IN THE SUPREME COURT OF JUDICATURE
ON APPEAL FROM THE HIGH COURT OF JUSTICE
CHANCERY DIVISION
MR JUSTICE WARREN
High Court Reference CH/2005/APP/0053
Royal Courts of Justice
Strand, London, WC2A 2LL
Before :
LORD JUSTICE CHADWICK
LADY JUSTICE ARDEN
and
LADY JUSTICE SMITH
Between :
THE CONDÉ NAST PUBLICATIONS LIMITED | Appellant |
- and - | |
COMMISSIONERS OF HM REVENUE & CUSTOMS | Respondent |
(Transcript of the Handed Down Judgment of
Smith Bernal WordWave Limited
190 Fleet Street, London EC4A 2AG
Tel No: 020 7421 4040 Fax No: 020 7831 8838
Official Shorthand Writers to the Court)
Mr Jonathan Peacock QC (instructed by Deloitte & Touche LLP, 180 Strand, London WC2R 1BL) for the Appellant
Mr Christopher Vajda QC and Miss Valentina Sloane (instructed by the Solicitor for HM Revenue & Customs, Somerset House, Strand, London WC2R 1LB) for the Respondent
Judgment
Lord Justice Chadwick :
This is an appeal from an order made on 10 June 2005 by Mr Justice Warren on an appeal by The Condé Nast Publications Limited (“CNP”) from the decision of the VAT & Duties Tribunal ( Dr David Williams, chairman, and Mr Praful Davda) released on 7 December 2004. The judge dismissed that appeal. This appeal is brought with the permission of this Court (Lord Justice Jacob) granted on 25 July 2005.
The issue between the parties is as to the time within which a taxable person (within the meaning of section 3(1) of the Value Added Tax Act 1994) may make a claim for the repayment of under-deducted input tax under regulation 29 of the Value Added Tax Regulations 1995 (SI 1995/2518). It is convenient, at the outset, to set that regulation in context.
The statutory framework
Value added tax (“VAT”) is charged, so far as material, on the supply of goods or services in the United Kingdom – section 1(1) of the Value Added Tax Act 1994 (“VATA 1994”). VAT on supplies made by a taxable person is known as “output tax” – section 24(2) VATA 1994. “Input tax”, in relation to a taxable person, is VAT on the supply to him of goods and services. Put shortly, input tax is the VAT paid by the taxable person to his supplier: output tax is the VAT for which the taxable person has to account on supplies made by him to his customer.
Section 25(1) VATA 1994 requires that a taxable person shall account for and pay VAT by reference to such periods (“prescribed accounting periods”) at such time and in such manner as may be determined by regulations. Section 25(2) VATA 1994 provides that the taxable person is entitled at the end of each prescribed accounting period to credit for so much of his input tax as is allowable under section 26 of the Act, and then to deduct that amount from any output tax which is due from him. Input tax is allowable under section 26 VATA 1994 if it is attributable to taxable supplies made by the taxable person in the course of his business. In a case where no output tax is due at the end of the relevant prescribed period, or where the amount of the credit to which the taxable person is entitled under section 25(2) VATA 1994 exceeds the amount of the output tax for which he is accountable, the amount of the credit or (as the case may be) the amount of the excess is payable to the taxable person by the Commissioners.
Regulation 29(1) of the Value Added Tax Regulations 1995 requires that “subject to paragraphs (1A) and (2), and save as the Commissioners may otherwise allow or direct, either generally or specifically” a person claiming deduction of input tax under section 25(2) VATA 1994 shall do so on a return made by him for the prescribed accounting period in which the VAT becomes chargeable. Regulation 29(1A) is in these terms:
“29(1A) The Commissioners shall not allow or direct a person to make any claim for deduction of input tax in terms such that the deduction would fall to be claimed more than 3 years after the date by which the return for the prescribed accounting period in which the VAT became chargeable is required to be made.”
Regulation 29(1A) was introduced by the Value Added Tax (Amendment) Regulations 1997 (SI 1997/1086). It has effect from 1 May 1997. There were no transitional provisions. At first sight, therefore, the effect of regulation 29(1A) was to exclude claims made after 1 May 1997 in respect of under-deduction of input tax in cases where the return for the prescribed accounting period in which the VAT became chargeable was required to be made more than three years before that date: that is to say, in cases where the claim to deduct, if made in accordance with regulation 29(1), would have been made in a return which was, itself, required to be made before 1 May 1994. But it is important to have in mind that, in giving effect to the provisions of domestic law in VATA 1994 and the subordinate regulations, regard must be had to Community law.
Community law
The provisions of VATA 1994 to which I have just referred were enacted to give effect to the right to deduct conferred by the Sixth Council Directive (77/388/EEC) on the harmonisation of the laws of Member States relating to turnover taxes – in particular, by article 17 of that Directive. Article 18 of the Directive is in these terms, so far as material:
“18.2 The taxable person shall effect the deduction by subtracting from the total amount of value added tax due for a given tax period the total amount of the tax in respect of which, during the same period, the right to deduct has arisen . . .
18.3 Member States shall determine the conditions and procedures whereby a taxable person may be authorised to make a deduction which he has not made in accordance with the provisions of [paragraph 2].”
The power conferred on Member States by article 18.3 must be exercised in a manner which is compatible with Community law principles. In particular, the exercise of that power must be compatible with the principle of effectiveness and the principle of the protection of legitimate expectations – see the decision of the Court of Justice in Marks and Spencer plc v Customs and Excise Commissioners (Case C-62/00) (“the Marks and Spencer case) [2002] STC 1036, at paragraphs [34] to [46].
The question which had been referred to the Court of Justice in the Marks and Spencer case arose in the context of an amendment to section 80(4) and (5) VATA 1994 which had been made under the Finance Act 1997. Section 80 VATA 1994 made provision for the recovery of overpaid VAT. As enacted sub-section (4) provided that no amount might be claimed under the section after the expiry of six years from the date when it had been paid. That was subject to sub-section (5), which allowed an extension of that period where the amount had been paid by reason of a mistake. Section 47 of the Finance Act 1997, which was given effect from 18 July 1996, reduced the period in section 80(4) VATA 1994 from six years to three years and repealed section 80(5) VATA 1994. The Court of Justice held that “national legislation retroactively curtailing the period within which repayment may be sought of sums paid by way of VAT collected in breach of provisions with direct effect of the Sixth Directive . . . is incompatible with the principles of effectiveness and of the protection of legitimate expectations”.
The Fleming case
In Fleming (trading as Bodycraft) v HM Revenue and Customs [2006] EWCA Civ 70; [2006] STC 864 this Court had to consider the effect of the decision in the Marks and Spencer case in the context of a claim for repayment of input tax under regulation 29.
The underlying facts in Fleming are summarised in paragraph [2] of the judgment of Lady Justice Arden:
“[2] In 1989 and 1990 Mr Fleming, the sole proprietor of a business engaged in the purchase and servicing of quality cars, purchased thirteen Aston Martin motor cars for the purpose of his business. He duly reclaimed the input tax on ten of the thirteen cars, and Customs & Excise paid his claim. However, he did not receive VAT invoices at the time of purchasing the remaining three cars. He needed those invoices to make a claim to recover input tax on purchase of the three remaining three cars. In the event, it was not until 23 October 2000 that Mr Fleming made a claim for repayment of VAT paid on these three cars and his claim was refused.”
