The Prudential Assurance Company Ltd & Ors v Revenue & Customs

This is another in the series of test cases in which the United Kingdom’s former rules on the taxation of dividends have been challenged as being contrary to the fundamental freedoms of establishment and free movement of capital previously contained in Articles 43 and 56 of the EC Treaty (now Articles 49 and 63 respectively of the Treaty on the Functioning of the European Union).

The present case is concerned with the UK’s treatment of dividends received by UK-resident insurance companies on shareholdings which were held by them as investments and allocated to their pension business and life assurance business. The holdings in question were all of less than 10% (and usually of much less than 10%) of the shares of the relevant companies. In no instance was the shareholding of sufficient size to enable the owner of the shares to control or definitely influence the business of the company or the conduct of its affairs. For that reason, dividends of this type are conveniently referred to as “portfolio dividends”.

For the same reason, Article 43 (freedom of establishment) is no longer in issue, and the breaches of EU law that are now complained of are breaches of Article 56 only. The Court of Justice of the European Union (which I will continue to refer to by its commonly used English abbreviation as “the ECJ”) has held in a number of recent cases that the right of establishment of a person resident in a member state is engaged only where that person’s shareholding in the capital of a company established in another member state gives him definite influence over the company’s decisions and allows him to determine its activities: see, for example, Cadbury Schweppes Plc v IRC (Case C–196/04) [2006] ECR I-7995, [2007] Ch 30, (“Cadbury Schweppes”) at paragraph 31 of the judgment, and Test Claimants in the Thin Cap Group Litigation v IRC (Case C-524/04) [2007] ECR I-2107, [2007] STC 906, (“Thin Cap”) at paragraph 27 of the judgment. Article 56, by contrast, is in principle capable of applying not only in respect of portfolio holdings in companies resident elsewhere in the EU and EEA, but also in respect of portfolio holdings in companies anywhere else in the world (“third countries”).

The first of the present test claimants, The PrudentialAssuranceCompany Limited, issued its claim form in the Chancery Division of the High Court on 8 April 2003. By subsequent amendments made in 2004, two further companies in the Prudential Group, Prudential Holborn Life Limited and Scottish Amicable Life Plc, were added as further claimants.

On 30 July 2003 a group litigation order (“GLO”) was made by Chief Master Winegarten to constitute the “CFC and Dividend Group Litigation”. In broad terms, the common issues of law raised in this GLO concerned (a) the lawfulness of the UK provisions on dividend taxation, and (b) the lawfulness of the UK legislation on controlled foreign companies (“CFCs”). Apart from questions about the compatibility of the relevant legislation with EU law, the GLO raised numerous allied and consequential issues relating to such matters as procedure and jurisdiction (in what forum could or should the disputes be resolved?), limitation, quantum (how should compensation or relief be assessed?) and remedies.

The CFC issues may for present purposes be ignored. This is not because they are unimportant, but because the question of the compatibility of the UK’s CFC legislation with EU law was considered by the ECJ in Cadbury Schweppes, although not as it happens in the context of the present GLO, but on a preliminary reference by the Special Commissioners on an appeal against a notice of liability served in accordance with that legislation. As I understand it, the relevant parts of the CFC GLO in the High Court are still in abeyance while the implications of the decision of the ECJ in Cadbury Schweppes are being worked out.

As to the compatibility of the UK dividend provisions with EU law, most of the relevant ground has in the event already been covered by other references to the ECJ in the Franked Investment Income (“FII”) and Thin Capitalisation (“Thin Cap”) Group Litigation. The latter of these references led on 13 March 2007 to the judgment of the ECJ in Thin Cap which I have already mentioned. The former led to its judgment, some three months earlier on 12 December 2006, in Test Claimants in the FII Group Litigation v IRC (Case C-446/04) [2006] ECR I-11753, [2007] STC 326 (“FII”). Yet further issues were referred in Class IV of the ACT Group Litigation, which led to the judgment of the ECJ, also on 12 December 2006, in Test Claimants in Class IV of the ACT Group Litigation v IRC (Case C-374/04) [2006] ECR I-11673, [2007] STC 404 (“ACT Class IV”).

Before any of those judgments had been given, an order for reference in the present group litigation was made by Park J on 18 March 2005. Leaving aside the questions relating specifically to the CFC legislation, the questions referred were as follows.

Question 1 asked whether it was contrary to Articles 43 or 56 EC for a member state to keep in force and apply measures which:

“(a)

exempt from corporation tax dividends received by a company resident in that Member State (“the resident company”) from other resident companies; but which

(b)

subject to corporation tax dividends received by the resident company from a company resident in another Member State and in particular a company controlled by it resident in another Member State and subject to a lower level of taxation there (“the controlled company”), after giving double taxation relief for any withholding tax payable on the dividend and for the underlying tax paid by the controlled company on its profits?”

Although asked with particular reference to controlled companies, this question also raised in general terms the issue whether it was lawful for a member state (here the UK) to exempt from corporation tax dividends received from a resident company (the relevant exemption was at all material times contained in section 208 of the Income and Corporation Taxes Act 1988 (“ICTA 1988”)), but to subject to corporation tax dividends received from non-resident companies (under section 18 of ICTA 1988 and Case V of Schedule D). The question appears to have assumed that relief would be given against such tax not only for any withholding tax payable on the dividend, but also for the underlying tax paid by the non-resident company on the relevant part of its profits. However, this was an over-simplification, and the ECJ correctly dealt with the question on the footing that, where the holding was a portfolio holding of less than 10%, the only relief afforded by the UK legislation was for withholding tax paid on the dividend. No relief was given for any underlying corporation tax paid by the company making the distribution.

Question 2 asked whether Articles 43, 49 or 56 EC precluded national tax legislation such as that in issue in the main proceedings under which, before 1 July 1997:

“(a)

certain dividends received by an insurance company resident in a Member State from a company resident in another Member State (“the non-resident company”) were chargeable to corporation tax; but

(b)

the resident insurance company was allowed to elect that corresponding dividends received from a company resident in the same Member State should not be chargeable to corporation tax, with the further consequence that a company which had made the election was unable to claim payment of the tax credit to which it would otherwise have been entitled?”

