated for the purposes of this section (including any further application of this subsection) as if it were advance corporation tax paid in respect of distributions made by the company in the next accounting period.”
[93]Section 239(5) explained how the set-off operated under both subsections (1) and (4). It provided:
“Effect shall be given to subsections (1) and (4) above as if on a claim in that behalf by the company and, for that purpose, a return made by the company under section 11 of the Management Act containing particulars of advance corporation tax or surplus advance corporation tax which falls to be dealt with under those subsections shall be treated as a claim.”
Company A could also surrender its ACT to its subsidiary in accordance with section 240, with the result that the ACT would be treated as ACT paid by the subsidiary and set against the subsidiary’s liability to pay MCT.[94]HMRC’s case is simple. They argue that if a taxpaying company included relevant details of ACT paid in its tax return, sections 239(1) and (5) mandated an automatic set-off of the ACT against the company’s liability for MCT. If, on a proper understanding of the law, the company did not owe sufficient MCT in the relevant accounting period, the ACT remained surplus and available to be set off in the next accounting period under section 239(4). In other words, HMRC argue that the law treats an unlawful MCT charge as a nullity, with the result that there is no set off under section 239(1) and no enrichment of HMRC by the payment of the ACT, which remained available to offset the taxpaying company’s lawful MCT in other accounting periods.[95]PAC opposes the grant of permission to HMRC on this issue and submits that it is a detailed issue of computation which is likely only to affect the appeal in PAC’s case if PAC’s approach to Issue V is correct. If this court were to give permission to appeal, PAC advances three arguments. First, it submits that the court’s approach should be governed by the principle that the taxpayer should be entitled to recover unduly levied tax. Secondly, it argues that, because the CJEU has characterised ACT as “nothing more than a payment of corporation tax in advance” (eg FII ECJ I para 88), ACT could only lawfully be charged where it is itself a pre-payment of a lawful charge to MCT. As a result, it contends that the correct analysis is that a payment of ACT, which is subsequently set against an excessive liability to MCT, is an advance payment of an excessive tax liability and is itself the payment of an excessive tax liability. As such, it is liable to be recovered in a claim in restitution. Alternatively, PAC contends that the payment of the ACT relates directly to the unlawfully levied MCT and so is recoverable in a claim in restitution. In support of those contentions PAC relies on dicta of the CJEU in FII ECJ I and FII ECJ II. PAC’s third argument is that, if it had been aware that it did not have any liability for a substantial part of the MCT, it would have not have paid the ACT. PAC was a subsidiary of Prudential plc and it had no subsidiaries of its own which generated profits giving rise to a liability to corporation tax against which PAC’s ACT could have been used. It would therefore have paid dividends to its parent company within a group income election so that the ACT was paid at the level of the parent company and would have been available for set-off against the MCT of other subsidiary companies within the group. This, it submits, would have been the only sensible course to avoid the ACT being stranded in PAC’s accounts.