The Commissioners rejected the claim, relying on the three year time limit for making repayment claims imposed by regulation 29(1A) of the 1995 Regulations. On Mr Fleming’s appeal to the Vat & Duties Tribunal it was held ([2004] V&DR 172) that the Commissioners could not rely on the three year time limit: Mr Fleming had an accrued right to deduct input tax under Community law which could not be taken away by legislation which had retrospective effect. Nevertheless, the tribunal dismissed Mr Fleming's appeal on the ground that if the matter were sent back to the Commissioners they would inevitably reject the claim in exercise of their discretion; and that they would be entitled to do so.
Mr Fleming appealed to the High Court. The Commissioners did not seek to uphold that tribunal’s decision on the basis of the reason which it had given. Mr Justice Evans-Lombe ([2005] EWHC 232 (Ch) [20]; [2005] STC 707, 712) accepted that, to the extent that regulation 29(1A) would otherwise prevent a taxable person with accrued rights from recovering under-deducted input tax – say, in a case where, at the time the regulation limit took effect, he had an accrued claim but the three year period had already expired – the principle of effectiveness precluded reliance upon the regulation. But he went on to observe that the principle of effectiveness did not have the effect of striking down the regulation: the national court must apply the regulation save in so far as the effect of doing so would be incompatible with the Community principle He referred to the judgment of Mr Justice Lawrence Collins in Local Authorities Mutual Investment Trust v Customs and Excise Commissioners [2003] EWHC 2766 (Ch) [67]; [2004] STC 246, 259 and to the judgment of the Court of Justice in Grundig Italiana SpA v Ministero delle Finanze (Case C-255/00) [2002] ECR I-8003, [41]. He said this (ibid [24]; 713):
“[24] The Court of Justice [in Grundig] was considering a case where, what was in issue, was the sufficiency of transitional provisions. It seems to me, however, that the principles, highlighted in that passage from the judgment, are equally applicable where the relevant time limit imposed by the national legislature is not, as in the case of reg 29(1A), accompanied by any such transitional provisions. The effect of what the Court of Justice is saying in this case is that, even in the case of individuals whose claims have accrued before the time limits were imposed and who may therefore be in a position to require the national court to disapply the time limits to their claims, if brought within a reasonable time after the imposition of the limits, the imposition of the time limits does not continue indefinitely thereafter. If they allow too long a period to go by before making a claim the national court may properly conclude that the principle of finality or legal certainty requires it to refuse to disapply the limitation provisions.”
He held that Mr Fleming had been capable of making a claim for repayment of input tax since 1990. He had delayed for three years and five months after regulation 29(1A) had come into force (in May 1997). In those circumstances the Commissioners were justified in relying on the regulation.
Mr Fleming’s appeal from the order of Mr Justice Evans-Lombe was allowed by this Court. But the members of the Court differed as to the reasons which led to that conclusion. Lord Justice Ward and Lady Justice Hallett took the view that it was impossible, as a matter of construction, to imply or read into regulation 29(1A) a transitional period (whatever a proper transitional period might be). As Lord Justice Ward put it, at paragraphs [80] and [81] of his judgment ([2006] EWCA Civ 70; [2006] STC 864):
“[80] . . . Whilst a benign purposive construction of an inadequate transitional period may possibly lead to the implication of a period of grace which is judged to be reasonable, I simply cannot see how one can construe something out of nothing. If it were possible, Marks & Spencer would surely have been decided differently.
[81] In my judgment the failure of the Commissioners to include any transitional time limit in regulation 29(1A) is fatal to their case. The regulation as it is without any transitional period fails the Marks & Spencer test. Consequently, in my judgment, it must be disapplied. The appeal must be allowed accordingly.”
Lady Justice Hallett agreed. After referring to the speech of Lord Steyn in Ghaidan v Godin-Mendoza [2004] UKHL 30, [48]; [2004] 2 AC 573, 576, she said this:
“[65] I derive from Ghaidan the principle that however strong and radical the obligation on a court to interpret legislation, there is a line the courts may not cross. . . .
[66] The duty is to read legislation in such a way that it gives proper effect to Community rights ‘so far as possible’. I emphasise ‘so far as possible’ or as Lord Oliver put it at page 559E of Litster [Litster v Forth Dry Dock & Engineering Co Ltd [1990] 1 AC 546] the duty to construe legislation in this way arises ‘if the legislation can be reasonably construed so as to conform’. For my part, I cannot accept that this is a Litster type situation. I cannot accept that the legislation here can be reasonably construed as incorporating a transitional period where none was provided or contemplated. In my judgment it is simply not possible to do so however broad or robust an approach the court is obliged to adopt.”
Lady Justice Arden rejected the contention that no transitional period could be implied or read into regulation 29(1A). As she pointed out (at paragraph [44] of her judgment) the effect would be that taxpayers whose claims arose before 1 May 1997 could continue to make claims under regulation 29 until that regulation were amended to include a transitional period; notwithstanding that those claims were submitted outside whatever period might properly have been allowed. She said this:
“[42] The Court of Justice in the Grundig case does not suggest that in determining the adequacy of the transitional period for this purpose the individual circumstances of taxpayers in Grundig's situation should be taken into account, or that the particular knowledge which they might be expected to have of the legal situation is relevant. If the circumstances affecting a particular taxpayer are not relevant, it is difficult to see how the principles of Community law would prevent the implication of a transitional period where none has been provided for by the legislature as a matter of interpretation. It will be recalled that under article 10 of the Treaty on European Union:
‘Member States shall take all appropriate measures, whether general or particular, to ensure fulfilment of the obligations arising out of this Treaty or resulting from action taken by the institutions of the Community. They shall facilitate the achievement of the Community's tasks’
[43] This obligation is owed by every organ of the Member State concerned within the sphere of its competence and the judiciary is an organ of the Member State for these purposes: Simmenthal 1978 [ECR] 629.
[44] In other words, if a transitional period is required to give effect to Community rights, in my judgment it follows from the obligations imposed on national courts by the EU treaty that the national court must read such a period into its national legislation under the principle of conforming interpretation (as to which, see, for example, Litster v Forth Dry Dock & Engineering Co Ltd [1990] 1 AC 546). The very basis of this principle is that the courts may be obliged to read legislation as containing a provision which the legislature ought to have made under Community law but failed so to do. The principle of conforming interpretation, or Marleasing principle, was considered recently by this court (Pill, Latham and Arden LJJ) in Customs & Excise v IDT Card Services Ireland Ltd ([2006] EWCA Civ 29). It was there held that to apply the principle the court might have to write in words. This court further held, relying on the principles applying under section 3 of the Human Rights Act 1998 and explained by the House of Lords in Ghaidan v Godin-Mendoza [2004] 2 AC 557, that the principle could not be applied if its application overrode some cardinal feature of the legislation, or raised policy issues that the court could not resolve. But that is not the situation in the present case. There is nothing in regulation 29 which states or implies that there is to be no transitional period if that is what is necessary for the United Kingdom to fulfil its obligations under the EU Treaty. Moreover, the Court of Justice itself determined the length of an adequate transitional period in the Grundig case, thus indicating that where the legislature has failed to provide such a period the exercise of determining such a period is intended to be done through the normal judicial process. The role of the legislature is not excluded. It can itself determine the length of the transitional period or indeed confer a longer transitional period than one which is merely adequate. However, alternatively (if this is different) regulation 29(1A) can be disapplied as regards persons entitled to the benefit of a transitional period. . . .”