Translated into UK terms, this question was concerned with the validity of the right of election contained in section 438(6) of ICTA 1988. I will not at this stage attempt to explain the nature of the right, beyond noting that it applied in the computation of the profits of pension business carried on by an insurance company, and enabled the company to elect to exclude FII from the computation, and thus to reduce the pension business profits chargeable to corporation tax, at the price of foregoing the tax credits that the company would otherwise have been entitled to claim and have paid to it if the FII were included in the computation. In all normal circumstances, it was advantageous to the company to make such an election, because the rate of corporation tax thereby avoided was higher than the rate of tax reflected in the tax credits foregone. However, by virtue of its confinement to FII, which was a product of the UK’s partial imputation system of dividend taxation and did not apply to foreign dividends, the election could not be made in respect of foreign dividend income, whether from the EU or third countries. Section 438(6) was repealed by the Finance (No. 2) Act 1997, which received the Royal Assent on 31 July 1997, in relation to accounting periods beginning on or after 2 July 1997.

Question 4 asked whether the answers to questions 1 or 2 would be different if the controlled company (in question 1) or the non-resident company (in question 2) was resident in a third country.

Questions 6 to 12 asked questions about classification of the claims under EU law and remedies in broadly similar terms to the questions raised in FII, Thin Cap and ACT Class IV.

Since the reference raised questions similar to those referred for a preliminary ruling in Cadbury Schweppes, FII, Thin Cap and ACT Class IV, the proceedings were stayed pending the decision of the ECJ in those cases. On 3 April 2007, after the decisions in those cases had been given, the ECJ asked by letter whether the High Court still wished to maintain its reference for a preliminary ruling. By its reply dated 12 June 2007, the High Court informed the ECJ that it was maintaining its reference.

Against this background, the ECJ decided, pursuant to Article 104(3) of its Rules of Procedure, to dispose of the reference by means of a reasoned order. So far as material, Article 104(3) provides as follows:

“Where a question referred to the Court for a preliminary ruling is identical to a question on which the Court has already ruled, or where the answer to such a question may be clearly deduced from existing case-law, the Court may, after hearing the Advocate General, at any time give its decision by reasoned order in which reference is made to its previous judgment or to the relevant case-law. The Court may also give its decision by reasoned order, after informing the court or tribunal which referred the question to it, hearing any observations submitted by the persons referred to in Article 23 of the Statute and after hearing the Advocate General, where the answer to the question referred to the Court for a preliminary ruling admits of no reasonable doubt.”

In the present case the Court followed the former of these procedures, as paragraph 34 of its reasoned order makes clear. The order was made by the Fourth Chamber of the ECJ, presided over by Judge Lenaerts who was also the Rapporteur. He was very well versed in the issues, having been the Juge Rapporteur in all four of the previous cases to which I have referred. The Advocate General (Trstenjak) had not been involved in any of the earlier cases. In accordance with the reasoned order procedure he did not deliver an opinion, but the Court still had the benefit of hearing him. The reasoned order is reported as a Note at [2008] STC 1513, and I will refer to it as “the Reasoned Order”.

In paragraphs 6 to 10 of the Reasoned Order, the ECJ set out its understanding of the relevant UK legislation relating to the taxation of dividends:

“6.

Under section 208 of ICTA, where a United Kingdom-resident company receives dividends from a company that is also resident in that Member State, it is not liable to corporation tax in respect of those dividends.

7.

When a United Kingdom-resident company receives dividends from a company resident outside the United Kingdom, it is liable to corporation tax on those dividends. In such a case, the company receiving those dividends is not entitled to a tax credit and the dividends paid do not qualify as franked investment income. However, under sections 788 and 790 of ICTA, it is entitled to relief for tax paid by the company making the distribution in the State in which the latter is resident. Such relief is granted either under the legislation in force in the United Kingdom or under a double taxation convention (“DTC”) concluded by the United Kingdom with the other State.

8.

Thus, the national legislation allows withholding taxes paid on dividends from a non-resident company to be offset against the liability to corporation tax of a resident company receiving the dividends. Where a resident company receiving dividends either directly or indirectly controls, or is a subsidiary of, a company which directly or indirectly controls not less than 10% of the voting power in the company making the distribution, the relief extends to the underlying foreign corporation tax on the profits out of which the dividends are paid. Relief on that foreign tax is available only up to the amount due in the United Kingdom by way of corporation tax on the income concerned.

9.

Certain specific provisions concern the taxation of investment income, particularly dividends, received by insurance companies on assets allocated to pensions business and life assurance business.

10.

Section 208 of ICTA does not apply, as a rule, to either pensions business or overseas life assurance business, with the result that dividends from portfolio investments linked to such business are subject to United Kingdom tax calculated in accordance with the principles applicable to the computation of trading profits arising from underwriting. However, prior to 1 July 1997, a life insurance company could, as an exception to that principle, elect that that section should apply as regards dividends received from residents, as part of its pensions business. If it did so elect it could not claim payment of the tax credits attached to dividends. Such an election was not, on the other hand, possible as regards dividends received from non-resident companies, as part of such business.”

The Reasoned Order then referred to the UK legislation on CFCs, set out the background of the present group litigation, and quoted the questions in the order for reference. The Court then considered the questions in turn. It answered the first question exclusively by reference to its decision in FII. The steps in its reasoning may be summarised as follows. First, the Court repeated its key holding of principle that Community law does not in principle prohibit a member state from avoiding the imposition of a series of charges to tax on dividends received by a resident company by exempting those dividends from tax when they are paid by a resident company, but applying an imputation system when they are paid by a non-resident company: see FII, paragraph 48 of the judgment. In other words, there is no objection in principle to a dual exemption and imputation system such as that which the UK operated until 1997. Secondly, there is no breach of Article 43 where the shareholding in question is sufficiently large to engage that Article, “provided that the tax rate applied to foreign-sourced dividends is not higher than the rate applied to nationally sourced dividends and that the tax credit is at least equal to the amount paid in the Member State of the company making the distribution, up to the limit of the tax charged in the Member State of the company receiving the dividend”: see FII, paragraph 57 of the judgment. Thirdly, with respect to dividends paid by a company in which the resident shareholder holds 10% or more of the voting rights, but the holding is not sufficiently large to engage Article 43, there is again no breach of Article 56, provided that the same conditions about the rate of tax and the amount of the tax credit are satisfied: see FII, paragraph 60 of the judgment.

The Court then dealt with portfolio holdings as follows:

“40.

As regards, thirdly and lastly, resident companies which received dividends from companies in which they hold fewer than 10% of the voting rights, after noting that nationally- sourced dividends are exempt from corporation tax, whilst foreign-sourced dividends are subject to that tax and are entitled to relief only as regards any withholding tax charged on those dividends in the state in which the company making the distribution is resident (Test Claimants in the FII Group Litigation, paragraph 61), the Court held that the difference in treatment arising from legislation such as that at issue in the main proceedings as regards dividends received by resident companies from non-resident companies in which they hold fewer than 10% of the voting rights constitutes a restriction on the free movement of capital which is, in principle, prohibited by Article 56 EC (Test Claimants in the FII Group Litigation, paragraph 65).

