Investment Trust Companies v HM Revenue and Customs

Reduced to its simplest terms, the nature of the problem which arises in the present case may be described as follows. An end customer of goods or services has for many years been charged VAT at the standard rate by the supplier of those goods or services. The customer has paid the VAT, and the supplier has duly accounted for it as output tax to HMRC (a term which I will use to include their statutory predecessors, the Commissioners for HM Customs and Excise). It then transpires, as a result of a ruling by the Court of Justice of the European Union (“the ECJ”), that the supplies in question have at all material times been exempt, with the consequence that the VAT was unlawfully charged. To the maximum extent permitted by UK domestic law, the supplier then recovers from HMRC the VAT which it has overpaid, and passes on the benefit of that recovery to the customer. There are two reasons, however, why the supplier may be unable to recover from HMRC the full amount of the unlawful VAT paid by the customer. The first reason is the statutory three year limitation period applicable at the material time to claims by a taxable person to recover overpaid VAT from HMRC. The second reason is that the supplier will have paid to HMRC in the first place only a net amount representing the difference between the output tax on the supply to the customer and any associated input tax incurred by the supplier. The amount of the overpaid tax which the supplier can recover from HMRC, and thus pass on to the customer, is therefore confined to the net amount of tax which the supplier has actually paid to HMRC, as well as being subject to the three year limitation period.

In these circumstances, the question arises whether the end customer, who has borne the full economic burden of the unlawful VAT, can bring a direct claim against HMRC to recover the balance of the tax which it paid to the supplier. That is the fundamental issue in these lead claims, which were the subject of legal argument before me on questions relating to liability over four and a half days in May 2011.

In a little more detail, the customers (and therefore the claimants) in the cases before me are all closed-end investment funds constituted as limited companies, or “investment trusts” as they are generally known. The term “closed-end” denotes that the trust is established with a fixed number of shares in issue and a term date when it will be wound up and its assets will be distributed to the shareholders. The relevant services were investment management services provided to them by management companies (“the Managers”), under the terms of investment management agreements which typically provided for the Managers to be remunerated by the payment of fees plus VAT “if applicable” (or words to similar effect). Under the UK VAT legislation then in force, it was generally understood that these services did not qualify for exemption, although there was from 1990 an express exemption for investment management services supplied to authorised unit trust schemes, and in 1997 this exemption was extended to open-ended investment companies (i.e. investment trusts without a fixed termination date). Accordingly, despite the existence of these exemptions, the Managers continued from 1990 onwards to charge VAT at the standard rate on the management services which they rendered to the claimants, and the claimants paid the VAT on the assumption, and in the belief, that it was legally due.

The failure of the UK legislation to provide an exemption for fund management services supplied to closed-end investment trusts was eventually challenged on the ground that it was incompatible with Article 13B(d)(6) of the Sixth VAT Directive, which provided that the “management of special investment funds as defined by Member States” should be exempt from VAT. One important issue was whether this wording gave Member States a wide discretion to decide which management services to exempt, or whether the area of choice was more narrowly confined, for example by the principle of fiscal neutrality which requires similar economic operations to be treated in the same way for VAT purposes. Another issue was whether closed-end investment trusts were “special investment funds” within the meaning of the exemption. A further issue was whether Article 13B(d)(6) had direct effect, so that it could be relied upon by a taxpayer in a national court.

The challenge was made by an investment trust like the claimants in the present action, J P Morgan Fleming Claverhouse Investment Trust, with the backing of the Association of Investment Trust Companies. On 28 June 2007, following a reference by the VAT and Duties Tribunal in London on 19 September 2005, the ECJ ruled in Case C-363/05, J P Morgan Fleming Claverhouse Investment Trust Plc and another v HMRC, [2007] ECR I-5517, [2008] STC 1180, (“Claverhouse”) that:

the words “special investment funds” in Article 13B(d)(6) were capable of including closed-end investment trusts;

the discretion afforded to Member States in defining the funds in their territory to be exempted

Article 13B(d)(6) had direct effect, so that it could be relied on by a taxable person before a national court in order to challenge the application of national legislation alleged to be incompatible with its provisions.

At this stage, it is unnecessary to say more about the judgment in Claverhouse. Although the ECJ left it to the national court (see paragraph 49 of the judgment) to determine whether the exclusion of closed-end investment trusts from the exemption then contained in Items 9 and 10 of Group 5 of Schedule 9 to the Value Added Tax Act 1994 (“VATA 1994”) was consistent with the objective of Article 13B(d)(6) and the principle of fiscal neutrality, the guidance given by the court left no real room for doubt about the answer (see in particular paragraphs 50 to 54). It can therefore have occasioned no surprise when on 7 November 2007 HMRC published Business Brief 65/07, in which they commented on Claverhouse and said:

With effect from 1 October 2008, VATA 1994 was amended (by the Value Added Tax (Finance) (No. 2) Order 2008, SI 2008/2547) to take account of Claverhouse and to comply with EU law. Thus Items 9 and 10 of Group 5 in Schedule 9 now exempt:

an authorised open-ended investment company; or

an authorised unit trust scheme;

As a result of Claverhouse, it became clear that:

the supplies of investment management services by the Managers to the claimants should have been exempt from VAT under Article 13B(d)(6) from 1 January 1990 onwards;

Article 13B(d)(6) had direct effect at all material times;

between 1 January 1990 and 1 October 2008 the United Kingdom failed to transpose Article 13B(d)(6) correctly into its national legislation; and

the Managers were entitled to make claims to HMRC pursuant to section 80 of VATA 1994 for the crediting and repayment of sums accounted for and paid by them in error by way of VAT on investment management services.

Before describing the repayment claims, I will first say a little more about the position of the claimants.

The nine claimants are all investment trusts (a term which I will use hereafter to mean closed-end investment trusts, unless the contrary is stated). They are all in members’ voluntary liquidation, not due to insolvency but because their term dates have arrived and this is the means by which their assets are realised and distributed to the shareholders. The joint liquidators are in each case Richard Setchim and Ian Oakley Smith, who are respectively a partner and director of PricewaterhouseCoopers LLP. The present claims are brought by the liquidators on behalf of the claimants, and if they succeed the proceeds (after deduction of costs and expenses) will be distributed to the shareholders.

Before going into liquidation, each of the investment trusts carried on the business of investing funds subscribed by the public in various asset portfolios, with a view to obtaining capital and/or income returns. As I have already said, the Managers supplied investment management services to the claimants, in return for fees on which VAT was charged and paid at the standard rate. The Manager was often, but not always, a company associated with the investment trust.

Four of the claimants were at all material times registered in the UK for VAT purposes, but the other five were not. The reason for this distinction is explained by Malcolm Richardson, who has been Head of Tax at the M & G Group of companies since 2008, in his witness statement. Where an investment trust invests only in stocks, shares and other securities within the EU, its activities are exempt from VAT and it makes no taxable supplies. In practice, there is no requirement for such a business to register for VAT in the UK. If, however, the investment trust invests outside the EU, it is entitled to recover UK input tax in respect of that part of its activity, which is not exempt but is zero-rated. In order to exercise this right, it is necessary for the investment trust to register for VAT. The amount of input tax that the investment trust is entitled to recover is, broadly speaking, the tax on those of its purchase invoices which are directly attributable to the non-EU portfolio plus an appropriate proportion of general expenditure on overheads etc. In the case of a trust which has both EU and non-EU portfolios, the portion of the input tax which the investment trust is entitled to recover is expressed as a percentage known as its partial exemption rate.

By a consent order of Chief Master Winegarten made on 25 January 2010, it was directed that the claims of the third, seventh and eighth claimants should proceed as lead claims, and that the claims of the other claimants should be stayed until further order. Of the three lead claimants, Kleinwort Overseas Investment Trust Plc (“Kleinwort Trust”) was registered for VAT during the claim period, whereas the other two (F & C Income Growth Investment Trust Plc (“F & C Trust”) and M & G Recovery Investment Trust Plc (“M & G Trust”)) were not. The partial exemption rate of Kleinwort Trust is agreed to have been an average of 58.4%, with the consequence that its claim is for 41.6% of the VAT which it paid to its Manager. F & C Trust and M & G Trust, on the other hand, were unable to recover any input tax from HMRC and therefore bore the full burden of the VAT which they were charged on the management services supplied to them. Their claims are accordingly for the full amount of the VAT they were charged on those services, less recoveries to date.

Details of the relevant payments made by each of the lead claimants are set out in the schedules to Mr Setchim’s first witness statement. For present purposes, the details do not matter. The essential pattern of the payments is helpfully summarised in the claimants’ written submissions, as follows:

The Manager provided investment management services and then invoiced them to the investment trust, including VAT on the invoice.

The investment trust paid the invoice, together with the VAT.

In the case only of Kleinwort Trust, it made periodic VAT returns in which it reclaimed (on average) 58.4% of that VAT as input tax.

The Manager made periodic VAT returns in which it:

accounted for the VAT charges on the investment management services as output tax; and

reclaimed as input tax VAT which it had paid to third parties for supplies attributable to the provision of the investment management services.

The Manager paid HMRC the net difference between the output tax and the input tax shown on its VAT return. These were global returns, and the payments made were of the difference between the Manager’s overall output tax and input tax; however, this included the differences between the output and input tax referable to the investment management services.

The payment flows summarised above are illustrated by two diagrams which the claimants appended to their written submissions, and which I append to this judgment. A few brief comments on these diagrams are called for. The first diagram is a simplified representation of the position of F & C Trust and M & G Trust. The figures of £100 for output tax on the Manager’s services to the investment trust and £25 for the Manager’s associated input tax are, of course, both symbolic and illustrative, and the net payment of £75 made by the Manager to HMRC in its VAT return would in practice vary according to the amount of the input tax. Secondly, the diagram assumes for simplicity that the input tax of “£Y” incurred by the Manager’s own supplier is in fact nil, with the result that there is no deduction from the payment of £25 made by that supplier to HMRC. Thirdly, while agreeing that the diagrams provide a useful illustration of the VAT position, HMRC point out that the Manager’s right to deduct the £25 as input tax is not dependent or conditional upon the Manager’s supplier accounting for that sum to HMRC as output tax, or HMRC receiving any sum from the supplier (the conditions for exercise of the right to deduct input tax are set out, for example, in Case C-152/02, Terra Baubedarf-Handel GmbH v Finanzamt Osterholz-Scharmbeck [2004] ECR I-5583, [2005] STC 525). Finally, the second diagram is a simplified representation of the position of Kleinwort Trust.

Section 80(1) of VATA 1994, as amended by the Finance (No.2) Act 2005, provides that:

has accounted to the Commissioners for VAT for a prescribed accounting period (whenever ended), and

in doing so, has brought into account as output tax an amount that was not output tax due,

I will need to return later to the full provisions of section 80 and the statutory scheme of which it forms part, but at this stage the important points to note are the obligation on HMRC to credit a taxpayer with output tax which has wrongly been brought into account, and the three year time limit in subsection (4) on making a claim for that purpose. There are also complex provisions which give HMRC a defence if the crediting of such an amount would unjustly enrich the claimant. Prima facie, this defence would apply to repayment claims made by the Managers, because the burden of the unlawful VAT was wholly passed on to the investment trusts and the Managers were not themselves left out of pocket. However, the defence is disapplied where “reimbursement arrangements” are made, pursuant to regulations and supported by such undertakings as HMRC may require, with the purpose of ensuring that the taxpayer is not unjustly enriched by the repayment of any amount pursuant to the claim, and that the payment is used to reimburse the customers of the taxpayer who have borne the burden of the tax. In the context of the present case, this meant that the Managers could only claim repayment of VAT which they had overpaid by entering into reimbursement arrangements which ensured that the benefit of the repayments would be passed straight on to the investment trusts, which had borne the burden of the VAT and were not themselves in a position to make repayment claims under the statutory scheme.

It is also relevant to note that, at all material times, section 78(1) of VATA 1994 provided for payment of simple interest on amounts of output tax for which a person had accounted due to an error on the part of HMRC.

A further complication follows from subsection (2A) of section 80, which provides as follows:

as a result of a claim under this section by virtue of subsection (1) … above an amount falls to be credited to a person, and

after setting any sums against it under or by virtue of this Act, some or all of that amount remains to his credit,

The history of the repayment claims made by the three lead claimants is briefly as follows.

In early 2004, when the Claverhouse litigation began and was publicised, the Managers of F & C Trust and M & G Trust claimed refunds in respect of VAT accounting periods from 2001 to 2004. Claims were not made in relation to any earlier accounting periods because of the three year limitation period. No claim was made in respect of Kleinwort Trust, because it had gone into liquidation and stopped using investment management services in 1998.

Following the decision of the ECJ in Claverhouse in 2007, HMRC allowed the above claims and repaid the relevant net amounts with simple interest pursuant to section 78.

In the light of the decision of the House of Lords in Fleming v Revenue & Customs Commissioners [2008] UKHL 2, [2008] 1 WLR 195, HMRC accepted that the introduction of the three year limitation period on 4 December 1996, without any transitional arrangements, had been incompatible with EU law, with the consequence that it had to be disapplied in respect of rights which had accrued at that date. HMRC therefore invited further refund claims to be made before 1 April 2009 in respect of VAT paid in relation to accounting periods ending before 4 December 1996. The limitation period previously applicable to such claims had been six years, with a proviso extending the period in cases of mistake similar to section 32(1)(c) of the Limitation Act 1980, so the effect of this change was to allow further claims to be made going back at least to 1990.

