BAT Industries PLC and Others v Revenue and Customs (INCOME TAX/CORPORATION TAX : European law) [2017] UKFTT 558 (TC)

FTT-Tax

BAT Industries PLC and Others v Revenue and Customs (INCOME TAX/CORPORATION TAX : European law)

[2017] UKFTT 558 (TC) · 2017-05-12

[1]This decision is divided into the following subject headings: Introduction [2] – [6] The background [7] – [17] The FII GLO and the compound interest award [18] – [34] Introduction of Part 8C CTA [35] – [43] The 45% rate – policy and methodology [44] – [67] Policy considerations [45] – [54] The HMRC methodology [55] – [67] BAT’s position [68] – [110] BAT’s surplus ACT problem [69] – [70] BAT’s modelling [71] – [110] Mr Cohn’s basic model [72] – [83] Mr Cohn’s revised model [84] – [91] Mr Cohn’s second revised model [92] – [110] Third party evidence [111] – [122] Expert evidence [123] – [153] Section A – an assessment of HMRC’s methodology in establishing the 45% rate for restitution interest [127] – [136] Section A2 – the appropriateness of KAI’s methodology [137] – [141] Section B – assessment of the effective rate of tax (i.e. the rate of tax actually paid after reliefs and allowances) at which UK corporate taxpayers paid corporation tax in each of the fiscal years in the period 1973-2016 [142] – [144] Section C – an assessment of the advantages and disadvantages in setting the tax rate by reference to either the nominal (statutory) rate of tax or the effective rate of tax and how each approach would reflect the economic realities of business [145] – [153] Discussion [154] – [255] The principle of effectiveness [160] – [195]a. A. The 45% rate [183] – [195]

b. B. Ring fencing [196] – [198] The principle of sincere cooperation [199] The principles of legal certainty and legitimate expectation [200] – [207] The principle of proportionality [208] – [216] Conclusion on principles of EU law [217] – [220] Fundamental rights [221] – [253] Fair hearing and effective remedy [223] – [231] Right to/protection of property [232] – [253] Constitutionality and the rule of law [254] Remedies [255] Reference to the CJEU [256] – [259] Decision [260] Application for permission to appeal [261] Appendix A – Part 8C, Corporation Tax Act 2010 A Appendix B – Experts’ Joint Statement of Issues of Agreement and Disagreement (12 December 2016) B Appendix C – Extracts from the European Convention on Human Rights (“ECHR”) and the Charter of Fundamental Rights of the European Union (“CFREU”) C Introduction

[2]This appeal, which consolidates appeals made by members of the BAT group, raises the question of the lawfulness of certain UK primary legislation. The legislation in question is that which was inserted, with effect from 26 October 2015, by s 38 of the Finance (No 2) Act 2015 into the Corporation Tax Act 2010 (“CTA”) as Part 8C of that Act. Part 8C introduced a charge to corporation tax at the rate of 45%, in place of what would otherwise be the ordinary charge to tax, on restitution interest arising to a company, a charge which is “ring-fenced” by not being capable of being offset by reliefs and set-offs, and as part of the machinery of that charge, a withholding tax at 45% on payments of restitution interest.

[3]Restitution interest is defined, by s 357YC CTA, to mean interest (not being limited to interest at a statutory rate) paid by HMRC in respect of a claim by the company for restitution, either by reference to the payment of an amount to HMRC under a mistake of law relating to a taxation matter, or the unlawful collection by HMRC of an amount in respect of taxation. Put shortly, the charge is on interest, other than purely simple interest at a statutory rate, awarded as part of a restitutionary remedy. That includes both awards of compound interest, and mixed awards of compound interest and simple interest at a statutory rate.

[4]It is not necessary to make any extensive reference to the statutory provisions themselves. I include those provisions for ease of reference in Appendix A to this decision. Much of Part 8C is taken up with mechanical or anti-avoidance provisions. Those substantive provisions which are particularly relevant to this appeal are: s 357YA – which imposes the charge to corporation tax on restitution interest arising to a company s 357YC – which defines “restitution interest” s 357YE – provides for the period in which amounts are to be brought into account as restitution interest s 357YK – sets the rate (the “restitution payments rate”) of corporation tax on restitution interest at 45% s 357YL – takes restitution interest out of the total profits of a company subject to the usual charge to corporation tax and “ring fences” the tax on restitution interest by excluding the application of reliefs and set-offs s 357YO – provides for tax at the full restitution payments rate to be deducted at source by HMRC when making a payment of restitution interest s 357 YQ – provides for assessment of tax charged on restitution interest s 357YV – excludes restitution interest charged to corporation tax under Part 8C from any other charge, including under the loan relationships provisions

[5]The appeals, which are brought in this tribunal by virtue of s 357YS CTA, are against the deduction by HMRC of withholding tax, in the sum of £261,420,591.92, when making a payment to the appellants (which I shall collectively describe as BAT) on 10 November 2015 of the balance of a judgment debt of £1,184,107,228.48, including compound interest, awarded to BAT as test claimants in the Franked Investment Income Group Litigation (“the FII GLO”).

[6]The substantive issue in this consolidated appeal is whether or to what extent Part 8C CTA is incompatible with(a) BAT’s directly effective EU law rights, primarily the principle of effectiveness, the principle of protection of legitimate expectation and the principle of proportionality,

(b) BAT’s rights under the European Convention of Human Rights, namely under Article 1 of the First Protocol (A1P1) and Article 6 (right to a fair trial) and

(c) BAT’s rights under Articles 17(1) and 47 of the Charter of Fundamental Rights of the European Union (2000/C 364/01). In addition, there arise issues of the rule of law and the separation of powers between the judicial function in determining an effective remedy in a particular case and the legislative function. The background

[7]Although the parties were unable to reach full agreement on a proposed statement of agreed facts, such a statement, marked to show continuing areas of disagreement, was produced. It is not, however, necessary for me to refer to much of that statement by way of background. To the extent there are areas of factual dispute which are material to the issues in this appeal, I will address them at the relevant points in this decision. For the time being, a brief summary of the background will suffice, if only to identify some of the material judgments which form the backdrop to the current appeal.

[8]The claims in the FII GLO arose out of the UK’s partial imputation system in relation to distributions of profits. Under that system, until its abolition in 1999, dividends paid by a UK resident company were subject to a charge to advance corporation tax (“ACT”), and shareholders were entitled to a tax credit. Dividends paid by one UK resident company to another were exempt from corporation tax, and in addition, if paid subject to ACT, carried a tax credit equal to the ACT, and the dividend and the tax credit together constituted “franked investment income”, which to that extent enabled dividends to be paid by the recipient company without payment of ACT. By contrast, dividends received from a company resident outside the UK, including those resident in other member states of the EU, were not exempt. In certain cases relief for foreign withholding tax and underlying tax was available.

[9]Where a UK resident company received such foreign dividends, surplus ACT was likely to arise. This was because receipt of the dividend did not give rise to a tax credit with the result that the UK resident company had to pay the full amount of ACT on the dividend paid (with no deduction), and because relief for foreign tax reduced or eliminated the UK mainstream corporation tax liability on the dividend, the result was that there was less tax against which the ACT could be offset. To ameliorate that position, a foreign income dividend (FID) regime was introduced in 1994. ACT was payable on a FID, but recoverable to the extent that it matched the foreign income dividend received. The corollary was that, unlike ordinary dividends, a FID carried no repayable shareholder credit, although individual shareholders were treated as having received income which had borne tax at the lower rate.

[10]Claims were brought by UK resident companies, including BAT, which held shares in companies resident in another member state or in a non-member country, seeking repayment of and/or compensation for losses arising from the application of the partial imputation system. The claims centred upon the argument whether the material aspects of that system were contrary to EU law.

[11]The starting point for BAT, so far as its claims were concerned, was the claim of the BAT Group against HMRC in the Chancery Division of the High Court (Claim no. HC03C02223) which was issued on 18 June 2003 and enrolled in the FII GLO made on 8 October 2003. The BAT claim sought restitution for payments of ACT and corporation tax on income charged to tax under Schedule D, Case V of s 18 of the Income and Corporation Taxes Act 1988 (“DV income”) made by mistake, on the ground that those payments were contrary to EU law. The claim concerned tax paid from 1973 to 1999 (ACT was repealed for distributions on and after 6 April 1999). The BAT claim became in due course the test claim in the FII GLO.

[12]The FII GLO litigation was tortuous. The following is only a very brief summary of the principal elements of that litigation. When the original trial of the test claims commenced in June 2004, questions were immediately referred to the Court of Justice of the European Union (“ECJ”). Following an opinion of Advocate General Geelgoed delivered on 6 April 2006, the ECJ delivered its judgment in Test Claimants in the FII Group Litigation v Inland Revenue Commissioners (

Case C-446/04

)

[2007] STC 326

(“ FII(ECJ)1 ”). The matter then returned to the High Court to consider a number of issues, including a factual issue which would determine whether or not the UK corporation tax regime for taxation of foreign dividends under Sch D, Case V was compatible with EU law, and questions of remedies.

[13]Mr Justice Henderson (as he then was) gave his judgment in Test Claimants in the FTT Group Litigation v Revenue and Customs Commissioners

[2009] STC 254

(“ FII(HC)1 ”) on 27 November 2008. That judgment was the subject of appeals to the Court of Appeal which, in its judgment on 23 February 2009 reported at Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners

[2010] STC 1251

(“ FII(CA)1 ”), reversed FII(HC)1 in certain respects, in particular in relation to limitation and jurisdiction, and itself decided to refer certain further questions to the ECJ. The reference was remitted to be made by Henderson J. It included a further issue for reference added by the Supreme Court following the granting of permission to appeal from FII(CA)1 .

[14]The Supreme Court considered and determined certain questions of jurisdiction and limitation in its judgment of 23 May 2012: Test Claimants in the Franked Investment Income Group Litigation v Revenue and Customs Commissioners

[2012] STC 1362

(“ FII(SC) ”). It decided that one provision, s 107 of the Finance Act 2007, which disapplied the extended limitation period under s 32(1)(c) of the Limitation Act 1980 in respect of certain mistake of law claims relating to tax, was contrary to EU law and decided to make a third reference to the ECJ in respect of one issue, namely whether a retrospective amendment by s 320 of the Finance Act 2004 to the limitation period for a cause of action based on mistake was compatible with EU law. That reference was made on 25 July 2012.

[15]Following the opinion of Advocate General Jääskinen on 19 July 2012, the ECJ delivered its judgment on 13 November 2012 in Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners (

Case C-35/11

)

[2013] STC 612

(“ FII(ECJ)2 ”). Subsequently, in respect of the third reference, and after an opinion of Advocate General Wathalet on 5 September 2013, the ECJ issued on 12 December 2013 a further judgment at

[2014] STC 638

(“ FII(ECJ)3 ”), finding that s 320 of the Finance Act 2004 was contrary to EU law. Following FII(ECJ)3 , the Supreme Court, on 16 April 2014, made an order allowing BAT’s appeal from FII(CA)1 .

[16]In the meantime, in a judgment issued on 29 November 2013, Test Claimants in the FII Group Litigation v Commissioners of Inland Revenue

[2013] EWHC 3757 (Ch)

(“ FII(HC-Pleadings1) ”), Henderson J dismissed an application by HMRC to amend their defence and ordered HMRC to serve proposed amendments to their defence to take account of that judgment. That judgment was appealed by HMRC to the Court of Appeal (and the appeal was ultimately dismissed in Test Claimants in the FII Group Litigation v Inland Revenue Commissioners

[2014] EWCA Civ 1214

(“ FII(CA)2 ”)), but on 20 December 2013 HMRC sought to introduce new defences, including something referred to as the “hypothetical tax saving defence”:

“In any event, the measure of the benefit enjoyed by the Defendants must, for each year, take into consideration the fact that if the Claimants did overpay tax, they have accordingly borrowed more than they otherwise would have done, leading to greater (tax deductible) interest payments on their debts, thereby reducing any benefit received by the Defendants. In addition, overpayments of tax led to a diminution of taxable profits which the Claimants would otherwise have received thereby reducing any benefit received by the Defendants.”

That application was dismissed by Henderson J in an unreported judgment of 30 January 2014 (“ FII(HC-Pleadings2) ”) on case management grounds.

[17]Quantification issues then came before Henderson J in the High Court. On 18 December 2014 judgment was handed down in Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners

[2015] STC 1471

(“ FII(HC)2 ”). That judgment having been appealed by HMRC, and there having been a further appeal against summary judgment given by Henderson J in respect of the FIDs elements of the FII GLO, on 24 November 2016 the Court of Appeal handed down its judgment on those appeals in Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners

[2017] STC 696

(“ FII(CA)3 ”). There matters currently rest, subject to an outstanding application by HMRC to appeal to the Supreme Court. The FII GLO and the compound interest award

[18]As a consequence of this lengthy litigation, it was concluded that BAT, as test claimants, could rely on their mistake-based claims to recover the tax which they had overpaid, with interest, dating back in some cases to 1973. All remaining issues as to liability and quantification came before Henderson J in FII(HC)2 .

[19]In respect of the calculation of unlawful ACT, Henderson J considered rival methodologies put forward by BAT and HMRC. He determined that the methodology that had ultimately been propounded by BAT – the “CT61” method – was, subject to certain revisions to ensure that the method more faithfully replicated the UK’s franked investment income system, to be preferred. The method’s title referred to the specified form on which companies made their quarterly ACT returns, reflecting the fact that the method was based on the machinery of the ACT system.

[20]The CT61 method is set out by Henderson J at [151] of his judgment. It entailed envisaging a virtual CT61 which assumed an entry for deemed credits in respect of EU dividends to reflect the lawful position under EU law. That credit was then treated in the same way as franked investment income, passing through the group and ultimately being set against ACT paid in respect of dividends in the same way as a domestic tax credit. In broad terms, and subject to certain adjustments to meet particular features of the available credit for EU dividends, that excess ACT so relieved was the ACT unlawfully levied.

[21]As ACT had been set against mainstream corporation tax without distinguishing between lawful and unlawful ACT, it was necessary to determine how and when the unlawful ACT should be treated as having been utilised for the purpose of the claims. That required the making of certain assumptions. Again, rival approaches were put forward, as described by Henderson J at [191] to [206]. In this case it was a combination of the two approaches that was preferred. HMRC’s pro rata approach, which essentially treated all the payments or applications of ACT which actually took place as having been comprised of lawful and unlawful ACT on a pro rata basis, was taken as the starting point, but Henderson J accepted certain criticisms of exceptions introduced by HMRC to the general pro rata approach, holding, at [205], that all payments, surrenders and applications of ACT which actually took place were to be regarded as having been composed of both lawful and unlawful ACT on a pro rata basis across the board, with the limited exception that actual payments and repayments under the self-contained FIDs regime should be taken into account.

[22]It is worth noting that, in rejecting BAT’s primary approach, which sought to establish an order of priority for the use of lawful and unlawful ACT on an assumption as to how the company would have utilised ACT if it had been aware of the distinction between lawful and unlawful ACT, Henderson J, at [204], rejected what he described as the hopeless enterprise of attempting to reconstruct what would have happened in the absence of unlawful charges which were unknown to everybody at the time, and which would involve cash flows which could never have been replicated in the real world. That has resonance when I come to consider the various “counterfactuals” – or hypotheses - as to how BAT would have operated in the relevant period which formed a significant part of the evidence, both factual and expert, before me.

[23]Having dealt with issues of quantification, Henderson J turned, at Section VI of his judgment, to remedies. He noted, at [259], that as a matter of EU law the restitution required to give full effect to a San Giorgio [1] claim may go significantly beyond the purely subtractive remedy which would be required by the English law of unjust enrichment. He noted that the payment of interest had to provide the claimant with an “adequate indemnity” for its loss; it had to represent the time value to the claimant (not the defendant, in this case the State) of the loss of the use of the sums overpaid. The remedy, which Henderson J described at [261] as having a hybrid character in terms of classification under English law, and could conveniently be referred to as “restitutionary compensation”, is one which under the EU principle of effectiveness the English law of unjust enrichment must provide by being moulded or adapted accordingly. At [260], Henderson J, whilst recognising that the remedy remained essentially restitutionary, in the sense that its core component is reimbursement of the unlawful tax, and repayment of amounts directly linked with it, any wider loss being recoverable, if at all, in a Factortame [2] claim for damages, made clear that the measure of the restitution is “the loss occasioned to the claimant by the payment of the tax, not the enrichment of the State by its receipt”.

[24]In FII(HC)2 , HMRC argued that, although the market value of money over a period of time is normally measured in the modern world by compound interest, it was open to a defendant to show that the actual benefit to him from the use of the money was less than its market value, thereby reducing the restitutionary amount. That argument was rejected by Henderson J on a number of grounds. In particular, Henderson J held that such an argument would be precluded under EU law, concluding, as he had done in Littlewoods Retail Ltd and others v Revenue and Customs Commissioners

[2014] STC 1761

, at [302], that the reference to “adequate indemnity” by the ECJ in Littlewoods Retail Ltd v Revenue and Customs Commissioners (

Case C-591/10

)

[2012] STC 1714

, at [29], required payment of an amount of interest which is broadly commensurate with the loss suffered by the taxpayer of the use value of the tax which he has overpaid, running from the date of payment until the date of repayment.

[25]At [433], referring to restitution of the time value of the prematurely paid (or utilised) ACT, Henderson J held that the entitlement of the claimants ran from the dates of payment to the dates of utilisation, in other words according to the pro rata utilisation of the lawful and unlawful ACT, and that the time value of the claims was to be measured by reference to compound interest.

[26]In respect of the interest rates to be applied, BAT were, as Henderson J recorded at [450], content to take the cost of government borrowing as a reasonable proxy for the time value of its own loss. Accordingly, those were the rates which the court found should be applied to all of the time value claims. In the result, the parties were able to agree the relevant computations, and they were reflected in the court’s order of 18 April 2015.

[27]The judgment of the High Court having been appealed by HMRC to the Court of Appeal in FII(CA)3 , that appeal was, so far as is material to this case, dismissed. In concluding that, as a matter of EU law, HMRC’s arguments that there should be reductions in the amount of restitution by reference to shareholder tax credits, the actual benefit to HMRC having been less than the objective use value of the prematurely paid ACT, and a “change of position” defence, were precluded, the court said, at [335]:

“… we think it is now reasonably clear, as the judge said, that where domestic law allows two possible remedies to vindicate a San Giorgio right the EU law principles of equivalence and effectiveness require that both those remedies are moulded so as to vindicate the Claimants' San Giorgio rights to recover the overpaid tax and an adequate indemnity for the losses occasioned (see para 29 in Littlewoods CJEU ). As we have pointed out already, it is hard to 'mould' a domestic restitutionary claim which looks only to the enrichment of the defendant to satisfy such an unequivocally compensatory right. But that is what must be done.”

[28]In neither the judgment in FII(HC)2 nor the judgment in FII(CA)3 was there any question of the quantification of the claimant’s loss taking account of tax on the award or tax which would have been paid on the interest if it were treated as having accrued over the period for which it was awarded. In other words, the measure of the loss to BAT and the computation of the compound interest, on the basis described above, was on a gross basis.

[29]There is accordingly, in my judgment, no basis for the submission made by Mr Aaronson that tax benefits must be taken to have been factored in to the amount of the compound interest awarded. The fact that it was found, both in the High Court and the Court of Appeal, that the domestic law remedy must be moulded or, as the Court of Appeal put it at [202], that there must be a “total transformation” of the domestic law remedy in order to provide full compensation for overpaid tax, does not in my view lead to the conclusion that any remedy determined upon must be regarded as having taken into account all the tax effects for the claimant. I accept Miss Foster’s submission that the moulding or transformation of the English law remedy of restitution was directed towards the rejection, as a matter of EU law, of the arguments put by HMRC with a view to reducing the amount for which restitution was to be required.

[30]In this regard, however, Mr Aaronson pointed to the refusal of Henderson J, on case management grounds, to permit HMRC to amend their defence to include what is described as the “tax saving” defence. I have referred earlier to the proposed amendment, put forward as part of a new paragraph 31 of HMRC’s defence. For ease of reference I repeat the text here:

“In any event, the measure of the benefit enjoyed by the Defendants must, for each year, take into consideration the fact that if the Claimants did overpay tax, they have accordingly borrowed more than they otherwise would have done, leading to greater (tax deductible) interest payments on their debts, thereby reducing any benefit received by the Defendants. In addition, overpayments of tax led to a diminution of taxable profits which the Claimants would otherwise have received thereby reducing any benefit received by the Defendants.”

[31]In his judgment in FII(HC-Pleadings2) , Henderson J refused HMRC permission to amend the defence on the ground that it was simply too late, on any reasonable view, to allow points of that nature to be pleaded.

[32]That, in my judgment, takes the matter no further. It is evident that this was no more than a further example of HMRC seeking to argue that the restitutionary award should be reduced because the actual benefit to the State should take into account the reduction in tax which, on the assumptions put forward by HMRC, would have been received by the State as an indirect consequence of the overpayment of ACT and Case V tax. There was no argument as to the effect of taxation of the interest, whether actual in the sense that the interest award would, as was common ground, absent Part 8C have been subject to corporation tax in the usual way (see Shop Direct and others v Revenue and Customs Commissioners

[2014] STC 1383

, in the Court of Appeal), or notional, if the interest was assumed to have accrued over the relevant period.

[33]Mr Aaronson sought to argue that, if HMRC had wished to make a case that the basis used to quantify the claims would produce a “windfall” for the claimants by reference to the higher tax that HMRC claimed would have arisen if the interest had been taxed on an assumed accruals basis over the period, that case ought to have been made at the trial before the High Court. As this was not done, Mr Aaronson argued that the judgment must be taken as representing the amount of the claimant’s loss.

[34]I do not agree. As there was no question of the tax effects, whether deemed or actual, being taken into account, it is clear that the amount of the award was determined on a gross basis. There was no grossing up for tax which was payable on the award of interest, and no account of the net of tax loss if the interest was to be treated as accruing over the period. The judgments make clear that the award represented, albeit by way of proxy, the time value of BAT’s loss, but that time value was a gross value, with no tax benefits (or charges to tax) being factored in to the amount of the compensation awarded. Introduction of Part 8C CTA

[35]Part 8C CTA took effect in general from 21 October 2015 and from 26 October 2015 in relation to withholding. Part 8C was given effect by a House of Commons resolution on 26 October 2015, pursuant to the Provisional Collection of Taxes Act 1968. It was subsequently enacted on 18 November 2015 by s 38 of the Finance (No 2) Act 2015.

[36]The following explanation of the rationale for the legislation was given in the Explanatory Note to what was then new clause 8 of the Summer Finance Bill 2015:

“29. Under the law as it currently stands, payments of restitution paid by HMRC to companies, which relate to interest, are subject to Corporation Tax at the standard rate. However, the interest payments targeted by this clause will be subject to a higher rate, to reflect the particular circumstances of these awards. These include the number of years over which the overpayments were made, the fact that any such awards would be calculated on a compound basis and the historic corporation tax rates that applied during the years to which these claims relate.”

[37]A Tax Information and Impact Note (“TIIN”), published on 21 October 2015 stated:

“ Policy objective The interest element of any restitution award to a company is liable to Corporation Tax. That means that the interest would be taxed at the current historic low rate of Corporation Tax. This measure ensures that the rate of Corporation Tax applicable to payments of restitution interest made by HMRC reflects both the rates of Corporation Tax over the period to which typical awards relate, and the effect of compounding interest not taxed in the year to which it relates. This is a unique set of circumstances and this measure ensures that recipients of such restitution interest payments do not enjoy an unfair tax advantage at the expense of the public purse.”

[38]I was also taken to the Hansard report of the report stage of the Finance Bill in the House of Commons on 26 October 2015, and to the following remarks of the Financial Secretary to the Treasury (Mr David Gauke MP) (Hansard, Monday 26 October 2015, Vol 601, No 57, at cols 52-53):

“New clause 8 addresses an unfairness whereby in certain claims for repayment of tax and restitution through interest payments, taxpayers might receive a significant additional benefit at the expense of the public purse. The vast majority of interest payments that are paid by [HMRC] are made under the relevant Taxes Act. These will continue to be subject to the normal rate of corporation tax. However, the interest payments targeted by this clause arise from claims made under common law, which stretch over a large number of years – in some cases, going back to 1973 – and represent a unique set of circumstances. As it stands under current law, any payments will be taxed at the low corporation tax rate that applies at the time the payments are due to be made. Since the interest payments targeted by the clause have accrued over years when the rate of corporation tax was much higher than companies currently enjoy, those making the claims received a significant financial benefit. In addition, such payments may have to be calculated on a compound basis, further improving the advantage gained at the expense of the public purse. … … that measure is targeted at very specific circumstances in which compound interest may have to be paid in relation to claims which, as I have said, potentially date back to 1973 … … such payments may have to be calculated on a compound basis, which would increase the advantage gained at the expense of the public purse. To address that unfairness, the Government are ensuring that an appropriate amount of tax, set at a rate of 45%, is paid on any such awards. That rate reflects the long period over which any such interest accrued, the higher rate of corporation tax which applied during the period, and the compounding effect of such potential awards … New clause 8 will ensure that a principled and targeted system is in place to address a potential unfairness whereby a few businesses receive significant benefits resulting from the unique nature of this litigation at the expense of the public purse.”

[39]In his evidence to the tribunal, Mr Martyn Rounding of HMRC confirmed that there had been no public consultation before the introduction of Part 8C in this way. There was no evidence of any prior representations: the reference in Hansard, at col 54, to the remarks made by the Financial Secretary to there having been “no shortage of representations”, which were put to Mr Rounding by Mr Aaronson, clearly related to a different provision of the Bill, namely certain technical changes to the tax treatment of “carried interest”.

[40]In its closing submissions, by way of a note on the evidence, BAT sought to submit that the stated purpose set out in these various materials could not have been HMRC’s true objective. This was based in part on an assertion that Part 8C was the latest in a series of attempts by HMRC to deprive the FTT GLO claimants of the “full fruits of their victories” in the ECJ by the introduction of unlawful legislation, namely s 320 of the Finance Act 2004, s 107 of the Finance Act 2007 and s 234 of the Finance Act 2013 (restrictions on interim payments in proceedings relating to taxation matters).

[41]It is not necessary to examine this submission in detail. I do not accept it for two reasons. The first is that the claim was not put to any witness of HMRC. Mr Rounding was taken to the various statements as to the objective of Part 8C, to which I have referred above. He confirmed that those statements represented the rationale for the introduction of the new law. He was asked about the lack of consultation, and accepted that certain changes had to be introduced to deal with the position of insurance companies and charities, after those issues had been raised. But it was not suggested to him that HMRC might have had a different motivation, apart from that expressed in Parliament, for the introduction of Part 8C.

[42]The second reason is more fundamental. As Miss Foster submitted, it is not for this tribunal to go behind the published statements to enquire into the reason why particular legislation has been introduced. What is relevant to this appeal is the effect of the legislation and whether, having regard to its effect, it is lawful as a matter of EU law or the law on human rights. It is well-established that the function of the court and the tribunal is to consider and apply the enactments of Parliament and that it is not appropriate to seek to impugn the validity of a statute by seeking to establish that Parliament, in passing it, was misled by fraud or otherwise (see British Railways Board v Pickin

[1974] AC 765

[43]At any event, the submission is not made out. I do not accept that there can be any inference to be drawn from prior legislation having been introduced to curtail the rights of claimants or that legislation having been held to be unlawful, or that the policy decisions in relation to Part 8C appeared to coincide broadly with the introduction of a simplified HMRC debtor and creditor interest rate which, BAT submitted, favoured HMRC over claimants. Nor can purpose be determined by reference to whether it can be shown – a question which is at the core of this appeal – that Part 8C is incapable of justification. I will also address the significance of the fact that the decision that reliefs and allowances of individual companies should not be taken into account was one that was taken at the outset, and before any work had been carried out on methodology. But, again, that does not in my view point towards the purpose of the Government in introducing Part 8C being anything other than that which it set out. The 45% rate – policy and methodology

[44]I had evidence with respect to the policy underlying the introduction of Part 8C, the scope of the analysis that had been determined and the methodology employed in arriving at the tax rate of 45%. Policy considerations

[45]Evidence of the underlying policy was given by Mr Rounding, an Assistant Director of HMRC with overall responsibility for litigation matters within the Corporation Tax, International and Stamps (“CTIS”) division. Mr Rounding has overall responsibility and oversight for the provision of advice to HMRC on how Part 8C should apply to restitution-based interest and for development of any future policy changes in that respect.

[46]Mr Rounding’s evidence explained the twin “windfall” elements which HMRC considered attached to recipients of restitution interest. The first was the fact that such interest would be taxed at the then current, historically low, rate of corporation tax (20% at the time of introduction of Part 8C, falling to 19% for years commencing 1 April 2017, 2018 and 2019 and at 18% for the year starting 1 April 2020), rather than at the higher rates that applied during the periods to which the awards of interest related (a rate of corporation tax of 52% applied between 1973 and 1982) and which would have been charged had the interest been paid in each of those earlier years. The second was that the claimant benefitted by the fact that the award would have been calculated by compounding on a gross basis, and not net of the tax that would have been paid on an accruals basis over the periods in question. That would result in a substantially increased award, especially over a typical length of claim period, as compared with compounding net.

[47]That, explained Mr Rounding, was the rationale for the decision to redress the identified windfall by the imposition of a special rate of tax. The objective of the proposed legislation was to place the claimants in broadly the same position as a taxpayer company who had received the same amount of interest in each relevant accounting period and had paid tax on that interest at the relevant due and payable dates for each accounting period. In this way, according to Mr Rounding’s evidence, the net effect of any remedy was intended to remain fair and adequate as required by EU law, but would not include the identified windfall at the expense of the public purse.

[48]Mr Rounding’s evidence was to the effect that the legislation as developed did not target any particular company. Although, as Mr Aaronson put to Mr Rounding, there was an element of targeting, in that the constituency of those affected by Part 8C was limited to those entitled to an award of restitution interest, and the Financial Secretary had referred to the measure being targeted to the very specific circumstances of those awards, I find that the targeting was of those circumstances and not of any individual company. The comparator was not intended to be any specific company, but a hypothetical company that had paid tax on an assumed accrual of interest over the claim period.

[49]Mr Rounding candidly admitted that the original policy had failed to take account of certain special classes of company that might receive restitution interest. That included life insurance companies to the extent that the interest would be taken into account, under s 73 of the Finance Act 2012, in the I-E calculation as referable to basic life assurance and general annuity business and which represented policyholder income, which historically had had a zero, or low, rate of taxation. That interest was recognised as falling outside the mischief of Part 8C, and was excluded, along with an exclusion for charitable companies, by the Corporation Tax Act 2010 (Part 8C) (Amendment) Regulations 2017. (Those amendments are to be found in s 357YDA and s 357YA(2) and s 357YB(2) respectively.) On the other hand, companies with substantial ACT carried forward and unutilised against mainstream corporation tax were not considered to be a special class in the same way as life insurance companies or charitable companies, nor were companies, such as those with FID claims (where from 1999 onwards interest had been calculated at simple interest at the statutory rate, but which, as part of a wider award including compound interest, was within the scope of Part 8C).

[50]In setting the policy with respect to the proposed legislation, one consideration was whether the special rate should be a single rate of tax or a tailored rate, that is one that was tailored to the specific circumstances, including the availability of reliefs, in the relevant period. Mr Rounding’s evidence was that the possibility of a tailored rate had not been adopted, principally because the information necessary for HMRC to have a complete understanding of the tax affairs of claimants was not available to HMRC. It was not considered reasonable for Parliament to have set a rate based on individual company circumstances.

[51]Mr Rounding identified what he described as significant difficulties in drafting legislation to express the very complex formulae which he considered would be necessary to have a tailored as opposed to a single headline rate in a way that was not opaque or delegated too much power as regards rate-setting to the executive, effectively arriving at individual rates for companies by means of negotiation. One problem was that there was a significant period over which restitution interest could arise, and consequently the period for which information could be required. Full or partial repayments or set-offs of tax would take place at different dates, and any algorithm or formula would have to take account of every identifiable period of set off to cover all scenarios. Mr Rounding said that he and his colleagues had calculated that for payments between 1973 and 2000 something in the region of 1,000 separate rates would be required in the legislation. If calculations were more accurately done on a monthly basis, to take account of monthly set-off, thousands of rates would be required.

[52]By contrast, a special single rate, based on a considered and principled methodology, was considered to have the advantage of simplicity, avoiding the need for complex calculations. In calculating that rate, the methodology employed did not take into account the existence of available reliefs or the special circumstances of all companies that were affected by the legislation. This, explained Mr Rounding, was because it was considered important to carry out the underlying analyses on a like-for-like basis, ignoring variables such as the existence and utilisation of reliefs by specific companies. It was considered impracticable to second guess decisions of claimants in using reliefs and/or in re-opening closed periods to disclaim allowances and set-offs.

[53]By way of further explanation, Mr Rounding said that a factor in determining that reliefs should not be taken into account was that the exercise involved the setting of a rate by reference to an examination of a significant number of years. Although a relief may occur in a particular year, and there may be excess reliefs for a particular year, there is a substantial likelihood that such reliefs will be set off in subsequent years. Evidence in that respect was also given by Mr Paul Lane, an HMRC officer in the Large Business Directorate. Mr Lane noted in his evidence that, as a general matter, in practice if reliefs were used to cover additional profits in one year those reliefs would no longer be available for offset in the following year, and that in those circumstances the utilisation of losses might only create a timing effect. This would depend on the facts of an individual case, as Mr Lane accepted, including that of BAT where the position with respect to the utilisation of reliefs was known.

[54]Mr Rounding also addressed an alternative way of introducing a tailored rate, namely to introduce a headline rate and to use regulations to tailor that rate to the specific circumstances of each individual taxpayer. He viewed this as extremely complex to draft, and inappropriately so, as companies would not know what rate would apply to them without complex calculations and extended discussions with HMRC. That, Mr Rounding said, would have led to material uncertainty for companies as to what tax rate would be applied. The need for legislation allowing HMRC officials to set the individual rate would, in Mr Rounding’s view, have probably led to an unprecedented delegation of power, and was not therefore considered to be a sensible way to legislate. The HMRC methodology

[55]The analysis to arrive at a rate of tax to be applied to restitution interest was carried out, on the instructions the CTIS team, by the Knowledge Analysis and Information directorate (“KAI”) at HMRC. I heard evidence from two members of KAI, Ms Rebecca Richmond and Mr Neil Duncan.

[56]The instructions given to KAI were directed at devising a methodology from which would be produced a set of data to indicate, based on information gleaned from a sample of cases, the rate (or a range of rates) at which tax would have to be set in each sample case to fairly mitigate the effects of the “windfall” factors (lower than historic rates and gross compounding). Ms Richmond explained that the first task was to produce a range of calculated rates and then to produce a weighted average corporation tax rate that would be relevant to all claims, in that it would, on average, remove any windfall otherwise enjoyed by the claimants.

[57]In essence, the methodology compares the interest compounded over the whole period (up to 31 March 2015, although claims may extend beyond that date) without any deductions of corporation tax with a “counterfactual” where corporation tax is deducted each year at the relevant nominal rate from the interest, reducing the amount compounding over the period. Where a principal sum is repaid to a claimant, that sum is deducted from any amount outstanding in the relevant year, and thus reduces the compounding effect. Likewise any statutory simple interest is deducted.

[58]The initial calculations were done on an annual basis, reflecting the financial year. It was assumed for this purpose that tax was paid two years after the end of the financial year, and that the recipients of restitution interest would not pay by quarterly instalments. As that delayed the assumed tax deduction, it resulted in a lesser deduction for net, as opposed to gross, compounding. An annual calculation, involving assumed payments and repayments as having been made at the same time during the year was not considered overall to have any material effect on the rate: the use of an annual calculation may have increased the estimated rate slightly, but the modelling of the repayment lag was likely to reduce the rate.

[59]During initial testing of the methodology on a range of cases, KAI used the 10-year rolling average rates of 10 year gilts for the FII GLO claims, corresponding with the rate used to calculate the restitution in those claims. Following certain refinements in relation to VAT claims, the actual rate of interest in the lead claim was employed.

[60]Ms Richmond’s evidence explained how the methodology applied to the facts of BAT’s appeal. The data used in the calculations was the agreed data for the purpose of the calculation of the award to the BAT claimants. Payments and repayments within each financial year were aggregated in order to simplify the model; those simplifications, after compounding by reference to the 10-year gilt rate, resulted in a figure of £1,220,272,098 which was 6.6% higher than the aggregate compound interest amount actually awarded to the BAT claimants (which had employed a more precise daily calculation).

[61]The counterfactual case of calculating compound interest net of tax employed the same simplifications. The methodology was as follows:(1) For each year of the claim, the following are calculated:

a. A. The net difference between overpayments and repayments in the financial year (in-year overpayments minus in-year repayments)

b. B. The cumulative total outstanding (total tax overpaid less total tax repaid plus interest due (net of tax repayments))

c. C. In-year interest accruing on the cumulative total (cumulative total multiplied by the interest rate for that year)

d. D. Tax accruing on the in-year interest (in-year interest multiplied by the in-year tax rate)

e. E. Tax to deduct in-year. This is lagged as follows: a. 1973/74 until 1988/89: tax accruing on interest two financial years previously (reflecting 15 month (or 13 month in 1988/89) gap between the end of an accounting period and the date on which corporation tax was due; b. 1989/90: tax accruing on interest two years previously (due to introduction of 9 month lag between the end of an accounting period and the date on which corporation tax was due); c. 1990/91 until 2013/14: tax accruing on interest in the previous financial year (to reflect 9 month lag between the end of an accounting period and date on which the corporation tax was due)

f. F. Total outstanding (B plus C minus E)

g. G. Interest to be paid (C minus E)

(2) For BAT, the total counterfactual compound interest after tax produced by this calculation is £426,996,061. After deduction of the final year’s accrued tax (due the following year) the overall figure is £423,586,833. Comparing that figure with the gross amount of compound interest calculated as above produces a reduction of 65.29%. When adjusted to reflect actual dates of payments and repayments (rather than those calculated on an annualised basis), the reduction is calculated at 66.26%.

[62]The basic methodology, as applied to each claimant considered, was also helpfully summarised, uncontroversially, by the expert witness instructed by HMRC, Professor Michael Devereux, in his first report, in the following way:(1) First, for any unlawful tax payment in any year, calculate the current value (for example in 2015 or 2016) of that payment in the absence of tax by cumulating forwards using compound interest. For example, consider an unlawful payment of £100 at the beginning of 2013-14, and suppose that the interest rate is fixed at 10%. Then the current value of the unlawful payment would be £110 at the end of 2013-14 and £121 at the end of 2014-15.

(2) Second, deduct the original unlawful payment to find the current value of the cumulated interest. In the example at (1) above, at the end of 2014-15 this would be £21.

(3) Third, aggregate the current values of all such cumulated interest for a particular taxpayer – call this value X.

(4) Fourth, repeat the previous three steps, but allowing for corporation tax that would have been paid on the hypothetical interest due in each period – call this Y. To continue the example, and ignoring any lags in tax payments, at a tax rate of 20%, tax of £2 would be due on the £10 of hypothetical interest received at the end of 2013-14. This means that the current value of the unlawful tax payment would then be only £108 at the end of 2013-14. Following the same approach, at the end of 2014-15, the current value of the unlawful tax payment would be £116.64. The current value of the cumulated hypothetical interest after tax would therefore be £16.64.

(5) Fifth, calculate the “implicit tax rate” as the percentage difference between £21 and £16.64, which is 20.8%.

[63]As Professor Devereux goes on to explain, this approach measures the percentage difference between the current value of all restitution interest, with and without corporation tax being applied in the relevant period. The resulting calculated implicit rate depends on:(a) the corporation tax rate applied to the hypothetical interest in each period (in some cases going back to the 1970s),

(b) the length of time over which compound interest is calculated, and

(c) the interest rate used in each period. Applying tax in each period reduces the net rate of compound interest, which can have a significant effect on the resulting implicit tax rate.

[64]This methodology was used on a representative group of cases to produce a weighted average corporation tax rate that would take into account all claims and would be designed to remove the windfall otherwise enjoyed by the claimants. All available large cases were included in the sample, along with a representative number of smaller claims. The relevant groups were described as follows: the FII GLO, Large VAT claims (settled), Large VAT claims (still in litigation), Small VAT claims, CFC Dividend GLO and Tax Act claims.

[65]A number of possible weighting methods were considered. That chosen was to weight by the size of the principal sums involved in the claim, but not taking into account repayments of principal or utilisation. It was preferred to simply taking the average across the rate calculations for each case, and weighting by size of the claim as a whole (including interest); that would give greater weight to cases covering longer time periods.

[66]A number of adjustments were made. First, for the smaller cases that had been sampled, the weights were grossed up to represent the larger group. Secondly, for Tax Act claims, grossing up was based on an estimated 34% of such claims which could potentially attract compound interest. Thirdly, the FII GLO cases were grossed up based on the number of cases for which data was available to the total number of expected claims. Data was available on nine claims and a total of 40 claims was expected, so that the appropriate multiplier of the principal sums was a factor of 40/9 = 4.44, when calculating the weighted rate. (In fact, as set out in the evidence of Mr Duncan, an error in the calculation in September 2015 meant that the actual multiplier employed had been 49/9 = 5.44. Once the error was discovered, in September 2016 there had been a recalculation, resulting in a very small reduction in the overall weighted rate from 44.77% (as originally calculated) to 44.29%.) Finally, as it was uncertain to what extent certain VAT cases would involve a compound interest claim, the weights applied to those cases were reduced by 50%.

[67]The weighted average rate arrived at on the basis of the original calculations, and on which the 45% tax rate was based, was 44.77%. As noted above, but for the arithmetical error in applying weighting to the FII GLO claims, the average rate would have been 44.29%. BAT’s position

[68]Witness evidence of fact (as opposed to expert evidence) was given in relation to BAT’s tax and financial position by the following:(1) Mr Anthony Cohn, a chartered accountant and chartered tax adviser, who is the Regional Tax Manager for the BAT Group with responsibility for Eastern Europe, the Middle East and Africa;

(2) Mr Neil Wadey, the Group Treasurer of the BAT Group;

(3) Mr Kenneth Hardman, a retired chartered accountant who was the group head of tax of the BAT Group until his retirement on 31 December 2014; and

(4) Mr David Woods, a corporate tax specialist in the tax department of the BAT Group. I also had a number of witness statements made by Mr Simon Whitehead, a partner in Joseph Hage Aaronson LLP, which amongst other things provided background information about the proceedings and participants in the FII GLO. Those statements were unchallenged. BAT’s surplus ACT problem

[69]Mr Hardman provided a witness statement which was, in the end, unchallenged. He explained that since 1973 BAT had been a global business operating in local markets through local subsidiaries. The BAT Group’s policy was to repatriate profits where possible. Those dividend receipts were liable to corporation tax under Case V of Schedule D, but the liability was reduced to a great extent by double tax relief. The result was that there was insufficient mainstream corporation tax liability against which the ACT on onward dividends could be offset. Surplus ACT was created. This was written off through BAT’s profit and loss account to the extent that it was treated in effect as a permanent tax.

[70]Mr Hardman’s statement gave detail as to how the ACT problem was managed by the group, including the rationale for the payment of FIDs, which from the introduction of the regime from July 1994 could be paid so long as they matched dividend income received from foreign subsidiaries carrying a full rate of double tax relief. FIDs attracted ACT in the normal way, but the ACT could be reclaimed and was repaid and not therefore offset against mainstream corporation tax. The statement also explained the rationale for so-called “enhanced FIDs” which compensated shareholders for the fact that the FIDs carried no shareholder tax credit (unlike ordinary dividends). It is not necessary for me to make any more detailed findings in those regards. BAT’s modelling

[71]Evidence of the calculations made by BAT to ascertain, on the basis of its own assumptions, the percentage difference between the net benefit to the BAT group on the basis of the counterfactual approach in HMRC’s model as against the amount of the interest award, was given by Mr Cohn. Mr Cohn made six witness statements, and his evidence was of an iterative nature. Over time he adjusted his computations to take account of points raised, both by HMRC’s witnesses and by the expert witnesses. In oral evidence he provided a very helpful interactive run-through of the salient spreadsheets and their underlying calculations. Mr Cohn’s basic model

[72]Mr Cohn’s starting point, in terms of historic corporation tax rates, was to annualise those rates to correspond with BAT’s accounting periods which ran to 30 September until 1979 and were thereafter calendar year ends. Instead of using estimates of the restitution interest as calculated for the purpose of HMRC’s model, Mr Cohn used the actual amounts of interest awarded to BAT. In essence, the data used by Mr Cohn in his calculations was derived from the agreed tax computations or the “CT61 method” used for the calculation of the award.

[73]Unlike the HMRC model, Mr Cohn’s computations take into account reliefs available to BAT in the relevant periods. He applies the tax provisions applicable at the time to establish how much of the notional tax liabilities under HMRC’s model remained for payment after the utilisation of reliefs.

[74]The first step in the calculation is for Mr Cohn to take the actual amount of restitution interest that was awarded and break that down into each accounting period. This exercise was done in relation to the relevant individual companies in the group, and aggregated per accounting period.

[75]To those aggregate amounts for each accounting period were applied the annualised tax rates described above, to give a tax liability, before the set off of any reliefs.

[76]The next step, in relation to the period from 1973 to 1998 (prior to the abolition of ACT), was to identify the largest ACT capacity of a single company in the BAT group. The purpose in doing so was to demonstrate that if there was sufficient capacity in that one company, there would be sufficient capacity in the group as a whole, and it would not be necessary to trace ACT capacity further. Mr Cohn explained BAT’s ACT capacity as having arisen over the whole period of the ACT regime largely as a result of its foreign income (which was not franked investment income and in respect of which mainstream corporation tax was reduced or eliminated by foreign tax credits) exceeding its UK domestic income. The dividends, on which ACT arose, were largely funded out of foreign income. Only a small part of that ACT could be set against mainstream corporation tax and the surplus built up over time.

[77]Notably, at this stage in the iteration, Mr Cohn did not seek to distinguish between lawful and unlawful ACT. All ACT that had actually been paid by the BAT group, whether lawfully or unlawfully, was counted as part of the available ACT to the extent that it had not already been utilised to offset mainstream corporation tax on profits.

[78]Mr Cohn then assumed that where there was available ACT in that year, that ACT was utilised against the notional additional liability to tax on the notional accrued interest. In addition, and prior to any ACT offset, Mr Cohn identified in relation to certain group companies capital allowances that had not been claimed. He assumed for the purpose of this exercise that those allowances had been claimed. He also assumed that the claims for group relief surrendered to those companies would have been correspondingly reduced, thus increasing the group relief available to be used by or surrendered to the company in which the restitution interest was treated as arising.

[79]In that period 1973-1998, only in the last three years (1996, 1997 and 1998) was there calculated to be a residual tax liability. The effect of that residual tax liability was that, with effect from 1997, the compounding was calculated on the net of tax amounts, thus reducing the notional net interest accrual for that and subsequent years.

[80]With the abolition of ACT in 1999, for BAT the tax on the marginal profit represented by the accruing restitution interest was no longer offset by ACT. For the period from 1999 to 2015, as Mr Cohn explained it, the BAT group as a whole was broadly loss-making and there was no ACT available to offset the additional tax liability. However, relief was available by means of a combination of tax losses and capital allowances. Mr Cohn’s evidence, and the accompanying spreadsheets in respect of this basic model, showed that there were tax losses carried forward in every year from 1999 to 2015, culminating at the end of 2015 with losses of almost £787 million. Furthermore, in the same period there was a carried forward pool of capital allowances in each year, with the pool of available allowances being about £1.1 billion at the end of 2015.

[81]Those reliefs were then applied by Mr Cohn to offset the tax on additional restitution interest in each year, making certain assumptions as to some losses being surrendered by way of group relief to the applicable group company. That resulted in a residual tax liability for the group as a whole only in 1999 and 2001, which was again taken into account in reducing the compounding effect in respect of the restitution interest.

[82]Having completed that exercise, two particular figures are produced. One (RI) is the reduction in the amount of restitution interest by reason of compounding having been revised to reflect the actual tax paid for years when the available reliefs were insufficient to offset the tax on assumed interest accruals in full. The other (RT) is the total for all the relevant years of the residual tax liability on that interest. The sum of RI and RT effectively gives the overall reduction, on the basis of the counterfactual, in BAT’s net receipt of restitution interest. This is the value of Y in the comparison calculation. That value of Y is then compared with the actual (gross) receipt of the award of restitution interest (X) to ascertain the percentage reduction and consequently the notional tax rate on actual receipt that would be needed to achieve the same reduction. That, in essence, applies the same principle of comparison as HMRC’s model, although with the differences that the actual interest awarded is taken as the base case, the calculations are done for BAT alone (and there is no weighted average) and, crucially, the BAT model takes into account reliefs from taxation whereas HMRC’s model does not. On the figures as originally presented, RI was £70,594,470 and RT was £56,166,051, giving an aggregate amount of £126,760,521 (Y). That was compared with the aggregate restitution interest of £1,145,674,166 (X). The result was to show that the difference, and consequently the required rate, was 11.06%

[83]That was Mr Cohn’s basic model. It was, however, adjusted on a number of occasions to reflect comments received from HMRC and in HMRC’s witness evidence produced in these proceedings. The first revision was to bring Mr Cohn’s calculations into line with those of HMRC with respect to tax being paid on a quarterly, and not an annual, basis. That had the effect that, for accounting periods 1999 and 2001 only, some tax was paid earlier than on Mr Cohn’s original basic model. This accelerated tax had an effect on the net compounding of the restitution interest, reducing the aggregate amount to some extent. The net effect was to increase the percentage difference (and the notional tax rate) from 11.06% to 11.18%. Mr Cohn’s revised model

[84]The next stage in the iterative process derived from the first witness statement made by Mr Lane. Mr Lane had been asked, amongst other things, to review Mr Cohn’s first witness statement (and thus Mr Cohn’s basic model). At that time, of course, that basic model had produced a notional tax rate of 11.06%.

[85]In his witness statement, Mr Lane challenged that result. Applying Mr Cohn’s hypothesis, that of setting reliefs against notional interest arising and setting available ACT (both lawful and unlawful) against the notional corporation tax on that interest, Mr Lane arrived at a notional tax rate of 25.35%. Principally, this was on the basis that Mr Cohn’s basic model had not taken into account the fact that if unutilised ACT could be set against the notional tax, the unlawful ACT would then be utilised, with the result that, applying the time value of money calculation that had been applied by the court in arriving at the calculation of the compound interest award, the value of the claim for compound interest would have been reduced. On Mr Lane’s calculations, if ACT had been utilised sooner, as suggested by Mr Cohn’s basic model, the BAT claimant’s total claims in restitution would have been £160,004,439 less than they actually were.

[86]After making certain other adjustments for ACT that Mr Lane’s computations showed could not have been utilised, Mr Lane arrived at a revised amount for Y of £290,444,729. Comparing that to the aggregate of restitution interest received (X) gives a difference, and thus a notional tax liability, of 25.35%. That calculation did not exclude unlawful VAT from the equation, but it took account of its earlier utilisation in reducing, for the purpose of the counterfactual, the amount of compound interest that would have been earned.

[87]Mr Lane’s first witness statement in fact went further. He explained that he had hypothesised that, having had regard to the dividend policies of the BAT group over the relevant period, BAT would have paid out a proportion of its notional additional profits, represented by the restitution interest treated as accruing annually, and assuming that there was no tax on that interest, as dividends in all the years from 1979 to 2015. He reasoned that, because the payment of a dividend would decrease the assets of the group, this would reduce the compounding effect on the restitution interest in the same way as the payment of tax would do so. The net effect of that hypothesis was that the compounded value of the retained interest would be 25.8% of the value if compounded gross. Mr Lane’s conclusion was that the windfall to the BAT claimants on that basis would be 74.2% of the restitution award.

[88]It is unnecessary to dwell on this particular aspect of Mr Lane’s evidence. In my judgment, in common with similar conclusions I have reached with respect to the scope of the counterfactual when considering the expert evidence, it has no relevance to the question of the calculation of a notional tax rate. The payment of a dividend, even if it might be one possible use of the notional additional funds available to the BAT group, is not in the same category of a charge on the restitution interest in the way a residual tax liability is, in terms of reducing the amount of restitution interest required to be compounded. That element of the methodology, which is common to both Ms Richmond’s and Mr Cohn’s model (Ms Richmond on a gross basis, and Mr Cohn after taking account of reliefs), and which flows directly from one of the mischiefs intended to be addressed by the Part 8C charge, is entirely appropriate in seeking to determine the windfall element. There is no warrant for further hypothesis, including as to the payment of dividends. Nor, I might add, would it be right to assume both a dividend and the consequent increase in payment of lawful ACT, with the increased ACT capacity that would produce, as suggested by Mr Cohn in giving his oral evidence. That only serves to illustrate in my view the endless hypothesis that such an approach would lead to, and why it is not appropriate in this case.

[89]Nor, for essentially the same reason, do I consider that it would be a legitimate hypothesis to seek to recompute the total claim in restitution by reference to the earlier utilisation of unlawful ACT. It is not the earlier utilisation of the unlawful ACT that is the flaw in Mr Cohn’s basic model; it is the application of unlawful ACT at all, in circumstances where it logically has to be assumed, for the purpose of applying reliefs against the restitution interest, that the unlawful ACT would either not have been paid, or if paid (and due to be repaid because it was unlawful) could not lawfully have been used to offset mainstream corporation tax. For this purpose it is immaterial that the judgment debt was calculated on the basis (which reflected what had actually happened) that the unlawful ACT was offset against mainstream corporation tax. That was appropriate, not only to ascertain to what extent there was unutilised unlawful ACT and thus what restitution had to be made in that respect, but also to compute the time value of the unlawfully paid ACT up to when it was in fact utilised. That is not relevant to a counterfactual which simply looks at the net of tax restitution interest and compares that to the gross amount of that interest. The way that interest has been calculated has ceased to have any relevance to that hypothesis.

[90]That leads on therefore to the next iteration of Mr Cohn’s model; his revised model. As Mr Cohn explained, his revised approach was to remove from the pool of ACT available to be utilised against mainstream corporation tax all unlawful ACT. Only lawful ACT is assumed to be capable of being utilised against the additional tax notionally arising on the accrued restitution interest. On that basis, the available ACT is exhausted by 1996, so that for that year and for 1997 and 1998 there are residual tax liabilities. There are also such liabilities for 1999 and 2001. The years in which residual tax liabilities arise accordingly match those in Mr Cohn’s basic model, although the numbers are different. The differences arise in a number of ways. First, in 1997, the residual tax liability is greater because of the non-availability of the unlawful ACT. Secondly, and on the other hand, that gives rise to a greater reduction in the compounding of the restitution interest from that year onwards. Thirdly, notional tax is chargeable on the notional interest as so reduced, and the amount of the notional tax for years 1997, 1998 and 1999 is lower than in the basic model. The net effect is that, on this hypothesis, RI is calculated to be £70,114,808 and RT is £53,230 giving an aggregate amount for Y, marginally lower than on the basic model, of £123,344,831, which is 10.77% of the actual gross amount of residual interest (X) of £1,145,674,166. On this basis, therefore, the notional tax rate would be 10.77%.

[91]That revised model, however, itself suffered from a flaw. As Mr Cohn states in his fourth witness statement, he having considered those computations in the light of the report by Professor Devereux, the revised model, as well as ignoring unlawful ACT as a relief in respect of the restitution interest, also ignored the lawful mainstream corporation tax against which it was utilised. That mainstream corporation tax would have to have been paid or met with lawful ACT. The revised model did not account for that. Secondly, the revised model had the effect of utilising lawful ACT against unlawful Schedule D, Case V tax, a consequence which in the revised model Mr Cohn had ignored for the sake of simplicity. Mr Cohn’s second revised model

[92]Mr Cohn’s second revised model accordingly took as its premise the counterfactual suggested by Professor Devereux, namely that, as a first stage, the unlawful tax would not have been paid, but would have been invested in an asset earning a risk-free interest rate. To reflect that, Mr Cohn assumed a notional account bearing interest at the rate assumed for the claim, namely the 10-year moving average of the average annual yield of 10 year gilts compounded on an annual basis. Into that notional account it is assumed that unlawful payments of ACT are deposited on the date those payments were made. Similarly, unlawful payments of Case V tax and unlawful Case V liabilities against which lawful ACT was utilised are deposited on the payment dates for those liabilities.

[93]As with the earlier revised model, it is assumed that the unlawful ACT is not available for utilisation. It has therefore been removed from the pool of available ACT. Lawful ACT which was utilised against unlawful Case V tax has also been removed. The only ACT to be used in the notional calculation is therefore lawful ACT. To the extent that lawful ACT is exhausted, notional tax liabilities arise as before. Those liabilities are treated as having been withdrawn from the notional account on the relevant payment dates, adopting the quarterly payment structure.

[94]The notional account is not therefore directed solely to the restitution interest. It includes both principal (in the form of the unlawful payments of ACT, the unlawful payments of Case V tax and the lawful ACT utilised against unlawful Case V tax) and interest. Withdrawals from that account are likewise not confined to the notional tax on restitution interest, but also include all notional tax that is calculated to have arisen if the ACT utilisation were to have been restricted to lawful ACT.

[95]After further revision of the figures to reflect a technical change from 1987 in the extent to which ACT capacity could be utilised, the result derived from this second revised model is that the total balance in the notional account, including compound interest of £1,218,898,817, is £1,150,771,475 (reflecting a negative principal amount of £68,127,342). That total balance (Y) is compared by Mr Cohn against X, which is the original value of the BAT group’s claim in the FII GLO, namely £1,184,074,056. That comparison gives a difference, and notional tax rate, of 2.81%.

[96]Although it was put to Mr Cohn that his second revised model might have failed to eliminate all of the unlawful ACT, I am satisfied, from my own review of the spreadsheets, that it did so. A particular concern was raised in cross-examination of Mr Cohn that his second revised model had failed to exclude unlawful ACT, in that certain elements of the calculation appeared to HMRC to take account of unlawful ACT. That, from consideration of the relevant spreadsheets and Mr Cohn’s evidence, seems to me to have been a misunderstanding of Mr Cohn’s calculations. It is true to say that Mr Cohn had calculated, in a table (Table 6B – Summary of ACT utilisation) the total lawful ACT paid in each relevant year available for utilisation, namely the total of the lawful ACT paid in the year less lawful ACT that had been utilised against unlawful Sch D, Case V tax, and that he had deducted from that amount the total ACT (lawful and unlawful) that had been in fact utilised against mainstream corporation tax or refunded. But that did not mean that unlawful ACT was being taken into account. It meant the opposite. What it achieved was a reduction in the amount of lawful ACT available to offset tax on the assumed accruing interest by reference to the total amount of ACT that had been used for offset. Instead therefore of unlawful ACT being used for offset (which would have the effect of preserving an equivalent amount of lawful ACT), the effect of the calculation was to assume that the whole offset was by means of lawful ACT, thereby ensuring that only the net amount of lawful ACT (if any) was available to be utilised against the notional interest.

[97]On the other hand, the notional tax rate of 2.81% was arrived at by comparing a value for Y, which included both principal and interest, with a value for X, which was the original value of the restitution interest claim. That, it seems to me, was a flaw in the calculation, and potentially a serious one, as my view is that the only proper comparison could be by reference to interest only, on both sides of the equation. However, although the inclusion of both principal and interest in Y was noted by Professor Devereux in his second report, albeit in a footnote (footnote 2), and was referred to by him in his oral evidence, no particular point was taken in this regard, Professor Devereux referring to it in his evidence as being “relatively minor”. For that reason, although I continue to harbour my own reservations, I have pursued it no further.

[98]By contrast, Professor Devereux identified in his second report what he regarded as a fundamental flaw in the second revised model. He considered that the model used what he described as a “hybrid” methodology, applying at the same time the approach of the court to the calculation of the compensation, recognising the use of unlawful ACT pro rata with lawful ACT (in arriving at a value for X), and a counterfactual in which, in the calculation of the hypothetical interest net of tax (the value for Y) the unlawful ACT, and its effects, was stripped out. In Professor Devereux’s opinion, in order to calculate an implicit (or notional) tax rate it was necessary to calculate both X and Y on a consistent basis. Thus, if Y was calculated so as to exclude unlawful ACT from the equation (assuming therefore that it had never been paid and never been used to offset mainstream corporation tax), then X could not be the amount of the award itself, as that had been computed by reference to the payments and offset of unlawful ACT. X would therefore have to be re-computed on the same basis. Professor Devereux did not produce for the tribunal any detailed calculations in this respect, but by adapting Mr Cohn’s second revised model he arrived at a value for X for comparison of £2.01 billion so as to yield an implicit tax rate of 48.6%. Professor Devereux’s figures were in turn challenged by Mr Cohn in his fifth witness statement.

[99]I confess to having had some conceptual difficulty with this approach, as it seemed to me initially that if what is being sought to be ascertained is the “windfall” actually achieved by a claimant from receiving a gross amount of restitution interest at the end of a period instead of accruing that amount over the course of that period and being taxed on it on an accruals basis, with the consequent effect on compounding, the natural starting point (and the value therefore of X) would be the actual amount of the restitution interest. However, I have concluded that that holds good only to the extent that the counterfactual calculates the interest in a manner consistent with the court award, but on a net as opposed to gross basis. If the counterfactual interest is calculated in a different way, it necessarily invalidates any comparison with the interest award. The counterfactual in Mr Cohn’s second revised model no longer reflects simply the tax effect on the interest award on an accruals basis, and the effect of compounding those amounts; it affects the underlying calculation of the compound interest itself, producing a different result in principle in that respect than the court’s calculation.

[100]To adjust for that difference, I respectfully agree with Professor Devereux that the same adjustment must be made in respect of X, in order that there can be a valid comparison for the purpose of establishing the implicit tax rate. It is not possible, without detailed calculations, to verify the figure of 48.6% suggested by Professor Devereux, but I agree with him that the implicit tax rate of 2.81% arrived at by Mr Cohn in his (adjusted) second revised model cannot be relied upon. That rate has not been calculated on a reasonable basis and should be disregarded.

[101]Mr Cohn’s fifth witness statement, which compared the after tax calculation of the hypothetical claim value put forward by Professor Devereux as the proper comparison with Mr Cohn’s second revised model outcome with the after tax amount of the judgment debt (ignoring Part 8C) both in 2015 and in 2018 when that debt is expected to be recognised in BAT’s profit and loss account, and which was argued to show that there was no windfall, does not alter my conclusion with respect to the 2.81% implicit tax rate. Nor does it show that there was no windfall with respect to restitution interest, as, first, it includes both principal and interest, and secondly it takes as its base case a hypothetical claim value. The figure for the after-tax hypothetical claim value does not represent, as Mr Cohn’s witness statement asserted, the amount required to compensate BAT for having made unlawful payments of tax. That amount has been conclusively determined by the court. The hypothetical value cannot therefore be compared with the after-tax amount of the judgment debt to seek to illustrate the absence of a windfall.

[102]Although I consider those are the proper conclusions with respect to the implicit tax rate in the second revised model, I do not accept Mr Aaronson’s submission that Professor Devereux’s preferred methodology of assuming that no unlawful tax was paid, with the consequence that there was no resulting unlawful ACT to be set off against corporation tax, could not be applied to that extent to the award actually made by the court. To be fair to Mr Cohn, his second revised model was an attempt, in the absence of a full description by Professor Devereux of his preferred methodology, to construct a methodology to give effect to those principles. It is that methodology, however, and not the principle of eliminating the unlawful ACT, that renders the calculation of the implicit rate unreasonable.

[103]Nor do I accept that Professor Devereux was confused or mistaken as to the nature of the award to BAT. I do not consider that he assumed that compensation had been awarded for the entire sum of the unlawful ACT from the moment of each payment of unlawful ACT down to the date when BAT was actually paid the restitution interest. I agree with the submissions for HMRC, that any computations which include unlawful ACT involve double-counting, and are flawed to that extent. The award to BAT gave compensation only for the time value of the unlawful ACT between the dates of each payment of unlawful ACT until the date when it was treated as being set off against mainstream corporation tax. From the perspective of an award of compensation that was clearly the right approach, and the reason it was right is that it was necessary to reflect that, by offsetting the unlawful ACT against mainstream corporation tax, BAT had received value (from the Government in terms of not having to pay the mainstream corporation tax) at that time for the unlawful ACT, and it was not appropriate therefore for that element of the value to be compensated again.

[104]It is the case, therefore, that all the unlawful ACT has been compensated for either by being offset against mainstream corporation tax or repaid, and likewise the time value of all of the unlawful ACT up to the time of offset or repayment has been compensated for by restitution interest. In those circumstances, any application of unlawful ACT in any calculation of the notional tax on assumed accruing interest would double count that unlawful ACT.

[105]Although the implicit tax rate produced by the second revised model cannot be relied upon, what the second revised model does show, on the other hand, and this was accepted by Mr Cohn, is that in the majority of the years from 1973 to 1998 (when the ACT regime was in force), once unlawful ACT had been stripped out, BAT would have paid tax. The spreadsheet setting out, in respect of this iteration of the model, the summary of ACT utilisation (which also took into account the tax effect of additional capital allowances claimed) shows that in 18 of the total of 26 years either all of the restitution interest (treated as marginal income) would have been taxed (15 of the years) or part of it would have been taxed (3 of the years – 1983, 1985 and 1987). In only 8 of those years (1973, 1979-80 and 1989-93) would there have been no tax because of the application of reliefs. On the other hand, the spreadsheet dealing with the offset of losses against restitution interest for the years 1999 to 2015, a time when the group was broadly loss-making, showed residual tax liabilities on part of that restitution interest only in the years 1999 and 2001.

[106]It is convenient at this stage to refer to the evidence of Mr Neil Duncan, another HMRC officer in the KAI, with regard to his investigation of how typical it is for groups of companies within the charge to UK corporation tax to be consistently non-taxpaying or low-taxpaying over multiple consecutive years, the theory being that such groups would be more likely that other groups to have reliefs available to offset additional profits for those years. The years for which data was extracted were those from 2000-01 to 2012-13.

[107]Subject to certain limitations, including the lack of accurate group structure data for years before 2012-13, and that the analysis only considered groups which filed a return in all the relevant years, Mr Duncan’s conclusion was that the analysis suggested that in the relevant years it was relatively rare for a group to be non-taxpaying or low tax paying in multiple successive years, with only 4% of groups filing a return every year from 2000-01 to 2012-13 being non-taxpaying in every year of this period, and the results remaining relatively stable when also including groups paying small amounts of tax.

[108]As Mr Aaronson pointed out in cross-examination of Mr Duncan, his analysis was necessary only a partial picture, and crucially excluded all the years for which ACT had been applicable prior to its abolition in 1999. It was, on the other hand, supported as a matter of principle by the expert witness instructed by BAT, Mr Nicholas Forrest, who regarded Mr Duncan’s empirical findings as broadly consistent with his own. The longer the time frame of the analysis, Mr Forrest found, the lower will be the proportion of groups which are continually non-taxpaying in every year.

[109]Equally, however, Mr Forrest concluded that it would be uncommon for groups to be consistent taxpayers for many consecutive years. Professor Devereux considered the impact of a period of taxable losses on the calculation of an implicit tax rate. He concluded that relatively short periods of tax losses would have an insignificant effect on the implicit tax rate. A four-year period of tax losses in the period 2000-01 to 2003-14, for example, would result in a decrease in the implicit tax rate of just 1%. Overall, it was suggested that even relatively long periods of taxable loss have only a relatively small impact on the implicit tax rate. Professor Devereux found that there were two offsetting effects of a period of tax loss. First, tax is deferred to a future period, which reduces the implicit tax rate. But secondly in the interim the company would hypothetically earn higher interest, due to compounding on a gross basis; that higher interest would in turn be subject to tax, a factor that tends to increase the implicit tax rate. That, I accept, is a general principle only. As Mr Forrest said in his oral evidence, there will be cases where a company has such substantial losses that the prospect of tax becoming payable recedes indefinitely. But those cases are very much the exception and do not affect the general principle identified by Professor Devereux.

[110]I should also add that the evidence of Mr Lane as to the use of losses in certain circumstances representing a timing difference only, to which I have referred earlier, was accepted by him not to refer to BAT specifically, but to be a general observation. I do not accept therefore Mr Lane’s suggestion in his first witness statement that by failing to take account of the possible future use of reliefs that are notionally used to offset restitution interest against other profits, Mr Cohn’s computations failed to take account of all relevant issues. Third party evidence

[111]In support of its appeal, BAT also provided evidence from other companies within the scope (or originally within the scope) of Part 8C CTA. I had witness statements from: Mr Delme Davies, the senior tax adviser of Tata Steel UK Limited Mr Paul Henry, formerly the tax and treasury manager at the Perkins Group Mr Massimo Di Cesare, the group head of tax at Richement International SA, of the Richemont Group Mr Michael Smalley, a director of Imperial Chemical Industries Limited (“ICI”) and the head of tax for Akzo Nobel NV Mr Mark Downey of The Prudential Assurance Company Limited Mr Simon John Rivara, a UK tax manager of the Schneider Electric Group, of which a group company, Invensys International Holdings Limited, was a claimant in the FII GLO Ms Anna Pack, a vice President of Tax for the InterContinental Hotels Group, of which (amongst others) a group company, Six Continents Overseas Holdings Ltd, was a member of the CFC and Dividend Group Litigation Mr Jonathan Greenwood, Group Tax Director of Inchcape Plc Ms Sheelpa Shah, the UK tax manager for the Solvay UK Group Ms Stacey-Ann Jackson, the UK Finance Manager of the Evonik UK group of companies Mr Michael Shaw, formerly the head of taxation at the GKN group

[112]The principal case of HMRC with respect to this third party witness evidence was that it was not relevant to BAT’s own appeal. With one exception, the evidence was challenged only as to relevance, and not as a matter of fact. The single exception was the witness evidence of Mr Smalley of ICI, which was subjected to cross-examination.

[113]Mr Smalley’s evidence was that from 1982 to 1990 he had no record of reliefs and allowances which might have been available to shelter further income (as in the restitution interest). He had concluded therefore that in that period any accruing restitution interest would have incurred corporation tax on the corporation tax liability dates for those accounting periods or in slightly later accounting periods.

[114]For later periods, Mr Smalley said that from 1991 there was an amount of ACT carried forward, and from 1992 ACT payments became permanently surplus. Furthermore, from 1995, the ICI group has had substantial tax losses arising from pension fund top-up payments. A combination of such losses, surplus ACT and capital allowances was considered to be more than sufficient to meet any additional tax liabilities on the accruing restitution interest. After adjustment to take account of the fact that for accounting periods beginning on or after 3 June 1986 double tax relief available to the company would not affect the ACT utilisation capacity on notional receipts of interest (because of effective “ring-fencing” of the interest income), so that tax would have been paid on the difference between the corporation tax rate and the gross ACT rate, the difference between the judgment debt before tax and the debt re-computed to assume that interest was taxed annually was 16%, thus implying a required notional tax rate of that percentage.

[115]In cross-examination, Mr Smalley accepted that the accruing interest that had formed part of the calculations exhibited to his witness statement formed only part of a larger claim. When re-examined, however, he explained that the larger claim had now been valued and that, on the assumption that the major part of the restitution interest would arise in respect of the later years in the period of the claim, a combination of surplus ACT and the very substantial carried forward tax losses of the group would be available to be set off against any additional interest income.

[116]Mr Smalley’s statement was dated 2 September 2016, and the calculations attached to it as exhibits dated from that time. Mr Smalley confirmed that, as far as he was aware, those computations had been made on the same basis as for BAT. Subject only to the adjustment made to reflect the “ring-fencing” of interest income from 1986, the computations must have reflected the methodology adopted up to that date by Mr Cohn. At that time, the methodology was that contained in Mr Cohn’s basic model, revised only to reflect the taking into account of quarterly, as opposed to annual, tax payments. It had not yet reached the iteration of seeking to exclude unlawful ACT. There is nothing in the calculations exhibited to Mr Smalley’s statement to suggest that unlawful ACT had been stripped out.

[117]I accept that, in the particular case of ICI, with its limited reliance on surplus ACT and greater reliance on tax losses generally, particularly in the later years, the use of Mr Cohn’s basic model may not have had any significant impact on the result. But I can make no finding as to what the position would have been if the model had been adjusted, as I consider it would have been necessary to do, to reflect the fact that unlawful ACT could not be treated as offsettable against mainstream corporation tax on the assumed accruing restitution interest.

[118]For this reason, I do not give any significant weight to the evidence of Mr Smalley to the extent that it goes further than the facts of the surplus ACT and tax loss position of ICI itself. In the case of ICI, the claim period was from 1982-89. Prior to 1991, tax would have been paid on the assumed interest income. Between 1991 and 1994, ACT would have been available to be utilised against additional tax, but I can make no finding in this respect as to the result if unlawful Act were to be excluded. After 1994, I accept that the losses of ICI were more than sufficient to meet any additional liability. But I can make no finding with respect to the calculations for ICI which seek to establish a particular notional tax rate.

[119]The evidence of third parties indicated that, like ICI, those companies were either not reliant, or not wholly reliant, on surplus ACT to offset tax on restitution interest that was assumed to have accrued. Thus, Tata Steel had excess losses over the whole of the relevant period, that is losses that had not been claimed by way of group relief or against Tata Steel’s own liabilities. Ford had disclaimed capital allowances which, if claimed, would offset the additional taxable profit. Perkins Group could utilise losses as well as surplus ACT. Richemont had losses in a number of years and disclaimed capital allowances. Inchcape had, from 1997 to 2014, substantial losses derived from non-trading loan deficits, excess expenses of management and disclaimed capital allowances. Solvay UK Group had group losses and disclaimed capital allowances as well as some surplus ACT. Evonik would have been able to employ available losses or surplus ACT. GKN had trading losses and surplus non-trading deficits as well as some surplus ACT. Invensys (Schneider Electric group) had surplus ACT up to 1999 and losses from 2000. IHG was tax paying in the UK from 1994 to 2003, and from 2004 onwards had carried forward losses available for offset against accruing restitution interest.

[120]As Mr Forrest found when considering the third party witness statements, these showed a number of companies which were in a non-tax paying position for a number of years. Many were taxable for individual years, but few were non-taxpaying for the whole period of 1973 to 2015. That supports the general premise, as explained by Mr Rounding, that over an extended period if there are excess reliefs for a company in a particular year, as a general proposition those reliefs are likely to be available for set off in a subsequent year.

[121]A question arose as to whether the calculations with respect to the third parties’ positions included any offset of unlawful ACT. As a matter of principle, I can accept BAT’s submissions in this respect that to the extent that there was no offset of ACT at all (which appears to be the case with Ford and IHG), there could be no question of the use of unlawful ACT. In other cases, disregarding Prudential which occupies a special position in respect of its life assurance business (and which for that reason is now outside the scope of Part 8C), although in principle it can be accepted that a claim that was restricted to tax on Schedule D, Case V income and one that was restricted to ACT on FIDs which had been repaid in the ordinary course would not give rise to offsetable ACT, I agree with HMRC that, as I have summarised above, a number of the third party computations use ACT to set against restitution interest and it is not possible to conclude from the evidence that only lawful ACT has been included.

[122]As this third party evidence was not challenged, I accept it for the purpose of this appeal as representing the factual tax position of those companies and groups of companies over that period. In so far, however, as the computations of notional tax rates rely on the methodology in Mr Cohn’s basic model, include an offset for ACT but do not split out unlawful ACT, I do not accept the results of those calculations. I nonetheless for completeness set out below, but without making any factual finding in those respects, the percentage difference as calculated for each relevant claimant (other than BAT) between the judgment debts (before tax) and the debt re-computed to assume that interest was taxed annually, as revised to take account of the ring-fencing of the interest: Claimant Percentage difference Evonik 13% Ford 0% GKN 15% ICI 16% IHG 0% Inchcape 6% Invensys 12% Perkins Rhodia (Solvay) Richemont Tata Steel 8% 10% 18% 0% Expert evidence

[123]I have referred above to the fact that expert evidence was provided by Professor Devereux on the instructions of HMRC. Professor Devereux is a highly-qualified academic whose expertise lies in the field of the economics of business taxation. Most recently, since 2006, he has been Professor of Business Taxation at the University of Oxford and Director of the Oxford University Centre for Business Taxation.

[124]I have also referred to the expert evidence provided by Mr Forrest on the instructions of BAT. Mr Forrest is a professional economist who has considerable professional experience in his specialised field of economic and financial analysis in the context of regulatory, valuation, competition and dispute purposes. He has since 2009 been a Director, Economics and Policy, Consulting with PricewaterhouseCoopers UK.

[125]Each expert produced three individual reports and together they produced a helpful Joint Statement of Issues of Agreement and Disagreement (“the Joint Statement”). As that will form the cornerstone of my discussion of the expert evidence, and to avoid simply duplicating it in my own words, I have appended it in full as Appendix B to this decision. In my discussion, however, there is some inevitable repetition. Each of Mr Forrest and Professor Devereux gave oral evidence, answering questions from both parties and from me. I am grateful to them both for the assistance they gave to the Tribunal.

[126]In discussing the expert evidence, I shall adopt the approach of the experts in their joint statement, namely to consider each of three questions as directed by the tribunal. Section A – an assessment of HMRC’s methodology in establishing the 45% rate for restitution interest

[127]The experts divided this question into three. The first is a consideration of the appropriate counterfactual, the second is appropriateness of the KAI methodology and the third is the calculations and data used by KAI.

[128]As to the third of those sub-questions, the experts have no material disagreements. They agree that Ms Richmond’s mechanical calculations appear to be accurate, and that the data has been obtained from appropriate sources.

[129]Whilst the experts also have a measure of agreement as to the principle of the use of a “counterfactual” in determining the appropriate tax rate, it is on entering the world of the counterfactual that they begin to disagree. The starting point for the experts, however, is that there must be a counterfactual, and that it is one in which the unlawfully paid tax was not paid to the tax authority.

[130]The HMRC approach, as I have described, was to assume that the company invested the amount of the unlawful tax at a risk-free rate. That approach is broadly supported by Professor Devereux as a standard approach used in comparing values of assets and income at different points in time; as Professor Devereux notes, the issue is not to re-evaluate the amount of the award but to determine how the restitution interest should be taxed.

[131]Mr Forrest takes a different approach which he, in turn, considers best reflects standard economic theory. He postulates a different counterfactual, namely to assume that the unlawful tax would have been employed by the company in scaling-up its activities, on the basis that the unlawful tax can be regarded as having introduced a distortion into the company’s profit maximisation decisions.

[132]It is difficult to understand the economic basis for a counterfactual such as that proposed by Mr Forrest without appreciating the effect on the company’s profits which it is said to produce. There is an overlap here with the separate question addressed by the experts on the respective merits of the use of a nominal rate or an effective rate as the basis for the setting of the rate of tax for restitution interest. It is necessary, in any comparative exercise, for a defined amount of income from the use or exploitation of the unlawful tax assumed not to have been paid to be capable of being ascertained. That is not apparent from a counterfactual which merely anticipates that the company’s business would be scaled up. It is necessary, as Mr Forrest does when examining the question of nominal and effective rates, to make the further assumption, which Mr Forrest considers to be a reasonable one, that the business would be scaled-up in such a way, including as to profitability and the generation of reliefs and allowances, that the tax on the marginal profit would be the same as the effective rate of tax on the company’s profits in the absence of the counterfactual.

[133]It is clear that the parallel world of the counterfactual can conceptually take almost any form. But in approaching what is appropriate in any given case it is in my view necessary for the counterfactual to reflect by proxy the reality of what it is that requires a counterfactual for comparison. In this case it is a receipt of interest. Logic would suggest that any counterfactual must itself postulate an interest return. If, as the experts have agreed is appropriate, the starting-point of the counterfactual is that the unlawfully paid tax has not been paid, but has been retained by the company, it is consistent with that logic to regard it as having been invested at the rate of return which was used in the real world to determine the amount of the restitution interest.

[134]In constructing such a counterfactual it is not necessary, as the premise for Mr Forrest’s view was, that the counterfactual should reflect the economic realities of business. It is necessary for it to be reflective of the interest receipt in the real world; but it is a hypothetical world in which the economic reality of how cash might be used is suspended. In fact, as Mr Forrest acknowledged, the evidence as to economic reality in the case of BAT did not support his counterfactual or his view of the economic realities of business. Mr Wadey, the BAT Group Treasurer, gave evidence that, if the amount of unlawful tax had not been paid BAT would, as a matter of common sense, have reduced its borrowings, reducing first the most expensive borrowing. Referring to BAT’s access to the capital markets, Mr Wadey specifically disavowed any notion that the payment of unlawful tax would have hampered the group’s commercial operations and ventures; it would merely have had the result that the group would have borrowed more than it otherwise would have.

[135]In just the same way as I discount the notion that any counterfactual could properly postulate that the additional available cash in the group would have been used to pay increased dividends (as to which see my discussion above of Mr Lane’s evidence), so too do I consider that Mr Forrest’s scaling-up counterfactual must be flawed. I agree with Professor Devereux’s reasons, set out at paragraph 19 of the Joint Statement, why Mr Forrest’s approach is unreasonable. There is in my view no justification for seeking to construct a counterfactual that goes beyond producing a return analogous to the real return at issue, namely in this case an interest return.

[136]I do not accept the view of Mr Forrest that a counterfactual which postulates a risk-free interest return involves any assumption other than that such a return is generated. In particular, I can see no warrant for it requiring to be assumed that the risk-free investment has to be kept secret from the business in order to prevent the business from making different decisions as to the deployment of incremental risk-free assets. Were that to be a necessary – and by implication impermissible – assumption, it would it seems to me follow that economic theory would never permit the use in a business context of an assumed risk-free interest return on available cash as a counterfactual. I do not regard that as a tenable proposition. Section A2 – the appropriateness of KAI’s methodology

[137]The experts were in large measure agreed on a number of issues with respect to HMRC’s methodology. Put briefly, they agreed that a marginal tax rate was appropriate, that calculating claimant-specific tax rates would, conceptually, be the most accurate means of putting individual claimants in the position of a company receiving interest on an accruals basis in each relevant accounting period and paid tax accordingly, and that the feasibility of calculating such a claimant-specific rate would depend, for example, on the availability of information. They also agreed that if, in contrast, an averaging approach was to be taken, a weighted average was appropriate, and that such a weighted average would differ from a claimant-specific rate for most claimants.

[138]The experts disagreed, however, on the feasibility of calculating claimant-specific rates and the appropriate weighting if average rates were to be used.

[139]The question of feasibility was not really one that was susceptible to economic analysis. Mr Forrest candidly considered that this was not a question on which he could comment, except to the extent that he had demonstrated the possibility for claimant-specific rates to be calculated. Professor Devereux did not dispute the possibility of such rates being determined if all the required information were available, but he regarded it as unlikely that it would be for every claimant. It would not, in Professor Devereux’s view, be reasonable to use a weighted average of claimant-specific rates based only on claimants for whom it was possible to calculate such a rate.

[140]The question in this regard is not whether it would be feasible for there to be a claimant-specific rate, but whether, if it were, it would be unreasonable for the legislation to have been developed so as to provide for a single global rate. That is a question that falls to be considered, as a matter of law, after analysing the applicable law, which I will turn to later.

[141]There was little disagreement on the second aspect, namely the appropriate weighting. Mr Forrest expressed the view that the approach of Ms Richmond, in the context of her methodology, in weighting the implicit tax rate by the amount of unlawfully paid tax, was appropriate. The difference between the experts was in relation to an alternative weighting approach proposed by Professor Devereux, which was that the weight for each company should be based on the amount of pre-tax restitution interest, on the ground that this is the tax base for each company. Both Ms Richmond’s weighting and that of Professor Devereux give more weight to the larger claims; Professor Devereux’s gives more weight to older claims. There seems to be little in principle to separate them. The most that can be said is that there is nothing wrong in principle with the weighting in the KAI model. It may well be a preferable method, as there does not seem to me to be any reason in principle why older claims should be given greater weight, and any weighting approach which did so might well require adjustment to remove such an effect. Section B – assessment of the effective rate of tax (i.e. the rate of tax actually paid after reliefs and allowances) at which UK corporate taxpayers paid corporation tax in each of the fiscal years in the period 1973-2016

[142]Unsurprisingly, the experts agreed on the general propositions that, first, there is often a material difference between the effective rate of corporation tax and the nominal rate, and secondly that the difference between the nominal and effective rates has narrowed over time due to the reduction in the rate of capital allowances and other factors.

[143]The experts also agreed with the methodologies employed by Mr Forrest to calculate effective tax rates and that the data he used was reasonable.

[144]The context in which such calculations fall to be made assumes that a claimant-specific rate is, for whatever reason, unavailable or inappropriate. Mr Forrest’s analysis focuses not on the effective tax rate of one claimant, but on a broader population of companies. Mr Forrest reviewed four methodologies based on the following:(1) National accounts. Using this methodology, a measure of corporate profitability from the national accounts is used to estimate an economy-wide effective tax rate.

(2) HMRC statistics (for which Mr Forrest expresses a marginal preference). A measure of corporate profitability is taken from the corporation tax data tables provided annually by HMRC.

(3) Accounting data. Effective tax rates are calculated based on a sample of annual company accounts. For a representative sample of companies, profitability and taxation figures are extracted from company income and cash flow statements

(4) Tax legislation. This category calculates marginal and average effective rates for a hypothetical investment in certain types of asset, for example an investment in industrial buildings. The effective tax rates are calculated by applying parameters from tax legislation (such as the nominal rate and available capital allowances), and by making certain assumptions about economic parameters. Section C – an assessment of the advantages and disadvantages in setting the tax rate by reference to either the nominal (statutory) rate of tax or the effective rate of tax and how each approach would reflect the economic realities of business

[145]The experts agree that the use of a nominal rate has the advantage of simplicity. They agree that a disadvantage of the use of a nominal rate is that it does not take into account that surplus ACT, losses, group relief and disclaimed capital allowances (and other reliefs) could reduce the effective rate to below the nominal rate in a given year, although this effect would be likely to vary considerably across claimants. A disadvantage of the use of the effective rate is that it requires a judgment as to the most appropriate data sources and empirical methodologies.

[146]The first, and key, disagreement is on the question whether the effective tax rate measures employed by Mr Forrest (see Section B above) are ever appropriate to use in calculating an appropriate tax rate on restitution interest.

[147]Having considered the respective experts’ opinions, I have reached the clear conclusion that, in a circumstance in which a claimant-specific rate is ruled out, no other measure of effective rate could reasonably be used. I regard that as the case whether one is considering the effective rate of tax applicable to an individual claimant company (absent any hypothesis as to accruing interest or additional resources by not having to pay unlawful ACT), or a more general effective rate found by employing one or more of the methodologies advocated by Mr Forrest.

[148]As to the individual effective rate, that is the rate arrived at by taking a measure of profit for a period, such as accounting profit, and comparing that with the actual tax paid for that period. That will provide an effective tax rate for that period, and it is possible to arrive at an average effective rate over a number of periods. But it will only rarely reflect the marginal rate which should be applied in the case of marginal income such as the interest income assumed to accrue for each period in seeking to determine the appropriate rate of tax on restitution interest. That may be the case where, for example, a company or group is consistently loss-making and has considerable available reliefs over the whole period to shelter any marginal income. But in the more normal case where over a period a company or group is tax-paying to some extent, it is axiomatic (subject to one relevant exception) that if any tax is paid in a period a part at least of the profits must have been subjected to tax at the marginal rate which will usually be the nominal rate. If that is the case, the point made by Mr Aaronson that the high rates of tax in the relevant period were accompanied by more generous reliefs will make no difference. The exception is in the case of a company with surplus ACT, utilisation of which was restricted. The increased profit would release a further part of that surplus ACT for offset against the marginal income [3] . But that would not normally result in the same effective rate as that applicable to the company in the absence of the marginal income and such an effective rate could not therefore properly be applied to that marginal income.

[149]The only circumstance in which in principle the effective rate of tax on marginal income would be the same as the effective rate on the actual profits of the company or group would be if Mr Forrest’s broader counterfactual – which hypothesised a scaling-up of the business – were to be taken as the appropriate counterfactual. For the reasons I have given, I do not consider that such a counterfactual would be appropriate. It would follow that the use of the effective tax rate of the individual company or group would be equally inappropriate.

[150]Turning to the more general calculations of effective tax rate using the four methodologies identified by Mr Forrest, it is the case that Professor Devereux agreed that taxation affects (such as surplus ACT, losses, group relief and disclaimed capital allowances) should be considered. But Professor Devereux nonetheless considers that effective tax rates ascertained by those methodologies are not appropriate for identifying the tax due on marginal interest income. He contrasts those effective tax rates with the need to determine a marginal rate on hypothetical interest income, and concludes that the average rates which the methodology would produce would take account a range of factors that are irrelevant to interest income.

[151]A similar concern underlies Professor Devereux’s rejection of the use of such effective tax rates as a means of taking into account the effect of tax reliefs. The methodologies are not, in his view, reasonable substitutes for applying the tax in the counterfactual taking into account the details of each claimant’s tax position.

[152]I agree with the views expressed in this regard by Professor Devereux. Like him, I do not consider that, absent an analysis that was capable of identifying, by reference to the position of the individual company or group, what the actual tax would be on an accruing amount of interest in every period, having regard to the application of available reliefs, the ascertainment of average effective tax rates for a sample of other companies and groups can reasonably be regarded as an appropriate proxy for the actual individual position. It is interest income that is hypothesised in the counterfactual, and that income (which, subject to the possible increased use of surplus ACT, generates no reliefs of its own) is marginal income taxed at the marginal rates applicable to the company or group. For any period in which the company or group is taxpaying, that marginal rate will generally be the nominal rate of corporation tax.

[153]Having reached that conclusion, it is not necessary to consider in detail the further areas of disagreement identified by the parties on this issue. They are effectively only sub-sets of the main question as to the suitability of the use of Mr Forrest’s effective tax rate, and there is considerable overlap with the points I have already discussed. For completeness, however, I should say that I agree with the reasons given by Professor Devereux (at paragraphs 41 to 49) for concluding, contrary to the view taken by Mr Forrest, that in the context of a counterfactual of an interest return which is marginal income the effective tax rate is not more reflective of the economic realities of business. Discussion

[154]The question at issue in this appeal is the lawfulness of Part 8C CTA. It is not a question of whether one counterfactual is preferable to another, or even if there is a “correct” counterfactual. My own findings illustrate the illusory nature of the quest for the perfect counterfactual; once the world of the counterfactual is entered it is all too easy, and beguiling, to extend the hypothesis beyond that which is required.

[155]As Mr Aaronson submitted in closing, the counterfactual needs to go as far as, but no further than, to assume that the restitution interest granted in the court’s award accrues year by year and is subject to tax in each year on an accruals basis, and to assume that the tax chargeable during the accrual period was actually paid so as to reduce pro tanto the amount of interest available to be carried forward and so compounded over the next year, and so on for each year.

[156]As Mr Aaronson noted, that was essentially the approach on Mr Cohn’s basic model and the KAI model. The difference between them was that Mr Cohn’s model, and this was consistently the case with the further iterations of that model, took into account reliefs and allowances, including surplus ACT, and HMRC’s model simply applied the relevant nominal rates and did not take reliefs into account. The question for this tribunal resolves itself into whether in that context the legislation applying the rate of 45% arrived at through HMRC’s methodology, and “ring-fencing” the application of that rate, is lawful.

[157]None of Mr Cohn’s calculations has shown conclusively what the precise implicit tax rate in BAT’s particular case ought properly to be, if reliefs were appropriately taken into account. Each of his models, painstakingly prepared and impressively presented as they were, suffered from the flaws I have identified above. Similarly, for the reasons I have given, it is not possible to say with certainty what the implicit tax rate ought to be in the case of most of the third party claimants who provided evidence. It is not necessary, however, to be able to reach precise conclusions in those respects in order to appreciate that, in principle, the rate ascertained by applying a nominal rate alone to the restitution interest treated as accruing over the period will be likely to be higher than a rate computed by applying that same rate to the accrued interest but subject to available reliefs, even if, as I consider to be necessary, those reliefs exclude unlawful ACT. The only circumstance where that will not be the case is where it is assumed (as the experts agree HMRC’s methodology does) that the claimant is always in a fully taxpaying position.

[158]In this respect I accept, as a general proposition, that the majority of companies with accounting profits in the period 1973 to 2005 paid corporation tax at an effective rate lower than the applicable statutory rate. That was the conclusion of an expert report prepared by Mr John Whiting in March 2008 on the instructions of the solicitors for the test claimants in the FII GLO, which was referenced by Mr Forrest in his first report, having regard to a survey of 4,055 corporation tax computations prepared by PricewaterhouseCoopers in that period. However, except where the effective rate is zero, it can also be concluded that a part of the profits of the companies would have been taxed at the nominal rate (thereby, after application of reliefs, giving a positive effective rate of tax), and that accordingly any further marginal taxable income, in the form of assumed accruing interest, can likewise be assumed to be taxable at the nominal rate. An effective rate could not therefore be an appropriate rate on such marginal income.

[159]I also accept the agreed position of the experts that claimant-specific rates would, conceptually, be the most accurate approach. That approach would involve the assumed application of applicable reliefs in each individual case. I have concluded, however, that this could not properly be achieved by the proxy method of seeking to apply either effective tax rates applicable to the individual company or a more general effective rate determined in the manner suggested by Mr Forrest. It is in that context that I turn to consider the question whether Part 8C is lawful as a matter of EU law and the law on human rights. The principle of effectiveness

[160]It was common ground that the general principles of EU law, including the principles of effective legal protection, protection of legitimate expectations and proportionality, emanate from the rule of law on which the EU is founded (see Article 2 of the Treaty on the Functioning of the European Union; Unión de Pequeños Agricultores v EU Council (C-50/00 P)

[2002] ECR I-6677

[2003] QB 893

, ECJ). Member states must provide remedies which are sufficient to ensure “effective legal protection”, in other words an “effective legal remedy” enabling injured parties “to obtain reimbursement of the tax unlawfully levied upon them and the amounts paid to that member state or withheld by it directly against that tax” ( FII(ECJ)1 ).

[161]Equally, the separation of powers, as between the executive and the judiciary, characterises the rule of law. As the ECJ explained in DEB Deutsche Energiehandels - und Beratungsgesellschaft GmbH v Bundesrepublik Deutschland (C-279/09)

[2010] ECR I-13849

, at [58]:

“EU law does not preclude a Member State from simultaneously exercising legislative, administrative and judicial functions, provided that those functions are exercised in compliance with the principle of separation of powers which characterises the operation of the rule of law.”

[162]It is “for the national courts to determine, in the light of the facts of each case” what constitutes an effective remedy (see, for example, Société Comateb v Directeur general des douanes et droits indirects and related references (Joined cases C-192/95 to C-218/95)

[1997] STC 1006

). BAT’s case is that Part 8C CTA interferes with the role of the courts in determining what would constitute an effective remedy, and that, by enacting Part 8C, Parliament has usurped a quintessentially judicial function, and has deprived BAT of an effective legal remedy by way of indemnity for the loss it suffered as a result of paying unlawful tax.

[163]The question of an effective legal remedy is therefore central to this case. That remedy is a right to an adequate indemnity for the loss occasioned through the undue payment of tax. The principles are summarised by the ECJ in Littlewoods Retail Ltd and others v Revenue and Customs Commissioners (

Case C-591/10

)

[2012] STC 1714

, at [24] – [30]:

“24. It is settled case law that the right to a refund of charges levied in a member state in breach of rules of EU law is the consequence and complement of the rights conferred on individuals by provisions of EU law as interpreted by the court (see, inter alia, Amministrazione delle Finanze dello Stato v SpA San Giorgio (Case 199/82) [1983] ECR 3595 , para 12, and Metallgesellschaft Ltd v IRC, Hoechst AG v IRC (Joined cases C-397/98 and C-410/98) [2001] STC 452 , [2001] ECR I-1727 , para 84). The member state is therefore in principle required to repay charges levied in breach of Community law ( Société Comateb v Directeur General des Douanes et Droits Indirects and related references (Joined cases C-192/95 to C-218/95) [1997] STC 1006 , [1997] ECR I-165 , para 20; Metallgesellschaft (para 84); Weber's Wine World Handels-GmbH v Abgabenberufungskommission Wien (Case C-147/01) [2005] All ER (EC) 224, [2003] ECR I-11365 , para 93 ; Test Claimants in the FII Group Litigation v IRC (Case C-446/04) [2007] STC 326 , [2006] ECR I-11753 , para 202). 25. The court has also held that, where a member state has levied charges in breach of the rules of Community law, individuals are entitled to reimbursement not only of the tax unduly levied but also of the amounts paid to that state or retained by it which relate directly to that tax. That also includes losses constituted by the unavailability of sums of money as a result of a tax being levied prematurely ( Metallgesellschaft (paras 87 to 89), and Test Claimants in the FII Group Litigation (para 205)). 26. It follows from that case law that the principle of the obligation of member states to repay with interest amounts of tax levied in breach of EU law follows from that law. 27. In the absence of EU legislation, it is for the internal legal order of each member state to lay down the conditions in which such interest must be paid, particularly the rate of that interest and its method of calculation (simple or 'compound' interest). Those conditions must comply with the principles of equivalence and effectiveness; that is to say that they must not be less favourable than those concerning similar claims based on provisions of national law or arranged in such a way as to make the exercise of rights conferred by the EU legal order practically impossible (see, to that effect, San Giorgio (para 12); Weber's Wine World (para 103); and MyTravel plc v Customs and Excise Comrs (Case C-291/03) [2005] STC 1617 , [2005] ECR I-8477 , para 17). 28. Thus, according to consistent case law, the principle of effectiveness prohibits a member state from rendering the exercise of rights conferred by the EU legal order impossible in practice or excessively difficult ( R (on the application of Wells) v Secretary of State for Transport, Local Government and the Regions (Case C-201/02) [2005] All ER (EC) 323 , [2004] ECR I-723 , para 67, and i-21 Germany GmbH v Bundesrepublik Deutschland (Joined Cases C-392/04 and C-422/04) [2006] ECR I-8559 , para 57). 29. In this case, that principle requires that the national rules referring in particular to the calculation of interest which may be due should not lead to depriving the taxpayer of an adequate indemnity for the loss occasioned through the undue payment of VAT. 30. It is for the referring court to determine whether that is so in the case at issue in the main proceedings, having regard to all the circumstances of the case…”

[164]BAT submits that Part 8C substantially deprives it of the indemnity for its loss as determined by the national courts, and that consequently it infringes the principle of effectiveness and BAT’s right to an adequate indemnity. It points to the fact that all organs of the State, including the legislature, the executive and the judiciary, are bound to comply with EU law and its principles. The principle that member states are obliged to make good damage caused to individuals by breaches of EU law attributable to the State applies where it is the legislature that is responsible for the breach. As the ECJ said in Brasserie du Pecheur SA v Federal Republic of Germany; R v Secretary of State for Transport, ex p Factortame Ltd and others (No 4) (Joined Cases C-46/93 and C-48/93)

[1996] ECR I-1029

, at [34] – [35]:

“34. … in international law a state whose liability for breach of an international commitment is in issue will be viewed as a single entity, irrespective of whether the breach which gave rise to the damage is attributable to the legislature, the judiciary or the executive. That must apply a fortiori in the Community legal order since all state authorities, including the legislature, are bound in performing their tasks to comply with the rules laid down by Community law directly governing the situation of individuals. 35. The fact that, according to national rules, the breach complained of is attributable to the legislature cannot affect the requirements inherent in the protection of the rights of individuals who rely on Community law and, in this instance, the right to obtain redress in the national courts for damage caused by that breach.”

[165]The fact that Part 8C is primary legislation does not therefore preclude the application of the principle of effectiveness or restrict its scope. It is first necessary to examine, in broad terms, the scope of that principle. Numerous examples may be cited, the common thread being that a member state is precluded from rendering the exercise of rights conferred by the EU legal order impossible in practice or excessively difficult. Thus, as BAT submitted, the principle has been held to preclude such matters as unlawful limits on compensation ( Marshall v Southampton and South West Hampshire Area Health Authority (No 2) (

Case C-271/91

)

[1993] ECR I-4367

; Metallgesellschaft Ltd and others v Inland Revenue Commissioners and another; Hoechst AG and another v Inland Revenue Commissioners and another (Joined cases C-397/98 and C-410/98)

[2001] ECR I-1727

), peremptory reduction of limitation periods (for example in FII(ECJ)3 itself) and rules on evidence which precluded review by the courts ( Johnston v Chief Constable of the Royal Ulster Constabulary (

Case C-222/84

) [1986] ECR 1651 ) as well as a number of procedural impediments.

[166]For HMRC, Miss Foster submitted that the principle of effectiveness does not concern substantive questions, such as whether and to what extent an award made by a national court is taxable in the hands of the recipient. She argued that the principle is concerned with national procedural rules including the form or cause of action by which the underlying EU law rights may be vindicated. In this connection, it is submitted, BAT’s rights have already been vindicated by its common law claims in the FII GLO and by the awards of what I have found to be gross compound interest, which is taxable. In these circumstances, it is said, EU law has no further role.

[167]As to that, I do not accept that the principle of effectiveness should be constrained by any limitation to procedural rules. It is true that it is in the context of procedural rules inhibiting the making of a claim for breach of EU law that the principle of effectiveness has typically been applied. Thus, for example, in Weber's Wine World Handels-GmbH and others v Abgabenberufungskommission Wien (

Case C-147/01

) [2005] All ER (EC) 224, the ECJ referred, at [86], to the principle that a national legislature may not, subsequent to a judgment of the ECJ from which it follows that certain legislation is incompatible with EU law, adopt a procedural rule which specifically reduces the possibility of bringing proceedings for recovery of taxes which were levied though not due under that legislation. With specific regard to the principle of effectiveness, the ECJ reiterated, at [111], that national rules which place on a taxable person the burden of proving that a charge was not passed on to third parties, or any presumption that the charge was passed on would not be consistent with EU law.

[168]But whilst the existence or creation of procedural hurdles to the claiming of an effective remedy for breach of EU law will be the most common context for the application of the principle of effectiveness, the essential focus of the principle is not on the procedure itself, but on the substantive question whether an impediment has been introduced to the exercise of EU law rights which has rendered it virtually impossible or excessively difficult to exercise those rights. That is the implication of the relevant part of the judgment of Lord Sumption in FII(SC) when, having referred to the fact that the ECJ itself would not determine the nature of the domestic remedy, he said, at [146]:

“This is, however, subject to the overriding requirement derived from the Treaty and referred to in the passage which I have quoted from Rewe I , that national legal systems should provide a minimum standard of protection for EU law rights. In the case law of the Court of Justice, the standard of protection required is embodied in two principles which are restated in almost every decision on the point. First, the substantive and procedural provisions of national law must be effective to protect EU law rights (the 'principle of effectiveness'). Their enforcement in national law must not be subject to onerous collateral conditions or disproportionate procedural requirements. They must not render 'virtually impossible or excessively difficult' the exercise of rights conferred by EU law. Secondly, the relevant provisions of national law must not discriminate between the rules and procedures applying to the enforcement of EU law rights, and those applying to the enforcement of comparable national law rights (the 'principle of equivalence'). There is a third principle which features less prominently in the case law on this subject but is of considerable importance because it informs the approach of the Court of Justice to the first two. This is the principle of legal certainty, which lies at the heart of the EU legal order and entails (among other things) that those subject to EU law should be able clearly to ascertain their rights and obligations. One aspect of that principle is that within limits EU law will protect within its own domain legitimate expectations adversely affected by a change in the law.”

[169]It is right, as Miss Foster submitted, that Lord Sumption’s reference to “substantive and procedural provisions” was followed by the conventional description of the principle. But it is impossible to think that he meant nothing in referring specifically to substantive provisions of national law as well as to procedural provisions. It is thus the case that, whilst the principle of effectiveness will prevent procedural requirements from making it “virtually impossible or excessively difficult” to exercise rights conferred by EU law, the effect of the principle must be one that goes to the substance of the right. That is not to confuse the principle of effectiveness with the substantive right itself. I accept, as Miss Foster submitted, that the principle comes into play only once the substance of the relevant EU right has been defined, and does not create any separate right. It is, in that respect, secondary and adjectival. But once the right has been defined, it is the substance of that right that must be protected.

[170]The case of Ilie Nicolae Nicula v Administraţia Finanţelor Publice a Municipiului Sibiu Adminisţia Fondului pentru Mediu (

Case C-331/13

)

[2015] 1 CMLR 977

concerned a motor vehicle pollution tax which had been held to be incompatible with Article 110 TFEU, as it was levied on imported second-hand vehicles, including those imported from other member states. Mr Nicula was held to be entitled to recover that pollution tax in respect of a vehicle originally registered in Germany, and which he had re-registered, with payment of the pollution tax, in Romania.

[171]For procedural reasons, Mr Nicula’s case was referred back to the court of first instance in Romania for a fresh ruling. After the case had been re-entered on the roll of that court, Romania introduced a new tax, the environmental stamp duty, which was imposed on all vehicles, whether or not imported. In Mr Nicula’s case, the environmental stamp duty in respect of his motor vehicle exceeded the amount of pollution tax he had paid (and was due to be refunded), with the result that he was no longer entitled to recover the pollution tax and the amount he had paid in respect of the pollution tax was, under the new law, retained by the tax and environmental authorities.

[172]The new law provided for a system of repayment of the incompatible pollution tax which imposed a condition that the amount of the pollution tax had to exceed the new environmental stamp duty, and that only the excess was repayable. The balance was retained in lieu of the environmental stamp duty which would become due on the next registered transfer of the vehicle. The ECJ thus had to examine whether such a system of repayment by offsetting made it possible for individuals, such as Mr Nicula, to exercise the right available to them under EU law to seek repayment of the pollution tax unlawfully paid.

[173]That system of repayment was held to be in breach of the principle of effectiveness. The Court said, at [36]:

“As stated by the European Commission, it follows that a system of repayment such as that in issue in the main proceedings has the effect, in the case of a second hand vehicle imported from another Member State, of restricting or, as in the main proceedings, completely eliminating the requirement to pay the pollution tax levied in breach of EU law, which perpetuates the discrimination established by the Court on the judgments in Tatu ( EU:C:2011:219 ) and Nispeanu ( EU:C:2011:466 ).”

[174]Nicula was not concerned with a procedural requirement, but with a system of repayment which impinged upon the right to recover the unlawful tax, or part of it, and accordingly perpetuated the discrimination against the importers of vehicles from other member states which had given rise to the unlawfulness of the pollution tax. The system went not to the procedural means whereby a person could seek repayment of the unlawful tax, but to the substance of the right of recovery itself.

[175]The difficulties with the Romanian pollution tax did not end with Nicula . In Silvia Georgiana Câmpean v Serviciul Fiscal Municipal Mediaş and Administraţia Fondului pentru Mediu (

Case C-200/14

) EU:C:2016:495 , provision was introduced into Romanian law for sums determined by judicial decision relating to the refund of the pollution tax, interest until the date of full payment and legal costs to be paid over a period of five years by annual payments of 20% of the amount payable. Repayment was also contingent on the availability of funds to be received from the new environmental stamp duty, over which claimants had no control.

[176]The finding of the Court was that such a system of repayment was in breach of the principle of effectiveness. The individual claimant was put into a “prolonged period of uncertainty” as to the date when he would obtain a full refund of the improperly paid tax, without any means to compel the public authority to execute its obligation if it did not do so voluntarily, either for reasons related to lack of funds or other reasons.

[177]Câmpean is an example of a case where the principle of effectiveness operated to preclude a national rule which purported both to defer the time at which repayment of an unlawful tax and to introduce an element of uncertainty as to the enforcement of the right to repayment. That rule operated in relation to a claim that had been vindicated by an award in the courts. But the fact that a remedy for the breach of EU had been established in the courts did not preclude the continuing application of the principle of effectiveness as regards the enforcement of that remedy, and the effective exercise by the claimants of their EU law rights. I do not accept, therefore, Miss Foster’s submission that the restitutionary remedy awarded to BAT means that BAT has exercised its directly-effective San Giorgio rights and that the relevant principles, including that of effectiveness, have been given effect to and exhausted. The principle of effectiveness goes beyond the making of the award itself, and operates to preclude the erection of further hurdles to the proper substantive exercise of a claimant’s EU law rights.

[178]The question, then, is whether Part 8C is such a provision which offends the principle of effectiveness with respect to the restitution interest which BAT has been awarded in vindication of its EU law rights in that respect. There can be no doubt that, in an ordinary case, the incidence of taxation on the profits arising from an award would not render the exercise by a claimant of its EU law right either virtually impossible or excessively difficult. In such a circumstance, it is the vindication of the right itself which has given rise to the taxation of the profit, according to the ordinary corporation tax rules. That conclusion is supported by the recent decision of the Upper Tribunal in Coin-a-Drink Limited v Revenue and Customs Commissioners

[2017] UKUT 211 (TCC)

, which was decided shortly after the hearing of this appeal. Although Coin-a Drink in the Upper Tribunal was on the question of the taxation of the repayment of VAT wrongly paid, and not in relation to taxation of the award of statutory interest (there having been no appeal from the decision of the First-tier Tribunal in that respect), the argument for the taxable person was that since the sum in question was levied in breach of EU law, the taxable person was entitled under EU law to the whole of the restitutionary award and that any reduction for taxation would undermine the EU law requirement for an effective remedy. That argument, which was rejected by the Upper Tribunal, could equally have been applied to a restitutionary award of compound interest, which itself satisfies an entitlement to an adequate indemnity under EU law.

[179]Having regard to the nature of Part 8C, and its introduction, it cannot be equated to the ordinary taxation of the VAT repayment (or the statutory interest) at issue in Coin-a-Drink . It is necessary to consider whether the imposition of the 45% tax charge and its ring-fencing from available reliefs itself contravenes the principle of effectiveness. The question is whether the taxation provided for by Part 8C has rendered it virtually impossible or excessively difficult for BAT to exercise its EU law right to compound interest as an adequate indemnity for its loss.

[180]In my judgment, Part 8C does not contravene the principle of effectiveness. Although that principle can operate after a judgment, as it did in Nicula and in Câmpean , what it is concerned with is the substantive exercise of the claimant’s EU law rights. Both Nicula and Câmpean are examples of cases where it is the award itself in vindication of the claimant’s EU law rights that is denied to the claimant or in respect of which the claimant’s rights are inhibited. By contrast, the incidence of taxation on a compensatory award is predicated on that award having been made and on the benefit of that award having been received by the taxpayer in a form that is subject to tax.

[181]That is not to say that there are no circumstances in which taxation of an award could breach the principle of effectiveness. If it were to be found that, although taking the form of taxation, the measure in question was in substance confiscatory, that in my judgment could equally be held to be in breach of the principle of effectiveness. But it is necessary to have regard to all the circumstances to determine whether that is the case. It is plain that the mere incidence of taxation in the ordinary course does not amount to a confiscatory measure. In my judgment, in applying the principle of effectiveness to a case of this nature, a determination has to be made whether, viewed objectively, the measure, albeit one of taxation, is confiscatory in nature. There has to be a dividing line, and in my view that line is one of rationality and proportionality; a taxation provision may be confiscatory if and to the extent that it does not form a rational and proportional basis for taxation. In other words, such a provision must be rational and proportionate as a taxation measure.

[182]I have rejected BAT’s submission that the purpose of HMRC in introducing Part 8C was anything other than to eliminate the perceived windfall, which was considered to have arisen due to the twin effects of the award being subject to a lower rate of corporation tax than those which would have applied to accruals of interest in the relevant period, and the fact of the award having been on the basis of the compounding of interest gross and not net of that accruals based taxation. The 45% rate was, however, arrived at by means of a model constructed by HMRC on a basis that ignored the reliefs and allowances which individual companies likely to be subject to the 45% rate would have been entitled to in the relevant years. Furthermore, the 45% rate was to be applied to the restitution interest “ring-fenced”, thereby precluding companies from offsetting any of that tax by means of present reliefs which would have been available for offset according to normal corporation tax principles. Those two features require careful consideration. A. The 45% rate

[183]I accept, as Mr Aaronson submitted, that EU law requires the national courts to provide claimants, such as BAT, with an adequate indemnity, and that must be based on their individual circumstances. At the same time, as was accepted by BAT, that adequate indemnity is limited to full restitution, and does not include any windfall element. BAT argued that the evidence showed that, having regard to the reliefs and allowances available to it at the relevant times, it did not receive a windfall. I do not agree, for the reasons I have explained above. But in any event, whether BAT, or any other individual claimant, could show that there was no actual windfall, or a windfall of a lesser amount, if such reliefs and allowances were brought into account, that could not be decisive. The question is not whether there is a different measure of windfall which could have been employed in establishing, either individually or generally, a rate of tax to eliminate that windfall, but whether the approach of HMRC in setting a single rate applicable in every relevant case was a rational one.

[184]If the only rational approach were to tailor a tax rate in each individual case so as to eliminate the windfall on a case by case basis, then I accept, as Mr Forrest and Professor Devereux agreed, that a claimant-specific rate would, as a matter of economic theory, be a better way of achieving that objective. Mr Aaronson submitted that a “one-size-fits-all” tax rate could not be consistent with the requirements of EU law, and that the only approach that could fully satisfy the requirements of EU law, namely to ensure that each claimant received a full indemnity for its payment of unlawful tax but without thereby gaining a windfall advantage, was the claimant-specific approach. But the mischief at which the Part 8C provisions were aimed was not focused on individual cases. It was a mischief perceived, and rationally perceived, to apply generally by virtue of the fact that restitutionary awards for the time value of money over a long period, at a time when nominal rates of tax were historically high, and compounded gross, would produce a higher gross amount than if the interest had been treated as taxed at a marginal rate equal to the nominal rate, and compounded net of that taxation.

[185]It was not irrational, therefore, to seek to determine a tax rate that applied generally to redress this imbalance. I accept, as Miss Foster submitted, that the objective of Part 8C was to place claimants, such as BAT, in the same position as a taxpayer company which had received the same amount of interest in each relevant accounting period and had paid tax on that interest at the relevant due and payable dates for each such period. But the taxpayer company to which Miss Foster refers is not BAT, nor is it any other claimant company. It is a notional company which is assumed to have paid tax (at a marginal rate which must be the nominal rate). Because it is a notional company, no assumptions can be made as to whether it has any available reliefs which could be employed to offset the notional tax without cost to the notional company.

[186]That, in my judgment, in the circumstances that had presented themselves, was an entirely rational approach. It was not irrational for the Government to determine that a single rate should apply to all affected companies. As has been demonstrated, there is no compelling evidence which would favour an approach of seeking to apply an individual effective rate of tax, or any other effective rate deduced from various sources to operate as a proxy for an individual effective rate. It cannot in my view be established that such an approach would have been the only rational one, or indeed a more rational one than the approach of identifying, by the means adopted, a single rate of tax. It was rational for the HMRC model to have been constructed to isolate the value attributable to those identified features of the circumstances of the awards that were perceived, in general terms, to produce a windfall. It was rational to take the view that these were the elements that were common to the claimants, and common to the awards of restitution interest.

[187]It was equally rational, therefore, for a model which was designed to apply generally not to take account of the infinitely variable circumstances of the individual claimants, including the availability of reliefs in any particular company. As Miss Foster submitted, any approach which took into account reliefs or allowances available to individual companies would have to have regard to whether a particular relief or allowance had in reality been used to offset tax in a subsequent year. If so, some adjustment would be necessary. In his evidence, Mr Forrest accepted that, if a single tax rate was to be specified, it would be very difficult to take into account the timing effects of certain reliefs being used to offset tax on restitution interest and then not being available to offset future profits on which tax might subsequently otherwise be payable. Adjustments would also be needed to unravel group relief surrenders that might have been made. Fundamentally, any choices that would need to be made with respect to the use of reliefs (and Mr Cohn confirmed in his evidence that part of the role of the tax professional was to maximise the use of reliefs, which required judgments to be made) would have to be made by reference to circumstances which had arisen many years in the past.

[188]The imposition of a tax which applies generally to those entitled to the particular class of income on which the tax is to apply, and which applies a single rate of tax to all, is not in my judgment by virtue of those features in breach of the principle of effectiveness to the extent that it applies to an element of an award in vindication of a claimant’s San Giorgio right. I reject therefore Mr Aaronson’s argument that it is not possible, in the context of a restitutionary award in respect of a breach of a claimant’s EU law rights, for there to be a “one size fits all” tax. Although it is necessary to have regard to the rights of an individual claimant and the loss suffered by the individual claimant in determining the adequate remedy in respect of any claim, that does not in my judgment preclude a rational system of taxation designed to address the perceived windfall by reference to the taxation effects over the period of accrual of the restitution interest.

[189]In Littlewoods Ltd and others v Revenue and Customs Commissioners

[2015] STC 2014

, the Court of Appeal considered the question of the measure of interest to be paid in compliance with the principle of effectiveness, which required that a taxpayer should not be deprived of an adequate indemnity for its loss. It had been held by Henderson J in the High Court that sections 78 and 80 of the Value Added Tax Act 1994 (“VATA”) violated the principle of effectiveness and should be disapplied to enable the claimants to bring their claims and that the correct approach to quantification of the claims was to ascertain the objective use value of the overpaid tax, which was properly reflected in an award of compound interest.

[190]One of the arguments raised on appeal by HMRC was that the principle of effectiveness did not require cases to be considered on an individual basis, and that it was open to a member state to set up a system of rules even if those rules might operate harshly on the facts of an individual case. The system needed only to provide a fair balance of the interests of the individual taxpayer and the interests of society as a whole which is consistent with the principle of effectiveness ( Littlewoods , at [91](v)). That argument was rejected. In addressing the question whether, by restricting interest to simple interest, s 78 VATA violated the principle of effectiveness by depriving Littlewoods of an adequate indemnity for its loss, and giving the judgment of the court, Arden LJ said, at [103]:

“We are also unable to accept HMRC's submission that one should approach the question of whether s 78 affords an adequate indemnity by looking at the system as a whole and ignoring 'hard cases' or 'outliers'. First, the EU law right is, in the terms in which it is expressed in the case law, a private or personal right of the taxpayer. National law must give effect to that right, and it is no answer to the individual taxpayer's claim that national law has done so for other taxpayers, or even for the vast majority of them. Secondly, it is clear from the way in which the court [4] expressed itself at paras 29 to 31 of its judgment that what it envisaged in the present case was an assessment of the position of the individual taxpayer, and not a generic assessment of the overall functioning of the section. Thus the court asked whether the taxpayer 'in the case at issue' and 'having regard to all the circumstances of the case' had been deprived of an adequate indemnity.”

[191]It was thus no answer, in the case of an individual claimant for whom an adequate indemnity required something more than the simple interest provided for by s 78 VATA, to be able to point to the generality of claimants for whom such a remedy would have been adequate. But whilst an argument of that nature cannot operate to prevent something that would make the exercise by a claimant of its EU law right either virtually impossible or excessively difficult, and thus be in breach of the principle of effectiveness, from being such a breach, it is not the case that the mere application of a provision which has general effect, and is not tailored to the individual circumstances of a claimant, necessarily has the consequence that it is such a breach. That must be determined having regard to all the circumstances, including the rationality and proportionality of the measure in question.

[192]The requirement to consider individual circumstances when framing an adequate indemnity does not have the consequence that rationally-conceived tax legislation in the form of Part 8C, which applies a single 45% rate of tax, and does not provide for a claimant-specific rate, is in breach of the principle of effectiveness, even if it could be shown that the application of that rate (when taken together with the ring-fencing) resulted in a tax charge greater than a “windfall” calculated by taking into account individual reliefs and allowances. The individual circumstances of the case that fall to be considered include not only the circumstances of the claimant, but also the nature of the provision which is said to be in contravention of the principle of effectiveness.

[193]Although Mr Aaronson argued that the particular claimants in the FII GLO would have as a common feature surplus ACT, which could be used, at least in part, to offset notional tax on assumed marginal income, and that this could have been taken into account in the modelling, it was not in my view irrational for HMRC not to have followed this course. It was concerned with a broader constituency than those claimants in the FII GLO. Furthermore, as the evidence shows, it is not clear that the basic premise, an ACT mountain, can hold good in all cases, once unlawful ACT has been stripped out.

[194]In fact, HMRC’s model did take account of the particular class of claimant in the FII GLO, along with other classes of claimant. It developed a weighted average by doing so. There was no challenge to the data or logic applied by KAI; Mr Forrest had no material issues with those aspects of the calculations. The result of that process was the production of the 45% rate. I consider that the process by which that rate was arrived at was both logical and rational as were the underlying assumptions that underpinned the rate. It produced a rational basis of taxation, and not confiscation. In contrast to Nicula , it does not perpetuate the unlawful situation. It does not deprive BAT of an adequate remedy; it properly addresses a factor of the restitutionary award that would tend towards over-compensation.

[195]Furthermore, although HMRC’s modelling did have regard to the classes of claimant likely to be affected by the charge to tax on restitution interest, it was not targeted at any individual claimant, or any individual award. It was not, to use the expression found in Marks & Spencer v Customs and Excise Commissioners (

Case C-62/00

)

[2002] STC 1036

, a case concerning the retrospective curtailment of rights to recover unlawful tax, “intended specifically to limit the consequences of a judgment of the [ECJ] to the effect that national legislation concerning a specific tax is incompatible with Community law”. It lacks the necessary element of confiscation. B. Ring fencing

[196]The other material feature of Part 8C is the ring-fencing of the restitution interest, which means that the tax upon it is not capable of being offset by any current reliefs or allowances. That approach was sought to be justified in a number of ways. It was said, for example, that it was the natural corollary of the fact that individual reliefs and allowances were not taken into account in the setting of the 45% rate. Logically, however, ring-fencing would have been even more appropriate in a case where a claimant-specific rate was established by reference to individual reliefs and allowances, or if the rate had reflected reliefs and allowances more generally by use of effective, rather than nominal rates. Ring-fencing in those circumstances would be required to avoid the double application of reliefs. Where, by contrast, the 45% rate was set without reference to such reliefs and allowances, I do not therefore consider that ring-fencing was dictated by the approach to the setting of the rate.

[197]In my view, however, ring-fencing is a natural, and logical, feature of any tax that is applied to present income or profit by reference to an assumed tax effect in earlier periods with a view to restoring the taxpayer to the position of a hypothetical taxpayer which paid tax in those earlier periods. It is natural and logical in those circumstances because the effect would otherwise be to permit current reliefs effectively to be carried back, and would not replicate the position of the hypothetical taxpayer paying tax at the relevant times. The use of current reliefs would thus be objectionable irrespective of the approach taken to the ascertainment of the appropriate tax rate. The one does not justify the other; each must be objectively considered on an independent basis.

[198]In these circumstances, therefore, I consider that the ring-fencing element of Part 8C is a rational and proportionate legislative response. It does not breach the principle of effectiveness. The principle of sincere cooperation

[199]Nor in my judgment is there any breach of the principle of sincere cooperation, reflecting the treaty obligation on member states to take all measures to ensure fulfilment of their obligations under EU law, which precludes a member state from adopting provisions making repayment of a tax held to be contrary to EU law by a judgment of the ECJ (or whose incompatibility with EU law results from such a judgment) subject to conditions relating specifically to that tax which are less favourable that those that would have applied, in their absence, to such a payment (see Francovich v Italian Republic; Bonifaci and others v Italian Republic (Joined cases C-6/90 and C-9/90)

[1991] ECR I-5357

, at [36]; Câmpean , at [44]). Part 8C does not amount to the imposition of such conditions in respect of any award; it is nothing more than a rational scheme of taxation in the context of the particular circumstances of the award of restitution interest. The principles of legal certainty and legitimate expectation

[200]There is in this case, as BAT’s submissions acknowledged, an overlap between the principles of legal certainty and legitimate expectation and the principle of effectiveness. The principle of legal certainty requires that the law should be clear and predictable. That of legitimate expectation in this context requires that those who act in good faith on the basis of the law as it is or seems to be should not be frustrated in their legitimate expectations.

[201]As BAT accepted, a trader can have no legitimate expectation that there will be no change in the law (see, for example, FII(SC) , per Lord Sumption at [152] and per Lord Reed at [241] – [242]). Nor will the mere fact of a disadvantage under the changed law be capable of founding a complaint by a trader of the disappointment of a legitimate expectation ( ATB v Ministerio per le Politiche Agricole (

Case C-402/98

)

[2000] ECR I-5501

[202]There is, on the other hand, as Advocate General Slynn put it in Mulder v Minister van Landbauw en Visserij (

Case C-120/86

) [1988] ECR 2321 , at p 2341), a dividing line between what is merely “hard business luck” and what is unreasonable treatment.

[203]Mulder itself was a case in which it was held that a milk producer who had taken advantage of a Community measure encouraging the suspension of milk production for a limited period in the general interest and in return for a premium could legitimately expect not to be made the subject of specific restrictions on resumption of production arising precisely from the fact that the producer had taken advantage of the option offered by community legislation.

[204]Other cases where legitimate expectations have been held to have arisen have concerned the retrospective effect of legislation. Thus, for example, in Grundstückgemeinschaft Schloßstraße GbR v Finanzamt Paderborn (

Case C-396/98

)

[2000] ECR I-4279

the ECJ held that a taxable person who commenced trading in good faith and who was entitled to an immediate right of deduction in respect of input VAT in relation to the letting of immovable property could not be deprived of that right retrospectively by a legislative change (removing the right to waive exemption) made after the supply had been made.

[205]The cross over between the principle of legitimate expectation and the principle of effectiveness can be observed in FII(ECJ)3 , where it was held that, whilst the claimants had no legitimate expectation that Parliament would not shorten the limitation period for mistake claims, they could legitimately expect that it would not be curtailed peremptorily and without notice, as that would render the vindication of their EU law rights excessively difficult or impossible, and thus be a breach of the principle of effectiveness.

[206]In so far as any reliance is sought to be placed on an argument that Part 8C operates with retroactive effect, I reject such a proposition. Part 8C applies only to restitution interest in respect of which HMRC’s liability to pay the interest becomes final (by determination or final settlement agreement) on or after 21 October 2015, or as regards withholding such interest which is paid on or after 26 October 2015. I also note that, although s 357YW contains a power to amend Part 8C by regulation, that power cannot be used to introduce retrospection (see s 357YW(2)(d)). The fact that Part 8C is designed to eliminate a perceived windfall over a period which can extend back as far as 1973 does not introduce any element of retrospectivity. Nor is there any element of retrospection by virtue of the fact (which is not the case for BAT, as I understand from Mr Cohn’s evidence) that the interest might have been recognised for accounting purposes in an earlier accounting period, notwithstanding that it was not at that stage final, and consequently fell into charge to tax under the loan relationship provisions of Parts 5 or 6 of the Corporation Tax Act 2009. The recognition of a loan relationship credit in an earlier period does not result in a retrospective effect for a provision that operates only with regard to conditions which must be satisfied after the date of its announcement in the TIIN. Section 357YV CTA 2010 ensures that there is no double taxation in those circumstances.

[207]While it is correct to say that a person may have a legitimate expectation that the principle of effectiveness will not be breached in relation to the vindication by that person of its EU law rights, that can take BAT no further in this case. I have concluded that Part 8C does not breach the principle of effectiveness; there can therefore have been no breach of any legitimate expectation on the part of BAT in that respect. The principle of proportionality

[208]I have found that HMRC’s approach satisfies the test of rationality. It is a proper response to what was a legitimate concern in the light of the making of a restitutionary award in respect of a prior period with the accompanying consequences.

[209]It was in turn, in my view, a proportionate response to those concerns. Proportionality is a general principle of EU law. It was explained by the Supreme Court in R (Lumsdon and others) v Legal Services Board

[2016] AC 697

in the following way (at [33]: Proportionality as a general principle of EU law involves a consideration of two questions: first, whether the measure in question is suitable or appropriate to achieve the objective pursued; and secondly, whether the measure is necessary to achieve that objective, or whether it could be attained by a less onerous method. There is some debate as to whether there is a third question, sometimes referred to as proportionality stricto sensu: namely, whether the burden imposed by the measure is disproportionate to the benefits secured. In practice, the court usually omits this question from its formulation of the proportionality principle. Where the question has been argued, however, the court has often included it in its formulation and addressed it separately, as in R v Minister for Agriculture, Fisheries and Food, Ex p Fedesa (

Case C-331/88

)

[1990] ECR I-4023

[210]It is true, as BAT submitted, that the principle of proportionality is in general terms applied more strictly as a ground of review of national measures that it is with respect to a review of EU measures ( Lumsdon , at [37]). But a less strict approach is adopted where the subject matter lies within an area of national rather than EU competence. There is a distinction, therefore, between a case where the measure in question would amount to a confiscation of a part of an award in respect of a breach of EU law, and thus within the scope of EU competence, and unharmonized taxation of such an award, which is essentially a matter of national competence.

[211]In my judgment, whatever approach is taken to proportionality in this case, the measure adopted by the government in the form of Part 8C, and the process by which it was arrived at satisfy the test of proportionality. First, for the reasons I have given, I am satisfied that Part 8C was a suitable response to the identified mischiefs, and that the approach was a rational and reasonable one in the circumstances.

[212]Secondly, the fact that there may be possible alternative courses of action does not render the approach actually taken by HMRC unjustified. The test is not one of strict necessity (see, in a human rights context concerning deprivation of property, R (on the application of the Public and Commercial Services Union and others) v Minister for the Civil Service

[2011] EWHC 2041 (Admin)

, at [46] referring to James v UK

(1986) 8 EHRR 123

, at [51]). James was a case in the European Court of Human Rights (“ECtHR”) concerning the domestic leasehold enfranchisement legislation. The relevant passage from the judgment of the ECtHR is at [51]:

“According to the applicants, the security of tenure that tenants already had under the law in force provided an adequate response and the draconian nature of the means devised to give effect to the alleged moral entitlement, namely deprivation of property, went too far. This was said to be confirmed by the absence of any true equivalent to the 1967 Act in the municipal legislation of the other Contracting States and, indeed, generally in democratic societies. It is, so the applicants argued, only if there was no other less drastic remedy for the perceived injustice that the extreme remedy of expropriation could satisfy the requirements of Article 1. This amounts to reading a test of strict necessity into the Article, an interpretation which the Court does not find warranted. The availability of alternative solutions does not in itself render the leasehold reform legislation unjustified; it constitutes one factor, along with others, relevant for determining whether the means chosen could be regarded as reasonable and suited to achieving the legitimate aim being pursued, having regard to the need to strike a 'fair balance'. Provided the legislature remained within these bounds, it is not for the Court to say whether the legislation represented the best solution for dealing with the problem or whether the legislative discretion should have been exercised in another way.”

[213]The availability of alternative solutions is accordingly a factor, but it is not decisive. In this case, HMRC gave proper consideration to legislating for a claimant-specific rate. Their decision to proceed as they did, in the circumstances of the case and having regard to the mischief at which the provision was to be aimed, was a rational one, suitably aimed at redressing the consequences of the award and, in my judgment, a proportionate response to the issues which was justified in all the circumstances. Although it can be accepted that a different legislative approach, one that took account of reliefs available during the relevant period, would in principle be less onerous, to the extent that it would result in a lesser burden of tax, that factor does not persuade me that Part 8C fails to strike a “fair balance” between the public interest in addressing the perceived windfall available to claimants in respect of awards of restitution interest. In my judgment, Part 8C is reasonably necessary for the achievement of its objectives, and it is not obliged to be the least intrusive means of doing so. It represents in my view “sensible and practical decision-making in the public interest” in the circumstances of this case (see R (Clays Lane Housing Co-operative Ltd) v Housing Corporation

[2004] EWCA Civ 1658

, per Maurice Kay LJ, at [25]). Nor, as regards proportionality stricto sensu, for the same reasons do I consider that it has been demonstrated that the burden imposed by Part 8C is disproportionate to the benefits secured.

[214]Finally, as Lumsdon makes clear, at [64], where a relevant public interest is engaged in an area where EU law has not imposed complete harmonisation, the member state possesses “discretion” (or a “margin of appreciation”) in choosing an appropriate measure. As I have decided that Part 8C does not amount to a confiscation such as to be in breach of the principle of effectiveness, but is a provision of unharmonized taxation, that margin of appreciation is applicable to this case. In my judgment, the policy decisions made by HMRC, its approach to ascertaining the proper rate of tax on restitution interest and its structuring of Part 8C accordingly were all proper exercises by HMRC of their discretion and well within the margin of appreciation.

[215]BAT’s submission that Part 8C fails the tests of rationality is based on the assumption that the only rational means of addressing the mischief identified with respect to the restitution interest is to adopt a claimant-specific approach, and to take reliefs into account. It is on that basis that it is argued that Part 8C fails properly to address the “unfairness” that is said to arise if restitution interest were left to be taxed at current corporation tax rates because, first, it applies uniformly to all claimants, even though their positions differ, and secondly, it is said, systematically results in an excessive charge so that “everyone is a loser”. However, for the reasons I have given, I do not accept that the matter is to be judged on the basis of such an assumption. Part 8C is a rational response, and the means whereby the 45% rate was arrived at also meets the test of rationality. It is rational to apply the tax uniformly, and not to have regard to the infinite variety of the individual tax positions of the claimants. It is rational not to take account of individual reliefs, including the cost to individual claimants of applying those reliefs. It is thus nothing to the point of proportionality that an individual claimant may be able to show that, if reliefs were to have been taken into account, it would have been in a better position if interest had been taxed over the period on an accruals basis than it would be after the application of Part 8C.

[216]BAT also submit in this connection that Part 8C is unnecessary because the impact of taxation could properly be taken into account by the national courts in determining the amount of the tax. Even if that is right, it does not seem to me to get BAT very far. BAT’s claim has been vindicated by an award which, as I have described, did not take tax into account in the way that Part 8C is designed to do. If a court were to take tax into account, it would no doubt do so with regard not only to the historic tax position, but also the tax on the award itself. The effect of Part 8C would be a factor in any such net award. But that is not a reason for Part 8C to be held to be either unnecessary or disproportionate. Conclusion on principles of EU law

[217]My conclusion is that both the setting of the 45% rate and the ring-fencing element of Part 8C was the result of a rational and defensible process undertaken by HMRC. Its approach in not taking into account individual reliefs and allowances, whether by seeking to legislate for a claimant-specific rate, or by some other means such as the use of effective rates, was rational as a matter of principle and a proportionate response to the identified mischief. It is not necessary, in my view, for HMRC to seek to justify its approach by reference to practical difficulties, such as the difficulty in verifying individual calculations by reason of absence of information available to HMRC, or by reference to any perceived difficulty in legislating for a claimant-specific rate (as to which there was general consensus in the evidence that thousands of rates might be required). Part 8C in my judgment does not contravene the principle of effectiveness, or any other principle of EU law. As there is no breach, no justification for any breach is required.

[218]Were any question of justification to have arisen, Mr Aaronson submitted that the unavailability of data could not be a justification for domestic tax laws that breached the principles of EU law. He referred in this respect to FII(ECJ)1 , in particular to [70], where the ECJ held that difficulties which might arise in determining the tax actually paid in another member state could not justify a restriction on the free movement of capital such as that arising in that case, where domestic dividends from companies in which the recipient held fewer than 10% of the voting rights were exempt, but corresponding dividends from companies in other member states were subject to UK tax, with relief only for withholding tax and not for underlying tax. Were, contrary to my conclusion, Part 8C to have been found to be confiscatory and thus to breach the principle of effectiveness, I agree that reliance on practical difficulties would not excuse or eliminate that breach. But that does not arise in this case.

[219]As a further reason why the unavailability of information to HMRC could not justify a breach, if there were one, Mr Aaronson submitted that the compensatory nature of the restitutionary remedy necessarily places the responsibility for establishing the amount of the loss suffered by a claimant on the claimant itself. I also accept that proposition. But, for the reasons I have given, the question of justification does not arise in this case, and the fact that a claimant might be able to provide the relevant information in verifiable detail (although I do not accept that BAT’s calculations have succeeded in doing so) cannot affect the rationality of Part 8C.

[220]Mr Aaronson also submitted that HMRC could not seek to justify their rejection of a claimant-specific rate on the basis that this would require extremely long and complex legislation. Mr Aaronson is right to point out that the evidence of Mr Rounding was that there had been no approach to HMRC’s legal experts or Parliamentary draftsman before a decision was taken to reject a claimant-specific approach. But the question is not whether there is any alternative to HMRC’s approach, but whether their approach was a rational one. I have found that it was rational. Mr Aaronson helpfully prepared a number of drafts of an alternative Part 8C to illustrate his submission that a claimant-specific tax could be enacted in reasonably short and intelligible legislation. He acknowledged, rightly in my view, that these drafts were not the finished article, and that the Parliamentary draftsman could be expected to improve them, but irrespective of whether it is right that something workable could be produced, in my judgment that does not assist BAT’s case. Fundamental rights

[221]BAT’s case also rests on what it claims is the infringement by Part 8C CTA of certain of its fundamental rights. The rights in question are conveniently subdivided into two parts. First are the rights provided for by, respectively,

Article 6 of the Convention

for the Protection of Human Rights and Fundamental Freedoms (European Convention on Human Rights or “ECHR”) (right to a fair hearing) and Article 47 of the Charter of Fundamental Rights of the European Union (“CFREU”) (right to an effective remedy and right to a fair hearing). The second are the rights provided by Article 1, Protocol 1 (A1P1) of the ECHR and Article 17(1) of the CFREU (right to/protection of property).

[222]It will not be necessary to address the particular drafting of those provisions in any detail, but for completeness I include the relevant texts at Appendix C to this decision. Fair hearing and effective remedy

[223]The first question to be addressed is one of the engagement of each of

Article 6 ECHR

and Article 47 CFREU.

[224]As regards

Article 6 ECHR

, HMRC’s case is that it is clearly established by the ECtHR in Ferrazzini v Italy

(2001) 31 EHRR 19

that Article 6 does not apply to tax disputes such as that in the present case because those disputes are neither criminal in nature (except as regards certain penalties) nor do they involve the determination of civil rights and obligations. Despite reservations that have been expressed, both judicially (a substantial minority reached the contrary view in Ferrazzini itself) and academically, that principle has been followed and applied by the ECtHR in subsequent cases, including as recently as 20 December 2016 in the case of Lindstrand Partners Advokatbyrå AB v Sweden (Application no 18700/09,

[2016] ECHR 1139

) (see [110] – [113]).

[225]Except in special circumstances, the domestic courts, including this tribunal, are bound to follow the clear and consistent jurisprudence of the ECtHR. Recently, in R (on the application of APVCO 19 Ltd and others) v Her Majesty’s Treasury and another

[2015] STC 2272

, in the context of a question whether certain retrospective legislation targeting certain avoidance schemes in relation to stamp duty land tax (SDLT) violated the ECHR, the Court of Appeal held that, in reliance on Ferrazzini , the determination of the efficacy of the schemes was not a civil right or obligation in respect of which the claimants had been denied a fair and public hearing by the legislative changes (see, per Vos LJ, at [68]). That principle is binding on this tribunal.

[226]Mr Margolin, making submissions for BAT in this respect, accepted the hurdle which Ferrazzini represents, but sought to circumvent it by arguing that the present dispute should not be characterised as a tax dispute, but as a dispute concerning the framing of an essentially confiscatory provision as tax legislation, where the award said to have been the subject of confiscation was itself the product of a case concerning civil rights and obligations, namely the right of BAT and the other claimants in the FTT GLO to restitution in respect of unlawful tax paid.

[227]I do not accept that the nature of the proceedings in the FII GLO can have the effect that Article 6 can be applied to the present case. The appeal to this Tribunal is undoubtedly a tax dispute; the Tribunal has jurisdiction under s 357YS CTA only as regards the deduction of the withholding tax from the payment of restitution interest. In so far as Mr Margolin’s submission was premised on Part 8C being confiscatory, and not a normal taxation measure, I have for the reasons I have outlined above rejected that characterisation. Ferrazzini applies, and Article 6 is not engaged.

[228]On the other hand, there is no reference in Article 47 of the CFREU to “civil rights and obligations”. The rights, principles and freedoms in Article 47 extend beyond those contained in

Article 6 of the ECHR

. This has been confirmed by the Court of Appeal in ZZ v Secretary of State for the Home Department

[2011] EWCA Civ 440

, per Maurice Kay LJ at [16], referring to the Explanations in the CFREU with respect to Article 47. Ferrazzini is inapplicable to the engagement of that Article.

[229]Miss Foster submitted that Article 47 is nonetheless not engaged in this case because the case concerns the setting of a national tax rate by primary legislation in the field of an unharmonized tax within the competence of Parliament. She relied in this respect on the further observations of Maurice Kay LJ in ZZ , where he said at [16] – [17]:

“… However, what the Charter does not and cannot do is to give birth to rights, freedoms and principles in areas in which the Treaties claim no rule-making competence but acknowledge the exclusive competence of Member States. This is spelt out in Article 51.2 of the Charter, as to which the Explanations state: ‘[Article 51.2] confirms that the Charter may not have the effect of extending the competences and tasks which the Treaties confer on the Union. Explicit mention is made here of the logical consequences of the principle of subsidiarity and of the fact that the Union only has those powers which have been conferred on it … [It] also confirms that the Charter may not have the effect of extending the field of application of Union law beyond the powers of the Union as established in the Treaties … it goes without saying that the reference to the Charter in Article 6 of [TEU] cannot be understood as extending by itself the range of Member State action considered to be ‘implementation of Union law’.’ In other words, a Member State is not to be taken to be acting ‘in the implementation of Union law’ if it is acting within an area which, under the Treaties, is not allocated for Union legislation. 17 It follows that the potential of Article 47 as a legal peg upon which the appellant might hang his claim to procedural fairness derived from EU law has to be assessed by reference to the allocation of competences by the Treaties…”

[230]I do not agree with Miss Foster in this respect. In my view the scope of this appeal, although undoubtedly an appeal concerning taxation in an unharmonized field, includes questions on the application of EU law principles which are, as I have found, apt to be considered in determining the lawfulness of Part 8C. As those principles are within the competence of the EU Treaties, Article 47 CFREU is engaged. As Mr Aaronson and Mr Margolin submitted, the underlying rights which form the context for this appeal are EU law rights, and the issues before the tribunal go to the implementation of EU law. That is the essence of the dispute before the tribunal, even if it arises in the context of a tax appeal.

[231]Having said that, it is difficult to see how BAT’s reliance on Article 47 can take it further than the principle of effectiveness. That is the essential principle of EU law which falls to be applied. The CFREU does not operate to extend the competencies of the EU or the application of EU law. Article 47 is concerned with ensuring that the underlying EU law rights are effectively protected by ensuring compliance with the right to an effective remedy and to a fair hearing (see Lesoochranárske zoskupenie VLK v Obvodný úrad Trenčín (

Case C-243/15

) ECLI:EU:C:2016:838 ). Consistently with this, essentially the same case is made by BAT in relation to Article 47 as was made with respect to the principle of effectiveness. It is argued that Part 8C is an unjustified interference with the right of BAT to recover the full amount of the judgment debt it has been awarded. It is submitted that Part 8C is disproportionate, applying the four-stage analysis outlined in Bank Mellat v HM Treasury (No 2 )

[2014] AC 700

, per Lord Reed at [74] of the judgments in the substantive appeal. I have considered each of those arguments in relation to the principle of effectiveness and there is nothing to add in relation to Article 47. I would observe only that BAT’s reliance on Bank Mellat in this context is misplaced. As has been made clear in Lumsdon , at [26], although there is some common ground the four-stage analysis of proportionality as explained in that case with respect to the ECHR is not applicable to proportionality in EU law. There is therefore nothing further to add to the earlier discussion based on Lumsdon . Right to/protection of property

[232]No jurisdictional questions arise with respect to A1P1 of the ECHR or Article 17 of the CFREU. BAT’s right to restitution interest is a possession within each provision. Part 8C will deprive BAT of a part of that possession, and needs to be justified as being in the public interest. There is no difference between the scope of A1P1 and Article 17, and the case law relevant to A1P1 is thus applicable to both.

[233]The relevant test is set out by the ECtHR in National & Provincial Building Society v United Kingdom (117/1996/736/933-935; 23 October 1997)

(1998) 25 EHRR 127

. That case had a complex history, but the essential point to which the A1P1 debate related was whether the interference with the possessions of certain building societies through certain legislation with retrospective effect which brought to an end claims of those societies to recovery of amounts paid to the Inland Revenue, was justified. The principles are described by the Court at [80]:

“According to the Court's well-established case law, [5] an interference, including one resulting from a measure to secure the payment of taxes, must strike a “fair balance” between the demands of the general interest of the community and the requirements of the protection of the individual's fundamental rights. The concern to achieve this balance is reflected in the structure of Article 1 as a whole, including the second paragraph: there must therefore be a reasonable relationship of proportionality between the means employed and the aims pursued. Furthermore, in determining whether this requirement has been met, it is recognised that a Contracting State, not least when framing and implementing policies in the area of taxation, enjoys a wide margin of appreciation and the Court will respect the legislature's assessment in such matters unless it is devoid of reasonable foundation…”

[234]Against the background of that wide margin of appreciation, the Court in National & Provincial Building Society considered the facts in that case. It reasoned that the legislation in question had been enacted with retroactive effect to restore the original intention of Parliament in bringing into force certain regulations designed to forestall the building societies from enjoying a windfall on a change to the regime for charging tax on interest earned on deposits with the societies. Those regulations had been held (by the House of Lords in R v Inland Revenue Commissioners, ex parte Woolwich Equitable Building Society

[1990] STC 682

) to be ultra vires on technical grounds. The Court in National & Provincial Building Society held (at [81]) that in those circumstances the ultimate aim of the measure was not without reasonable foundation having regard to the public interest considerations which underpinned the proposal to legislate with retroactive effect and Parliament’s endorsement of that proposal. The Court also observed that there was an obvious and compelling interest to ensure that private entities do not enjoy the benefit of a windfall in a changeover to a new tax payment regime and do not deny the Exchequer revenue simply on account of inadvertent defects in the enabling tax legislation.

[235]The tests have been consistently repeated and consistently applied, both by the ECtHR and by the domestic courts. More recently, AXA General Insurance Ltd and others v HM Lord Advocate and others

[2012] 1 AC 868

, in the Supreme Court, concerned the passing of primary legislation to reverse the effect of case law establishing that certain conditions did not constitute any injury capable of giving rise to a claim to damages. On (amongst other things) an A1P1 challenge by certain insurers, who might be liable to indemnify employers in respect of such claims, and having held that the insurers had an A1P1 possession, Lord Hope referred, at [31], to the approach of the EctHR:

“The approach that the Strasbourg court takes to this matter was explained in James v United Kingdom [6] , para 46, in which the court said: ‘Because of their direct knowledge of their society and its needs, the national authorities are in principle better placed than the international judge to appreciate what is ‘in the public interest’. Under the system of protection established by the Convention, it is thus for the national authorities to make the initial assessment both of the existence of a problem of public concern warranting measures of deprivation of property and of the remedial action to be taken. Here, as in other fields to which the safeguards of the Convention extend, the national authorities accordingly enjoy a certain margin of appreciation. ‘Furthermore, the notion of ‘public interest’ is necessarily extensive. In particular, as the commission noted, the decision to enact laws expropriating property will commonly involve consideration of political, economic and social issues on which opinions within a democratic society may reasonably differ widely. The court, finding it natural that the margin of appreciation available to the legislature in implementing social and economic policies should be a wide one, will respect the legislature's judgment as to what is ‘in the public interest’ unless that judgment be manifestly without reasonable foundation.’ This formula has been repeated in many cases since that date: see, for example, Broniowski v Poland (2005) 40 EHRR 21 , para 149; Maurice v France (2006) 42 EHRR 42 , para 84. In Draon v France 42 EHRR 40 , para 76 the court said that the notion of ‘public interest’ is necessarily extensive as it will commonly involve consideration of political, economic and social issues. The court will, it said, respect the legislature's judgment as to what is in the public interest unless that judgment is manifestly without reasonable foundation.”

[236]In another case outside the field of tax, but like National & Provincial Building Society also concerning perceived over-compensation, Affaire Sud Parisienne de Construction v France (Application No 33704/04; 11 February 2010), the ECtHR considered whether the enactment of a law and regulation retrospectively reducing the rate of default interest on public contracts constituted an unjustified interference with the claimant’s A1P1 rights. The law had been enacted after the termination of the relevant contract, and after proceedings had been brought for recovery of the amount of an invoice for work carried out before cancellation and default interest at the rate provided for in the contract (17%).

[237]That was a clear interference in a possession of the claimant for the purpose of A1P1. The question was whether it was justified. The justification given by the French government was that the legislative scheme was based on a number of public interest imperatives. These were described as the correction of a major economic dysfunction due to the upheaval of monetary conditions and the very sharp fall in the rate of inflation. The aim was to restore legal and financial coherence to the rate of default interest due on public contracts so that the rate was determined on the basis of a reasonable ratio to inflation and approaching the rates actually applied in the market for short-term financing, and that equal treatment could be ensured between contract holders.

[238]Those reasons were considered by the Court to be relevant, sufficient and convincing. The law was justified by “imperative reasons of public interest”. In reaching that conclusion, the court reiterated that, having regard to their direct knowledge of society and its needs, national authorities are in principle better placed than the court to determine what is in the public interest. The national authorities enjoy a certain margin of appreciation. A legislator normally has considerable discretion in pursuing economic and social policy, and the court will respect the judgments of the national authority in such respects, unless that judgment is manifestly lacking a reasonable basis. This then returns to the question of rationality that I have considered earlier.

[239]The court considered whether the legislative interference in the rate of default interest with retroactive effect had the effect of nullifying the main proceedings that had been brought by the claimant. Such an interference would be disproportionate. It concluded that was not the case. The interference related to only part of the default interest, since it concerned only the fixing of the rate. The claimant had been entitled to succeed in the main proceedings. Its right to compensation had not been affected; but the law had corrected on a reasonable basis by reference to inflation a difference, or anomaly, which had arisen by reason of a change in monetary conditions.

[240]The principles were reiterated by the ECtHR in the case of NKM v Hungary (Application no 66529/11; 14 May 2013)

[2013] STC 1104

, where at [48] to [51] the Court said:

“48. It follows that, in addition to being in accordance with the domestic law of the contracting state, including its Constitution, the legal norms upon which the deprivation of property is based should be sufficiently accessible, precise and foreseeable in their application (see Guiso-Gallisay v Italy (App no 58858/00, [2005] ECHR 832 ) (8 December 2005, unreported), paras 82–83). The court would add that similar considerations apply to interferences with the peaceful enjoyment of possessions… 49. The court would, moreover, reiterate the finding in its settled case law that the national authorities are in principle better placed than an international court to evaluate local needs and conditions. In matters of general social and economic policy, on which opinions within a democratic society may reasonably differ widely, the domestic policy-maker should be afforded a particularly broad margin of appreciation (see, for example, Stec v UK 20 BHRC 348 , para 52). 50. In so far as the tax sphere is concerned, the court's well-established position is that states may be afforded some degree of additional deference and latitude in the exercise of their fiscal functions under the lawfulness test (see National & Provincial Building Society v UK [1997] STC 1466 , 25 EHRR 127 , paras 75 to 83; OAO Neftyanaya Kompaniya YUKOS v Russia [2011] STC 1988 , 54 EHRR 599, para 559). 51. Moreover, since in the present case the interference with the applicant's peaceful enjoyment of possessions was incarnated by a tax measure, it is convenient to point out that retroactive taxation can be applicable essentially to remedy technical deficiencies of the law, in particular where the measure is ultimately justified by public-interest considerations. There is in fact an obvious and compelling public interest to ensure that private entities do not enjoy the benefit of a windfall in a changeover to a new tax-payment regime (see National etc, paras 80 to 83). However, no such deficiency of the law has been demonstrated in the circumstances of the present case. Therefore, the court considers that particular caution is called for when assessing whether or not the impugned measure was 'lawful' for the purposes of art 1 of the First Protocol.”

[241]Those principles were helpfully distilled (albeit obiter) by Proudman J in Lobler v Revenue and Customs Commissioners

[2015] STC 1893

, at [84], where she summarised the position as being that the interference, that is with the peaceful enjoyment of possessions, could be justified if it satisfied three tests:(a) the legislation must be sufficiently accessible, precise and foreseeable in its application;

(b) the legislation must pursue a legitimate aim in the public interest; and

(c) the interference with the right to peaceful enjoyment must be proportionate in the sense that it strikes a fair balance between the demands of the general interest of the community and the requirements for the protection of the individual’s fundamental rights.

[242]It is the third of these tests which carries the greatest weight. As the Court in NKM said, at [60]:

“Even if [the interference with A1P1] has taken place subject to the conditions provided for by law—implying the absence of arbitrariness—and in the public interest, an interference with the right to the peaceful enjoyment of possessions must always strike a ‘fair balance’ between the demands of the general interest of the community and the requirements of the protection of the individual's fundamental rights. In particular, there must be a reasonable relationship of proportionality between the means employed and the aim sought to be realised by the impugned measure…”

[243]In NKM , the question before the Court concerned the enactment by the Hungarian Parliament of a new tax on certain payments for employees in the public sector whose employment was terminated. Severance payments and certain other amounts became taxable, with some retrospective effect, at the rate of 98% above a certain limit, translated into an effective rate on the severance payment of 52%, as contrasted with the ordinary income tax rate of 16%. The bill preceding the Act justified the tax with reference to "public morals” and the unfavourable budgetary situation of the country. It was held that the “sense of social justice of the population” in combination with an interest to protect the public purse and to distribute the public burden satisfied the ECHR requirement of a legitimate aim, notwithstanding the broad nature of that aim. But the tax failed to proportionality test, in that it did not strike a fair balance. In the particular circumstances of the case, the measure was found to entail an excessive and individual burden on the applicant, and consequently to have violated A1P1.

[244]Mr Margolin argued that this was a compelling analogy when considering BAT’s case. The Court in NKM had drawn attention to the fact that, although not decisive in itself, the tax rate in question exceeded by a considerable margin the rate applicable to all other revenues ( NKM , at [67] - [68]) and to the fact that it targeted only a certain group of individuals (with the majority of citizens not being obliged to contribute, to a comparable extent, to the public burden) ( NKM , at [72]). Further, the Court noted, at [73], that the tax was directly deducted by the employer (in other words, the State) without any individualised assessment of the applicant’s situation.

[245]On this basis, Mr Margolin submitted that Part 8C and the application of a 45% tax rate is similarly discriminatory, in that it targets only a specific cadre of companies, that is to say recipients of restitution interest, within the wider body of corporation tax payers, it allows for no consideration of the circumstances of individual taxpayers and the applicable tax is substantially higher than the ordinary corporation tax burden. It is accordingly submitted that the 45% tax imposes an excessive and disproportionate burden on BAT, and its stated purpose does not provide sufficient justification for its application in terms of the public interest.

[246]I do not agree. As Miss Foster submitted, the facts of NKM are very different from those in this case, and I do not accept that the parallels sought to be drawn by Mr Margolin can in fact be drawn. In any event, the test to be applied is of a fair balance having regard to the circumstances of an individual case; the search for parallels with other cases, themselves decided by reference to their own facts, is of little assistance in the application of the relevant principles to the facts at hand.

[247]I have already concluded, contrary to BAT’s submission on the facts, that BAT did become entitled to a windfall in terms of the gross award for restitution interest that was made in its favour. That gross award was greater than it would otherwise have been if the calculation had assumed that the twin mischiefs identified by the Government, namely the fact that tax at the historically higher rates had not been applied to the restitution interest during the period over which it had been assumed to accrue and that consequently interest had been compounded gross rather than by reference to interest net of such tax, had been eliminated. I have also rejected BAT’s submission that the purpose of HMRC in introducing Part 8C was anything other than to eliminate the perceived windfall. I accept, for the reasons I have given above when considering the principle of effectiveness, Miss Foster’s submission that the reason for the introduction of Part 8C is rationally connected to the particular circumstances giving rise to the windfall. That can be contrasted with NKM , where none of the reasons provided by the Hungarian Government could be said to have been directly applicable to the applicant, who only received compensation that was due by statute and who could not have been made responsible for the fiscal problems, or excessive risk-taking in the financial services sector, which the Government intended to remedy. On the principles set out in the cases I have referred to, and as applied by the ECtHR in Sud Parisienne by way of example, the interference in BAT’s possession was justified by reason of public policy, to remedy the perceived windfall effects. It did not interfere with the right of BAT to compensation, but merely rationally addressed an anomaly.

[248]I have also found, again in the context of the application of the principle of effectiveness, that the Part 8C tax is not confiscatory. In that regard, although BAT referred to cases in the ECtHR such as Agurdino SRL v Moldova (Application no 7359/06,

[2011] ECHR 1455

; 27 September 2011) and Scordino v Italy (No 1)

(2007) 45 EHRR 7

, neither case can assist. In the former, there was a quashing of a judgment by which the applicant had been absolved from paying an amount to the tax inspectorate. In the latter, there was an enactment, during the course of proceedings, of a law which had the effect of depriving the claimants retrospectively of a substantial part of the compensation previously recoverable from the State for the expropriation of land. In this case, I do not consider that Part 8C falls into the category of a confiscatory provision. It is a rational measure of taxation to address matters of taxation which have given rise to anomalous results. It cannot be considered to be outside the State’s margin of appreciation, nor can it be described as devoid of reasonable foundation. It is not necessary, in arriving at that conclusion, to seek to assess the extent or breadth of the margin of appreciation by reference to the nature of the decision or the identity of the decision-maker, in this case Parliament through the enactment of primary legislation. Such an approach, although adopted by the Court of Appeal in R (on the application of Sinclair Collis Ltd) v Secretary of State for Health

[2011] EWCA Civ 437

, was doubted by the Supreme Court in Lumsdon (see Lumsdon , at [81]). But in my view, whatever margin of appreciation is applicable, the rational and proportionate response provided by Part 8C falls well within it.

[249]It is true to say that Part 8C affects only a particular class of corporate taxpayer. But it does so for the perfectly rational reason that it is that class of taxpayer which is entitled to the restitution interest in question, to which Part 8C is to be applied, again on an entirely rational basis as I have described. It is no different in that respect from any other tax on a particular category of income which necessarily affects only those in receipt of such income.

[250]There is an element of retrospection in the purpose of Part 8C, although not in its application. However, that element of retrospection is itself rational in that it seeks to address an anomaly which arises from the fact that the restitution interest is a present remedy for a past wrong. Unlike the position in NKM , where the tax was imposed on income related to activities prior to the material tax year but which arose in the tax year of the applicant’s dismissal, so that tax at a considerably higher rate than that which was in force when those earlier activities had been carried out, the rationale of Part 8C is to the opposite effect, namely to reflect, in what I have found to be a rational and proportionate way, the tax rates and the tax effect of applying those rates to restitution interest in the period in respect of which it has been calculated.

[251]In AXA , at [122], Lord Reed explained the significance of the retroactive effect of legislation in the civil sphere in the following way:

“The Strasbourg court has recognised that the fact that legislation in the civil sphere has retroactive effects does not necessarily mean that it is incompatible with the rule of law or the Convention. In relation to A1P1, in particular, the court has considered retroactive effects in its assessment of proportionality rather than when considering the lawfulness of the interference, and has found such effects to be objectionable only in particular circumstances where they imposed an “individual and excessive burden” upon the applicant. In Mellacher v Austria (1989) 12 EHRR 391, for example, which concerned the introduction of rent controls that were applicable to existing leases, the court stated (para 51), in its consideration of proportionality, that in remedial social legislation, and in particular in the field of rent control, it must be open to the legislature to take measures affecting the further execution of previously concluded contracts in order to attain the aim of the policy adopted. In Zielinski v France (2001) 31 EHRR 19 , which concerned a retrospective change in employment law and was brought under article 6(1), the court stated (para 57) that while in principle the legislature is not precluded in civil matters from adopting new retrospective provisions to regulate rights arising under existing laws, the principle of the rule of law and the notion of fair trial enshrined in article 6 preclude any interference by the legislature—other than on compelling grounds of the general interest—with the administration of justice designed to influence the judicial determination of a dispute. In Bäck v Finland (2005) 40 EHRR 48 , which concerned legislation enabling courts to authorise arrangements under which a debtor's pre-existing obligations to his creditors were modified, the court stated (para 68) that neither the Convention nor its Protocols preclude the legislature from interfering with existing contracts.”

[252]For the reasons I have given, I do not consider that Part 8C is disproportionate, nor that it imposes an “individual or excessive burden” on BAT. The aim of Part 8C was a rational aim and the measures taken by the Government in enacting Part 8C to attain that aim were themselves rational and proportionate. They were not designed to influence the judicial determination of a dispute nor, I might add, to do anything other than seek to address the effect of the twin mischiefs that had been identified with the award of compound interest in respect of the period in question on a gross basis. They were not designed to deprive BAT of its adequate indemnity to compensate it in vindication of its San Giorgio right.

[253]For those reasons, I conclude that BAT’s challenge to Part 8C under A1P1 of the ECHR and Article 17(1) of the CFREU must fail. Constitutionality and the rule of law

[254]I can deal with this quite shortly. As expressed in argument, the rule of law and the essentially different roles of the executive and the courts formed less of a substantive ground in their own right than a context in which the challenges to Part 8C under EU and human rights law were made. It was accepted by BAT that this Tribunal is bound, not least by British Railways Board v Pickin , to give effect to primary legislation, except in so far as it might be held to be unlawful by reference to those laws. Remedies

[255]As I have found that Part 8C CTA does not breach EU law or any ECHR or CFREU right of BAT, it is not necessary to consider the question of remedies. Although the parties made submissions in that respect, in particular as to whether it would be possible to read down Part 8C, and its 45% rate of tax, as a matter of conforming interpretation, so as to impose an appropriate rate of tax on each claimant’s restitution interest, I do not consider that it would assist any higher court or tribunal if I were to offer what would, in these circumstances, be nothing more than tentative, and what is more theoretical, observations on what is a question of law. Reference to the CJEU

[256]I was urged by Mr Aaronson, in the event that I had any reasonable doubt about the conclusion as to the breach of the EU law principle of effectiveness (at least in so far as I did not accede to his submission that I should have no reasonable doubt that there had indeed been such a breach), I should refer the case immediately to the Court of Justice of the European Union (“CJEU”). Mr Aaronson pointed out that in the current state of affairs in relation to Brexit, and given that references made before the UK’s exit from the EU will be heard by the CJEU, the sensible course would be to refer the case immediately.

[257]The proper approach to a question of whether to make a reference has been authoritatively summarised by Sir Thomas Bingham MR in R v International Stock Exchange of the United Kingdom and the Republic of Ireland Ltd ex parte Else (1982) Ltd and another

[1993] QB 534

at page 545. In essence, unless the court or tribunal, faced with a material question of EU law, can with complete confidence resolve that issue, or if it is in any real doubt in that respect, it should ordinarily refer.

[258]A distinction must be drawn between issues of EU law, and issues as to the application of an aspect EU law which has already been the subject of judicial clarification by the ECJ or the CJEU. As the ECJ itself observed, in CILFIT (Srl) v. Ministry of Health (Case 283/81)

[1982] ECR I-3415

, no purpose is served by the making of a reference where previous decisions of the ECJ have already dealt with the question of law. Where there is no doubt about the scope and interpretation of the EU law itself, but the matter is one of application of that law to the facts of a particular case, that is a material factor to be taken into account in determining the appropriateness of a reference.

[259]That, I consider, is the position in this case with respect to the arguments on EU law. Those arguments have focussed, rightly in my view, on the application of established principles of EU law as regards which, and as has been clearly demonstrated by the number and nature of the judgments of the ECJ and the CJEU which have been cited, there is no absence of authoritative guidance on the applicable principles. In the light of that guidance, I consider that I have been able with complete confidence to resolve all relevant issues of EU law, and to apply that law to the facts of this case. In those circumstances, I do not consider it would be appropriate for me to make a reference to the CJEU. Decision

[260]I dismiss BAT’s appeal. Application for permission to appeal

[261]This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application must be received by this Tribunal not later than 56 days after this decision is sent to that party. The parties are referred to “Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and forms part of this decision notice. ROGER BERNER TRIBUNAL JUDGE RELEASE DATE: 12 JULY 2017 APPENDIX A Part 8C, Corporation Tax Act 2010 357YA Charge to corporation tax on restitution interest(1) The charge to corporation tax on income applies to restitution interest arising to a company.

(2) In subsection (1) the reference to a company does not include a charitable company. 357YB Restitution interest chargeable as income (1) Profits arising to a company which consist of restitution interest are chargeable to tax as income under this Part (regardless of whether the profits are of an income or capital nature). (1A) In subsection (1) the reference to a company does not include a charitable company. (2) In this Part references to “profits” are to be interpreted in accordance with section 2(2) of CTA 2009. 357YC Meaning of “restitution interest” (1) In this Part “restitution interest” means profits in relation to which Conditions A to C are met. (2) Condition A is that the profits are interest paid or payable by the Commissioners for Her Majesty's Revenue and Customs in respect of a company's right (or possible right) to restitution with regard to either of the following matters (or alleged matters)— (a) the payment of an amount to the Commissioners under a mistake of law relating to a taxation matter, or (b) the unlawful collection by the Commissioners of an amount in respect of taxation.

(3) Condition B is that— (a) a court has made a final determination that the Commissioners are liable to pay the interest, or (b) the Commissioners have in final settlement of a claim in respect of the right (or possible right) mentioned in subsection (2) entered into an agreement under which a person is entitled to be paid, or is to retain, the interest.

(4) Condition C is that the interest determined to be due, or agreed upon, as mentioned in subsection (3) is not limited to simple interest at a statutory rate (see section 357YU).

(5) Subsection (4) does not prevent so much of an amount of interest determined to be due, or agreed upon, as represents or is calculated by reference to simple interest at a statutory rate from falling within the definition of “restitution interest”.

(6) For the purposes of subsection (2) it does not matter whether the interest is paid or payable— (a) pursuant to a judgment or order of a court, (b) as an interim payment in court proceedings, (c) under an agreement to settle a claim, or (d) in any other circumstances.

(7) For the purposes of this section— (a) “interest” includes an amount equivalent to interest, and (b) an amount paid or payable by the Commissioners as mentioned in subsection (2) is “equivalent to interest” so far as it is an amount determined by reference to the time value of money.

(8) For the purposes of this section a determination made by a court is “final” if the determination cannot be varied on appeal (whether because of the absence of any right of appeal, the expiry of a time limit for making an appeal without an appeal having been brought, the refusal of permission to appeal, the abandonment of an appeal or otherwise).

(9) Any power to grant permission to appeal out of time is to be disregarded for the purposes of subsection (8). 357YD Further provision about amounts included, or not included, in “restitution interest” (1) Interest paid to a company is not restitution interest for the purposes of this Part if— (a) Condition B was not met in relation to the interest until after the interest was paid, and (b) the amount paid was limited to simple interest at a statutory rate (2) Subsection (1) does not prevent so much of a relevant amount of interest determined to be due, agreed upon or otherwise paid as represents or is calculated by reference to simple interest at a statutory rate from falling within the definition of “restitution interest”. (3) In subsection (2) “relevant amount of interest” means an amount of interest the whole of which was paid before Condition B was met in relation to it. (4) Section 357YC(7) applies in relation to this section as in relation to section 357YC. 357YDA Life insurance companies: amounts representing policyholder income (1) This section applies if— (a) an amount of interest paid or payable by the Commissioners for Her Majesty's Revenue and Customs would (but for this section) be restitution interest arising to a company, and (b) were this Part not to have effect, that amount would be taken into account under section 73 of FA 2012 (the I-E basis) as income chargeable for an accounting period of the company that is referable to its basic life assurance and general annuity business. (2) So much (if any) of the amount as represents policyholder income is to be treated for the purposes of this Part as if it were not restitution interest. (3) To determine how much (if any) of the amount mentioned in subsection (1) (amount “A”) represents policyholder income, take the following steps— Step 1 (a) Take so much of amount A as consists of non-ACT interest (“the non-ACT amount”). (b) Determine how much (in total) of the non-ACT amount is to be assigned to with-profits funds (one or more) of the company. Call this total amount “P”. In this step “non-ACT interest” means interest which is not interest in respect of advance corporation tax. Step 2 Determine how much of P is to be assigned to each of the with-profits funds concerned. This is the “assignable amount” in the case of each fund. Step 3 In the case of each fund mentioned in step 2, determine in what proportions profits of the fund concerned are to be divided between policyholders and shareholders under the distribution policy for the fund. Step 4 Express the policyholders' proportion (as determined under step 3) as a percentage of the whole. This is the “policyholder percentage” for the fund. Step 5 Multiply each assignable amount by the policyholder percentage for the fund in question. The result is the “policyholder amount” in the case of each fund. Step 6 Amount A “represents policyholder income” so far as it does not exceed the total policyholder amounts found under step 5. (4) For the purposes of subsection (3) “the distribution policy for the fund” means the basis on which the company has decided profits of the fund are to be divided between policyholders and shareholders. (5) The distribution policy for a with-profits fund is to be determined as at the time when the interest arises, and with particular reference to— (a) any relevant information in the company's articles of association, and (b) any relevant information or document published by the company in connection with obligations under the FCA Handbook. (6) In this section— “the FCA Handbook” means the Handbook made by the Financial Conduct Authority under the Financial Services and Markets Act 2000, and “interest” has the same meaning as in section 357YC. 357YE Period in which amounts are to be brought into account (1) The amounts to be brought into account as restitution interest for any period for the purposes of this Part are those that are recognised in determining the company's profit or loss for the period in accordance with generally accepted accounting practice. (2) If Condition A in section 357YC is met, in relation to any amount, after the end of the period for which the amount is to be brought into account as restitution interest in accordance with subsection (1), any necessary adjustments are to be made; and any time limits for the making of adjustments are to be disregarded for this purpose. 357YF Companies without GAAP-compliant accounts (1) If a company— (a) draws up accounts which are not GAAP-compliant accounts, or (b) does not draw up accounts at all, this Part applies as if GAAP-compliant accounts had been drawn up. (2) Accordingly, references in this Part to amounts recognised for accounting purposes are references to amounts that would have been recognised if GAAP-compliant accounts had been drawn up for the period of account in question and any relevant earlier period. (3) For this purpose a period of account is relevant to a later period if the accounts for the later period rely to any extent on amounts derived from the earlier period. (4) In this section “GAAP-compliant accounts” means accounts drawn up in accordance with generally accepted accounting practice. 357YG Restitution interest: appeals made out of time (1) This section applies where— (a) an amount of interest (“the interest”) arises to a company as restitution interest for the purposes of this Part, (b) Condition B in section 357YC is met in relation to the interest as a result of the making by a court of a final determination as mentioned in subsection (3)(a) of that section, (c) on a late appeal (or a further appeal subsequent to such an appeal) a court reverses that determination, or varies it so as to negative it, and (d) the determination reversing or varying the determination by virtue of which Condition B was met is itself a final determination. (2) This Part has effect as if the interest had never been restitution interest. (3) If— (a) the Commissioners for Her Majesty's Revenue and Customs have under section 357YO(2) deducted a sum representing corporation tax from the interest, or (b) a sum has been paid as corporation tax in respect of the interest under section 357YQ, that sum is treated for all purposes as if it had never been paid to, or deducted or held by, the Commissioners as or in respect of corporation tax. (4) Any adjustments are to be made that are necessary in accordance with this section; and any time limits applying to the making of adjustments are to be ignored. (5) In this section— “final determination” has the same meaning as in section 357YC; “late appeal” means an appeal which is made by reason of a court giving leave to appeal out of time. 357YH Countering effect of avoidance arrangements (1) Any tax advantages that would (in the absence of this section) arise from relevant avoidance arrangements are to be counteracted by the making of such adjustments as are just and reasonable in relation to amounts to be brought into account for the purposes of this Part. (2) Any adjustments required to be made under this section (whether or not by an officer of Revenue and Customs) may be made by way of an assessment, the modification of an assessment, amendment or otherwise. (3) For the meaning of “relevant avoidance arrangements” and “tax advantage” see section 357YI. 357YI Interpretation of section 357YH (1) This section applies for the interpretation of section 357YH (and this section). (2) “Arrangements” include any agreement, understanding, scheme, transaction or series of transactions (whether or not legally enforceable). (3) Arrangements are “relevant avoidance arrangements” if their main purpose, or one of their main purposes, is to enable a company to obtain a tax advantage in relation to the application of the charge to tax at the restitution payments rate. (4) But arrangements are not “relevant avoidance arrangements” if the obtaining of any tax advantages that would (in the absence of section 357YH) arise from them can reasonably be regarded as consistent with wholly commercial arrangements. (5) “Tax advantage” includes— (a) a repayment of tax or increased repayment of tax, (b) the avoidance or reduction of a charge to tax or an assessment to tax, (c) the avoidance of a possible assessment to tax, (d) deferral of a payment of tax or advancement of a repayment of tax, or (e) the avoidance of an obligation to deduct or account for tax. (6) In subsection (5)(b) and (c) the references to avoidance or reduction include an avoidance or reduction effected by receipts accruing in such a way that the recipient does not bear tax on them as restitution interest under this Part. 357YJ Examples of results that may indicate exclusion not applicable (1) Each of the following is an example of something which might indicate that arrangements whose main purpose, or one of whose main purposes, is to enable a company to obtain a tax advantage are not excluded by section 357YI(4) from being “relevant avoidance arrangements” for the purposes of section 357YH— (a) the elimination or reduction for the purposes of this Part of amounts chargeable as restitution interest arising to the company in connection with a particular claim, if for economic purposes other or greater profits arise to the company in connection with the claim; (b) preventing or delaying the recognition as an item of profit or loss of an amount that would apart from the arrangements be recognised in the company's accounts as an item of profit or loss, or be so recognised earlier; (c) ensuring that a receipt is treated for accounting purposes in a way in which it would not have been treated in the absence of some other transaction forming part of the arrangements. (2) In this section “arrangements” and “tax advantage” have the meaning given by section 357YI. 357YK Corporation tax rate on restitution interest (1) Corporation tax is charged on restitution interest at the restitution payments rate. (2) The “restitution payments rate” is 45%. 357YL Exclusion of reliefs, set-offs etc (1) Under subsection (3) of section 4 (amounts to which rates of corporation tax applied) the amounts to be added together to find a company's “total profits” do not include amounts of restitution interest on which corporation tax is chargeable under this Part. (2) No reliefs or set-offs may be given against so much of the corporation tax to which a company is liable for an accounting period as is equal to the amount of corporation tax chargeable on the company for the period at the restitution payments rate. (3) In subsection (2) “reliefs and set-offs” includes, but is not restricted to, those listed in the second step of paragraph 8(1) of Schedule 18 to FA 1998. (4) Amounts of income tax or corporation tax, or any other amounts, which may be set off against a company's overall liability to income tax and corporation tax for an accounting period may not be set off against so much of the corporation tax to which the company is liable for the period as is equal to the amount of corporation tax chargeable at the restitution payments rate. 357YM Assignment of rights to person not chargeable to corporation tax (1) Subsection (4) applies if— (a) a chargeable company (“the transferor”) transfers to a person who either— (i) is not a company, or (ii) is a non-qualifying company, a right in respect of a claim, or possible claim, for restitution, (b) the transfer is made on or after 21 October 2015, and (c) conditions A and B are met. (2) Condition A is that the main purpose, or one of the main purposes, of the transfer is to secure a tax advantage for any person in relation to the application of the charge to tax on restitution interest under this Part. (3) Condition B is that as a result of that transfer (or that transfer together with further transfers of the rights) restitution interest arises to a person who either— (a) is not a company, or (b) is a non-qualifying company. (4) Any restitution interest which arises as mentioned in Condition B is treated for corporation tax purposes as restitution interest arising to the transferor. (5) For the purposes of this section a company is a “chargeable company” if it meets the first and second conditions. The first condition is that the company is UK resident or carries on a trade in the United Kingdom through a permanent establishment in the United Kingdom. The second condition is that the company is not a charitable company and would not be exempt from corporation tax on restitution interest (were such interest to arise to it). (5A) For the purposes of this section a company is a “non-qualifying company” if— (a) it is non-UK resident, or (b) it is a charitable company, or would be exempt from corporation tax on restitution interest (were such interest to arise to it). (6) In this section “tax advantage” has the meaning given by section 357YI. 357YN Migration of company with claim to restitution interest (1) This section applies where— (a) restitution interest arises to a non-UK resident company, (b) the rights in respect of which the company is entitled to the restitution interest had (to any extent) accrued when the company ceased to be UK resident, and (c) the company's main purpose, or one of its main purposes, in changing its residence was to secure a tax advantage for any person in relation to the application of the charge to tax on restitution interest under this Part. (2) The company is treated as a UK resident company for the purposes of the application of this Part in relation to so much of that restitution interest as is attributable to relevant accrued rights. (3) “Relevant accrued rights” means rights which had accrued to the company when it ceased to be UK resident. (4) The company is to be treated for the purposes of sections 185 and 187 of TCGA 1992 as not having disposed of its assets on ceasing to be resident in the United Kingdom, so far as its assets at that time consisted of rights to receive restitution interest. (5) Any adjustments that are necessary as a result of subsection (4) are to be made; and any time limits for the making of adjustments are to be ignored for this purpose. 357YNA Transfer of rights: restitution interest arising after a winding up or dissolution (1) Subsection (2) applies if an amount of restitution interest which is paid or payable to a person would be treated under section 357YM(4) as arising to a company (“the transferor”) but for the fact that the company no longer exists at the time when the restitution interest arises. (2) If an officer of Revenue and Customs gives a related company a notice under this subsection in respect of the restitution interest, the restitution interest is treated for corporation tax purposes as restitution interest arising to that company. (3) Subsection (4) applies if an amount of restitution interest which is paid or payable to a person would apart from this section be treated by virtue of section 357YM(4) as arising to a company which has been wound up (“the transferor”). (4) If an officer of Revenue and Customs gives a related company a notice under this subsection in respect of the restitution interest, the restitution interest is treated for corporation tax purposes as restitution interest arising not to the transferor but to that company. (5) A notice under subsection (2) or (4) must specify— (a) the amount of the restitution interest, and (b) the date on which it is paid or payable. (6) A notice under subsection (2) or (4) in respect of an amount of restitution interest must be given by the later of— (a) the date on which the amount is paid or payable, or (b) the time when any notice under section 357YQ(2) in respect of the amount is given to the related company. 357YNB Meaning of “related company” (1) A company is a “related company” for the purposes of section 357YNA(2) if at any time in the relevant period (see subsection (5)) that company was a member of the same group as the transferor (see section 357YNA(1)). (2) A company is a “related company” for the purposes of section 357YNA(4) if at any time in the relevant period (see subsection (6)) that company was a member of the same group as the transferor (see section 357YNA(3)). (3) For the purposes of this section two companies are members of the same group if— (a) one is a 51% subsidiary of the other, or (b) both are 51% subsidiaries of a third company. (4) In subsection (1) “the relevant period” means the period which— (a) begins— (i) if the transferor was not wound up before it was dissolved, at the beginning of the 12 months ending with the date on which the company is dissolved, (ii) if the transferor was wound up before it was dissolved, at the beginning of the 12 months before the commencement of the winding up, and (b) ends when the amount mentioned in section 357YNA(1) is paid or becomes payable (whichever is later). (5) In subsection (2) the “relevant period” means the period which— (a) begins at the beginning of the 12 months before the commencement of the winding up of the transferor, and (b) ends when the amount mentioned in section 357YNA(3) is paid or becomes payable (whichever is later). 357YO Duty to deduct tax from payments of restitution interest (1) Subsection (2) applies if the Commissioners for Her Majesty's Revenue and Customs pay an amount of interest in relation to which Conditions 1 and 2 are met and— (a) the amount is (when the payment is made) restitution interest on which a company is chargeable to corporation tax under this Part, or (b) a company would be chargeable to corporation tax under this Part on the interest paid if it were (at that time) restitution interest. (2) The Commissioners must, on making the payment— (a) deduct from it a sum representing corporation tax on the amount at the restitution payments rate, and (b) give the company a written notice stating the amount of the gross payment and the amount deducted from it. (3) Condition 1 is that the Commissioners are liable to pay, or have agreed or determined to pay, the interest in respect of a company's claim for restitution with regard to— (a) the payment of an amount to the Commissioners under a mistake of law relating to a taxation matter, or (b) the unlawful collection by the Commissioners of an amount in respect of taxation. (4) Condition 2 is that the interest is not limited to simple interest at a statutory rate. In determining whether or not this condition is met, all amounts which the Commissioners are liable to pay, or have agreed or determined to pay in respect of the claim are to be considered together. (5) For the purposes of Condition 1 it does not matter whether the Commissioners are liable to pay, or (as the case may be) have agreed or determined to pay, the interest— (a) pursuant to a judgment or order of a court, (b) as an interim payment in court proceedings, (c) under an agreement to settle a claim, or (d) in any other circumstances. (6) For the purposes of subsection (2) the restitution payments rate is to be applied to the gross payment, that is to the payment before deduction of a sum representing corporation tax in accordance with this section. (7) For the purposes of this section— (a) “interest” includes an amount equivalent to interest, and (b) an amount which the Commissioners pay as mentioned in subsection (1) is “equivalent to interest” so far as it is an amount determined by reference to the time value of money. 357YP Treatment of amounts deducted under section 357YO (1) An amount deducted from an interest payment in accordance with section 357YO(2)— (a) is treated for all purposes as paid by the company mentioned in section 357YO(1) on account of the company's liability, or potential liability, to corporation tax charged on the interest payment, as restitution interest, under this Part, and (b) is accordingly to be treated for corporation tax purposes as going towards the discharging of the company's liability to pay, for the accounting period concerned, tax charged under this Part (as calculated under paragraph 2 of the fifth step of paragraph 8(1) of Schedule 18 to FA 1998). (2) Subsections (3) and (4) apply if— (a) the Commissioners for Her Majesty's Revenue and Customs have, on paying an amount which is not (when the payment is made) restitution interest, made a deduction under section 357YO(2) from the gross payment (see section 357YO(6)), and (b) a company becomes liable to repay the net amount to the Commissioners, or it otherwise becomes clear that the gross amount cannot, or will not, become restitution interest. (3) If the condition in subsection (2)(b) is met in circumstances where the company is not liable to repay the net amount to the Commissioners, the Commissioners must— (a) repay to the company the amount treated under subsection (1) as paid by the company, and (b) make any other necessary adjustments; and any time limits applying to the making of adjustments are to be ignored. (4) If the condition in subsection (2)(b) is met by virtue of a company becoming liable to repay to the Commissioners the amount paid as mentioned in subsection (2)(a)— (a) this Part has effect as if the company were liable to repay the gross payment to the Commissioners, and (b) the amount deducted by the Commissioners as mentioned in subsection (2)(b) is to be treated for the purposes of this Part as money repaid by the company in partial satisfaction of its liability to repay the gross amount. (5) Subsections (3) and (4) have effect with the appropriate modifications if the condition in subsection (2)(b) is met in relation to part but not the whole of the gross amount mentioned in subsection (2)(a). (6) In this section “the net amount”, in relation to a payment made under deduction of tax in accordance with section 357YO(2), means the amount paid after deduction of tax. 357YQ Assessment of tax chargeable on restitution interest (1) An officer of Revenue and Customs may make an assessment of the amounts in which, in the officer's opinion, a company is chargeable to corporation tax under this Part for a period specified in the assessment. (2) Notice of an assessment under this section must be served on the company, stating the date on which the assessment is issued. (3) An assessment may include an assessment of the amount of restitution income arising to the company in the period and any other matters relevant to the calculation of the amounts in which the company is chargeable to corporation tax under this Part for the period. (4) Notice of an assessment under this section may be accompanied by notice of any determination by an officer of Revenue and Customs relating to the dates on which amounts of tax become due and payable under this section or to amounts treated under section 357YP as paid on account of corporation tax. (5) The company must pay the amount assessed as payable for the accounting period by the end of the period of 30 days beginning with the date on which the company is given notice of the assessment. 357YR Interest on excessive amounts withheld (1) If an amount deducted under section 357YO(2) in respect of an amount of interest exceeds the amount which should have been deducted, the Commissioners for Her Majesty's Revenue and Customs are liable to pay interest on the excess from the material date until the date on which the excess is repaid. (2) The “material date” is the date on which tax was deducted from the interest. (3) Interest under subsection (1) is to be paid at the rate applicable under section 178 of FA 1989. 357YS Appeal against deduction (1) An appeal may be brought against the deduction by the Commissioners for Her Majesty's Revenue and Customs from a payment of a sum representing corporation tax in compliance, or purported compliance, with section 357YO(2). (2) Notice of appeal must be given to Her Majesty's Revenue and Customs— (a) in writing, (b) within 30 days after the giving of the notice under section 357YO(2). 357YT Amounts taxed at restitution payments rate to be outside instalment payments regime For the purposes of regulations under section 59E of TMA 1970 (further provision as to when corporation tax due and payable), tax charged at the restitution payments rate is to be disregarded in determining the amount of corporation tax payable by a company for an accounting period. 357YU Interpretation (1) In this Part “court” includes a tribunal. (2) In this Part “statutory rate” (in relation to interest) means a rate which is equal to a rate specified— (a) for purposes relating to taxation, and (b) in, or in a provision made under, an Act. 357YV Relationship of Part with other corporation tax provisions (1) So far as restitution interest is charged to corporation tax under this Part it is not chargeable to corporation tax under any other provision (including Part 2 of FA 2012: but see also section 357YDA). (2) This Part has effect regardless of section 464(1) of CTA 2009 (priority of loan relationship provisions). 357YW Power to amend (1) The Treasury may by regulations amend this Part (apart from this section). (2) Regulations under this section— (a) may not widen the description of the type of payments that are chargeable to corporation tax under this Part; (b) may not remove or prejudice any right of appeal; (c) may not increase the rate at which tax is charged on restitution interest under this Part; (d) may not enable any provision of this Part to have effect in relation to the subject matter of any claim which has been finally determined before 21 October 2015. (3) Subject to subsection (2), regulations under this section may have retrospective effect. (4) For the purposes of this section a claim is “finally determined” if a court has disposed of the claim by a final determination or the claimant and the Commissioners for Her Majesty's Revenue and Customs have entered into an agreement in final settlement of the claim. (5) Section 357YC(8) (which defines when a determination made by a court is final) has effect for the purposes of this section as for the purposes of section 357YC. (6) Regulations under this section may include incidental, supplementary or transitional provision. (7) A statutory instrument containing regulations under this section must be laid before the House of Commons. (8) The regulations cease to have effect at the end of the period of 28 days beginning with the day on which they are made unless, during that period, the regulations are approved by a resolution of the House of Commons. (9) In reckoning the 28-day period, no account is to be taken of any time during which— (a) Parliament is dissolved or prorogued, or (b) the House of Commons is adjourned for more than 4 days.

(10) Regulations ceasing to have effect by virtue of subsection (8) does not affect— (a) anything previously done under the regulations, or (b) the making of new regulations. APPENDIX B Experts’ Joint Statement of Issues of Agreement and Disagreement (12 December 2016) Introduction 1 . In relation to the dispute between BAT and HMRC, Mr Nicholas Forrest has produced an expert report dated 9 September 2016, and a supplementary report dated 14 October. Professor Michael Devereux has produced an expert report, dated 31 October 2016. Both reports addressed the following three questions asked by the First tier Tribunal (Tax chamber):

a. An assessment of the methodology and data used in the witness statement of Rebecca Richmond to establish the 45% special CT rate for restitution interest;

b. An assessment of the effective rate of tax (i.e. the rate of tax actually paid after reliefs and allowances) at which UK corporate taxpayers paid corporation tax in each of the fiscal years in the period 1973-2016; and

c. An assessment of the advantages and disadvantages in setting the tax by reference to either the nominal (statutory) rate of tax or the effective rate of tax and how each approach would reflect the economic realities of business. 2. This document presents the matters on which the experts, Mr Nicholas Forrest and Professor Michael Devereux, agree and disagree in relation to the three questions asked by the tribunal. 3. This document is structured in accordance with the above questions, with the agreements, disagreements and further discussion separated accordingly. Section A - An assessment of the methodology and data used in the witness statement of Rebecca Richmond to establish the 45% special CT rate for restitution interest 4. It is helpful to divide this question into three sub-sections. The formulation of a counterfactual is important to the application of the methodology used in the determination of the appropriate tax rate to apply to restitution interest, so this is discussed first in Section Al. The extent to which Ms Richmond's methodology is appropriate in the determination of the appropriate tax rate to apply to restitution interest is discussed in Section A2. The calculations and data used by Ms Richmond in the context of her methodology are discussed in Section A3. Section Al 5. With regard to the appropriate counterfactual to use in the determination of the appropriate tax rate to apply to restitution interest, the experts agree on the following issues: a. It is appropriate to use a counterfactual approach when determining the appropriate tax rate to apply to restitution interest; b. The counterfactual should be one in which the unlawfully paid tax is not in fact paid by claimants to HMRC (or, in earlier years, to Inland Revenue or HM Customs and Excise); and c. The HMRC methodology assumes that, in its counterfactual, the claimant is always in a fully taxpaying position and also assumes that the claimant does not change its business in any way as a consequence of not paying the unlawful tax. 6. With regard to the appropriate counterfactual to use in the determination of the appropriate tax rate to apply to restitution interest, the experts disagree on the following issues: a. The extent to which it is appropriate to specify a counterfactual in which, instead of paying the unlawful tax, the company invested the same amount in an asset earning a risk-free rate of interest. b. The existence of an alternative counterfactual to that specified in 6a that is consistent with the award of restitution interest, better reflects the economic realities of business, and is appropriate to use in the determination of the appropriate tax rate to apply to restitution interest. 7. The respective positions of the experts and the points of disagreement are described in more detail below. 8. Regarding the disagreement in 6a, Mr Forrest does not consider that a counterfactual where the company invested the same amount as it paid in unlawful tax in an asset earning a risk-free rate is reasonable, in contrast to the counterfactual proposed by HMRC and supported by Professor Devereux. Mr Forrest contrasts the selection of an appropriate interest rate and the prescribed use of funds in the counterfactual. In a number of important cases, the High Court determined that the appropriate interest rates to be applied to restitution claims were typically a measure of the cost of UK government borrowing over the claim period and that “the claimants are content to take the cost of government borrowing as a reasonable proxy for the time value of their own loss” [7] . In Mr Forrest’s view, the use of a risk free rate is appropriate to make the necessary time-value conversion, as the High Court ruled [8] , and consistent with economic and finance theory. However, it is not the case that this necessarily implies that the appropriate counterfactual to assume is one in which the claimants invest in risk-free assets, nor did the court specify that this was the appropriate counterfactual to use. 9. Mr Forrest considers that a counterfactual whereby claimants invest in risk-free assets separate from the business is not reflective of the economic realities of business. Moreover, in order to disregard business effects, it is necessary to make the further assumption that this investment is kept secret from the business (so that no business decisions are made which rely upon the knowledge of incremental risk-free assets). This represents another unreasonable assumption that does not reflect the economic realities of business. In the counterfactual for the claim calculation, the claimants would not have known about the tax they could have otherwise unlawfully paid (in the actual case) and therefore would not have been able to identify this unlawfully paid tax as anything other than a fungible part of the income earned by the company. This means that neither the unlawfully paid tax nor the interest subsequently earned would have been identified by the company as separable or indeed marginal. The company would, therefore, have treated it no differently to the rest of its income. There is no particular reason, Mr Forrest suggests, to suppose that the unidentifiable income would have been used for any specific purpose, such as the purchase of risk free assets. It is Mr Forrest's contention, therefore, that such a narrowly defined counterfactual is inappropriate for the calculation of the appropriate tax rate to apply to restitution interest. 10. Regarding the disagreement in 6b, Mr Forrest contends that one appropriate counterfactual can be constructed as follows. The unlawfully paid tax would have been an unidentifiable and fungible part of the income earned by the company, Therefore, a neutral assumption would be that, in the counterfactual scenario, the company simply scaled up its activities in proportion to the unlawfully paid tax. Given no particular reason to suppose a company would have used this unidentifiable income for any specific purpose [9] , it is reasonable to adopt an approach that does not presume such a specific purpose or use. By “scale up”, Mr Forrest is simply referring to the fact that, in the counterfactual scenario he is proposing, the claimant company, and its constituent activities, would have been bigger when compared to the firm in the actual scenario. 11. Mr Forrest proposes that such an approach best reflects standard economic theory. In the counterfactual scenario whereby the claimant did not pay the unlawful tax, its (post-tax) marginal revenue would have been higher than in the actual scenario. In these terms, it is clear that the unlawful tax introduced a distortion into the firm's profit maximisation decisions. A profit-maximising firm (with no financing constraint) would produce a greater level of output in the counterfactual scenario, where its marginal revenue is higher, than in the actual scenario. Therefore, the firm in the counterfactual scenario would have “scaled up” its activities to generate this extra output. It is Mr Forrest's contention that this counterfactual is a better representation of the economic realities of business than that outlined in 6a. 12. Professor Devereux believes that the issues in 6a and 6b go to the heart of the disagreement between the two experts. Professor Devereux believes the approach taken by HMRC is the standard approach used in the academic economics and finance literature, as well as in practice, to compare values of assets and income at different points in time. It is also the basis on which the amount of restitution interest to be paid has been determined by the court. Professor Devereux notes that the issue at hand is how that restitution interest should be taxed, not to re-evaluate how much should be paid. 13. Professor Devereux believes that there are several scenarios broadly consistent with the approach taken by HMRC, and which do not involve any change in the activities of the company, or have any impact on the allowances which it could claim. One is that the unlawful tax paid would otherwise have been invested in a risk-free asset. An alternative is that the unlawful tax paid would otherwise have been used to reduce borrowing. Mr Wadey in his statement of June 10, 2016, suggested in paragraph 14 that, if the unlawful tax had not been paid by BAT, “the logical use for the excess cash created by not paying the amounts ... would have been to reduce borrowing”. If, in general, claimants would otherwise have used the funds paid in unlawful tax to reduce borrowing, this would also imply that it would be appropriate to assume that the other activities of the business are unaffected. 14. Professor Devereux believes that another alternative scenario is that the unlawful tax would otherwise have been use by the company to pay a higher dividend to its shareholders. This is consistent with assuming that the company had exhausted all profitable investment opportunities that earned a higher rate of return than that otherwise available to shareholders. Under this interpretation, the amount of restitution interest compensates the shareholders for not having received the dividend in an earlier period. 15. Beyond that point, Professor Devereux points out that there are countless ways in which a company might have behaved had the unlawful tax not been paid. Other counterfactuals might include the company paying higher remuneration to its senior management, or making a donation to charity. It seems fruitless to speculate on what might have happened. 16. Professor Devereux believes that the specific alternative counterfactual proposed here by Mr Forrest is not reasonable. Mr Forrest seeks to draw a distinction between the interest applied in the counterfactual scenario and the use of funds. Instead of the interest rate being the return to investing in a risk-free asset, Mr Forrest suggests that this risk-free interest rate might be the net return from the business investing the funds in the same way as its existing activities - that is, it might simply “scale up” the business. 17. The rate at which the hypothetical interest is calculated is based on a risk-free rate determined by the court for the payment of restitution interest. Professor Devereux believes this is a reasonable approach where the counterfactual involves investing the unlawful tax in an interest-bearing account, or in the other scenarios discussed in paragraphs 13 and 14 above. But he believes it is not a reasonable rate to use under the counterfactual proposed by Mr Forrest, for reasons set out below. 18. The importance of whether the counterfactual can reasonably be defined in the way proposed by Mr Forrest is because this may determine whether it is reasonable to take into account the normal allowances (such as capital allowances) on business investment, and hence provide a justification for the use of an effective tax rate, referred to in Section B below. Effective tax rates are generally lower than the normal statutory rate. 19. Professor Devereux considers the approach proposed by Mr Forrest to be unreasonable for several reasons. · First, Mr Forrest argues that the company's investment would have been reduced in practice by the expectation of paying the unlawful tax. By contrast, he argues that, had the company known that it would not have had to pay the tax, then its investment would have been higher. Professor Devereux argues that, the academic economics and finance literature has generally concluded that taxes on dividend payments (such as ACT, and surplus ACT) do not affect corporate investment decisions financed by retained earnings. [10] The reason is that when funds are retained in the company to finance an investment, the dividend payment is reduced, and the shareholder saves the tax due on the foregone dividend, so the net cost to the shareholder is lower than the amount of the investment. When the return from that investment is subsequently paid out in the form of a dividend, then tax is paid. These two tax effects net out; they do not affect the required rate of return on an investment, and consequently they do not affect investment decisions. [11] · Second, Professor Devereux further argues that even if it were true that the company's investment was reduced to some extent by the prospect of paying the unlawful tax, there is no necessary connection between the amount of the reduction in the investment and the amount of the unlawful tax, so no reason to suppose that the reduced investment would be equal to the amount of unlawful tax. One cannot therefore infer the converse, namely that the whole of the unlawful tax would otherwise have been used for investment and “scaling up” the business. · Third, to assume that the whole of the unlawful tax would have been used to finance additional investment must imply that the business did not take up profitable investment opportunities because of a lack of available funds — that is, if the unlawful tax had not been paid, the business would have used the funds to take up profitable investment opportunities. This is inconsistent with the evidence of Mr Wadey in his evidence of June 10, 2016 at paragraph 13 that “In the period of the claim (that is from 1973) BAT has always had access to capital markets for its funding requirements”, and his comment, cited above, that BAT would have reduced borrowing. · Fourth, the usual assumption in economic analysis is that a business would take up all available investment projects that are expected to earn at least the required rate of return, but none that are expected to earn less than the required rate of return. The "marginal" investment is therefore expected to earn just the required rate of return. It is this "marginal" investment that would be available to a business if it were presented with available funding (as would be the case if the unlawful tax had not been paid). Business investment is generally risky, and the required rate of return should be higher to reflect that risk. This is not consistent with Mr Forrest's argument that higher business investment would earn only the risk-free rate of return. · Fifth, as long as at least one existing project earned more than the required rate of return, then the average rate of return should exceed the marginal rate of return available for new projects. It is therefore not reasonable to apply an effective tax rate to a new project where that effective tax rate is based on the average rate of return of all projects. Section A2 20. With regard to considerations of the extent to which Ms Richmond's methodology is appropriate in the determination of the appropriate tax rate to apply to restitution interest, the experts agree on the following issues: a. A marginal tax rate is the appropriate tax rate to apply to the hypothetical receipt of interest (or other income) in calculating the appropriate tax rate to apply to restitution interest receipts; b. HMRC's methodology uses the nominal statutory tax rate as its measure of the marginal tax rate; it therefore disregards any effects arising from surplus ACT, group relief, unused losses, and disclaimed capital allowances; c. Calculating claimant-specific tax rates for each claimant would be, conceptually, the most accurate approach to meet HMRC's objective of placing “the claimants in broadly the same position as a taxpayer company who had received the same amount of interest in each relevant accounting period and had paid tax on that interest at the relevant due and payable dates for each accounting period” [12] ;

d. The feasibility of calculating any claimant-specific tax rate depends, among other things, on the availability of information for that claimant over the entire period covered by the claim; the actual tax position of the claimant in each year would need to be adjusted under the hypothetical position in which the unlawful tax were not paid;

e. If an averaging approach is to be taken, it is reasonable to apply weights to the tax rates of individual claimants; that is, to use a weighted average;

f. Any weighted average tax rate will differ from a claimant-specific tax rate for most claimants. 21. With regard to the extent to which Ms Richmond's methodology is appropriate in the determination of the appropriate tax rate to apply to restitution interest, the experts disagree on the following issues: a. The extent to which it is feasible to calculate claimant-specific tax rates for all companies; and b. The appropriate weights to apply to the tax rates of individual claimants in the stratified sample if an averaging approach is taken. 22. The respective positions of the experts and the points of disagreement are described in more detail below. 23. Regarding the disagreement specified in paragraph 21a, Mr Forrest considers that he is not best placed to comment on whether it is feasible to calculate claimant-specific tax rates for all claimants. He would like to draw the tribunal's attention to the fact, however, that Table 4.1 in his report demonstrates that it is possible for claimant-specific tax rates to be calculated for individual claimants and one reasonable approach could use a weighted average of these rates where it is not possible to calculate claimant-specific tax rates for all claimants. 24. Professor Devereux addressed the issue of the feasibility of using claimant-specific tax rates in his report in paragraphs 20 to 24. Briefly, he believes that it would be very difficult to set out a policy that clearly described the counterfactual, or draft legislation to specify clearly how to calculate the claimant-specific rate, given that the counterfactual requires the actual tax position to be adjusted on the assumption that the unlawful tax had not been paid. In addition, according to HMRC, the relevant information for calculating claimant-specific rates is not available for all claimants. Professor Devereux has not been able to verify the claims in Mr Forrest's Table 4.1. However, as set out in his report at paragraphs 67 to 75, Professor Devereux does not accept the calculations for BAT made by Mr Cohn in his statement of June 8, 2016. While it should be possible in principle to estimate a claimant-specific rate given the required information, this seems unlikely to be possible for all claimants. Professor Devereux does not agree with the suggestion that it would be reasonable to use a weighted average of claimant-specific tax rates based only on claimants for whom such a claimant-specific tax rate could be calculated. This would ignore the position of other claimants. 25. Regarding the disagreement specified in 21b, Mr Forrest notes that, regardless of the weights used, a weighted average approach will not yield the most accurate tax rate for the majority of claimants. However, he acknowledges that if an averaging approach is to be taken, it is appropriate to use a weighted average. 26. Mr Forrest regards the weights proposed by Ms Richmond that is, weighting the implicit tax rate of claimant companies by the amount of unlawfully paid tax, in the context of her methodology, as appropriate. He also regards the alternative weights proposed by Professor Devereux as reasonable. 27 Mr Forrest argues that each weighting approach has its advantages and disadvantages. For example, he notes that while weighting the individual tax rates by the amount of pre-tax restitution interest, as proposed by Professor Devereux, will yield an accurate result at the aggregate level, there are situations when such an approach would result in potentially undesirable consequences. For example, it could mean that older, larger claims become more important in setting the tax rate and other typically younger, smaller claims could bear tax rates which are very different to the rate they would have paid on a bespoke or claimant-specific basis [13] . Mr Forrest suggests such differences could be greater in the context of Professor Devereux's proposed approach. 28. Professor Devereux proposes that the weight for each company should be based on the amount of pre-tax restitution interest, on the grounds that this is the tax base for each company. He agrees that this gives more weight to older claims. Both his proposal and the weights used by Ms Richmond would give more weight to larger claims. Both weighting schemes would result in a weighted average tax rate that would in general be different from claimant-specific rates. In any weighting scheme, giving a higher weight to one claimant inevitably moves the weighted average closer to the position of that claimant, and away from the position of other claimants. Section A3 29. With regard to the calculations and data used by Ms Richmond in the context of her methodology, the experts agree on the following issues: a. The mechanical calculations undertaken by Ms Richmond to reach the special CT rate for restitution interest appear to be accurate (neither expert has undertaken a review of every calculation performed); and b. While neither expert has had access to the underlying data used by Ms Richmond to establish the special CT rate for restitution interest, data has been obtained from appropriate sources. 30. With regard to the calculations and data used by Ms Richmond in the context of her methodology, the experts have no material disagreements. Section B - an assessment of the effective rate of tax (i.e. the rate of tax actually paid after reliefs and allowances) at which UK corporate taxpayers paid corporation tax in each of the fiscal years in the period 1973-2016. 31. With regard to an assessment of the effective rate of tax, the experts agree on the following issues: a. The methodologies Mr Forrest uses to calculate effective tax rates are reasonable; b. The data Mr Forrest uses to calculate effective tax rates are reasonable; c. There is often a material difference between the effective rate of corporation tax and the nominal statutory rate of corporation tax; d. The difference between the nominal and effective rates has narrowed over time due to the reduction in the rate of capital allowances, and other factors. 32. The experts have no material disagreements on the measurement of the commonly-used measures of effective tax rates set out by Mr Forrest in his report. Section C - An assessment of the advantages and disadvantages in setting the tax by reference to either the nominal (statutory) rate of tax or the effective rate of tax and how each approach would reflect the economic realities of business 33. With regard to the advantages and disadvantages of setting the tax by reference to either the nominal (statutory) rate of tax or the effective rate of tax, and how each approach would reflect the economic realities of business, the experts agree on the following issues: a. An advantage of setting the tax by reference to the nominal (statutory) rate is that the nominal rate is set out in tax legislation. The nominal rate is therefore a simpler rate to use in the calculation of the special CT rate to apply to restitution interest; b. A disadvantage of using only the nominal (statutory) rate is that the ability to account for surplus ACT, losses, group relief and disclaimed capital allowances can reduce the appropriate marginal rate to apply to the hypothetical receipt of interest to below the nominal statutory rate in a given year. The impact of these reliefs and allowances is an empirical question, and is likely to vary considerably across claimants; c. A disadvantage of the effective rate is that it requires judgements to be made as to the most appropriate data sources and empirical methodologies to use in its calculations. 34. With regard to the advantages and disadvantages of setting the tax by reference to either the nominal (statutory) rate of tax or the effective rate of tax, and how each approach would reflect the economic realities of business, the experts disagree on the following issues: a. The key disagreement between the experts is in whether the effective tax rate measures referred to in Section B above are ever an appropriate rate to use in calculating the appropriate tax rate on restitution interest. b. The extent to which the agreed disadvantage of the nominal (statutory) rate listed in 34b represents a corresponding advantage of the effective rate; c. The scope of restitution interest to result in the generation of new reliefs and allowances (e.g. new capital allowances) and whether this is a disadvantage of using the nominal rate and/or an advantage of using an effective rate; d. The use of active corporation tax management strategies as a disadvantage of using the nominal (statutory) rate and an advantage of using the effective rate; and e. Whether the best estimate of the appropriate tax rate to apply to restitution interest is almost certainly below the rate calculated using nominal statutory tax rates. 35. The respective positions of the experts and the points of disagreement are described in more detail below. 36. Regarding the disagreement in 34a on the appropriateness of using the effective tax rates for identifying the tax due on marginal interest income, Mr Forrest considers that this depends upon the precise formulation of the counterfactual scenario: a. In the counterfactual scenario used by HMRC, and supported by Professor Devereux, there are limited grounds for incorporating business effects, but Professor Devereux agrees that, in principle, taxation effects (such as surplus ACT, losses, group relief and disclaimed capital allowances) should be considered. This means the appropriate overall marginal rate will (save for exceptional circumstances) be somewhere between that calculated using nominal and effective rates. The nominal rate will be an upper bound. Given the magnitude of reliefs and allowances and the timeframe of the claim period, it highly likely that the appropriate marginal rate will be significantly below the upper bound provided by the nominal rate. b. In the broader counterfactual, which Mr Forrest proposed in paragraph 10, it is reasonable to incorporate business effects such as a different level of investment, which would result in different levels of reliefs and allowances. Where the business in the broader counterfactual is scaled up, but the same in all other respects, then it is reasonable to suggest that it would have the same effective tax rate as compared to the actual scenario. This means that the appropriate marginal rate, comparing these two scenarios, is the average effective rate. 37. Regarding the disagreement specified in 34a, Professor Devereux believes that the effective tax rates described in Section B above are not appropriate for identifying the tax due on marginal interest income. Specifically, under the counterfactual that the company hypothetically receives interest, then the appropriate tax rate is a marginal tax rate on interest income. The effective tax rates discussed in Section B are generally average tax rates that take into account a wide range of factors that are irrelevant to interest income. 38. Professor Devereux agrees that, in the calculation of a claimant-specific tax rate, it would in principle be reasonable to take into account issues of surplus ACT, losses, group relief and possibly disclaimed capital allowances. However, he believes that the measures of “effective tax rates” referred to in Section B are not suited to making adjustments for these factors. Instead, they incorporate a wide range of factors that are not relevant. They are therefore not reasonable substitutes for applying the tax in the counterfactual even on a claimant-by-claimant basis, taking into account the details of each claimant's tax position. 39. Professor Devereux believes that, even if the counterfactual proposed by Mr Forrest in paragraph 10 were accepted, the measures of effective tax rates referred to in Section B would still not be applicable. That is partly because such measures depend crucially on the rate of profit earned and are based on average rates of profit. But additional investment undertaken by a claimant would not generally have had access to such high rates of profit. In addition, the measures of effective tax rates are typically of the form of a tax liability expressed as a proportion of before-tax profit. The value of the measure therefore depends crucially on the way before-tax profit is measured. Almost by definition, this is different from the taxable profit of a company. The resulting measure therefore depends crucially on the accounting treatment of the returns to whatever investment the company is deemed to have made. 40. In sum, Professor Devereux believes that the effective tax rates referred to in Section B can never be a reasonable rate to use in the calculation of the appropriate marginal tax rate on hypothetical interest income or the returns on hypothetical marginal investment. 41. Regarding the disagreement specified in 34b, Mr Forrest contends that the disadvantage of the nominal rate listed in 33b corresponds to an advantage of the effective rate. This is because companies use the allowances and reliefs specified in 33b and while the nominal rate does not take into account this reality, the effective rate does. Thus Mr Forrest contends that the effective rate is in this instance better reflective of the taxation realities of business. 42. Professor Devereux disagrees, for the reasons given above in paragraphs 37 to 40. Specifically, he believes that the disadvantage referred to in 33b does not justify the use of an “effective tax rate”. That is because measures of effective tax rates: (a) allow for a wide range of factors other than those referred to in 33b; (b) are unlikely to deal properly with the specific issues mentioned in 33b; and (c) measure average, rather than marginal, tax rates. 43. Regarding the disagreement specified in 34c, Mr Forrest contends that the generation of new allowances and reliefs (e.g. capital allowances) in the counterfactual scenario is a consideration when determining the appropriate marginal rate at which to tax restitution interest. This is because new allowances and reliefs can reduce the appropriate marginal rate applying to restitution interest below the nominal statutory rate in a given year. 44. Mr Forrest argues that in the counterfactual proposed by Professor Devereux, there is scope for restitution interest to result in the generation of new allowances and reliefs in a given year. Even in a counterfactual scenario where the unlawfully paid tax is used to purchase assets that generate a risk-free return, it is appropriate to consider business effects. It is possible for the business to make different decisions, such as investments, in full knowledge of the more valuable assets owned by the company. In short, in any reasonable counterfactual, a firm that owned a large amount of risk free assets would operate differently to one that did not. Mr Forrest considers that, in order to suppose an absence of business effects, it would be necessary to make the further, unreasonable, assumption that the assets in which the unlawfully paid tax is invested are kept secret from the rest of the business. As such, Mr Forrest argues that even under a “narrow” counterfactual, it is appropriate to consider the possibility of the generation of new allowances and reliefs when determining the appropriate marginal rate to apply to restitution interest. 45. Mr Forrest further argues that under the counterfactual proposed in paragraph 10, it is a reasonable contention that new allowances and reliefs would be generated in proportion to the larger business in this counterfactual. This contention requires no assumption regarding the specific use of the unlawfully paid tax. Mr Forrest also notes that under this counterfactual, the effective tax rate will be the same as in the actual case, and therefore the marginal rate (the difference between the counterfactual and actual scenarios) at which to tax restitution interest would be exactly equal to the effective average tax rate in each year, and thus a blended effective tax rate applying over the whole claim period would be the most appropriate tax rate to levy on restitution interest. 46. Mr Forrest argues, therefore, that it is a disadvantage of setting the special CT rate with reference to the nominal (statutory) rate that such a rate does not take into account the generation of new allowances and reliefs. Mr Forrest notes that this is an advantage of the effective rate, which in this instance is better reflective of the economic realities of business. 47 Professor Devereux disagrees strongly with this analysis. Under the counterfactual used by the court to identify the amount of the restitution interest, and also used by HMRC, it is assumed that the business earns additional interest, but that the other activities of the business are unaffected. In this case, any allowances and reliefs that are generally available to businesses for their normal investment activity are simply irrelevant. Even if the counterfactual proposed by Mr Forrest were accepted, then the use of effective tax rates referred to in Section B would still not be reasonable, for reasons set out in paragraph 39. 48. Regarding the disagreement in 34d, and as he argues in section 5.7d of his report, Mr Forrest notes that taking lawful steps to reduce tax burdens are part of the ordinary course of business for large UK companies. In this regard, the nominal (statutory) rate misses an important and relevant factor, which is captured by the use of an effective tax rate. in this regard, the effective rate is more reflective of the economic realities of business. 49. Professor Devereux disagrees, and can see no reason why the claims of Mr Forrest in this regard might be true. If Mr Forrest is referring to, say, the possibility that a multinational company may shift profit overseas, then that would reduce both the tax liability and measured pre-tax profit; the effective tax rate may rise, fall or even stay the same as a result. If such tax planning is a relevant factor, then in Professor Devereux's view, it is another argument against using effective tax rates. 50. Regarding the disagreement in 34e, Mr Forrest argues that, because many companies benefitted from allowances and reliefs during certain years of the claim period, it is the case that the appropriate tax rate to apply to restitution interest lies below that calculated with reference to the nominal (statutory) rate. This is the case even under the “narrow” counterfactual used by Professor Devereux. 51. Professor Devereux agrees that, for companies that were in a position of surplus ACT, losses, group relief and possibly disclaimed capital allowances, the nominal statutory tax rate would generally lie above the relevant marginal tax rate. But he does not accept that this is “because many companies benefitted from allowances and reliefs”. 52. This statement accurately sets out the issues on which we are agreed, the issues on which we disagree and, where appropriate, the reasons for our disagreement. (signed) Nicholas Forrest Date: 12/12/2016 (signed) Professor Michael Devereux Date: December 12, 2016 APPENDIX C Extracts from the European Convention on Human Rights (“ECHR”) and the Charter of Fundamental Rights of the European Union (“CFREU”) ECHR Protocol Paris, 20.III.1952 Article 1 Protection of property Every natural or legal person is entitled to the peaceful enjoyment of his possessions. No one shall be deprived of his possessions except in the public interest and subject to the conditions provided for by law and by the general principles of international law. The preceding provisions shall not, however, in any way impair the right of a State to enforce such laws as it deems necessary to control the use of property in accordance with the general interest or to secure the payment of taxes or other contributions or penalties. ARTICLE 6 Right to a fair trial 1. In the determination of his civil rights and obligations or of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law. Judgment shall be pronounced publicly but the press and public may be excluded from all or part of the trial in the interests of morals, public order or national security in a democratic society, where the interests of juveniles or the protection of the private life of the parties so require, or to the extent strictly necessary in the opinion of the court in special circumstances where publicity would prejudice the interests of justice. 2. Everyone charged with a criminal offence shall be presumed innocent until proved guilty according to law. 3. Everyone charged with a criminal offence has the following minimum rights: (a) to be informed promptly, in a language which he understands and in detail, of the nature and cause of the accusation against him; (b) to have adequate time and facilities for the preparation of his defence; (c) to defend himself in person or through legal assistance of his own choosing or, if he has not sufficient means to pay for legal assistance, to be given it free when the interests of justice so require; (d) to examine or have examined witnesses against him and to obtain the attendance and examination of witnesses on his behalf under the same conditions as witnesses against him; (e) to have the free assistance of an interpreter if he cannot understand or speak the language used in court. CFREU Article 17 Right to property 1. Everyone has the right to own, use, dispose of and bequeath his or her lawfully acquired possessions. No one may be deprived of his or her possessions, except in the public interest and in the cases and under the conditions provided for by law, subject to fair compensation being paid in good time for their loss. The use of property may be regulated by law in so far as is necessary for the general interest. 2. Intellectual property shall be protected. Article 47 Right to an effective remedy and to a fair trial Everyone whose rights and freedoms guaranteed by the law of the Union are violated has the right to an effective remedy before a tribunal in compliance with the conditions laid down in this Article. Everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal previously established by law. Everyone shall have the possibility of being advised, defended and represented. Legal aid shall be made available to those who lack sufficient resources in so far as such aid is necessary to ensure effective access to justice. [1] Amministrazione delle Finanze dello Stato v SpA San Giorgio (Case 199/82) [1983] ECR 3595 ,

[1985] 2 CMLR 658

. A San Giorgio claim is broadly one for repayment of taxes levied in breach of EU law. [2] Brasserie du Pêcheur SA v Germany; R v Secretary of State for Transport, ex p Factortame Ltd (No 4) (Joined cases C-46/93 and C-48/93)

[1996] QB 404

[1996] ECR I-1029

[3] This was the effect of s 239(2) of the Income and Corporation Taxes Act 1988:

“The amount of ACT to be set against a company’s liability for any accounting period under s 239(1) shall not exceed the amount of ACT that would have been payable … in respect of a distribution made at the end of that period which, together with the ACT so payable in respect of it, is equal to the company’s profits charged to corporation tax for that period.”

In his first report, at Appendix 8, para 7, Professor Devereux describes the effect in the following way: “In general, the limit for set-off is equal to the rate of ACT multiplied by UK taxable profit. A rise in UK taxable profit due to the hypothetical receipt of interest would raise the limit for the ACT set-off. For example, if the ACT rate were 35% (or 35/65 of the cash dividend), and taxable profit rose by £100, then the company would have been able to set an additional £35 of ACT against the corporation tax charge.” [4] The ECJ in Littlewoods Retail Ltd v Revenue and Customs Commissioners (

Case C-591/10

)

[2012] STC 1714

[5] [Original footnote]: See, among other authorities, the Gasus Dosier- und Fördertechnik v. Netherlands , op. cit. , para. 62. [6]

(1986) 8 EHRR 123

[7]

[2014] EWHC 4302 (Ch)

, Paragraph 450 [8] The reason why Mr Forrest considers the risk-free rate is an appropriate rate for translating historical amounts of overpaid tax into compensating amounts at the date of the award is because the claimant has been denied access to the overpaid tax over the period of the claim, but these amounts have not been exposed to business risk and therefore the rate of compensation only needs to incorporate inflation and real time preference, without any risk premium. The risk-free rate, which can be measured from the yields on government bonds, is a good proxy for inflation and real time preference. [9] Mr Forrest notes the evidence of Mr Wadey which, in his role as BAT group treasurer, addressed the question of what BAT would have done with the unpaid tax from a treasury perspective. He suggests the company could have repaid borrowings with the unlawfully paid tax. Mr Forrest considers this is one possible way the additional post-tax revenues could have been used, but there are alternative economic uses, consistent with the company being larger in the counterfactual scenario. [10] This is generally known as the “new view” of dividend taxation, though it was first set out in 1979. [11] There can be an effect if the tax rate changes over time, which would occur when a company moves into, or out of, a position of surplus ACT; in these cases, the required rate of return can rise or fall depending on the tax position. [12] Witness statement of Martyn Rounding, Paragraph 16 [13] Take, for example, four claimants with one large claim spanning many years and three smaller claims covering a more recent time period. If the larger claim is given an even greater weight then the revised weighted average will be closer to the appropriate individual rate for the large claim and further away from the appropriate individual rates for the 3 small claimants. Such distribution effects are also important in the consideration of averaging techniques.

BAT Industries PLC and Others v Revenue and Customs (INCOME TAX/CORPORATION TAX : European law) [2017] UKFTT 558 (TC)

Cited in 2 later judgments