41.

The Court went on to hold that the mere fact that it is for a Member State to determine for such holdings whether, and to what extent, the imposition of a series of charges to tax on distributed profits is to be avoided does not of itself mean that it may operate a system under which foreign-sourced dividends and nationally-sourced dividends are not treated in the same way (Test Claimants in the FII Group Litigation, paragraph 69), and that, irrespective of the fact that a Member State may, in any event, choose between a number of systems in order to prevent or mitigate the imposition of a series of charges to tax on distributed profits, the difficulties that may arise in determining the tax actually paid in another Member State cannot justify a restriction on the free movement of capital such as that which arises under the legislation at issue in the main proceedings (Test Claimants in the FII Group Litigation, paragraph 70).

42.

Therefore, the Court held that Article 56 EC precludes legislation of a Member State which exempts from corporation tax dividends which a resident company receives from another resident company, where that State levies corporation tax on dividends which a resident company receives from a non-resident company in which it holds less than 10% of the voting rights, without granting the company receiving the dividends a tax credit for the tax actually paid by the company making the distribution in the State in which the latter is resident (Test Claimants in the FII Group Litigation, paragraph 74).”

It will be seen, therefore, that the ECJ held that the UK system of taxation of portfolio dividends infringed Article 56 because the resident company in receipt of the dividend was not granted “a tax credit for the tax actually paid by the company making the distribution”. In other words, the limitation of the tax credit, in the case of portfolio dividends, to withholding tax alone could not be justified. Had a credit for the underlying tax been allowed, portfolio dividends would then have been in the same position as the second category of dividends considered by the Court, so presumably the conditions laid down by the Court in dealing with that category would also have had to be satisfied.

The ECJ addressed the second question in paragraphs 44 to 69 of the Reasoned Order. It began by holding that only Article 56 was engaged, because the dividends in question were all portfolio dividends (paragraphs 45 to 52). The Court then described the way in which the election under section 438(6) of ICTA 1988 operated, in substantially the same general terms as it had already employed in paragraph 10 (see above), and continued as follows:

“55.

Such a system would be contrary to Article 56 EC, if dividends paid by companies established in another Member State to insurance companies established in the United Kingdom were treated, for tax purposes, less favourably than those paid by companies established in the United Kingdom (see, to that effect, Verkooijen, cited above, paragraphs 34 to 38, and Test Claimants in the FII Group Litigation, paragraph 64).

56.

In that regard, it is not apparent from the decision making the reference whether, in light of the fact that the permitted election, as regards dividends of national origin, entailed the waiver of tax credits, a company receiving dividends of foreign origin, which could not exercise such an election, was treated less favourably because of that fact alone.

57.

It is for the national court to establish whether such was the case.

58.

On the other hand, since it is clear from the decision making the reference that companies with a less than 10% shareholding in the company making the distribution were not entitled to relief in respect of corporation tax paid by that company in the Member State in which it was resident, those companies were, contrary to Article 56 EC, subject to less favourable tax treatment.”

Thus the Court repeated its well-established governing principle that there would be a breach of Article 56 if the effect of the legislation was to treat dividends paid by a company established in another member state less favourably than dividends paid by a UK company. In view of the fact that the making of an election under section 438(6) would involve the foregoing of tax credits, it was left to the national court to determine whether the confinement of the election to UK dividends did in fact entail less favourable treatment. The Court also pointed out that there was in any event a breach of Article 56 by reason of the UK’s failure to allow a tax credit for underlying corporation tax paid by the company making the distribution.

In the remainder of its discussion of question 2 the Court rejected various defences relied upon by the UK government, and in paragraph 69 it gave its answer to the question as follows:

“69.

Accordingly, the reply to the second question must be that Article 56 EC is to be interpreted as meaning that it precludes legislation of a Member State which allows an exemption from corporation tax for certain dividends received from resident companies by resident insurance companies but excludes such an exemption for similar dividends received from non-resident companies, in so far as it entails less favourable treatment of the latter dividends.”

This was, of course, an answer to the Court’s reformulated version of question 2 in paragraph 44, rather than an answer to the original question with its specific focus on a right of election such as that in section 438(6). However, the Court’s answer to the specific question can be found in paragraphs 55 to 58, quoted above.

The Court dealt with the fourth question in paragraphs 87 to 97 of the Reasoned Order. The Court began by referring to recent authority which established that third country issues arose only in relation to Article 56, which (together with Articles 57 and 58) could be relied on before national courts in relation to the movement of capital between member and non-member states (paragraphs 88 to 91). The Court then continued as follows (the references to “A” are to Skatteverket v A (Case C-101/05) [2007] ECR I-11531):

“92.

Admittedly, it is clear from the case law of the Court that the extent to which the Member States are authorised to apply certain restrictive measures on the movement of capital cannot be determined without taking account of the fact that movement of capital to or from third countries takes place in a different legal context from that which occurs within the Community. Accordingly, because of the degree of legal integration that exists between Community Member States, in particular by reason of the presence of community legislation which seeks to ensure co-operation between national tax authorities … the taxation by a Member State of economic activities having cross-border aspects which take place within the Community is not always comparable to that of economic activities involving relations between Member States and third countries (Test Claimants in the FII Group Litigation, paragraph 170).

93.

It may also be that a Member State will be able to demonstrate that a restriction on the movement of capital to or from third countries is justified for a particular reason in circumstances where that reason would not constitute a valid justification for a restriction on capital movements between Member States (A, paragraphs 36 and 37).

94.

As regards the grounds advanced by the United Kingdom government to justify the national measures to which the first and second questions refer, particularly the need to ensure the cohesion of the tax system, that Government put forward no evidence explaining how those grounds justify those measures in relations between a Member State and non-member countries.

95.

Moreover, as regards the difficulties associated with the verification of compliance with certain requirements by companies established in third countries, the Court decided, in the context of the free movement of capital, that, where the legislation of a Member State makes the grant of a tax advantage dependent on satisfying requirements, compliance with which can be verified only by obtaining information from the competent authorities of a third country, it is, in principle, legitimate for that Member State to refuse to grant that advantage if, in particular, because that third country is not under any contractual obligation to provide information, it proves impossible to obtain such information from that country (A, paragraph 63).

96.

Therefore, it follows from that judgment that Articles 56 EC to 58 EC are to be interpreted as not precluding the legislation of a Member State under which a tax concession in respect of dividends may be granted only if the distributing company is established in a State within the European Economic Area or a State with which a taxation convention providing for the exchange of information has been concluded by the Member State imposing the tax, where that concession is subject to conditions compliance with which can be verified by the competent authorities of that Member State only by obtaining information from the State of establishment of the distributing company (see, to that effect, A, paragraph 67).

