The Commissioners for HM Revenue and Customs v GMAC UK PLC (formerly General Motors Acceptance Corporation (UK) PLC and British Telecommunications v The Commissioners for HM Revenue and Customs: [2012] UKUT 279 (TCC) PLC) [2012] UKUT 279 (TCC)

UPPER TRIBUNAL
TAX AND CHANCERY CHAMBER
[2012] UKUT 279 (TCC)Case No FTC/52/2010
Between
The Commissioners for HM Revenue and CustomsAppellantGMAC UK PLC (formerly General Motors Acceptance Corporation (UK) PLC and British Telecommunications v The Commissioners for HM Revenue and CustomsRespondent
Before
Mr Justice Warren, Chamber PresidentJudge Charles HellierDate 3 August 2012Category: Tax
[1]There are two matters before us. The first is an appeal against the decision of the Tax Chamber (Judge Wallace and Miss S C O’Neil) (“the Decision” and “the Tribunal”) which allowed the appeal of General Motors Acceptance Corporation (UK) PLC (“GMAC’) against HMRC's refusal of VAT bad debt relief in relation to supplies of cars made on hire-purchase (“hp”) terms between 1978 and 1997. The second is the determination of 3 preliminary issues within a tax appeal by British Telecommunications PLC (“BT”) currently proceeding in the Tax Chamber. These issues have been agreed between the parties and have been referred to the Tax and Chancery Chamber pursuant to the relevant procedural rules. The preliminary issues raise some of the questions which arise in the appeal. It has been convenient for the appeal and the preliminary issues to be dealt with in one continuous hearing. In both cases, Paul Lasok QC and Eleni Mitrophanous appear for HMRC; in the appeal, Roderick Cordara QC appears for GMAC together with Jessica Wells and in the BT matter, he appears for BT together with Lyndsey Frawley.[2]The appeal has been before us on a previous occasion when HMRC sought a reference to the ECJ (as we will refer to the European Court of Justice as it used to be called and to the Court of Justice, as it is now called, being part of the Court of Justice of the European Union) of a number of preliminary points. We declined to make a reference at that stage. We considered that if a reference were to prove necessary at all, the appropriate time to make one would be after the hearing of the appeal when we would have a better idea of precisely what question or questions should be referred. We will return to this aspect of the appeal in due course. If a reference is needed in the appeal, a reference may be needed in the BT case too. The GMAC appeal[3]We deal with the appeal first. For the purposes of introducing the appeal, the facts can be taken very shortly. GMAC supplied cars on hp terms. The customer would agree to pay for the car by instalments. If the customer defaulted, GMAC would repossess the car and sell it at auction. The auction proceeds reduced the balance of the instalments due from the customer. We take by way of example the same example as that provided by the Tribunal at [30] of the Decision (in turn taken from the decision of Field J in earlier litigation to which we will come).a. a customer agrees to pay £15,000 (plus interest) over the period of an hp agreement; 4b. the customer stops paying when £10,000 is outstanding (having paid only £5,000);c. the car is repossessed and sold at auction for £6,500;d. under the terms of the hp agreement the £6,500 is set against the outstanding £10,000 and the customer owes £3,500;e. the customer then does not pay the £3,500.[4]At the time of the repossession, GMAC would already have accounted for VAT on the full (VAT-inclusive) purchase price of £15,000. It is entitled to adjust its accounts in respect of the reduction in the consideration, that is to say the amount at which the car is resold at auction, pursuant to Regulation 38 of the Value Added Tax Regulations 1995 (“the VAT Regulations 1995”). That adjustment does not, however, cover the amount remaining outstanding, and unpaid, by the customer, that is to say the £3,500. The appeal concerns whether GMAC is entitled to some form of VAT bad debt relief in respect of the £3,500 it does not receive. II. The Issues[5]The appeal raises three issues which can be described as the Compatibility Issue, the Windfall Issue and the Time-limit Issue which we now describe. A. The Compatibility Issue[6]During the period 1978 to 1997, the domestic VAT bad debt relief provisions imposed one or both of two conditions which had to be satisfied before bad debt relief was available (see paragraphs 31-37 below).a. the Property Condition: that on a supply of goods the property had passed; andb. from 1978 to 1990 only, the Insolvency Condition which required that the debtor was formally insolvent and, for part of that period, that the taxpayer had proved in the insolvency.[7]Under the hp agreements in question, property in a repossessed car had not passed to the customer prior to the repossession. GMAC was therefore unable to satisfy the Property Condition at any time in the period. Further, since GMAC did not in all cases itself wish to commence insolvency proceedings against a defaulting customer, it did not, in many cases, satisfy the Insolvency Condition. Moreover, in some cases, the amount outstanding was less than the relevant bankruptcy or insolvency limit so that bankruptcy or winding-up proceedings could not be instituted. The result was that GMAC was not entitled in any of these cases to bad debt relief under UK domestic legislation. But GMAC’s case now is that the Property Condition and the Insolvency Condition were incompatible with the Sixth VAT Directive 77/388/EC (“the Directive”). It asserts that the Directive 5 provided a directly enforceable right to relief for non-payment on which it can now rely.[8]On this issue the Tribunal held that the Insolvency Condition and the Property Condition for bad debt relief in the domestic legislation were incompatible with the Directive and fell to that extent to be disapplied. B. The Windfall Issue[9]The supply of a car on hp terms is treated for VAT purposes as a supply of goods. As noted above, VAT became payable on the supply of the car on the total amount due (excluding the finance charge). Thus in the example VAT became payable on £15,000 on the consumer taking possession of the car (we use VAT “on” an amount as shorthand for “in respect of" that amount).[10]If the car was repossessed and sold, then, by virtue of the Value Added Tax (Cars) Order 1992 (“the Cars Order”) and similar predecessor provisions (see paragraph 28 below), the sale was, for UK domestic purposes, not treated as a supply: it was "desupplied". Thus the £6,500 in the example received at the auction for the sale of the car was not subject to VAT.[11]Since 1 January 1990, successive regulations (Regulation 7 VAT (Accounting and Records) Regulations 1989 and then Regulation 38 of the VAT Regulations 1995) have provided for a reduction in the amount of the consideration for a supply which is subject to VAT when there was a decrease in the consideration. Prior to 1990, there was no UK domestic enactment of Article 11C(1) of the Directive.[12]It had always been accepted by HMRC that if there was a consensual termination of an hp agreement following which the car was sold (eg for £6,500 as in the example assuming a consensual termination rather than a repossession) and the outstanding debt reduced by the sale proceeds (as in the example to £3,500), Regulation 38 applied, with the result that GMAC was to be treated as having made the hp supply for a reduced consideration (in the example, for £8,500 being the original consideration of £15,000 less the reduction of £6,500), but until the decision of Field J in C&E Commissioners v GMAC [2004] STC 577 (“GMAC No 1”), they did not accept that the same treatment applied where the customer defaulted and under the terms of the agreement the car was repossessed with the remaining consideration due being reduced in the same way. Applying what Field J had held in GMAC No 1, to the example the setting of the £6,500 auction proceeds against the original £15,000 price of the car was a reduction in consideration which, contrary to HMRC’s argument, engaged Regulation 38.[13]He also held that the Cars Order applied as well. Taken together, the Cars Order and Regulation 38 produced, in the case of a default, what HMRC rightly say is a surprising result since, at this stage, if the customer in fact paid the remaining balance, GMAC would have received £15,000 (£8,500 from the customer and £6,500 auction proceeds) but would pay VAT on only £8,500. On the consumption of a total supply worth £15,000, VAT is payable only on £8,500. 6 This is the "windfall". HMRC say that it is inconsistent with the result intended by the Directive.[14]Giving GMAC bad debt relief if the customer did not pay the outstanding £3,500 would, as Mr. Lasok put it, rub salt into this wound. If bad debt relief were available on the default of £3,500, then GMAC (having received £11,500) would be liable to account for VAT in respect of only £5,000.[15]HMRC say that GMAC cannot at the same time benefit from the windfall which arises as a result of a combination of the Cars Order and Regulation 38, and then seek the benefit of the direct application of the Directive further to reduce its tax liability. GMAC must take the burden of the Directive (the loss of the windfall) if it wants the benefit of bad debt relief.[16]On this issue the Tribunal held that GMAC’s rights to relief under the directive were not defeated or curtailed by the result of a combination of that relief with other domestic provisions. C. The Time-limit Issue[17]The hp supplies in question took place between 1978 and 1997. The bad debts arose between 1978 and sometime after 1997 (since hp agreements may last several years and defaults may occur some time after the agreements are made) but GMAC did not make its claim for bad debt relief until 20 February 2006. HMRC say that the part of GMAC’s claim which relates to the period 1978 to 1989/1990 is time-barred as a result of section 39(5) the Finance Act 1997 (“FA 1997”) which provided that domestic claims for bad debt relief under section 22 of the Value Added Tax Act 1983(“VATA 1983”) (applying to supplies made before 26 July 1990) could no longer be made after 19 March 1997.[18]On this issue the Tribunal held that section 39(5) retrospectively curtailed GMAC’s rights under the Directive without an adequate transitional period and accordingly that it should be disapplied so that no time bar applied to the claims in relation to supplies in those years. It is part of GMAC’s case that section 39 is of no relevance. Its claim has never been under the domestic provisions; it seeks simply to assert its directly enforceable rights under the Directive to which no domestic time limits are relevant.[19]HMRC seeks to raise a new point which was not raised before the Tribunal and did not feature in its notice of appeal. HMRC now submit that as a result of the judgment of the General Court of the European Court of Justice, Case T-433/10P Allen v Commission 14 December 2011 (“Allen”), GMAC had no exercisable rights under EU law at the time when it made the claim giving rise to the appeal to the Tribunal and now to us. We will consider this new point, and whether HMRC should be allowed to raise it in the appeal, later in this decision. We understand that there is to be no review of the decision by the Court of Justice. III. The legislative history.

(a) EU provisions

[20]The Directive was adopted on 17 May 1977. Article 11A(1) of the Directive provided for the taxable amount to be everything which constitutes consideration from the customer.[21]Article 11C(1) provided:
"In the case of cancellation, refusal or total or partial non-payment, or where the price is reduced after the supply takes place, the taxable amount shall be reduced accordingly under conditions which shall be determined by the Member States. However, in the case of total or partial non-payment, Member States may derogate from this rule."
[22]The first part of this provision permits conditions to be placed upon the reduction in the consideration. The second part permits member states to derogate but the derogation is limited to the case of total or partial non-payment.[23]The distinction between a condition for the application of a provision and a derogation from the rule is not always easy to discern although it is clear that the article requires a distinction to be drawn. It seems likely that the difference is between an administrative condition which relates to something which must be done to claim relief, and a denial of the right in certain circumstances (see the Advocate General in EC v Germany C-427/98 [2003] STC 201).[24]In 1994 Article 26a was added to the Sixth Directive. This contained specific rules for arrangements applicable to second-hand goods. The rules were designed to deal with the anomaly which previously arose when a taxable dealer sold secondhand goods on which, because his supplier was not a taxable person, he was not entitled to an input tax deduction although output tax was charged on the full consideration for his supply. This resulted in double charging of VAT on the same economic consumption. It also resulted in a distortion in the market since a nontaxable person selling in the same market did not have to bear VAT. Article 26a provided for VAT to be charged only on the dealer’s profit margin. (b) UK domestic provisions.[25]The UK domestic legislation which had the effect of implementing Article 11C(1) contained two sets of provisions. The first applied where there was a reduction in the consideration and the other applied, and gave relief, where there was total or partial non-payment (ie bad debt relief). Prior to GMAC No 1 it had been HMRC’s position that the former only applied where there was a consensual termination of an hp agreement and not where there was a default followed by repossession and sale.

(i) reduction in consideration 8

[26]The domestic provisions relating to reduction in consideration were, from 1995, to be found in Regulation 38 (read with Regulation 24) of the VAT Regulations 1995. They provided that where there was a decrease in the consideration for a supply which included an amount of VAT then the taxpayer should adjust his VAT account by making a negative entry for the relevant amount of VAT. For these purposes, a decrease in consideration was recognised only if it was evidenced by a credit note or other document having a similar effect. Similar regulations had applied in the period 1990 to 1995.[27]But, at times relevant to the appeal - and also before the adoption of the Directive - the UK also provided relief on the sale of second-hand cars. This relief has been provided in materially identical terms in successive statutory instruments. We will refer to the provisions of the Value Added Tax (Cars) Order 1992 (“the Cars Order”) although we have been referred also to the VAT (Cars Order) 1980. Regulation 8 of the Cars Order effectively provided that the VAT when a car dealer sold a second-hand car was to be charged on an amount equal to the dealer's margin. This latter provision was consistent with Article 26a of the Sixth Directive.[28]The Cars Order also dealt with repossessed cars. Regulation 4 provided for the desupply of repossessed cars:
"(1) Each of the following descriptions of transactions shall be treated as neither a supply of goods nor a supply of services -- (a) the disposal of a used motor car by a person who repossessed it under the terms of a finance agreement, where the motor car is in the same condition it was when it was repossessed….."
[29]The desupply of repossessed second-hand cars was not specifically authorised by the Directive before or after the adoption in 1994 of Art 26a but it may have fallen within the latitude given to Member States in relation to the implementation of previous directives. The effect of the Cars Orders was, Mr Lasok told us, considered by HMRC to be the same in economic terms as reducing the consideration for the initial supply. This was because, ignoring the effect of Regulation 38 which HMRC had contended did not apply, if the car was sold at auction (remaining with the example) for £6,500, VAT was collected in respect of the £15,000 on the original supply of the car, and that was the economic value of the consumption of the car: without the Cars Order there would (leaving aside Regulation 38) have been double taxation since tax would have been levied in respect of £21,500 (ie £15,000 plus £6,500).[30]The successive statutory instruments, including the Cars Order itself, were not amended following the increased harmonisation required by the Directive or even after the insertion of Article 26a. But, after the decision in GMAC No 1 (which held that relief was available both under Regulation 38 and under the Cars Order - giving rise to the windfall), the Cars Order was amended in 2006 by disapplying the desupply when a car was repossessed under a finance agreement (whether pursuant to a default or by agreement). 9 (ii) bad debt relief[31]For supplies made between 2 October 1978 and 19 March 1997 (and for supplies made prior to 2 October 1978 giving rise to a bad debt after that date) bad debt relief in domestic legislation was available under what was called the Old Scheme or the New Scheme. Claims could be made under the Old Scheme in relation to supplies made between 2 October 1978 and 26 July 1990, and under the New Scheme for supplies made between 1 April 1989 and 19 March 1997. There was thus an overlap period, between 1 April 1989 and 26 July 1990 when a claim could be made under either (but not both) schemes. The Old Scheme contained the Property Condition and the Insolvency Condition (in one of two differing forms); the New Scheme contained the Property Condition only. There were other differences between the two schemes such as the time which had to elapse from the date of the supply before a claim could be made and whether the debt had to be written off in the taxpayer’s accounts. The following table sets out the position: Period Scheme Property Condition Insolvency Condition 2 October 1978 to 31 March 1986 Old Yes Yes: Mark I 31 March 1986 to 1 April 1989 Old Yes Yes: Mark II 1 April 1989 to 26 July 1990 Old -and- New Yes Yes: Mark II No 26 July 1990 to 18 March 1997 New Yes No The Old Scheme[32]This was first enacted by section 12 of the Finance Act 1978 (“FA 1978”), and then re-enacted in section 22 VATA 1983. We will refer to these as “section 12” and “section 22” and will do so without distinction. Section 22 provided:
"(1) Where - (a) a person has supplied goods or services for consideration in money and has accounted for and paid tax on that supply; and 10 (b) the person liable to pay any outstanding amount of the consideration has become insolvent, then, subject to subsection (2) and to regulations under subsection (3) below, the first mentioned person shall be entitled, on making a claim to the Commissioners, to a refund of the amount of tax chargeable by reference to the outstanding amount. (2) A person shall not be entitled to a refund under this section unless - (a) he has proved in the insolvency and the amount for which he has proved is the outstanding amount of the consideration less the amount of his claim; (b) the value of the supply does not exceed its open market value; and (c) in the case of a supply of goods, the property in the goods has passed to the person to whom they were supplied…"
Under section 22(4) an individual was insolvent for the purposes of the section if he was adjudicated bankrupt or if the court made an order for the administration in bankruptcy of his estate. A company was insolvent if it was subject to a creditor’s voluntary winding-up or a compulsory winding-up and “the circumstances are such that the company is unable to pay its debts”.[33]Section 22(1)(b) and (2)(a) contained the Mark I Insolvency Condition – a requirement that the creditor had become insolvent and that the debtor has proved in the insolvency. An amendment to the section – to create the Mark II Insolvency Condition – by the Finance Act 1985 which applied to persons becoming insolvent after 31 March 1986, provided a reformulation (with a slight tightening) of the definition of insolvency (bankruptcy or liquidation or administration with negative net assets) but removed the requirement that the debtor prove in the insolvency. The New Scheme[34]Section 11 of the Finance Act 1990 (“FA 1990”) created the New Scheme and provided that the Old (section 22) Scheme should not apply to supplies made after 26 July 1990. It also provided that a claim under the New Scheme could not be made if a claim had been made under the Old Scheme. A taxpayer had therefore to elect between the two schemes in relation to supplies made in the overlap period.[35]Regulations were made under the provisions of the enabling sections. These are set out below in section ‘VI. Time Limits’.[36]The New Scheme applied to supplies made after 1 April 1989. It was re-enacted in section 36of the Value Added Tax Act 1994 (“VATA 1994”) with, for present purposes, only one material change (that in subs (1)(c): see below). Section 36 provided: "(1) Subsection (2) below applies where – 11(a) a person has supplied goods or services for a consideration in money and has accounted for and paid VAT on the supply,(b) the whole or any part of the consideration for the supply has been written off in his accounts as a bad debt, and(c) a period of six 6 months [replacing the period of two years in the provisions of section 11 FA 1990] (beginning with the date of the supply) has elapsed. (2) Subject to the following provisions of this section and to regulations under it the person shall be entitled, on making a claim to the Commissioners, to a refund of the amount of VAT chargeable by reference to the outstanding amount. [Subsection (3) provides a definition of the "outstanding amount".] (4) A person shall not be entitled to a refund under subsection (2) above unless - (a) the value of the supply is equal to or less than its open market value; and (b) in the case of the supply of goods, the property in the goods has passed to the person to whom they were supplied or to a person deriving title from, through or under that person....” Subsection (5) provided for the making of Regulations. We refer to these later in relation to the time limit issue.[37]In 1994 VATA 1983 was replaced by VATA 1994. Schedule 13 VATA 1994 provided Transitional Provisions and Savings relating principally to the continuity of VAT law which are referred to below. It contained specific provision in paragraph 9 relating to bad debt relief.[38]We now turn to address the three issues described in section II above. IV. The Compatibility Issue[39]In Goldsmiths (Jewellers) Ltd v Customs & Excise Commissioners C-330/95 [1997] STC 1073, (“Goldsmiths”) the appellant (which had suffered from a failure by the counter-party to a contract to provide certain non-monetary consideration for a supply) challenged the limitation of the UK domestic bad debt provisions to supplies which had been made for consideration in money. The ECJ noted that Article 11A(1)(a) embodied “one of the fundamental principles of the directive” and said: “16. In accordance with that principle, the first subparagraph of art 11C(1) of the Sixth Directive defines the cases in which the member states are required to ensure that the taxable amount is reduced accordingly, under such conditions which are to be determined by the member states themselves. That provision therefore requires member states to reduce the taxable amount and, consequently the amount of VAT payable by the taxable person whenever, 12 after a transaction has been concluded, part of the consideration has not been received by the taxable person.[17]Nevertheless, the second subparagraph of article 11C(1) of the Sixth Directive permits member states to derogate from the above-mentioned rule in the case of total or partial non-payment.[18]The power to derogate, which is strictly limited to the latter situation, is based on the notion that in certain circumstances and because of the legal situation prevailing in the member state concerned, non payment of consideration may be difficult to establish or may be only temporary. It follows that the exercise of that power must be justified if the measures taken by the member states for its implementation are not to undermine the object of fiscal harmonisation pursued by the Sixth Directive.[19]With regard to section 11 of the 1990 Act, the United Kingdom seeks to justify the refusal to refund the tax on the ground that there is a greater risk of evasion where the underpaid consideration is not expressed in money.[20]That justification is unacceptable for two reasons.”[40]The two reasons given by the ECJ were, briefly, these:a. Measures intended “to prevent tax evasion or avoidance” (see [21]) may not in principle derogate from the basis for charging VAT “except within the limits strictly necessary for achieving that specific aim”. The exclusion by the UK legislation of all non-money transactions from the refund of VAT went beyond what was strictly necessary.b. No distinction was to be drawn between money and other consideration in Article 11A(1)(a) or Article 11C(1). The distinction drawn in the treatment of the different transactions under UK legislation was discriminatory and restricted traders from choosing the contract which they consider to be most suitable to their economic interests.[41]Thus the power to derogate is not unfettered. The second part of Article 11C(1) provides a power of derogation from a fundamental principle, namely that the chargeable consideration is what is received by the supplier. Exceptions from the operation of fundamental principles are to be narrowly circumscribed so that the exercise of the power of derogation must be justified. That means, at the very least, that the national measure in derogation of the ordinary principle must be directed at a permissible objective. As Mr Cordara puts it, a Member State does not have “a free-wheeling discretion” as to whether to operate the provision, referring not only to Goldsmiths but to other cases including Minister Finansów v Kraft Foods Polska SA (Case C-588/10).[42]What objectives are permissible in this context? The Advocate General in Goldsmiths at [19] of his opinion said that the “member states must use the discretion so provided in such a way as to comply with the aims of harmonisation 13 and the underlying principles of the legislation.” Those principles include the principle of fiscal neutrality, preventing evasion and avoidance of tax and the fundamental principles of Article 11. The ECJ itself, at [18] of the judgment, described the power to derogate as based on the notion that non-payment may be difficult to establish and may be temporary, from which it followed that the derogation had to be justified. The power to derogate is focused on the legitimate aim of the Member State to ensure that there has been a total or partial nonpayment of consideration. It seems to us that any derogation must, in accordance with that approach, be justified by reference to that factor.[43]Mr Lasok suggested that the ECJ, in stating the basis of the derogation, may not have intended exhaustively to specify the permissible aims of the derogation. That indeed seems to be the case at least in so far as the ECJ regards the prevention of evasion or avoidance of VAT as a permissible aim, an aim which is not expressly encompassed in its identification of uncertain and temporary non-payment. However Mr Lasok did not suggest that any other objective could be found other than respect for the fundamental principles of legal certainty, legitimate expectation and fiscal neutrality including non-discrimination.[44]Does the derogation engage considerations of proportionality? Mr Lasok submits that considerations of proportionality so far as concerns Article 11C(1) are limited to measures directed against evasion and avoidance. He therefore submits that proportionality has no part to play in relation to the Property Condition and the Insolvency Condition adopted by the UK. He submits in effect that a derogation is justified if it would not be unreasonable for the Member State to regard it as a measure aimed at a permissible purpose, a formulation not without considerable difficulty in the light of the need to identify what is or is not reasonable and from whose perspective.[45]In the context of proportionality, the discussion of justification by the AG in Goldsmiths at [12] to [26] repays careful reading. We note [17]:
“That is why the derogation has to be justified. And it will be justified, to my mind, only if it is made in accordance with the principles and rules of Community law relevant to the legislation in the case in point……”
[46]At [18] and [19], the AG referred to different powers to derogate, which had been considered in the cases, concluding that, however different those cases were from the case of Goldsmiths itself, Members States were given a power “to be exercised within the bounds of, and thus in keeping with, the Sixth Directive. It is clear from the case law that member states must use the discretion so provided in such a way as to comply with the aims of harmonisation and the underlying principles of the legislation”. Then, at [20] he said this:
“……Attention must be given, from the specific angle of proportionality, to the question of whether there is any justification for the derogation introduced 14 into the United Kingdom system which, for the purposes of tax exemption, discriminates between money transactions and barter transactions.”
