Birmingham Hippodrome Theatre Trust Ltd v The Commissioners for HM Revenue and Customs: [2013]UKUT 057 (TCC) [2013] UKUT 057 (TCC)

UPPER TRIBUNAL
TAX AND CHANCERY CHAMBER
[2013] UKUT 057 (TCC)Case No FTC/40/2011
Between
Birmingham Hippodrome Theatre Trust LtdAppellantThe Commissioners for HM Revenue and Customs: [2013]UKUT 057Respondent
Before
MRS JUSTICE PROUDMAN DBEMR CHARLES HELLIER (TRIBUNAL JUDGES)Philippa Whipple QC and Brendan McGurk (instructed by the General Counsel and Solicitor to HM Revenue and Customs) for RespondentsDate 14 February 2013Category: Tax
[1]Birmingham Hippodrome Theatre Trust Ltd (the “Theatre”), represented by Mr Milne QC and Mr Yates, appeals against the decision of the Commissioners for Her Majesty’s Revenue and Customs (“HMRC”, represented 5 by Mrs Whipple and Mr McGurk) not to make a repayment of overpaid output tax. That decision was based on two grounds: first that such a repayment would give rise to a result contrary to the provisions of the Sixth Directive (HMRC described the claim as “abusive”), and secondly because that was the effect of the set-off provisions of s. 81(3A) (“s. 81 10 (3A)”) of the Value Added Tax Act 1994 (“VATA”).[2]These issues arose because for a number of VAT periods between 1990 and 2004 the Theatre had accounted for output tax on supplies which should have been treated as exempt but between 2000 and 2001 had received a repayment of input tax, which, because its outputs should have been treated as exempt, should not have been made. 15 In 2006 the Theatre made a claim for the repayment of its overpaid output tax for those VAT periods in which its claim was not barred by time limits. Those VAT periods happened to fall between 1990 and late 1996. But HMRC were, by that stage, out of time to claim repayment of the input tax wrongly repaid. The repayments exceeded the overpaid output tax. HMRC sought to deny the claim on the alternative 20 bases set out above.[3]The history of the legislative provisions and the facts are admirably set out in the decision of the First tier Tribunal ("FTT"). There would be no advantage in gilding that lily so we treat paragraphs 1-35 and 46-51 of the decision as incorporated here. 3[4]The FTT dismissed the appeal holding that(i) there was "not a generally applicable principle that taxpayer claims that happened to produce an advantageous result for the taxpayer constitute ‘abusive practices’" which would permit the claim to be denied; but that(ii) s. 81 (3A) did entitle HMRC to offset the input tax erroneously repaid for 2000 to 2001 against the claim in respect of 5 1990 to 1996 so as to reduce the amount payable to the Theatre to nil. Terminology.[5]HMRC characterised the Theatre's claim as "abusive". By this they did not mean 10 that it was fraudulent or that it was abusive in the sense used in Halifax & others v HMCE Case C-255/02 [2006] STC 919 (they accept that it was not), but that it was an opportune claim whose satisfaction would result in a state of affairs which was not consistent with the outcome intended by the Sixth Directive (“the Directive”).[6]In order to avoid the baggage which normally accompanies the word "abusive" 15 we have not used it in this decision, but instead we have referred to HMRC's argument as being that the claim was “incongruent” with the Directive: the issue being whether such incongruity must cause the Theatre’s claim to fail.[7]Both Mrs Whipple and Mr. Milne referred to the doctrine of "effectiveness", but they meant somewhat different things by it. Mr. Milne used the word to describe the 20 one-sided principle described in the following passage by the European Court of Justice in Edilizia Industriale Siderurgica Srl v. Ministero dele Finanze (“Edis”) (Case C-231/96) [1998] ECR I-4979 at [19]: 4 "…it is for the domestic legal system of each Member State to designate the courts and tribunals having jurisdiction and to lay down the detailed procedural rules governing actions for safeguarding the rights which individuals derive from Community law, provided, first that such rules are not less favourable than those governing similar domestic actions ( principle of equivalence) 5 and, secondly, that they do not render virtually impossible or excessively difficult the exercise of rights conferred by Community law (principle of effectiveness)…”[8]The court is there concerned with the rights which individuals draw from the 10 Directive; not the rights of the member state.[9]Mrs Whipple, on the other hand, used it to mean a two-edged principle on which both the State and the taxpayer could rely. It was used this way, at least at one stage, in the judgment of the ECJ in Amministrazione dell’Economia e della Finanze v Fallimento Olimpiclub Srl [2009] ECR 1-750. 15 10. In most of the reported cases however the principle of effectiveness has been used to describe the one-sided principle available to protect the taxpayer's rights under Community law. To avoid confusion we use the term in that sense only; and we use "conforming application" to refer to the principle Mrs Whipple advances that the Directive should be given full effect for the benefit of the state. 20 S. 81 (3A).[11]The subsection provides:
"(3A) Where- (a) the Commissioners are liable to pay or repay any amount to any 25 person under this Act, 5 (b) that amount falls to be paid or repaid in consequence of a mistake previously made about whether or what or to what extent amounts were payable under this Act to or by that person, and (c) by reason of that mistake a liability of that person to pay a sum by way of VAT, penalty, interest or surcharge was not 5 assessed, was not enforced or was not satisfied, any limitation on the time within which the Commissioners are entitled to take steps for recovering that sum shall be disregarded in determining whether that sum is required by subsection (3) above to be set against the amount mentioned in 10 paragraph (a) above."