97.

Having regard to the foregoing considerations, the reply to the fourth question referred for a preliminary ruling must be that Articles 56 EC to 58 EC are to be interpreted as not precluding the legislation of a Member State which grants a corporation tax concession in respect of certain dividends received from resident companies by resident companies but excludes such a concession for dividends received from companies established in a non-member country particularly where the grant of that concession is subject to conditions compliance with which can be verified by the competent authorities of that Member State only by obtaining information from the non-member country where the distributing company is established.”

In general terms, the thrust of this passage appears to be that a restriction on the movement of capital to or from a third country may be justified, even though an equivalent restriction within the EU would infringe Article 56, if to grant the same treatment would entail the satisfaction of requirements which could be verified only by obtaining information from the competent authorities of the third country, and where no right to obtain such information was conferred under the relevant double taxation convention. With particular reference to the breach of Article 56 identified by the Court in its answer to question 1, questions would thus arise whether the UK Revenue was able to obtain information under a DTC concluded with a third country about underlying corporation tax paid by a company resident in that country which made a portfolio distribution. If the UK Revenue had no contractual right to obtain such information, the Court’s reasoning appears to lead to the conclusion that a refusal to grant a credit for such tax would be justified in respect of dividends paid from the relevant country.

The Court took questions 6 to 12 in the order for reference together, and answered them in terms very similar to the answers which it had already given in FII and Thin Cap. There are no specific points in these answers to which I need to draw attention at this stage.

After the ECJ had made the Reasoned Order, the Prudential test claims proceeded towards trial in the usual way. The Reasoned Order had answered a number of questions, but it had also raised new issues (for example in relation to third countries), it had remitted at least one question to the national court (see paragraph 57), and a host of difficult questions remained to be resolved in working out the implications in domestic law and procedure of the breaches of Article 56 (both actual and potential) which the ECJ had identified.

In May 2008 the Revenue formally conceded, in the case of one of the participants in the present GLO 9General Reinsurance UK Limited) that liability had been established in relation to portfolio dividends received from EU-resident companies, and consented to summary judgment in terms equivalent to paragraph 42 of the Reasoned Order and paragraph 74 of the judgment in FII.

In March 2009 a case management conference took place before me, at which directions were given for the service of amended statements of case, disclosure, exchange of witness statements, and further preparations for the trial which had already been fixed to begin in mid-November 2009.

These directions were substantially complied with, but the claimants also began to give thought to widening the scope of the issues to be determined at the forthcoming trial. The new issues which they wished to introduce related to payments of advance corporation tax (“ACT”) which had been made by the test claimants as a result of the onward distribution of profits arising out of dividends received from non-UK portfolio holdings. In addition, they wished to contend that they were entitled to treat their non-UK portfolio dividends as FII, on the footing that this would be the most appropriate way to remedy the UK’s breach of Article 56. If the latter contention were correct, further technical issues would then arise about the computation of the claimants’ taxable profits. These further issues were peculiar to life insurance companies, which are subject to rules of notorious complexity and difficulty, but the basic point was that an increase in the amount of FII would lead, in one way or another, to a reduced overall tax liability.

Over the course of the summer the claimants, through their solicitors, proposed amendments to the GLO issues and sought to persuade the Revenue that the new questions relating to ACT and FII could not sensibly be divorced from the forthcoming hearing. In early October 2009 they submitted a draft of proposed amendments to the Prudential test claim, and on 19 October the amended particulars were formally sealed and served pursuant to a procedure laid down in an order of Mr Justice Park dated 12 December 2003. The Revenue maintained their stance that these new issues should not be dealt with at the forthcoming trial, and by an application notice dated 2 November 2009 they sought directions about what issues were to be tried at the hearing. I heard this application on 9 November and decided, with some hesitation, that the least unsatisfactory course was to permit the claimants to raise and pursue the new issues at the forthcoming trial, and not to adjourn any part of the proceedings. The new issues were essentially questions of law; it was agreed that there would be ample to time to deal with them at the hearing, for which up to five days had been allotted; and in view of my own commitments and those of counsel it was far from clear when an adjourned hearing could conveniently be arranged.

A further complicating factor, which I have not yet mentioned, was that the Court of Appeal had in October heard argument for nine days on the appeal and cross-appeal against my lengthy decision in the FII Group Litigation, following the decision of the ECJ in FII: see Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners, [2008] EWHC 2893 (Ch), [2009] STC 254 (“FII (Chancery)”). Everybody recognised that the decision of the Court of Appeal in this case would be highly material to many of the issues in the present case, including in particular the new issues relating to FII and ACT. With the benefit of hindsight, it would probably have been preferable if I had decided to adjourn the present trial at least until the Court of Appeal had given judgment. However, that is not the course which I followed, and instead I indicated that if the Court of Appeal gave its judgment before I had delivered my judgment after the hearing, the parties would be at liberty to apply to restore the matter for further argument in the light of the Court of Appeal’s judgment.

In the event, that is what then happened. I heard argument on all the issues for a day and a half on 18 and 19 November, and then reserved judgment. A few weeks earlier I had also heard argument on another GLO case, in the ACT Class IV Group Litigation, and that was the case to which I gave priority when I had time to begin work on a judgment: see Test Claimants in the ACT Group Litigation (Classes 2 and 4) v Revenue and Customs Commissioners [2010] EWHC 359 (Ch), [2010] STC 1078. Thus it came about that I had still not begun work on a judgment in the present case when the Court of Appeal released to the parties its draft judgment in the FII case. The judgment was then formally delivered on 23 February 2010: see Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners [2010] EWCA Civ 103, [2010] STC 1251 (“FII (CA)”).

In a number of important respects the Court of Appeal disagreed with my reasoning and conclusions in FII (Chancery). In the next section of this judgment I will briefly describe the main areas of such disagreement. In consequence, it is no surprise that both sides wished to re-open the hearing in the present case, and in due course a further hearing took place on 20 and 21 May 2010. By this stage the Revenue’s primary submission, advanced by Mr David Ewart QC, was that the present proceedings ought to be adjourned generally pending any appeal to the Supreme Court in the FII litigation. The claimants, however, submitted that I should hear argument on the further matters arising out of the judgment of the Court of Appeal, and that I should then proceed to give judgment in the usual way. Since both sides had come prepared to argue the further points arising from the Court of Appeal’s judgment, and since I took the view that the adjournment issue was a difficult one which needed to be resolved in the context of the argument as a whole, I decided not to treat it as a preliminary issue but rather to hear argument on all the points which the parties wished to raise. I would then decide, with the benefit of that argument, whether to accede to the Revenue’s primary application, or whether to proceed and give judgment on some or all of the contested issues. The argument again took no more than a day and a half, although it covered a lot of ground, and at the conclusion of the adjourned hearing I again reserved judgment.