[47]We see no reason to question the AG’s approach. There is certainly nothing in the judgment of the ECJ or in any other case law of which we are aware that casts doubt on what he said. It is to be noted that, at this stage of his Opinion, the AG was dealing with justification perfectly generally; and it was in that general context that he saw proportionality as an aspect “which is crucial to resolution of the dispute” (see the first sentence of [20]). It was only later that he addressed the specific circumstances of the avoidance and evasion of tax.[48]It is also worth referring to what he said at [26] of his Opinion in addressing the justification for the UK legislation put forward by the UK government, as recorded in [24]: “26. I for one consider that this case should be considered with regard to the limits and conditions which the member states are bound to observe in adopting derogating measures. Nor can I overlook the fact that the court has applied the principle of proportionality, which is one of the general principles of law underlying the Community order, when defining the scope of derogating provisions. Accordingly, in Johnston v Chief Constable of the Royal Ulster Constabulary (Case 222/84) [1986] ECR 1651, para 38, it held: ‘That principle requires that derogations remain within the limits of what is appropriate and necessary for achieving the aim in view’”.[49]The principle of proportionality, it can be seen, was regarded by the Advocate General as one of the general principles of law underlying the Community order. It is not consistent with that approach to say, in the context of the Directive and, in particular, the derogation under Article 11C(1), that proportionality only impacts on evasion and avoidance aspects. But, having said that, it is to be noted that limits are set by what is “appropriate and necessary”.[50]The principle of proportionality can be found at work in many decisions of the ECJ in the context of VAT. See for instance:a. Garage Molenheide BVBA v Belgium (Case C-286/94) [1998] STC 126, the ECJ, having held that a national measure was not in principle precluded by Article 18(4) of the Directive (concerning the right of deduction) concluded at [46] and [48] of the judgment as follows: “46. Thus, in accordance with the principle of proportionality, the member states must employ means which, whilst enabling them effectively to attain the objective pursued by their domestic laws, are the least detrimental to the objectives and the principles laid down by the relevant Community legislation…. ………… 15 48….the principle of proportionality is applicable to national measures, which, like those at issue in the main proceedings, are adopted by member states in the exercise of its powers relating to VAT, since, if those measures go further than necessary in order to attain the objective, they would undermine the principles of the common system of that and in particular the rules governing deductions which constitute an essential component of that system.”b. In Ampafrance SA v Directeur des Services Fiscaux de Maine-et-Loire and Sanofi Sythelabo v Directeur des Services Fiscaux deVal-de-Marne (joined cases C-177/09 and C-181/94) [2000] ECR I-7013 (concerning a purported derogation in French legislation pursuant to an apparent authorisation under a Council Decision and held to be invalid) the ECJ said this at [60] of the judgment: “60. Second, it should be observed that in order for a Community measure concerning the VAT system to be compatible with the principle of proportionality, the provisions which it embodies must be necessary for the attainment of the specific objectives which it pursues and have the least possible effect on the objectives and principles of the sixth directive." And, as was pointed out at [61] of the judgment, the fact that “means less detrimental….can be contemplated or already exist in the national order” meant that the measure was not necessary.[51]Similarly, the principle can be seen at work also in decisions of the UK courts. CR Smith Glaziers (Dumferline) Ltd v C&C Commissioners [2003] STC 419 is such a case. It related to UK provisions which deprived certain services of the exemption afforded by mandatory provisions of Art 13B(a) of the Directive (concerning the mandatory exemption of insurance-related services from VAT). Lord Hoffmann said at [23] of his speech that the only way in which the UK could justify rules depriving such services of exemption was by reliance on Article 13B which permitted Member States to lay down conditions “for the purpose of ensuring the correct and straightforward application of the exemption and of preventing any possible evasion, avoidance or abuse”. He then referred to an opinion of the Advocate General that prevention of evasion could justify a domestic provision and said: “The Advocate General did not enlarge upon what kind of conditions might be regarded as appropriate for this purpose. But in general European law would require them to satisfy the principle of proportionality in its broad sense, which, following German law, is divided into three sub principles: first, a measure must be suitable for the purpose for which the power has been conferred; secondly, it must be necessary in the sense that the purpose could not have been achieved by some other means less burdensome to the persons affected and thirdly, it must be proportionate in the narrow sense, that is, the burdens imposed by the exercise of the power must not be disproportionate to 16 the object to be achieved. In the particular instance of conditions for allowing a VAT exemption, The Court of Justice of the European Communities has recently said that such conditions must be "necessary for the attainment of the specific objective which [the legislation] pursues and have the least possible effect on the objectives and principles of the Sixth of the Directive" reference being made to Ampafrance SA.[52]It is true that in all of those cases, the national provision in question was directed at avoidance, evasion or abuse. And to that extent, the decisions in those cases are not inconsistent with Mr Lasok’s submissions. But what the AG said in Goldsmiths is, in our view, inconsistent with that submission. And as we read the judgments of the ECJ in Ampafrance SA and Garage Molenheide BVBA, there is nothing to restrict what appear to be perfectly general statements about the limits which are to be placed on any power relating to VAT. Similarly, although Lord Hoffmann’s remarks were made in the context of value shifting, which HMRC considered to be avoidance or an abuse, the nature of the principles he describes, and in particular his first sub-principle that the measure must be suitable to the purpose for which the power was conferred, provide no indication that the principle of proportionality is limited to restrictions on general principles justified by the prevention of evasion or avoidance. He described the principle in perfectly general terms and did not so much as hint that the principle, in the context of EU VAT law, was restricted to cases of evasion, avoidance or abuse.[53]The cases, we think, lend a great deal of support to this proposition: that when a Member State is permitted to derogate from a provision of a directive, its power to do so must be exercised to give effect to a permitted objective which can be drawn from the general principles of EU law and the context of the directive, so that any national provision in pursuit of such derogation must be limited to that which is proportionate in the light of those principles and that context. That requirement is not limited, in the context of VAT, to derogation aimed at evasion or avoidance.[54]Mr Lasok challenges, as we have said, the idea that proportionality is relevant. Returning to the basis of the power to derogate identified by the ECJ at [18] of its judgment in Goldsmiths, he asks how the concept of proportionality fits in. To answer his question, we need to say a little more about Article 11C(1) itself. On its face, it appears to require a reduction in the taxable amount whenever there is a total or partial non-payment whether or not the debtor is in fact able to pay. A simple failure to pay where the debtor is able to pay and where the debt can be easily enforced is not what would commonly be regarded as giving rise to a bad debt. Rather, we think that the concept of a bad debt is one which has become payable and remains unpaid and which the debtor either cannot pay (because he has no means to do so) or which he simply refuses to pay in circumstances where it is in commercial terms not reasonable for the creditor to enforce. The phrase “bad debt relief” is not used in Article 11 at all but there can be no doubt, we think, that national legislation which makes no provision for any reduction in relation to total or partial non-payment other than in cases of bad debts (in the sense we have just described) is permissible. Whether this is because the concept 17 of total or partial non-payment carries with it the concept of a bad debt (in that sense) or because it is permitted by the derogation may not matter.[55]It is, of course, no part of the concept of bad debt as we have described it that the debtor is subject to bankruptcy or insolvency proceedings. It is not even part of the concept that the debtor is insolvent. For instance, in the case of a small debt, it may not be commercially sensible to pursue the debtor. Indeed, in the present case the Tribunal recorded at [41] of the Decision Mr Cordara’s submission (which we assume to have been supported by the evidence since Mr Lasok has made no challenge) that for between 90% and 95% of repossessions, the customer was not subject to bankruptcy or insolvency. That, of course, does not necessarily mean that all of those cases resulted in a bad debt but it is reasonable to assume that many cases did so. In that respect, the findings of fact in [35] of the Decision are of interest:
“[35] GMAC took a commercial view on the recovery of debts, often accepting a compromise payment from the customer to limit costs. GMAC was very circumspect in seeking court orders for money judgments, only incurring the expense if there was a substantial debt involved and there were definite means of enforcement such as an attachment of earnings order. Legal action was only taken in respect of around 10 per cent of the shortfall, in which cases GMAC would expect to recover around 40 per cent of the debt subject to such action over a two or three year period.”
[56]The ability to derogate, however, does not simply allow the Member State to refuse repayment where there is a bad debt as we have described that concept. According to the ECJ, the derogation is, to repeat, “based on the notion that in certain circumstances and because of the legal situation prevailing in the member state concerned, non payment of consideration may be difficult to establish or may be only temporary”. What this means is that a Member State can legislate to ensure that bad debt relief is not given in cases where there may be doubt about whether a debt is in fact a bad debt or not. It is the very nature of the derogation that the Member State can legislate in such a way that, in some cases of an actual bad debt, the reduction in the taxable amount which Article 11 would otherwise mandate will not be obtained. And, of course, there will also be the possibility, depending on how the legislation is drawn, that a reduction will be obtained when there is no non-payment at all.[57]And so Mr Lasok argues that the Insolvency Condition, even in its original Mark I form, was aimed at identifying with certainty debts which were bad debts and, in taking the strict approach which the Insolvency Condition adopted, was also designed to ensure that debts which might not be bad debts did not attract the relief. Once there is a departure from the Insolvency Condition, then there will inevitably be a weaker test for what is to count as a bad debt and the greater the likelihood that relief will be given for debts which are not bad debts at all. It is a matter of policy for the member state where to draw the line. A derogation is justified if it reflects a policy which is not unreasonable for determining what should be taken to be non-payment in the context of Article 11C(1). 18[58]On this approach, Mr Lasok submits that the Insolvency Condition manifestly related to the purpose of the power to derogate because it was actually directed at the question of what is or is not a bad debt or, to use the language of the ECJ, at the question of whether non payment has been established or may only be temporary. And it is in that context that he answers his own question by saying that proportionality simply does not fit with this particular power to derogate at all.[59]In our view, the thrust of the authorities, both in the ECJ (including the Opinion of the AG in Goldsmiths) and in the UK courts is entirely against Mr Lasok’s submissions. Lord Hoffmann, in CR Glaziers (Dumferline) Ltd, saw proportionality in the context of a power to derogate as a general principle of Community law. In our view, the Insolvency Condition and the Property Condition can be upheld only if they are proportionate responses to the aim identified by the ECJ in Goldsmiths. We will return to the consequence of that view in a moment.[60]But even if, contrary to our view, Mr Lasok is right in saying that proportionality has no part to play, he has to meet Mr Cordara’s arguments(i) that the Insolvency Condition and the Property Condition infringe another fundamental principle, that of non-discrimination(ii) that the Insolvency Condition (insofar as it requires a taxpayer to bring bankruptcy or insolvency proceedings and then to prove for the debt) and the Property Condition each render the rights conferred by Article 11C(1) impossible or excessively difficult to exercise and(iii) that even applying Mr Lasok’s approach, each of the Insolvency Condition (in both its Mark I and Mark II forms) and the Property Condition went beyond anything which was justifiable in pursuance of the object of the derogation as already explained. In particular in relation to (iii), and questions of proportionality apart, there is the argument, recorded in [41] of the Decision, that “in 90 to 95 per cent of repossessions there was no bankruptcy or insolvency. The vast majority of customers were individuals. Where a customer was insolvent there was the need to prove in the insolvency. In very large part the insolvency conditions could not be satisfied as a matter of ordinary business practice. The insolvency conditions were a serious inroad into the UK’s implementation of Article 11C(1)”.[61]As to fiscal neutrality and non-discrimination, the principle can be broadly stated, although perhaps oversimplified, in this way: transactions which are economically and commercially speaking the same are to be treated for the purposes of the Directive in the same way. As the Advocate General said in his opinion in Goldsmiths (see at [28]), the principle of non-discrimination and the principle of fiscal neutrality are intimately bound up. Equal treatment is required for similar economic activities in order to avoid distortions in the more general EU VAT system caused by the drawing of unimportant and unjustified distinctions. Thus, in Finanzamt Uelzen v Armbrecht (Case C-291/92) [1995] STC 997 the ECJ emphasised that the taxability of a transaction is not a matter for the national law: 19 “…the national law applicable to the main proceedings cannot provide the answer to the question raised, which concerns not the civil law applicable to supply but whether the transaction is subject to the tax”.[62]Mr Cordara has referred to a number of cases in the ECJ concerning the principle of fiscal neutrality. For present purposes, we need refer only to the case which found its way to the ECJ as Revenue and Customs Commissioners v Rank Group Plc (Joined Cases C-259/10 and C-260/10) [2012] STC 23 (“Rank”) which is the most recent case to provide guidance about the interpretation of the principle of fiscal neutrality.[63]Rank concerned the difference in VAT treatment between different types of mechanical cash bingo (“MCB”) and slot machines. The claims were based essentially on the argument that the different types of MCB and slot machines were treated differently for the purposes of exemption from VAT and that accordingly the fact that certain machines were subject to VAT infringed the principle of fiscal neutrality. So far as concerned the MCBs, it was common ground that exemption was available only if the stake was less than 50p and the prize less than £25. It was also common ground, it is important to note, that the two types of MCBs were identical from the customer’s point of view. The factual position in relation to the slot machines was far more complex and we do not need to go into it. We do, however, note that the questions referred to the ECJ in relation to the slot machines included this: whether the principle of fiscal neutrality had to be interpreted as meaning that a difference in treatment for VAT purposes of two supplies of services which were identical or similar from the point of view of the consumer was sufficient to establish an infringement of that principle, or whether such an infringement required in addition that the actual existence of competition between the services in question or distortion of competition because of the difference in treatment could be established.[64]At [32] of the judgment, the ECJ formulated the principle in this way:
“According to settled case law, the principle of fiscal neutrality precludes treating similar goods and supplies of services, which are thus in competition with each other, differently for VAT purposes, so that those goods or supplies must be subjected to a uniform rate……”
. The ECJ went on to add at [33] and [34] as follows: “[33] According to that description of the principle the similar nature of two supplies or services entails the consequence that they are in competition with each other. [34] Accordingly, the actual existence of competition between two supplies of services does not constitute an independent and additional condition for infringement of the principle of fiscal neutrality if the supplies in question are identical or similar from the point of view of the consumer and meet the same needs of the consumer”.[65]The answer to the question which we have identified above was therefore that the principle of fiscal neutrality must be interpreted as meaning that a difference in 20 treatment for purposes of VAT of two supplies of services which are identical or similar from the point of view of the consumer and meet the same needs of the consumer is sufficient to establish an infringement of the principle. Such an infringement thus does not require in addition that the actual existence of competition between the services in question or distortion of competition because of such difference in treatment be established.[66]At [43] and [44] of the judgment, the ECJ said something about what is necessary in order to determine whether two supplies of services (and we have no reason to think that the approach to a supply of goods is any different) are similar. Account must be taken of the point of view of a typical consumer avoiding artificial distinctions based on insignificant differences. Two supplies are therefore similar where they have similar characteristics and meet the same needs from the point of view of consumers, the test being whether their use is comparable, and where the differences between them do not have a significant influence on the decision of the average consumer to use one such service or the other. The conclusion to be drawn from [56] of the judgment is that the determination whether two supplies are similar is for the national court to make. In the context of the facts of Rank, that determination was to be made from the point of view of the average consumer and was to take account of the relevant or significant evidence liable to have a considerable influence on his decision to play one game or the other.[67]The decision of Norris J at an earlier stage of the litigation, HMRC v Rank [2009] EWHC (Ch) 1244, [2009] STC 2304, contains a helpful passage at [31] and [32] of his judgment:
“[31] ……a UK taxpayer who establishes that he provides goods or services which are similar (in the sense required by the principle of fiscal neutrality) to those supplied by another economic operator, and are thus in competition (in the sense in which that is to be understood for the purposes of the principle of fiscal neutrality), but whose services are treated differently for VAT purposes, has established that there is a distortion of competition (and that the principle of fiscal neutrality is engaged). He has established that he has been subjected, without justification, to a different tax regime. [32] ….But the principle of fiscal neutrality recognises that there may be cases in which different treatment is justified and does not violate the principle of fiscal neutrality. In my judgment it would be for the authority imposing the different treatment to establish the ground which justifies that difference, not for the taxpayer to eliminate all conceivable grounds of difference in order to show that his treatment was without justification…….”