[12]We start with a consideration of this section. In argument Mrs Whipple put the emphasis of HMRC’s case principally on the basis of the incongruity argument, and on the basis of s. 81 (3A) in the alternative. We think it is better to start with s. 81 15 (3A). Indeed it seems to us that if, properly construed, s. 81 (3A) has the result that the Theatre has no right to repayment, then there is no room in this appeal for the application of HMRC's incongruity argument: if the result intended by the Directive was that the Theatre should have no claim, and the effect of the domestic legislation is that the Theatre has no claim, the Directive is given full effect by the domestic 20 legislation and the result of the claim is not incongruent with the Directive.[13]Of course it is possible (because the Directive does not contain the time limits found in domestic legislation) that the effect of s. 81 (3A) and domestic time limits might be to reduce the Theatre’s Claim to nil, whereas the outcome envisaged by the Directive might be that the taxpayer was liable to make a payment to HMRC. But 25 HMRC did not suggest, and we do not believe, that any principle of conforming application would have the effect that domestic time limits permissibly imposed should be set aside. 6[14]Mrs Whipple said that each of the conditions in s. 81 (3A) were satisfied (or absent any conforming application argument would be satisfied):(a) the Commissioners were liable to repay overpaid tax on the Theatre’s supplies of ticket sales from 1 January 1990 to November 1996;(b) that liability arose out of a mistake: that mistake was to treat 5 the Theatre’s supplies of theatre services as taxable, when they were exempt; and(c) by reason of the same mistake the Theatre deducted input tax for the period between January 2000 and November 2001 when it could only have made a deduction on the footing that its outputs were taxable. 10[15]Mrs Whipple does not concede that condition (a) is satisfied: she says that the incongruity argument means that HMRC were not "liable to pay or repay any amount". It seems to us that this approach mistakes mechanism for result: the result Mrs Whipple seeks is a nil net liability; the mechanism through which the UK 15 legislation might achieve that is by setting against a liability of HMRC a resurrected right of HMRC against the taxpayer. What matters for the purpose of the incongruity argument must be the result obtained by the domestic legislation, not the method by which it is obtained.[16]Five issues arose in relation to the provisions of s. 81 (3A):(1) whether HMRC 20 may pick and choose between past out of time periods;(2) whether the set-off should be limited to amounts connected in some way to the claim for repayment;(3) time limits;(4) whether there had been one mistake or two; and(5) whether there should have been a transitional period before s. 81 (3A) came into force. (1) Picking and choosing 25 17. This concerns the extent to which HMRC are permitted under s. 81 (3A) to pick and choose between out of time periods. For example if a taxpayer had, by virtue of 7 the same mistake, (a) an in time claim for the current year, (b) an out of time claim for year X and (c) an out-of-time repayment liability for year Y, does s. 81 (3A) permit HMRC to reduce the amount otherwise payable to the taxpayer by the repayment liability for year Y without taking into account the out of time claim of the taxpayer 5 for year X?[18]HMRC say that s. 81 (3A) does not operate in this way. Properly read it requires all the consequences of the mistake to be taken into account "by restoring the tax situation to that which would have prevailed had the mistake as to the treatment of supplies not been made". It thus requires a global calculation not a selective one. This 10 is the way HMRC say they operate s. 81 (3A) in practice.[19]Mr. Milne says that there is nothing in the words of s. 81 (3A) which requires this result: if the section permits any out-of-time period to be resurrected for the purposes of the set-off then HMRC can pick and choose. The fact that, as an administrative matter, HMRC does not do so is not sufficient to avoid a breach of Community law 15 (Mulligan v Minister of Agriculture Case C-313/99 [2012] ECR 1-5719 at [47]). He says that any ability of HMRC to pick and choose breaches the principles of effectiveness (since it undermines the taxpayer’s right to recover), legal certainty (because it leaves the taxpayer exposed in relation to old periods), and equality (since the taxpayer and the State are treated differently without justification). 20 8 (2) Whether the set-off should be limited to amounts connected in some way to the claim for repayment[20]HMRC say that s. 81 (3A) merely requires those amounts which would have arisen by virtue of mistake to be brought into account.[21]Before the FTT the Theatre had advanced the argument 5 that s. 81 (3A) was limited to set-off only within VAT accounting periods. That was certainly how HMRC's original internal guidance had originally viewed the provision: "Assuming the normal three-year assessment provisions apply, we are advised that the claimed refunds cannot be set off against the under declarations if they 10 are in different accounting periods. For example, if a refund claim has been submitted for the periods 02/96 and 05/96, but under-declarations have been discovered for the periods 11/95 and 02/96, we can only set off the underdeclaration against the over-declaration for the period 02/96. The under declaration in 11/95 would escape." 15[22]The FTT rejected that argument and Mr. Milne did not pursue it before us. Instead he advanced the argument that the set-off could only be between amounts which were relevantly connected: that is to say where one was a cost component of the other. Mr. Milne says that the Directive and Community case law make it a 20 fundamental condition for the deduction of input tax that there be a direct and immediate connection between the related input and a taxable output, or in other words that the input be a cost component of the output: Article 2 of the First Directive provided that tax should be chargeable "after the deduction of the amount of value added tax borne by the various cost components."; Article 17 of the (Sixth) Directive 25 provided for the right to deduct "in so far as goods and services are used for the purpose of [the taxpayer’s] taxable transactions"; and the ECJ case law had 9 emphasised the necessity of a direct and immediate link, (see for example Skatteverket v AB SKF Case C-29/08 [2010] STC 419 at [57–60], among others).[23]Against that background Mr Milne says that unassessed VAT which arose by virtue of the "same mistake” must be limited to VAT attributable to mistakes referable to transactions one of which was the cost component of the other. 5 That approach was supported by the original version of HMRC's internal guidance: "10.4. Section 81 (3A) was introduced with the specific intention of allowing directly connected input tax and output tax to be set off against each other ..." 10 and by the revised guidance: "however, section 81 (3A) brings into the equation, for example, input tax that was applicable to the supplies in respect of which output tax is being claimed where it was deducted as a result of the same mistake ..."; 15 and by the FTT in Laing The Jeweller Limited v HMRC VAT Decision 18841 where the FTT applied the s. 81 (3A) set-off between accounting periods, noting it applied to "a closely connected cross liability".[24]Mr Milne says that the fact that reciprocity between elements claimed and reclaimed was the intention of Parliament in enacting s. 81 (3A) was also clear from 20 the Explanatory Notes to the Finance Bill 1997. Those notes were an accepted aid to interpretation of the statute (see below). In the Explanatory Notes (set out in [49] of the FTT decision) at paragraph 5 it is said: "Previously, where a taxpayer has overpaid tax, but has as a consequence simultaneously overclaimed input tax, Customs cannot set the overclaimed input 25 tax against the refund due if they are out of time to issue an assessment for the overclaimed input tax. For example if supplies have been charged with tax, and 10 the supplies are subsequently ruled to be exempt, Customs can reduce the repayment to take account of input tax originally attributable to the supplies which is now no longer claimable. However litigation can sometimes take a long time to settle and the limits for making an assessment can sometimes expire ..." 5 (3) Time Limits.[25]The effect of HMRC's interpretation of s. 81 (3A) is to permit set-off of amounts which arise in earlier accounting periods without limit of time. Thus HMRC could go back to the beginning of recorded VAT time (1973) to dredge up amounts which 10 could be offset.[26]This, said Mr. Milne, was contrary to the three fundamental Community law principles of legal certainty, effectiveness and equality.[27]First, it was contrary to the principle of legal certainty. Taxpayers are required to keep records for only six years: if HMRC dredge up a claim from 1973 a taxpayer 15 would have no realistic prospect of defending it. The result would be "a state of uncertainty as regards their rights in an area governed by Community law" (Mulligan at [47]) which was proscribed by the principle of legal certainty.[28]For the same reason it offended the principle of effectiveness, namely that domestic rules must not "render virtually impossible or excessively difficult the 20 exercise of rights conferred by Community law". That argument, it seemed to us, had additional force when it is recalled that before UK tribunals the onus is generally on the taxpayer to prove his claim, and an almost certain lack of records from very old periods would make defending a set-off in respect of such periods contended for by HMRC excessively difficult. 11[29]And it was contrary to the principle of equality - that “similar situations should not be treated differently unless differentiation is objectively justified”, (see [30] of the ECJ’s judgment in Societe Financiere D’Investissements SPTL v Belgium Case C- 85/97 [2000] STC 164 (“SFI”)) - because, whilst the taxpayer was prevented by the domestic time limit in s. 80 from going back more than three 5 years, s. 81 (3A) permitted the State to go back to 1973. Whilst it might be argued that some form of extended limitation period for set-off in response to a taxpayer claim for a refund might be justified, the lack of any time limits whatsoever could not be justified by reference to the difference between the position of the tax authority and the taxpayer: 10 see Ecotrade Spa v Agenzia della Entrate Case C-95/07 [2008] STC 2626 and SFI. The effect of such an interpretation of s. 81 (3A) was, said Mr. Milne, grossly disproportionate.[30]Time limits on the actions of both the taxpayer and the State were, Mr. Milne said, not only permitted by Community law (see for example Ecotrade) but actually 15 required. In Alstom Power Hydro v Valsts ienemumu dienests Case C-472/08 [2010] STC 777, the ECJ stated at [16] that the principle of legal certainty:
"requires the tax position of the taxable person, having regard to his rights and obligations vis-à-vis the tax authority, not to be open to challenge indefinitely (Ecotrade SpA v Agencia della Entrate –Ufficio di Genova 3 C-95/07 and C- 20 96/09 [2008] STC 2626, [2008] ECR I-3457 paragraph 44)."