There are at least five areas in which the Court of Appeal has taken a significantly different view from myself, although as will be seen the last three of them are connected. As matters now stand, therefore, the law on these matters is the law as stated by the Court of Appeal, and the contrary views which I expressed in FII (Chancery), although they briefly held the field at the date of the first hearing in the present case last November, have now been overruled.

The lawfulness of the Case V charge

In FII (Chancery) I held that the charge under Case V of Schedule D on dividends received by UK-resident parent companies from their EU resident subsidiaries infringed Article 43 (paragraphs 39 to 66). The issue turned on the correct analysis of the reasoning of the ECJ and the answer to a single question, framed in potentially ambiguous terms, which the ECJ had remitted to the national court for determination. It was common ground that if, as I held, there was an infringement of Article 43, there would likewise be an infringement of Article 56, for the same reasons, in relation to portfolio dividends, provided that the factual situation were one where Article 56 was capable of applying (paragraphs 67 to 73).

The Court of Appeal was divided on the question of the correct analysis of the ECJ’s reasoning: Etherton LJ agreed with my approach, but the majority (Arden and Stanley Burnton LJJ) preferred a different analysis which would lead to the conclusion that there was no breach of Article 43 (see FII (CA) at paragraphs 25 to 43, and Annex 3 to the judgment). All three members of the Court agreed, however, that the matter was not acte clair, and that it should be referred back to the ECJ so that its decision on this critical issue could be clarified. The Court did not make the reference itself, but said that the reference would most appropriately be made by the High Court (paragraph 270). No application has yet been made to me for that purpose, presumably because the Court directed (as foreshadowed in paragraph 270) that a reference should be delayed until the Supreme Court had dealt with the applications for permission to appeal which both sides have now made and until any subsequent appeal had been determined.

The present position, therefore, is that the key issue of the lawfulness of the Case V charge remains open, pending either a decision of the Supreme Court or a further decision of the ECJ.

Conforming interpretation

In FII (Chancery) it was common ground that it was impossible to construe the domestic ACT provisions in a way which achieved conformity with Article 43 (paragraph 144 of my judgment), so the focus of the argument was on disapplication of the offending provisions. For the reasons which I gave in paragraphs 145 to 153, I concluded that the simplest and most appropriate remedy would be to remove the UK territorial limitation on FII in section 231(1) of ICTA 1988, in its application to the relevant foreign dividends, so that such dividends would then generate FII in the hands of the parent company in the same way as domestic dividends. I rejected the Revenue’s argument that the appropriate remedy would simply be to grant a credit for foreign corporation tax which had actually been paid by the company making the distribution, mainly because this seemed to me to run against the legislative scheme of ACT, but also because of the practical difficulties that the claimants would inevitably encounter in obtaining the necessary information over a period of up to 25 years, thus involving a possible breach of the Community law principle of effectiveness.

The Court of Appeal dealt with these questions in FII (CA) at paragraphs 97 to 109. They seem, with the greatest respect, to have proceeded under the misapprehension that the arguments advanced to me, and my decision, had been based on the doctrine of conforming construction rather than disapplication (see paragraphs 100 to 103). In any event, they held that a conforming construction was possible, and that it could be achieved (paragraph 107):

“simply by reading in words that make it clear that it is not just resident companies that can claim a credit under section 231 but also other persons entitled to do so by Community law to the extent that they are so entitled.”

The Court said that the extent of this entitlement should be investigated when the section falls to be applied. Issues of effectiveness and evidential difficulties were irrelevant to the question of interpretation, and should be considered only at the time when the legislation came to be applied (paragraphs 105 and 107).

I would add that I returned to the issues of conforming interpretation and disapplication, and dealt with them in rather more detail, in the context of the Thin Cap Group Litigation. I did this after the Court of Appeal had delivered its important judgment in Vodafone 2 v Revenue and Customs Commissioners [2009] EWCA (Civ) 446, [2010] BLR 96, [2009] STC 1480, but before the decision of the Court of Appeal in FII (CA): see Test Claimants in the Thin Cap Group Litigation v Revenue and Customs Commissioners [2009] EWHC 2908 (Ch), [2010] STC 301, (“Thin Cap (Chancery)” at paragraphs 76 to 99). My decision in Thin Cap (Chancery) has itself been the subject of an appeal to the Court of Appeal, on which the Court has recently heard argument (in October 2010) and judgment is awaited.

Restitutionary remedies

These issues include some questions of great interest and difficulty, at the frontiers of the law of restitution. I dealt with the general restitutionary issues that arose in paragraphs 192 to 276 of my judgment in FII (Chancery). For present purposes, and at the risk of over-simplification, the most important points were:

The Court of Appeal disagreed, and held:

The Court then continued:

“174.

If, however, contrary to our conclusion, 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 San Giorgio principle requires our domestic law to provide an effective remedy for the repayment of unlawful tax. Neither the domestic Woolwich cause of action (if it is dependent on there having been an unlawful demand) nor the cause of action for monies paid under a mistake satisfy San Giorgio, the latter cause of action because it depends on the claimant having been mistaken when the payment was made. In these circumstances, the cause of action for the repayment of a tax unlawful under Community law must be held not to require an unlawful demand. In effect, that is Woolwich without the requirement of an unlawful demand.”

I would again add that I had returned to some of these themes in Thin Cap (Chancery): see paragraphs 196 to 236. I would draw attention in particular to what I said in the context of Mr Ewart’s submission that the Woolwich principle was all that was needed to satisfy the claimants’ San Giorgio claims:

“223.

These submissions were attractively advanced, but I do not find them convincing. In the first place, it is in my view wrong in principle to treat the Community law principle of effectiveness as though it were a limiting one, and to look for the minimum remedies in UK law that are needed to satisfy it. The true principle is, rather, that it stipulates a minimum content that UK law must provide if San Giorgio claims are to be satisfied. If, however, English law provides two alternative causes of action to which a claimant may have recourse, it is no part of the principle of effectiveness to say that only the more restrictive of those causes of action is needed. Thus I do not see how it would avail the Revenue to establish that the Woolwich cause of action would, by itself, or with appropriate extensions, be sufficient to satisfy the claimants’ San Giorgio claims, when they also have the alternative open to them of making a mistake-based restitution claim. On the contrary, my understanding is that where domestic law goes further than is strictly necessary to give effect to Community law rights, Community law requires the full range of domestic remedies to be made available, and not to be curtailed without due notice.