Accordingly, on various dates before 1 April 2009 all three Managers made further claims to HMRC for refunds in respect of accounting periods ending before 4 December 1996. These claims were again allowed with simple interest and the appropriate repayments were made to the Managers, who in turn passed on the repayments to the investment trusts.

It follows from this procedural history that no claims could be made by the Managers in relation to periods ending after 4 December 1996 but before the dates in 2001 that were three years before the date of their first section 80 claims. This period was referred to in argument as “the dead period”. The dead period ends:

in the case of Kleinwort Trust, on 20 March 1998 (the day on which it was put into liquidation);

in the case of F & C Trust, on 6 April 2001; and

in the case of M & G Trust, on 1 April 2001.

The particular significance of the dead period is, of course, that the claimants have been unable to make any recovery from the Managers, through the medium of the Managers’ section 80 claims, in respect of the unlawful VAT which they paid during the dead period. The sums claimed by the lead claimants in relation to VAT incurred in the dead period are: in the case of Kleinwort Trust, £353,478; in the case of F & C Trust, £262,289; and in the case of M & G Trust, £1,790,850. The total of the dead period claims of the other six claimants is estimated to be £4,844,817.

The only claims now pursued in the present action by Kleinwort Trust and F & C Trust relate to payments of VAT during the dead period. Their claims were originally more extensive, but subsequent recoveries from their respective Managers in respect of periods other than the dead period have reimbursed them in full. M & G Trust, however, is in a different position. Apart from its dead period claim, it also makes claims in relation to unrecovered VAT (equivalent to its Manager’s input tax deductions) for periods going back to 1992, and for periods extending from the end of the dead period until 26 March 2002. The amount of its claim under this head is £121,594.

The total claims of the other six claimants for periods outside the dead period are estimated to amount to £2,939,183.

To summarise, therefore, the unrecovered VAT which the lead claimants seek to recover from HMRC in the present action comprises the aggregate of (a) the VAT charges paid by all of them during the dead period, less the input tax reclaimed by Kleinwort Trust; and (b) that proportion of the VAT charges paid by M & G Trust in respect of other periods from 1992 to 26 March 2002 which has not been refunded by HMRC, and which is equal to the input tax brought into account by M & G Trust’s Manager in relation to those periods.

On 15 July 2010 it was ordered by Chief Master Winegarten that all issues relating to liability should be tried separately from, and before, issues relating to quantum. The parties were directed to agree a list of liability issues. Initial attempts to agree a list were unsuccessful, but by the date of the trial the parties had been able to agree one. The first section of this document sets out common ground, most of which I have incorporated in the previous sections of this judgment. It is, however, worth recording the parties’ agreement that all the section 80 claims made by the Managers were made, determined and satisfied in accordance with the requirements of VATA 1994, and that in particular:

HMRC rejected any such claims to the extent that they were time-barred by reason of the three year “cap” in section 80(4), read with section 121(1) of the Finance Act 2008; and

in relation to all “uncapped” section 80 claims, HMRC repaid to the Managers the net difference after subtracting from the gross amount of output tax over-declared to HMRC the greater of (i) the sum deducted by the Manager as attributable input tax, and (ii) that part of the VAT charged by the Manager to the investment trust that the investment trust had recovered as input tax.

In respect of the dead period, and by reference to diagram 1, it is agreed that the dispute is whether the investment trusts have, or are required to be given, any claim at all in respect of the unrecovered VAT; and (if so) whether, in principle, F & C Trust and M & G Trust should be able to recover either:

the full amount of the unrecovered VAT, represented in diagram 1 by the £100 paid by the investment trust to the Manager and accounted for by the latter as output tax; or

the £75 paid by the Manager to HMRC, after deducting from the £100 as input tax the £25 paid by the Manager to its supplier.

It is common ground, by reference to diagram 2, that Kleinwort Trust’s unrecovered VAT is represented by the £41.60, being the difference between the £100 VAT paid and the £58.40 input tax recovered.

In respect of the other periods, it is agreed that the dispute concerns M & G Trust’s unrecovered VAT, which is represented by the £25.

Against that background, the agreed issues are formulated as follows:

Do the Investment Trusts (in principle) have mistake-based restitution claims/causes of action against [HMRC] for the Unrecovered VAT? This includes consideration of the following:

Were [HMRC] enriched as a result of the VAT Charges that were paid by the Investment Trusts to the Managers and accounted for by the Managers to [HMRC]?

If so, what is the extent of that enrichment?

Do [HMRC] remain enriched by the amounts of the Unrecovered VAT, taking into account the repayments made by [HMRC] under section 80 [VATA 1994]?

If [HMRC] were and remain so enriched, was and is that enrichment at the expense of the Investment Trusts?

If [HMRC] were and remain enriched at the expense of the Investment Trusts, was and is that enrichment unjust?

If the Investment Trusts have any mistake-based restitutionary claim/cause of action against [HMRC] as a matter of English law, is that cause of action excluded by statute, namely by section 80 [VATA 1994]?

If the Investment Trusts have no mistake-based restitutionary claim as a matter of English law (or they do but that claim is excluded by statute), does EU law require that the Investment Trusts should be able to claim the Unrecovered VAT from [HMRC] (by means of a directly effective right to reimbursement or otherwise)?

In the circumstances of this case, does the statutory scheme contained in [VATA 1994] (section 80, etc) provide a remedy that satisfies the principle of effectiveness as regards the protection of the Investment Trusts’ EU law rights (if any )?

If EU law requires that the United Kingdom should provide the Investment Trusts with a claim for reimbursement against [HMRC], is the statutory exclusion (if any) of such claims to be disapplied to the extent necessary to allow the Claimants the mistake-based restitutionary cause of action they assert?

Is there any need for one or more questions to be referred to [the ECJ] in respect of any of the EU law issues that arise in this case?”

In more general terms, it can be seen that the issues fall under three broad headings. The first main question is whether the claimants have a remedy under the English law of restitution. Counsel for the lead claimants made it clear in their written submissions, and Mr Rabinowitz QC repeated in his opening address, that this is their primary case, and that there is no need for them to rely on any EU rights. The second main question, assuming the first question to be answered in the claimants’ favour, is whether the prima facie common law claim against HMRC is excluded by the provisions of VATA 1994, and in particular by section 80(7) which provides as follows:

The third main question, which arises only if by reason of section 80(7), or for any other reason, the claimants do not have an English law claim for restitution, is whether they have an enforceable EU law right to reimbursement, and (if so) how English law should give effect to that right.

The question of EU law turns to a large extent on an examination of two decisions of the ECJ, one of them very recent. The first decision is that in Case C-35/05, Reemtsma Cigarettenfabriken GmbH v Ministero delle Finanze, [2007] ECR I-2425, [2008] STC 3448 (“Reemtsma”). The second decision is that in Case C-94/10, Danfoss A/S, Sauer Danfoss ApS v Skatteministeriet, (“Danfoss”), in which the ECJ gave judgment on 20 October 2011. At the date of the hearing before me in May, the Advocate General (Kokott) had given her opinion on 24 March 2011, but no official English translation of it was yet available, and the judgment of the Court was still awaited. In those circumstances, it was agreed that I should not deliver this judgment until after the ECJ had given its decision, and that there should be an opportunity for the parties to make further submissions in the light of the decision should they wish to do. In the event, the decision of the Court did not depart significantly from the Advocate General’s opinion, and the parties were content to make brief further submissions in writing, without a further oral hearing. One unavoidable consequence of this is that the gap between the conclusion of the oral argument before me on 24 May 2011 and the handing down of this judgment has been considerably longer than I would otherwise have wished.

It has now become conventional to consider the question whether English law recognises a right to restitution by reference to the four questions identified by Lord Steyn in Banque Financière de la Cité v Parc (Battersea) Limited [1999] 1 AC 221 at 227A-B, namely:

Has the defendant been benefited, in the sense of being enriched?

Was the enrichment at the claimant’s expense?

Was the enrichment unjust?

Are there any defences?

I have no quarrel with this basic conceptual structure. It needs to be remembered, however, that the four questions are no more than broad headings for ease of exposition. They should not be approached as if they had statutory force. There may also be a considerable degree of overlap between the first three questions. I would add that the development of a coherent set of defences, including in particular the defence of change of position, is important, because it should encourage the court not to adopt an unduly narrow or restrictive approach in considering whether the first three questions are satisfied.

At first sight, the answer to this question might seem obvious. The VAT which the investment trusts paid to the Managers was either paid by the Managers directly to HMRC as part of the net quarterly payments which they made (i.e. the £75 in the diagrams), or it reduced the amount of those quarterly payments by being set against and absorbing the related input tax which HMRC would otherwise have been obliged to pay or credit to the Managers (the £25 in the diagrams). Thus HMRC were enriched, as to £75 directly and as to £25 indirectly.

On behalf of HMRC, Mr Swift QC did not dispute that HMRC were enriched by the net payments of £75 which they actually received, but he argued that there was no enrichment in relation to the £25. His main submission here was that the £25 was properly payable to HMRC in any event, because it represented output tax properly payable by the Managers’ own suppliers, and it was so payable regardless of the Claverhouse mistake which led to the unlawful imposition of VAT on the onward supplies of investment management services made by the Managers to the investment trusts. Mr Swift emphasised that the right to deduct input tax is one of which a trader does not have to take advantage, so it would be wrong, he said, to characterise the deduction, when it is claimed by the Manager, as a set off of existing debts or obligations within the principle of Spargo’s case (In re Harmony and Montague Tin and Copper Mining Company (1873) LR 8 Ch App 407) as Mr Rabinowitz sought to do on behalf of the claimants. The principle in Spargo’s case was stated as follows by Mellish LJ at 414:

It is convenient at this point to set out the relevant provisions in section 25 of VATA 1994 relating to the payment of output tax and credit for input tax:

A taxable person shall –

in respect of supplies made by him …

Subject to the provisions of this section, he is entitled at the end of each prescribed accounting period to credit for so much of his input tax as is allowable under section 26, and then to deduct that amount from any output tax that is due from him.

If either no output tax is due at the end of the period, or the amount of the credit exceeds that of the output tax then, subject to subsections (4) and (5) below, the amount of the credit or, as the case may be, the amount of the excess shall be paid to the taxable person by the Commissioners; and an amount which is due under this subsection is referred to in this Act as a “VAT credit”.

These domestic provisions reflect and implement the corresponding principles which at the material times were principally contained in Articles 17(1) and (2), 18(2) and 22(5) of the Sixth Directive. The right to deduct input tax is conferred by Article 17; Article 18 contains rules governing the exercise of the right to deduct; and Article 22(5) says that:

In the light of this statutory scheme, I do not find it especially helpful to think in terms of a set off in accordance with Spargo’s case. The statutory scheme is governed by its own rules, both for claiming a deduction for input tax and for obtaining payment of the balance from HMRC if the deduction exceeds the output tax. But as to the existence of the obligation on HMRC to repay the excess, there can in my opinion be no doubt. Nor can there be any doubt about the obligation on HMRC to allow attributable input tax to be deducted from the output tax for which the trader has to account in his quarterly returns, subject only to the necessary claim being made for that purpose. In practice, the claim is made simply by including the relevant figures for input tax in the prescribed form of quarterly VAT return.

These provisions make it clear, in my judgment, that HMRC should indeed be regarded as enriched by the £25, because although the £25 was not paid to HMRC by the Managers, it was nevertheless used to give them a credit, at HMRC’s expense, for the input tax attributable to their investment management services which was wrongly thought to be deductible on the footing that the services were not exempt from VAT. HMRC therefore ended up out of pocket to the extent of the input tax. It is simply irrelevant to this analysis that the input tax is in principle the same as tax for which the Managers’ own suppliers were liable to account as output tax at the previous stage in the supply chain.

The other arguments which Mr Swift advanced under this heading seem to me to belong more comfortably under the next heading, which is whether the relevant enrichment (to the extent of the full £100) was at the expense of the claimants. I will therefore move on to that question.

As with the first question, common sense might be thought to suggest that the answer to this question was obvious. It is common ground that the unlawful VAT was paid in full by the claimants on the invoices submitted to them by the Managers. With the partial exception of the input tax reclaimed by Kleinwort Trust, the full economic burden of the unlawful tax (the £100) was therefore borne by the claimants. This entirely accords with the fundamental nature of VAT as a tax on the supply of goods or services, the burden of which is borne by the final consumer. Equally clearly, it is HMRC who ultimately benefited from the imposition of the tax: that is the very purpose of taxation, to transfer money from the taxpayer to the Crown. How, then, can it seriously be disputed that the enrichment of HMRC was at the expense of the claimants, who actually paid the tax, and were unable to pass it on to anybody else?