Her conclusion that Community law would recognise, or imply, or both, a notional transitional period, not specifically set out in legislation, led Lady Justice Arden to address the related issue “by reference to what event should that period be fixed.” On that issue, she expressed agreement (ibid, [51]) with the first of the two options under what Mr Justice Warren had described as “the second possible approach” at paragraph [39] of his judgment in the present case ([2004] EWHC 1167 (Ch), [39]; [2005] STC 1327, 1339f):
“[39] The second possible approach is that the principle of effectiveness requires that a taxpayer should be entitled to enforce his Community law right until the time has been reached when he could first have been expected to assert that right; and that he could not be expected to do so as long as that right had not been established and was subject to challenge in the ECJ by the Member State concerned. He should, therefore, have a reasonable time in which to assert his Community law right once that right had been established . . .”
On that basis, the event by reference to which the transitional period during which Mr Fleming was entitled to enforce his Community law right was the handing down of the judgment of the Court of Justice in the Marks & Spencer case. As Lady Justice Arden put it (ibid, [49]) :
“[49] The question of the availability of knowledge (or foresight) is relevant here primarily for the purpose of the principles of legal certainty and legitimate expectations. Those Community principles are in my judgment not satisfied in the circumstances of this case unless taxpayers, who are entitled to claims that would otherwise be capped, have a transitional period for making late claims whose expiry is fixed by reference to the date of the handing down of the judgment of the Court of Justice in the Marks & Spencer case. . . . That would mean that Mr Fleming's claim could not be rejected by reason only that it was made later than the expiration of an adequate transitional period commencing on 1 May 1997 . . .”
The judgment of the Court of Justice in the Marks & Spencer case was handed down on 11 July 2002. Mr Fleming had made his claim to recover input tax on the three outstanding invoices on 23 October 2000. That was within the notional transitional period. So she would have allowed the appeal on that basis.
The repayment claim in the present case
CNP’s claim for repayment in the present case is in respect of input tax on supplies made to CNP relating to staff entertainment. The claim was set out in a letter of voluntary disclosure dated 27 June 2003. There had been under-deduction of input tax in each of the prescribed accounting periods from 1 April 1973 to 31 March 2003; although, for the purposes of this appeal, it is only the claim in respect of periods up to and including the period ended 30 April 1997 – that is to say, in respect of periods before regulation 29(1A) took effect on 1 May 1997 - that is in issue. The amount claimed is in excess of £115,000.
That element of the claim was rejected by a decision of the Commissioners contained in a letter dated 31 October 2003. CNP appealed to VAT & Duties Tribunal from that decision. The tribunal dismissed that appeal. It accepted, at paragraph 23 of its decision (released on 7 December 2004) that “CNP would, as the law is now understood, have been entitled to repayment of any input tax incurred on relevant forms of staff entertainment expenditure during the period to 1999 if it had made a timely repayment claim”. But it posed the question whether CNP, on the basis of the facts which the tribunal had found, had a right in law to claim in 2003 a refund of the input tax which it had failed to claim by way of deduction in 1973.
The tribunal’s reasoning
Put shortly, the tribunal took the view that, but for the initiative of CNP’s advisers, Deloitte & Touche LLP – who, it was said, would take a percentage of any input tax repayment made – CNP would not have chosen to make the claim even if there had been a transitional period available. The relevant findings of fact are at paragraphs 20 and 21 of its decision:
“20. [The tribunal] finds that – as it now turns out, wrongly – all those involved with CNP VAT liability were thinking in terms of claims under section 80 [VATA 1994] rather than Regulation 29 at the relevant times. It finds that CNP was aware of under recovery of input tax on this form of expenditure at the relevant time from at least receipt of the VAT Notes in 1994. It finds that the only relevant information of which CNP was aware before the Price Waterhouse notice of September 1997 was in the VAT Notes sent out by the respondents. It finds that at no time did the individual responsible for making VAT claims and returns for CNP – Mr Michael – take any initiative to make any claim, nor make any enquiry save those to his immediate superior, related to this issue. Further, the approach taken by Mr Michael on that issue was one for which a clear justification has been identified, that justification being entirely reasonable in the particular circumstances of CNP. It further finds that, as a matter of considered management policy, CNP had taken no steps to see if it could recover the under recovered tax at any time before 2003. Nor did the tribunal see any evidence that CNP had put in any other protective claim in connection with VAT at any time. While the tribunal finds that CNP did not consider the possibility of putting in a "global" claim for all past years until 2003, CNP did consider the question whether to claim before there was a time limit, and rejected the opportunity to do so. It also rejected the opportunity to do so for a three year period after the time limit came in. It did not react, until the current claim, to notification by the respondents or from any other source of transitional periods.
21. The Deloittes letter contends that, nonetheless, it is ‘inconceivable’ that CNP would not have put in an uncapped claim if given a transitional period in which to do so in 1997. The tribunal is unable to agree that this was ‘inconceivable’. When the current claim was made it was by the initiative of a third party talking to Mr Michael's superior on terms that did not involve Mr Michael in any decision making, staff time, or specific expenditure prior to the claim being made by the third party. The only basis for concluding that a failure to claim by CNP was ‘inconceivable’ would be that it was clear on the evidence beyond all doubt that such a third party would make such an approach to Mr Michael or his superior. The tribunal is unable to make that finding. Even using the standard of the balance of probabilities, the tribunal is unable to conclude otherwise than that a claim may have been made, and would probably have been made if there had been such an approach as was in fact made by Deloittes some years later. There is no evidence that advice would have been sought, or would have come in any other form than it did come, from CNP's retained advisers. The tribunal has no evidential basis on which to consider whether such an approach by an outsider would have been probable.”
On the basis of those findings of fact the tribunal concluded, at paragraph 48 of its decision, that:
“48. . . . there is no evidence that the amendments introduced to regulation 29 stopped the appellant making any claim that they were minded to make, or had done any work on making, at that time, nor any clear evidence of any other detriment to this appellant's consideration of any claim for unrecovered input tax at that time. There is therefore no disadvantage shown to this appellant in the introduction of the time limit without a transitional period.”
It followed, in the view of the tribunal, that it was not open to CNP to revisit, in 2003, its earlier decisions not to claim in respect of the under-deduction of input tax.
The judge’s view
CNP appealed to the High Court. The appeal came before Mr Justice Warren in May 2005 – that is to say, after the decision of Mr Justice Evans-Lombe in the Fleming case but before the appeal in that case had come before this Court. It is of interest to note that the approach which was (subsequently) to find favour with the majority in this Court in Fleming was not pursued – or, at the least, not pressed – by counsel for CNP. The judge identified the point at paragraphs [22] and [24] of his judgment ([2004] EWHC 1116 (Ch), [22], [24]; [2005] STC 1327):
“[22] First, consistent with the decision in [the Marks & Spencer case] transitional provisions should have been included in relation to Regulation 29 just as they should have been included in relation to section 80 [VATA 1994]. The absence of an adequate transitional period renders the amendment made by Regulation 29(1A) unlawful since the amendment then constitutes an unacceptable fetter on CNP's right to recover input tax. Regulation 29(1A) in its current form is in breach of Community law and provides no legitimate basis on which the Commissioners can refuse repayment.
. . .
[24] Although [counsel] maintained his first principled submission, it seemed to me that, taken to its logical conclusion, it leads to the result that, since no transitional provision has, even to this day, been adopted (whether by statute or by practice) in relation to Regulation 29, it would still be open to a taxpayer who had not yet made a claim now to make one in relation periods prior to 1 May 1997. [Counsel], recognising that such a result would be extreme, accepted that it would now be too late to make a claim.”
The judge observed, at paragraph [24], that in order to reach the conclusion that “it would now be too late to make a claim” it was necessary either (i) to reject the submission to which he had referred in paragraph [22] or (ii) to qualify that submission so as to limit the period within which claims could be brought so that that period allowed taxpayers a reasonable time (and no more) to consider and digest the decisions of the Court of Justice in Marks & Spencer and Grundig.
After examining the judgments of the Court of Justice in the Marks & Spencer and Grundig cases with care and at some length, the judge held (at paragraph [36] of his judgment) that the failure, when introducing regulation 29(1A), to include an appropriate transitional period during which a trader could make a claim under regulation 29 in relation to periods prior to the new time-limited period, was a breach of Community law which gave the trader directly enforceable rights. But he observed that neither Marks & Spencer nor Grundig “deals clearly with how matters are to be dealt with after the expiry of what would have been a reasonable transitional period had one been properly introduced”. As to that, he identified the two distinct approaches which, as he thought, it would be possible to adopt. He set those out at paragraphs [38] and [39] of his judgment:
“[38] The first possible approach is that the new national time limit can be relied on by the Member State once a reasonable time has passed since its introduction; and this is so regardless of whether the taxpayer knew that he had, or believed that he might have, a Community law right which he could enforce notwithstanding the failure to provide for a proper transitional period. . . .
[39] The second possible approach is that the principle of effectiveness requires that a taxpayer should be entitled to enforce his Community law right until the time has been reached when he could first have been expected to assert that right; and that he could not be expected to do so as long as that right had not been established and was subject to challenge in the ECJ by the Member State concerned. . . .”
If the second of those approaches were adopted, it would be necessary to decide whether the time within which the taxpayer could first have been expected to assert his right was to be determined by reference to the date at which that right had been established (the first option) or the date at which “the generality of taxpayers and advisers appreciated that such a right might subsist” (the second option).
The judge then turned to examine the decision of Mr Justice Evans-Lombe in the Fleming case. The judge held (correctly, in my view) that Mr Justice Evans-Lombe had adopted the first of the two approaches which he, the judge, had identified. [Note: the judge refers, at paragraph [49] of his judgment to “the first approach which I have set out briefly in [39]ff above”. That is, I think, an error. The “first approach” is identified at paragraph [38]]. He took the view that, unless satisfied that Mr Justice Evans-Lombe was clearly wrong, he should follow that decision. He was not prepared to hold that Mr Justice Evans-Lombe was clearly wrong; indeed, as he accepted “he may be right”.
The effect, in the present case, of adopting the first approach – as Mr Justice Evans-Lombe had done in Fleming – had been spelt out by Mr Justice Warren at paragraph [38] of his judgment:
“[38] On that approach, it may be that a longer period should be allowed for enforcement of the directly enforceable right than the minimum period which could have been expressly provided. In the present case, and assuming that a 6 month transitional period for the purposes of Regulation 29 would have been appropriate, the time limit for making a claim would have expired 6 months after either 26 March or 1 May 1997, long before the claim was in fact made by CNP on 27 June 2003. Even allowing a longer period for enforcing Community law rights, a reasonable period would have expired long before that date.”
So, if that approach were adopted, the appeal had to be dismissed.
The judge went on to consider what the position would be if the second approach were adopted. As he said, he saw “a great deal of force in favour of the second approach”. He observed, at paragraph [54] of his judgment, that:
“[54] In the absence of a transitional period in the legislation, the most that a taxpayer could expect, even under the second approach, would be a reasonable time within which to make a claim under Regulation 29. . . . The sole question is, I consider, the duration of the reasonable period from, at latest 5 August 2002 (when the Community law requirement that there should be a transitional period was first recognised by the Commissioners in the first Business Brief), during which a Regulation 29 claim (not subject to the new time limit) could be made. In my judgment, CNP's claim, made in late June 2003, was made after the expiry of a period which would have been reasonable.”
So, again, if the second approach were adopted, the outcome would be the same: the appeal had to be dismissed.
The judge reminded himself, at paragraph [55] of his judgment, that it had been the Commissioners’ principal submission – and the point on which the tribunal had decided the appeal in favour of the Commissioners – that “the taxpayer must be able to show that he would have made a claim if transitional provisions had in fact been introduced”. He observed, at paragraph [60], that
“[60] The real question, then, is whether a requirement that the taxpayer should need to prove that he would have made a claim if proper transitional provisions had been included is compatible with the principle of effectiveness. I remind myself of the principle which in short is that national rules must not render virtually impossible or excessively difficult the exercise of Community Law rights.”
He answered that question at paragraphs [61] and [62]:
“[61] In my judgment, it is not possible to impose this requirement on taxpayers. The principle of effectiveness is designed to protect a person's rights. As a matter of Community law, taxpayers in the position of CNP had, prior to the introduction of Regulation 29(1A), a right to reclaim input tax. That right has been curtailed. The jurisprudence of the ECJ tells us that such rights must be protected for a transitional period. I can detect nothing in the language used in the judgments of the ECJ which describes the purpose of the protection as to put the taxpayer in the same position which he would have been in if the transitional provision had been included. The language is that of protecting rights: if those rights are not properly protected during a transitional period by national laws, then Community law disapplies the national law to the extent necessary to preserve those rights.
[62] If it is correct that the second approach discussed in [39]ff above is correct (and it is only if it is correct that the question now under consideration arises - if the first approach is correct, the claim has, in fact, to be made within the reasonable period and no question arises about what the taxpayer might have done since we will know what he in fact has done), then the Community law right which the taxpayer had during the transitional period is not to be taken away from him for a reasonable period after he could first reasonably be expected to have asserted it. The fact that he did not have the opportunity to exercise it during the transitional period is entirely the fault of the Member State in failing to comply with its Community law obligations. It would, against that background, make it excessively difficult, in my judgment, for him to exercise his Community law right if it were the rule that a taxpayer had to prove something which might, in its nature, be very difficult to prove, namely that he would have exercised his right had a transitional period been included.”
The issues
I have set out the judge’s reasoning at some length as an aid to identifying the issues which, following his judgment, had emerged in this litigation. There is no dispute as to the starting point. The decision of the Court of Justice in the Marks & Spencer case has established that transitional provisions should have been included in relation to regulation 29(1A). In this context there can be no difference in principle between the recovery of over-declared output tax under 80 VATA 1994 and the recovery of under-deducted input tax under regulation 29. The failure to make provision for an adequate transitional period has the effect that regulation 29(1A) is inconsistent with the need to give direct effect to article 17 of the Sixth Directive. Starting from that premise, the issues which had emerged in this litigation may, I think, fairly be summarised as follows:
Whether the lack of any transitional provisions in regulation 29(1A) leads to the conclusion that that regulation provides no legitimate basis on which the Commissioners can refuse repayment of past under-deducted input tax. That point was taken before the judge on behalf of CNP (but not, it seems, pressed), as he noted at paragraphs [22] and [24] of his judgment. [The issue was, of course, subsequently determined by this Court in the Fleming case. This Court has held that regulation 29(1A) does, indeed, provide no legitimate basis for the refusal to repay under-deducted input tax on a claim arising before 1 May 1997.]
Whether the domestic court is required, by the Community law principle of effectiveness, to read into regulation 29(1A) transitional provisions which allow for a reasonable period in which to bring claims which had arisen before 1 May 1997, as Mr Justice Evans-Lombe had held in Fleming. [That was the approach adopted by Mr Justice Warren in the present case; and that, in principle, was the approach favoured by Lady Justice Arden in Fleming.]
Whether the transitional provisions which are to be read into regulation 29(1A) in order to give effect to the principle of effectiveness are those which it would have been appropriate to include (or those provisions which, if included, would have allowed the minimum period necessary to give effect to the principle) if, at the time when the regulation was introduced, transitional provisions had been included. That was what Mr Justice Evans-Lombe had held in Fleming. [Mr Justice Warren adopted that approach in the present case because (as he said) he thought it right to follow Mr Justice Evans-Lombe; but that approach was rejected by Lady Justice Arden at paragraph [49] of her judgment in Fleming and by Lord Justice Ward, at paragraph [82] of his judgment in that appeal.]
Whether the transitional provisions which are to be read into regulation 29(1A) are those which do, in fact, give effect to the principles of effectiveness and the protection of legitimate expectations in the light of the events which have happened since the regulation was introduced; and, in particular, the fact that it was not until the decision of the Court of Justice in Marks & Spencer (on 11 July 2002) that those with claims which had arisen before 1 May 1997 knew that (notwithstanding the absence of transitional provisions in regulation 29(1A)) those claims could be pursued. [That approach attracted Mr Justice Warren in the present case. It received approval from Lady Justice Arden at paragraph [49] of her judgment in Fleming.]
Whether, if the correct approach would otherwise be that described in (4) above, it is necessary or appropriate to adopt that approach in a case where there is no evidence that – and the tribunal was not prepared to hold that – the taxable person would have taken advantage of whatever transitional provisions might, properly, have been included in regulation 29(1A) if at the time when the regulation was introduced, transitional provisions had been included. That was the point on which the Commissioners had succeeded before the tribunal in the present case. [That approach was rejected by Mr Justice Warren at paragraphs [61] and [62] of his judgment in the present case.]
Whether, if the correct approach is that described in (4) above, the transitional provisions which are to be read into regulation 29(1A) must be taken to provide that the period within which claims must be brought under those provisions did not expire before 30 June 2003 - that being the end of the period allowed by transitional provisions introduced to deal with the comparable problem which arose in relation to claims under section 80 VATA 1994 in respect of over-declaration of output tax. [That approach was rejected by Mr Justice Warren at paragraph [54] of his judgment in the present case. The need to address that point did not arise on the facts in Fleming.]
This appeal
CNP filed an appellant’s notice on 24 June 2005. The grounds of appeal seek to suggest that there is doubt as to the basis upon which the judge dismissed the appeal: that is to say, it is said to be unclear whether he had, on a true analysis, followed the approach adopted by Mr Justice Evans-Lombe in the Fleming case (the first approach) or had chosen his own approach (the second approach, described in paragraph [39] of his judgment) and had reached his conclusion for the reason given at paragraph [54] of his judgment.
In my view those doubts are misplaced. It is, I think, clear that – whatever reservations he may have had as to what he described as the first approach – Mr Justice Warren did follow Mr Justice Evans-Lombe and that it was on that basis that he dismissed the appeal. Be that as it may, CNP challenged that approach (“the First Conclusion”) on the ground that: “In holding . . . that time started to run against the Appellant from the point in time at which the UK introduced a three year cap to such claims (and thus that the Appellant was out of time), rather than the point in time at which the Appellant ought reasonably to have known that the three year cap contravened Community law, the Judge had insufficient regard to the principle of effectiveness”.
The second approach would have led the judge to conclude – for the reasons set out in paragraph [54] of his judgment - that a reasonable period for making a claim to repayment of under-deducted input tax was to be reckoned from 5 August 2002 (when the Community law requirement that there should be a transitional period was first recognised by the Commissioners in the first Business Brief) and would have expired before CNP’s claim was made in late June 2003. That conclusion was challenged, also, on the basis that “the Judge had insufficient regard to the principle of effectiveness.” It was said (in the skeleton argument filed shortly after the appellant’s notice) that, even if the judge was right to treat 5 August 2002 as the date from which a reasonable period for making a claim under regulation 29 (which CNP did not accept) the period could not end before 30 June 2003 – that being the end of the transitional period allowed by non-statutory transitional provisions introduced to deal with the comparable problem which arose in relation to claims under section 80 VATA 1994 in respect of over-declaration of output tax. I shall need to return to that point later in this judgment. It is sufficient to note, at this stage, that CNP made its repayment claim just within a period which ended on 30 June 2003.
The Commissioners filed a respondents’ notice, on 24 August 2005, in which they sought to uphold the judge’s order on the additional ground that the tribunal had been correct, or (at least) entitled, to find that that CNP would not have made a claim for repayment in any transitional period and so had not, in fact, been prevented from an effective exercise of its Community rights by the manner in which regulation 29(1A) had been introduced. As I have explained, that contention had been rejected by the judge as a basis for dismissing the appeal from the tribunal.
The grounds advanced in the appellant’s notice and the respondents’ notice were overtaken by the decision of this Court in Fleming. As I have said, this Court held, by a majority, that Community law requires that regulation 29(1A) must be disapplied: the Commissioners cannot rely upon that regulation and so there is no basis upon which they can reject a repayment claim made under regulation 29(1). None of the points raised in the appellant’s notice or in the respondents’ notice affect that position.