224.

Secondly, the principle that a claimant is normally entitled to choose between alternative remedies is one that is well-recognised in English law …”

These views are, I think, incompatible with the approach adopted by the Court of Appeal in FII (CA), but the Court of Appeal has not yet explained why they are wrong.

Time limits and limitation

These are questions of great practical importance, in view of the extended limitation period for mistake-based claims afforded by section 32(1)(c) of the Limitation Act 1980. As I explained in FII (Chancery) at paragraphs 405 to 413, Parliament sought to disapply section 32(1)(c) in relation to mistakes of law relating to taxation matters under the care and management of the Revenue, first by enacting section 320 of the Finance Act 2004 (which applied prospectively to actions brought on or after 8 September 2003) and then by enacting section 107 of the Finance Act 2007 (which extended the disapplication retrospectively to any action brought before 8 September 2003 for relief from the consequences of such a mistake of law). Neither enactment contained any transitional provisions. The effect of the legislation, if valid, was to confine the period for claiming repayment of overpaid tax to 6 years from the date of the original payment, even in cases where the payment had been made under a mistake of law which the claimant did not discover (or could not with reasonable diligence have discovered) until a later date.

On the strength of my conclusion that a mistake-based cause of action in restitution was needed in order to satisfy the claimants’ San Giorgio claims, I held that sections 320 and 107 breached EU law in so far as they curtailed the limitation period applicable to such claims without providing any transitional arrangements (paragraphs 414 to 427). However, I went on to say (in paragraph 429) that if the Revenue were to succeed in persuading a higher court that the Woolwich cause of action alone was sufficient to satisfy the claimants’ rights to recover overpaid tax under EU law, I was unable to see any answer to Mr Ewart’s submission that the sections would be unaffected by EU law in the context of that case.

The Court of Appeal, having held that the Woolwich cause of action alone was sufficient, agreed (in effect) with what I had said in paragraph 429 and held that there was nothing in EU law which prevented the application of the two sections to mistake claims (paragraphs 217 to 229). The reason for this, as they explained in paragraph 229, was

“because English law provides a good San Giorgio remedy in the form of the Woolwich cause of action, which is not affected by sections 320 and 107.”

Section 33 of the Taxes Management Act 1970

The fundamental issue here is whether the “error or mistake” provisions contained in section 33 of the Taxes Management Act 1970, and (since 1998) in its corporate equivalent (paragraph 51 of Schedule 18 to the Finance Act 1998), are compatible with EU law. Since the decision of the Court of Appeal in Monro v Revenue and Customs Commissioners [2008] EWCA Civ 306, [2009] Ch 69, it has been clear that section 33 provides an exclusive remedy for repayment of overpaid tax to which it applies, and that the section cannot be circumvented by framing a claim in restitution under the Woolwich principle. However, as a matter of general principle section 33 must be read and applied subject to the overriding requirements of EU law, including the principle of effectiveness.

In FII (Chancery) I held that section 33 did not apply to claims to recover ACT, because the section applied only to tax paid under an assessment or self-assessment (paragraph 436). I also held that, where the tax in question was paid under an assessment, the section must yield to the EU principle of effectiveness (paragraph 439), by which I meant that it should be disapplied to the necessary extent where the claimant was seeking to enforce a San Giorgio claim. To the best of my recollection, neither side submitted to me that section 33 could be construed compatibly with EU law in relation to a San Giorgio claim. Mr Ewart’s main argument was, rather, that mistake-based restitution claims were not required by the San Giorgio principle, with the consequence that nothing prevented section 33 from applying to them in the usual way. Had the issue of construction been raised before me, I would probably have said that in my view the section could not be construed and applied as it stood to a mistake-based San Giorgio claim, because the section expressly precluded the grant of any relief where the relevant tax return (or self-assessment) had been made “on the basis or in accordance with the practice generally prevailing at the time when the return was made”. The principle of effectiveness would have required this proviso to be disapplied in relation to a San Giorgio claim.

The Court of Appeal agreed that section 33 did not apply to ACT claims (FII (CA) at paragraph 250). However, the Revenue had by now submitted, and the Court of Appeal accepted, that section 33 could be interpreted in accordance with the Marleasing principle in such a way as to be compatible with EU law (paragraphs 256 to 268). Relying on the presumption that the United Kingdom intended to comply with its obligations under the EC Treaty, and on the breadth of the Marleasing principle, the Court held that the “settled basis” proviso should be read “as subject to the limitation that it applies only if and to the extent that the United Kingdom can consistently with its Treaty obligations impose such a restriction” (paragraph 261). The Court went on to consider whether such a conforming interpretation would go against the grain of the legislation, and concluded that it would not.

An important consequence of the Court of Appeal’s decision on this question is that all claims for the recovery of corporation tax paid by mistake must be brought in accordance with the statutory procedure laid down in section 33, and not by means of a claim in the High Court: see FII (CA) at paragraph 264 and the principles expounded by the majority of the House of Lords in Autologic Holdings Plc v IRC [2005] UKHL 54, [2006] 1 AC 118 (“Autologic”). As the Court said in paragraph 264:

“The effect of our interpretation of section 33 is consistent with [Autologic] since it means that claims for repayment on the grounds of mistake must also be brought in accordance with the statutory scheme for determining disputes with respect to other claims for the repayments of tax.”

Submissions

Against this background, the Revenue’s arguments in favour of adjourning the present trial are advanced on both procedural and substantive grounds.

The procedural grounds, shortly stated, are that in the light of the judgment in FII (CA) all of the claims in the present action are either time-barred or are claims which in accordance with Autologic fall within the exclusive jurisdiction of the Tax Chamber of the First-tier Tribunal (formerly the Special Commissioners). In more detail, the Revenue submit:

In these circumstances, submits Mr Ewart, the court would be obliged to dismiss all of the claims if the hearing proceeded, and any views which the court expressed on the substantive issues would inevitably be obiter and unnecessary to the disposal of the case. At first blush, it might be thought that an immediate dismissal of all the claims could only benefit the Revenue. Mr Ewart realistically accepted, however, that the claimants would inevitably appeal, with the result that in practice all of the issues (both procedural and substantive) would then have to be dealt with by the Court of Appeal, which would find itself confronted with essentially the same case management problem as I am. Far better, says Mr Ewart, for me to grapple with the case management issues myself, and defer any further hearing of the present trial at least until the likely appeal to the Supreme Court in FII has run its course.