In support of this common sense view of the matter, Mr Rabinowitz argued that the role of the Managers was merely that of tax collectors, who were in effect acting on behalf of HMRC, and that they could helpfully be visualised as conduits through whom the VAT flowed from the claimants to HMRC. He referred in this connection to what Neuberger J (as he then was) said in University of Sussex v Customs and Excise Commissioners [2001] STC 1495 at [48]:

In the passage which I have just quoted Neuberger J referred to what he called “the classic statement” by the ECJ in Case C-317/94, Elida Gibbs Ltd v Customs and Excise Commissioners [1996] ECR I-5339, [1997] QB 499, at paragraphs 19 and 22 of the judgment of the Court. In view of the importance attached to this passage by Mr Rabinowitz, and because it provides a very clear statement of the basic working of the VAT system, I will quote paragraphs 18 to 24 of the judgment, which come under the heading “General Considerations”:

In my judgment this passage makes good the contention, at a fairly high level of generality, that for the purposes of VAT the final consumer is properly to be regarded as the taxpayer, and that the role of the intermediate taxable persons in the chain of supply is to collect the tax and account for it to the tax authorities. On the other hand, I am satisfied that there are dangers in pressing this point too far, because the final consumer is not in any ordinary sense a taxpayer vis-à-vis HMRC. It is the supply of goods or services to the final consumer which triggers the liability to pay and account for VAT on the supply, and the supplier is liable accordingly, whether or not he passes on the tax charge to his customer. Conversely, the customer is under no liability to HMRC for the VAT, even if the supplier fails to pay it, and the customer’s obligation to pay the VAT to the supplier rests only in contract. If, for whatever reason, the supplier fails to charge VAT to the customer, he will have no remedy against the customer unless the customer is contractually bound to pay it. Essentially for these reasons, I am not happy with the claimants’ conduit metaphor, which suggests that the tax flows from the final consumer to HMRC, and that the supplier is in some sense acting as the customer’s agent in paying it to HMRC. It needs to be clearly understood that the liability of the supplier to pay and account for output tax is a primary one, and in legal terms the taxable person is the supplier, not the customer. Nor is there any relationship of agency, or anything resembling agency, between them.

Building on these distinctions, Mr Swift correctly submitted that there was no direct payment of the unlawful VAT by the investment trusts to HMRC, with the consequence that HMRC’s enrichment cannot be said to have been at the direct expense of the claimants. The significance of this point lies in the fact that in most cases where the courts have held an enrichment to be at the claimant’s expense, there has been a direct benefit (or the direct conferral of a benefit in some other form) by the claimant to the defendant. Accordingly, some (but by no means all) academic writers have framed the requirement in terms of a general rule, while acknowledging that there are a number of apparent exceptions to it whereby the claimant can obtain a restitutionary remedy against the indirect recipient of a benefit.

So, for example, Professor Graham Virgo, in The Principles of the Law of Restitution, 2^nd edition (2005), says that the requirement of direct benefit, which he calls “the privity principle”, is “fundamental to our understanding of the action to reverse the defendant’s unjust enrichment” (page 106), and submits that careful analysis of the so-called exceptions “shows that there are few true limits on the operation of the principle” (page 107). In support of the privity principle, he cites the decision of Goulding J in Re Byfield [1982] 1 Ch 267 and the observation of Morritt LJ, in Kleinwort Benson Ltd v Birmingham City Council [1997] QB 380 at 400F, that the words “at the expense of the plaintiff”, which he traced back to the American Law Institute’s Restatement of the Law of Restitution (1937), “do no more than point to the requirement that the immediate source of the unjust enrichment must be the plaintiff”.

On the other hand, Professor Birks took the view that a causal link between the claimant’s payment and the defendant’s enrichment was in principle sufficient, and that a remote recipient will anyway have the protection of all the usual defences, such as bona fide purchase for value and change of position. See generally the second edition of his work on Unjust Enrichment, Oxford (2005), chapter 4, at pp 89-98, and his paper “At the expense of the claimant”: direct and indirect enrichment in English law in Unjustified Enrichment: Key Issues in Comparative Perspective, edited by David Johnston and Reinhard Zimmermann, Cambridge (2002), at pp 493 - 525.

Occupying a somewhat intermediate position is Professor Burrows, who favours a general rule whereby only the direct provider of a benefit is entitled to restitution, but recognises numerous exceptions to it, and is agnostic on the question whether the list of exceptions should be extended: see The Law of Restitution, 3^rd edition (2011), at pp 69-85.

Faced with this wide divergence of academic opinion, it is natural to look for such guidance as one can find in the authorities; but the guidance is limited and the illumination sporadic, because none of the cases to which I was referred appear to me at all close to the factual and legal position which I now have to consider. The nearest that one finds to a full discussion of the subject is in the judgment of Morritt LJ in Kleinwort Benson Ltd v Birmingham City Council, loc.cit, at 395-402, with which Saville LJ agreed (see 394E). The issue in that case, however, was whether English law should recognise a non-statutory defence of passing on, in circumstances where a bank which had entered into a void swap transaction with a local authority had successfully hedged its position by a valid back-to-back swap with a different counterparty. The Court of Appeal held that no such defence should be recognised for a variety of reasons, but it is fair to say that each member of the Court adopted a relatively strict approach to what “at the expense of” requires.

Evans LJ, delivering the leading judgment with which Saville LJ also agreed, accepted the submission of counsel for the bank that the central issue was whether the phrase “at the expense of” had to be interpreted by reference “to the payer/payee relationship alone”, as distinct from other parts of the overall transaction. He then said at 393A:

Saville LJ said at 395A:

I have already quoted Morritt LJ’s view that the phrase points to the requirement that “the immediate source” of the unjust enrichment must be the claimant, but to put his observation in context I should quote the whole of the paragraph in his judgment beginning at 400F:

I do not read these statements as part of the ratio of the decision, because no issue of indirect enrichment was before the Court. Nevertheless, the strict approach to the requirement adopted by each member of the Court is clearly of strong persuasive force. It is also salutary to be reminded of the need to avoid importing contractual notions such as compensation and mitigation into the law of restitution, and of the practical problems which could arise if recovery from anybody other than the immediate payee were to be permitted.

Perhaps recognising the force of some of these points, counsel for the claimants placed at the forefront of their written and oral submissions on this part of the case a more recent, unreported, decision of the Court of Appeal where a more liberal approach to “at the expense of” appears to have been adopted, albeit obiter. The case is Filby v Mortgage Express (No. 2) Ltd [2004] EWCA Civ 759 (“Filby”). The facts in Filby were briefly as follows. Mr Filby fraudulently induced Mortgage Express to make a loan to himself and his wife, to be secured by a first legal charge on their matrimonial home, of which they were co-owners. Mrs Filby had no knowledge of the transaction, and unknown to Mortgage Express her signature had been forged on the relevant documents by her husband, with the result that the legal charge was ineffective against her. The solicitors acting for Mortgage Express and (as they thought) for both of the Filbys then paid off an existing charge on the property, as required by Mortgage Express, and on Mr Filby’s directions transferred the balance of the money from their client account to Midland Bank, in partial discharge of the debt owing on the couple’s unsecured joint loan account. When Mr Filby failed to repay Mortgage Express, and the facts came to light, Mortgage Express brought proceedings in which it claimed to be entitled, as against Mrs Filby, to be subrogated to the charge, and also to Midland Bank’s personal rights in respect of the unsecured loan account. Mrs Filby conceded the right of Mortgage Express to be subrogated to the former charge on the property, but disputed its entitlement to be subrogated to Midland Bank’s personal rights.

Mrs Filby conceded that she had been enriched to the extent that the Midland Bank loan was discharged through the use of money advanced by Mortgage Express. She argued, however, that this enrichment had been at the expense, not of Mortgage Express, but of Mr Filby. In particular, she argued that, once the former charge had been paid off, the balance remaining in the solicitors’ client account had been beneficially owned by Mr Filby, because he had received it under a contract which, although voidable by reason of his fraud, Mortgage Express had never rescinded. The money had therefore become his, and Mrs Filby’s enrichment was at his expense.

The Court of Appeal rejected this argument. The primary ground of decision was that Mortgage Express had throughout remained the beneficial owner of the money held in the solicitors’ account, with the consequence that Midland Bank had been paid with money belonging to Mortgage Express; the money was traceable, and on established principles the remedy of equitable subrogation was available (see [21] to [31]). This conclusion was sufficient to dispose of the appeal, but May LJ (who delivered the leading judgment, with which Hooper and Kennedy LJJ agreed) went on to consider the case more broadly, in the light of recent authority which had explained the nature of non-contractual subrogation as being part of the law of restitution (including, in particular, the speeches of Lord Steyn and Lord Hoffmann in Banque Financière de la Cité v Parc (Battersea) Ltd [1999] 1 AC 221). Addressing the four questions posed by Lord Steyn, May LJ said this at [46]:

As May LJ had explained in [41], in Banque Financière the first and second defendants were companies within the same group, and Mr Herzig was general manager of the group’s holding company. In 1990, he had negotiated a refinancing loan with the plaintiff bank, but in order to circumvent Swiss banking regulations on disclosure, the transaction was effected by the plaintiffs making their loan to Mr Herzig, who took steps to ensure that an equivalent sum was paid directly by the plaintiffs to the first defendants. The first defendants provided no security for the loan, but arrangements were put in place whereby the plaintiffs took an assignment from Mr Herzig of a promissory note given to him by the first defendants, and a letter of request signed by Mr Herzig confirmed that no group company would demand any repayment of loans to the first defendant until there had been repayment in full of the plaintiffs’ loan to Mr Herzig.

Mortgage Express had made the loan in Filby in the mistaken belief that it was to be secured by a first legal charge, so the enrichment of Mrs Filby at the expense of Mortgage Express satisfied the test of being unjust. May LJ then continued at [50]:

It is clear, in my judgment, that the “reality” which May LJ was invoking in paragraphs [50] and [62] of his judgment in Filby was not confined to a strictly legal reality, but could in appropriate circumstances include a broader underlying commercial or economic reality. Quite where the boundaries of this more generous version of reality should be drawn is unclear, but Mr Rabinowitz submitted that the claims in the present case fall squarely within the principle because it would be pure formalism to regard the interposition of the Managers as making any difference, and the underlying reality is that the VAT was paid by the claimants and accounted for to HMRC.

Mr Rabinowitz also argued that support for this broader approach could be found in an earlier group of subrogation cases in which a similar result was reached. Those cases are: Bannatyne v D & C MacIver [1906] 1 KB 103 (CA); Butler v Rice [1910] 2 Ch 277; and B Liggett (Liverpool) Ltd v Barclays Bank Ltd [1928] 1 KB 48, as explained in Re Cleadon Trust Ltd [1939] Ch 286 (CA). These are the main authorities on which Professor Birks relied in support of his thesis that a restitutionary remedy can be sought against a second or subsequent recipient: see in particular his Unjust Enrichment, 2^nd edition, at pp.96-98. It appears that Professor Birks was unaware of Filby, but I agree with Mr Rabinowitz that, if he had been aware of it, he would have regarded the case as providing further support for his views.

I must now draw the threads together, and state my conclusions on this difficult question. In the first place, I agree with Mr Rabinowitz that there can be no room for a bright line requirement which would automatically rule out all restitutionary claims against indirect recipients. Indeed, Mr Swift accepted as much in his closing submissions. In my judgment the infinite variety of possible factual circumstances is such that an absolute rule of this nature would be unsustainable. Secondly, however, the limited guidance to be found in the English authorities, and above all the clear statements by all three members of the Court of Appeal in Kleinwort Benson Ltd v Birmingham City Council, suggest to me that it is preferable to think in terms of a general requirement of direct enrichment, to which there are limited exceptions, rather than to adopt Professor Birks’ view that the rule and the exceptions should in effect swap places (see “At the expense of the claimant”: direct and indirect enrichment in English law, loc.cit., at page 494). In my judgment the obiter dicta of May LJ in Filby, and the line of subrogation cases relied on by Professor Birks, provide too flimsy a foundation for such a reformulation, whatever its theoretical attractions may be, quite apart from the difficulty in framing the general rule in acceptable terms if it is not confined to direct recipients.

The real question, therefore, is whether claims of the present type should be treated as exceptions to the general rule. So far as I am aware, no exhaustive list of criteria for the recognition of exceptions has yet been put forward by proponents of the general rule, and I think it is safe to assume that the usual preference of English law for development in a pragmatic and step by step fashion will prevail. Nevertheless, in the search for principle a number of relevant considerations have been identified, including (in no particular order):

the need for a close causal connection between the payment by the claimant and the enrichment of the indirect recipient;

the need to avoid any risk of double recovery, often coupled with a suggested requirement that the claimant should first be required to exhaust his remedies against the direct recipient;

the need to avoid any conflict with contracts between the parties, and in particular to prevent “leapfrogging” over an immediate contractual counterparty in a way which would undermine the contract; and

the need to confine the remedy to disgorgement of undue enrichment, and not to allow it to encroach into the territory of compensation or damages.