[31]Likewise in Ze Fu Fleishhandel GmbH Case C-201/10 (5 May 2011) the ECJ had held, in a situation in which a directive provided a time limit but also an option for a member State to extend the limit, that an extension to 30 years was precluded by the 25 principle of proportionality. 12[32]Mr Milne submitted that some reasonable time limit was required. That must be provided by a requirement to limit the application of s. 81 (3A) to inputs and outputs between which there was a direct and immediate connection. Whereas in most cases an input preceded the output of which it was a cost component it would be only by a short period; the capital goods scheme envisaged a 10 year 5 period at most; if exceptionally an input was incurred after an output of which it was a cost component, the period would be short and the taxpayer would be in a position to defend the input. Where one was a cost component of the other, a reasonable time limit was required by the principle of legal certainty. 1033. Mrs Whipple responds to these arguments thus. She says that the principle of effectiveness is about protecting the rights of the taxpayer under Community law. The right to be protected must therefore first be identified. S. 81 (3A) does not impinge on the right of a taxpayer to make claims for input tax or to account properly for output tax in current periods; it affects only claims for the adjustment of previous mistakes. 15 The taxpayer’s Community right is to be subject to tax in accordance with the Directive - that is not a right limited to the particular period in which a particular claim arises but is a broader right which takes into account all relevant periods in which a mistake was made. HMRC's interpretation of s. 81 (3A) secures that right effectively. 2034. She says that the principle of legal certainty does not require the imposition of time limits. The offsetting in s. 81 (3A) is there to ensure an effect which is congruous with the Directive. In this circumstance the principle of effectiveness (congruity) trumps legal certainty and requires that there be no time limit. So far as the principle 13 of equality is concerned she says that the taxpayer and the state are not in the same position. (4) One mistake or two?[35]S. 81 (3A) limits the set-off to liabilities (or amounts which 5 would but for a time limit have been liabilities) which arose by reason of the same mistake as gave rise to HMRC’s liability to pay a claim. The question debated before us was whether in the Theatre’s case there had been one mistake or two:(1) there were two mistakes if Parliament’s failure to implement Article 13A 10 (1)(n) was different from HMRC’s mistaken interpretation of the voluntary management condition in Note 2 (c) to Group 13 (in Notice 701/47); and(2) there was one mistake if the failure to exempt cultural services and the administrative failure in relation to Note (2) (c) were regarded as a single failure. 15 36. If there was only one mistake then, on HMRC’s approach to s. 81 (3A), the repayments could be set against the output tax claimed reducing the net claim to nil.

(n) 14

[37]It was clear, and neither party contended otherwise, that the terms of the legislation in Group 13 which from 1 June 1996 exempted cultural services complied with the Directive. In Note (2) (c) of that legislation the UK had taken advantage of the option provided for by the second indent of Article 13 (2) (a), and the terms in which it did so, in Note (2) (c), complied with that Article. The problem 5 was that for a period after 1996 the UK's administrative practice did not comply.[38]It was also common ground that the adoption of domestic legislation correctly implementing a directive does not exhaust the effects of that directive. That was made plain by the ECJ in Marks and Spencer Ltd v Customs and Excise Case C-62/00 10 [2002] STC 1036 (“M&S 2002”) at [27]:
"27. Consequently, the adoption of national measures correctly implementing a directive does not exhaust the effects of the directive. Member States remain bound actually to ensure full application of the directive even after the adoption of those measures. Individuals are therefore entitled to rely in the national courts, 15 against the State, on the provisions of the directive which appear, so far as their subject matter is concerned, to be unconditional and sufficiently precise whenever the full application of the directive is not in fact secured, that is to say, not only where the directive has not been implemented or has been implemented incorrectly, but also where the national measures correctly implementing the 20 directive are not being applied in such a way as to achieve the results sought by it."
[39]We address the parties’ arguments on this question in more detail later in this decision. 25 (5) Transitional periods.[40]S. 81 (3A) was introduced by s. 48 (1) Finance Act 1997 at the same time as the three-year cap. It was stated to apply: 15 "for determining the amount of any payment or repayment by the Commissioners on or after [18 July 1996] including a payment or repayment in respect of a liability arising before that date."[41]In their skeleton argument Mr. Milne and Mr Yates suggested 5 [at 42] that if s. 81 (3A) applies to require set-off across accounting periods, its implementation without any transitional period in which accrued claims could be made would "give rise to the same concerns about breaching the principle of effectiveness as occurred in Fleming and Conde Nast [Fleming t/a Bodycraft v. HMRC and Conde Nast v. HMRC [2008] 10 STC 324] ... in relation to the introduction of the three-year cap."[42]Mrs Whipple says that s. 81 (3A) does not remove any Community law right: it merely removes or reduces a domestic right by making it subject to set-off. The rights considered in M&S 2002 were removed; here the right under the Directive to pay the proper amount of tax (on her interpretation of s. 81 (3A)) is preserved. 15 Discussion[43]We start by considering the plain words of s. 81 (3A), setting aside so far as possible its context in the implementation of the VAT directives and considerations of Community law. 20 The wording of s. 81 (3A).[44]The words do not to our minds require any restriction of the periods in which the liabilities or amounts payable arise: the subsection applies in relation to "any amount" 16 HMRC are liable to repay, and "a" liability to pay "a" sum which was not assessed, enforced or satisfied; it permits the offset of these amounts under s. 81 (3) to be determined without regard to time limits; and there is nothing in s. 81 (3) which confines offsets to amounts arising in the same period or by reference to cost 5 components. Explanatory Notes.[45]Both parties referred us to the Explanatory Notes published prior to the enactment of the Finance Act 1997. In Westminster City Council v National Asylum Support 10 Service [2002] 1 WLR 2956 Lord Steyn explained the function of Explanatory Notes in interpreting the provisions to which they related:
"4. In 1999 a new system was introduced. It involves publishing Explanatory Notes alongside the majority of public bills introduced in either Houses of Parliament by a Government minister ... the texts of such notes are prepared by 15 the Government department responsible for the legislation. The Explanatory Notes do not form part of the bill, are not endorsed by Parliament and cannot be amended by Parliament. The notes are intended to be neutral in political tone: they aim to explain the effect of the text and not justify it. The purpose is to help the reader to get his bearings and to ease the task of assimilating the law. ... 20 “5. The question is whether in aid of the interpretation of the statute the court may take into account the Explanatory Notes and, and if so, to what extent. The starting point is that language in all legal texts conveys meaning according to the 17 circumstances in which it can be used. ... It is therefore wrong to say that the court may only resort to evidence of the contextual scene when an ambiguity has arisen ... Again, there is no need to establish an ambiguity before taking into account the objective circumstances to which the language relates. Applied to the subject under consideration the result is as follows. In so far 5 as the Explanatory Notes cast light on the objective setting or contextual scene of the statute, and the mischief at which it is aimed, such materials are therefore always admissible as aids to construction. They may be admitted for what logical value they have. "6. If exceptionally there is found in Explanatory Notes a clear assurance by the 10 executive to Parliament about the meaning of a clause, or the circumstances in which a power will not be used, that assurance may in principle be admitted against the executive in proceedings in which the executive places a contrary contention before the court. This reflects the actual decision in Pepper v Hart [1993] AC 593. What is impermissible is to treat the wishes and desires of the 15 Government about the scope of the statutory language as reflecting the will of Parliament. The aims of the Government in respect of the meaning of the clauses as revealed in the Explanatory Notes cannot be attributed to Parliament. The object is to see what is the intention expressed by the words enacted."