Turning to the substantive grounds, Mr Ewart’s main point is that the general issue of the compatibility of the Case V charge with EU law is still open, following the decision of the Court of Appeal to refer it back to the ECJ. It is true that the Case V charge on portfolio dividends has already been held unlawful by the ECJ in the Reasoned Order, on the specific ground that no credit was allowed for underlying corporation tax paid by the company making the distribution. Nevertheless, it is still important to know whether this is the only defect in the UK legislation, or whether the charge is unlawful for other reasons as well. Until this critical question has been answered, it is impossible to reach any concluded view on the questions whether a conforming interpretation of the UK legislation is in principle possible; if so, what form it should take; and if not, how and to what extent the offending provisions should be disapplied.

For similar reasons, submits Mr Ewart, the technical ACT and FII-related claims could only be dealt with at this stage on a provisional basis, because their definitive resolution will inevitably be dependent on the ultimate outcome of the FII litigation.

For the claimants Mr Aaronson QC and Mr Cavender QC, each of whom made submissions on this question, do not dispute that, on the basis of the judgment of the Court of Appeal in FII (CA), nearly all of the claims in the present action are either time-barred or must be pursued in the Tribunal. They submit, however, that one category of claims, of significant monetary value, is still in time. These are claims which relate to the use of lawful ACT to discharge unlawful mainstream corporation tax (“MCT”). Mr Cavender submits that claims of this nature were adequately pleaded in early amendments to the particulars of claim which were made on 2 September 2003. He accepts that there is no explicit reference in this pleading to a claim to recover lawful ACT used to discharge unlawful MCT, but says that such a claim is subsumed in the claim which was then made for restitution of MCT paid pursuant to the unlawful Case V charge. To the extent that the MCT liability was discharged by the statutory set-off of ACT which had previously been lawfully paid, the ACT has to be treated in the eyes of EU law as a pre-payment of the relevant part of the MCT, and the Court of Appeal has now expressly held that, as such, it can be recovered in a Woolwich claim: see FII(CA) at paragraphs 148 and 151. As Mr Cavender’s vivid but perhaps dubious addition to the legal lexicon would have it, the ACT at that point “morphed” into unlawful MCT and became recoverable accordingly. If the foregoing submissions are correct, the claimants have now produced figures which show that MCT of approximately £25.5 million due and payable by them in December 1996, 1997 and 1998 was discharged by set-off of ACT, less than six years before the amended pleadings of 2 September 2003.

It is convenient at this point to note Mr Ewart’s answer to these submissions. There are two separate reasons, he says, why the claimants’ analysis cannot be correct. First, on the assumption that the MCT was unlawfully levied, no statutory set-off of ACT can have occurred, because the charge was an unlawful one and therefore a fiscal nullity. The right analysis, he submits, is that nothing of any legal significance happened at that stage, and the ACT simply remained unutilised ACT.

Secondly, if the first submission is wrong, and there was a real set-off, the result is only that part of the liability to MCT was paid by means of the set-off. This follows from section 239(1) of ICTA 1988, which provides that:

“… [ACT] paid by a company (and not repaid) in respect of any distribution made by it in an accounting period shall be set against its liability to corporation tax on any profits charged to corporation tax for that accounting period and shall accordingly discharge a corresponding amount of that liability.”

Accordingly, submits Mr Ewart, what was paid was all MCT, and the ACT did not “morph” into anything different. It simply discharged part of the claimants’ MCT liability. Thus their claims can only be for recovery of MCT, and as such they fall within the scope of section 33 and must be pursued, if at all, in the Tribunal.

In reply, Mr Cavender submitted that the holding of the Court of Appeal on this point was clear and unambiguous, and it was my duty to apply it. In paragraph 148 the Court of Appeal said this:

“We consider that the principles in paragraphs 205 to 207 of the ECJ’s judgment are clear. Paragraph 205 holds that individuals are entitled to reimbursement of the unlawfully levied tax and any other amounts paid to or retained by the member state which relate directly to that tax. That will include interest, under the Hoechst principle, and penalties. It also will include, in our judgment, ACT set against the unlawful Case V charge since the jurisprudence of the ECJ treats ACT as an advance payment of MCT.”

Similarly, in paragraph 151 the Court stated its conclusion that:

“… the claim in respect of ACT set against the unlawful Case V charge is also a San Giorgio claim.”

This language leaves no room for doubt, says Mr Cavender, that the subject-matter of the San Giorgio claim is ACT, not MCT, and that it is recoverable because EU law regards it as an advance payment of the unlawful MCT. Since the ACT was lawfully paid in the first place, it can only have become unlawful, and thus recoverable, at the point of set-off: hence the change of character, or “morphing”, upon which the claimants rely.

Although the claimants accept that none of their claims in the present action, apart from the ACT claims which I have just mentioned, could succeed if the judgment of the Court of Appeal in FII (CA) were the last word on the subject, Mr Aaronson nevertheless urged me to proceed to hear and decide the substantive issues. He pointed out that everybody had come to court prepared for a full argument on all of the issues, and I have of course now heard that argument. He submitted that the Supreme Court was almost certain to grant permission to appeal in FII, at least on the issues concerning remedies and limitation, bearing in mind that the Law Society, the Chartered Institute of Taxation and the Association of Chartered Accountants in England and Wales have all intervened to ask the Supreme Court to hear the appeals on those issues. Accordingly, submitted Mr Aaronson, I could properly proceed on the hypothetical basis that parts of the judgment of the Court of Appeal may in due course be overturned, and I could consider the position on various alternative assumptions as to liability. He also suggested that I should in any event decide the main issues of liability and quantification on their merits, because the whole purpose of the GLO procedure is to obtain the court’s ruling on the common issues of law set out in the order, and directions for trial of those issues had already been given. A segregated issue by issue approach would therefore be justifiable in a way that might not be appropriate in ordinary litigation.

With regard to section 33, Mr Aaronson argued that even if most of the claims would have to be dismissed because they had been brought in the wrong forum, my views on the merits of the claims would still be of assistance to the parties and to the Tribunal, and might also avoid the need for hearings in the Tribunal and subsequent appeals, or at least greatly reduce the likely areas of dispute. Indeed, Mr Aaronson went so far as to suggest that the Supreme Court might find it helpful to have the benefit of my views on the issues which are peculiar to the present case, or which might cast fresh light on issues which arise in other cases. He instanced, in particular, the question of the correct approach to conforming interpretation, and the advantage to the Supreme Court of having a complete picture of the full range of dividend taxation issues when it considers the FII case.

Conclusions

These submissions were persuasively advanced by Mr Aaronson and Mr Cavender, and I have not found the question of how best to proceed an easy one. In the end, however, I have come to the clear conclusion that the Revenue are basically right, and that the better course is to adjourn the present trial at least until the Supreme Court has heard and disposed of any appeal in FII.