Many of these considerations present no difficulty in the present case. There is no risk of double recovery, because the claimants have in effect exhausted their remedies against the Managers. The Managers have obtained the maximum repayments from HMRC available under the domestic statutory scheme, and have passed on those repayments in full to the claimants. I am also satisfied that no claim for breach of contract could lie against the Managers at the suit of the claimants. Although the possibility of such claims was mooted at various stages in the oral argument, I agree with Mr Rabinowitz that there has been no breach by the Managers of their contracts with the investment trusts, and that the terms in the investment management agreements which required payment of VAT “if applicable”, or words to similar effect, did not impose any warranty or obligation to ensure that the VAT charged was in fact lawfully due. The only remedy of the investment trusts against the Managers in respect of the overpaid tax was therefore a restitutionary one, based on mistake. If, however, any such claim were now to be brought, the Managers would have a cast iron defence of change of position, having accounted to HMRC for the entirety of the tax as output tax, and having retained no benefit from it.

Similarly, it is not suggested that the present claims against HMRC would conflict in any way with the contractual arrangements between the investment trusts and the Managers; and the claims are limited to disgorgement of the unlawful tax by which HMRC have been enriched.

The requirement of causation, however, is more problematic. As I have already pointed out, there is no strict causal connection between the payment of the VAT element of the invoices submitted by the Managers to the claimants, and the payment of the VAT by the Managers to HMRC. The Managers were liable to account for the VAT to HMRC once they had supplied the relevant services, and the obligation of the claimants to pay the Managers was purely contractual: see paragraph 50 above. It cannot even be said that the VAT was paid or accounted for to HMRC out of the money paid by the claimants to the Managers, or that the VAT would not have been paid but for the payments by the claimants to the Managers.

On the other hand, the scheme of VAT, as explained by the ECJ in Elida Gibbs and echoed by Neuberger J in the Sussex University case, is to impose the burden of the tax on the final consumer, and to make the suppliers of the goods or services the collectors of the tax on behalf of the tax authorities. In other words, VAT is a tax on the consumer, collected by the supplier, and paid or accounted for to HMRC. Viewed in this way, the nexus between the consumer and HMRC could hardly be closer or stronger, and in economic terms the person at whose expense unlawful VAT is paid to HMRC is indubitably the consumer. I remind myself at this point that “at the expense of” is not a statutory requirement, and (as the subrogation cases show) it can be satisfied by reference to the underlying commercial reality of a transaction. To recognise that the test is satisfied in the present case would not, as Mr Swift submitted, be to dismiss the structure of the VAT legislation as mere formalism, but rather to give due weight to the economic reality which explains and underpins that structure.

In the end, I come back to where I started. In my judgment there is no convincing answer to the common sense proposition that the enrichment of HMRC was indeed at the expense of the claimants, and I would therefore so hold.

The next question is whether the enrichment of HMRC at the expense of the claimants was unjust. As Professor Burrows explains in The Law of Restitution, 3^rd edition (2011), chapter 5, the traditional common law approach to this question requires the claimant to establish the existence of an “unjust factor” which caused the payment that the claimant seeks to recover. English law has not, at least so far, moved to a single unifying concept of “absence of basis”, although cautious encouragement for such a development may be found in the observations of Lord Walker of Gestingthorpe in Deutsche Morgan Grenfell Group Plc v IRC [2006] UKHL 49, [2007] 1 AC 558, (“DMG”) at [154] to [158]. In the present case, the relevant unjust factor is mistake. The claimants’ simple contention is that, but for their mistaken belief that VAT was lawfully due on the investment management services rendered to them by the Managers, they would not have paid the VAT to the Managers, and HMRC would not have been indirectly enriched by it.

The decision of the House of Lords in DMG established that a taxpayer who has wrongly paid tax under a mistake of law is entitled to a restitutionary remedy against the Crown, and that the taxpayer may choose to pursue this remedy (with the favourable limitation period afforded by section 32(1)(c) of the Limitation Act 1980), even if there is a concurrent restitutionary cause of action open to him (in the present context, the right to recover unlawfully levied tax pursuant to the Woolwich principle) which has a shorter limitation period.

The mistake in DMG concerned the availability of a group income election when dividends were paid by the UK resident taxpayer company to its German resident parent. Under the UK legislation then in force, such an election could be made only when the parent company was also resident in the UK; but in the Hoechst case (Joined Cases C-397 and 410/98, Metallgesellschaft Ltd v IRC and Hoechst AG v IRC, [2001] ECR I-1727, [2001] Ch 620) the ECJ had held that this restriction infringed the freedom of establishment of the parent company under EU law, and that UK subsidiaries which had paid advance corporation tax (“ACT”) on the dividends were in principle entitled to restitution of the overpaid tax (or, where the tax had subsequently been set off against mainstream corporation tax, to compensation for the premature payment of the ACT).

There was a difference of opinion between the members of the Appellate Committee in DMG about the precise nature of the mistake which had been made, and which grounded the taxpayers’ right to restitution. Lord Hope of Craighead pointed out at [58] that, if the correct approach was to look only at the system laid down by the UK legislation, there was no mistake, because no group income election had been made, with the result that ACT was payable. He continued, however:

At [62] Lord Hope expressed his agreement with Park J, who had held at first instance that the relevant mistake was as to the availability of group income elections, and that the payment of ACT was a secondary consequence of this mistake. He then said:

By contrast, Lord Hoffmann regarded Park J’s view of the nature of the mistake as “rather sophisticated”, and said he agreed with the Court of Appeal “that the mistake was about whether DMG was liable for ACT”, and that the election provisions were purely machinery which the taxpayer would undoubtedly have used to enforce its right to exemption from liability: see [32]. Lord Walker also thought that Park J had been “rather over-analytical in his approach”, and at [143] expressed his agreement with what Neuberger J had said in Nurdin & Peacock Plc v D B Ramsden & Co Ltd [1999] 1 WLR 1249, 1272:

Lord Brown of Eaton-under-Heywood agreed with the speeches of Lord Hoffmann, Lord Hope and Lord Walker, without adding anything on the analysis of the relevant mistake. Lord Scott of Foscote dissented, taking the view that the ACT had not been paid under any mistake of law, because the taxpayer (in the absence of a group income election) was under a legal obligation to make the payments: see [89].

In their amended defence in the present case, HMRC make some important admissions. They admit:

that in so far as any VAT may have been paid by an investment trust to a Manager in circumstances where the VAT was not properly charged and due as a matter of law, the VAT was paid by mistake;

that if the investment trusts have any claim or cause of action against HMRC for relief from the consequences of a mistake, they are entitled to rely on section 32(1)(c) of the Limitation Act 1980 in respect of such claim or cause of action; and

that the investment trusts did not discover, and could not with reasonable diligence have discovered, their mistakes until the ECJ delivered its judgment in Claverhouse on 28 June 2007.

The terms of the first of these admissions leaves room for an argument that, despite the unlawfulness of the VAT as established in Claverhouse, it was nevertheless still due as a matter of contract between the investment trusts and the Managers, with the consequence that they were obliged to pay it in any event, and the payments were not caused by the mistake.

An argument to this effect was developed by counsel for HMRC at some length in their written submissions, and a panoply of counter arguments against this and other possible objections was deployed by counsel for the claimants in their opening written and oral submissions. In the event, however, the objections all but evaporated in the course of the hearing, and when he came to address the court Mr Swift expressly accepted on behalf of HMRC that the contracts between the claimants and the Managers did not, on their true construction, oblige the claimants to pay VAT to the Managers if it was not legally due. In other words, he accepted that, on the hypothesis that the unlawfulness of the VAT charges had been known to the investment trusts and the Managers, the investment trusts would not have been under any contractual obligation to pay the charges, had the Managers in fact sought to impose them.

For good measure, unchallenged evidence was also given on behalf of two of the claimants (F & C Trust and M & G Trust) to the effect that, in such a hypothetical situation, the relevant Managers would not have sought to increase the price of their investment management services, and they would have been content to absorb the costs of their now irrecoverable input tax as a business expense. I see no reason to suppose that the position would have been any different in respect of Kleinwort Trust, or indeed any of the other investment trusts. Mr Richardson made it clear in his statement on behalf of M & G Trust that he was aware of opinion throughout the industry, through his active participation in a number of groups such as the Investment Management Association, and he was not aware of any Manager which chose to charge VAT in the aftermath of the ECJ decision in Claverhouse, even though HMRC’s Business Brief 65/07 gave them the option to do so during the period before the UK legislation was amended. Mr Richardson also said that he would have reacted in exactly the same way before the commencement of the Claverhouse litigation in 2004, and expressed the belief that this would have been true for the majority of industry participants.

Since there is no longer any live issue about the construction of the contracts between the claimants and the Managers, it is unnecessary for me to set out the relevant provisions or to discuss the rival arguments. It is enough to say that I agree with the conclusion that, as a matter of construction of the agreements, there was no contractual obligation on the claimants to pay the VAT charges to the Managers, because the VAT was not lawfully due on the services supplied by the Managers. This appears to me to be the natural construction of an obligation to pay VAT “if due”, or words to similar effect, and in my view it would require very clear words to impose a liability to pay VAT if it was not in fact due. Thus the position was that the claimants paid the VAT charged on the invoices in the mistaken belief that it was legally due, but when they were in fact under no obligation to do so.

In those circumstances, there can in my judgment be no doubt, and Mr Swift ended up accepting, that the VAT paid by the investment trusts to the Managers was paid under a mistake, and there was a sufficient causal connection (on a “but for” basis) between the mistake and the payments to ground a cause of action in restitution as between the investment trusts and the Managers. The position is, indeed, clearer than in DMG, because in that case there was nothing unlawful about the basic charge to ACT in the absence of a group income election, and from one point of view (as Lord Hope recognised, and Lord Scott held) it could be said that the dividends in question were not paid under a mistake at all, because they were paid pursuant to a subsisting legal obligation. In the present case, by contrast, once the ECJ had held in Claverhouse that open and closed-end investment trusts had to be treated alike, it was clear that the VAT exemption for the former had to be extended to the latter, and that the VAT charged by the Managers was therefore unlawful.

Having yielded all this ground, Mr Swift took his final stand on the proposition that there was no sufficient causal link between the mistake which prompted the payments by the claimants to the Managers, and the payments which the Managers made to HMRC. This argument, however, was in substance the same as the argument that HMRC were not enriched at the expense of the claimants, which I have already rejected. In the result, I am satisfied that the same mistake about the lawfulness of the VAT charges caused both the payments by the investment trusts to the Managers and the corresponding enrichment of HMRC. I am also satisfied that, if the true legal position had been appreciated, no such payments would have been made by the investment trusts, and the enrichment of HMRC would not have occurred. It follows, in my judgment, that all the basic ingredients of a restitutionary cause of action by the claimants against HMRC are made out, and that subject to any available defences the claims should succeed as a matter of English domestic law.

The only defence relied on by HMRC, on the assumption (as I have held) that the necessary ingredients of a restitutionary claim under English law are made out, is the statutory exclusion contained in section 80(7) of VATA 1994.

I have already referred to many of the relevant provisions in section 80, but it is convenient at this point to set them out again in the form in which they were in force when the claim form was issued on 28 August 2009:

Where a person –

has accounted to the Commissioners for VAT for a prescribed accounting period (whenever ended), and

in doing so, has brought into account as output tax an amount that was not output tax due,

have assessed a person to VAT for a prescribed accounting period (whenever ended), and

in doing so, have brought into account as output tax an amount that was not output tax due,

an amount that was not output tax due being brought into account as output tax, or

The Commissioners shall only be liable to credit or repay an amount under this section on a claim being made for the purpose.

as a result of a claim under this section by virtue of subsection (1) or (1A) above an amount falls to be credited to a person, and

after setting any sums against it under or by virtue of this Act, some or all of that amount remains to his credit,

The Commissioners shall not be liable on a claim under this section –

to credit an amount to a person under subsection (1) or (1A) above, or

to repay an amount to a person under subsection (1B) above,

It is common ground that, for taxpayers who have themselves accounted to HMRC for output tax that was not due (“undue VAT”), section 80 provides a code for the recovery of the undue VAT which is both exhaustive and excludes other remedies (such as a common law claim for restitution). I discussed this question, in the context of claims to recover compound interest at common law on undue VAT, in F J Chalke Ltd v Revenue & Customs Commissioners [2009] EWHC 952 (Ch), [2009] STC 2027, at [57] and following, concluding at [68]:

It is also common ground that the investment trusts could never have made a claim under section 80 in respect of the VAT which they paid to the Managers, because it was the Managers, and not the investment trusts, who paid or accounted for the tax to HMRC. Accordingly, it was only the Managers who could bring themselves within the terms of subsection (1). The critical question, therefore, is whether (as the claimants submit) the exclusion of other remedies in subsection (7) applies only to taxpayers who would in principle be able to claim a refund of undue VAT under subsection (1), or whether (as HMRC submit) the exclusion in subsection (7) is potentially wider in scope, and applies to the facts of the present case. It is agreed that the question is one of statutory construction, to be resolved in the usual way and without preconceptions about what the answer ought to be.

I will first approach the question without reference to the statutory history of section 80.