[46]Paragraph 5 of the Notes explains the mischief at which the provision is aimed. 20 The second sentence gives the example of input tax attributable to a supply becoming uncreditable because the supply has become exempt. There will be many cases where, for example, goods acquired in one accounting period will be supplied in the next. This example does not suggest that the mischief is limited to within-period set-off. It 18 does however provide an example where the link is between a cost component and the supply.[47]The first sentence of the same paragraph provides the greatest support for some limitation on the periods from which liability to pay and a right to repayment may be drawn - the mischief is that the taxpayer has "simultaneously" 5 with his overpayment overclaimed input tax. At first sight that strongly suggests that the two mistakes must happen at the same time - that is to say in the same VAT return. But that word too must be read in the context of the Note as a whole. The context of the example discussed above, the first sentence of paragraph 6:
"the taxpayer will be back in the 10 position he would have been if the mistake had not occurred”, and of paragraph 1: "all the consequences of the earlier mistakes are taken into account before any repayment is made" (added emphasis), all suggest that the mischief was not an inability to set off within an accounting period but a wider problem. "
Simultaneous" must mean, as Mrs Whipple suggested, by the same mistake not at the same time. Thus we do not draw 15 from the Notes the conclusion that the mischief at which s.81 (3A) is aimed is inability to net within a period.[48]Nor do we see in the Notes, or in particular in the example in paragraph 6, a clear indication of mischief limited to the link between a particular input and the output of which it is a cost component. That is an example, but "all the consequences" of an 20 earlier mistake are not so limited.[49]Finally we note that there is no clear assurance in the Notes given by the executive to Parliament about the meaning of clauses within the comments Lord Steyn makes in the quotation above. 19[50]Thus we find nothing in the Explanatory Notes which moves us to consider that the intention or meaning of that provision is, as regards the VAT periods in relation to which s.81 (3A) operates, other than that which its words convey. 5 Statutory context.[51]Nor do we find anything in the statutory context relevant to matters within s.81 (3A) which suggests a different interpretation.[52]S. 25 (2) VATA permits the deduction of input tax against output tax at the end of each VAT period. If there is an excess of input tax it is payable by HMRC to the 10 taxpayer; s. 25 (3) defines a "VAT credit" as the amount so due.[53]S. 73 provides for assessment by HMRC:(i) where they consider VAT returns incomplete or inaccurate; and(ii) where for a VAT period a repayment of VAT or a VAT credit has been paid which ought not to have been. 15 54. In the case of (ii) the section provides that HMRC assess the amount "as being VAT due" for the period. We return to this point below. Subsection (6) provided time limits for such assessments which were normally two years from the end of the VAT period.[55]For claims made after 26 May 2005 subsections (1), (1A), (1B) and (2A) of s. 80 20 provide, on the making of a claim, for: 20(i) credit for output tax which was not due but had been accounted for (subsection (1)) or had been assessed (subsection (1A)); and(ii) the repayment of any such credit after setting “any sums against it under or by virtue of" the Act (subsection 80 (2A)). Subsection (2A) was inserted with effect for any claim made after 25 May 2005 whenever 5 the claim arose;(iii) repayment of amounts which were paid to HMRC by way of VAT which, otherwise than as a result of (i) or of failing to deduct input tax were not VAT due (subsection (1B).[56]S. 81 (3) provides for the set-off of any sum due by way of VAT from a taxpayer 10 against any amount due from HMRC under the Act.[57]Before the FTT HMRC had argued that an interpretation of s. 81 (3A) which restricted its operation to offset within a VAT period made it redundant in view of the offset in s. 80 (2A) which clearly permitted offset in the same accounting period. Before us HMRC say that s. 81 (3A) is "plainly intended to apply to sums which are 15 now out of time to assess and that can only mean sums which arise in accounting periods different from those of" the claim.[58]The Theatre says that the meaning of s. 81 (3A) should not be analysed by reference to s. 80 (2A) since that subsection was enacted in 2005, whereas s. 81 (3A) was inserted in 1997; and that in any event the two subsections have different 20 functions.[59]It seems to us that, since s. 80 (2A) was enacted to have application in relation to claims made after 25 May 2005, and the Theatre's claim was made in 2006, it is part 21 of the statutory setting for consideration of the application of s. 81 (3A) to the Theatre’s claim. But subsections 80 (2A), 81 (3) and 81 (3A) all perform different functions. S. 80 (2A) transforms the rights to "credits" given by s. 80 (1) and 80 (1A) into rights to repayment after setting off sums under the Act; s. 81 (3) is a wider setoff of sums due under the Act against any sums due from HMRC. 5 That will include sums due from HMRC as a result of the operation of s. 80 (2A) but also sums due from HMRC under s. 80 (1B) and as the result of a VAT return showing an excess of input over output tax. S. 80 (1B), dealing with repayment, contains no intrinsic setoff. S. 81 (3A) operates only for the purpose of s. 81 3): thus a sum which is time10 barred cannot be brought into the set-off under 80 (2A) but can, if s. 81 (3A) is satisfied, be brought into the wider s. 81 (3) set-off. Subsections 80 (2A) and 81 (3A) perform different functions in relation to sums of different natures (credits and repayments). We can see nothing in this matrix which requires any particular construction of s. 81 (3A). 15 Sunningdale and HMRC's first internal interpretation.[60]On 14 May 1997 the London VAT Tribunal gave its decision in Sunningdale Golf Club v Customs and Excise Commissioners [1997] V & DR 79 (VATD 14,899). The hearing was on 9 April 1997. The case concerned the setting off of claims to 20 repayment of output tax wrongly paid against input tax wrongly credited. The insertion of subsection (3A) into s. 81 was effected by s. 48 Finance Act 1997 which was enacted on 19 March 1997, before that appeal was heard. 22[61]The relevant extract from the earlier version of HMRC's manual is set out in [50] of the FTT’s decision.[62]From this extract it appears that:(1) HMRC considered that s. 81 (3A) was enacted against the background of 5 Sunningdale type claims;(2) HMRC considered that the intention of s. 81 (3A) was to allow "directly connected” input tax and output tax to be offset against one another; and(3) HMRC had been advised that the effect of s. 81 (3A) was limited to inperiod set-off. 10 63. So far as the last of these points is concerned, it is of little assistance to us: the reasoning for that advice is not explained; it simply indicates that someone knowledgeable has taken that view; we should consider carefully whether it was correct.[64]The second point is a statement which may evidence the government’s intent in 15 promoting s. 81 (3A); but it is weak evidence of that intention and by no means clear that it was communicated to Parliament. Even if it were, "what is impermissible is to treat the wishes and decisions of the government about the scope of the statutory language as reflecting the will of Parliament" (Lord Steyn at [6] in National Asylum). The most that can be derived from this statement is that permitting the potential out of 20 time netting of such directly connected input and output tax may have been an intention of Parliament in enacting s. 81 (3A). 23[65]So far as the first, Sunningdale, point is concerned, it seems to us to rest on a misunderstanding of Sunningdale. As we understood it the appellant’s argument in that case was that although there was an EU right to recover improperly paid output tax, the club's right to deduction of income tax was a domestic right under VATA which, unlike the domestic obligation to pay output tax, was 5 unaffected by the Directive: the tribunal held that the appellant’s right was only to the net amount. Thus it was not a case which turned on time limits.[66]Thus to our minds the reference to the Sunningdale decision is not helpful in construing s. 81 (3A) and all it shows is that there was a concern which might be 10 derived from the Explanatory Note: that "all the consequences of an earlier mistake be taken into account before payment is made." We find nothing here which affects our approach to the construction of s. 81 (3A).