I begin with the obvious point that the normal rules of precedent apply to litigation conducted under a GLO in the same way as they do to ordinary litigation. I am therefore bound by the judgment of the Court of Appeal in FII (CA), and if I were to deliver a final judgment now I would have to do so on the basis of the law as it has been declared by the Court of Appeal. Since this would admittedly lead to the conclusion that the great majority of the claims had to be dismissed for reasons of limitation and/or procedure, it would in my view be inappropriate for me to rule on the substantive merits of those claims. As the law now stands such rulings would be both unnecessary and obiter. I would also have to spend a considerable amount of further time in preparing and writing this judgment; and, more importantly, my judgment would set in train an inevitable appeal process, which would then occupy a substantial amount of the Court of Appeal’s valuable time and resources, first of all in dealing with case management issues and then in hearing and disposing of such parts of the appeals as were permitted to proceed at this stage.

More generally, it seems to me that the existing appeals in FII and Thin Cap, together with any further references to the ECJ that may be ordered in the course of those appeals, will in due course probably resolve the great majority, if not all, of the issues which arise in the present case; and even if they do not directly resolve some of the issues, such as the technical questions of life office taxation, they will in all likelihood point the way to a solution clearly enough for a further hearing to be unnecessary. In those circumstances, I consider that the balance of convenience comes down firmly in favour of adjourning the present trial to await further developments, rather than pressing ahead and adding to the already very heavy appellate burden in the FII and Thin Cap litigation. The court should in my view think long and hard before launching another GLO missile into the judicial stratosphere, when a “wait and see” strategy is available as an alternative.

I would also reject Mr Aaronson’s argument that I should proceed to give judgment on at least the issues which are not time-barred but to which section 33 applies, with a view to assisting the Tribunal or perhaps obviating the need for a hearing before it. I am prepared, where possible, to adopt a reasonably pragmatic approach to the division of functions between the High Court and the Tribunal, but I cannot ignore the principles laid down by the House of Lords in Autologic, and it seems to me that to accept Mr Aaronson’s invitation would be to circumvent the procedure required by section 33 without any justification in either statute or EU law for so doing. The Tribunal is, in any event, the specialist body appointed by Parliament for the resolution of virtually all tax disputes, and in cases of sufficient difficulty or complexity arrangements could always be made for a High Court judge to sit as a member of the First-tier Tribunal, or for the appeal to proceed straight to the Tax and Chancery Chamber of the Upper Tribunal.

Still less am I influenced by Mr Aaronson’s flattering suggestion that the Supreme Court might be interested in what I had to say in the present case. As I have already had occasion to point out in a different case, the GLO procedure is not intended to provide claimants with a jurisprudential laboratory for research into new and difficult issues of law, but rather to promote the efficient and cost-effective disposal of cases which involve common issues of law or fact. If the test claimants and their advisers think that there are special features of the present case which could usefully inform the debate in the Supreme Court, if and when an appeal in FII or Thin Cap reaches it, steps can no doubt be taken to bring those features to the attention of the Supreme Court in due course and to make submissions about them.

Nor do I consider that there is any particular prejudice to the test claimants, or to the other GLO members, in adjourning the present case. The companies concerned are relatively few in number; they are all members of large insurance groups; the claims relate to historical legislation which is no longer in force; and no evidence of any need for urgency in resolving the dispute has been placed before me. Furthermore, since an appeal would inevitably follow any judgment of mine, it is hard to see how the ultimate resolution of the case would be accelerated by my following that course. What is certain, however, is that there would be a great increase in the costs of the litigation, and a very substantial expenditure of judicial time and court resources in dealing with questions which are more than likely to disappear altogether once the existing appeals in FII and Thin Cap have run their course and the ECJ has ruled on such further questions as may be referred to it.

In the light of all these considerations, I do not think it particularly matters, at this stage, whether the claimants are correct in their contention that some of their ACT claims are not time-barred. Even supposing that to be right, it would make no difference to my conclusion that the case should be adjourned. I therefore prefer to leave the point open, but in case it is helpful I will give an indication of my preliminary views on it.

In the first place, the only one of the three test claimants which could rely on the point is the PrudentialAssuranceCompany Limited, because it was the only claimant when the relevant amendments to the particulars of claim were made on 2 September 2003. The other two test claimants were not joined as parties until 2004.

Secondly, it seems to me that the critical question is whether the tax which Prudential now seeks to recover is “tax charged under an assessment” which it “paid” within the meaning of section 33(1). In my view it is, because as a matter of English statutory construction I can see no good answer to Mr Ewart’s argument that it was MCT paid by way of set-off pursuant to section 239(1) of ICTA 1988. The ACT was set off against, and discharged a corresponding part of, Prudential’s liability to MCT. The ACT was not transmuted into anything else. By virtue of section 239(1) it discharged, and therefore paid, part of the MCT liability.

If that is right, I would be inclined to accept that the sums claimed fall within the claim in paragraph 7(b) of the September 2003 particulars of claim for “restitution of … monies paid or liable to be paid … by the Claimants … pursuant to a mistake of law”. But in that case, it follows that the claims have to be dealt with, as a matter of English procedural law, under the section 33 regime, although with the modifications to the section laid down by the Court of Appeal in its application to a San Giorgio claim. The fact that EU law would arguably classify the claim as a San Giorgio claim to recover the ACT is, I think, irrelevant to the separate question of how the claim should be analysed for the domestic procedural purposes of section 33.

My provisional conclusion, therefore, is that the claims in question are not time-barred, but they fall within the scope of section 33 and must therefore be dealt with, if at all, in the Tribunal.

The Revenue submit that the proceedings should now be adjourned until after the Supreme Court has heard any appeal in FII, or until it has refused permission to appeal. They do not at this stage ask for any longer adjournment. I agree that this is the appropriate adjournment to order at this stage, so that the question can be kept under regular review. The logic of the views which I have expressed in this judgment is that it may in fact be advisable to prolong the adjournment until after any further references to the ECJ have been made and answered. However, there are still so many uncertainties that I can see no point in trying to be prescriptive at this stage. Much will depend on the amount of guidance which the Supreme Court is able to give, and the precise nature of the questions referred.

I would say, finally, that I naturally regret not giving the parties an answer at this stage to questions which were fully argued before me. I hope, however, that my reasons for declining to do so will be clear from what I have said in this judgment. Furthermore, I have now covered much of the background which I would need to set out in any final judgment, and I also have the benefit of the parties’ written arguments, my own notes, and the transcript. So if it does become clear that the time is ripe for me to proceed to give judgment on any of the issues, I should be able to do so without undue further difficulty or delay.