The claimants submit that the whole structure of section 80 shows that it is concerned, and concerned only, with the crediting or repayment of undue VAT to a taxpayer who has paid or accounted for the tax in the first place. The claimants place particular emphasis on the use of the verb “repay” in subsection (7), which they say must mean to pay back to the original payer, both as a matter of ordinary English usage and because the word is used throughout the section in contexts where it can only refer to the original payer: see subsections (1B), (2), (2A) and (4). I think it is also worthy of note, although Mr Rabinowitz did not mention it, that in subsection (2A) the draftsman used both of the verbs “pay” and “repay” in relation to HMRC, presumably intending the former to cover the position where the taxpayer was a repayment trader who had not made any actual payment to HMRC.

Mr Rabinowitz also sought support for his submissions in certain observations of Neuberger J in the University of Sussex case. The legal issues in that case were complex and difficult, but the basic facts were relatively simple. The University had registered for VAT in 1973. Its main supplies were of education, and were exempt, but it also made some taxable supplies. Between 1973 and 1996 the University chose not to deduct certain amounts of input tax in its VAT returns, and paid VAT accordingly. In November 1996 the University submitted a claim for repayment of the input tax which it had failed to claim over the period from April 1973 to July 1996, relying on regulation 29 of the Value Added Tax Regulations 1995, which as it then stood was not subject to any express time limit, and on the right to deduct input tax provided by Articles 17(1) and (2) and 18 of the Sixth Directive. HMRC argued that the University’s claim fell within section 80, and was therefore subject to the three year cap. This argument was rejected by the judge, for a number of reasons, and he concluded that the claim was validly made under regulation 29 and should have been accepted by HMRC. For present purposes, most of the details do not matter, and the important point is that the judge held that section 80, in its then form, did not apply to claims for the recovery of input tax. The passage on which Mr Rabinowitz particularly relied is in paragraph [57] of the judgment:

To similar effect, Neuberger J began his conclusions with this statement at [73]:

The reasoning of Neuberger J in the University of Sussex case is authority for the proposition that, if a claim to recover undue VAT could be brought under section 80, the time limit in subsection (4), and the exclusion of other remedies in subsection (7), will apply to the claim, even if some alternative way of framing the claim might appear to be available. However, it is not authority for the converse proposition that section 80(7), on its true construction, cannot apply to claims which, for whatever reason, cannot be brought under subsection (1). In other words, it does not decide that the scope of subsection (7) is in all respects co-extensive with that of subsection (1). The fundamental problem for HMRC was that, even on a broad interpretation of subsection (7), it could not apply to the University’s claims to recover input tax, because as it then stood the subsection only relieved HMRC from liability “to repay an amount paid to them by way of VAT”. On the judge’s analysis of the statutory scheme, the simple point was that failure to claim a credit for input tax did not result in an overpayment of VAT, but merely left the input tax available to be claimed on a subsequent occasion against a later liability for output tax.

On behalf of HMRC, Mr Swift submitted that the purpose of subsection (7) is to exclude liability for claims of any kind that are, in substance, claims to recover undue VAT, unless the claim is brought under subsection (1). There is no reason, he argued, to assume that the exclusion in subsection (7) was directed only to persons who could bring section 80(1) claims, particularly bearing in mind that the prohibition in subsection (7) is a prohibition on liability. If one asks what it is in respect of which there is to be no liability, the answer is an amount of money accounted for or paid to HMRC by way of VAT. The exclusion therefore attaches to any claim that seeks to extract that sum of money from them.

As to the use of the word “repay”, Mr Swift submitted that from the point of view of HMRC there is a repayment of the VAT, even if it is made to somebody other than the original payer, because HMRC would be paying back money which they had received in the first place. Furthermore, since a restitutionary claim by the taxable person would admittedly be excluded, it would be very strange if the end consumer could be in a better position. This is particularly so, because one only reaches the present stage in the argument if it has been held in the claimants’ favour that HMRC were enriched at their expense, and that they were in economic substance the taxpayers. It would be inconsistent, said Mr Swift, for the claimants to take advantage of the economic substance of the matter in order to found their restitutionary claim, and then to disown it when seeking to escape from the clutches of subsection (7).

By way of authority, Mr Swift sought to draw support from the decision of the Court of Appeal, affirming the Chancellor of the High Court, in Monro v Revenue & Customs Commissioners [2008] EWCA Civ 306, [2009] Ch 69 (“Monro”). The issue in Monro was whether the taxpayer could maintain a restitutionary claim for capital gains tax overpaid in consequence of a mistake of law, in circumstances where it was not open to him to make an “error or mistake” claim under section 33 of the Taxes Management Act 1970 because the tax had been paid in accordance with the prevailing practice at the time (a defence expressly provided by subsection (2A) of section 33). Section 33 contains no provision equivalent to section 80(7), so the argument for HMRC had to be that the common law restitutionary remedy which the taxpayer wished to claim was, by necessary implication, excluded. This argument was accepted, both by the Chancellor and by the Court of Appeal, with the result that the claim failed.

The leading judgment in the Court of Appeal was given by Arden LJ. In paragraph 15 of her judgment, she recorded the argument for the taxpayer, Mr Monro, that subsection (2A) of section 33 applied only to exclude claims “under this section”, and that it did not deal comprehensively with all claims for recovery of tax paid under a mistake, but only those which were brought under section 33. This argument, submits Mr Swift, was similar to the argument advanced by the claimants in the present case, but it was rejected in Monro because, in essence, the statutory regime contained features (including the settled practice defence) which were inconsistent with the continued existence of any wider common law remedy. As Sir Andrew Morritt C pithily put it in paragraph [31] of his judgment at first instance:

More generally, Mr Swift relies on Monro for the purposive approach taken to the construction of a statutory scheme comparable with that in section 80 of VATA 1994, and for the diagnostic test of whether the right asserted would be inconsistent with the purpose of the statutory scheme.

I begin my discussion of these submissions with the obvious point that this is not a case, like Monro, where any exclusion of a concurrent common law remedy has to be found by a process of necessary implication. Instead, there is an express exclusion in subsection (7), and the only issue is how far it extends.

My next point is that there is nothing in the wording of subsection (7) which expressly makes its ambit co-extensive with that of subsection (1). On the contrary, subsection (7) provides in apparently unqualified terms that, except as provided by section 80, HMRC shall not be liable “to credit or repay any amount accounted for or paid to them by way of VAT that was not VAT due to them”. It is true that these words are most naturally and easily read as referring to the taxable person who paid or accounted for the overpaid VAT in the first place; and this impression is strengthened by the references to crediting and repayment of undue tax earlier in the section. Nevertheless, I consider it at least possible to read the words “repay any amount” as including repayments by HMRC to somebody other than the taxable person himself. The first meaning of “repay” in the Oxford English Dictionary, second edition, is: “To pay back (money or its equivalent); to refund, return (a sum or amount owed); to give money or goods in discharge of (a debt or loan).” In my judgment, a payment made by HMRC to the claimants, in response to a common law restitutionary claim, of an amount equal to the undue tax received from or accounted for by the Managers, could be described without any abuse of language as a “repayment” by HMRC of the undue tax by which they had been directly or indirectly enriched. I agree with the submission of Mr Swift that enough force can be given to the notion of payment back inherent in the prefix “re” by looking at the matter from HMRC’s point of view, without any need to insist that the recipient of the repayment should be the same person as the original payer. I also see no insuperable difficulty in treating the concept of repayment as extending to the full amount of the enrichment, even though the amount actually paid to HMRC was the £75, not the full £100. I therefore conclude that a construction of subsection (7) which would include within its ambit claims by end consumers such as the claimants in the present case is linguistically an available one, even if it is not the most natural way of reading the words.

At this point, purposive considerations appear to me to be decisive. The evident purpose of section 80, so far as taxable persons are concerned, is to provide exhaustive and exclusive machinery for the recovery of undue VAT, subject to a relatively strict time limit for the making of claims. It is thus common ground that the Managers could not make restitutionary claims against HMRC in respect of VAT overpaid by them during the dead period, although in the absence of section 80 there would be nothing to prevent them from advancing such claims, with the benefit of the usual six year limitation period and mistake-based extensions to it pursuant to section 32(1)(c) of the Limitation Act 1980. Given that Parliament has decided to enact this limited regime in relation to the taxable persons by whom the undue VAT was paid or accounted for to HMRC, it seems to me inconceivable that Parliament could have intended a more generous regime to be available to the end customers by whom the economic burden of the unlawful tax was actually borne. It would make no sense to limit recovery by the tax collector, but to expose the Exchequer at the same time to far more extensive claims by the “real” taxpayer. Furthermore, it could not plausibly be suggested that the position of end customers was somehow overlooked, because the section contains a defence of passing on, and (as I have already explained) regulations make elaborate provision for the benefit of repayments to suppliers to be passed on to their customers. It would be wholly inconsistent with this limited and carefully regulated scheme if claims by the end customers fell outside its scope.

In the light of these purposive considerations, I am firmly of the view that subsection (7) should be construed as extending to claims of the present type, with the consequence that, as a matter of English law, the claims must fail. Since I consider the position to be clear, and since an amended statute should be construed in the first instance without regard to its statutory antecedents, it is unnecessary, and arguably inappropriate, to investigate the statutory history of section 80 for any light that it may throw on the problem. However, since both sides made some submissions on the subject, I will briefly refer to it.

Section 80 was first enacted as section 24 of the Finance Act 1989. Omitting commencement and transitional provisions, it provided as follows:

It will be seen that the basic structure, and much of the wording, of the section have remained essentially unchanged since its introduction. The main changes have been twofold. First, following the decision in the University of Sussex case, which exposed the fact that the section did not apply to claims arising from unclaimed input tax, the section was restructured so that it reflected the way in which VAT has to be accounted for by traders, and made the primary remedy a credit for overstated VAT rather than a repayment of overpaid VAT. These changes were introduced by section 3 of the Finance (No.2) Act 2005, and had effect whenever the event occurred in respect of which the claim was made. Secondly, the original limitation provisions in subsections (4) and (5), which mirrored those in the Limitation Act 1980 and included an extension of time in cases of mistake, were replaced with the three year cap on claims, without any extension for mistake. I have already explained how the failure to provide for transitional claims was held to infringe EU law. None of these changes, in my view, has any bearing on the question which I now have to consider, and they throw no useful light on the meaning of “repay” in subsection (7).

In his opening submissions, Mr Swift pointed out that section 24 of the Finance Act 1989 provided for the recovery of overpaid VAT at a time before the English law of restitution had recognised a right to recover payments made under a mistake of law (which came only with the decision of the House of Lords in Kleinwort Benson Ltd v Lincoln City Council [1999] 2 AC 349), and also before the Woolwich cause of action to recover unlawfully demanded tax had first been recognised by the House of Lords in 1992. Mr Swift suggested that, by providing a remedy in circumstances where (as the law was then understood) none would otherwise have been available, Parliament could not have contemplated the possibility of claims for the recovery of overpaid VAT being made otherwise than through the medium of section 24, and that this should encourage the court to give a wide interpretation to the exclusion in subsection (7).

I am not sure that I follow the logic of this submission, bearing in mind the observation of Arden LJ in Monro at [3] that “the court cannot adopt a different approach to the interpretation of a statute simply because the common law has developed since the statute was enacted or because Parliament may not have fully appreciated recent developments in the common law”. But in any event, any force which the point might otherwise have had is in my judgment blunted by the fact that, as Mr Rabinowitz was able to establish by reference to the decision of the House of Lords in Customs and Excise Commissioners v Fine Art Plc [1989] AC 914, section 24 did not appear in a legislative vacuum. The House of Lords held in that case that paragraph 2(4)(c) of schedule 7 to the Value Added Tax Act 1983 gave the Commissioners specific power to make regulations for the correction of errors, whether of law or fact, and that the prescribed form of VAT return, read with the relevant requirements in the Value Added Tax Guide, gave a taxpayer a legal right to deduct from current liability to VAT past overdeclarations which had been made through an error of either law or fact. Fine Art Plc had taken advantage of this machinery in order to deduct VAT which it had overdeclared between 1981 and 1983, pursuant to a direction given by the Commissioners which a subsequent decision of the ECJ had shown to be void. It thus appears that section 24 was enacted in order to provide a formal framework for the exercise of this right in primary legislation, to impose a time limit for its exercise, and to introduce a defence of passing on.

To conclude, even if reference to the statutory history of section 80 were permissible, it would not in my view assist in the resolution of the present question. I therefore hold, as a matter of English domestic law, that the claims in the present case are all barred by section 80(7), and that subject to the issues of EU law to which I will now turn, they must be dismissed.

I have already set out in paragraph 34 above the parties’ agreed formulation of the EU law issues. In the event, however, Mr Rabinowitz approached the matter in a rather different way in his oral submissions on behalf of the claimants, and began with what may aptly be termed a threshold issue, namely whether the claimants in their capacity as end customers could in principle have any enforceable EU law rights upon which they could rely against HMRC in the English court, whether pursuant to the direct effect of Article 13B of the Sixth Directive or on any other basis. Clearly, if that question is answered in the negative, the enquiry need proceed no further: the case would have no EU law dimension, and (subject to any appeal) the dismissal of the claimants’ domestic law claims would be conclusive.