[68]How is this approach to similarity to be given effect in the context of the present case? Mr Cordara submits that hp agreements, conditional sales and credit sales are all similar for this purpose, and this is so whether or not the relevant contract contains a retention of title provision (or Romalpa clause as it has been referred to in these proceedings). Accordingly, the Property Condition is discriminatory and an infringement of the principle of fiscal neutrality. It must also be part of his 21 case (although we are not certain that he put it quite this way) that all transactions within any one of those classes are similar so that it is not possible to discriminate between different contracts within the class. It is only on that basis, as we see it, that he can argue, as he does, that the Insolvency Condition is discriminatory; the submission being that creditors owed debts less than the minimum amounts required for bringing bankruptcy or company insolvency proceedings would never be able to bring themselves within the Insolvency Condition unless some other creditor had petitioned successfully for a bankruptcy order or a winding-up order.[69]Mr Lasok, in contrast, contends that there is a material difference, and one which justifies a different VAT treatment, between contracts which do, and do not, contain Romalpa clauses. On that footing, the Property Condition is not discriminatory at all but, even if it is, it is a justifiable difference in treatment between the two classes of contract. He also contends that there is no inadmissible discrimination so far as concerns the Insolvency Condition. As between different classes of contract, there is no discrimination because all are treated in the same way. Within a single class (eg all hp agreements) he says that there is no relevant discrimination.[70]Although the Tribunal addressed in appropriate detail(i) proportionality and the ways in which it was submitted by Mr Cordara that the Insolvency Condition and the Property Condition went beyond a permissible derogation and(ii) the arbitrary and anomalous results which were said to arise, they said very little indeed expressly about why contracts with and contracts without Romalpa clauses were similar for the purposes of the principal of fiscal neutrality. They identified, at [37] of the Decision, common types of arrangements other than hp agreements under which customers could obtain credit to buy a car, namely (i) credit sale agreements and bank loans, noting that GMAC did not enter into credit sale agreements, although some competitors including Toyota Finance Ltd did and (ii) conditional sale agreements with a retention of title clause, which they noted “had no practical difference from HP finance in regulation under the Consumer Credit Act”. But they did not explain why they considered all of these to be relevantly similar. All we find is what was said at [62] of the Decision:
“[62] We do not accept Mr Lasok’s submission that in order to rely on the difference in treatment between hire purchase and credit sale transactions, GMAC needed to produce evidence that competitors had engaged in credit sales. The transactions were essentially similar…..”
[71]Mr Lasok says, in any case, that the Tribunal was inaccurate in asserting that the difference was between hp agreements and credit sale agreements. The statutory distinction is between contracts under which property passes and contracts under which property does not pass. Credit sale agreements containing Romalpa clauses would fall within the second category. That is true, but as is apparent from [38] of the Decision, the Tribunal had in mind the relevant distinction, although we suspect that the second reference in that paragraph to “conditional sale agreements” might have been intended as a reference to credit sales. It is also apparent from [61], where the distinction being drawn was between contracts 22 where title did and did not pass. The Tribunal clearly saw a contrast between hp agreements and credit sale agreements in that respect. Accordingly, we consider it to be clear that the Tribunal saw the different types of contracts as similar.[72]However, it should be remembered that Mr Lasok had made submissions, recorded at [45] of the Decision, that credit sale agreements were obviously different in their nature from hp agreements with different legal results. His submission was that, although the Advocate General in Goldsmiths at [28] of his opinion had referred to “unimportant and unjustified distinctions”, the transfer of title was an objective difference. GMAC chose to structure its arrangements to ensure that it retained title for commercial reasons and must take the consequences, see Customs and Excise Commissioners v Cantor Fitzgerald (Case C – 108/99) [2001] STC 1453 at [31] and BLP Group plc v Customs and Excise Commissioners (Case C – 4/94 [1996] 1 WLR 174 at [25]-[26]. He said that where transactions are genuinely different, there is no breach of the principle of fiscal neutrality if they are treated differently.[73]The Tribunal did not engage with that argument other than to say, in that single sentence in [61] of the Decision, that the transactions were similar. However, we are bound to say that we do not consider that those two cases referred to by Mr Lasok provide much, if any assistance. If Mr Lasok is right in saying that the two types of transaction are not similar, then questions of fiscal neutrality and discrimination do not arise. The challenges to the Insolvency Condition and the Property Condition succeed or fail by reference to other arguments. In contrast, if the two types of transaction are similar, then the difference in treatment must be justified (as Goldsmiths shows); and in that context, we do not see that the fact that traders have a choice about which type of transaction to adopt is a sufficient justification by itself, although it may be a factor to be taken into account.[74]We see some force in Mr Lasok’s argument that contracts where property passes immediately and contracts where property does not pass until payment in full is made are indeed objectively different. As a matter of (purely domestic) law, what he says is obviously correct whether viewed from the perspective of the seller or the purchaser (the “average consumer” in the language of Rank). As a matter of commercial and economic result, however, the position is not clear to us. We do not know enough about how each of these different purchase finance arrangements operate to be able to say whether they are “similar” or not and, even if we did, we could not take into account matters which were not in evidence. As to the evidence, we do not know what was before the Tribunal to allow them to form any sort of view on the question. And we have to say that we are reluctant to treat that single sentence in [62] of the Decision as determinative of this question of fact.[75]As to the Insolvency Condition, we find it difficult to see how this infringes the principle of fiscal neutrality. The insolvency condition applies alike to hp agreements, credit sale agreements and conditional sale agreements. The only suggested discrimination is, as we have already noted, the inability of a creditor with a debt below the relevant bankruptcy or insolvency limit to bring bankruptcy 23 or insolvency proceedings, making it impossible for him to satisfy the Insolvency Condition unless a third party obtains a bankruptcy or winding-up order. It is a different question whether this factor means that the Insolvency Condition cannot be justified, but that is nothing to do with infringement of the principle of fiscal neutrality.[76]We do not consider that there is anything in Mr Cordara’s appeal to the principle of effectiveness which, according to him, is infringed by both the Insolvency Condition and the Property Condition. If those Conditions are within the power of derogation, then the principle of effectiveness, to the extent it would otherwise apply, is validly overridden. If those Conditions are not within that power, there is no scope for the application of the principle of effectiveness in the first place. We accept, of course, that the exclusion of access to bad debt relief as a result of either of those Conditions will be a factor in determining whether they can be justified in accordance with Goldsmiths and in particular in relation to proportionality if that concept is applicable. But we reject any argument that the principle of effectiveness would render invalid either of the Conditions if, otherwise, the Condition would be valid.[77]We come then to the third obstacle identified in paragraph 60 above which Mr Lasok faces namely that neither the Insolvency Condition nor the Property Condition can be justified even if he is right about proportionality and even if there is no infringement of the principles of fiscal neutrality and nondiscrimination. We find it convenient to consider this obstacle together with the application of the principle of proportionality.[78]In relation to the Insolvency Condition, we need to digress for a moment to say something about the operation of insolvency law in England and Wales. So far as concerns individual insolvency, amendments were made to the definition of insolvency for the purposes of VAT- these are found in the amended section 22 VATA 1983 as part of the Mark II form - following the introduction of individual voluntary arrangements and company voluntary arrangements by the Insolvency Act 1985 (provisions re-enacted in the Insolvency Act 1986). Section 22(6) in its amended form provided a regulation-making power which would have enabled the terms of any Individual Voluntary Arrangement to have been the basis for establishing a bad debt. We know of no regulations exercising the power in that way. Accordingly, it remained necessary under section 22(5) for a creditor actually to prove in an individual bankruptcy before a bad debt could be established.[79]In the case of a company, the position was more complicated. Following amendments to section 22 VAT Act 1983 by FA 1985, the position was that a company became insolvent for the purposes of that section if(a) it went into liquidation at a time when its assets were insufficient to meet its debts, liabilities and expenses of winding up or(b) an administrator or administrative receiver had been appointed and he had issued a certificate that, if the company went into liquidation, the assets would be insufficient “to cover the payment of any dividend in respect of debts which are neither secured nor preferential”. Section 22(5) did 24 not apply in such a case and the certificate was enough to found a bad debt relief claim.[80]Under the Insolvency Act 1985 and 1986, proposals for a company voluntary arrangement could be made by the directors of a company. There was nothing in section 22 equivalent to the position in relation to individual voluntary arrangements and therefore no power, under section 22(6), to make regulations equivalent to those we have just described in relation to individuals.[81]The position in relation to both individual and company insolvencies was, in the absence of any regulations in the former case, that no provision was made for establishing bad debts in the case of voluntary arrangements. And yet the typical result of any such arrangement is that creditors will only ever receive a proportion of what is owing to them. The present case is not concerned with voluntary arrangements; the fact that bad debt relief cannot be obtained in a case where there has been a voluntary arrangement is not fatal to the validity of the Insolvency Condition in other cases. But the absence of such relief leads, we think, to the conclusion that the Insolvency Condition is not compliant with Article 11C(1) in such a case.[82]After that digression, we return to the main discussion. It is important to note that, in the case of an individual insolvency, a bankruptcy petition can only be presented if the debt on which the petition is based exceeds a statutory minimum (now £750 but previously £200 for the part of the period with which the appeal is concerned). The UK in effect, imposed a blanket exclusion from bad debt relief on creditors of individual debtors whose debts were less than the statutory minimum.[83]In relation to the insolvency of a registered company, it is possible to present a petition on the basis that the company is unable to pay its debts (see now section 122(1)(e) Insolvency Act 1986: the position was the same before that Act). One of the ways of demonstrating that a company is unable to pay its debts is to serve a statutory demand but there was and is a monetary limit (now £750: see section 123(1)(a) Insolvency Act 1986). That comparatively simple route is not open where the debt is less than £750. In contrast with personal insolvency, it has always been possible for a creditor, however small his debt, to establish an inability on the part of a company to pay debts by demonstrating to the satisfaction of the court that the company is unable to pay its debts as they fall due. It might be said, therefore, that even in the case of small debts of UK companies, it was always open to a creditor to fulfil the Insolvency Condition.[84]However, it has never been a cost-free exercise to present a petition against either an individual or a company. Quite apart from court fees, there are the costs of instructing professionals and the inherent risk that these fees and costs will never be recovered. A creditor, it is obvious, is bound to consider very carefully whether to take proceedings in respect of any debt. This is so where the debt exceeds the statutory limit for personal insolvency or for a corporate statutory demand; a fortiori in the case of a company debt of an amount below that limit. It 25 is entirely unsurprising and entirely reasonable that an entity such as GMAC should have clear policies about its approach to the enforcement of unpaid debts through the courts and to take a commercial view about the cost-effectiveness of attempting to do so. Even if it is thought to be reasonable and indeed proportionate to require that a judgment be obtained for the debt (eg in the small claims court) it is an entirely different question whether it is proportionate or even reasonable to require the invocation of insolvency procedures and the proof of debt before bad debt relief can be claimed in respect of small debts.[85]We do not consider that it was justifiable to exclude from bad debt relief debts owing by an individual which do not exceed the statutory minimum. That exclusion was not a proportionate response to the aim identified in [18] of the judgment in Goldsmiths or, indeed, of any other admissible aim. Further, we find it hard to categorise the derogation as reasonable on any possible meaning of that word. There is no material of which we are aware which would go anywhere near justifying the total exclusion from bad debt relief where the debt was less than the statutory limit in force at the relevant time. This, we add, has nothing to do with discrimination between cases which fall below the limit as compared with cases which fall above the limit (although there is such discrimination) but has everything to do with the scope of the power to derogate.[86]Further, we do not consider that it was a proportionate response, either, to require proof in an insolvency as a condition for relief in all cases which are not altogether excluded as a result of the statutory limit. This we consider is manifestly so in the case of a small debt, say £100 or even less, owed by a company debtor. We do not consider that a regime of that sort is justifiable as a proportionate response to the legitimate aims of the UK government. There is nothing in the findings of the Tribunal, and, as far as we are aware, there was no evidence before them on the basis of which such justification could be established. Further, we do not, as in the case of debts below the statutory limit owing by an individual, even consider that it could be reasonable let alone proportionate to impose such a stringent requirement.[87]Having said that, it may be the case that where there is a large debt, it would be both reasonable and proportionate to require proof in insolvency. Where the line could be drawn could be a matter of dispute but it is not one we need to address and we decline to say anything about what could be said to be “large”. The point here is that the domestic legislation does not attempt to draw the line. As we see it, we must rule on whether the Insolvency Condition as a whole is valid or invalid. We have no doubt on that basis that it is invalid. It is certainly not a proportionate response to the legitimate aims of the UK; we do not even consider that it is a reasonable response.[88]As to the Property Condition, we do not understand what the justification for this is said to be nor how it is any more or less difficult to say that a debt is a bad debt depending on whether property has passed or not. Even if the principle of proportionality has no part to play, and even if there is no issue of breach of the principles of fiscal neutrality or non-discrimination, we do not consider that there 26 can be one policy in relation to cases where property has passed and another policy in cases where property has not passed, given that both policies must derive from the same requirement to justify the derogation. Further, given the basis for the power to derogate stated in [18] of the judgment in Goldsmiths, such a derogation would not satisfy the principle of proportionality. Accordingly, assuming that this is a matter for the national court, we consider that the Property Condition was not authorised by the power to derogate.[89]Whether the principle of proportionality applies is a matter of EU law. Accordingly, the views which we have just expressed are not necessarily conclusive unless we consider that the matters are acte clair. We consider it to be clear that EU law requires any derogation under Article 11C(1) to be proportionate. On that basis, we also consider that, in relation to both the Property Condition and the Insolvency Condition, the matter is acte clair.[90]For completeness, we should deal with the suggestion made by Mr Cordara that because the UK later modified and then removed the relevant requirements, those requirements cannot have been necessary in the first place and were therefore not proportionate. Mr Lasok submitted that if this argument was accepted, it would effectively destroy the discretion which was intended to be given to Member States by the Directive and was therefore not a permissible approach. We agree with Mr Lasok. As with so many aspects of the law generally, there is a range within which a person or body of persons or even a State can operate reasonably. So far as concerns proportionality, the test appears to be that the national measure must, as Lord Hoffmann puts it, be necessary “in the sense that the purpose could not have been achieved by some other means less burdensome to the persons affected”. But this cannot be seen as an absolutely rigid, black line, test. The word “necessary” is not to be construed strictly so as to be given the meaning which a logician or mathematician might ascribe to it. There is, we think, even in relation to necessity, a spectrum, albeit it is narrowly confined, within which different reasonable and objective minds can take different views about what is or is not necessary. And, of course, there may be two different ways in which a permissible objective can be achieved, each of which carries a different detriment for the persons affected. It could not be said that either was strictly necessary although it could be said that necessity entailed the selection of one of them. Accordingly, we see nothing wrong with the proposition that a Member State could take the view that a particular measure was necessary, but nonetheless modify, or replace, that measure in the light of experience.[91]This chimes with the fact that the power to derogate is precisely that: it is a discretionary power to refuse bad debt relief where there is in fact a bad debt in the sense that the debt will never be paid, something which the Advocate General recognised at [15] of his opinion in Goldsmiths. As Mr Lasok puts it, the whole point about the discretion is that bad debt relief schemes can be organised in different ways with different trade-offs between their advantages and disadvantages. . 27 V The Windfall Issue[92]The Windfall Issue arises only if the Insolvency Condition and the Property Condition are invalid. If either of them is valid, then GMAC has no directly enforceable right to bad debt relief under the Directive. We proceed, therefore, on the basis that they are both invalid. Our discussion is carried out in the context of a sale of a new car to the customer of a dealer where the new car is first purchased by GMAC from the dealer and is then sold on by GMAC to the customer pursuant to an hp agreement.[93]We have already described in general terms at paragraph 24 above the rationale for the introduction of Article 26a. The avoidance of a double charge to VAT on the same economic consumption was also the rationale behind the Cars Order (given the approach taken to Regulation 38 by HMRC prior to GMAC No 1). Thus, to return to the example, the full £15,000 purchase price on the sale by GMAC would bear VAT. It would be wrong in principle for the resale, following repossession of the car, to bear VAT on the resale consideration at least to the extent that GMAC recovered on the resale amounts not exceeding the amounts owing under the hp agreement.[94]After GMAC No 1, the resulting position was not one contemplated either by domestic legislation or by the Directive alone. Under the Directive, GMAC was entitled to obtain relief as a result of Art 11C(1) in respect of the shortfall in the payments which it received from the purchaser, with the Directive providing for nothing equivalent to the Cars Order on a subsequent resale. Under UK domestic legislation as understood by HMRC prior to the decision in GMAC No 1, the economic effect of the Cars Order may have been broadly equivalent to the giving of bad debt relief in relation to an amount equal to the resale price and, in that sense, it can be said that the Directive would have been adequately implemented at least in relation to that part of the non-payment by the purchaser of the original consideration. But the interaction of Regulation 38 and the Cars Order after GMAC No 1 would have resulted (even assuming the validity of the Insolvency Condition and the Property Condition) in less VAT ultimately being payable than would have been the case if the Directive had been properly implemented. HMRC cannot, of course, complain about that; the remedy, as in fact eventually implemented, was to change the legislation for the future. But what they do say is that the Cars Order was inconsistent with the Directive given that a reduction in the taxable amount was effected by Regulation 38.[95]Mr Cordara, however, submits that the Cars Order was not inconsistent with the Directive even in the light of GMAC No 1. In other words, he says that Regulation 38 and the Cars Order, taken together, represent an effective and proper implementation of the Directive. He says:(i) it ensures there is no distortion of the second hand cars market: GMAC would otherwise be competing in a market in which other sellers were not charged VAT;(ii) without the desupply there would be double taxation, and(iii) EC Commission v Italian Republic (Case C-45/95) [1997] STC 1062 supports the treatment. 28[96]We are unable to accept those reasons. As to the second-hand market, we note what was said by the ECJ in EC Commission v Ireland (case C-17/84) [1985] ECR 2375, at [13] and [14] of the judgment in relation to the law as it then stood prior to Article 28a: “ 13. If the consumer subsequently supplies the goods to another nontaxable consumer, no tax is charged or deducted in respect of that transaction. If the consumer supplies the goods to a taxable trader, such supply does not give rise to a charge to tax either, but where the goods are resold by the taxable person an amount of VAT proportional to the resale price is charged without that taxable person being entitled to any deduction of the VAT which the goods have already borne. 14 Second-hand goods which are reintroduced into commercial circulation are therefore taxed once again, whereas second-hand goods which pass directly from one consumer to another remain burdened by the tax imposed on the occasion of the first sale to a non-taxable consumer. Especially where the rate of VAT is high, that difference in treatment distorts competition between direct sales from one consumer to another and transactions passing through ordinary commercial channels, and thus places at a disadvantage branches of trade in which a large number of transactions involve second-hand goods, such as the motor trade in particular."[97]By contrast it seems to us that where both Regulation 38 and the Cars Order apply a distortion of competition or advantage to commercial traders arises in the case of repossessions of cars. In relation to the original sale, GMAC charges its customer VAT but will be entitled to a deduction for its input tax on the acquisition of the new car from the dealer. The consumer obtains a car burdened (to use the language of the ECJ) by the VAT. If the customer subsequently sells the car (having paid the full price and obtained title) the car remains burdened with the VAT. In contrast, where a car is repossessed, it is not burdened in the same way with all the VAT payable on the original sale. This is because on a resale, the Cars Order effects a desupply of the car so that there is no VAT payable, but GMAC obtains a reduction in its VAT liability on the original sale under Regulation 38 (and bad debt relief pursuant to its directly enforceable rights under the Directive if GMAC is correct in its claim).[98]Thus if GMAC provides a car for instalments totalling £15,000 and the car is repossessed after paying £5,000 and the is car is sold at auction for £10,000 (we take this unlikely scenario to eliminate the need for consideration of bad debt relief) the total amount of VAT payable by GMAC would in the absence of the Cars Order be in relation to a VAT-inclusive figure of £15,000 being(i) £5,000 (=£15,000- £10,000) in relation to the hp plus(ii) VAT on the auction sale in relation to £10,000. Thus the car in the hands of the auction purchaser would be burdened with VAT on £15,000. But after the effect of the Cars Order, which eliminates the VAT in relation to the £10,000, the total VAT burden would be in relation to £5,000 only. The car is thus burdened with a smaller amount of VAT in the case of GMAC’s sale than it would be in the case of a sale by a consumer 29 who is not a trader and who had purchased the car new and outright and placed it in the auction himself. GMAC is thus in a better net position than a non-trader. This is not the elimination of a distortion of the second-hand market but the creation of one by the combined effect of the Cars Order and Regulation 38. GMAC is competing in a market in which other sellers, although they do not charge VAT on their sales are dealing in cars which are burdened with VAT on the full £15,000 of consumption.[99]Nor do we see how the absence of the Cars Order would give rise to double taxation given the interpretation of Regulation 38 in GMAC No 1. The same analysis shows that there is less tax payable by the time the purchaser at the auction has acquired his car than there would have been if the original purchaser had honoured his obligations. 100.So far as the Italian Republic case is concerned, what that did was to explain that if a person did not get a deduction for input tax in respect of goods used for the purposes of his exempt business he should not tax his sale of those goods because, by reason of the input tax block, the consumption of the goods had already borne tax. We are afraid that we do not entirely understand the point which Mr Cordara is making, but since GMAC does get input tax credit, this case seems to us to be irrelevant. 101.Accordingly, we perceive the combination of Regulation 38 (as interpreted by GMAC No 1) and the Cars Order as an ineffective implementation of the Directive, giving an excessive relief from VAT inconsistent with the objective of the Directive and thus not compliant with EU law.[102]That being so, HMRC say that GMAC cannot seek to rely on the direct effect of the Directive whilst at the same time seeking to rely on national legislation, the combination of which gives rise to a result inconsistent with the aims and intentions of the Directive.[103]Returning to the argument, we summarise Mr Lasok’s submissions in the following paragraphs.[104]First, and very importantly, he focuses on the overall result which GMAC seeks to achieve by reliance on the Directive. He submits that the overall result must be compatible with the Directive and that this Tribunal, as an emanation of the State, is bound to secure the result intended by the Directive and not some other result. As it is stated in the current Treaty on the Functioning of the European Union (“TFEU”), Article 288: “A directive shall be binding, as to the result to be achieved, upon each Member State to which it is addressed, but shall leave to the national authorities the choice of form and methods”.[105]But seeking to achieve some other result is, he says, precisely what GMAC is doing. On the one hand, it has relied in previous claims against HMRC on 30 domestic legislation (Regulation 28 and the Cars Order in particular) but now it seeks to rely on the direct effect of the Directive, the combined effect of which is a result inconsistent with the scheme of the Directive. It is not possible to describe the fiscal result which it is attempting to achieve by reference to the Directive alone or by reference to domestic legislation alone.[106]Thus, looking at domestic legislation alone, there is no valid claim for bad debt relief for the relevant periods because the Property Condition and the Insolvency Condition prevent it, which is why reliance is placed on the direct effect of the Directive. But if the Directive alone is taken into account, the position, going back to the example, is this:a. the sale of the car by the hp company to the customer gives rise to a VAT charge on the purchase price (a VAT-inclusive figure of £15,000);b. the resale of the car for £6,500 gives rise to an adjustment of the consideration for the purposes of Article 11C(1);c. there is an entitlement to bad debt relief in relation to the original sale in respect of the unpaid balance of £3,500; andd. the resale of the car is subject to VAT; in effect, the £6,500 is a VAT-inclusive figure.[107]And yet, by picking and choosing as Mr Lasok would say, GMAC is able to obtain a result which is better than either domestic legislation or the Directive taken alone, and that is HMRC’s complaint. He submits that it is wrong to approach the matter on a transaction by transaction basis and then to apply the domestic legislation or the Directive separately to each transaction. Instead, a more holistic approach must be taken; the connected transactions are to be seen as a unified whole and the same legislation must be applied across those connected transactions. On this approach, it is wrong to treat the original sale of the car to the customer and its resale at auction as independent of each other for the purposes of VAT. In particular, it is wrong to apply the Directive to the unpaid consideration (the £3,500 in the example) but to apply domestic legislation in relation to the desupply, by the Cars Order, on the resale of the car. So Mr Lasok says that GMAC is entitled to the treatment afforded by domestic legislation and that, we add, is so even if domestic legislation incorrectly transposes the Directive and gives the taxpayer more favourable treatment than the Directive allows. If the taxpayer complains that the domestic treatment is not what the Directive provides, then so be it; but the taxpayer must then take the whole of the Directive and not combine bits of it with domestic law.[108]This argument is, we think, critically dependent upon an approach which treats the original sale and the subsequent resale as connected in a relevant way so that it is permissible, for VAT purposes, to treat them together and thus to qualify the usual approach which is to treat two transactions (ie the original sale and the subsequent resale) as separate for VAT purposes. If the sale and the resale are not linked in some way relevant to VAT, there is no reason why GMAC should not 31 elect to apply the Directive in relation to the first transaction and at the same time elect to apply the Cars Order to the second transaction.[109]There obviously is a connection as a matter of fact. The same car is the subject matter of the original sale and the resale and the contractual provision (clause 9 of the standard contract) which results in a reduction in the consideration only applies where the car has been repossessed. Property in the car does not pass to the original purchaser so that GMAC has never parted with property in that which it disposes of on the resale. On one view, therefore, the car has never left the economic chain from manufacturer to end consumer, via dealer, GMAC and original purchaser. But on another view, the car left the relevant economic chain for VAT purposes when it was sold to the original purchaser; this is reflected in the fact that the whole of the VAT is due notwithstanding that GMAC receives it from the purchaser only as and when he pays contractual instalments. The critical question, it seems to us, is therefore how these factually related transactions fall to be treated under the Directive.[110]Mr Lasok, however, approaches the matter from a different direction. He starts with what the Tribunal said at [84] of the Decision where it accepted Mr Cordara’s submission “that there is no precedent in EC law for limiting the exercise of directly effective rights in the manner suggested”. In reaching that conclusion, they appear to have relied on what was said by Lord Hoffmann in Customs and Excise Commissioners v Robert Gordon’s College [1995] STC 1093 at 1099: see at [82] of the Decision where they rejected Mr Lasok’s submission that the approach of Lord Hoffmann in that case - as they put it that it is necessary to analyse matters transaction by transaction - was overtaken by Halifax plc v Customs and Excise Commissioners (Case C – 255/02 [2006] STC 919. Mr Lasok challenges that conclusion and asks rhetorically “What are the purposes of the Directive? Is it consistent with the purposes of the Directive for its provisions to be exploited selectively so as to produce a result inconsistent with what the Directive intends to achieve?”[111]He answers his own questions, unsurprisingly, by saying that it is pretty obvious on basic principles that it was not within the intendment of the authors of the Directive that its provisions should be used selectively to produce a result inconsistent with what it was seeking to achieve. And so it would be correct to refuse the ability of GMAC to rely upon a provision of the Directive in order to produce a result inconsistent with what the Directive intended. We do not disagree with those general statements. But what is entirely unobvious is what the intendment of those authors, to be ascertained from the Directive itself and any other admissible material, actually was. If the intention was to treat separate supplies entirely separately, then, subject to special considerations of which Halifax is an example, the linkage for VAT purposes between the original sale and the resale cannot be established. In those circumstances, GMAC would be entitled to apply the Directive to the original sale and domestic legislation to the resale. 32[112]If, in contrast, different supplies can be linked for VAT purposes, then there is considerable force in Mr Lasok’s submission that it cannot have been the intention of the Directive that the linked transactions should have the VAT consequences for which GMAC contends. But it leaves hanging the question: When can different supplies be so linked?[113]At this point, we digress for a moment to consider the position if the intention of the Directive is, subject to limited exceptions such as Halifax, to treat separate transactions independently, so that the sale and the resale need to be addressed separately. We have said that GMAC should be able to rely on the Directive in relation to the original sale; indeed, Mr Lasok’s own submissions would lead to that result provided that GMAC does not at the same time seek to rely on domestic legislation in relation to the original sale so as to give it an advantage over the Directive.[114]In that context, it seems to us that Clause 9 of the standard contract would bring about a situation in relation to Article 11C(1) where “the price is reduced after the supply takes place”. As to that, Field J’s decision in GMAC 1 was not appealed. One of his findings was that Clause 9 resulted in a decrease in the consideration for the original supply of the car for the purposes of Regulation 38. If he was right in that conclusion, it is not easy to see why a different result would be reached in relation to Article 11C(1). And if the result is not different, there is no headroom for HMRC to argue that, in order to obtain relief in relation to the reduction of the consideration (£6,500 in the example), it is necessary for GMAC to rely on domestic legislation and that, in accordance with their main argument, it cannot at the same time obtain bad debt relief on the balance outstanding (£3,500 in the example).[115]That conclusion is consistent with the approach which has been taken in subsequent domestic legislation. Thus the amendment effected in order to reverse the effect of GMAC 1 (the Value Added Tax (Cars)(Amendment) Order 2006 SI 2006/874) operates by taking a case out of the Cars Order where it falls within Regulation 38 (among other provisions) and Regulation 38 is clearly seen as reflecting the first limb of Article 11C(1). We do not decide this point, however, since it has not been argued; there are arguments that Field J was wrong and that his decision is not binding in relation to Article 11C(1).[116]After that digression, we return to Article 288 TFEU, Mr Lasok submits that the choice given to the national authorities about the form and method of implementation of the binding obligations imposed by a directive justified the approach taken by the UK in combining the Cars Order with the statutory provisions dealing with, for example, bad debt relief. Combining the different provisions (as they had been understood by HMRC prior to GMAC 1) and assuming that the bad debt relief provisions complied with the Directive, produced the end result that the Directive intended. That, he says, was one permissible configuration of the legislation, notwithstanding that a different approach could 33 have been taken, for instance by giving more extensive bad debt relief and dispensing with the Cars Order.[117]In support of his submission, Mr Lasok places considerable reliance on the decision of the ECJ in Becker v Finanzamt Munster-Innenstadt (Case 8/81) [1982] ECR 53. This was a case where the German government had not implemented the Directive in time and, during the period in which it had not been implemented, Mrs Becker claimed tax treatment for her supplies on the basis of the Directive which was inconsistent with the basis specified in the German legislation. Under the Directive, her supplies were exempt; under German legislation, they were taxable. At [25] of the Judgment, the court stated its conclusions about when direct effect can be given to the provisions of a directive:
“Thus, wherever the provisions of a directive appear, as far as their subject matter is concerned, to be unconditional and sufficiently precise, those provisions may, in the absence of implementing measures adopted within the prescribed period, be relied upon as against any national provision which is incompatible with the directive or in so far as the provisions define rights which individuals are able to assert against the State.”
[118]Following that, the Court addressed the direct effect of Article 13 of the Directive, saying this at [27]:
“Inasmuch as it specifies the exempt service and the person entitled to the exemption, the provision [Article 13], taken by itself, is sufficiently precise to be relied upon by an individual and applied by a court. However, it remains to be considered whether the right to exemption which it confers may be considered to be unconditional, having regard to the general scheme of the directive, to the context in which Article 13 is placed and also to the particular characteristics of the system of taxation within which the exemption is to apply.”
[119]Mr Lasok explains that passage by saying that the ECJ was not saying that direct effect simply involves looking at one particular provision of a directive. Instead, account has to be taken of the general scheme, the context and the particular characteristics of the system of taxation within which the provision needs to apply.[120]Following that passage, the ECJ then went on to consider the system of VAT and certain arguments put forward by the German government based on the features of VAT, namely the chain of transactions with which VAT is concerned and the consequences of severing that chain by the effect of exemption; the concern was that reliance on direct effect might create confusion because some people might be relying on the direct effect of the Directive and some people might be relying on German law. Particular emphasis was placed by the German government on the disruption caused by the fact that an exemption might be claimed a posteriori, to the detriment of taxpayers who, in a business relationship 34 with the person exempted from tax, either follow or precede him in the chain of transactions. As to that, the ECJ said this:
“In that regard it should be pointed out that the scheme of the directive is such that on the one hand by availing themselves of an exemption persons entitled thereto necessarily waive the right to claim a deduction in respect of input and on the other hand, having been exempted from the tax, they are unable to pass on any charge whatsoever to the person following them in the chain of supply, with the result that the rights of third parties in principle cannot be affected.”
[121]Mr Lasok places reliance on the words “by availing themselves……claim a deduction”. This he says is precisely the position which would arise if a taxpayer attempted to combine the two systems. He also places reliance on [45] and [46] of the Judgment. In [45], the ECJ dismisses the expressed concerns about disruption saying that these concerns are “unfounded where a taxpayer has expressed his intention to avail himself of the exemption conferred by the directive and moreover bears the consequences of his choice.” There, he says, the ECJ is again envisaging an election: if the taxpayer expresses an intention to avail himself of the exemption, he has to bear the consequences of his choice. A similar point is made in relation to [46] which we do not need to articulate at length.[122]The next case on which Mr Lasok relies is Kloppenburg v Finanzamt Leer (Case C-70/83) [1984] ECR 1075. The issue in this case was, in essence, Mrs Kloppenberg’s ability to rely on the direct effect of the Directive. The situation here was that the Directive had come into full effect but had not been implemented in Germany (as in Becker). A later Community measure postponed the date of entry into effect of the Directive. It was held that the measure would not have retroactive effect so as to deprive of their rights persons relying on the direct effect of a directive in the meantime, unless the later measure expressly so provided.[123]The ECJ saw the situation as identical to that in Becker and in another case to which we will come in a moment, Grendel GmbH v. Finanzamt für Körperschaften, Hamburg (Case C-255/81) [1982] E.C.R. 2301, saying this at [9] of the Judgment: “It follows that, during the said period, a credit negotiator who had not passed on the tax to persons following him in the chain of supply was justified in relying on the exception provided for by Article 13B(d)1 of the Sixth Directive and a member-State which had failed to fulfil its obligations could not claim, as against that person, that it had not implemented the directive.” 35[124]Mr Lasok interprets this by saying that Kloppenburg is a case in which the ECJ considered that for somebody to rely on the Directive, it is not good enough simply to say that it has not been implemented domestically; rather, their ability to rely on the Directive is contingent on their acting consistently with the implications of their claim that it should apply.[125]The next case we refer to is Grendel. We do not need to go into the facts of that case. At [11] of the Judgment, reference is made to Becker where the ruling in that case is described in this way:
“….it was possible for the provision concerning the exemption from turnover tax of transactions consisting of the negotiation of credit contained in Article 13 of Directive 77/388 to be relied upon, in the absence of the implementation of that directive, by a credit negotiator where he had refrained from passing that tax on to persons following him in the chain of supply, and that the State could not claim, as against him, that it had failed to implement the directive.”
[126]Mr Lasok draws attention to the words “where he had refrained from passing that tax on to persons following him in the chain of supply”. He submits that here the ECJ is indicating that, due to the close connection between different provisions of the Directive and the fact that one is dealing with chains of supply, a person wishing to rely on the Directive must act consistently with what the Directive seeks to achieve. In other words, the result which the person relying on the Directive seeks to achieve must be the result envisaged by the Directive.[127]He relies on two decisions of the VAT and Duties Tribunal in support of this approach. They are Sunningdale Golf Club v Commissioners of Customs and Excise [1997] V&DR 79 and Barclays Bank plc v Commissioners of Customs and Excise [2001] UKVAT V18410. What both of these cases show is that a taxpayer cannot take the benefit of direct effect without the burden. As it was put in [29] of the decision in Barclays Bank, “A trader has no Community Law right or legitimate expectation to elective application of direct effect”. Both of these case were, however, concerned with the VAT consequences of identified transactions or chains of transactions. In the Sunningdale Golf Club case, the taxpayer relying on the direct effect of Article 13A(m) (exemption for supply of certain services linked to certain sporting provision) also had to accept the application to him of Article 17 (the right to deduct) on the basis that his supplies were exempt. In the Barclays Bank case, the same result was reached, the exemption in that case being under Article 13B(d)(6).[128]Accordingly, Mr Lasok submits that GMAC, having relied on domestic law following GMAC 1, in relation to the periods of account now under appeal, cannot rely on the direct effect of the Directive. Or, if it does, it can only do so on the footing that the Directive applies, to the exclusion of domestic law, in relation to the resale of the cars concerned, with the result that the Cars Order would not apply. 36[129]Mr Cordara says that this is all wrong. He identifies a central question in this way. Can a Member State seek a judicial suspension of a directly effective right [“A”] because it claims to have enacted a domestic law right [“B”] which, it says, is incompatible with some of its other EU obligations (ie other than its obligation to furnish right A to its taxpayers)? Or, to put it another way, can a Member State plead (as a defence to an otherwise plainly directly enforceable right) that it has broken EU law by having a domestic law B available in an allegedly improper combination with EU law right A (ie can it rely on its own supposed wrong as a defence)?[130]If those questions are the correct questions to ask, then Mr Cordara is well on the way to establishing his case. Those questions do not, however, address Mr Lasok’s central contention which is that it is impermissible to mix and match rights under the Directive and rights under domestic law: one must take the whole package of the Directive so that, where inconsistent with domestic law, the Directive applies. Mr Cordara’s formulations beg the very question at issue, namely whether GMAC does, indeed, have a directly enforceable right to bad debt relief under the Directive which is independent of its obligation, applying the Directive alone, to account for VAT on the resale price.[131]That brings us back to what we have described as the critical question at the end of paragraph 109 above. The starting point is what we hope is an uncontroversial proposition that ordinarily the VAT system operates at the level of individual supplies. There is a now a well-established qualification to that where the transactions concerned constitute an abuse of rights such that the VAT treatment falls to be redefined in accordance with the principle in Halifax. But apart from cases of abuse, the focus of the Directive is on each supply as a separate economic activity.[132]Mr Cordara puts it this way: the three cardinal elements of the system (identity of supply, amount of consideration and scope of input tax recovery) are all analysed and supervised by reference to individual supplies and not groups of supplies. This, he says, is so within a single supply chain and is, a fortiori, so between two supply chains. He says that very little is required to create two supply chains, referring us to Kuwait Petroleum v Customs & Excise Commissioners (Case C-48/97) [1999] ECR-I-2323, [1999] STC 488 where, at [43] the ECJ said this:
“In reality, it is not possible to treat as a single economic transaction a series of events consisting of two distinct transactions; sale of fuel coupled with the supply of stamps and the subsequent supply of redemption goods for those stamps.”
[133]And so, in the present case. Mr Cordara says that there are clearly two distinct transactions: the original sale of the car to the customer by the hp company and the resale of the car by the hp company at auction. Those transactions are to be treated entirely independently for VAT purposes: there are two supplies. There is 37 nothing whatsoever to prevent GMAC from asserting the direct effect of the Directive in relation to the original sale and in relying on the Cars Order to desupply the resale. He relies on case-law. As well as Kuwait Petroleum, he refers to what Lord Hoffmann said in Customs and Excise Commissioners v Robert Gordon’s College [1995] STC 1093 at 1099, citing BLP Group plc v Customs and Excise Commissioners (Case 4/94) [1995] STC 424, to the effect that it is necessary to analyse matters transaction by transaction. It is not permissible to take a global view of a series of transactions in the chain of supply. Like the Tribunal (see [82] of the Decision) we do not consider that the transaction by transaction approach explained by Lord Hoffmann has been overtaken by Halifax. Mr Cordara also refers to Optigen Ltd v Commissioners of Customs & Excise and other cases (Joined cases C-354/03; C-355/03; C-484/03) [2006] ECR-I-483 to establish that it not permissible to conflate more than one set of supplies at a time. As it was put in [54] of the Judgment in Optigen (which must, of course, be read subject to what was said about abuse in Halifax):
“The Court has consistently held that, according to the fundamental principle which underlies the common system of VAT, and which follows from Article 2 of the First and Sixth Directives, VAT applies to each transaction by way of production or distribution after deduction of the VAT directly born by the various cost components…..”
[134]We would add also a reference to Card Protection Plan Ltd v Customs and Excise Commissioners (Case C-349/96) [1990] STC 270 at [29] where it was made clear that it followed from Article 2(1) of the Directive that “every supply of a service must normally be regarded as distinct and independent”.[135]Mr Cordara does not rely simply on the case-law. Looking at the provisions of the Directive as a whole, and at Article 11 in particular, he says that those provisions are framed in terms of individual supplies. Again, as a general proposition, that is no doubt correct. On that analysis, the result is clear. The original sale is one supply in relation to which GMAC is entitled to assert the direct effect of Article 11C(1); and the resale is a different supply in relation to which GMAC is entitled to assert the provisions of the Cars Order with the result that no VAT is due. The combination of Regulation 38 (following GMAC 1) and the Cars Order may not be compliant with the Directive, but that is not a matter which HMRC can pray in aid against GMAC in relation to the resale viewed in isolation. Mr Cordara has addressed to us a multitude of different and subtle arguments. But in essence, they are all, we think, based on the simple approach that the original sale and the resale are separate supplies which fall to be treated entirely independently for VAT purposes. It follows on that approach, as we have already indicated, that GMAC can elect to apply the Directive (including for this purpose those parts of domestic law which are consistent with it) to the original sale and can elect to apply the Cars Order to the second sale.[136]There is, however, one aspect of Mr Cordara’s arguments which we do address. He says, correctly, that HMRC cannot ignore the fact that the original 38 supply is treated as a supply of goods by GMAC and that the whole of the consideration is brought into account when the supply is treated as made. And this is so notwithstanding that property in the car does not pass to the purchaser. To be fair to HMRC, they do not contend otherwise. However, what Mr Cordara says is that HMRC then blow hot and cold: on the one hand they require that the whole of the consideration be brought into account in that way (as would be the case of an outright sale), but on the other hand they say that the car did not leave the economic chain on the original sale thus seeking to prevent the recognition of the obligation to refund part of the tax under the principles of Article 11 and Elida Gibbs Ltd v Customs and Excise Commissioners (C-317/94) [1996] E.C.R. I- 5339; [1996] STC 1387. Mr Cordara’s essential point here is that the result of the original sale is that, for VAT purposes, the car left the supply chain and entered into consumption.[137]As to Elida Gibbs, it is necessary only to make these observations: The ECJ said that the “basic principle of the VAT system is that it is intended to tax only the final consumer” and that “taxable persons [other than the final consumer] do not bear the burden of VAT”, but “collect the tax on behalf of the tax authorities and account for it to them”. Thus, VAT is, in general, charged on the consumers in each transaction in the chain, and is recoverable by taxable consumers in the chain, provided they make taxable supplies, so that the tax is ultimately borne by the final consumer.[138]In the real world, rather than the VAT world, things are rather different. In the real world, the car has been provided (we avoid the use of the word “supply” which has a special VAT meaning) to the customer by GMAC on the terms of the hp agreement. The customer obtains possession of the car and is entitled to use it. But GMAC retains title and a significant economic interest in the car. When the car is repossessed, GMAC is effectively reasserting its retained title and is able to resell the car. The car does not leave the chain of provision by GMAC unconditionally on the original sale: it does so only provisionally, with the final departure taking place only when title passes to the customer on payment of the purchase instalments in full.[139]As we understand Mr Lasok’s argument, he does not challenge the proposition that there are two separate supplies, first on the sale and secondly on the resale. He certainly does not argue that the combination of transactions is to be treated as a single supply of the car (a difficult concept which would involve a single supply to two separate unrelated purchasers). On the contrary, he would apply to the separate transactions precisely the treatment which Mr Cordara says is required by the Directive and the case-law of the ECJ. But what he says is that the treatment of both must be applied by reference either to the Directive or to domestic legislation; what is inadmissible is to pick and choose.[140]The issue then comes down, in essence, to ascertaining the purposes which it is the objective of the Directive to achieve. In answering that question, Mr Cordara focuses on each separate transaction. His case has to be that the purpose 39 of the Directive (absent abuse of the Halifax type) is to tax, or not tax, each transaction in a chain and not to look at the overall effect of the transactions in that chain: a fortiori, it is not appropriate to consider the overall effect of separate transactions which are not even in the same economic chain. Mr Lasok, too, looks at each transaction separately. But his case has to be that, where the transactions are relevantly connected, it is appropriate to require the Directive to be applied to both transactions. If the taxpayer, by reference to an incorrect transposition of the Directive into domestic law, gains an advantage which the Directive does not contemplate in relation to one of the connected transactions, it cannot at the same time seek the advantage given by the Directive in relation to the other transaction.[141]In the present case, the argument is that the sale and the resale are, to use our phrase, relevantly connected because(i) the same goods are concerned in the original sale and resale(ii) the original sale and the resale are both sales by GMAC(iii) property in the goods never left GMAC(iv) the goods are available for resale by GMAC only because of the terms of the original hp agreement and(v) there has, in economic terms, been provision of the car for a consideration including the resale price.[142]If Mr Lasok is right then the next question is: What sort of connection is required to create such a relevant connection? Is it necessary or sufficient that the same goods are involved – so that for example successive hirings of goods to different consumers would be enough to activate the principle? Is a connection established if the consideration given by one customer is affected by the consideration given by another? Since VAT is a tax borne by a consumer, does the principle apply only if the same consumer is involved? Is the connection established simply because two transactions are effected by the same trader?[143]In the light of this discussion, the resolution of the Windfall Issue requires us to determine whether GMAC has a directly enforceable right under Article 11C(1) for bad debt relief in respect of the original sale in circumstances where it has sought to rely on the Cars Order in respect of the resale; or, putting the issue slightly differently, whether GMAC is prevented from relying on any directly enforceable right under Article 11C(1) by reason of its ability to take the benefit of the Cars Order. That is an issue of EU law. Mr Cordara submits that the Windfall Issue is acte clair in favour of GMAC. HMRC cannot, he would say, rely on the incorrect transposition of the Directive by the UK to deprive GMAC of its clear rights under Article 11C(1). We do not consider that point to be acte clair. Assuming against GMAC that the sale and the resale are relevantly linked, then the purposes of the Directive can be seen to include the taxation of both the sale (subject to relief pursuant to Article 11C(1)/Regulation 38) and the resale; to allow relief under Article 11C(1) without bringing into account the absence, as a result of domestic law, of VAT on the resale would be to bring about a situation which, as a result of the application of the Directive itself, would be contrary to the purpose of the Directive. 40[144]The issue could, we suppose, be articulated the other way round. Mr Lasok’s case appears to be that GMAC has a choice of which regime – the Directive or domestic legislation – should apply. Suppose that GMAC had not yet sought to apply domestic legislation and were to seek bad debt relief pursuant to its directly enforceable rights under Article 11C(1). What would be the position if GMAC subsequently sought to apply the Cars Order in relation to the resale? That might be said to be a question of domestic law, in effect, whether it would be possible to disapply domestic legislation. That, we think, would be an impossible argument. But what would be possible is that, having subsequently applied the Cars Order to the resale, HMRC would be able to claim that the relief given pursuant to Article 11C(1) should then be retrospectively withdrawn. That again throws one back onto a question of EU law.[145]We do not consider that we can deal with the Windfall Issue without a ruling from the ECJ. Whether we should make a reference is something which we will come to later in this decision. VI. The Third Issue: Time limits Further legislative history 146.We have explained above how the Old Scheme for bad debt relief was brought in by FA 1978 and became settled in section 22 VATA 1983; that in 1990 it was replaced by the New Scheme; and that there was an overlap period between 1 April 1989 and 26 July 1990 when a taxpayer had a choice of which scheme to use. We now need to describe in more detail the statutory provisions making the changes which form the background to section 39(5) FA 1997 which prevented further claims under the Old Scheme after 19 March 1997. 147.It must be borne in mind that the periods to which the two statutory schemes related were defined by reference to the time when a supply was made rather than the date of any claim to relief: a bad debt might arise some time after a supply had been made so that, for example, a bad debt arising after 26 July 1990 by reference to a supply made in say 1988 was eligible for relief under the Old Scheme only. This is particularly relevant in the case of hp agreements under which payment obligations may last for many years. Accordingly, to talk of the Old Scheme being repealed and being replaced by the New Scheme is slightly misleading. The New Scheme never applied to supplies made before 1 April 1989: the New Scheme did not replace the Old Scheme in relation to such supplies: instead, the eventual total repeal of the Old Scheme resulted in no bad debt relief being available thereafter for such supplies where a claim for bad debt relief had not been made, even in cases where the bad debt had not yet arisen by the time of the repeal. 148.It is not difficult to envisage cases where this could have led to relief being unavailable. For example, consider a 5-year hp agreement entered into with GMAC on 31 March 1989 so that a bad debt, should one materialise, could only fall within the Old Scheme (ignoring for the purpose of this example the Property Condition). Suppose that the contract was terminated in the final year (say on 1 41 January 1994). In order to claim bad debt relief, GMAC would have had to prove in the insolvency of the customer. GMAC may have had perfectly sound commercial reasons for not pursuing, itself, insolvency proceedings. But even if it had done so, it may well not have been able to have obtained a bankruptcy order or a winding up order and to have proved in the bankruptcy or insolvency by 18 March 1997, let alone to have submitted a claim for relief. 149.Section 11 FA 1990, having set out the New Scheme for supplies made on or after 1 April 1989, provided in subsection (9) that section 22 “shall not apply to any supply” made after 26 July 1990, the date of the passing of the Act. As we have already stated, a claim could not be made under the New Scheme relating to a supply in the overlapping period if relief was claimed under the Old Scheme: see section 11(8). Mr Lasok has referred on a number of occasions to the repeal of section 22 by FA 1990. The relevant repeal in FA 1990 is to be found in section 132 together with Part III Schedule 19. Section 22 is repealed to the extent specified in Schedule 19 but this applies only to supplies made after 26 July 1990. There was no general repeal of section 22, which remained fully in force in relation to supplies made before 26 July 1990. The continued application of section 22 to supplies made before 26 July 1990 cannot, we think, be seen as intended to be simply a temporary arrangement so far as concerns those supplies. We do not consider that the provisions of FA 1990 can be seen as a signal that the provisions of the Old Scheme would be likely, in the future, to be repealed rather than to be allowed to run their course. 150.In 1994, the VAT legislation was consolidated in the VAT Act 1994. This Act reenacted the New Scheme in section 36. Schedule 13, entitled Transitional Provisions and Savings, provided at paragraph 9(1) as follows:
"(1) claims for refunds of VAT relating to supplies made before 27 July 1990 may continue to be made in accordance with section 22 of the 1983 Act notwithstanding the repeal of that section by the Finance Act 1990"
[151]The closing words of that provision are slightly odd given that the repeal of section 22 (pursuant to section 132 and Part III Schedule 19 FA 1990) had effect only in respect of supplies made after 26 July 1990. However, the VAT Act 1994 (see Schedule 15) repealed the whole of the VAT Act 1983 and thus repealed section 22 to the extent that it was still in force. The saving provision in paragraph 9(1) Schedule 13 VAT Act 1994 clearly allowed claims to be made under section 22 notwithstanding the general repeal of the VAT Act 1983. It is correct, no doubt, to describe this as a transitional provision in the sense that it provides for the Old Scheme to apply in relation to old supplies and would, in the course of time, become exhausted. But that is not to say that it was transitional in the sense of being merely temporary in the expectation that it would in due course be abrogated. As with FA 1990, we do not consider that the provisions of the VATA 1994 can be seen as a signal that the provisions of the Old Scheme would be likely, in the future, to be repealed rather than to be allowed to run their course. 152.And then, in 1997, section 39(5) FA 1997 proscribed further claims under the Old Scheme: 42 "(5) No claim for a refund may be made in accordance with section 22 of the Value Added Tax Act 1983 (old scheme for bad debt relief) at any time after the date on which this Act is passed [19 March 1997 ]". 153.FA 1997 also formally repealed the provisions in paragraph 9(1) Schedule 13 VATA 1994 under which claims for supplies before 27 July 1990 could continue to be made under the Old Scheme. 154.Thus up to 19 March 1997 the taxpayer could make a claim under the Old Scheme in respect of a supply made before 27 July 1990 if the debt on that supply went bad (and the relevant conditions of the regulations were satisfied), and thereafter he could not. If the supply was made before 1 April 1989 (in which case a claim could not be made under the New Scheme), he therefore ceased to be able to claim bad debt relief at all under domestic legislation. Section 121 FA 2008 155.Section 121 FA 2008 provided:
“(1) The requirement in section 80(4) VATA 1994 that a claim under that section be made within 3 years of the relevant date does not apply to a claim in respect of an amount brought into account, or paid, for a prescribed accounting period ending before 4 December 1996 if the claim is made before 1 April 2009.” 156.Section 80 VATA 1994 provided: "(1) Where a person -- (a) has accounted to the Commissioners for VAT for a prescribed accounting period (whenever ended), and (b) in doing so, has brought into account as output tax an amount that was not output tax due, the Commissioners shall be liable to credit the person with that amount. (2) the Commissioners shall only be liable to repay an amount under this section on a claim being made for the purpose. … (4) the Commissioners shall not be liable on a claim under this section [to credit or repay an amount] if the claim is made more than three years after the relevant date." The Advertisement of the Withdrawal of the Ability to make a Section 22 claim 157.Mr Lasok places reliance on the advertisement of the demise of the Old Scheme: 43 a. He records that VAT Notes No 2, 1990 published in September 1990 (after the passing of the Finance Act 1990 and after the end of the overlap period on 26 July 1990) provided that "The new scheme will replace the existing arrangements for bad debt relief, but for an interim period both will operate" It might be said that this indicates that the Old Scheme will be available only for an interim period in the sense that, if a claim is not made within that period, it will cease to be available. Indeed, Mr Lasok submits that that is precisely what is indicated. He relies, among other things, on the use of the word “will” which, he says, shows that the interim period cannot be a reference to the overlap period which had already come to an end. However, it is far from clear that that is a proper reading of the Note. What it actually says following the sentence just quoted is this: “…..The time when you made the supply which led to the bad debt will decide whether you use the old or the new scheme. This means that (a) any supply you made before 1 April 1989 will be eligible for relief under the old scheme……; (b) any supply you made after 26 July 1990… will be eligible for relief only under the new scheme……; (c) a supply you made between 1 April 1989 and 26 July 1990 will be eligible for relief under either the old or the new scheme…..”
[emphasis in original] We find it entirely unclear whether the “interim period” was intended to refer to the overlap period identified in the third bullet point or to some other period. It is perfectly possible to read it in the former sense, reading “for an interim period” as “in respect of an interim period”. On that approach, the overlap period is an “interim period” in the sense that bad debts arising out of supplies made during that (interim) period can be dealt with under either Scheme. We do not consider that that approach is precluded by the use of the word “will” in the phase “but for an interim period both will operate”. Another possible reading, however, focuses on the period during which the Scheme operates. The Scheme will, of course, operate so as to provide bad debt relief as and when a bad debt is identified in the future, that is to say in some period extending past the date of the Notice itself. However, even if that is the correct reading, it does not follow that the “interim period” is one which was intended to come to an end whilst any potential claims to bad debt relief under the Old Scheme remained. There is nothing in the Notes, read as a whole, to suggest that the relief applicable to supplies made before 1 April 1989 was liable to be 44 withdrawn leaving nothing – the New Scheme would not be applicable – in its place. The use of the phrase “interim period” would be no more than an acknowledgement that old supplies would be dealt with under the Old Scheme and that, in the course of time, the Old Scheme would be exhausted. But even if that latter approach is correct, it can hardly be said that the Notes are a clear statement that the ability to make a claim under the Old Scheme in respect of supplies made before the end of the interim period would be withdrawn at some time in the future. b. Before the 1996 budget (which gave rise to FA 1997) there was a public consultation on bad debt relief. There was nothing however before us to illuminate the scope of that consultation. A Budget News Release of 26 November 1996 explained the changes to the scheme which would be made: there would be a clawback of input VAT from those defaulting payers who had claimed it; the provision which required the six months to run from the time of supply before relief could be claimed would be amended to run from the time the debt was due to prevent relief being claimed too early; and there would be other changes:
"designed to help businesses and clarify the law. These will be effective from the time the Finance Bill receives Royal Assent. These changes will: [remove the Property Condition]; [change the method of claiming relief]; “introduce a definition of “bad debt”; clarify the VAT period in which bad debt relief can be claimed; cancel the VAT Regulations covering the old (pre-1989) scheme of Bad Debt Relief."
Note 7 of the Notes for Editors stated:
“Details for traders are available in Budget notice BN 48/96”. c. We do not know the date of Budget Notice 48/96 but it is likely to have been 26 November as well: it was clearly available by then at latest. It contained the same list of changes as the Budget News Release. Paragraph 4 stated that a number of the changes (including the cancellation of the Regulations) would come into effect “on Royal Assent [in the event, 19 March 1997] to the Finance Bill”
. And paragraph 5 stated that "amendment will be made to VAT Act s 36, Sch 13 para 9 and the VAT Regulations 1995 Parts VIII, XIX".[158]We consider that a reasonably well-informed reader of the Budget News Release and the Budget Notice would have read them as an indication that the Old Scheme was to be brought to an end. He might, on a very careful reading, reflect that a different reading was possible. He might then have wondered whether the 45 repeal of the Regulations affecting the Old Scheme would be followed by their replacement by new administrative mechanics aligned to those of the New Scheme and whether the amendments to paragraph 9 schedule 13 VATA 1994 might not have involved the repeal of the Old Scheme.[159]However, the Finance Bill referred to in the Budget News Release and the Budget Notice was published a week later on 3 December 1996. The Bill, as amended, was enacted as FA 1997. Clause 39 of the Bill dealt with bad debt relief. Clause 39(5), dealing with termination of the Old Scheme, was reflected in section 39(5) of FA 1997. Anyone reading the Budget News Release and the Budget Notice would have known that details would be found in the Finance Bill and anyone reading and understanding the Bill would have known that the Old Scheme was to come to an end once the Bill passed into law. Discussion[160]In order to address the issue of the time limits (if any) applicable to GMAC’s claim, we find it helpful to consider the question how a person who asserts a directly enforceable right under EU law is to make his claim. The national court has to provide a remedy in one way or another and it will often be able to do so by an appropriate moulding of domestic legislation. Mr Lasok submits that the domestic provisions and procedural rules which apply to bad debt claims, that is to say section 22 and relevant regulations, are to be applied to GMAC’s directly enforceable rights. In contrast, Mr. Cordara argues that GMAC's claims should not be considered to lie under section 22 at all, but are to be treated as separate direct claims to rights under the Directive. That argument is based on the premise that section 22 cannot be read so as to cover the current claim.[161]It is convenient at this stage to consider one of the authorities on which Mr Lasok relies, Rewe v Landwirtschaftskammer Saarland (Case 33/76) [1976] ECR 1989. In that case the taxpayer had paid certain charges for inspection of its imports (French apples imported to Germany) which were regarded as equivalent to customs duties which meant that the charges had been illegally imposed. The taxpayer therefore had a Community Law right to claim repayment of the illegally levied charges. But the exercise of that right was barred by a domestic time limit. The ECJ held that a domestic limitation period could be a defence to the exercise of a Community right so long as the domestic rules satisfied the principles of equivalence and effectiveness: it was "for the domestic legal system of each Member State to designate the Court having jurisdiction and to determine the procedural actions at law intended to ensure the protection of the rights which citizens have from the direct effect of community law, it being understood that such conditions cannot be less favourable than those relating to similar actions of a domestic nature." The judgement of the ECJ does not (and nor does the opinion of the Advocate General) go so far as to say that where domestic legislation offers a system for enforcing a right parallel to a community right, then that system must be adopted for the right under Community Law but it recognises that a domestic system can apply to such a right. 46[162]It is argued, on the basis of that decision, that UK law applied, through section 22 and regulations, a procedural mechanism, including time limits, for making bad debt claims in cases where the conditions of section 22 were fulfilled. Those same provisions should apply, appropriately moulded if necessary, to the directly enforceable right. Section 22 VATA 1983 163.We start by asking whether GMAC’s claim is properly to be treated as made under section 22. That section contained in subsections (1)(b) and (2)(a), the Insolvency Condition and, in subsection (2)(c), the Property Condition. On the footing that those conditions are incompatible with EU law, the question then is whether GMAC’s directly enforceable rights should be given effect through the mechanism of section 22 but with some appropriate “moulding” to reflect the invalidity of those Conditions. 164.Where domestic legislation is inconsistent with EU law, in the sense that the legislation provides for something which EU law does not allow, the wellestablished jurisprudence (see for instance Autologic plc v Inland Revenue Commissioners [2006] 1 AC 118 citing R v Secretary of State for Transport, Ex p Factortame Ltd [1990] 2 AC 85 ,140, and Imperial Chemical Industries plc v Colmer (No 2) [1999] 1 WLR 2035, 2041), demonstrates that the relevant legislation is to be read as being without prejudice to any directly enforceable right under EU law. 165.In the present case, UK domestic legislation, section 22, did provide for a form of bad debt relief in the shape of the Old Scheme. The Old Scheme was intended to reflect the obligations of the UK under Article 11C. On the hypothesis now under consideration, the Property Condition and the Insolvency Condition were impermissible restrictions on the right to relief. But that does not detract from the fact that section 22 was the mechanism by which bad debt relief was to be delivered. We see no reason why the approach of the cases just mentioned should not be applied to GMAC’s claims for bad debt relief. As we see it, section 22 can, and should, be read as providing the mechanism for giving effect to GMAC’s directly enforceable claims. Where there is a partial or total non-payment within the meaning of Article 11C(1), section 22 is to be read as providing relief in relation to that non-payment by disapplying the Property Condition and the Insolvency Condition. 166.Mr Cordara says in his skeleton argument that there is no way in which section 22 can be read as covering GMAC’s claims. That succinct submission no doubt reflects what the Tribunal recorded him as saying, namely that section 22 could not be made compliant with EU law by reading in words such as “without prejudice to the directly enforceable community rights” because that would go against the grain of section 22. Like the Tribunal, we consider that section 22 can properly be seen as the mechanism by which effect is to be given to GMAC’s directly enforceable rights to bad debt relief under the Directive. 47 167.But that is not the end of this part of the discussion. Domestic legislation provides for a claim to be made in accordance with regulations (see section 22(3)), which may provide for the matters set out in paragraphs (a) to (e) of that subsection. It is worth noting that paragraph (b) as originally enacted authorises, by way of example, regulations to require a claim to be evidenced and quantified by reference to such records and other documents preserved for such period, not exceeding 3 years, from the making of the claim as may be specified. Regulations might specify certain records and other documents which would be inappropriate to a claim based on a directly enforceable right under the Directive and any requirement to provide such records or documents would have to be disapplied. But equally, regulations might specify other records and documents which would be as relevant to a directly enforceable claim as to a claim actually falling within the section. It would be curious if a taxpayer in the position of GMAC were able to avoid having to provide those records and documents on the footing that section 22 and the regulations do not apply in any respect to directly enforceable claims. 168.Turning to the regulations themselves, the first set (relevant to the predecessor of section 22, section 12 Finance Act 1978) was the Value Added Tax (Bad Debt Relief) Regulations 1978. These provided: "3.(1) Save as the Commissioners may otherwise allow or direct, the claimant shall make a claim to the Commissioners by including in box 8 of the return prescribed in paragraph (2) of this regulation the correct amount of the refund.(2) the return shall be the one ... for the prescribed accounting period during which [the claimant] received the document mentioned in subparagraph (a) of regulation 4 ...[4]Save as the Commissioners may allow, the claimant before he makes a claim, shall hold --(a) a document issued to him by the person with whom he proves in the insolvency ...(b) [an invoice or similar document] and(c) [records showing he had accounted for and paid the tax]. 169.The section 22 regulations also dealt with the treatment of set off and the allocation of monies paid between different supplies. They provided for an obligation to repay the bad debt refund if record-keeping conditions were breached. 170.In 1981 (by the Value Added Tax (General and Bad Debt Relief)(Amendment) Regulations 1981, SI 1981/1080,) the requirement to make an entry in box 8 of the VAT return was replaced by a requirement to include the sum in the box on the return "Over declarations of tax made in a previous return". The return had at this time a box to be ticked if bad debt relief claims were included. The amendment 48 did not change the description of the relevant return period. The 1978 and 1981 regulations were revoked by the Value Added Tax (Bad Debt) Regulations 1986, SI 1986/335. Regulations 4 and 5 effectively continued the previous provisions. 171.The regulations thus set out a procedure intimately connected with the Insolvency Condition - the claim had been made in the VAT return for the VAT period in which the claimant received the document evidencing his proof in the insolvency. It is only by reference to that procedure that any time-limit for making the claim is imposed. The regulations contain no other provisions relating to any time limit for the making of a claim. That procedure cannot, of course, apply to the directly enforceable claims which GMAC assert since the Insolvency Condition is not fulfilled: there is not, and never will be, a document evidencing proof in an insolvency. 172.Accordingly, the requirement to make a claim in the way specified in the regulations will have to be disapplied and adapted (to use the word used by Lord Nicholls in Autologic at [17]) in some way to give effect to GMAC’s directly enforceable rights. In that context, it might be said that the provisions of the regulations concerning VAT returns do apply to a directly enforceable claim at least to this extent, namely that the claim must be made by including the correct amount of the refund in the box identified. But that does not lead one anywhere in terms of time limits since what those provisions do not do is to identify which return is the appropriate return in the case of a directly enforceable claim. It would go far beyond any permissible adaptation of the statutory provisions to treat them as identifying some other VAT period as the relevant period and could lead to anomalies. 173.Furthermore, subject to the impact of Allen¸ to which we will come in due course, we do not consider that there is to be implied, as a matter of domestic law, any limitation period for the making of a claim. So although GMAC’s claims can, and in our view should, be treated as made under section 22, neither that section nor the regulations made under it result in the imposition of a time limit within which the claims must be made. 174.We conclude that section 22 does provide the mechanism for giving effect to GMAC’s directly enforceable rights; but it does not result in the imposition of any time-limit for the making of a claim. We find some support for the first of those conclusions in Rewe which we have already considered. 175.If we are right in those conclusions, it follows that section 83(1)(h) VATA 1994 provided the jurisdiction for the Tribunal to deal with GMAC’s appeal in relation to supplies made before 1 April 1989. This was the conclusion which the Tribunal themselves reached; and we agree with them. Section 83(1) provides for an appeal to lie to the First-tier Tribunal in relation to the matters listed. Paragraph (h) refers to “a claim for a refund under section 36 or section 22 of the 1983 Act”. Once it is accepted that section 22 provides the mechanism through which GMAC’s bad debt relief claims are to be recognised, the case falls, in our view, within paragraph (h). The claims are ones for refunds “under…. section 49 22…”. Even if, contrary to our view, the claims are not strictly under section 22, but are claims which are given effect to as if section 22 applies as a result of the necessary adaptation, there should, we consider, be a parallel adaptation of paragraph (h) to bring within its scope the claims which GMAC now makes. So far as concerns bad debt claims made in respect of later supplies, the position was covered for the relevant years under appeal by section 36 VATA 1994. There is no equivalent, for these claims, to section 39(5) FA 1997 so that, subject to the argument based on Allen that EU law itself imposes a reasonable time-limit, there is no time-limit issue on which HMRC can succeed. 176.At this stage, we turn to consider the alternative jurisdictional gateways on which Mr Cordara relies. These arise not only in relation to HMRC’s own appeal, but are central to GMAC’s cross-appeal. Mr Cordara relies on paragraphs (b) and (t) of section 83 VAT 1994. Paragraph (b) relates to “the VAT chargeable on the supply of goods…..” and paragraph (t) relates to “a claim for the crediting of an amount under section 80…..”. He relies on these provisions because, on his case, section 22 and section 36 are not applicable; he therefore needs to find another route by which (a) substantive effect is given to GMAC’s directly enforceable rights and (b) the First-tier Tribunal is given jurisdiction over a dispute about those rights. And the reason why he seeks a different route for GMAC to assert its claims is so that the difficulties presented by the repeal of the Old Scheme can be avoided 177.As to paragraph (b), Mr Cordara submits that this looks at the nature of the underlying tax dispute. On his approach, the current dispute is simply a dispute about the application of Article 11C(1), which is headed “Taxable Amount”. A dispute about the taxable amount is precisely a dispute about the VAT chargeable on the supply of goods. The substantive right which GMAC seeks to enforce is simply its claim to bad debt relief under the Directive; jurisdiction to resolve the dispute is to be found in section 83(1)(b) and there is no need to find a substantive right under domestic legislation to found that jurisdiction. We see the force of this argument given the premise that section 22 is not applicable. But if section 22 is applicable, as we think it is, it really takes the matter no further. This is because when one addresses the dispute about the amount of VAT chargeable for the purposes of section 83(1)(b), one is thrown back to the provisions of section 22 itself: it makes no difference whether jurisdiction is founded on paragraph (b) or paragraph (h). The issue is then what, if any, claim GMAC has when it comes to applying section 22 in the light of the bringing to an end of the Old Scheme pursuant to FA 1997. 178.This meets Mr Cordara’s argument based on section 121. He would, no doubt, be correct in saying that there was no time limit, before 1 April 2009, for the making of a claim if the claim was properly made as one for “output tax not due”. But that is not the proper basis of the claim which is one under section 22 as appropriately adapted and moulded. 179.As to paragraph (t), we have set out the relevant parts of section 80 at paragraph 176 above. Section 80 applies only if GMAC (a) has accounted for tax and (b) in 50 doing so, has brought into account as output tax an amount not due as output tax. Mr Cordara submits that section 80 is not focussed on any category of underlying dispute but merely focuses on situations in which a refund of tax is due for whatever reasons. Thus, GMAC’s directly enforceable claims result in a reduction of the amount chargeable to VAT so that GMAC has brought into account as output tax an amount that was not output tax due. As in the case of paragraph (b), and for the same reasons, this does not take matters further once it is accepted that section 22 does apply, paragraph (t), like paragraph (b), throws one back to section 22 and the effect of the bringing to an end of the Old Scheme. 180.Mr Lasok submits that section 80 does not apply to claims for bad debt relief. He makes the point that bad debt relief under domestic law results in a refund of VAT so that, according to him, GMAC is wrong in asserting that, as it did not make a claim for bad debt relief when it arose, it has overpaid VAT. It is wrong because if no claim for bad debt relief is made, no refund can be granted and there is therefore no overpayment. Mr Cordara submits that the point about a bad debt claim resulting in a refund goes nowhere, suggesting that a claim under section 80 itself is, or can be, a claim for a refund. It is true that unless a refund claim is made, no bad debt relief can be granted but that, he says, is equally true of section 80 itself. He points to section 80(5) (now repealed) which provided that where an amount had been paid by reason of a mistake, a claim for the repayment of the amount under section 80 itself could be made within 6 years of discovery of the mistake (subject to a reasonable diligence test). However, that point gets Mr Cordara nowhere. If there was a mistake, then the output tax was never properly due and the case fell squarely within section 80(1)(b). The position in relation to bad debt relief is rather different. And to demonstrate why that is so, we need to say a little more about section 80. 181.Section 80 is concerned with cases where a taxpayer has brought into account as output tax an amount that was not output tax due. When GMAC made its supplies and accounted for the full amount of output tax, it accounted for an amount of output tax which was then due: it is only the subsequent failure of the customer to pay which has resulted in any possible claim for bad debt relief. It does not seem to us that later circumstances giving rise to a bad debt for the purposes of Article 11C(1) and which results in a reduction in the chargeable amount renders the amount which was actually paid retrospectively incorrect in the sense that it can be said that the amount actually paid was “not output tax due” within section 80. It was, when paid, output tax which was due; and remained such until a bad debt arose. 182.However, if we are wrong about the applicability of section 22, then GMAC would have paid more output VAT than was due. Section 80 (2) provides:
"The Commissioners shall only be liable to repay an amount under this section on the claim being made for the purpose."
If a claim is made before 1 April 2009, section 121 FA 2008 provides that the 3 year time-limit under section 80(4) does not apply. The appeal to the Tribunal related to the ruling given in HMRC’s letter dated 18 July 2006 in response to GMAC’s bad debt relief claim made on 20 February 2006. Although GMAC’s letter was concerned with section 36 of the 51 VAT Act 1994, reference was made also to the Old Scheme. It seems clear that the claims being made were made as claims for bad debt relief under the relevant statutory provisions relating to bad debt relief. No claim was made in that letter under section 80 of the VAT Act 1994: indeed, the letter contains the following paragraph:
“In our opinion, there is no time limit that currently precludes the making of Bad Debt Relief claims in respect of supplies made between the first introduction of a Bad Debt regime from 2 October 1978 to 1 May 1997. The claim therefore falls outside section 80 (“Credit for, or repayment of, overstated overpaid VAT”) because, in principle, it does not relate to an overpayment of VAT (that is, a payment by way of VAT that was not VAT due to the Commissioners at the time.)” 183.In our view, GMAC’s letter cannot be read as an alternative claim to credit or repayment under section 80(2). We know of no other claim. It is therefore too late to rely on section 121. Whether there are any bad debts remaining which can be made the subject of a claim within the section 80(4) time limit, we do not know. 184.Once it had become apparent that the taxable amount should be reduced pursuant to Article 11C(1), it would be open to the taxpayer to claim appropriate relief. If, as Mr Cordara submits, section 22 does not apply, there is no domestic provision which indicates how or when the relief is to be given. But it is obvious, we think, that the onus is on the taxpayer to make a claim; in the absence of a claim, HMRC would have no way of knowing that a bad debt had arisen. It follows, unless and until a claim is indicated, that it cannot be said that any relief is to be afforded and that it cannot be said that any amount has been brought into account as output tax that was not output tax due. Accordingly, section 80 does not, in our judgment, in terms apply to GMAC’s claims. 185.That is not to say that section 80 could not be moulded or adapted in order to give effect to GMAC’s directly enforceable rights just as we have decided that section 22 can be moulded or adapted. The necessary adaptation would be to include within the words “not output tax due” an amount by which output tax is reduced following the giving of bad debt relief pursuant to Article 11C(1). We do not consider adaptation of section 80 to be as appropriate a way of giving effect, under domestic law, to GMAC’s directly enforceable rights as the adaptation of section 22 which we have already addressed. Section 39(5) FA 1997 186.The provisions of section 39(5) apply to everyone with a potential claim under section 22. To the extent that such persons had claims which they could at a future time have made under section 22, section 39(5) provided that they could not be made after 19 March 1997. Two issues arise out of this provision in relation to GMAC’s claims. The first is whether the bringing to an end of relief under section 22 automatically brought to an end GMAC’s right under Article 11C(1) to 52 claim bad debt relief for the period in question in this appeal (that is to say from 1978 to 1 April 1989). It is to be remembered that at this stage of the debate, we are proceeding on the basis that domestic legislation fails to provide GMAC with relief and that it is therefore necessary to rely on the direct effect of Article 11C(1). The second issue, assuming that GMAC could, prima facie, no longer bring such claims, is whether it was necessary for taxpayers to have been given notice of the termination of the Old Scheme and, if so, whether adequate notice was given. 187.As to the first of those issues, we have already decided that section 22 and the regulations are to be adapted so as to provide the mechanism, coupled with section 83(h) by which GMAC’s rights are given effect under domestic law. Section 39(5) provided that no claim could be made under section 22 after 19 March 1997. It is clear, we think, that, subject to the second issue, section 39(5) precludes GMAC’s claim being made through the mechanism of section 22. 188.In this context, we should re-iterate the important point that GMAC’s claim to bad debt relief was one which it could first have asserted under EU law many years ago. Those claims would have been subject to domestic time-limits had any been imposed. Where there is such a domestic time-limit, it will apply, subject to the principle of effectiveness (which, we would add, applies to a domestic claim too) to directly enforceable rights under EU law even if those rights have not been established by a decision of the Court and are disputed by the Member State concerned. This was the position in Case C-188/95 Fantask A/S and Others v Industriministeriet (Erhvervsministeriet) [1997] ECR I-6783. That case, like the present case, concerned (assuming GMAC succeeds on the Compatibility Issue) the incorrect transposition of a directive. 189.The question then is whether the effect of section 39(5) in relation to GMAC’s claims is to bring the possibility of such claims to end in the same way as it brings to an end a claim which did fall within the letter of section 22; or whether, in contrast, the termination of the mechanism provided by section 22 and the regulations (appropriately adapted) for GMAC to make a claim in relation to its directly enforceable rights leaves those rights extant with the need, therefore, for the tribunals to provide another means of enforcement. In our judgment, the first of these alternatives is correct. Once it is accepted (as is our decision) that GMAC’s rights are to be effected under section 22 and the regulations, appropriately adapted, the ending of the possibility of claims under section 22 should apply as much to directly enforceable rights as it does to purely domestic rights. 190.We turn next to the second issue, the question of notice. Was it necessary for taxpayers generally to be given notice of the termination of the Old Scheme and if so was adequate notice given? So far as domestic law is concerned, it is clear when the claim would first arise, namely when the Insolvency Condition was fulfilled. In the case of a directly enforceable claim, it is less clear when the claim would first arise. It would, we suppose, be when the facts first fell within Article 11C(1) absent any derogation by the Member State concerned but it is not at all 53 clear to us when that would be. It is, of course, a factual question in any particular case, but what would be sufficient in order for a taxpayer to establish a directly enforceable claim is not, at least to us, obvious. A few things are, however, clear. The first is that many of GMAC’s bad debts giving rise to a directly enforceable claim (assuming the invalidity of the Property Condition and the Insolvency Condition) arose many years before the passing of the FA 1997. The second is that in relation to supplies made before 1 April 1989, many, if not most, directly enforceable claims to relief under Article 11C(1) would have arisen well before 18 March 1997. We do not know the details here and they probably do not matter. The third, related to the second, is that, at least in theory and probably in reality, there will, in respect of such supplies, have been some bad debts which arose for the first time shortly before or even after 18 March 1997. Some hp agreements or other agreements within section 22 can give rise to obligations running over several years and the possibility of a bad debt arising for the first time shortly before or after 18 March 1997 cannot be dismissed. 191.In relation to the question of notice, there is a difference in philosophy between the parties about the nature of the effect of section 39(5). Mr Lasok submits that the issue is not one about time-limits at all. He says that section 39(5) is simply the final stage of a change from the Old Scheme to the New Scheme. He views the provisions of section 11 FA 1990 and Schedule 13 VATA 1994 as transitional provisions continuing the Old Scheme for a temporary period with the final termination of the Old Scheme – something which everyone knew would be coming at some time – being effected by section 39(5). Mr Cordara, in contrast, submits that section 39(5), viewed purposively, is simply a provision which imposes a time limit on the making a claim in respect of a subsisting right. He submits that the case is really no different from Fleming (t/a) Bodycraft v HMRC [2008] UKHL 2 (“Fleming”), a decision which we will turn to in a moment. 192.It is, of course, true as Mr Lasok points out that the UK government was entitled to change the conditions under which bad debt relief was available pursuant to its powers of derogation from Article 11C(1). He submits that this is precisely what it did by substituting the New Scheme for the Old Scheme. Taxpayers in general and GMAC in particular had no legitimate expectation that the Old Scheme would remain in force unchanged. However, that, it seems to us, is only half the story. It was never the case (except in relation to supplies made during the overlap period from 1 April 1989 to 26 July 1990) that the New Scheme was substituted for the Old Scheme. In relation to supplies made before 1 April 1989, it was the Old Scheme alone which applied; the New Scheme could never apply. The termination of the Old Scheme therefore had the result that it ceased to be possible to obtain bad debt relief in relation to supplies made before that date; there was no question of some alternative scheme of relief being available. So far as those supplies are concerned, the position was this: immediately before the coming into effect of section 39(5), GMAC had claims (under section 22 and the regulations as appropriately adapted) for bad debt relief. Some of those claims were old, in the sense that they could have been made years earlier, but others were more recent and, indeed in relation to supplies made towards the end of the relevant period (say in early 1989) there may have been debts which transpired to be bad debts 54 only shortly before or even on or after 19 March 1997. After that date, GMAC had no claims at all if section 39(5) is to be given its apparent temporal effect. 193.We turn now to Fleming. This case related to claims for input tax repayment under regulation 29 of the VAT Regulations 1995. Those regulations, it is to be noted, required a claim to be made in the period in which the VAT became chargeable, but (in the words “save as the Commissioners may otherwise allow or direct”) reserved a discretion to the Commissioners. 194.A new regulation 29A was introduced which removed the right under regulation 29 to claim a deduction more than three years after the return date for the period in which the VAT became chargeable. The question was whether this was permissible without the creation of a sufficient transitional period in which outstanding claims could be made, and whether the way in which the change had been advertised created such a period. Lord Hope (see at [1]), Lord Carswell (see at [77]) and Lord Neuberger (see at [103]), and as such, a majority of the House, all appear to have considered that the communication of a transitional period might be made by the Administration rather than the legislature, but said that it must be widely disseminated. 195.On the basis of Fleming and the decisions of the Court referred to, it might be thought that just as the effect of the enactment of regulation 29A was to prevent a taxpayer exercising an EU right given by domestic conforming legislation so too, in the present case, the effect of the enactment of section 39(5) was to prevent a taxpayer exercising such accrued EU rights as he may have under domestic legislation adapted or moulded so as to conform with the Directive, as well as precluding altogether claims in relation to bad debts arising after 18 March 1997. On that footing, proper notice of the change effected by section 39(5) would need to have been given in relation to accrued rights. And a serious question would arise whether the termination of relief under section 22 was permissible at all in relation to post 18 March 1997 bad debts in respect of pre-1 April 1989 supplies without an alternative scheme of relief applicable to such debts being created. 196.Mr Lasok submits that the present case is different from Fleming. The present case is not one of “a retrospective imposition of a time-limit when before there was none” to use the words of the Tribunal at [102] of the Decision, because it concerns the discontinuance of a legal regime applicable during a particular period of time, on the expiry of a transitional period, rather than a change in the limitation period applicable to claims made under that regime. In Fleming, the disapplication of the (invalid) statutory time-limit removed a bar on the exercise of a right to make a claim under the Regulation 29 regime; in the present case, there is no similar bar on the exercise of any right that GMAC might have had; instead, the underlying legal regime itself has changed. It is, we acknowledge, correct that the regime has changed, but as we have already observed, so far as supplies made before 1April 1989 are concerned, the regime has not changed, rather it has been abolished without replacement. 55 197.Mr Lasok says the decision in Fleming was based on the proposition that transitional provisions were needed to protect legitimate expectations. There being no legitimate expectation that any particular bad debt regime would continue unaltered, he submits that Fleming has no scope for application. He goes on to submit that the present case is more akin to cases such as Cases C- 487/01 and C-7/02 Gemeente Leusden and Holin Grop v Staatsecretaris va Financien [2004] ECR I-5337 and Case C-201/08 Plantanol GmbH & Co KG v Hauptsollant Darmstadt [2009] ECR I-8343. In these cases the exercise by member states of options permitted by the Directive on no (or short) notice was upheld by the ECJ. He says that Article 11C(1) permitted derogations and conditions, and that the Member States could vary the derogations and conditions which they adopted from time to time: the discontinuation of one regime (the Old Scheme) and its replacement by another (the New Scheme) was merely such an action. The notice of the change was adequately disseminated. 198.In Gemeente Leusden the Dutch government took advantage of the permission given to Member States in the Directive to permit taxpayers to opt for supplies of letting to be taxable rather than exempt. Holland then withdrew the right to opt in certain circumstances with no effective transitional period. This disadvantaged Gemeente Leusden because it had arranged its affairs on the basis that its letting income would be taxable in the future (and it would get input tax credit). The ECJ acknowledged the Dutch government’s right under the Directive to change its option and addressed the question whether principles of legitimate expectation and legal certainty precluded the effect which the changes would have on Gemeente Leusden [65]. It noted that a change in the State’s option under the Directive could not be regarded as unforeseeable and thus that no legitimate expectation arose that it would not change [69]. It then considered whether a sudden and unexpected change breached those principles; it held that: "a taxable person cannot rely upon there being no legislative amendment, but he can only call into question the arrangements for the implementation of such an amendment". The ECJ held that the only question was whether or not the change in the national legislation breached the principles of legitimate expectations and legal certainty (see [70]): “in that, without taking account of a legitimate expectation of taxable persons which had to be protected, it suddenly and unexpectedly withdrew the right to opt for taxation of lettings of immovable property, when the objective to be attained did not require it, without allowing taxable persons bound by leases current at the time of entry into force of the law the time to adjust to the new legislative situation.” going on to hold that the notification given (by press release) was adequate. 199.Likewise in Plantanol, the sudden withdrawal of a favourable tax regime for biofuel was held not to offend the principle of legitimate expectation where a 56 prudent and circumspect taxpayer could have foreseen the possibility of such a change. 200.We agree with Mr Lasok that a taxpayer could not have a legitimate expectation that any particular scheme of derogation from the Directive adopted by a Member State would remain in force unchanged. As it was said in Plantanol, economic operators have no legitimate expectation in the continuance of an existing situation which is capable of being altered by the national authorities in the exercise of their discretionary powers. But even if he is right in saying that the present case is more akin to Gemeente Leusden and Plantanol than it is to Fleming, it also has to be recognised, as we think it was in Gemeente Leusden, that there would be a breach of legitimate expectations if a Member State suddenly and unexpectedly withdrew a particular regime in relation to a particular subject matter; in particular if the UK had suddenly and unexpectedly withdrawn the Old Scheme (particularly if, in doing so, it did not replace the Old Scheme so far as concerns old supplies). 201.In both Gemeente Leusden and Plantanol, it is to be noted that the regime which resulted from the changes in each case were permissible end results in accordance with EU law. In contrast, the absence of any bad debt relief in the present case in relation to supplies made before 1 April 1989 was not in accordance with EU law save to the extent the right to claim such relief had become time-barred by some permissible temporal limitation imposed under national law (or under EU law itself: this is the Allen point to which we will come later). As is well established, the imposition of a reasonable limitation period is not incompatible with EU law. 202.In our view, the position in the present case is much closer to Fleming than it is to Gemeente Leusden and Plantanol. First of all, in both the present case and in Fleming, the taxpayer has a claim (in the present case, a directly enforceable right under Article 11C(1) on the footing that the Insolvency Condition and the Property Condition are incompatible with EU law) to reduce its liability for tax. In contrast, the impact of the changes in Gemeente Leusden and Plantanol was on the amount of tax which would become payable because of the way the transactions in question would be charged. Secondly, the total exclusion of bad debt relief for supplies made before 1 April 1989 – in contrast with the adoption of a replacement scheme applicable to such supplies – cannot, we consider, be justified as a condition or derogation within Article 11C(1). The only justification, as we see it, is that the elimination of claims under section 22 (as appropriately adapted for directly enforceable claims) is the imposition of a reasonable time limit within which such claims must be made. The fact that a replacement scheme might have been adopted (and in such a case, the Gemeente Leusden and Plantanol approach might well be correct) is not an answer in the case where the section 22 claim was altogether abolished. A taxpayer in the position of GMAC with an accrued directly enforceable claim was, we consider, in substantially the same position as Mr Fleming. In our view, GMAC had a legitimate expectation that the period during which it would be able to make a claim for bad debt relief in the absence of any replacement scheme would not be brought to an end without an adequate opportunity being given to make a claim. 57 Thus, just as the introduction of a shortened time limit without a transitional period in Fleming breached the principles of effectiveness and legitimate expectations, so, in the present case, the termination of the right to make a claim under section 22 without an adequate opportunity to make a claim would breach those principles unless an adequate transitional period was provided for. 203.That conclusion leads on to the question of notice and transitional provisions. In addressing that question, it is important to see precisely what it was that the House of Lords decided in Fleming. That case concerned the retrospective shortening of a time limit for making a claim. It applied in respect of accrued rights. It was held, that to be compliant with EU law (i) the new time limit had to be fixed in advance so as to give legal certainty (ii) where the new time limit was retrospective, there had to be an adequate transitional provision so that those with accrued rights had a reasonable time within which to make their claims before the new time limit applied, it being for Parliament, or HMRC by means of an announcement disseminated to taxpayers to introduce prospectively an adequate transitional period (iii) that where a new time limit was introduced without any, or any reasonable, transitional period, it would be a breach of EU law to enforce the new time limit in relation to accrued rights at least for a reasonable period (iv) that the adequacy of the transitional period was to be determined by reference to the principles of effectiveness and legitimate expectations, so that the period was not so short as to render it practically impossible or excessively difficult for a person with an accrued right to make his claim and (v) where the national court decided that the transitional period was inadequate, it had to fashion the remedy necessary to avoid an infringement of EU law which would normally be to disapply either permanently or temporarily the operation of the retrospective application of the new time limit. As to (iii), the reasonable period must itself be certain. This appears most clearly from the speech of Lord Neuberger at [88] and [90], although it is implicit in the speeches of Lords Hope and Scott and of Lord Carswell too. 204.On the basis of those holdings, it can be argued as follows: (i) section 39(5) imposed a retrospective time limit in that it precluded a person with an accrued right prior to 18 March 1997 from making a claim (ii) it did not contain any transitional provision nor has any announcement been made since the commencement of section 39(5) about the period within which claims could be made (iii) although clear notice of an impending change in the legislation might be sufficient to satisfy the requirements of the principles of certainty, legitimate expectations and effectiveness, there was no sufficient notification on the facts so that (iv) the operation of section 39(5) must be suspended until certainty is provided either by legislation or an appropriate announcement. 205.There are three differences between Fleming and the present case which we would highlight: a. The first is that, in Fleming, the complaint of the taxpayer was that his accrued rights were adversely affected by the shortening of the period within which he had to make a claim. HMRC attempted to correct the position with announcements (made in the Business Briefs referred) made after the relevant 58 legislation had come into force. In the present case, GMAC’s complaint, on the footing that section 22 applied, is again that its accrued rights have been adversely affected by the bringing to an end of relief under that section. But in this case, HMRC’s position is that, if (contrary to their primary case) notice is relevant at all, adequate notice was given by the VAT Notes, the Budget Notice and the Finance Bill to which we have referred before the legislation was passed. It is said that these, whether taken separately or cumulatively, were sufficient to put taxpayers generally on notice that the Old Scheme was coming to an end so that, on the footing that section 22 applied, they had the opportunity to make claims in respect of bad debts arising from supplies made from 1978 to 1989. This is particularly so given that the possible ending of the Old Scheme had been signalled, according to Mr Lasok, by the repeal of section 22 contained in FA 1990 and its continuance in VAT Act 1984 under the heading “Transitional Provisions and Savings”. b. The second is that, in Fleming, the taxpayer made his claim within a reasonable time of the date of the start date identified by the House of Lords. We use the words “start date” to refer to what Lord Neuberger referred to as “the start of the end of disapplication period”
. That is to say the date from which it is appropriate to start measuring the appropriate period within which the taxpayer must make his claim. In the present case, GMAC did not in fact make its claims until some time in 2006, around 9 years after the Old Scheme was finally terminated. c. The third is that, in Fleming, the taxpayer’s substantive rights were clear: the case was concerned with the introduction of time limits. In the present case, GMAC’s substantive rights are disputed by HMRC who assert that the Property Condition and the Insolvency Condition are valid and that the Windfall Issue provides an answer to the claim even if those Conditions are invalid. 206.As to the third of those differences, we do not think that it helps GMAC’s case. As we have explained, Fantask shows that a domestic time limit can be effective against a claim under EU law. We see no reason to apply a different approach to the introduction of a time limit for the first time or to the shortening of an existing time limit in the case of a right which has been validly transposed into domestic law and one which has not. Thus, suppose that the UK had transposed the right which GMAC claims into domestic law without imposing any time limit for making a claim and suppose that it then sought to introduce such a time limit. Fleming would apply. We see no material difference between that case and a case where the right has not been properly transposed into domestic law. In such a case, the protection afforded to taxpayers as a result of Fleming must be observed. But provided that this is observed, the introduction of a time limit should be as effective in relation to the directly enforceable right as it is in relation to the domestic right; and it should be effective notwithstanding that the existence of that right is not appreciated by the taxpayer and that HMRC deny its existence. 59 207.We do not consider that the first two of those differences lead to the conclusion that Fleming is materially distinguishable. It is no doubt the case, that an announcement made prior to the passing of legislation to the effect that the Old Scheme would be withdrawn from a specified future date is capable of satisfying the principles of certainty, effectiveness and legitimate expectations in relation to accrued rights at the date of the announcement and to some extent in relation to rights which accrue between that date and the date of the withdrawal. The announcement would fulfil the same function as the sort of communication which the House of Lords in Fleming acknowledged would be sufficient to bring about an effective disapplication of the offending statutory provisions. Further, we consider that the period thereby afforded to the taxpayer to make his claim could be certain enough to satisfy the legal requirement for certainty. Although the end date of the period (the commencement of the legislation) could not be known for sure at the date of the announcement, taxpayers and their advisers would have a very good idea of when an announced proposal to be included in the next Finance Bill would be likely to become law. The decision in Allen which we will come to lends support to that last conclusion if any support is needed. 208.Whether the principles of effectiveness and legitimate expectations were fulfilled thus depends on the length of notice given of the termination of the Old Scheme and thus of the right to make a claim in respect of accrued rights. The point was made in Fleming (see for instance Lord Walker at [67] and Lord Neuberger (perhaps making a slightly different point) at [84]) that the adequate period required pursuant to Grundig II within which a taxpayer should be allowed to make a claim in the absence of the provision of an express transitional period did not equate with the adequate period of an express transitional relief. In our view, the present case is one where, in contrast with Fleming, the adequacy of the notice should be judged by reference to the period appropriate to an express transitional period and not by reference to a reasonable period of disapplication. This is because the period from the date of the announcement to the date when the legislation comes into effect is analogous to an express transitional period. To use Lord Walker’s own analogy, the announcements in the present case tell us when, within a small time-frame, the last bus or tube will be departing so you know when public transport stops (the small time-frame reflects the fact that the precise date of the commencement of the proposed legislation is not known at the date of the announcement). 209.The outstanding questions are then whether the announcements made were effective notice of the ending of the Old Scheme and, if so, whether that notice was of adequate duration. 210.It is to be noted that the focus of our discussion of the Time-limit Issue so far has been on GMAC’s directly enforceable rights on the footing that the Property Condition and the Insolvency Condition are incompatible with EU law. However, there may well have been other cases where a taxpayer fulfilled the Property Condition and would have been able without too much difficulty to satisfy the Insolvency Conditions and, if he had done so, to claim bad debt relief strictly in accordance with domestic legislation. Consider, then, the case of such a taxpayer 60 who, immediately prior to 26 November 1996, owed a debt which the taxpayer considered would never be paid. Assuming that the taxpayer knew of the announcements and the Finance Bill, he would have realised that he needed to satisfy the Insolvency Condition to be able to make a claim under domestic legislation. It is clear, in our view, that if it was necessary, as a matter of EU law, for such a taxpayer to be provided with a transitional period sufficient to enable him to satisfy the Insolvency Condition, the period from 26 November 1996 to 18 March 1997 would have been insufficient. In the case of a debtor who is an individual, it would have been necessary to serve a statutory demand which could not be prepared overnight. It would then have been necessary to present a bankruptcy petition on the basis of that demand once the time for payment had expired. Even if the debtor did not challenge the debt or the petition, that would all take time. After that, the creditor would have had to prove in the bankruptcy following the appointment of a trustee in bankruptcy. Possibly with the application of adequate resources and the devotion of sufficient time, it would have been possible to reach the position where the Insolvency Condition was satisfied by 18 March 1997 but that is not the same thing as saying that a transitional period expiring before that date would be adequate. In our view, it clearly would not be. Similar, although not identical, considerations apply in the case of corporate insolvency. 211.Given our rejection of Mr Lasok’s submission that there is no scope for the application of Fleming, we consider that such a taxpayer who wished to take advantage of the domestic legislation cannot simply be deprived of that opportunity without an adequate transitional period, especially given that he was not provided with any replacement bad debt regime at all, as we have explained. The UK government had provided a scheme of bad debt relief in the shape of the Old Scheme which it alleged, and continues to allege, was compliant with EU law. HMRC cannot now, when it comes to determining whether such a taxpayer is entitled to a transitional period, and if so what period, rely on our decision that the Old Scheme was not in fact compliant, to deprive the taxpayer of that opportunity. 212.GMAC is not, of course, such a taxpayer since it did not, and cannot now, fulfil the Property Condition. But the fact that there may be other taxpayers who could seek to rely on a transitional period sufficient to allow the Insolvency Condition to be satisfied has an impact, in our view, on the appropriate transitional period in respect of GMAC’s directly enforceable claims. It would be wrong in principle, we consider, for GMAC (and others in the same position) to be required to bring a claim within a shorter period than that applicable to persons potentially entitled to make such a claim under domestic legislation. A person seeking to enforce a claim under EU law must not be treated in a less favourable way than a person seeking to enforce an equivalent right under domestic law. Accordingly, we conclude that the period from 26 November 1996 to 18 March 1997 was not an adequate transitional period within which GMAC could be required to assert its directly enforceable rights. 213.It might be said that it is wrong to identify the appropriate transitional period for GMAC’s directly enforceable claims by reference to the period appropriate to a 61 taxpayer seeking to rely on domestic legislation. But even if we are wrong to do so, the Budget News Release, the Budget Notice and the Finance Bill which we have referred to at paragraphs 157b above did not, in our view and on the basis of the evidence before the Tribunal and the arguments which we have received, give adequate notice of the end of the Old Scheme so as to provide sufficient time for those with accrued directly enforceable rights to exercise them even viewing those rights in isolation. We are influenced in reaching that conclusion by what others – the Advocate General and the ECJ in Grundig and the House of Lords in Fleming - have said about the adequate length of a transitional period. In Grundig, it is true that the Advocate General had said (see [27] of his Opinion) that it was not possible to determine whether the 90 day transitional period provided for in the Italian legislation was adequate without having regard to all the factual and legal requirements which the domestic legal order imposed for bringing actions for recovery, matters within the competence of the Italian courts alone. But the ECJ itself seemed willing to assess an adequate transitional period as 6 months. In Fleming, Lord Neuberger considered that, on the basis of the limited argument and evidence which the House had received on the point, the duration of a transitional period (not, it is to be noted the (longer) period of disapplication) would be between 6 months and 12 months. In the present case, we, too, have had only limited evidence and we bear in mind that the necessary transitional period is to be determined in relation to very different circumstances from those in Fleming and Grundig. But with those steers from the ECJ and the House of Lords, it appears to us that a period of 16 weeks (from, at earliest 26 November 1996 when the announcements were made) until 19 March 1997 (from which date it became impossible to make claims under the Old Scheme) was inadequate. 214.Even if we had concluded that the transitional period was adequate in relation to accrued rights subsisting on 26 November 1996, the period for making a claim in respect of bad debts arising after that date became shorter as time went by and disappeared altogether on 18 March 1997. Given our conclusion about the adequacy of the transitional period actually provided for, we do not need to say any more about the difficulties which would face HMRC about identifying precisely where the dividing line is to be drawn. 215.On the basis of our conclusion that there was an inadequate transitional period, we do not consider that HMRC can say that, nonetheless, GMAC’s claims are barred because they were not brought within a reasonable disapplication period. In Fleming, it had been argued that the start date should be taken as the date of one of the decisions considered by the House of Lords, namely Grundig II, M&S II and University of Sussex. Those potential start dates were rejected by the majority since the ordinary reasonably well-informed taxpayer would not have appreciated the effects of those decisions and would not have appreciated his right to make a claim within a reasonable disapplication period. That aspect of the decision is binding on us. We cannot rely on the decisions referred to as providing the start date. It might be argued that Fleming itself does provide an answer to that particular reason for the decision in favour of Mr Fleming and would thus provide a start date in relation to GMAC’s claims. But even if that is 62 so, it does not provide a defence to those claims since they were in fact made before the decision in Fleming. 216.Subject to the impact of Allen, which we come to next, our conclusion is that GMAC’s claims are not barred by lapse of time. Allen v Commission Case T - 433/10P 14 December 2011 (“Allen”) 217.Ground 5 of the HMRC’s grounds of appeal asserts that, by the time it made its claims, GMAC was barred from asserting any EU law right which it might have had. This claim is based on the decision in Allen. We should note, at this point, that Allen is relied on not only to establish this free-standing ground of appeal. But Mr Lasok says that it is relevant also to the approach to the fixing and length of appropriate transitional periods and that it has an impact on what the House of Lords said in Fleming. 218.Allen is a decision of the General Court. Since the hearing of this appeal, we have been informed that Allen is not going to be reviewed by the ECJ. It therefore represents the current state of EU law in relation to the subject matter of the decision and is, so Mr Lasok submits, binding on us by virtue of section 3(1) European Communities Act 1972. It was held in Allen in the context of the subject matter there under consideration, that “…there is an obligation to act within a reasonable time in all cases except those where the legislature has expressly excluded or expressly laid down a specific time-limit” Further, the principle of legal certainty “precludes institutions and natural persons from acting without any time-limits”. And the length of the reasonable period for bringing a claim is to be determined by the courts: see [26] of the judgment. 219.In Allen the appellants sought compensation from the Commission for the loss they suffered because they had not been recruited by the Commission while they were working at the Joint European Torus project. This claim followed earlier decisions of the General Court in Eagle v Commission Case T-144/02 [2007] ECR II-2721 and Sanders v Commission Case [2007] T-45/01 ECR II-2665I in which the General Court had held in relation to other appellants that the Commission had acted unlawfully in the same situation, and had fixed damages. The Court said:
“[26] ...it must be held that the appellants’ argument that the absence of a time limit automatically means that it is possible to bring a claim for damages without any time-limit cannot succeed. It should be noted on that point, contrary to what the appellants contend, there is an obligation to act 63 within a reasonable time in all cases except those where the legislature has expressly excluded or expressly laid down a specific time limit. The legal basis for setting a reasonable time limit, in the absence of any statutory rule, is the principle of legal certainty, which precludes institutions and natural persons acting without any time limits thereby threatening to undermine the stability of legal positions already acquired (see to that effect, Case- T192/99 Dunnett and Others v EIB [2001] ECR II-813, paragraphs 51 to 53; Case T-281/01 Huygens v Commission [2004] ECR-SC I-A-203 and II- 903, paragraphs 46 and 47; and Eagle, paragraph 57). Thus, in the absence of any statutory rule it is for the judicature to decide on the length of the reasonable period for submitting a claim for damages, in the light of the circumstances of the case (see, to that effect, Dunnett and Others v EIB, paragraph 54; Huygens v Commission, paragraph 49; Eagle, paragraph 57; and Sanders, paragraph 58)." [italics added] The legislature here referred to is the EU legislature. 220.Then the appellants argued that if a time limit was to be imposed by the court when none had existed under the statutory provisions, there was an obligation to create a transitional period. In relation to this the Court said: “[50] "it must be said that, by the second argument, the appellants seek to draw an analogy between the powers of the legislature and those of the judicature. In that regard, it should be noted that in [Marks & Spencer and Grundig], the Court held that in a situation where a national legislature reduces a limitation period, the compatibility of the new period is subject to the condition that provision is made for it to apply in the future and for the transitional arrangements to be made in order to ensure that individuals are not denied time for lodging claims. However that rule cannot be applied in the present case. In the first place, in the case of judgements, the rule is that they come into force immediately ... Secondly, even if the appellant's situation were exceptional, the analogy is not relevant since in the order under appeal the civil service tribunal did not set a new limitation period by reducing an existing one, it merely reiterated the requirement that a claim for damages must be brought within a reasonable period ... thirdly it must be borne in mind that the interpretation which the European Union judicature gives of a provision of European Union law is limited to clarifying and defining the meaning and scope of that provision as it ought to have been understood and applied from the time of its entry into force. It follows that the provision as thus interpreted must be applied even to legal relationships which arose and were established before the judgement in question was delivered ..." 221.Mr Lasok submits that neither under the Old Scheme nor under the New Scheme (so far as the claim period is concerned) did the relevant legislation, including the Directive, expressly exclude a time-limit or expressly lay one down. Accordingly, under EU law the position, by the time the Old Scheme was repealed (as Mr Lasok would have it) in 1990, was that the principle of legal certainty had barred GMAC from exercising any EU law right to bad debt relief which it might have 64 had before then for all of its claims save those which were still within a “reasonable time”; and only those claims that were still within a reasonable time could have benefited from the transitional period that preserved the Old Scheme after 1990. Further, by the time the transitional period (as he would have it) expired in 1997, GMAC could not have supported any of its pre-1 April 1989 claims by its claimed directly effective rights. Moreover, in relation to New Scheme claims arising in respect of the remainder of the Claim Period (up to 19 March 1997), the directly enforceable rights claimed by GMAC could have been invoked only in relation to those bad debt relief claims that had accrued within a “reasonable time” before the making of the claim: by the time the claim was actually made on 20 February 2006, GMAC had failed to act within a reasonable period of time in respect of the entirety of its claim. 222.Mr. Cordara says that Allen is a judgement which relates to the special selfcontained legal world of the Commission. That world is a creation of the Treaty and not subject to the jurisdiction of any state. The General Court was dealing in that world with a dispute effectively between an employer and its employees, not between a Member state and its subjects. It cannot sensibly be suggested that the established jurisprudence dealing with the effect of the Directive and with the principles of effectiveness, certainty and legitimate expectations has been significantly altered without reference to the wealth of case law. The General Court, it will be remembered, was dealing with a claim for damages against a Community institution, not with a claim by a citizen against a Member State for the benefit of a directive. In the latter case there are principles other than legal certainty at stake for instance, the harmonisation of indirect tax, the principle of effectiveness and the principle that the Member State cannot plead its own failure to implement the Directive. 223.Although Allen was, on its facts, dealing with the special situation of an EU institution where the legal relationship was governed by EU law in the absence of any relevant domestic law, that is not to say that it has nothing useful to say about EU law more generally as applied to the Directive and the relationship between a taxpayer and a Member State. EU law as applied within the institutions of the EU might be thought to represent the paradigm of “pure”
EU law establishing the legal order which should permeate all aspects of EU law as applied to Member States and their citizens. And just as what the General Court said in Allen may be of some assistance in understanding the relevant principles of EU law generally, so the established jurisprudence developed by the ECJ in the context of EU law as applied across Member States was taken account of in Allen. If that were not so, there would have been no reference to cases such as Marks and Spencer and Grundig. The General Court clearly felt the need to say not simply that those cases provided no analogy with the case before it, but to explain why it did not do so. 224.There is this significant difference, however, between Allen and a case concerned with a directive. In a case such as Allen, the relevant court (the General Court in that case) could look only to EU law; unless the court itself imposed a time-limit, one could not be found elsewhere and it was considered that the legal order 65 required the imposition of a time-limit (in that case for claiming damages) in the interests of legal certainty. In contrast, the ECJ has repeatedly said that timelimits are a matter for Member States so that, provided the principles of effectiveness and legitimate expectations are observed, the principle of legal certainty allows for the imposition of time limits in relation to rights arising under national law following proper implementation of a directive. Accordingly, the absence of a time-limit in a directive so far as concerns a right conferred by that directive does not have the same significance as the absence of a time-limit in an EU statute dealing only with rights between institutions of the Union among themselves or with individuals connected with them (as in Allen). 225.Consider the situation where a Member State properly transposes a directive but provides no time-limit in relation to the conferring and claiming of certain rights. The Member State could not then, of course, seek to impose a time-limit by reference to EU law. This, as we see it, would be because EU law does not, in the case of a directive, require the Member State to impose a time-limit: time-limits are left to the Member State. In other words, the transposition of the directive into domestic law without the time-limit would be entirely in accordance with EU law. The EU legal order does not require that Member States are obliged to provide certainty by imposing any, or any particular, time limit. In terms of the analysis in Allen, an express provision in the legislation of a Member State that there should be no time-limit in relation to a particular claim would be as unobjectionable as such a provision in EU legislation; and the absence of an express time-limit in a Member State which, as a matter of domestic law, would not imply such a timelimit, would be equivalent to an express exclusion of a time limit. 226.What, then, is the position if the Member State fails properly to transpose a directive and an individual seeks to rely on his directly enforceable rights? What time limit, if any, applies? Take the present case by way of example. Assuming that the Property Condition and the Insolvency Condition were, as Mr Cordara submits and as we have held, contrary to EU law. GMAC then has a directly enforceable right under Article 11C(1). Our view, for reasons already given at length, is that its claim, in respect of supplies made prior to 1 April 1989, is one to be made under section 22 appropriately adapted and moulded. In relation to that claim, there is, we also consider, no domestic time-limit. It follows that the directly enforceable claim is one in relation to which, under domestic law, there is no time-limit. We do not consider that some overriding EU time-limit comes into play. We do not consider that Allen leads to that conclusion.[227]Furthermore, Allen clearly does nothing to cast doubt on the proposition that it is necessary, when introducing a time-limit for the first time, or in shortening an existing time-limit, to observe the principles of effectiveness and legitimate expectation. What it does do, however, is to illustrate what the principle of certainty requires. We have no reason to think that what the General Court said about certainty applies only to cases such as Allen itself dealing with pure EU law divorced from any national context. But in relation to the need for a newly introduced limitation period to be “fixed in advance” (see Marks and Spencer”), it is not possible sensibly to think that Allen can be seen as leading to the conclusion 66 that this is not necessary. Indeed, the discussion in [50] of the judgment in Allen is not consistent with such a proposition. What might be more open to question is whether, as a matter of EU law, the period during which a new limitation period is to be disapplied, must also be fixed before it can be relied on by the Member State. Why, one might wonder, could the General Court fix the reasonable period during which claims for damages could be made in a pure EU law case, when, at the same time, EU law would require a Member State to provide a disapplication period which was fixed in advance. 228.In Fleming, it was decided by the majority (Lord Walker taking a different approach) that it was not for the court to fix a limitation period when the UK legislature had not done so. It was not for the court to fill the gaps. The decision of the majority in refusing to fix a limitation period did not, therefore, turn on the need for the period of disapplication of a new time limit to be fixed. Accordingly, we do not consider that Allen provides any basis for us to revisit the decision in Fleming even if, in principle, we would be able to do so in the light of the development of EU law. However, Allen does provide some reason for doubting the second reason given by Lord Neuberger for rejecting HMRC’s primary case. What he said at [88] was that the period of disapplication had to be fixed just as a valid limitation period must be fixed. Allen does lend support to the view that the fixing can be done by the domestic court and is not necessarily to be fixed by the legislature or other authorities such as HMRC. 229.We would add that Allen cannot, on any view, throw doubt on the decision in Fleming insofar as it decides that none of the decisions in Grundig II, M&S II and University of Sussex can be taken as providing the start date for the disapplication period. And see further at paragraph 239 below. 230.So far as New Scheme claims under section 36 VAT Act 1994 are concerned, that section, like section 22, requires adaptation and moulding in order to give effect to GMAC’s directly enforceable rights assuming the invalidity of the Property Condition. In relation to those rights, domestic legislation provided no time-limit. We do not consider that Allen can be relied on by HMRC to require us to give effect to some alleged overriding EU requirement in relation to the Directive that claims for bad debt relief must be made within a reasonable time (presumably a reasonable time from it becoming established that a bad debt has arisen). 231.If we are wrong in thinking that section 22 and section 36 provide the domestic mechanism for making GMAC’s claims, we have already expressed the view that section 80 does so, in which case section 80(4) applies to exclude claims in respect of debts which are now out of time under that subsection. Claims which are still in time can, of course, still be made. Allen has no impact. To the extent that claims are time-barred, there is no need to rely on Allen. To the extent that they are not time-barred, they are within a time-limit which is expressly laid down in domestic legislation. The UK could not rely on Allen, even if it were to be taken as laying down the principle for which Mr Lasok contends, to override section 80 in favour of the fisc. 67 232.We would add only this in relation to Allen. That case concerned a free-standing claim for damages as the result of an alleged breach of Community legislation. One can readily see why the General Court concluded that considerations of legal certainty required such claims to be brought within a reasonable time. In the present case, a claim for bad debt relief under Article 11C(1) ought in principle to be allowed without a time-limit in order to preserve the integrity of the VAT system. Time limits for the making of claims can, no doubt, be introduced by Member States in the interests of legal certainty; but such time-limits are not an absolute requirement in order to provide that certainty. 233.Interesting as it is, we do not think that there is sufficient merit in Mr Lasok’s argument in relation to Allen to justify a reference to the ECJ Conclusion on time limits and the making of claims 234.GMAC’s claims are not barred by lapse of time. Accordingly, claims in respect of such bad debts arising before 19 March 1997 in relation to supplies made before 1 April 1989 can still be made under section 22 notwithstanding its repeal. So far as concerns bad debts in relation to such supplies arising after 18 March 1997, these did not give rise to accrued rights which fell within section 22. Instead, neither section 22 (which was repealed) nor section 36 (which applied only to supplies made after 1 April 1989) could provide the mechanism for giving rise to a taxpayer’s directly enforceable rights in respect of such debts. Accordingly it is, in our view, section 80 which provides the mechanism for giving relief. But the result is then as set out in paragraphs 182 to 184 above. The Preliminary Issues in the BT matter 235.The Preliminary Issues raised in the BT matter raise substantially the same questions as those which we have discussed in relation to the GMAC appeal. There is an additional aspect which relates to the position concerning supplies made in the period 1 January 1978 to 1 October 1978. The former date is that date by which the Directive was to be implemented by Member States. The latter date marks the commencement of the Old Scheme. During that period, there were no domestic provisions in the UK implementing Article 11C(1). 236.The Preliminary Issues arise out of a decision by HMRC contained in a letter dated 11 January 2010 upheld on review. The review decision is to be found in a letter dated 1 April 2010. The review decision refused a claim for bad debt relief by BT in a letter dated 30 March 2009. Its claim was in an amount of nearly £92 million plus interest in respect of “unclaimed bad debt relief in the period 1 January 1978 to 31 March 1989”. We have already explained the relevance of the former date. The claim period ends on 31 March 1989, from which date the New Scheme was available to BT. In the claim letter, the view was expressed that BT could make “a claim for bad debts suffered with solvent customers in the period 1 January 1978 to 31 March 1989”. There are two points which arise out of that. The first is that we are concerned only with bad debts which arose on or before 31 March 1989: we are not concerned with cases where the supply was before that 68 date but the bad debt arose after that date (a point of distinction from GMAC). The second is that we take the reference to solvent customers as a reference to customers who had not been adjudged bankrupt or which had not been put into insolvent liquidation. There may have been insolvent customers who had not been subject to formal insolvency proceedings, but we see no distinction for present purposes between them and solvent individuals or companies. 237.The three Preliminary Issues all relate to the time within which claims must be made and are as follows:a. Issue 1: On the assumption that BT could otherwise have relied on an EU law right to bad debt relief, in respect of bad debts allegedly arising in the prescribed accounting periods running from 1 January 1978 to 31 March 1989, by virtue of Article 11C(1) of the Sixth VAT Directive, was the exercise of that right in 2009 barred in accordance with the general principles of EU law and/or subject to section 39(5) of the Finance Act 1997?b. Issue 2: If the answer to Question 1 is in the negative in relation to the general principles of EU law, but affirmative in relation to section 39(5), does section 39(5) fall to be disapplied, or construed, under EU law, in such a way as not to affect the exercise of BT's right under EU law?c. Issue 3: Do section 80 of the Value Added Tax Act 1994 and section 121 of the Finance Act 2008 apply to BT's claim irrespective of the answer to Question 1? 238.Issues 1 and 2: These issues are closely connected. It is convenient to deal with them together. It follows from our decision in the GMAC appeal that, in relation to a taxpayer’s directly enforceable rights in respect of supplies made during the period from 1 October 1978 to 1 April 1989, claims in respect of bad debts arising during that period are not time-barred as a result of section 39(5) FA 1997. On our analysis, although that section introduced an effective time bar to claims where there was not one previously, adequate notice was not given of the impending change with the result that claims which accrued prior to 1 April 1989 could still be made after 18 March 1997 unless an appropriate disapplication period had been specified and had expired. The approach in Fleming must be applied with the result that subject to one point, the claim made by BT in 2009 was still in time. We have rejected the submission that Allen would apply to impose a time-bar as a matter of EU law. 239.The point just mentioned is this. It might be argued that Fleming itself provides the start date for the running of the disapplication period and that, when BT made its claim in March 2009, that period had expired since, on any view, a period of over a year from the decision in Fleming (January 2008) and the making of the claim was more than adequate. The problem with that argument is that the majority in Fleming decided that a disapplication period had to be of a fixed duration and specified in advance. The only way out of that difficulty facing HMRC is to say that Allen shows that the principle of certainty does not require a 69 fixed period to be specified. But even if that is correct, on which it is not necessary to express a view, it would only be, at the very earliest, once the decision in Allen had been published that taxpayers would know that instead of there being the need for a fixed period, the mere passage of reasonable time could bar their claims. Just as there had to be a disapplication period in relation to an inadequate transitional period, it must be the case, in our judgment, that a reasonable period after the decision in Allen for bringing a claim must be allowed. The decision in Allen did not appear until December 2011; BT’s claims were made two years before this. Accordingly, we conclude that BT’s claims are not barred by lapse of time. 240.It was not necessary, in the GMAC appeal, to address supplies made in the period 1 January 1978 to 1 October 1978. There are two situations to consider. First, where the bad debt also arose in that period; secondly where it did so after 1 October 1978. In relation to the latter situation, it is clear that section 12 FA 1978 could apply even where the supply pre-dated 1 October 1978. That appears from section 12(6) which provides for the section to apply where the debtor becomes insolvent after that date: there is no reason to doubt that an insolvency taking place after that date would engage section 12 even where the supply was before that date. On that footing, the necessary adaptation or moulding of section 12 can be effected to give effect to a taxpayer’s directly enforceable rights in respect of supplies made before 1 October 1978. The resulting position is, in our view, no different from that which obtains in respect of supplies made after 1 October 1978. 241.So far as concerns bad debts arising before 1 October 1978, the position is marginally less clear. At the time when the directly enforceable right arose, section 12 was not on the statute book and effect could not, at that time, have been given to the directly enforceable right through the mechanism of section 12. However, once that mechanism came into being, we see no reason to conclude that it should not apply to the taxpayer’s directly enforceable rights. The position is then precisely the same as that addressed in the preceding paragraph. If we are wrong about that, then the mechanism has to be found elsewhere, an aspect we consider when dealing with Issue 3. 242.Issue 3: For the purposes of addressing this Issue, we will assume for the moment that BT had directly enforceable rights at the time when section 80 first came into force. For the reasons given in the GMAC appeal, we do not consider that section 80 applies at all, since the mechanism for giving effect to BT’s directly enforceable rights was to be found in section 22. There was no time-limit for making such a claim until section 39(5) came into force. If we are wrong in thinking that section 12 provided the appropriate mechanism for making a claim then some other route must be found. For reasons given in the GMAC appeal, one possible route is section 80 on the footing that BT has brought into account as output tax an amount which was not output tax due. So let it be supposed that, when section 80 first came into force, BT did have a directly enforceable right. The time limit for making a claim was originally 6 years subsequently reduced to 3 years. That time limit was subsequently made inapplicable in certain 70 circumstance as a result of section 121 FA 2008 which was treated as coming into force on 19 March 2008: see section 121(4). There were two conditions. First, the disapplication of the time limit only applied where the claim was in respect of an amount brought into account, or paid, for a prescribed accounting period ending before 4 December 1996. Secondly, it applied only where a claim was made before 1 April 2009. As to the first of those conditions, BT’s claims, if they subsisted under section 80 in the first place, were clearly within the first of those circumstances. But, so far as we can see, no claim was made by BT for relief under section 80 before 1 April 2009 or indeed at all. The claim letter dated 30 March 2009 asserted a right to a refund. The whole thrust of the letter was that the Insolvency Condition was invalid. Further, a mistaken view was taken of the repeal of section 22, the complaint being made that bad debt relief in respect of pre-1 April 1989 supplies was being abolished by FA 1990. That complaint was misconceived but the fact that it was made indicates that BT’s case was that relief should be granted as precisely that, namely bad debt relief. There was no suggestion that section 80 had any role to play or that there had been a payment of an amount which was not output tax due. 243.Our conclusions in relation to section 80 are therefore as follows. First of all, section 80 is not relevant at all because BT’s claims fall to be dealt within under section 22. Secondly, even if BT’s claims did fall within section 80(1), no claim under section 80(2) was made prior to 1 April 2009 so that the disapplication of the time limit in section 80(4) could not be relied upon. That leaves BT fixed with the ordinary time limit under section 80(4) and whether that is 6 years or 3 years, no claim was made within those periods and any claim would now be years out of time. It could not now be maintained, either, that the 3 or 6 year time limit should not be applied on the basis of the principle of effectiveness since section 121 provided the vindication of BT’s rights which EU law required. 244.The answers to the Preliminary Issues are therefore as follows: a. Issue 1: On the assumption that BT could otherwise have relied on an EU law right to bad debt relief, in respect of bad debts allegedly arising in the prescribed accounting periods running from 1 January 1978 to 31 March 1989, by virtue of Article 11C(1) of the Sixth VAT Directive, the exercise of that right in 2009 was not barred in accordance with the general principles of EU law but was no longer available as a result of section 39(5) FA 1997. b. Issue 2: Section 39(5) falls to be disapplied, or construed, under EU law, in such a way as not to affect the exercise of BT's right under EU law. This conclusion turns on our view that inadequate notice of the termination of the Old Scheme was given. Accordingly, BT’s claims were not time-barred when they made them in the claim letter dated 30 March 2009. c. Issue 3: Section 80 of the Value Added Tax Act 1994 and section 121 of the FA 2008 have no relevance to BT's claim on the footing that its claims arise under section 22. If its claims arise, instead, under section 80, those claims were not made before 1 April 2009 and are now time-barred. 71 Reference to the ECJ 245.We do not consider that either the Compatibility Issue or the Time-limit Issue raise any question which needs to be referred. Although we have not found the Time-limit Issue at all straightforward, our decision turns, ultimately, on whether the period from 26 November 1996 to 18 March 1997 was an adequate transitional period. A question of proportionality of that sort is a matter for the national court in accordance with the principles as explained and expounded by the ECJ. The principles are clear. The fact that they are difficult to apply and that different judges may take different views does not lead to the making of a reference. However, to succeed on its appeal GMAC also need to succeed on the Windfall Issue (a matter not relevant to BT). The Windfall Issue raises matters which are not acte clair and to which we would need to know the answer before being able to rule on it. We will consider whether or not a reference needs to be made on this issue at this stage of the proceedings in the light of further submissions. Mr Justice Warren Charles Hellier Chamber President Upper Tribunal Judge Release Date: 03 August 2012

Cited in 5 later judgments