[67]As the FTT record ([51]), the earlier HMRC internal manual statements were replaced in 2009 by VR 8200. That indicated that HMRC then regarded s. 81 (3A) as 15 capable of applying across VAT periods, but that it would not be so applied unless the claim sought an unjustified tax advantage. For reasons similar to those in relation to the third part of the earlier internal material, we find this of very little assistance. Liability to “a sum by way of VAT” 20 68. At this stage we should mention a concern with the words of s. 81 (3A) which taxed us during the hearing. S. 81 (3A) (c) relates to "a sum paid by way of VAT” which was not assessed, enforced or satisfied. 24[69]In this appeal the sum which HMRC seek to set against the Theatre's claim is not the amount of an assessment of a VAT liability (the crystallisation of an existing liability under the Act) but the reclaim of a repayment of VAT (a liability which is not strictly to VAT). As noted above, s. 73 (2) permits HMRC to assess an amount which was paid to any person as a repayment or refund of VAT 5 which ought not to have been paid, and the tailpiece of s. 73 (2) provides that HMRC may assess it "as being VAT due". Thus once so assessed the liability can be treated as “VAT due”.[70]In this case HMRC did not assess the Theatre for the overpaid repayment. They could not do so because s. 73 and s.77 provided a time limit for such an assessment 10 and they were too late. But without an assessment the sum (which otherwise is not a VAT liability) is not deemed "as being VAT due" by s. 73 (2). Thus that deeming did not make the recovery "by way of VAT due" within s. 81 (3A) (c), and as a result there was no time-barred "liability" of the Theatre to make payment. If assessment is required to create a "liability to pay a sum by way of VAT", then HMRC may set off 15 a latent reclaim to the recovery of an overpayment under s. 81 (3A) only if “liability” means "liability or any amount which would have been a liability if it had been assessed".[71]Mrs Whipple says that there is no magic about the form of an assessment. A decision letter informing the Theatre of liability would be enough, and the decision 20 letter in this appeal gave rise to an assessed liability because it notified the Theatre of HMRC's claims.[72]It does not seem to us that Mrs Whipple's argument provides a solution. If assessment could not be made because of s. 73 (6), then any document by which it 25 was purported to be made could not be an assessment. As a result the decision letter did not create a "liability ... by way of VAT due” for the purposes of s. 81 (3A).[73]However, in this context, we find the Explanatory Notes helpful. The mischief at which they show that s. 81 (3A) was aimed was "where a taxpayer has overpaid tax, but has as a consequence simultaneously overclaimed input tax, 5 Customs cannot set the overclaimed input tax against the refund due if they are out of time to issue an assessment of the overclaimed input tax.". Thus the section was intended to operate in the circumstance where an assessment for the tax could not be made. For the section to address that mischief, "liability ... to pay a sum by way of VAT", in (c) needs to be 10 given a meaning which does not require the amount to have been assessed. Without that it would not address the mischief at which it was aimed.[74]We conclude that (c) should be construed so that it has effect as if it included the words "or would have been such a liability had it been assessed in time". 15 The effect of Community law on the interpretation of s. 81 (3A)[75]There is no doubt that the Tribunal must interpret s. 81 (3A) so far as possible in conformity with the purpose of the Directive and principles of Community law: if authority is needed see [9] of the ECJ judgment in Marleasing SA v La Comercial Internacional del Alimentacion SA C-106/89 [1990] ECR 1-4135 and HMRC v IDT 20 Card Services Ireland Ltd [2006] STC 1252. Nor is there any doubt that robust techniques may be used in such interpretation: see Pill LJ in IDT at [144]. But this principle has its limitations: the conforming interpretation must go with the grain of 26 legislation being interpreted: see for example [33] of Lord Nicholls in Ghaidan v. Godin-Mendoza [2004] AC 557 quoted by Arden LJ in IDT at [86].[76]Likewise there is no doubt that individuals may obtain rights under a directive where the relevant provisions are unconditional and sufficiently precise, and that domestic courts must give full effects to those rights: Ursula 5 Becker v Finanzamt Munster-Innenstadt Case C-8/81 [1992] ECR 55.[77]We shall come to the fundamental principles of legal certainty, effectiveness and equality later, but we start by asking what, in the context of the accounting necessary between a taxpayer and the state - the context of s. 81 (3A) - is the intended effect of 10 the Sixth Directive, and what is the scope of any right afforded to an individual by it?[78]It is clear that the Directive has specific purposes in relation to the taxation of specific transactions and that those intentions and purposes confer rights on individuals that those transactions be taxed in accordance with the Directive. But s. 81 (3A) is in the context of accounting for tax payable and repayable. 15 79. The rules governing the exercise of the right to deduct are in Article 18. Article 18 (2) requires the taxable person to effect deduction "in the period the input arises" by subtracting it from the VAT due for a period. If the taxable person does not make a deduction in accordance with Article 18 (2), Article 18 (3) provides a licence for member states to determine the conditions and procedures whereby a taxable person 20 may be authorised to make a deduction he has not made in accordance with the preceding paragraph. In Ecotrade the ECJ held that those conditions could include imposition of a time limit. 27[80]Article 22 (5) provides that the taxable person will pay the net amount of VAT due on submitting its returns.[81]The right to obtain a refund of charges levied by a member state in breach of Community law was the consequence and complement of the rights conferred on individuals by Community provisions: see M&S 2002 at [30]. 5 That right must be exercisable in accordance with the principles of equivalence and effectiveness.[82]Thus the purpose of the Directive in relation to accounting for tax seems to be clear. It is that each taxable person shall be liable to pay the correct net amount of tax when delivering his return. If it pays less than it ought to have paid, then it may be 10 made liable for the excess; if it paid more than it should have paid, it is entitled to a refund under Community law. Because the intention is that it pays the right amount of tax, if it is reimbursed input tax which it should not have been reimbursed the member state must be free to make it liable to refund the reimbursement.[83]In Ecotrade the court held not only that the time limit imposed by the state was 15 permitted under Art 18 (3) but,"44. Furthermore, the possibility of exercising the right to deduct without any temporal limit would be contrary to the principle of legal certainty, which requires the tax position of the taxable person, having regard to his rights and obligations vis-a-vis the tax authority, not to be open to challenge indefinitely. 2045. Consequently the argument that the right to deduct may not be coupled with any limitation cannot be accepted." 28[84]And at [46] it was held that the effect of a limitation period was not incompatible with the Directive so long as the principles of equivalence and effectiveness were not compromised.[85]The effect of time limits is to curtail the objects of the Directive but such curtailment is permissible. The effect of s. 81 (3A) is to modify the 5 effect of otherwise absolute time limits in favour of the state. That modification must be done in a way which does not violate fundamental principles of Community law and is in conformity with the object of the Directive. If s. 81 (3A) permitted the state to pick and choose between out of time periods so that it could choose only those in which the amounts 10 were due to HMRC for the purpose of the set-off the result would not conform to that object. Thus if possible s. 81 (3A) should be construed so as to require all the amounts which would be due to or from HMRC if time limitations were disregarded to be taken into account for the purposes of this setting off.