The claimants’ argument for an affirmative answer to this question depends essentially on the decisions of the ECJ in three cases: Claverhouse, Reemtsma and Danfoss. I have already explained (see paragraph 37 above) that the decision of the Court in Danfoss was still awaited at the time of the trial before me in May 2011, and that the ECJ did not give its judgment until 20 October. A good deal of the argument before me was therefore devoted to an examination of the opinion of Advocate General Kokott, which was available only in an unofficial English translation, and to speculation about what the Court might in due course decide. Since, however, the Court has now pronounced, and since it seems to me that its reasoning clearly establishes that there are limited circumstances in which the claimants could in principle be entitled to bring a claim against HMRC for reimbursement of the unlawfully levied VAT, I find it convenient to begin with that case, and then to look more briefly at Reemtsma and Claverhouse.

The facts in Danfoss were briefly as follows. The case did not concern VAT, but a Danish excise duty on lubricating oil which was levied from 1992 in purported compliance with an EU Directive (Council Directive 92/81/EEC of 19 October 1992) which required member states to impose a harmonised excise duty on mineral oils, but subject to an exemption for mineral oils used for purposes other than as motor or heating fuels. The Danish legislation which transposed this Directive into national law imposed duty on all lubricants and hydraulic oils, but failed to provide for the exception in purported reliance on Article 3(2) of the Directive which authorised the subjection of the relevant products “to other indirect taxes for specific purposes”. In due course, decisions of the ECJ made it clear that this attempt to circumvent the exemption was ineffective, and in 2002 the Danish tax authorities abolished the duty with effect from 1 December 2001.

Between January 1995 and November 2001, Danfoss purchased lubricating oil from various Danish suppliers. The price included the duty, all of which was passed on by the suppliers to their customers, and for which the suppliers accounted to the Danish exchequer. When it transpired that the duty was unlawful, the suppliers took no steps to claim reimbursement of it from the Danish tax authorities presumably because they had passed on the full burden of it to their customers, including Danfoss. Danfoss therefore claimed reimbursement of the duty which it had borne, and also claimed compensation from the Danish State for the loss it had suffered as a consequence of the duty being levied. These claims were rejected by the Danish authorities. The reimbursement claim was rejected on the ground that the only person entitled under EU law to recover unlawfully paid tax was the person who was directly taxable, and not a customer further down the marketing chain. The compensation claim was rejected on the ground that no sufficiently direct causal link could be established in a situation where an unlawful tax had been passed on down the chain. Danfoss appealed against these decisions, and the national court referred the following questions to the ECJ for a preliminary ruling:

The Court considered the first question in paragraphs 19 to 29 of its judgment. In paragraph 19 it identified the essence of the first question as being

The Court then dealt with the second question in paragraphs 30 to 39 of its judgment. The Court recited the well-established principles that the legal classification of actions brought before the national court is a matter for the national court to determine, and that an action for damages may coexist with a San Giorgio claim. It then referred to the three conditions which have to be satisfied if a breach of EU law is to give rise to liability for damages on the part of a member state, namely (paragraph 33) “the rule of EU law infringed must be intended to confer rights on those individuals [i.e. those harmed by the breach]; the breach of that rule must be sufficiently serious; and there must be a direct causal link between the breach and the loss or damage sustained by the individuals”.

The Court went on to observe that, while it is in principle for the national court to ascertain whether there is a sufficiently direct causal link, the ECJ may nevertheless give the national court guidance on the question, and continued:

A central point which emerges from Danfoss is the identification, in paragraphs 23 and 25 of the judgment of the Court, of the purpose of the San Giorgio principle as being to help to neutralise the economic burden (my emphasis) which the unlawful tax has imposed on the person who, in the final analysis, has actually borne it. In the light of this purpose, the EU law principle of effectiveness may in certain circumstances require the national law of a member state to afford a remedy to a final consumer who has borne the burden of an unlawful tax, whether in the form of reimbursement of the tax or damages, if and to the extent that the final consumer is unable to obtain a remedy from the taxable person himself. The example given in paragraph 28 of circumstances where reimbursement by the taxable person would prove impossible or excessively difficult (i.e. where the principle of effectiveness would be infringed, if that were the only remedy afforded by national law) is the insolvency of the taxable person; but it is clear, in my judgment, that this is only an illustration, albeit the one probably most likely to arise in practice, and was not intended to be exhaustive. The governing principle is that national rules of classification of actions and procedure must respect the principle of effectiveness, and must therefore afford a remedy against the member state itself to a final consumer who has borne the economic burden of an unlawful tax, in circumstances where recovery from the taxable person proves impossible or excessively difficult.

This reasoning is in my judgment sufficient to show that the threshold issue must be resolved in the claimants’ favour. The claimants are investment trusts which have borne the full economic burden of an unlawful tax. They have exhausted their remedies against the Managers, who were the taxable persons, but have not thereby been able to recover the full amount of the unlawful VAT which they paid, and by which (as I have held) HMRC were unjustly enriched. If, therefore, the principle of effectiveness requires that the claimants should have a remedy for the otherwise irrecoverable balance of the unlawful VAT, or for any part of it, national law must provide an appropriate procedure for that purpose. In so doing, national law would be recognising and giving effect to a right derived by the claimants from EU law, and from no other source. The proposition that in no circumstances could the claimants have any enforceable EU law rights to recover or seek damages from HMRC in respect of the unlawfully levied VAT is therefore untenable.

I emphasise that at this stage I am examining only the threshold question, and deciding no more than that the claimants could in principle have rights under EU law upon which they could rely against HMRC. I am expressing no view at this stage on the questions whether, and (if so) to what extent, the principle of effectiveness does in fact require English law to provide the claimants with a remedy.

It also needs to be borne in mind that Danfoss was not a case about VAT, although the Court in its reasoning drew on its earlier decision in Reemtsma, which was a VAT case. The Court therefore did not have to consider the possible complications caused by the right of the taxable person to deduct input tax, and the limit which this placed on the amount that the Managers were able to recover from HMRC (the £75 as opposed to the £100). Nor did the Court have to consider any question of limitation, such as the three year cap. These questions are, however, highly relevant to the next main stage in the enquiry, which is essentially what remedy (if any) EU law requires English law to make available to the claimants in the particular circumstances of the present case.

Before moving on to the next stage, however, I will first examine the decisions in Reemtsma and Claverhouse to see if they throw any further light on the threshold issue.

Reemtsma was a cross-border case, and therefore involved the provisions of the Eighth, as well as the Sixth, Directive. In very broad terms, VAT is not payable under the Sixth Directive on cross-border supplies of goods or services, but the recipient of the goods or services, if established in another member state, is liable to account in its home state for output tax on the supply to it pursuant to the “reverse charge” mechanism in Article 21(1)(b) of the Sixth Directive, and is also entitled to deduct that VAT as input tax from its domestic output tax liability. The Eighth Directive, which applies only to taxable persons, contains detailed provisions for the refund of foreign input tax paid in cross-border transactions, its general purpose being to replicate in the case of cross-border supplies the system of deduction of input tax from output tax which applies in a purely domestic context.

Reemtsma was a cigarette manufacturer resident in Germany. It had no permanent establishment in Italy. In 1994 an Italian company provided Reemtsma with advertising and marketing services on which it charged VAT, which Reemtsma duly paid, and for which the supplier duly accounted to the Italian tax authorities. Reemtsma then sought a partial refund of the tax from Italy, pursuant to the Eighth Directive. It transpired, however, that the VAT had been charged in error by the Italian supplier, because the proper place of supply of the services was Germany, with the result that Reemtsma should instead have accounted in Germany for VAT on the supply under the reverse charge mechanism. In these circumstances, Reemtsma sought reimbursement of the VAT which it had paid from Italy, but this was refused on the ground that the relevant services had never been subject to VAT in Italy. This refusal was upheld in the Italian courts at first instance and on appeal, but the Court of Cassation referred two questions to the ECJ. A further point which needs to be noted is that, under Italian domestic law, the supplier was entitled to a refund of VAT unduly paid from the authorities, but a recipient established in another member state (like Reemtsma) could seek reimbursement only from the supplier.

The first question referred to the ECJ asked, in essence, whether unlawfully charged VAT which had been paid and accounted for to the tax authorities was refundable pursuant to the Eighth Directive. The second question, which is the one that matters for present purposes, was in the following terms:

The answer which the ECJ gave to the first question was that VAT invoiced by mistake was not refundable under the Eighth Directive: see paragraphs 17 to 28 of the judgment. This answer would have been sufficient to dispose of the case, but the Court nevertheless went on to consider the second question, which it broke down into three component parts. In paragraph 34 it identified the second of those components as being

The Court’s answer to the question thus reformulated appears sufficiently clearly from paragraphs 39 to 42 of the judgment:

With the benefit of hindsight, it is in my judgment clear that in these paragraphs the Court was intending to lay down principles of general application in the fields of VAT and indirect taxation. Those principles have now been applied and restated by the Court in Danfoss, and linked to the San Giorgio principle. It is also worth noting that the Court’s recognition of the possible need for a direct right of recovery from the tax authorities by a recipient of services who is unable to obtain reimbursement from his supplier stemmed from the acceptance by the Court of submissions made by the Commission: see Reemtsma at paragraph 41, and the fuller discussion of the subject in paragraphs 82 to 92 of the opinion of Advocate General Sharpston, who said at paragraph 84 that she found the Commission’s analysis “persuasive in its entirety”. Mr Swift advanced a number of submissions to the general effect that the relevant part of the decision in Reemtsma was tentative, obiter (a concept which I would comment is unknown to EU law jurisprudence), distinguishable, or confined in its scope to taxable persons. I do not find it necessary to review these submissions, however, because in my view Danfoss shows them to be untenable.

I can deal with Claverhouse much more briefly, because the main point that Mr Rabinowitz sought to extract from it in the present context was that the claimants are entitled to rely on the direct effect of Article 13B of the Sixth Directive, even though they did not themselves pay or account for the VAT to HMRC. Mr Rabinowitz relied on the fact that the claimants in Claverhouse were (as in the present case) investment trusts, not managers, and it was at their instigation that the Tribunal referred the question to the ECJ asking whether Article 13B(d)(6) had direct effect. It is true that the Court’s answer to this question, in paragraph 62 of the judgment, was that Article 13B(d)(6) had direct effect “in that it can be relied on by a taxable person before a national court in order to challenge the application of national legislation alleged to be incompatible with that provision” (my emphasis). This language might be thought to indicate that only the Managers could take advantage of the provision’s direct effect. However, Mr Rabinowitz was able to point to other passages in the judgment (at paragraphs 56 and 61) which suggest that the Court may not have intended its answer to be so narrowly confined. He also submitted that the investment trusts were, in any event, “taxable persons” within the meaning of the definition of that term in Article 4.1 of the Sixth Directive, as applied by the ECJ in Case C-8/03 Banque Bruxelles Lambert SA v Belgium, [2004] ECR I-10157, [2004] STC 1643, (“BBL”), where it held that an open-ended investment company established in Luxembourg could be a taxable person. If that was true of an open-ended investment company, said Mr Rabinowitz, it must also be true of a closed-end investment trust, given the equation of the two types of investment trust by the ECJ in Claverhouse.

In my judgment these arguments have considerable force, both singly and collectively. On the other hand, there are counter-arguments which cannot be ignored. In particular, the ECJ in Claverhouse does not seem to have directed its mind to any possible distinctions between the positions of investment trusts and managers, because the real dispute in relation to direct effect was whether the discretion afforded to member states in Article 13B(d)(6) meant that the exemption therein provided could not satisfy the test of being sufficiently precise and unconditional to be relied upon by individuals (i.e. the test in Case 8/81 Becker [1982] ECR 53, and subsequent case law). Furthermore, the context in which the ECJ held in BBL that an open-ended investment company could be a taxable person was far removed from the present case, and did not involve the assertion of any right to recover tax paid by an end consumer.

If it were not for the decisions in Reemtsma and Danfoss, I would probably have taken the view that the answer to the question whether the claimants in the present case could rely on the direct effect of Article 13B(d)(6) was unclear, with the result that a reference to the ECJ would have been necessary. But in my judgment those two decisions make all the difference. The possible right of action recognised in those cases by an end consumer against the State only makes sense on the assumption that the claimant would be entitled to rely on the direct effect of the relevant exemption, whether because the claimant is properly to be regarded as a taxable person (as in BBL), or because it is not necessary for the claimant to be a taxable person. If it were otherwise, the claimant would have no right to recover the tax under EU law to which the principle of effectiveness could apply.

Having decided the threshold issue in the claimants’ favour, I next have to examine the question whether EU law does in fact require a remedy to be made available to the claimants in the circumstances of the present case, and (if so) what form that remedy should take. I will be fairly brief, because in my view the combined effect of the decisions of the ECJ in Reemtsma and Danfoss leaves little room for doubt about most of the answers. In short, I consider that EU law requires English law to provide a direct remedy for the claimants to recover the whole of any otherwise irrecoverable VAT from HMRC, but subject to limitation defences analogous to those which would apply to claims by the Managers against HMRC. I do not consider it to be a tenable proposition that the claimants should be confined to a claim for damages on Francovich principles, which it is common ground they would (for various reasons) be unable to pursue.

The starting point in the analysis is to ask whether the EU principle of effectiveness is engaged in the claimants’ favour. This depends on the answer to the question whether reimbursement of the claimants by the Managers has proved impossible or excessively difficult: see Danfoss at paragraph 28 and Reemtsma at paragraph 41.