Analysis

[96]In our view it is appropriate to give HMRC permission to raise this issue as it is a point of law of general public importance in an appeal to this court in the context of a group litigation order. While PAC submits that it alone is likely to be affected by the determination of this issue, the court is not in a position to assess whether or not that is so. The matter also arises in the FII litigation. HMRC had applied for permission to appeal the Court of Appeal’s ruling on this issue in FII CA but the determination of that application was postponed by this court by orders dated 8 November 2010 and 9 May 2017. The issue, which will be of relevance to the final determination of the FII litigation, therefore comes to this court in this appeal before this court has addressed the application to appeal in that litigation.[97]In addressing this issue, the starting point is to recall that an entitlement to repayment or restitution in this context requires that there has been an unlawful charge to tax as a result of incompatibility with EU law: San Giorgio. The question we are asked to consider is in substance: have HMRC unlawfully levied ACT by setting it against MCT which has been unlawfully charged? But there is a logically prior question, which is whether there has been any set-off.[98]Company A may have received income which has funded its distribution from UK-resident companies and also from companies resident elsewhere in the EU. In this computational issue the court is not concerned with “unlawful” ACT, which has been charged on a distribution by company A derived from income which it has received from an overseas-resident company in the absence of sufficient credit for foreign tax on the latter company’s distributions. We are concerned with ACT which is unquestionably lawful but which has purportedly been set against an unlawful MCT charge on company A.[99]PAC relies on dicta in FII ECJ I and FII ECJ II to argue that this prima facie lawful charge on company A’s dividend is tainted by its being merely an advance payment of an unlawful MCT charge. But the CJEU, when it characterised ACT as constituting “a form of advance payment of corporation tax” (FII ECJ I para 88 and FII ECJ II paras 68 and 110), was well aware of the provisions of ICTA which allowed the taxpaying company to utilise the ACT which it had paid in different ways. Thus, in FII ECJ II at para 6, the CJEU stated: “A company had the right to set the ACT paid in respect of a distribution made during a particular accounting period against the amount of mainstream corporation tax for which it was liable in respect of that accounting period, subject to certain restrictions. If the liability of a company for corporation tax was insufficient to allow the ACT to be set off in full, the surplus ACT could be carried back to a previous accounting period or carried forward to a later one, or surrendered to subsidiaries of that company, which could set it off against the amount for which they themselves were liable in respect of corporation tax.” (The reference to the surrender of ACT to a subsidiary is a reference to section 240 of ICTA.)[100]As we have shown, section 239 of ITCA did not confine the MCT, against which the ACT could be set, to MCT due for the same accounting period as that in which the ACT was paid (“the same accounting period”). If there was insufficient MCT due in the same accounting period, the surplus ACT was carried forward automatically to the next accounting period, unless company A elected to use it otherwise, such as by carry back under section 239(3). If the company did not so elect, and if in the same accounting period and subsequent accounting periods company A did not have sufficient MCT to use up the ACT which it had paid, or if Company A did not surrender the ACT under section 240, the ACT was, in PAC’s words, “stranded”. But that stranding of the ACT, were it to have occurred, would not affect the lawfulness of the ACT charge.[101]In our view, HMRC are correct in their submission that, if an apparent charge to MCT was unlawful, that charge was a nullity. The ACT could not have been set against a nullity but remained available to be carried back if a claim were made under section 239(3) or for automatic set-off against lawful MCT in a subsequent accounting period under section 239(4) or otherwise to be utilised. Being so available, the lawful ACT did not directly relate to the unlawful MCT in the same accounting period in the sense that penalties and interest may relate to an unlawfully levied tax. Accordingly, HMRC in receiving payment of the lawful ACT did not receive unlawfully levied tax which gave rise to a San Giorgio claim.[102]Further, PAC was obliged by ICTA to pay the lawful ACT. The payment of the ACT did not entail a defective transfer of value which falls to be corrected: the ACT was due when it was paid, and was available to PAC to utilise thereafter. PAC’s loss in the context of Issue III (ie in relation to lawful ACT) is the result of the levying of unlawful MCT, and, through the misunderstanding of the law which it shared with HMRC, of its not having been able to set the unutilised ACT against its liability for lawful MCT in the same or other accounting periods or otherwise to utilise the ACT to reduce its liability to tax. Its loss in that sense does not support a claim in restitution: Investment Trust Companies v Revenue and Customs Comrs, especially paras 41-45 per Lord Reed.[103]We are informed that PAC’s corporation tax liabilities in its accounting periods from 1994 to 1998 are not finalised as PAC’s returns in those years are still open and that therefore it may be possible for PAC to carry forward unutilised ACT to set against its MCT liabilities in those periods. But, whether or not that is the case, in agreement with Henderson J in FII High Court 1 ([2008] EWHC 2893) we consider that an enquiry into whether, and if so how, surplus ACT would otherwise have been used within a group of companies cannot give rise to a claim in restitution but would form part of a claim for damages if the criteria for such a claim were met.[104]We therefore, in agreement with Henderson J in the FII litigation, answer the question raised in Issue III (“Does a claim in restitution lie to recover lawful ACT set against unlawful corporation tax?”) in the negative.

Issue V

[105]PAC seeks permission to cross-appeal if (as we have done) we grant HMRC permission to appeal on Issue III. Again, the issue arises in the context of a GLO and we are unable to assess its significance in other cases within the GLO. But it is closely connected with Issue III and has significant consequences for PAC’s claim for interest. It is appropriate that we address it in the context of this appeal. We therefore grant permission for the issue to be raised.[106]Issue V comprises two related questions concerning the utilisation of ACT on a hypothesis that an undifferentiated fund of lawful and unlawful ACT was purportedly set off against an amount of MCT which was in part lawful and in part unlawful.

(a) ) which the parties have raised is:

“Where ACT from a pool which includes unlawful and lawful ACT is utilised against an unlawful corporation tax liability, is the unlawful ACT regarded as a pre-payment of the unlawful corporation tax liability or is the ACT so utilised regarded as partly lawful and unlawful pro rata?”
[107]PAC contends that the unlawful ACT which company A has paid is to be regarded as utilised first against the unlawful MCT. HMRC have argued for a pro rata approach by which the unlawful MCT is regarded as having been met by the utilisation of lawful and unlawful ACT in the same proportion as the unlawful MCT bears to the overall MCT charge. The background is that in so far as unlawful ACT has been utilised against lawful MCT, HMRC have conceded that the time value of the prematurely-paid ACT is recoverable in compound interest, as explained earlier in the discussion of Issue II. In so far as the unlawful ACT has purportedly been utilised against unlawful MCT, the unlawful ACT which the taxpaying company has paid is recoverable together with interest under section 35A of the 1981 Act, as explained in relation to Issue II, as both the ACT charge and the MCT charge were nullities.[108]Henderson J in his second judgment in this case ([2015] EWHC 118 (Ch); [2015] STC 1119) addressed this issue at paras 34-37. He expressed an initial inclination to adopt the pro rata approach as everyone at the time had assumed that the whole of both the ACT and the MCT had been lawfully charged. He decided however that PAC’s approach was correct because, if the unlawful ACT was regarded as a prepayment of the unlawful MCT, the end result reflected precisely the credit for foreign tax which EU law required, whereas on HMRC’s approach company A would have an additional and unnecessary claim to recover the element of lawful ACT which had been utilised against the unlawful MCT. His ruling was made expressly on the basis that he was bound by the Court of Appeal’s ruling on Issue III above, a ruling which we have now overturned.[109]The Court of Appeal (paras 113-127) disagreed with Henderson J’s approach. It stated that the issue was how to determine the extent of the benefit for HMRC in money terms from the payment or bringing into account of an unlawful MCT charge for the purpose of determining the extent of HMRC’s unjust enrichment. The court looked for a fair way of determining that enrichment in a situation where an undifferentiated fund of lawful and unlawful ACT had purportedly been set against an apparent liability to MCT, which in fact comprised both lawful and unlawful MCT. The court attached weight to the fact that both PAC and HMRC were unaware of the meaning and effect of the relevant EU law at the time; neither was to blame for the situation; both were disabled by their ignorance of the true state of affairs from applying their minds at the time to the allocation of lawful and unlawful ACT as between the lawful and unlawful elements of MCT. As a result, the court sought to strike a fair balance between their interests by adopting an objective standard. That standard was the pro-rating approach which Henderson J had earlier favoured in his judgment in Test Claimants in the FII Group Litigation v Revenue and Customs Comrs [2014] EWHC 4302 (Ch); [2015] STC 1471 (“FII (High Court) Quantification”), para 205.[110]We are not able to reconcile the Court of Appeal’s ruling with our decision on HMRC’s appeal on Issue III, which is inconsistent with the ruling in FII CA by which the Court of Appeal in this case was bound. As HMRC have submitted and we have held under that issue, section 239 has the effect that lawful ACT is not set against unlawful MCT, which is a nullity. The pro rata method, which involves unlawful MCT being met in part by unlawful ACT and in part by lawful ACT, cannot therefore work. Instead, lawful ACT, which was not utilised against lawful MCT, remained available to be claimed against lawful MCT in the same or other accounting periods. The unlawful ACT, which company A paid, was not set against the unlawful MCT charge in a given accounting period because both the unlawful ACT charge and the unlawful MCT charge are nullities. The principled answer is therefore that the unlawful ACT, which company A has paid, must be treated as having been utilised first against the unlawful MCT charge. Where there is no unlawful MCT against which to set the unlawful ACT which has been paid, the residual unlawful ACT is to be treated as utilised against lawful MCT. Because both of the unlawful charges are nullities, the unlawful ACT is itself recoverable, unless it has been set against a lawful MCT charge. When unlawful ACT has been set against lawful MCT, company A has a claim for interest on the ACT so used, as stated in para 78 above.[111]The second question under Issue V relates to the carry back to an earlier quarter of domestic FII received in a later quarter in the same accounting period. To address this, it is necessary to explain the operation of paragraph 4 of Schedule 13 to ICTA. Rather than set out the provision we gratefully adopt the explanation of its effect which Henderson J gave in FII (High Court) Quantification at para 209:
“The effect of these rather densely worded provisions may be summarised by saying that FII received in a later quarterly return period must first be applied in franking any dividends paid by the company in that period, but that any surplus may then be carried back to frank unrelieved dividends paid in an earlier quarter, thus generating a repayment of ACT. If there has been a change of ACT rates in the meantime, the repayment is not to exceed the amount of the tax credit comprised in the FII which is carried back.”
If the excess FII was not so used in repayment of ACT paid in the earlier quarter, it was carried forward into the next annual accounting period to set against FPs in the same way (section 241(3)).[112]Issue V(b) asks:
“Where domestic FII was carried back to an earlier quarter is it to be regarded as having been applied to relieve lawful and unlawful ACT pro rata or only lawful ACT?”