That that was the effect of the decision of this Court in Fleming was, of course, recognised by the parties. Each has filed “post-Fleming” submissions in writing in order to address the position in the light of that decision. Put shortly, CNP’s submission is that its appeal must be allowed. Further consideration of the issues should be left to the House of Lords on the Commissioners’ pending appeal to that House in Fleming (permission for which has been granted by this Court). If necessary, a further appeal to the House of Lords in this case can be joined with that pending appeal.
The Commissioners’ primary submission is that the proper course is for this Court now to seek a preliminary ruling from the Court of Justice under article 234 of the EC Treaty on the question which they have formulated in these terms:
“Do the principles of effectiveness and of the protection of legitimate expectations preclude reliance on a new time limit which curtails, without providing a transitional period of at least six months duration as laid down in Case C-255/00 Grundig, the period within which payment of VAT incurred by way of input tax can be claimed where: (i) the right to deduct arose more than six years prior to the making of the claim, and/or (ii) the claim could not and/or would not have been made during such a transitional period, even if such a transitional period had been contained in the legislation.”
In advancing that primary submission the Commissioners point out that section 3(1) of the European Communities Act 1972 (as amended) provides that:
“3(1) For the purposes of all legal proceedings any question as to the meaning or effect of . . . any Community instrument shall be treated as a question of law and, if not referred to the European Court, be for determination as such in accordance with the principles laid down by and any relevant decision of the European Court or any court attached thereto.”
So, it is said, notwithstanding that this Court would otherwise be required under the doctrine of precedent ordinarily applicable under domestic law to follow and apply its earlier decision in Fleming, the overriding obligation, imposed by statute, is to determine the issues which arise on this appeal in accordance with the principles of Community law. If, as the Commissioners contend, the decision of the majority in Fleming was reached on the basis of a misunderstanding or misapplication of those principles, then this Court must refuse to follow it. And, in order to decide whether the decision in Fleming misapplies Community principles, the Court should seek a preliminary ruling under article 234 of the EC Treaty.
Should this Court seek a preliminary ruling from the Court of Justice
CNP oppose a reference to the Court of Justice. It has responded to the Commissioners’ post-Fleming submissions with a further skeleton argument. It points out: (i) that the principal question which the Commissioners seek to have referred – whether the principles of effectiveness and the protection of legitimate expectations preclude reliance on the new time limit imposed by regulation 29(1A) in the circumstances that that regulation curtails the period within which a claim can be made under regulation 29(1) without providing any transitional period – has been the subject of recent decision in this Court after a full review of the relevant Community law; (ii) that there has been no further development of those principles in the Court of Justice which calls for reconsideration of the conclusion reached in Fleming; (iii) that it was not suggested to this Court in Fleming – and, in particular, not suggested by the Commissioners – that that question could not be determined without a reference to the Court of Justice; and (iv) that it would be inappropriate for this Court now to refer that question to the Court of Justice for a preliminary ruling in circumstances when the same question is pending on an appeal before the House of Lords - an appeal for which permission was sought from and granted by this Court.
It is, indeed, a striking feature of this litigation that the Commissioners did not request this Court, before judgments were handed down in Fleming, to make the reference to the Court of Justice which they now say is necessary to reach a decision in the present appeal. Rather, the Commissioners chose (as domestic law enabled them to do) to seek the determination of the House of Lords on the principal question: whether the principle of effectiveness precludes any reliance on the new time limit imposed by regulation 29(1A). I understand that the Commissioners will invite the House, in the context of the appeal now pending in Fleming, to take the view that a reference is necessary in order to enable the House to reach its decision on that question. But that will be a matter for the House to consider in due course. It seems to me wrong in principle for this Court to take a course which would, in effect, pre-empt consideration by the House of the need for a reference in Fleming by referring, on this appeal, the same question to the Court of Justice in advance of that consideration. If the House of Lords take the view that there should be a reference in Fleming, there is nothing to be gained by an anticipatory, or parallel, reference in this appeal. If the House take the view that a reference is, in principle, unnecessary, they are unlikely to find it of assistance in their task of determining the Fleming appeal that there is already a pending reference of the same question in this appeal.
For those reasons, I am not persuaded that this is an appropriate case in which to refer to the Court of Justice, for a preliminary ruling under article 234 of the EC Treaty, the question whether the principle of effectiveness precludes any reliance on the new time limit imposed by regulation 29(1A). In the particular circumstances of this litigation – in which I include the Fleming case – I think we must decide that question for ourselves without the benefit of a preliminary ruling.
The question whether the principle of effectiveness precludes any reliance on the new time limit imposed by regulation 29(1A) is the first of the six issues which I have identified at paragraph [26] of this judgment. If the House of Lords determine that issue against the Commissioners – thereby upholding the decision of the majority of the Court of Appeal on that point – the remaining issues will no longer arise, either in Fleming or in the present case. If the House determine that issue in favour of the Commissioners it seems to me inevitable – or, at the least, very likely – that they will go on to consider the issues described under (2), (3) and (4) of that paragraph. I find it difficult to see how, if issue (1) were determined in favour of the Commissioners, the House could decide the appeal in Fleming without addressing those issues. But they will not need to consider the issues described under (5) and (6). Those issues do not arise in Fleming. I have considered, therefore, whether this Court should seek a preliminary ruling from the Court of Justice on those issues alone.
In my view that course is not open to us. It would be impossible, as it seems to me, for the Court of Justice to give a ruling on issues (5) and (6) save on the basis of a decision, or assumption, as to the questions raised by issues (1) to (4). I have explained why I take the view that it would not be appropriate, in the particular circumstances of this litigation, for this Court to refer issues (1) to (4) to the Court of Justice. It would be equally inappropriate to invite the Court of Justice to give a ruling on issues (5) and (6) on the basis of an assumption as to the answers which the House of Lords may give, in Fleming, to the question raised by issues (1) to (4).
I would refuse the Commissioners’ invitation to make a reference to the Court of Justice in this case.
What should this Court decide in the absence of a preliminary ruling
As an alternative to their primary submission that this is a case in which the Court should make a reference to the Court of Justice in advance of (and in order to inform) its own decision whether the majority in Fleming misapplied Community principles, the Commissioners submit that the appeal should be dismissed. There are, I think, two elements underpinning that submission.
The first is that this Court should decide for itself that the majority in Fleming were wrong in failing to recognise the need for what is described as the “nuanced approach” said to be required by the observations of Lord Nicholls of Birkenhead in Autologic Holdings plc v Inland Revenue Commissioners [2005] UKHL 54, [17]; [2006] 1 AC 118, 127B-C:
“[17] . . . Accordingly, if an inconsistency with directly enforceable Community law exists, formal statutory requirements must where necessary be disapplied or moulded to the extent needed to enable those requirements to be applied in a manner consistent with Community law. . . . So if the residence restriction is found to be inconsistent with Community law this provision will need adapting so as to give effect to the overriding Community rights. . . .”
The second element reflects the ground in the respondents’ notice. It is said that, having decided that the principle of effectiveness requires only that the three year time limit imposed by regulation 29(1A) should be disapplied to the extent necessary to ensure compliance with that principle, this Court should hold, on the basis of the findings of fact made by the tribunal in the present case, that the application of the three year time limit does not breach the principle of effectiveness, because there was no evidence that the amendments to regulation 29(1) introduced by paragraph (1A) “prevented the Appellant making any claim that it was minded to make at that time”. It is not, perhaps, clear whether the phrase “at that time” means “at the time when the three year time limit was introduced”; or “within the time limited by transitional provisions which ought have been included when the three year time limit was introduced” – it being assumed that those provisions would have allowed a reasonable time for the making of existing claims which would otherwise have been excluded by the three year time limit. If the latter, the unspoken assumption is that whatever time ought properly to have been limited by transitional provisions – and, in that context, the Grundig case provides some support for the view that that time could have been no more than six months – CNP would have chosen not to make a claim within that time. It is said that: “Conde Nast has slept on its rights and its claim now falls squarely within the lawful application of the time limit imposed by Regulation 29(1A)”.
I am content to assume that there may be circumstances in which the obligation imposed on courts by section 3(1) of the European Communities Act 1972 would require this Court to refuse to follow its own earlier decision as to the meaning and effect of a Community instrument – including, in the present context, the effect of a judgment of the Court of Justice. Those circumstances would, I think, include a case in which the judgment of the Court of Justice under consideration by this Court in the earlier case had been the subject of further consideration – and consequent interpretation, explanation or qualification - by the Court of Justice in a later judgment. But, as it seems to me, one constitution in this Court should not substitute its own view as to the effect of a judgment of the Court of Justice for the view which has been reached by an another constitution in this Court in an earlier case on consideration of the same judgment in circumstances in which there has been no opportunity for the Court of Justice to review that judgment. In those circumstances, if persuaded that there are strong grounds for thinking that the earlier decision is wrong (as a matter of Community law) this Court may think it right to refer the point to the Court of Justice for a preliminary ruling. Or it may follow the earlier decision and give permission to appeal. But it should not refuse to follow the earlier decision merely because, on the same material and the same arguments, it is satisfied that a different conclusion should have been reached.
The need for a disciplined adherence to precedent in a comparable (but not precisely analogous) field was emphasised by Lord Bingham of Cornhill (with whom the other six members of the House expressly agreed on this point) at paragraphs [40] to [45] of his speech in Kay and others v Lambeth London Borough Council, Leeds City Council v Price and others [2006] UKHL 10, [40]-[45]; [2006] 2 WLR 570, 598B-591E. After referring to the observation of Lord Hailsham of St Marylebone, Lord Chancellor, in Broome v Cassell & Co Ltd [1972] AC 1027, 1054, that “in legal matters, some degree of certainty is at least as valuable a part of justice as perfection”, Lord Bingham said this, at paragraph [43] (ibid, 590F):
“[43] . . . That degree of certainty is best achieved by adhering, even in the Convention context, to our rules of precedent. It will of course be the duty of judges to review Convention arguments addressed to them, and if they consider a binding precedent to be, or possibly to be, inconsistent with Strasbourg authority, they may express their views and give leave to appeal, as the Court of Appeal did here. Leap-frog appeals may be appropriate. In this way, in my opinion, they discharge their duty under the 1998 Act. But they should follow the binding precedent, as again the Court of Appeal did here.”
In the present case, therefore, I do not think that we are free to refuse to follow the decision of the majority of the Court of Appeal in Fleming on what I have described as issue (1). The reasons why the view taken by the majority of the Court in that case may be wrong were powerfully expressed in the judgment of Lady Justice Arden. In the circumstances that the issue will be considered by the House of Lords in the pending appeal, I do not think that any useful purpose will be served by extending the length of this judgment with my own contribution to that debate. I will add only that I have found Lady Justice Arden’s analysis persuasive; and that, if she is correct on issue (1), then – as it seems to me – she must be right, also, in her view that the approach described under issue (4) is the correct approach to the question what transitional provisions are to be read into regulation 29(1A). I agree with her view (and that of Lord Justice Ward, obiter) that the approach described under issue (3) should be rejected.
On the basis that we are not free to refuse to follow the decision of the majority in Fleming, it is unnecessary to decide issues (5) and (6). Those issues do not arise. But, for completeness – and because the view which we take on those issues may be relevant in the context of any application for permission to appeal from the order which we make on this appeal - I will address them shortly.
Issue (5): whether CNP’s failure to prove that it would have taken advantage of transitional provisions, had they been included in regulation 29(1A) when first introduced, is fatal to its claim
I agree with Mr Justice Warren, for the reasons which he gives in paragraph [61] of his judgment, that issue (5) arises only if the correct approach would otherwise be that described under issue (4). As he put it: “if the first approach [described under issue (3)] is correct, the claim has, in fact, to be made within the reasonable period and no question arises about what the taxpayer might have done since we will know what he in fact has done”. On the basis that the correct approach would otherwise be that described under issue (4), the right to reclaim under-deducted input tax is not to be taken away until the taxable person has had a reasonable period during which he could have been expected to exercise that right.
As the judge pointed out, the fact that a taxable person did not have the opportunity to exercise the right to reclaim under-deducted input tax from 1 May 1997 to 11 July 2002 “is entirely the fault of the Member State in failing to comply with its Community law obligations”. To require such a person to prove that he would have exercised the right during that period, or during some part of that period, or during such part of that period as would have fallen within proper transitional provisions if such provisions had been included in the regulation when it was introduced, would be inconsistent with the principle of effectiveness. It would be excessively difficult for the taxable person to exercise his Community right to reclaim under-deducted input tax if he were required “to prove something which might, in its nature, be very difficult for him to prove, namely that he would have exercised his right had a transitional period been included”.
It follows that I agree with the judge that the tribunal was wrong to hold that the absence of evidence that CNP would have taken advantage of whatever transitional provisions might, properly, have been included in regulation 29(1A) if at the time when the regulation was introduced, transitional provisions had been included was an answer to the repayment claim in the present case.
Issue (6): whether the transitional period to be read into regulation 29(1A) continued until 30 June 2003
I turn, therefore, to issue (6): whether, if the correct approach is that described under issue (4) above, the transitional provisions which are to be read into regulation 29(1A) must be taken to provide that the period within which claims must be brought under those provisions did not expire before 30 June 2003 - that being the end of the period allowed by non-statutory transitional provisions introduced to deal with the comparable problem which arose in relation to claims under section 80 VATA 1994 in respect of over-declaration of output tax. As I have said, that approach would enable CNP to succeed: its claim was lodged on 27 June 2003.
In order to understand the argument advanced on behalf of CNP on this point it is necessary to have in mind the following facts, which are set out more fully in the judgment of Lady Justice Arden in Fleming and in the judgment of Mr Justice Warren in the present case:
Prior to the decision of this Court in the conjoined appeals of Marks and Spencer plc v Commissioners of Customs & Excise and University of Sussex v Customs & Excise Commissioners [2003] EWCA Civ 1448; [2004] STC 1 ("the University of Sussex case"), the Commissioners took the view that claims for the repayment of under-deducted input tax fell within section 80 VATA 1994, rather than regulation 29: see Business Brief 4/02, issued on 22 February 2002. Section 80 VATA 1994 provides for taxable persons to claim overpaid VAT subject to an express time limit of three years. That period was reduced from the previous period of six years (and in some cases six years from the date of the discovery of the mistake) by a resolution of Parliament passed under the Provisional Collection of Taxes Act 1968. The change was made on 4 December 1996 with effect from 18 July 1996.
Following, and to take account of, the decisions of the Court of Justice in the Marks & Spencer and Grundig cases, to which I have referred earlier in this judgment, the Commissioners issued statements of practice in the form of two Business Briefs. In the first - Business Brief 22/02 of 5 August 2002 - it was stated that, subject to various conditions, if over-payments of VAT had been made before 4 December 1996 and claims to recover such over-payments had been discovered or made prior to 31 March 1997, a claim for repayment could be submitted or resubmitted up to 31 March 2003. In the second - Business Brief 27/02 of 7 October 2002 - the date by which the claimant must have become aware of the existence of the claim was extended to 30 June 1997 and the claim period was extended to 30 June 2003.
As I have said, at the time when those Business Briefs were issued, the Commissioners’ position was that section 80 VATA 1994 applied both to claims for repayment of over-declared output tax and to claims for repayment of under-deducted input tax. It was the Commissioners’ view that claims under either head were brought under section 80 VATA 1994. Although, in University of Sussex v Customs and Excise Commissioners [2001] STC 1495, Mr Justice Neuberger had held that view to be wrong, it was not until his decision on that point was upheld in this Court, in October 2003, that the Commissioners accepted that claims for repayment of under-deducted input tax fell under regulation 29 and not under section 80 VATA 1994. Indeed, CNP’s claim, when lodged on 27 June 2003, was, at first, treated as a claim under section 80 VATA 1994 – as appears from the letter written by Customs and Excise Business Services and Taxes on behalf of the Commissioners on 30 July 2003. On that basis the claim would have been within the transitional period allowed by Business Brief 27/02. It was not until the decision letter of 31 October 2003, written shortly after this Court had handed down its judgments in the University of Sussex case on 21 October 2003, that the Commissioners took the point that the Business Briefs did not apply. They wrote:
“The Court of Appeal has now held that a claim for input tax not claimed previously is not one made under section 80 of the VAT Act 1994 but under regulation 29 of the VAT Regulations 1995. As claims made under regulation 29 were not capped until May 1997, the University’s claim was not capped and is payable in full.
So, as the claim submitted for Conde Nast Publications Limited relates to input tax underclaimed, then it does not fall within the scope of s80 VATA and so the M&S Business Briefs cannot apply – the claim is capped under normal rules relating to Regn 29(1A).”
Put shortly, the submission made on behalf of CNP is that, if the approach described under issue (4) is correct – which is the hypothesis upon which this point arises – it is necessary to recognise: (i) that a taxable person with a claim to recover under-deducted input tax which had accrued before 4 December 1996 (and of which he had become aware before 31 March 1997) would not know, until 11 July 2002 (when the Court of Justice delivered its judgments in the Marks & Spencer case), that the Commissioners were required to meet that claim notwithstanding the absence of transitional provisions; (ii) that (notwithstanding the judgment of Mr Justice Neuberger in University of Sussex v Customs and Excise Commissioners), the Commissioners continued to take the position that the repayment claim fell to be made under section 80 VATA 1994; (iii) that on 5 August 2002, before a reasonable period had expired since 11 July 2002 in which to make that claim, the Commissioners had issued Business Brief 22/02, stating that capped repayment claims under section 80 VATA 1994 would be met if made before 31 March 2003; (iv) that, on 7 October 2002, before the period allowed under Business Brief 22/02 had expired, the Commissioners had issued Business Brief 27/02, extending the time for making capped repayment claims under section 80 VATA 1994 to 30 June 2003; and (v) that until 21 October 2003 the Commissioners’ position was and remained that claims to recover under-deducted input tax which had accrued before 4 December 1996 were capped repayment claims under section 80 VATA 1994 to which the Business Briefs applied. With those factors in mind, it is impossible to conclude that a reasonable time within which to make a claim to recover under-deducted input tax under transitional provisions had expired before 30 June 2003.
The judge rejected that submission. At paragraph [54] of his judgment he said this:
“[54] . . . I see no reason why, as a matter of Community law, the taxpayer should be entitled to see replicated the scheme which the Commissioners, in the Business Briefs, adopted in relation to section 80.”
He identified as “the sole question”: what was a reasonable period from 5 August 2002 during which a regulation 29 claim (not subject to the new time limit introduced by regulation 29(1A)) could be made. Without defining that period, he held that it must have expired before 30 June 2003.
In my view the judge was wrong to reach that conclusion. I agree that there was no reason why the taxable person was “entitled to see replicated the scheme which the Commissioners, in the Business Briefs, adopted in relation to section 80 [VATA 1994]” – in the sense that, had the Commissioners proposed or adopted a scheme in relation to regulation 29, they would have been required to replicate the scheme in the Business Briefs. But that, as it seems to me, misses the point. The Commissioners did not propose or adopt a scheme in relation to regulation 29 claims. Had they done so, they might or might not have replicated the scheme in the Business Briefs. The probability must be that they would have done so; but they were not required to do so. The fact is that, until October 2003, the Commissioners treated claims to repayment of under-deducted input tax as claims under section 80 VATA 1994; and so, for that reason, treated those claims as claims to which the Business Briefs did apply. A properly advised taxable person might have taken the view, following the decision of Mr Justice Neuberger in University of Sussex v Customs and Excise Commissioners, that the Commissioners were wrong to take the position that they did. But it cannot be said that it would have unreasonable for a properly advised person to take the view that, given the position taken by the Commissioners until October 2003, a capped claim to repayment of under-deducted input tax made before 30 June 2003 would be met. For the Commissioners, now, to refuse to meet that claim – in reliance on their belated acceptance that the position which they had taken until October 2003 was mistaken – would be inconsistent with the need, under Community principles, to protect legitimate expectations.
Conclusion
I would allow this appeal on the ground that, in Fleming, this Court decided the first issue against the Commissioners. As I have said, I do not think that we are free to refuse to follow that decision. But I would add that, even if we were to refuse to follow that decision, I would have held that the appeal should be allowed (at least, in relation to the claim to input tax prior to 4 December 1996) on the ground that, although the judge was correct to adopt the approach described under issue (4), he was wrong to reach the conclusion which he did on issue (6).
Lady Justice Arden:
I agree. In para. 52 of my judgment in the Fleming case, I said that I agreed with Warren J’s view that the transitional period should not be further postponed simply because Customs & Excise took the view that claims could not be made under regulation 29, only under section 80 of VATA 1994. But it is clear from the remainder of that para. that I contemplated that taxpayers might make protective claims under regulation 29 or claims under section 80. I think it must therefore follow from para. 52 read as a whole that the first sentence refers to postponement of the expiry of the transitional period for regulation 29 claims beyond 30 June 1993 until after the Commissioners changed their position in relation to regulation 29 claims on 21 October 2003.
Lady Justice Smith:
I also agree.