As I have already indicated, I can see no good reason to confine the concept of impossibility or excessive difficulty to insolvency of the Managers, or to similar external causes which impact on their financial ability to meet otherwise valid claims against them. From the perspective of the claimants, the amount for which they are prima facie entitled to claim reimbursement is the full amount of the unlawful VAT which they paid to the Managers, i.e. the £100. The fact that the Managers may be able to recover only £75 from HMRC has no bearing on the fact that the amount actually paid by the claimants was the full £100. That is also the amount, as I have held, by which HMRC have been unjustly enriched at the claimants’ expense. Thus, to the extent that the claimants are unable to recover the £100 from the Managers, it seems to me to follow that reimbursement of the claimants by the Managers has, as a matter of fact, proved impossible. I conclude, therefore, that the principle of effectiveness is at least potentially engaged in the claimants’ favour.

The next question is whether, on the facts, the requirements for application of the principle of effectiveness are satisfied. In relation to the £75 element of the claims, there can in my judgment be no doubt about the answer. It is not in dispute that the claimants have exhausted their remedies against the Managers in relation to this part of their claims, and that the Managers would have a cast iron defence of change of position in relation to the sums of £75 which they paid to HMRC and which they have been unable to recover from HMRC because of the three year limitation period. To the extent of the £75 element of the claims, therefore, recovery by the claimants from the Managers has indeed proved impossible for the dead period.

In relation to the £25 element of the claims, however, the position is less straightforward. Until the service of Mr Richardson’s supplementary witness statement on 12 May 2011, less than a week before the start of the trial, it appeared likely that it would also be common ground that any attempt by the claimants to recover the £25 from the Managers would likewise be fruitless. Thus, in paragraph 81 of their opening written submissions (prepared before the receipt of Mr Richardson’s further evidence), counsel for the claimants pointed out that, in determining the section 80 claims made by the Managers, HMRC had accepted that the associated input tax of £25 was a retrospective cost to the Managers which they had not passed on to the investment trusts, on the assumption that it was deductible for VAT purposes, but which (it was to be inferred) they would have passed on to the investment trusts had the supplies of investment management services been treated as exempt from the beginning. In other words, the natural inference was that the Managers, faced with irrecoverable input tax, would have decided to make a corresponding increase in the price which they charged to the claimants. On this footing, the claimants could not have claimed the £25 from the Managers, “technically because [the Managers] had a change of position defence … and in reality because [the Managers] would have included that amount in the form of a higher ex-VAT charge to [the investment trusts]”. The change of position defence would have arisen, on this view of the facts, because the Managers would have irrevocably deprived themselves of the opportunity to increase the prices which they charged to the investment trusts by the otherwise irrecoverable £25.

I have already summarised the effect of Mr Richardson’s last minute evidence in paragraph 84 above, the crucial point being that, had the true VAT position been known, the relevant Managers would not have sought to increase the price of their services to the claimants, and they would have been content to absorb the irrecoverable input tax as a business expense. Moreover, this would have been the position both before and after the commencement of the Claverhouse litigation. In the light of this evidence, which was not challenged, Mr Swift QC made it clear at an early stage of the hearing that it was not common ground that the Managers would have a change of position defence in relation to the £25 element of the claims. Later on, Mr Swift developed the submission that the claimants could in principle have brought, and indeed could still bring, successful restitutionary claims to recover the £25 from the Managers, even though the Managers were unable to recover the £25 from HMRC. The Managers would be unable to establish a change of position defence, in view of Mr Richardson’s evidence about what they would actually have done had the true position under EU law been known. Accordingly, said Mr Swift, there is no room for the principle of effectiveness to operate in the claimants’ favour, because English law provides an adequate cause of action for recovery of the £25 by the claimants from the Managers, and it cannot be said that such recovery would in principle be either impossible or excessively difficult.

Mr Rabinowitz’s answer to this submission was briefly as follows. He accepted that the claimants had a prima facie restitution claim against the Managers on the basis of mistake. He also accepted that Mr Swift’s analysis might have force if the £25 had always stayed with the Managers, and if HMRC had never been enriched by the £25. But, he said, HMRC were in fact enriched by the full £100, for all of which the Managers had duly accounted to HMRC, even though the payments which they actually made were of the net amounts of £75. Accordingly, the Managers had a good change of position defence in relation to the full £100, and any attempt by the claimants to recover the £25 from the Managers would fail.

I accept the argument for the claimants on this point. It seems to me that the Managers changed their position in relation to the entirety of the £100 when they accounted for it as output tax in their quarterly VAT returns. The fact that they also received credit for the associated input tax does not in my judgment alter the position. The receipt of the credit was simply a consequence of the operation of the VAT rules which everybody was operating on the mistaken assumption that the investment management services were not exempt. Nor do I consider it relevant to enquire what the Managers would have done in the hypothetical situation where it was known to all concerned that the services were in fact exempt, not least because in that event there would have been no payment of £100 by the investment trusts in the first place, and the question of recovering it from the Managers would therefore not have arisen.

For these reasons, I consider that the requirements for the principle of effectiveness to operate in the claimants’ favour are satisfied. I would only add that, even if I am wrong in considering it to be irrelevant what the Managers would have done in a hypothetical world where it was known that the services were exempt, I do not think that it would be reasonable, from an objective standpoint, to require the claimants to pursue potentially difficult and uncertain claims against the Managers to recover the £25 before it could properly be said that such recovery had become practically impossible. Although I did not hear argument on the point, my provisional view is that this requirement should not be interpreted too severely, and that an analogy could helpfully be drawn with the position where a claimant in a case of breach of contract is not required to mitigate his loss by embarking upon litigation with an uncertain outcome.

Having reached this stage, the next question is what remedy the principle of effectiveness actually requires English law to provide in the circumstances of the present case, bearing in mind the well established principle of national procedural autonomy.

It is common ground that EU law permits even San Giorgio claims for the reimbursement of unlawfully levied tax to be subject to reasonable national limitation periods. It is also common ground that the three year cap was, in itself, a reasonable limitation period which was fully compatible with EU law. The only respect in which it infringed EU law lay in the failure to provide adequate transitional arrangements. In the context of the present case, this failure was fully addressed by the further refund claims which HMRC allowed to be made for accounting periods ending before 4 December 1996: see paragraphs 23 and 24 above. It seems to me, therefore, that the principle of effectiveness should not be interpreted as requiring the UK to afford the claimants any right to recover VAT from HMRC which would not be subject to the same limitation period (by analogy) as that which would apply to claims by the Managers to recover the same tax themselves. If this were not so, the claimants would be placed in a more favourable position for limitation purposes than the Managers, even though the Managers actually paid or accounted for the tax in its entirety to HMRC. Since HMRC have an EU law-compliant limitation defence against the taxable persons themselves, it would in my judgment be anomalous and unreasonable if, in the name of effectiveness, EU law were to require the national court to make available to a person who has borne the economic burden of the tax a claim which would not be subject to the same defence.

Another way of expressing the same conclusion might perhaps be to say that the inability of the claimants to recover the £100 in respect of the dead period cannot sensibly be characterised for the purposes of EU law as a situation where reimbursement of the VAT has proved impossible, because the impossibility has been caused, and caused only, by the effect of national limitation rules which are themselves fully compliant with EU law. In other words, in relation to the dead period claims, the principle of effectiveness would not be engaged at all.

The difficulty with analysing the matter in this way, however, is that it might suggest a distinction should be drawn for the purposes of the dead period between the £75 element of the claim, which is the most that the Managers could have recovered from HMRC and passed on to the claimants in the absence of the three year cap, and the £25 element, which the Managers would still have been unable to recover. Such a distinction would in my opinion be unwarranted, when what is really required is the application by analogy of the three year cap in section 80(4) to the entirety of the reimbursement claims brought by the claimants. In my view such a result can be achieved only by recognising that the principle of effectiveness is engaged, and then disapplying or moulding the provisions of section 80 to the extent necessary to accommodate the claimants’ EU law rights within an analogous limitation period.

On the other hand, there should be no limitation defence to a claim by the claimants to recover the £25 for accounting periods which are not caught by the three year cap, and I consider that in principle they should be entitled to recover this part of their claim from HMRC. Limitation apart, it is only in this way that the claimants could obtain full reimbursement of the tax which was wrongly invoiced to them, and which they have paid in full. Equally, it is only in this way that HMRC could be made to disgorge the full amount of the unlawful VAT by which they have been unjustly enriched at the claimants’ expense.

The mechanism by which this result could be achieved is in my view relatively simple. As matters stand, and on the assumption that the conclusions which I have reached in the earlier parts of this judgment are correct, the claimants have a valid common law cause of action in restitution against HMRC, and the only thing which defeats it is section 80(7) of VATA 1994. In order to give the claimants an effective remedy to recover the £25 for accounting periods which would not be time-barred under an extended version of section 80(4), all that would be needed would be to disapply section 80(7) in relation to such claims, but no further. In principle, therefore, and subject to the issues discussed in the remaining sections of this judgment, this would in my judgment be the appropriate way in which to give effect to the claimants’ directly effective EU rights in the present case.

The next issue under this head is whether the claims in the present case are properly to be classified for the purposes of EU law as San Giorgio claims. In my opinion the judgments of the ECJ in Reemtsma and Danfoss are clear on this point: they are. In each judgment, the Court consistently uses the language of reimbursement of sums unduly paid in order to describe the nature of the claim by an end consumer to recover a sum passed on to it in the mistaken guise of excise duty or VAT. This is the appropriate language for a San Giorgio claim (see for example Danfoss at paragraph 20, and Reemtsma at paragraph 41), and such a classification would in my view entirely accord with the purpose of the right of recovery under EU law as identified in Danfoss at paragraphs 23 and 25. Furthermore, any possible doubt on this score is in my judgment dispelled by the fact that in Danfoss the national court asked separate questions in relation to the reimbursement and damages claims brought by the claimants, and the Court then dealt with them separately, making it clear that it regarded the former claim as being a species of San Giorgio claim. It is true that Danfoss was an excise duty case; but in this respect I cannot see any rational distinction between excise duty and VAT.

Such a conclusion also has the merit, to my mind, of according with common sense. The sums which the present claimants seek to recover are the amounts of the wrongly invoiced VAT, the full burden of which has been borne by them and the full benefit of which has accrued to HMRC. In economic terms, it is a claim by an end consumer to recover the tax which he has had to pay; and it is a claim brought against the fiscal authority which levied the tax. As a matter of ordinary language, it is a claim for repayment or reimbursement of the tax which has been borne by the claimants, who under the scheme of VAT are the “real” taxpayers. Furthermore, the quantum of the claim is defined and limited by reference to the VAT invoiced to the claimants. By contrast, the claim would not naturally be described as a claim for consequential loss occasioned by the imposition of the unlawful tax. Mr Swift’s main argument for saying that the claims should be classified as damages claims was that the claimants are seeking to recover from HMRC more than the amounts actually paid by the Managers to HMRC; but this point falls away, in my judgment, once it is recognised that HMRC have had the benefit, directly or indirectly, of the full amount of the tax, and that in principle all that the claimants wish to do is to recover that amount (no more and no less) from HMRC to the extent that they are unable to recover it from the Managers.

It might be thought that my analysis could safely end at this point, on the footing that the effective remedy which EU law requires English law to provide, so that the claimants can vindicate their San Giorgio claims, must entitle them to succeed on their mistake-based restitutionary claims for the uncapped periods. After all, I have already held that such claims are in principle valid claims under the English common law of restitution, and that the only obstacle to their successful prosecution lies in the statutory prohibition in section 80(7) of VATA 1994. Surely all that is needed, in order to satisfy the EU principle of effectiveness, is for the statutory prohibition to be overridden to the extent necessary to permit the otherwise valid claims to be brought? Such an approach would have the merit of simplicity, but HMRC submit that it is both wrong in principle and contrary to decided authority.

HMRC argue that the task which the English court has to perform is to identify the domestic cause of action which, on an objective appraisal, would provide the most appropriate remedy, or the “best fit”, for vindication of the claimants’ EU rights which have been infringed. The cause of action which fits best, they say, is the cause of action in the law of restitution to recover tax which has been unlawfully exacted. This cause of action was first recognised by the House of Lords in the Woolwich case (Woolwich Equitable Building Society v IRC [1993] AC 70). For many years it was widely thought that the existence of an unlawful demand for the tax was a necessary ingredient of the Woolwich cause of action, but the Court of Appeal has now held in clear terms that there is no such requirement: see Test Claimants in the FII Group Litigation v Revenue & Customs Commissioners [2010] EWCA Civ 103, [2010] STC 1251, (“FII (CA)”) at paragraphs [152] to [173] of the judgment of the Court. In my judgment at first instance in the same case, to which it is convenient to refer as “FII (Chancery)”, [2008] EWHC 2893 (Ch), [2009] STC 254, I had reached the opposite conclusion, taking the view that I was bound to do so by the judgment of the Court of Appeal in NEC Semi-Conductors Ltd v Revenue & Customs Commissioners [2006] EWCA Civ 25, [2006] STC 606. Had I been correct in this view, the Court of Appeal would of course have been bound likewise; but the Court of Appeal held in FII (CA) that the relevant passage in the judgment of Mummery LJ in NEC Semi-Conductors was obiter, with the result that the Court was not bound by it: see the judgment at paragraph [169].