[86]Whilst the turning of a debt due by HMRC into a sum due by the taxpayer would 15 be clearly outside the purpose and intent of s. 81 (3) and (3A) (and could well breach the principle of legal certainty: Ecotrade), the reduction of debt otherwise due from HMRC to the net amount they would owe if the Directive were fully implemented goes with the grain of those provisions. Construing those subsections to that effect may require some violence to their language but is in our view possible. 20 87. The same conclusion obtains if the question is approached as an emanation of the directly enforceable right of an individual under the Directive. That right is to bear no more tax than the Directive requires. The set-off against a within-time claim of an otherwise out-of-time claim by HMRC offends that right if out-of-time claims of the 29 individual are ignored because the individual may then end up paying more than the Directive requires. Thus a limitation of the claims for which the time bar is raised to claims against the taxpayer chosen only by HMRC could make the individual's right under the Directive impossible to exercise. That is cured by applying s. 81 (3) and (3A) so that all otherwise time-barred claims, whether of HMRC or 5 the taxpayer, are taken into account for the purposes of the set-off in s. 81 (3A).[88]At [6] in its judgment in Marleasing the ECJ reiterates that a directive may not impose obligations upon individuals The interpretation of s.81 (3A) proposed in the preceding paragraph does not do that: rather it limits the individual's rights against the 10 state to those intended by the Directive and Community law principles.[89]Having thus set out our understanding of the proper construction of s. 81 (3A) we turn to the remaining issues. Identification: a mistake 1590. The words of s. 81 (3A) hinge on a particular mistake being made. If we are right its effect, properly construed, is to release time limits in relation to all occurrences of that mistake in determining what may be set against the amount otherwise due to the taxpayer. But if the object of the Directive is that a taxpayer pay the right amount tax, should the restriction to errors arising out of the same mistake be deleted? 2091. That does not seem to us to be possible. S. 81 (3A) hinges around a particular mistake. That is a fundamental feature of the provision: see [33] of Lord Nicholls in 30 Ghaidan v. Godin-Mendoza, quoted, as we have said, by Arden LJ in IDT. Extending the ambit of the provisions to matters other than a particular mistake would not go with the grain of the legislation. Accordingly we cannot interpret s. 81 (3A) in that way.[92]It is true that the limitation of the effect of s. 81 (3A) to adjustments 5 by reason of a particular mistake may mean that a taxpayer pays more or less tax than it otherwise should have under the Directive. But that is the result of the operation of time limits from which s. 81 (3A) is an exception. Given that the time limits are not precluded either by the Directive or by EU principles, the loss of rights entailed by this 10 limitation of s. 81 (3A) cannot be precluded either. Time limits[93]Ecotrade, SFI, and M&S 2002 make clear that time limits, so long as they respect the principles of equality, effectiveness and legal certainty are permissible even 15 though their effect is to deny a right or to curtail the obligation under the directive: see also Alstom at [19].[94]If s. 81 (3A) is construed in the manner described above it requires HMRC to consider all previous periods in determining the set-off. Does that unlimited requirement conflict with legal certainty? Should it be subject to some time 20 limitation? If a mistake extended over a significant period, say 20 years, would the effect be that before becoming entitled to reclaim in respect of, say, year X, the taxpayer would have to be able to show that the net result of all the mistakes in 31 previous years was not in his favour? Does that mean that the taxpayer's right to reclaim in respect of year X becomes impossibly difficult to pursue?[95]Mr Milne rightly says that a limitation on the ambit of s. 81 (3A) so that it would only permit set-off where the item to be set off was a cost component of the item giving rise to the claim would avoid the damage to legal certainty 5 caused by the wider interpretation discussed in the preceding paragraphs.[96]But Mr. Milne’s approach strikes a balance which favours legal certainty over the full application of Community law. In Olimpiclub, a case which dealt with the question of whether the legal certainty afforded by the principle of res judicata could 10 take precedence over the application of particular Community law, the Advocate General said:
“[47] It is, nonetheless, also clear from the case law of the Court that the principle of legal certainty - and the finality of decisions, which flows from that principle - is not absolute in the sense that it prevails in every situation: rather it 15 must be reconciled with other values worthy of protection, such as the principles of legality and the primacy of Community law, and the principle of effectiveness.”
That statement was echoed by the court: "[28] it is therefore necessary to determine, specifically, whether the above20 mentioned interpretation of…the Italian Civil Code may be justified with a view to protecting the principle of legal certainty, in the light of its implications for the application of Community law." 32 The court found it was not. The application of legal certainty had to be balanced against the application of Community law.[97]It is clear however, as Olimpiclub shows, that domestic procedural rules, in the context of matters subject to Community law such as VAT, may also be affected by the primacy of Community law. Thus, whilst it could be expected 5 that within the period for which domestic law requires records to be kept a taxpayer should have the onus of showing the amount of net effect of the relevant mistakes, in periods before this the onus of proof should be on the authority alleging otherwise - which would entail showing the net result was in its favour, not that there were only some periods 10 in which that was the case. That approach we believe strikes the right balance between the full implementation of the Directive, effectiveness and legal certainty.[98]Because this approach to procedure could be required, it seems to us that it permits s. 81 (3A) to be construed as operating without the importation of the time limit on the basis that that procedural rule is applied at the same time. For this reason 15 it does not seem to us that legal certainty requires a time limitation which would trump the object intended by the Directive.[99]We should address Alstom at this stage. The case concerned a Latvian domestic provision which imposed a three year time limit on the ability of the taxpayer to reclaim overpaid VAT. The question before the ECJ was whether that was contrary to 20 Article 18 (4) which required either the repayment or carry forward of the excess input tax. The court said: 33 "15. It should, at the outset, be noted that although, as Alstom claims, Article 18 (4) of the Sixth Directive does not expressly lay down such a limitation period, that fact does not in itself permit the conclusion that the provision must be interpreted as meaning that the exercise of the rights to a refund of excess VAT that cannot be subject to 5 a limitation period. “16.First, by analogy with the situation applicable to the exercise of the right to deduct, the possibility of making an application for the refund of excess VAT without any temporal limit would be contrary to the principle of legal certainty, which requires the tax position of the taxable person, having regard to his rights 10 and obligations vis-a-vis the tax authority not to be open to challenge indefinitely (Ecotrade ...)…[19]With regard to the principle of effectiveness, it should be pointed out that the court has stated that it is compatible with Community law to lay down reasonable time limits for bringing proceedings in the interests of legal certainty which 15 protect both the taxpayer and the authorities concerned ... Such time limits do not make it impossible in practice or excessively difficult to exercise the rights conferred by European Union law."[100]In the light of these considerations the ECJ found that the three-year limitation in Latvian law was not precluded by Article 18. 20 101. Both Mr. Milne and Mrs Whipple sought support from paragraph 16: Mrs Whipple said that the analogy with the position of the individual permitted the State to impose time limits, and Mr. Milne pointed to the basic principle that legal certainty required the tax position of the taxable person not to be open to challenge indefinitely. 34[102]We agree with Mrs Whipple that paragraph 16 acknowledges that it is permissible for a state to provide temporal limitations on its obligations to repay overpaid VAT, or under claimed refunds. We agree with Mr. Milne that legal certainty (in the absence of any competing overriding or in the circumstances weightier principle) requires some limitations on the State’s ability 5 to claim. But it seems to us clear that there is a difference between a provision which limits the liability of the tax authority by deducting amounts representing an otherwise out-oftime net liability of the taxpayer (as we have construed s. 81 (3A)), and one which leaves the taxpayer exposed to liabilities without limit of time. The first does not 10 contravene the principle here stated, the latter does.[103]We turn to the principle of equality, "that similar situations should not be treated differently unless differentiation is objectively justified".[104]In SFI, the taxpayer was complaining of the fact that it was only given a right to claim deduction of the VAT within five years of the date on which the right arose 15 whereas the tax authorities were permitted five years from the date on which any tax return should have been made to make a claim.[105]The Court, having recalled that the principle of equality was one of the fundamental principles of Community law required for the interpretation of national rules falling within the scope of Community law [29], [30], said: 20 "[32] However, the position of the VAT authorities cannot be compared with that of the taxable person. The authorities do not have the information necessary to determine the amount of the tax chargeable and the deductions to be made until, at the earliest, the day when the return ... is made ... In the case of an inaccurate return, or where it turns out to be incomplete, it is therefore only from that time 25 that the authorities can start to recover unpaid tax. 35 “[33] Thus, the fact that the five-year limitation period begins to run against the tax authorities on the date on which the return should in principle be made, whereas an individual may exercise his rights to deduction only within a period of five years from the date on which that right arose is not such as to infringe the principle 5 of equality."[106]Mr. Milne said that the lack of any time limit whatsoever cannot be justified.[107]There is no equivalent to s. 81 (3A) for a taxpayer who seeks to make a claim under the VAT legislation, rather than a claim for damages as in Swedish State v. 10 Stockholm Lindopark AB (Case C-150/99) [2001] STC 103. If mistakenly the taxpayer fails to make a claim to input tax in year 1, and by reason of the same mistake becomes liable in year 6 to be assessed for undeclared output tax in relation to year 5, he cannot offset the out-of-time input tax claim against the claim by HMRC for year 5. If these are similar situations they are thus treated differently. 15 108. Mrs Whipple says that the State and the taxpayer are in different positions because the state has to protect public revenues in the public interest whereas the taxpayer has no such responsibility.[109]We accept Mr. Milne's comment that there is a difference between giving a state a longer limitation period (as in Ecotrade or SFI) where the state may need extra time 20 to gather information about a return and the position in relation to s. 81 (3A). But we accept Mrs Whipple's argument that the State is in a different position from the taxpayer as it has different responsibilities. Therefore, although with some misgivings, we think that Mrs Whipple is right on this issue. 36 Limitation to cost components.[110]We have recorded Mr. Milne's submission that Community law requires s. 81 (3A) to be limited to the set-off of otherwise out-of-time liabilities of the taxpayer which are cost components of the overpaid output tax (or theoretically also vice versa). The arguments for that approach were(1) that legal certainty 5 so required,(2) that it was the construction required by the Explanatory Notes and so far as relevant HMRC’s internal materials, and(3) that it accorded with the fundamental principle of the Directive in relation to the link between inputs and outputs. We have dealt with the first two of these in the sections above. So far as the last is concerned it seems to 10 us that the context of s.81 (3A) is not that of determining what sums are capable of deduction but the accounting between the taxpayer and state for amounts which arise from the application of the Directive; the requirement for a direct link has effect at an earlier stage in the process. On our construction of the subsection, Mr Milne’s limitation would not affect the outcome where a mistakenly claimed input preceded a 15 mistakenly taxed output of which it was a component; it would have effect only where the mistaken deduction for the input was either a component of a later output which had not been taxed, or where it was not a cost component of a taxable supply. We see no principle under the Directive which indicates that a deduction for such an input should not be recovered. 20 A Mistake: one or two[111]Mr. Milne says that there were two mistakes. The first was the failure to implement Article 13(1) (n) at all (remedied with effect from 1 June 1996), and the 37 second was the administrative failure in relation to Note(2) (c). In the first period the Theatre’s rights arose solely under Community law and the direct effect of Article 13; indeed in that period it could not even be said that there was an issue in relation to the voluntary management condition since the UK did not avail itself of that option. You could not make a mistake about something which was not 5 there. Thus the 1990 to May 1996 overpayments were not the result of the same mistake as the 2000-2001 reclaims. S. 81 (3A) could not therefore apply to permit the reclaims to be set off against the 1990 – May 1996 claims.[112]Mr Milne says that the only alternative, namely to regard a "mistake" as a 10 "mistake about the amount of the VAT payable" renders the language of s. 81 (3A) meaningless: there is a clear intention to limit the scope of the set-off in the terms of the legislation.[113]He says that the distinction between the first and second mistake is consistent with the ECJ's formulation in M&S 2002, quoted above at paragraph [38], where the 15 court distinguishes between(1) the failure properly to implement the Directive at all (which has its counterpart between 1990 and 1996) and(2) "where the national measures correctly implementing the directive are not being applied in such a way as to achieve the results sought by it" (which has its counterpart in 2000/2001).[114]In Ambulanter Pflegedienst Kugler GmbH v Finanzamt fur Korperschaften I in 20 Berlin (C 141/00) [2002] ECR I-6833, the ECJ considered inter alia the right of a taxpayer to rely upon the exemption afforded by another subparagraph of Article 13 (1), namely Article 13(1) (g) which had not been fully implemented by Germany. Paragraph (g) related to welfare services supplied by parties regarded as charities. As 38 is the case with paragraph (n) the exemption could be made subject to conditions. These included conditions which could limit the type of charity which could benefit from the exemption. The Court held that paragraph (g) had a direct effect and that a member state which had not adopted permitted conditions could not rely on its own omission in order to refuse entitlement to the exemption ([60]). 5 Mr. Milne relies on Ambulanter as showing that the UK could not have relied on the Note(2) (c) condition in the 1990/1996 period. There were two mistakes.[115]The right to deduct, says Mr. Milne, arises from Article 2 of the First Directive - the direct linkage of cost components with supplies - the same mistake is one which 10 affects the supply and the cost component.[116]Mrs Whipple says that there was only one mistake: to treat the Theatre’s supplies as taxable when they should have been exempt. Both the 1990/1996 overpayments and the 2000/2001 reclaims resulted from that mistake. That mistake was to tax what should have been exempt: to ignore throughout the whole period the Theatre’s 15 directly effective right to the effect of the exemption on which it could rely. Paragraph [27] M&S 2002, quoted above, showed that it was the obligation of the state to achieve the result envisaged by the Directive. The mistake was failing to achieve that result. All Ambulanter shows is that the taxpayer could claim the benefit of the state’s failure to implement properly. Article 13 A(1) (n) appears with and is closely 20 connected to Article 13A(2) (a) which permits the member state to specify conditions for defining charitable bodies: the two provisions are not to be considered in isolation.[117]We start by noting that s. 81 (3A) does not implement a Directive, nor is it required by Community law; rather it is part of the domestic time limit regime which 39 provides a domestic limitation to a Community law right. As a result the way in which judgements of the ECJ have dealt with transgressions by Member States is of help in construing the ambit of "mistake" only in so far as the concepts appearing in those judgments formed part of the context in which the provision was enacted. The context of s. 81 (3A), as it appears from the Explanatory Note, 5 owes little to EU jurisprudence. The concern was that in some cases a taxpayer would exercise the right to payment when the state was out of time to assess the liability. And, even if the approach taken by the ECJ in M&S 2002 could be seen as part of the context of the enactment of s. 81 (3A), we find no help therein in elucidating the ambit of "mistake” 10 in s. 81 (3A). Whilst the court says that the adoption of national measures correctly implementing a directive does not exhaust an individual's rights under the directive, it does not expressly elide the two concepts. The court’s interest is not in whether they are the same mistake, but in the rights of the individual in both circumstances. Such a right may be traversed in different ways: the court does not say, and does not need to 15 say, that they are the same.[118]If that is right, one is left with the customary tools of construction: gleaning what one can from the words, the statutory context and the mischief at which the provision was aimed. So far as the words are concerned we note that paragraph (b) requires the right to repayment to be "in consequence of a mistake"; it seems quite possible that 20 such a right might arise in consequence of the number of descriptions of circumstances. Thus, for example, input tax may not have been claimed because the taxpayer’s advisers mistakenly relied on the domestic legislation, or because the taxpayer's financial director mistakenly accepted that advice, or because Parliament mistakenly thought it did not have to implement the directive, or because HMRC 40 mistakenly advised taxpayers that their rights were limited to the domestic legislation. All of those could be mistakes satisfying (b). Paragraph (c) refers to "that" mistake. Thus, for s. 81 (3A) to operate, at least one of the possible descriptions of events must also have caused HMRC not to assess. But there is no requirement that only one of the possible descriptions of mistake was the same as that 5 which caused HMRC to have failed to assess. But, so long as there was at least one, s. 81 (3A) can take effect.[119]In this there is no dichotomy between one mistake and two mistakes: there is simply the question of whether a particular description of a series of events can properly be called a mistake for the purposes of s. 81 (3A). 10 120. That suggests that if two different circumstances giving rise to the requisite consequences can be given a description which embraces them both, then s. 81 (3A) may take effect. So that if one can describe the reasons for the overpayment of output tax and the erroneous repayment to input tax as the "failure to accord to UK taxpayers the exemption of ticket sales" as did the FTT, or as the mistaken belief of the Theatre 15 and HMRC that the Theatre's ticket sales were standard rated, that description might enable s. 81 (3A) to take effect even if one can also describe the separate events as two mistakes.[121]This is not to accept Mrs Whipple’s argument that a mistake must be judged by its results, and that, because the results of a failure to implement the Directive and the 20 administrative failure were the same, there was one mistake. In ordinary usage two different mistakes may have the same result: if I mishear what you say, the result may be that I erroneously believe you are insulting me; if I hear what you say correctly but misunderstand your intention, I may also erroneously think you are insulting me. In 41 that example there are two quite different mistakes with the same result; but as a matter of language it is yet possible to formulate a single description which embraces them both, and which is a "mistake": "I mistook what you said".[122]Mr. Milne says that if a wide meaning is given to "mistake", then the limitation to a mistake in s. 81 (3A) becomes meaningless. He says that the 5 only alternative to the two mistakes which he describes is "a mistake about the amount of VAT payable". This might not cover a mistake about repayment, but "a mistake about the operation of VAT" would be wide enough to make Mr. Milne’s point. This argument shows that it remains a question of the use of language: when does a description become so 10 general as not to fit the purpose of the statutory language?[123]In this appeal the treatment of the Theatre’s supplies as standard rated derived from the omissions and actions of two different bodies: Parliament omitted to legislate, and HMRC acted to interpret legislation wrongly. The errors were corrected in different ways: Parliament by legislating, HMRC by changing their practice. It is 15 not possible to describe each separate omission or action as the same mistake - they were plainly different mistakes; but is it possible to describe the circumstances of both as the failure of the member state (its legislature and executive) to treat the Theatre’s ticket sales as exempt - for all that s. 81 (3A) requires is a description of events which may be called a mistake by reason of which rights to repayment and failure to assess 20 arose?[124]It seems to us that it is possible to do so. If one had asked the Theatre why it was declaring output VAT or reclaiming input VAT, it would have said "because our 42 ticket sales are standard rated". That is a fair description of the mistake made even though further enquiry might reveal different reasons for different aspects of it.[125]We conclude that there was a mistake, namely treating the Theatre’s ticket sales as standard rated, which was the reason for both the output claim and the repayment. S. 81 (5 3A) therefore applies. “Amounts payable under this Act”[126]Lastly Mr. Milne said that the mistake must be as to "amounts payable under this Act". The 1990/1996 overpayments of output tax were not by virtue of a mistake 10 about the amounts payable under the Act (because it contained no exemption at all at that time) but about amounts payable under the direct effect of the Directive. Only after 1996 could the mistake relate to payments under the Act.[127]We disagree. The mistake was a mistake about “what amounts or to what extent amounts were payable under the Act”. The Act falls to be construed in accordance 15 with the Directive. If mistakenly it was construed to have the effect that VAT was payable under it, that was a mistake about the amount payable under the Act. Transitional Period[128]It does not seem to us that any transitional period was required for the 20 introduction of s. 81 (3A) as we have construed it. Such a provision would be required if it removed rights under the Directive, but on our construction s. 81 (3A) does not 43 remove rights under the Directive but limits a taxpayer’s claim to the amount of his right under the Directive. A transitional period is not needed for that. Conclusions on s. 81 (3A)[129]The discussion above addresses each of the five issues 5 raised before us:(1) properly construed s. 81 (3A) does not permit HMRC to pick and choose;(2) set-off is limited to the same mistake, not to items linked in any other way;(3) no time limitation is required, although some modification of normal procedural rules may be needed;(4) it is unhelpful to characterise the mistakes by asking the question ‘one 10 mistake or two?’ but the same mistake was made; and(5) no transitional period was required for the implementation of s. 81 (3A).[130]S. 81 (3A) is to be construed subject to the conditions,(1) that all relevant previous years need to be considered and(2) that in relation to the consideration of years in which the taxpayer is not required to keep records, the onus should be on 15 HMRC to show that any adjustment should be made. However the operation of s. 81 (3A) is not precluded by the principles of legal certainty, equality, equivalence, or the supremacy of Community law.[131]So applied the effect of s. 81 (3A) is not incongruent with Community law. The issues of incongruence raised by HMRC do not arise. 20 132. We therefore agree with the conclusion of the FTT in relation to s. 81 (3A) and dismiss the appeal. 44 MRS JUSTICE PROUDMAN MR CHARLES HELLIER 5 UPPER TRIBUNAL JUDGES 10 RELEASE DATE: 14 FEBRUARY 2013

Cited in 3 later judgments