[113]Henderson J in his second judgment in this case discussed the issue briefly in paras 40-43 after hearing full argument on the point. He decided, with considerable hesitation, that the FII was to be regarded as having been applied to relieve only lawful ACT in the earlier quarter because otherwise FII from UK-sourced dividends, which was entirely lawful, would be used to cancel out part of the credit which EU law requires on foreign income. In reaching this conclusion he departed from the view which he had reached on essentially the same issue in FII (High Court) Quantification at paras 207-211.[114]The Court of Appeal disagreed and (paras 128-133) adopted an approach similar to that which it took on Issue V(a). Again, the court laid stress on the fact that at the time nobody appreciated that the ACT against which the FII was carried back might comprise both lawful and unlawful elements and no-one was to blame. The fair course therefore was to adopt the pro rata approach which the court had taken in relation to Issue V(a). The effect of that approach would be that “the primary period of unjust enrichment of HMRC” through receipt of the unlawful ACT would be brought to an end and HMRC’s enrichment would be measured by the time value of the ACT payment. The court did not see this as cancelling out any part of the credit which EU law required on overseas-sourced dividends.[115]In this appeal PAC renews the arguments which Henderson J favoured. The UK tax system was unlawful because credits were not given under section 231 for tax on overseas-sourced dividends in order to relieve an ACT liability. The use of carried-back FII to relieve unlawful ACT deprived company A of the credits which it should have had for the overseas-sourced dividends. The carried-back FII should therefore be regarded as having been applied to relieve only lawful ACT.[116]HMRC’s answer in their written case is that EU law does not mandate a form of credit for overseas-sourced dividends. They quote the statement of the CJEU in para 72 of FII ECJ II:
“As is clear from para 62 [of the present judgment], the obligation presently imposed on the resident company by national rules, such as those at issue in the main proceedings, to pay ACT when profits from foreign-sourced dividends are distributed is, in fact, justified only in so far as that advanced tax corresponds to the amount designed to make up the lower nominal rate of tax to which the profits underlying the foreignsourced dividends had been subject compared with the nominal rate tax applicable to the profits of the resident company.”
HMRC, unexceptionably, interpret this statement as meaning that it is lawful to charge ACT on a dividend paid by company A only to the extent that it was lawful to charge MCT on the profits out of which that dividend was paid. But HMRC go on to say that the relief required was not in the form of a credit which was the equivalent of further FII.[117]We do not accept HMRC’s submission on Issue V(b) for the following three reasons.[118]First, it follows from the answer which we have given on Issue I that we reject the contention that no particular form of credit is mandated by EU law. What the CJEU said in para 72 of FII ECJ II must be construed in the light of what it said in paras 61-65 which we have quoted in para 26 above. That in turn falls to be understood against the earlier ruling of the CJEU in FII ECJ I, which we have quoted in para 7 above. In other words, EU law requires a tax credit by reference to the FNR to which the profits of the overseas company have been subject. As a result, the UK can charge ACT in relation to company A’s dividends so far as they comprise profits from overseas-sourced dividends only to the extent that there is tax due in respect of those dividends after it has given company A that tax credit.[119]Secondly, the consequence of this is that PAC is correct in its contention that HMRC’s approach would result in depriving company A of the tax credits on overseas-sourced dividends which EU law mandates. Using the example which PAC gave in its written case, suppose that company A paid ACT of £100 in the first quarter when it had received overseas-sourced dividends which (if EU law had been applied correctly) would have entitled it to a credit of £25. If EU law had been applied correctly in that quarter, the ACT paid would have been £75. Suppose then that in the third quarter company A received FII for UK-sourced dividends which carried credits of £75 which it carried back to the first quarter. On PAC’s approach, the carried back FII would result in the repayment of all the ACT which had properly been paid. If, as on HMRC’s approach, the £75 of FII, which is carried back from the third quarter, were utilised pro rata between the lawful and unlawful ACT which comprised the £100 paid in the first quarter, £18.75 (1/4 of the £75) would be attributed to the unlawful ACT, thereby cancelling to that extent the credit to which company A was entitled in EU law.[120]Thirdly, we are not persuaded by the arguments as to fairness which influenced the Court of Appeal in relation to both of Issues V(a) and V(b). As unlawful ACT is a nullity, the principled answer is that domestic FII carried back to an earlier quarter is to be regarded as having been applied to relieve only lawful ACT so that any excess FII remained available for carry forward under section 241(3).[121]We therefore answer Issue V(b) by holding that domestic FII which is carried back to an earlier quarter under paragraph 4 of Schedule 13 of ICTA is to be regarded as having been applied to relieve only lawful ACT.[122]In further written submissions HMRC and PAC disagree on factual matters which may affect the working out of the rulings which we have made. This court is not in a position to resolve these matters. We will invite submissions in response to our judgment as to how our rulings may be applied.

Conclusions

[123]For the foregoing reasons, we allow HMRC’s appeal on Issues II and III, and dismiss it on Issue I. PAC’s proposed cross-appeal on Issue IV does not arise, as a result of its success on Issue I, and it also succeeds in its cross-appeal on Issue V(a). In relation to Issue V(b), the court holds that FII carried back to an earlier quarter is to be treated as having been applied to relieve only lawful ACT.

Cited in 36 later judgments

and 27 more of the most senior