For present purposes, it will suffice if I cite a brief extract from the judgment of the Court in FII (CA) on this point:

After reviewing the reasoning of the majority in Woolwich, and referring to other relevant authorities including NEC Semi-Conductors, the Court then concluded:

It is also material to refer to what the Court of Appeal said at paragraph [225], dealing with the question whether the application of section 320 of the Finance Act 2004 and section 107 of the Finance Act 2007 (which excluded the extended limitation period under section 32(1)(c) of the Limitation Act 1980 in relation to mistakes of law relating to taxation matters under the care and management of HMRC) was precluded by reason of the EU law principles of effectiveness, equivalence, legal certainty and legitimate expectations:

It follows that the Woolwich cause of action, as elucidated in FII (CA), would appear to provide a sufficient domestic remedy for an English taxable person who had accounted for and paid unlawfully levied VAT. Moreover, it is clear from the reasoning of the Court of Appeal in paragraph 225 that, in the view of the Court, the requirements of the principle of effectiveness are satisfied, in the context of San Giorgio claims to recover unlawfully levied tax, by the Woolwich cause of action alone, and that the reach of the principle therefore does not extend to any mistake-based causes of action within the purview of sections 320 and 107.

Pausing at this point, I observe that there is still a bridge that needs to be crossed before it could be said that the Woolwich cause of action would be available to the claimants in the present case. It is true that their claims are to be characterised, for the purposes of EU law, as San Giorgio claims; but, so far at any rate, it has never been held that a Woolwich remedy is available to anybody other than the actual taxpayer who has paid the unlawful tax to HMRC. In cases of the present type, this means that (in the absence of section 80 of VATA 1994) the Managers, as the taxable persons, would be entitled to bring a Woolwich claim against HMRC to recover the unlawful VAT which they paid, but the investment trust claimants would not, because the tax was never exacted from them by the State, and their obligation to pay it to the Managers was purely contractual.

Mr Swift submits, however, that I should not be deterred from taking this further step. He argues that if it is appropriate to recognise a San Giorgio right under EU law for an end consumer to recover unlawfully levied VAT from the State – and one only reaches this stage in the argument if such a right has been recognised – then EU law must likewise require the domestic remedy for San Giorgio claims to be moulded to accommodate such cases, on the assumption that it does not do so already. He goes on to submit that, if I need encouragement to take this step, I can find it (by analogy) in FII (CA), where the Court of Appeal held (at paragraph [174]) that if, contrary to what it had held, a wrongful demand is an essential ingredient of the Woolwich cause of action in a purely domestic case, “that requirement must be displaced in the context of Community law”. The reasoning of the Court on this point, necessarily obiter, was as follows (ibid):

It may at first blush seem surprising to find HMRC arguing for an extension of the Woolwich cause of action, but the explanation for this lies in the crucial second stage of the approach adopted by the Court of Appeal in FII (CA). It is clear from the passages which I have cited that the Court of Appeal took a restrictive view of what is required by the principle of effectiveness. They held, in effect, that once a domestic cause of action had been identified which was in principle capable of satisfying the claimants’ San Giorgio claims, in the shape of the Woolwich remedy (shorn, if necessary, of the requirement for a demand), there was then no need to go any further, and in particular there was no need to allow the freedom which a claimant would enjoy in a purely domestic context to make a mistake-based restitutionary claim with the more generous limitation period afforded by section 32(1)(c) of the Limitation Act 1980. In Test Claimants in the Thin Cap Group Litigation v Revenue & Customs Commissioners [2009] EWHC 2098 (Ch), [2010] STC 301, (“Thin Cap (Chancery)”, in which I handed down judgment on 17 November 2009, I had ventured to suggest a more relaxed view of what the principle of effectiveness requires: see paragraph [223] of my judgment. It seems clear to me, however, that this view cannot stand with the approach and reasoning of the Court of Appeal in FII (CA), judgment in which was handed down a few months later on 23 February 2010.

The implications of this aspect of the judgment in FII (CA) came to the fore in Littlewoods (No.1), where Vos J had to deal with a number of issues arising from restitutionary claims for compound interest made by traders who had been repaid unlawfully levied VAT pursuant to section 80 of VATA 1994 together with the simple interest provided for by section 78. The claimants made both Woolwich and mistake-based restitutionary claims. Vos J held that all of these claims were excluded, as a matter of domestic English law, by the exhaustive statutory regime in sections 78 and 80. Having reached this conclusion, Vos J then had to consider whether the exclusion was contrary to EU law, and (if so) how and to what extent the provision for simple interest in section 78 fell to be disapplied. (There was a prior question whether the conflict with EU law, if there was one, could be resolved by adopting a conforming construction of section 78, but Vos J disposed of this possibility in paragraphs [74] to [76] of his judgment).

On the issue of the compatibility of section 78 with EU law, Vos J came to the preliminary conclusion that there was no conflict between section 78 and EU law, but recognised that the answer to the question was not clear. He therefore decided that three questions needed to be referred to the ECJ, in terms which he provisionally set out in paragraph [69] of his judgment.

On the issue of disapplication, which would of course arise only if the ECJ held in due course that section 78 did infringe EU law, Vos J heard and considered arguments which were similar in many respects to those addressed to me (by the same leading counsel) in the present case. The relevant passage in his judgment runs from paragraphs 77 to 96. After summarising the rival arguments, Vos J helpfully set out in paragraph [82] the basic assumptions upon which he was considering this issue. In paragraph [83], he referred to my observations about the principle of effectiveness in Thin Cap (Chancery) at paragraph [223], and to the express decision in FII (CA) at paragraph [225], cited above. He then continued:

In the light of this analysis, Vos J expressed his provisional conclusion in paragraph [92] as being that sections 78 and 80 of VATA 1994 “must be disapplied so as to allow only the Woolwich claims”. He then considered whether there was any need to refer the issue to the ECJ, and decided that it would be appropriate to do so, even though HMRC argued that the question was purely one of domestic law and procedure. Vos J’s reasons for taking this course appear from paragraph [94]:

In paragraph [95] Vos J set out the question that he considered it appropriate to refer, but said he would hear further argument on the precise formulation. After hearing such argument in early November 2010, he settled Question 4 in the Order for Reference as follows (see Littlewoods Retail v Revenue & Customs Commissioners [2010] EWHC 2771 (Ch), [2011] STC 171, (“Littlewoods (No. 2)” at 177h):

The current position on the reference is that the oral hearing before the ECJ took place in November 2011, and the opinion of Advocate General Trstenjak was delivered on 12 January 2012. The decision of the Court is awaited.

It is also relevant to note at this point that the Supreme Court granted permission to appeal to the claimants in the FII litigation from the decision of the Court of Appeal on the Woolwich-related issues. The hearing before the Supreme Court started on 21 February 2012.

Against this background, the submissions of HMRC on this part of the case are in summary as follows. Relying on FII (CA) and Littlewoods (No. 1), they submit that:

the appropriate domestic cause of action to satisfy the claimants’ San Giorgio rights to repayment of the VAT which they paid to the Managers is the Woolwich cause of action, extended so as to include claims by end consumers;

the Woolwich cause of action, thus extended, would by itself satisfy the EU principle of effectiveness, because it would provide the claimants with a remedy which was neither practically impossible nor excessively difficult to exercise;

there is accordingly no need, as a matter of EU law, for a mistake-based restitutionary cause of action also to be made available to the claimants; and

the prohibition in section 80(7) of VATA 1994, which as a matter of English domestic law would bar both causes of action, should therefore be disapplied only in relation to the extended Woolwich cause of action.

For their part, the claimants attack this reasoning and conclusion on a number of separate grounds. In the first place, they submit that, whatever the position may have been in FII, in the present case there is only one domestic cause of action which fits the facts of the case without modification, namely the mistake-based restitutionary claim. Secondly, they submit that the court should in any event decline HMRC’s invitation to extend the Woolwich cause of action, the ambit of which should be confined to claims by actual taxpayers. Thirdly, they submit that the narrow approach to the principle of effectiveness adopted by the Court of Appeal in FII (CA), and provisionally applied by Vos J in Littlewoods (No. 1), cannot be right as a matter of English law, as is demonstrated by the decision of the House of Lords in DMG (see paragraphs 167 to 169 below). Finally, they submit that, even if it comes down to making a choice between different possible causes of action, in cases of the present type the “best fit” is provided by the mistake-based restitutionary cause of action. It would be paradoxical and absurd, say the claimants, if, in the name of effectiveness, they were to be deprived of the domestic cause of action which fits the facts of the present case like a glove, and which would provide them with a remedy, while being confined to an extended version of the Woolwich cause of action, the existence of which nobody could have foreseen before the Court of Appeal gave judgment in FII (CA), and which would now leave them without any remedy at all.

The general nature of the first, second and last of these submissions should be clear enough from what I have already said about HMRC’s case, but the third submission (based on DMG) needs some further explanation. A convenient starting point is the summary of DMG given by Vos J in Littlewoods (No.1) at paragraph [34]:

Mr Rabinowitz submits that, in one crucial respect, Vos J misunderstood the significance of DMG. He says that DMG, properly analysed, did involve the consequences of the disapplication of a rule of English law in order to secure compliance with EU law. The rule of English law which had to be disapplied was the principle in President of India, as it was then understood, to the effect that there was no power to award interest on debts paid before proceedings were begun. In Hoechst, the ECJ had expressly ruled that this principle had to be disregarded by the national court in providing a remedy for the premature levy of ACT: see the judgment of the Court at paragraphs [79] to [80] and [87] to [89]. It is clear, submits Mr Rabinowitz, that the claims in Hoechst could not have succeeded without the disapplication of the President of India principle, and it is irrelevant for this purpose that the principle was subsequently overruled by the House of Lords in Sempra Metals Ltd v IRC [2007] UKHL 34, [2008] 1 AC 561. Despite the fact that the claims depended on the disapplication of a rule of English law, however, the House held that the claimants were nevertheless entitled to rely on a common law mistake claim as well as on a Woolwich claim, thus reflecting the usual willingness of English law to allow a claimant to choose between concurrent remedies. Nobody appears to have suggested, and the House certainly did not hold, that the EU law principle of effectiveness somehow required this freedom to be circumscribed.

If DMG is viewed in this way, submits Mr Rabinowitz, the fact that the House also held that section 33 of the Taxes Management Act 1970 did not provide an exhaustive remedy for the recovery of taxes paid by mistake becomes irrelevant. It is true that, for this reason, section 33 did not have to be disapplied; but it remains the case that the President of India principle did have to be disapplied, and that without such disapplication neither the Woolwich nor the mistake-based causes of action would have been available to the claimants.

I will say at once that I see considerable force in this argument, which unfortunately does not seem to have been addressed to, or considered by, the Court of Appeal in FII (CA). I will also say, more generally, that the rival submissions on this part of the case appear to me to raise issues of great difficulty and potential importance, in an area where the law is evolving rapidly, and where there is good reason to expect that authoritative guidance on at least some key aspects of the problem may be provided in the near future by either or both of the ECJ (on the reference in Littlewoods) and the Supreme Court (on the appeal in FII). In those circumstances, I see much attraction in the suggestion made by Mr Swift, in the course of his oral submissions, that I should defer deciding this part of the case until after the ECJ and the Supreme Court have given their respective judgments. Although that time seemed fairly distant when I heard argument in May 2011, the preparation of this judgment has been delayed for reasons which I have already explained, and there must now be a good prospect that both judgments will be available by mid-2012. In my view a further adjournment of a few months would be a relatively small price to pay for the real likelihood that guidance at the highest level on the scope of the Woolwich principle, and the requirements of the EU law principle of effectiveness, will become available in the meantime. There must also be a possibility that the guidance will dispel the current fog of uncertainty in such a way that no further hearing on this part of the case proves necessary. Finally, even if a further hearing is needed, it is most unlikely that the adjournment will significantly delay the final resolution of the case, because if I were now to give judgment on these issues, an appeal by the losing party would be inevitable.

Mr Rabinowitz urged me to decline Mr Swift’s invitation to defer my decision on this part of the case. Apart from the question of delay, which has to a large extent been overtaken by events, he argued that the ruling of the ECJ in Littlewoods would be unlikely to take matters forward, because Vos J had already said he was provisionally minded to confine the claimants to Woolwich claims, and the ECJ would not interfere in a matter of domestic English law. He also argued that the Supreme Court in FII would not be dealing with precisely the same issue as that which arises in the present case. He accepted, however, that the ECJ might have something relevant to say about the requirements of the principle of effectiveness as a matter of EU law, and that the Supreme Court might decide some points which would affect my decision on this issue. In my view the likelihood that relevant, and possibly decisive, guidance will be given by either or both of the ECJ and the Supreme Court is considerably greater than Mr Rabinowitz was prepared to acknowledge, and although I have taken his reservations into account I consider that the balance comes down in favour of adjourning this part of the case until after the ECJ and the Supreme Court have delivered their respective judgments.

For the reasons which I have given, my conclusions are in summary as follows: