Boake Allen Ltd & Ors v Revenue and Customs Rev 1

This appeal is from a judgment of Park J, one of a series of judgments arising from litigation about advance corporation tax (ACT) organised under a group litigation order, following a decision of the European Court of Justice in Hoechst v. Attorney-General, Metallgesellschaft v. Attorney-General, Joined Cases C-397/98 and C-410/98 [2001] ECR I-1727, [2001] Ch 620 (which I will call the Hoechst case).

The common feature of the litigation is that it concerns liability to pay ACT in connection with the payment of dividends by companies incorporated in the UK which are subsidiaries of other companies incorporated outside the UK. The distinctive characteristic of the present appeal is that the parent companies are established in countries outside the European Union.

ACT was introduced by the Finance Act 1972 and applied from April 1973 until it was abolished with effect from April 1999. The legislation required that, when a UK company paid a dividend, it also paid an amount of corporation tax, referred to as ACT, calculated by reference to the amount of the dividend. If resident in the UK, the shareholder to whom the dividend was paid received a tax credit for the corresponding amount of ACT. The company paying the dividend could set the ACT off against its ordinary liability for corporation tax on its income and capital gains, which is sometimes called mainstream corporation tax (MCT). A non-corporate shareholder could set the tax credit off against his or its liability for income tax. A corporate shareholder resident in the UK would have received a tax credit in the form of franked investment income which franked its liability to account for ACT in respect of dividends which it paid. In some cases a Double Taxation Convention entitled a shareholder resident in a foreign territory to receive a tax credit: this is true of the USA, but not Japan.

If the company was a subsidiary of another UK company, however, the two companies could jointly make a group income election. Once this had been made and accepted by the inspector of taxes, the subsidiary could pay dividends without paying ACT at the same time, and correspondingly the parent did not obtain a tax credit. In its turn the parent would pay ACT when it paid dividends to its shareholders.

In the Hoechst case the European Court of Justice held that the requirement that the parent be a UK company in order that a group income election could be made was a breach of Article 43 of the EC Treaty, which confers the right of freedom of establishment, because it operated to the disadvantage of parent companies based in other Member States. Following that decision many claims have been made by parents or subsidiaries or both, where the parent is based in another Member State, seeking restitution in respect of ACT which it is said should not have been paid. In cases where the ACT paid by the subsidiary was later set off against the MCT liability of the company (known as utilised ACT) the claim is in respect of the fact that the company had to pay it earlier than would otherwise have been the case. If and to the extent that the ACT was not set off (known as unutilised ACT, or surplus ACT) the claim is for the amount of tax.

In the present cases, the parent companies are based in Japan or in the USA. They are also typical of cases where the parent is based in Switzerland. There are two aspects to the claim. First, it is said that the inability of the non-resident parent to join in a group income election is in breach of provisions of a Double Taxation Convention incorporated into UK law. Secondly it is said that that inability is also a breach of EU law, in this case under Article 56.

Park J decided that the legislation was a breach of the relevant Double Taxation Conventions, but that these were not part of UK law, so that this afforded no remedy. He also decided that it was clear that the legislation was not in breach of Article 56 of the Treaty, and that he should not refer to the ECJ the question whether there was such a breach. In those circumstances he did not have to consider what the position would have been as regards the consequences if the Double Taxation Convention had been both breached and incorporated into UK law, as regards the Claimants’ cause of action and the remedy to which they might be entitled, and he did not deal with these questions in full. He did consider a question of limitation which arose on the pleadings, although because he dismissed the Claimants’ claim this had no practical impact on the result of the case before him.

Since Park J decided the present case, on 24 November 2003, appeals against other decisions of his concerning the impact of the Hoechst case on questions of corporation tax have come before this court on at least four occasions before this appeal. Three are reported as: Pirelli Cable Holding NV v. IRC [2003] EWCA Civ 1849 (Pirelli), Deutsche Morgan Grenfell Group plc v. IRC [2005] EWCA Civ 78 (DMG) and Sempra Metals v. IRC [2005] EWCA Civ 389 (Sempra). In the fourth, judgment was handed down by the House of Lords immediately after the conclusion of argument on this appeal: Autologic Holdings plc v. IRC [2005] UKHL 54. Pirelli and DMG are under appeal to the House of Lords as well. In each instance the cases brought before the court are sample claims typical of many others. The same is true of the present appeals, which concern four corporate groups. One of the things which we have to decide is whether to refer the questions concerning Article 56 to the European Court of Justice. Counsel’s arguments about that turned, in part, on whether such a reference should be made after or before a further appeal to the House of Lords.

I will start by setting out such relatively few facts as need to be recorded, which I take from statements of facts which were agreed for the purposes of the trial of these cases as test cases.

NEC Semi-Conductors Ltd is a UK company, and a wholly owned subsidiary of NEC Corporation, a company resident in Japan. The subsidiary paid four relevant dividends to the parent, between October 1994 and March 1996, and paid the relative ACT. The dividends amounted to £163.9 million and the ACT to £40.975 million. Just over £34.1 million of the ACT was used by way of set-off against MCT, leaving some £6.8 million as surplus ACT. The periods between payment of ACT and its set-off against MCT varied between 259 and 800 days. The parent company obtained no tax credits. If a group income election had been permitted on the part of the parent and subsidiary, such an election would have been made.

Bush Boake Allen Inc is a company resident in the USA, and the direct or indirect parent of three relevant companies resident in the UK. Between January 1996 and April 1999 the three subsidiaries paid dividends to their direct or indirect US parents, and paid ACT accordingly amounting in all to some £10.4 million. About £8.2 million of this was set off against mainstream corporation tax, leaving £2.2 million surplus ACT. The periods between payment and set-off varied between 225 days and 1357 days. By virtue of provisions in the UK/USA Double Taxation Convention the US parent received a tax credit in respect of the dividends paid. (The effect of the decision in Pirelli is that this tax credit is irrelevant to the claim for restitution of the ACT.) If a group income election had been permitted one would have been made as between the US parents or parents and the UK direct and indirect subsidiaries.

Gallaher Ltd is a company resident in the UK, and at all material times a wholly-owned subsidiary of a US company, ATIC Group Inc. Between July 1995 and January 1997 Gallaher paid dividends to its parent; on one occasion within that period it made a distribution in the form of shares to its parent which counted as a qualifying distribution and therefore attracted the same liability to pay ACT as did the payment of a dividend. It paid ACT accordingly, amounting in all to £153,762,615. All of this was set off against MCT. The periods between payment and set-off varied between 261 and 717 days. If Gallaher and its parent had been able to join in a group income election, they would have done. (The agreement as to facts makes no mention of any tax credit.)

Acushnet Ltd is a UK resident company, and a wholly-owned subsidiary of Acushnet International Inc, resident in the US. The subsidiary paid dividends to the parent from 1989 onwards up to and including 1999. On each occasion the subsidiary also paid ACT. A limitation issue arises in respect of the ACT paid on and before 14 January 1995. The ACT so paid was all set off against MCT, most of it within no more than 261 days. If the parent and subsidiary could have joined in a group income election they would have done. The parent obtained tax credits under the Double Taxation Convention in respect of the ACT paid.

The several claim forms were issued during 2001. The NEC claim was commenced on 12 January 2001, the Bush Boake Allen claim on 6 March 2001, the Gallaher claim on 11 July 2001, and the Acushnet claim on 22 November 2001. Each used a Part 8 Claim Form.

Already at that stage there were common features to the formulation of the claims. A relevant Group Litigation Order was first made in November 2001. Since then all the relevant cases have been under the control of Park J as regards case management. On 11 July 2003 he made an order bringing the details of the various claims into line in all respects, except as regards the individual variables, above all details of the payments made in each instance. It is this order which gives rise to the argument about limitation already mentioned.

Under the Group Particulars of Claim the relationship between the foreign parent and the UK subsidiary is alleged, according to the relevant individual factual details. It is alleged that, pursuant to sections 14, 208, 231, and 247 and Schedule 13 of the Income and Corporation Taxes Act 1988 (ICTA) (referred to as the ACT Provisions) the UK subsidiaries paid dividends to the foreign parents and paid corresponding amounts of ACT, and that if the foreign parent had been a UK resident it and its UK subsidiaries would have made joint group income elections as a result of which the subsidiaries would not have paid ACT in respect of the dividends. The fact that UK resident subsidiaries of parents resident in the USA, Japan or Switzerland had to pay ACT because they could not join in a group income election, whereas UK subsidiaries of a UK parent could have avoided the need to pay ACT is alleged to amount to both (1) a restriction on movements of capital, or payments, or both from the UK subsidiaries to the foreign parents, contrary to Article 56 of the EC Treaty, and (2) an “other or more burdensome” taxation or connected requirement than those imposed upon other similar enterprises, in breach of the relevant article of the Double Taxation Convention applicable in each case. The remedies claimed are (a) a declaration that the ACT provisions as they apply to the Claimants are contrary to the Double Taxation Convention and therefore illegal; (b) restitution or damages for breach or failure to comply with the Double Taxation Convention; (c) a declaration that the ACT provisions as they apply to the Claimants are contrary to Article 56 and therefore illegal; (d) restitution and damages for breach of a failure to comply with Article 56; (e) restitution of, and/or compensation or damages for, the ACT payments paid pursuant to a mistake of law or demands by the Defendants and the ACT provisions, those provisions being contrary to Article 56 and the Double Taxation Convention. (I ignore claims under the European Human Rights Convention which are not pursued.)

Corporation tax itself was first introduced in the Finance Act 1965, at the same time as capital gains tax. ACT followed in 1972. I will refer to the history of the legislation later. It is now set out in ICTA. By section 6 of that Act, corporation tax is to be charged on the profits of companies, which means income and chargeable gains. A company is not chargeable to income tax or to capital gains tax. By section 8(1) a company is chargeable to corporation tax on all its profits wherever arising, subject to such exceptions as are contained in the legislation. ACT appears first in section 14, as follows:

“(1)

Subject to section 247, where a company resident in the United Kingdom makes a qualifying distribution it shall be liable to pay an amount of corporation tax (‘advance corporation tax’) in accordance with subsection (3) below.”

Subsection (3) provides that ACT “shall be payable on an amount equal to the amount or value of the distribution” and provides for the rate at which it is paid. Sub-section (4) is as follows:

“(4)

The provisions of this Act as to the charge calculation and payment of corporation tax (including provisions conferring any exemption) shall not be construed as affecting the charge calculation or payment of advance corporation tax, and the Corporation Tax Acts shall apply for the purposes of advance corporation tax whether or not they are for the time being applicable for the purposes of corporation tax other than advance corporation tax.”

Section 231 provides for tax credits, which are for the most part limited to persons resident in the UK.

Under section 239, subject to any group income election, ACT paid by a company and not repaid in respect of any distribution made by it in an accounting period was to be set against its liability to corporation tax on any profits charged to corporation tax for that accounting period. This is subject to various qualifications. If the company has surplus ACT, that is to say ACT which cannot be set against liability to corporation tax on those profits, it can be set against the company’s liability to corporation tax in respect of profits of previous periods or subsequent periods within limits. There may, nevertheless, be ACT which cannot be set against MCT and which is therefore not merely an advance payment of tax that will eventually become due.

The provisions about group income elections which lie at the heart of this case are in section 247, supplemented by section 248. Section 247(1) (ignoring immaterial words) is as follows:

“Where a company (“the receiving company”) receives dividends from another company (“the paying company”), both being bodies corporate resident in the United Kingdom, and the paying company is (a) a 51 per cent subsidiary of the other … then, subject to the following provisions of this section, the receiving company and the paying company may jointly elect that this subsection shall apply to the dividends received from the paying company by the receiving company (“the election dividends”)”.

While an election is in force the election dividends are excluded from the ambit of sections 14 and 231, and are part of the group income of the receiving company. They are thus part of the profits of that company. The paying company may give notice excluding particular dividends from the election, and either company may give notice to the inspector revoking the election. The election is made by notice to the inspector of taxes, and is of no effect in relation to dividends paid less than 3 months after the giving of the notice and before the inspector is satisfied that the election is validly made, and has so notified the companies concerned; if within 3 months the inspector notifies the companies that the validity of the election is not established to his satisfaction, the election is of no effect. A decision that the validity of the election is not established is subject to appeal on the part of the companies concerned as if the decision were an assessment made on either company. Accordingly the provisions governing appeals in the Taxes Management Act 1970 apply. These provisions have been amended over time, but there is a time limit (of 30 days) under section 42 and, on the other hand, an extension of time may be permitted: see section 49. In the case of a disputed group income election a new period for appealing could presumably be secured by notifying a fresh election.

A consequence of this structure of provisions as regards ACT and group income elections, however, is that the companies concerned must wait for the validity of the group income election to be established, whether to the satisfaction of the inspector or by way of appeal, before they can take advantage of the provisions of section 247 so as to avoid having to pay ACT when paying dividends. Nor can the appeal affect the position except for the future. If, for commercial reasons, the UK company must pay a dividend to its parent while the position is unresolved, it must pay ACT, and will not be able to recover it even if it later establishes that the group income election ought to have been accepted.

Different provisions apply for the administration of ACT from those which apply in relation to MCT, as is inevitable, and recognised by section 14(4). The collection of ACT is governed by Schedule 13 of ICTA. Since the quantification of ACT depends primarily on the amount of any qualifying distributions during the quarter, it is a different and more straightforward process from that for the quantification of MCT, which depends on the amount of the company’s income and chargeable gains. Returns have to be made each quarter, within 14 days after the end of the quarter. If any ACT is due, it is to be paid by the time by which the relevant return has to be made, and without any assessment. If it is not so paid it may be assessed on the company.

I take the following description of the position as regards double taxation conventions from Park J’s judgment:

“9.

Double taxation agreements are treaties concluded between sovereign states. Under the law of the United Kingdom they are entered into in exercise of the prerogative power of the Crown. Under our law treaties are not ‘self-executing’: that means that, although they are binding in international law between the United Kingdom and the other State as soon as they are concluded (or ratified, if by their terms they require ratification), they do not then take automatic effect in domestic law as part of the law of the United Kingdom. The intervention of Parliament, either directly by statute or by statutory delegation authorising another person or body to bring the treaty into effect domestically, is needed. In the case of double taxation agreements Parliament has, as I will explain in more detail later, delegated to Her Majesty the power to bring the treaties into domestic effect by Order in Council.

10.

The United Kingdom’s double taxation agreements are all bilateral agreements with one other State, and there are a great many of them. However, nearly all of them are based on a draft produced, with a supporting commentary, by the OECD. It follows that many of the provisions of the United Kingdom’s double taxation agreements are identical between themselves, and indeed identical to provisions of double taxation agreements concluded between pairs of other nations which have also adopted the OECD model. Among the States which have concluded with the United Kingdom double taxation agreements containing non-discrimination articles are Japan, Switzerland and the United States, all non-EC countries. NEC Semi-Conductors Ltd, one of the companies which is acting as a test case for the present proceedings, is a subsidiary of a Japanese parent company. Several Bush Boake Allen companies are subsidiaries of a United States corporation, and they are another test case. The United Kingdom/Japan double taxation agreement and the United Kingdom/USA double taxation agreement are not identical in all respects, but they are identical so far as the present case is concerned, since they both contain non-discrimination articles, and those articles include identical sub-paragraphs which are the particular provisions relied on. The test cases do not include a United Kingdom subsidiary of a Swiss parent company, but there are such subsidiaries which are parties to the GLO. The decision in the present case should serve as a test case for them, because the United Kingdom/Switzerland double taxation agreement also contains a non-discrimination article which has the identical sub-paragraph within it.”

The current Double Taxation Convention between the UK and Japan was made on 10 February 1969, before ACT was introduced into UK tax law. It was given effect under UK law by an Order in Council made on 17 December 1970: the Double Taxation Relief (Taxes on Income) (Japan) Order 1970, SI 1970 No 1948. The Convention follows the OECD model, for the most part.

By Article 2 of the Double Taxation Convention the taxes which are the subject of the Convention include the corporation tax. Most of the provisions of the Convention deal with issues of double taxation, including Article 11 about dividends. The provision which is relied on in the present case, however, is known as the non-discrimination article, Article 25. I set out the whole article, though only paragraph (3) is of direct relevance:

“(1)

The nationals of a Contracting State shall not be subjected in the other Contracting State to any taxation or any requirement connected therewith which is other or more burdensome than the taxation and connected requirements to which nationals in that other State in the same circumstances are or may be subjected.

(2)

The taxation on a permanent establishment which an enterprise of a Contracting State has in the other Contracting State shall not be less favourably levied in that other State than the taxation levied on enterprises of that other State carrying on the same activities.

(3)

Enterprises of a Contracting State, the capital of which is wholly or partly owned or controlled, directly or indirectly, by one or more residents of the other Contracting State shall not be subjected in the first-mentioned Contracting State to any taxation or any requirement connected therewith which is other or more burdensome than the taxation and connected requirements to which other similar enterprises of that first-mentioned State are or may be subjected.

(4)

Nothing contained in this article shall be construed as obliging either Contracting State to grant to individuals not resident in that State any of the personal allowances, reliefs and reductions for tax purposes which are granted to individuals so resident, nor as conferring any exemption from tax in a Contracting State in respect of dividends paid to a company which is a resident of the other Contracting State.

(5)

In this article the terms “taxation” means taxes of every kind and description.”

The equivalent Double Taxation Convention for the USA is dated 31 December 1975. It has been amended since then, but not relevantly for present purposes. The Order in Council is SI 1980 No 568, made on 21 April 1980. The Double Taxation Convention follows the OECD model. Having been made after the introduction of ACT, it includes some provisions arising from that legislation. In particular Article 10, which deals with dividends, includes the provision noted above whereby US resident shareholders holding shares in a UK resident company are entitled to a tax credit in respect of the dividend. The non-discrimination article is Article 24. It includes some additional provisions, not relevant to the present questions, but the central provision, Article 24(5), is in the same terms as the OECD model and as Article 25(3) in the UK/Japan Double Taxation Convention, set out above.

Although it is not directly relevant to the cases under appeal, I note that the Double Taxation Convention between the UK and Switzerland also includes the same non-discrimination article, as Article 23(5).

The effect of a Double Taxation Convention under UK law as regards income and corporation tax is governed by ICTA section 788, and for present purposes by sub-sections (1) and (3). These are as follows:

“(1)

If Her Majesty by Order in Council declares that arrangements specified in the Order have been made with the government of any territory outside the United Kingdom with a view to affording relief from double taxation in relation to

(a)

income tax

(b)

corporation tax in respect of income or chargeable gains, and

(c)

any taxes of a similar character to those taxes imposed by the laws of that territory,

and that it is expedient that those arrangements should have effect, then those arrangements shall have effect in accordance with subsection (3) below.”

“(3)

Subject to the provisions of this Part, the arrangements shall, notwithstanding anything in any enactment, have effect in relation to income tax and corporation tax in so far as they provide -

(a)

for relief from income tax, or from corporation tax in respect of income or chargeable gains; or

(b)

for charging the income arising from sources, or chargeable gains accruing on the disposal of assets, in the United Kingdom to persons not resident in the United Kingdom; or

(c)

[irrelevant]; or

(d)

for conferring on persons not resident in the United Kingdom the right to a tax credit under section 231 in respect of qualifying distributions made to them by companies which are so resident.”

The Orders in Council giving effect to the several Double Taxation Conventions contain the words necessary to bring them into operation under this section.

The section deals only with income tax and corporation tax because that is the subject matter of ICTA. The relevant provisions of ICTA are applied by reference for the purposes of capital gains tax, by section 277 of the Taxation of Chargeable Gains Act 1992. Equivalent provision to that of section 788 is made for inheritance tax by section 158 of the Inheritance Tax Act 1984.

The appellants contend that the ACT provisions of ICTA are in breach of the non-discrimination article in the Double Taxation Conventions with Japan and with the USA, and that the non-discrimination article is given direct and overriding effect as part of UK law by virtue of the opening words of section 788(3). This gives rise to two questions: (1) is there a breach of the non-discrimination article? (2) if so, is the non-discrimination article incorporated as part of UK law by virtue of section 788? The judge held that there was a breach but that the non-discrimination article was not (relevantly) incorporated into UK law. In the course of the hearing before us some submissions were also developed on behalf of the Appellants as to the true nature of the breach of the Conventions, if there was one.

The debate on this point turns on the meaning of the following words in the relevant provision of the non-discrimination article: “other similar enterprises of that first-mentioned State”. The case concerns enterprises of one Contracting State (UK) the capital of which is wholly or partly owned or controlled, directly or indirectly, by one or more residents of the other Contracting State (Japan or the USA in this instance), and to the taxation to which the UK subsidiary is subjected in the UK. The relevant taxation and other connected requirements must not be “other or more burdensome” than those to which other similar enterprises of the UK are or may be subjected. What are the “other similar enterprises”? It cannot mean other UK enterprises owned or controlled by residents of Japan; that would not prohibit any discrimination against Japanese owned UK subsidiaries generally. For the same reason it cannot mean UK companies owned or controlled by residents of countries other than the UK generally.

Mr Aaronson Q.C., for the Appellants, submits that it must mean UK subsidiaries of UK parent companies. The judge accepted that submission. Comparison with the paragraph of the article about permanent establishments lends force to it: see Article 25(2) in the Double Taxation Convention with Japan.

Mr Glick Q.C., for the Respondent, argues that the non-discrimination article is concerned with the taxation of the UK company, not with that of its parent, wherever that may be resident, and that a comparison which requires attention to be given to the tax status of the parent is outside the scope of the non-discrimination article. He submits that the Appellants cannot identify an “other similar enterprise” of the UK which is treated more favourably than the UK subsidiaries of Japanese or US parents.

Reference was made before us and below to a Commentary issued by the OECD on the model non-discrimination article, of which the relevant paragraph is paragraph 6. The commentary is as follows:

“This paragraph forbids a Contracting State to give less favourable treatment to an enterprise, the capital of which is owned or controlled, wholly or partly, directly or indirectly, by one or more residents of the other Contracting State. This provision, and the discrimination which it puts an end to, relates to the taxation only of enterprises and not of the persons owning or controlling their capital. Its object therefore is to ensure equal treatment for taxpayers residing in the same State, and not to subject foreign capital, in the hands of the partners or shareholders, to identical treatment to that applied to domestic capital.”

We were also shown a Technical Explanation by the US Treasury Department of the UK/US Double Taxation Convention which, as regards the equivalent article, interprets “other similar enterprises of the first-mentioned State” as “enterprises of the first-mentioned Contracting State carrying on the same activities, the capital of which is wholly or partly owned or controlled by one or more residents of the first mentioned Contracting State”.

Mr Glick’s submission is that, even if it were a valid comparison to postulate a UK company which is owned or controlled by one or more UK residents, this would not be sufficient for the Appellants. They have to postulate a UK company which is owned or controlled to at least 51% by a corporation resident in the UK. If a UK company is owned by a UK resident individual, no question of group income election arises, nor does it if the corporate shareholding on the part of UK resident companies does not amount to at least 51% for any one company or group of companies.

Mr Glick submits in support of this proposition that the provisions about group income elections are not really to do with reliefs against tax, but with the allocation of a liability for tax within the group of companies. By means of a group income election the subsidiary can avoid having to pay ACT at the time when it pays a dividend, but the parent comes under a corresponding liability to pay ACT, if and when it pays dividends, whether to a parent company or to a general body of shareholders, wherever they may be resident for tax purposes. In that context, he submits, there is no irrationality or objectionable discrimination in limiting the availability of a group income election to parent and subsidiary who are both resident in the UK, since non-resident parents are not liable for ACT, and there is no good reason to allow the subsidiary to opt out of ACT by making a group income election if the parent is itself outside the scope of that tax.

The comparison which is required by the non-discrimination article includes, among several possibilities, a comparison with a UK company wholly owned or controlled by a single shareholder. It is true that the non-discrimination article does not deal with whether the sole shareholder is an individual or a corporation, but it seems to me that it would be irrational to exclude that as a relevant element in the comparison, where the circumstances of the actual case are such that the foreign parent is a corporation. As Park J said at paragraph 28, “it is necessary to assume that, if the actual company is a subsidiary of another company, then so is the hypothetical comparator company”.

If that is so, then Mr Glick’s argument would rest solely on the proposition that the UK and foreign parent cases are not really comparable because only in the one case can ACT arise at all, and it would be inappropriate to allow a comparison on the part of a non-ACT paying foreign parent company with a UK parent which is liable to pay ACT when it makes distributions. I agree with Mr Aaronson and with the judge below that this would introduce an illegitimate element into the comparison. It is not legitimate because it is concerned with the fiscal position of the parent, which is not what the non-discrimination article is about. The OECD Commentary makes this point, but it seems to me that it would be apparent even without that text.

Mr Glick next challenged the judge’s conclusion that the ACT provisions would, on their face, subject the UK subsidiary of a Japanese parent company to other or more burdensome taxation or connected requirements than the UK subsidiary of a UK parent company. He submitted that the group income election regime is not a taxation or other requirement for the subsidiary, because it relates to the taxation of the group, not of the subsidiary. Like the judge, I cannot accept that submission. To impose a liability to ACT by section 14 of ICTA, which is itself subject to section 247, but then to limit the relieving provisions of section 247 to UK groups, seems to me clearly to create different requirements in respect of taxation for the subsidiary of the foreign parent from those which apply to the subsidiary of the UK parent. For the same reason, the provisions concerning ACT are more burdensome for the subsidiary of the foreign parent than for a member of a UK group, because the former does not have the opportunity to avoid having to pay ACT by a group income election.

For these reasons, it seems to me that the judge was right to hold that the ACT provisions, on their face, are inconsistent with the non-discrimination article because the UK subsidiary of a foreign company cannot avoid having to pay ACT when it pays a dividend by entering into a group income election, unlike members of a UK group. To limit the availability of group income elections to subsidiaries of UK companies is a breach of the non-discrimination article of the Conventions.

The judge held that, even though the non-discrimination article was infringed by the ACT provisions, there was no remedy for this under UK law because the provisions of the Double Taxation Convention as regard ACT are not incorporated into UK law by section 788. Mr Aaronson challenges this.

Essentially the judge’s reasoning on this is that, while section 788 gives effect to reliefs from mainstream corporation tax by virtue of sub-section (3)(a), ACT is different and is outside the scope of section 788(3)(a). Mr Aaronson submits that this would make it possible for the UK to impose a higher rate of ACT on companies in respect of distributions payable to companies outside the UK (and the EU), which he says cannot have been intended. It matters not that there has never been any such provision, nor whether such a provision would be likely to be introduced. It would be an obvious act of discrimination, against which the non-discrimination article is aimed. It is hardly likely, he says, that the section was designed to allow that kind of discrimination, especially when the section does include one provision which arises directly from the ACT regime, namely sub-section (3)(d) dealing with tax credits for foreign companies.

The point arises from the use in sub-section (3)(a) of the phrase “corporation tax in respect of income and chargeable gains”. Mr Glick submits, and the judge held, that, although ACT is corporation tax, it is not corporation tax in respect of income or chargeable gains. It is corporation tax by reference to distributions. Of course, it is connected with corporation tax in respect of income and chargeable gains because it may be set against the liability for that corporation tax. But it is payable whether or not there are income or capital gains for the directly relevant period, or any other relevant period, and it is calculated by reference to the amount of the distribution.

Some attention was given during submissions to the history of these provisions. Corporation tax itself was first introduced into UK tax law by the Finance Act 1965, at the same time as capital gains tax. Corporation tax was charged on profits which were already defined as including both income and chargeable gains, but corporation tax on chargeable gains was dealt with separately from corporation tax on income. Thus, in the Income and Corporation Taxes Act 1970 section 497, the predecessor provision to section 788, the equivalent paragraph to section 788(3)(a) refers to “relief from income tax or from corporation tax in respect of income”. Corporation tax in respect of chargeable gains was dealt with separately.

ACT was introduced by the Finance Act 1972, as already mentioned. In the same Act the two aspects of mainstream corporation tax were brought into line, in particular as regards double taxation provisions. By section 100 of that Act, the double taxation provisions of ICTA 1970 applicable to corporation tax in respect of income “shall apply also to corporation tax in respect of chargeable gains”. Section 98(2) effected the introduction into section 497 of what has now become section 788(3)(d), referring to the right to tax credits, such as are provided for in the UK/USA Double Taxation Convention. Section 788(3)(a) has followed its predecessor, as amended, by referring to corporation tax in respect of income and chargeable gains, rather than just to corporation tax. The question is whether by using the composite phrase it means only mainstream corporation tax, or whether it really means no more than “corporation tax”, so as to include ACT. Clearly it is an inclusive provision, by contrast with the former phrase which distinguished between corporation tax on one type of profits and that on the other. Is that the only significance of the reference to income and to chargeable gains, or does the phrase refer specifically to mainstream corporation tax and leave ACT aside, the latter being dealt with in the double taxation provisions only by the specific provisions of section 788(3)(d)?

If Mr Aaronson is right, and the phrase really just means “corporation tax”, then it is odd that the longer phrase should be used, especially as the shorter phrase is also used, in the introductory words of section 788(3). The context there seems to be that “corporation tax” is used as a general category and “corporation tax in respect of income and chargeable gains” as a subset.

Mr Aaronson invited us to consider what policy reason there could have been for not including ACT among the provisions which would be given effect by section 788, and submitted that, even if (contrary to his argument) the natural meaning of sub-section (3) was that reliefs from ACT were not covered by paragraph (a), nevertheless the paragraph was capable of being construed so as to extend to ACT, and that it ought to be so construed since otherwise the UK would be in breach of its obligation under the Treaty represented by the Double Taxation Convention.

In support of the last submission he cited from the judgment of Diplock LJ in Salomon v. Customs & Excise Commissioners [1967] 2 QB 116, at 143:

“The convention is one of those public acts of state of Her Majesty’s Government of which Her Majesty’s judges must take judicial notice if it be relevant to the determination of a case before them, if necessary informing themselves of such acts by inquiry of the appropriate department of Her Majesty’s Government. Where, by a treaty, Her Majesty’s Government undertakes either to introduce domestic legislation to achieve a specified result in the United Kingdom or to secure a specified result which can only be achieved by legislation, the treaty, since in English law it is not self-operating, remains irrelevant to any issue in the English courts until Her Majesty’s Government has taken steps by way of legislation to fulfil its treaty obligations. Once the Government has legislated, which it may do in anticipation of the coming into effect of the treaty, as it did in this case, the court must in the first instance construe the legislation, for that is what the court has to apply. If the terms of the legislation are clear and unambiguous, they must be given effect to, whether or not they carry out Her Majesty’s treaty obligations, for the sovereign power of the Queen in Parliament extends to breaking treaties … and any remedy for such a breach of an international obligation lies in a forum other than Her Majesty’s own courts. But if the terms of the legislation are not clear but are reasonably capable of more than one meaning, the treaty itself becomes relevant, for there is a prima facie presumption that Parliament does not intend to act in breach of international law, including therein specific treaty obligations; and if one of the meanings which can reasonably be ascribed to the legislation is consonant with the treaty obligations and another or others are not, the meaning which is consonant is to be preferred. Thus, in case of lack of clarity in the words used in the legislation, the terms of the treaty are relevant to enable the court to make its choice between the possible meanings of these words by applying this presumption.”

As Mr Glick pointed out, the present case is different, because the legislative provision is a general one, under which, by subordinate legislation, a number of treaties may be given effect as part of domestic law, whereas the Salomon case was concerned with a specific provision which, though it did not say so, was clearly intended to enact a particular treaty. Moreover section 788 makes it clear that it intends to bring into direct effect only certain provisions of relevant treaties.

Mr Glick submitted that the policy underlying the limited application of section 788 to ACT could be found in the White Paper “Reform of corporation tax” which explained the new provisions, in April 1972. Paragraphs 30 to 32 of the White Paper dealt with international aspects. Paragraph 32 made it clear that tax credits were limited to UK resident shareholders. Its last sentence is as follows:

“Power is also being taken to entitle non-resident shareholders to the tax credit under double taxation agreements: the terms on which non-resident shareholders will be entitled to a tax credit in respect of a qualifying distribution under any agreement will be a matter for negotiation.”

So, Mr Glick submits, the limited recognition of the role of Double Taxation Conventions was only as regards what has become section 788(3)(d), as reflected in Double Taxation Conventions negotiated since then, including those with the USA and Switzerland.

That is the explanation for the specific provision. It does not of itself provide an explanation by way of policy for applying the non-discrimination article directly as regards mainstream corporation tax but not as regards ACT. It would leave a stark omission, albeit one which might not ever be likely to be of practical relevance, in Mr Aaronson’s example of differential rates of ACT as between a company which is a subsidiary of a Japanese parent and another which is part of a UK or EU group.

The judge concluded that the non-discrimination article is not made part of UK law insofar as it relates to ACT. His reasoning proceeded as follows. First, not everything in a Double Taxation Convention is enacted by section 788; its only relevant effect is that set out in the four paragraphs of section 788(3), because of the words “insofar as they provide for”. Secondly, given that the terms of section 788(3) are not designed necessarily to implement as part of UK law everything relating to income tax or corporation tax that might realistically be covered in a Double Taxation Convention, there is no reason to give the words that do define what is to be implemented a wide meaning rather than what the court concludes is their natural meaning. Thirdly, he held that the natural meaning of “corporation tax in respect of income or chargeable gains” was to refer to mainstream corporation tax only, and not ACT.

I agree with the judge on these points. The first is clear from the words of the section. As for the second, I agree that the Salomon case is not directly in point, because it concerned legislation introduced to give effect to a specific treaty obligation. Section 788 is a different kind of provision, which has its equivalents in other fiscal legislation, such as those relating to capital gains tax and to inheritance tax mentioned above. It provides for a class of treaty obligations to be given effect to by Orders in Council. Some such Orders were already in force when section 788 became law, and others could be expected to be introduced by further orders thereafter. I agree that the section gives effect only to certain provisions of Double Taxation Conventions. As the judge comments, Double Taxation Conventions apply to all forms of taxation and so they may need to be given effect in relation to different aspects of taxation by affecting different domestic legislation. It is entirely rational to find, in a double taxation provision in ICTA, a list referring to income tax and corporation tax, so as to leave capital gains tax or inheritance tax, for example, to be given effect in the other legislation relating to those taxes. But even as regards income tax and corporation tax, the drafting technique of section 788(3) does suggest that there are limits to the provisions in a Double Taxation Convention to which automatic and overriding effect is given by the section. Otherwise, as the judge says, the sub-section could simply have said: “subject to the provisions of this Part, the arrangements shall have effect in relation to income tax and corporation tax notwithstanding anything in any enactment”.

Since the section is selective as to what provisions of a Double Taxation Convention are given direct effect under UK law, I agree with the judge that it would not be right to approach the question of interpretation with a presumption that it was intended to give effect to all provisions of the Double Taxation Convention. I therefore agree with him that the only proper approach is to consider what is the natural meaning of the words used in section 788(3).

So then the question is what is the natural meaning of the words in paragraph (a): “corporation tax in respect of income and chargeable gains”? Is it equivalent to “corporation tax”, so as to include ACT, or does it refer only to mainstream corporation tax and exclude ACT? The history of the corporation tax legislation throws a little light on the question, by explaining why there might be reference to corporation tax on income and on chargeable gains, rather than just to corporation tax, but it does not seem to me that this history provides the answer to the present question. The presence of paragraph (d) in sub-section (3) shows that the draftsman did have the ACT provisions in mind when putting section 788(3) together as part of the consolidation of the income tax and corporation tax legislation. That, however, also fails to provide the answer. Mr Glick’s comment is that it would be extraordinary in that context, if paragraph (a) were intended to cover all forms of corporation tax, that it used the apparently limited phrase which was in fact used. Mr Aaronson’s comment is that it would be incomprehensible that special effect should be given to the tax credit provisions by paragraph (d) and the rest of the ACT provisions should be left outside the ambit of the section.

It seems to me that most powerful indication of the natural meaning of the phrase in paragraph (a) is to be found in the juxtaposition of the short and general phrase “corporation tax” in the opening words with the longer and more specific phrase “corporation tax in respect of income and chargeable gains” in paragraph (a). Like the judge, I conclude that the paragraph is intended to have a more limited scope, and that it does not extend to ACT. In my judgment, although ACT is certainly “corporation tax” (that is clear from section 14, with its words “an amount of corporation tax (“advance corporation tax”)”), it is not “corporation tax in respect of income or chargeable gains”. I come to that conclusion because of the separate use of the phrase, in contrast with the shorter phrase “corporation tax”. It also seems to me that, despite the connection between the two forms of corporation tax, ACT is not in respect of income or chargeable gains, but rather “in respect of”, or perhaps better “by reference to”, qualifying distributions. I would not myself attach importance to the terms of section 14(4), because corporation tax paid in advance, by reference to distributions, must necessarily have different provisions for its charge, calculation and payment from those which apply to mainstream corporation tax. That would not, in itself, show that the former was not corporation tax in respect of income or chargeable gains. Nevertheless, for the other reasons I have given, I agree with the judge that the natural meaning of section 788(3) is that the only provisions concerning ACT which are given direct effect as part of UK tax law are those referred to in sub-section (3)(d) as regards tax credits.

For those reasons, it seems to me that the judge was right to hold that the Appellants have no remedy under UK law as such, by reference to the Double Taxation Conventions.

Mr Aaronson’s alternative (and preferred) challenge to the ACT provisions is by way of EU law. He submits that the ACT provisions are inconsistent with Article 56 of the Treaty, and not saved by Article 57.1. The respective articles are as follows:

“56.

1. Within the framework of the provisions set out in this Chapter, all restrictions on the movement of capital between Member States and between Member State and third countries shall be prohibited.

2.

Within the framework of the provisions set out in this Chapter, all restrictions on payments between Member States and between Member States and third countries shall be prohibited.”

“57.

1. The provisions of Article 56 shall be without prejudice to the application to third countries of any restrictions which exist on 31 December 1993 under national or Community law adopted in respect of the movement of capital to or from third countries involving direct investment – including in real estate – establishment, the provision of financial services or the admission of securities to capital markets.”

He does not suggest that we ought to decide whether the ACT provisions are restrictions which are prohibited by Article 56 and not saved by Article 57.1, but rather submits that we ought to refer that question to the European Court of Justice. Mr Glick submits that we ought not to consider any such reference because he says it is clear that, even if the ACT provisions are restrictions within Article 56, they are protected by Article 57.1. Alternatively he submits that a reference ought not to be made until after the case has proceeded to the House of Lords (if, as both sides seem to anticipate, it will) because only then would it be clear what, if any, restriction there is under national law. Mr Aaronson opposes the latter course.

In order to put this issue in context, it is necessary to trace the history of what is now Articles 56 and 57.

In the original Treaty, of Rome, the nearest equivalent was Article 67, as follows:

“1.

During the transitional period and to the extent necessary to ensure the proper functioning of the common market, Member States shall progressively abolish between themselves all restrictions on the movement of capital belonging to persons resident in Member States and any discrimination based on the nationality or on the place of residence of the parties or on the place where such capital is invested.

2.

Current payments connected with the movement of capital between Member States shall be freed from all restrictions by the end of the first stage at the latest.”

The transitional period was 12 years, and the first stage the first 4 years of that period. This provision was not regarded as being of direct effect. A series of Directives was made in order to put it into effect. The last was Council Directive of 24 June 1988, 88/361/EEC. Article 1 of that Directive was as follows:

“1.

Without prejudice to the following provisions, Member States shall abolish restrictions on movements of capital taking place between persons resident in Member States. To facilitate application of this Directive, capital movements shall be classified in accordance with the Nomenclature in Annex 1.

2.

Transfers in respect of capital movements shall be made on the same exchange rate conditions as those governing payments relating to current transactions.”

Article 7 referred to third countries, but in less specific terms:

“1.

In their treatment of transfers in respect of movements of capital to or from third countries, the Member States shall endeavour to attain the same degree of liberalization as that which applies to operations with residents of other Member States, subject to the other provisions of this directive.

The provisions of the preceding subparagraph shall not prejudice the application to third countries of domestic rules or community law, particularly any reciprocal conditions, concerning operations involving establishment, the provision of financial services and the admission of securities to capital markets.

2.

Where large-scale short-term capital movements to or from third countries seriously disturb the domestic or external monetary or financial situation of the Member States or of a number of them, or cause serious strains in exchange relations within the Community or between the Community and third countries, Member States shall consult with one another on any measure to be taken to counteract such difficulties. This consultation shall take place within the Committee of Governors of the Central Banks on the initiative of the Commission or of any Member State.”

Annex 1 contains explanatory notes as well as the Nomenclature referred to in Article 1, which is not exhaustive. Direct investment includes the establishment of new branches or new undertakings belonging solely to the person providing the capital, and the acquisition in full of existing undertakings. The opening notes say that the capital movements listed are taken to cover “all the operations necessary for the purposes of capital movements: conclusion and performance of the transaction and related transfers”.

The next stage came with the Maastricht treaty, effective as from 1 January 1994 (hence the reference in the present Article 57.1 to restrictions existing on 31 December 1993). New articles were introduced into the previous text of the Treaty, of which Article 73b was in the same terms as the present Article 56, and Article 73c was in the same terms as the present Article 57. Then, with the Treaty of Amsterdam, the articles were reorganised and renumbered, and the relevant articles achieved their present numbering.

Thus, provisions designed to prohibit restrictions on payments and on movements of capital started off as limited to the Member States, and already at the outset a distinction was drawn between capital movements on the one hand and payments on the other, with a different time scale for the liberalization in relation to one and to the other. Later, provisions designed to achieve liberalization as regards third countries were also introduced, and later still these were made mandatory, but subject to a transitional protection for restrictions as regards capital movements which were already in place on 31 December 1993. The question would be, assuming that UK law does impose restrictions on payments between the UK and third countries (as Mr Aaronson submits the ACT provisions are, especially if they are to be construed as I have indicated, so that there is no remedy under the Double Taxation Convention), and accepting that these restrictions were in place already on 31 December 1993, are those restrictions saved by Article 57.1, as being relevant restrictions on capital movements?

Mr Glick submits that it is clear from a decision of the European Court of Justice that, if there are relevant restrictions under UK law, they are restrictions on capital movements, and that it follows that they are protected by Article 57.1. That decision is Staatssecretaris van Financiën v. Verkooijen Case C-35/98 [2002] ECR I-4071 [2002] STC 654 (Verkooijen).

The Verkooijen case concerned a resident of the Netherlands who owned shares in a company resident in Belgium and received dividends on those shares. If the dividends had been paid by a company resident in the Netherlands their treatment for income tax in his hands would have been more beneficial, because a limited exemption from income tax applied to dividends paid by Dutch companies. The first question for the court was whether the availability of this exemption only in relation to dividends paid by Dutch companies was inconsistent with Article 1 of the 1988 Directive. The Court held that the acquisition of shares in an undertaking was a capital movement within the scope of the Directive, that the receipt of dividends on such shares presupposed participation in the undertaking, and that therefore the receipt by a resident of one Member State of dividends on shares in a company resident in another Member State was within the Directive. It then held that the refusal of the first Member State to extend to such dividends an exemption from tax which applied in relation to dividends on shares in companies resident in the first Member State had the effect of dissuading nationals of the first Member State from investing their capital in companies established in other Member States and, conversely, presented an obstacle to such companies from raising capital from residents of the first Member State. Accordingly, such a provision constituted a restriction on the movement of capital prohibited by Article 1 of the Directive.

It seems to me that the ACT provisions are analogous in effect to the provision of Dutch law at issue in Verkooijen. If, therefore, the ACT provisions are to be regarded as restrictions on capital movements, Mr Glick seems to have substantial grounds for saying that, though they are within Article 56, they are also within the protective effect of Article 57.1. Park J accepted that submission and held that there was no sufficient doubt about it for it to be appropriate to refer a question about the application of the articles. Mr Glick supports that position.

Mr Aaronson accepts, as he must, that Verkooijen holds that a discriminatory tax provision as regards dividends is a restriction on capital movements, but he says it is also a restriction on payments, and only because of its indirect effect does it come within the prohibition of restrictions on capital movements as well. Of course, the effect of it being categorised in that way in Verkooijen was to invalidate the provision whereas, according to Mr Glick, such categorisation in the present case would have the opposite effect. Mr Aaronson submits that the purpose of Article 57.1 is to preserve the direct and indirect effect of a number of strictly limited provisions, which must be aimed at capital movements, and in particular at investment in undertakings, and that, where the dividend arises from capital movements which are not themselves restricted, there is no justification for an indirect restriction such as the ACT provisions constitute.

Without going deeply into the area of debate, it seems to me that, if the decision of this court as to the position under UK law were the last word on that point, it would be appropriate to refer to the European Court of Justice the question whether the ACT provisions constitute a restriction on capital movements or on payments or both, under Article 56, and, depending on the answer to that, whether they are within the scope of Article 57.1 so as to be of continuing effect despite Article 56. It does not seem to me to be altogether clear that Article 57.1 does apply so as to save the restriction, if that is what it is.

As it is, it may be that the decision of this court is not the last word on the position under UK law, because the Appellants will seek and may obtain permission to appeal to the House of Lords. It is therefore necessary to consider whether a reference ought to be made at this stage, even if the case may proceed to the House of Lords on the questions under the Double Taxation Conventions.

Mr Aaronson submits that a reference ought to be made now in any event, and that it can be made on a contingent basis as to what UK law may be under the Double Taxation Conventions and section 788, and therefore what may constitute the restriction under national law to which Article 57.1 might apply. He therefore invites the court to make a reference, and (unusually) to order, under CPR 68.4, that the proceedings should not be stayed as regards the remaining issues, which involve questions of domestic law. Mr Glick submits that it cannot be sensible to make a reference now rather than wait until the House of Lords has ruled on the other points in the case, because until then it will not be known whether there is any, and if so what, restriction under national law.

In this context he pointed out that the Appellants’ arguments include saying that, even if they are right on the interpretation of the Double Taxation Conventions and the effect of section 788, so that the non-discrimination article does override other inconsistent provisions of ICTA, such as the limitation to UK resident parent companies of the ability to join in a group income election, nevertheless the way in which the legislation was enacted and applied constituted a restriction under national law. He resisted that submission, but relied on it being made as justifying leaving it to the last stage of the domestic court procedure before referring to the European Court of Justice any question as to Articles 56 and 57.

Mr Aaronson supported his argument for an immediate reference both by reference to the desirability of getting on with the two aspects of the case at the same time, rather than successively, and also in reliance on the fact that a case has fairly recently been referred to the European Court of Justice from Austria which raises issues under Articles 56 and 57 (Case C-157/05 Holböck v. Finanzamt Salzburg Land), and in the hope that if the present case were referred now, it might either be heard together with that case or at least the fact that both were pending at the same time might have the effect that the consideration by the court of the one case would inform its consideration of the other.

At first sight, it might be thought that the questions arising on Articles 56 and 57 could be decided without knowing what the nature of the restriction is under national law which is said to be overridden by Article 56 and not justified by Article 57, so long as whatever restriction there is already existed on 31 December 1993. That is not supported by the way in which first drafts of the questions to be referred, produced by Mr Aaronson’s team at the court’s request, were framed. They propose a number of variations for the court, depending on the nature and basis of the restrictions.

It seems to me that, formally, the question whether a reference should be made at this stage, ought to wait until it is clear whether the case is to proceed to the House of Lords in any event. I would therefore leave the question of a reference by this court to be dealt with after any application for permission to appeal to the House of Lords. However, since the point about the timing of a reference has been argued, and to avoid a further debate at a later stage, I would say that, if the case is to proceed to the House of Lords, it would not be appropriate to order a reference at this stage. I accept that this would result in a delay before the case could come to a final conclusion, having regard to the time that may be necessary (a) for an appeal to the House of Lords and (b) thereafter for a reference to the European Court of Justice and (c) for the final determination of the case in the light of the answers given by the European Court of Justice. Nevertheless it seems to me that, particularly so in the present case where there are several points of UK law to be argued which are relevant to whether there is any restriction under national law, it would be undesirable to refer a question to the European Court of Justice on alternative bases, depending on the answer given in the House of Lords to a number of points of UK law.

For those reasons, for my part I hold that the judge was right on the question of the interpretation of the Double Taxation Conventions, and on the effect of section 788, so I would dismiss the appeal as regards those issues. I also consider that he was right not to refer a question under Articles 56 and 57 to the European Court of Justice, though not for the reason that he gave. If this case is to proceed further, to the House of Lords, on the questions of domestic law, I would not refer such a question at this stage. Otherwise I would do so.

I have read the judgment of Mummery LJ. I agree with him on the pleading point. Like him I would allow the Respondents’ cross-appeal on this point. I also agree with him that on the facts of these cases there was no unlawful demand, and that it is not open to us to hold that the Appellants have a claim at all if they cannot show that the tax was paid pursuant to an unlawful demand.

He considers that, if the Appellants could get over those hurdles, and all others in their way, they would be entitled to a restitutionary remedy in respect of utilised ACT, which they describe as a claim for the time value of money. I see the justice of such an entitlement, for the reasons that he mentions. But a similar entitlement, and even one calculated on the principles of compound interest, has often been regarded as merited by courts which have found themselves unable to recognise it as legally sound. If it were necessary for us to decide whether there is such an entitlement, I would find it a difficult matter to reconcile the decisions said to be relevant, several of them in the House of Lords.

Conceptually, the claim is restitutionary and seeks that the Revenue disgorge the benefit that they have had through the early payment of the utilised ACT. In practice, it is a claim which can only be measured by the same processes as are used for the calculation of interest. Such a claim is available, in certain circumstances, in equity, but this is not an equitable claim. In Westdeutsche Landesbank v. London Borough of Islington [1996] AC 669, where the claim was for the unravelling of interest rate swap transactions which had been held to be ultra vires, the Court of Appeal awarded compound interest on the basis of equity, but the House of Lords found that that basis of award was not available. Lord Goff and Lord Woolf would, even so, have awarded compound interest, but the majority would not do so, not least because there was a statutory jurisdiction to award interest, but only simple interest, under the Supreme Court Act 1981, section 35A, replacing the provision made by the Law Reform (Miscellaneous Provisions) Act 1934, section 3. It does not appear to me to be consistent with that decision to find a way of awarding restitution of the time value of money by reference to general principles of restitution. It is true, as Mummery LJ says, that this point was not directly at issue in that case. It is also true that, as the Court of Appeal later pointed out in I.M. Properties plc v. Cape & Dalgleish [1999] QB 297, the House of Lords, when awarding interest under section 35A, overlooked the fact that their award included interest on a sum which was not due at the commencement of the action, which the section does not permit: see Waller LJ at [1999] QB 463 A-E, and Hobhouse LJ at 465H to 466C. It is the case, what is more, that there were unforeseen time constraints in the argument on Westdeutsche Landesbank, as described at [1996] AC 679B-C and 691H, which might account for the overlooking of a point not central to the appeal.

It is established that section 3 of the 1934 Act, and now section 35A of the 1981 Act, apply to restitutionary claims. Goff J, as he then was, so decided in BP Exploration (Libya) Ltd v. Hunt (No. 2) [1979] 1 W.L.R. 783, and the Court of Appeal ([1981] 1 W.L.R. 232) and House of Lords ([1983] 2 AC 352) upheld the award. The entitlement which the Appellants claim would go further than that jurisdiction in three respects: first, it would be a matter of right, not of judicial discretion; second, it would not be limited to simple interest, or to a calculation equivalent to simple interest; third, it would not be limited to cases where a principal sum is due at the commencement of the proceedings.

In Wadsworth v. Lydall [1981] 1 W.L.R. 598, the Court of Appeal held that the provisions of the 1934 Act did not preclude the recovery of interest charges incurred by the Claimant by way of special damages for breach of contract, if the second limb of Hadley v. Baxendale (1854) 9 Exch 341 was satisfied in the relevant respect. That case was approved in Westdeutsche Landesbank. Mr Cavender submitted for the Appellants that what the Appellants claim is, similarly, outwith the scope of the common law rule recognised by London Chatham and Dover Railway Co v. South Eastern Railway Co [1893] AC 429 and by The Pintada [1985] AC 104. I see the logic of his contention, but it is striking that the point has not previously been taken. There may not have been a recent case in which the absence of a principal sum due at the outset of the proceedings made it necessary to consider the point (accepting as having been overlooked the extent to which this was a feature of the Westdeutsche Landesbank case itself) but the argument would, in principle, justify a calculation on the basis of compound interest, rather than just simple interest. Whether such an award was possible was the very point that was argued in Westdeutsche Landesbank. I would hesitate, to say the least, before holding to be available a basis for recovery of what is, in effect, interest, that could justify calculation on the principles of compound interest, which had not, it seems, been thought of by any of the members of the House of Lords who heard that case, nor by Counsel who argued the case. I am, therefore, relieved not to have to decide the point.

I agree with the conclusions set out by Lloyd LJ at §83 and with his reasons for arriving at them. I also agree with Mummery LJ that the claims fail in the absence of an unlawful demand. I agree too, for the reasons given by Mummery LJ, that the respondents’ cross-appeal on the pleading point succeeds.

Whether a freestanding cause of action for the time value of money exists seems to me, as to the other members of the court, highly problematical. In principle, if money has been unlawfully tied up the wrongdoer ought to compensate the person entitled to it for its temporary loss. In practice, this is what awards of interest do. One reason why the present claims lack this derivative element is that no compensable wrong has been done upon which a claim for interest can be built. If there is a wrong, it is the simple tying up of the claimants’ money.

It is relevant that both the absence of an unlawful demand and the want of a primary cause of action in cases like the present are functions of the taxation process itself: the taxpayer accounts for what he is advised is due, and if he thinks he is being overtaxed he appeals. Success in an appeal does not mean that the Revenue has made an unlawful demand in the Woolwich sense of relying on ultra vires powers of recovery and enforcement. It means simply that the taxpayer’s liability is less than the Revenue claims, and that if he has overpaid he can have his money back with whatever interest the law allows him.

Since the present claims seek to change all that, I take the liberty of commenting on some of the issues of legal policy which they throw up.

The route provided by law for a taxpayer who considers that he has been improperly taxed is a statutory appeal brought within a prescribed time. The reason given by Mr Aaronson for the failure of the present appellants to take this route, and for their bringing instead actions for restitution many years later, is that at the time nobody thought his case arguable. This, he contends, was a mistake of law which permits a claim now to be made for restitution – though of what is itself problematical, as Mr Glick has demonstrated.

There is something wrong with this syllogism: indeed there has to be something wrong with it if our system of law is to function coherently.

One thing which is wrong with it is, I believe, an inapposite use of the concept of unjust enrichment. True enough, the relationship between the taxpayer and the Crown, in whose right taxes are levied, is governed by fiscal laws, so that any impermissible taxation can be said to be based on a mistake of law. But it is not a relationship in which either party can relevantly be said to have been unjustly enriched by an underpayment or overpayment of tax. The obligation to pay taxes is an obligation of public law, and the collection of them is a function of the state. There is, I accept, no doctrinal or other reason for exempting state organs from the law of restitution: see, most cogently, Wilson J’s dissent in Air Canada v British Columbia (1989) 59 DLR (4th) 161. But a body collecting revenue on behalf of the state is required to disgorge unlawful imposts if, and only if, an independent appellate tribunal (absent which there would almost certainly be a violation of the European Convention on Human Rights) holds that it has no entitlement to the money. By parity of reasoning, the taxpayer is not called upon to make restitution of underpayments which the Revenue has failed at the proper time to demand. Failing such a decision, taxes are to be treated as properly levied. Arguably there has in such a situation been no relevant mistake of law and no unjust enrichment of the Crown or – in the latter example – the taxpayer. What would attract the remedy of restitution is a failure by the Crown to disgorge overpaid tax after an appellate ruling in the taxpayer’s favour.

Beyond this lies the question of time. The Limitation Act 1980 makes no provision for a time limit on actions for restitution. Rather, therefore, than unacceptably claim a limitation-free cause of action, litigants and the courts have in effect agreed that restitution claims are to be subject to the statutory time limit for contract actions (see Re Diplock’s Estate [1948] Ch 465; Kleinwort Benson Ltd v Sandwell BC [1994] 4 All ER 890). This is something of a paradox, since restitution is intended to deputise for contractual rights, not to enhance them. But it is by adopting this fiction that the present appellants are now pursuing a surrogate remedy against what they say is unlawful double taxation well beyond the date at which their statutory rights of appeal expired.

No argument was directed in this court to the question whether it is an abuse of process to ignore the primary mode of challenge provided by law and instead to bring an action which evades the controls on that mode. That it is ordinarily an abuse to do so was decided by the House of Lords in O’Reilly v Mackman [1983] 2 AC 237. There the claimants, instead of seeking judicial review of some prison adjudications, had proceeded by action or originating summons for declaratory relief. This had avoided, among other procedural constraints, the narrow time limit for judicial review. Since O’Reilly v Mackman was decided, occasional issues have arisen which could legitimately be pursued either by judicial review or by action, and the rigidity of the rule has accordingly been relaxed: see Wandsworth LBC v Winder [1985] AC 461; Roy v Kensington and Chelsea FPC [1992] 1 AC 624. But the fundamental principle remains intact, and I see no reason at present why it should not apply to cases like those now before us. 

I am reinforced in this tentative view by two things. One is that (as I suggested in Niru Battery Mfg Co v Milestone Trading Ltd (No.1) [2003] EWCA Civ 1446, §192) the law of restitution is a means of distributing loss among parties whose rights are not met by some stronger doctrine of law. It is in its nature, in other words, a residual remedy. The other is that the opinions of the House of Lords in Autologic Holdings [2005] UKHL 54, delivered the day after the conclusion of argument on these appeals, appear to endorse the proposition that an action at law designed to circumvent appeal to the Commissioners is in principle an abuse of the court’s process.

Their Lordships by a bare majority allowed the Commissioners’ appeals on the grounds given by Lord Nicholls, to which I will come in a moment. But it is also relevant that the dissenting opinions of Lord Hope and Lord Walker were based on the view that the actions (by contrast, it might be thought, with those before us) were not surrogate appeals on the companies’ liability to taxation but were claims as to the incompatibility of their admitted liability with EU law. But there was, as I read the speeches, unanimity as to what Lord Nicholls held (at §13):

“[T]he conclusion that the proceedings are an abuse follows automatically once the court is satisfied the taxpayer’s claim is an indirect way of seeking to achieve the same result as it would be open to the taxpayer to achieve directly by appealing to the appeal commissioners.”

We do not know whether the present appeals will go further, but if they do, I would observe that abuse of process appears to be an issue not only for the parties but for the court: see Hunter v Chief Constable of the West Midlands Police [1982] AC 529, per Lord Diplock at 536C-D.

 I stress that my observations are made without the benefit of argument and are in no way dispositive of the present appeals. These fail for the reasons given by Lloyd LJ. But I may perhaps add one further observation. The growing specialisation of sectors of the legal profession, particularly the Bar, is constricting and in some instances fragmenting the sources of law which are deployed in argument. Planning cases on issues of natural justice, for example, are repeatedly argued by reference only to other planning cases, when more often than not a body of wider and more useful principle exists. The argument on abuse of process at the bar of the House in Autologic Holdings appears to have been based entirely on other tax cases. The law of restitution is at risk of going down a similar path of self-referential jurisprudence, with consequences, among others, of the kind addressed in the earlier part of this judgment. Such habits can retard and even reverse the growth of the law as a consistent body of principle. Professor Burrows’ warning about this – ‘We do this at common law but that in equity’ [2002] OJLS 22 - deserves to be heeded.

I agree with Lloyd LJ. The appeal should be dismissed on the substantive issues of construction arising on the non-discrimination provisions of the Double Taxation Conventions and on section 788 ICTA. I also agree with Lloyd LJ’s conclusions on Articles 56 and 57 of the European Treaty.

As Park J held that the relevant provisions of the Double Taxation Conventions, so far as they applied to ACT, were not given effect in UK domestic law, it was unnecessary for him to reach any decision about the availability of domestic remedies for non-compliance with them.

In case, however, Park J was wrong on the issues covered in the judgment of Lloyd LJ, the parties made detailed submissions about important issues on domestic remedies: whether the appellants are entitled to restitution on the basis that payment of ACT by them was pursuant to an unlawful demand by the respondent or, alternatively, under a mistake of law; and whether the appellants are entitled to maintain claims for interest on, or for restitution for loss of the use of, the sums paid as ACT which were later set off against liability to MCT.

For the appellants the points were argued by Mr David Cavender, Mr Graham Aaronson QC having concentrated on the substantive points covered by Lloyd LJ in his judgment. For the respondent Mr Ian Glick QC argued all points, substantive, procedural and remedies, save for the pleading point, which was argued by Mr David Ewart.

It is unusual to argue pleading points in proceedings that have been on foot for 4 years or more. But this is not ordinary litigation. The pleading disputes were dealt with by Park J very briefly in paragraph 71 of his judgment. They are brought to this court by way of cross appeal in the respondent’s notice. They were not at the forefront of the skeleton arguments or of the oral submissions. Being of a technical character, arguments about amending the pleadings do not generate the same interest as novel arguments about the law of restitution.

Nevertheless, the pleadings matter. The pleadings are the starting point for defining the issues and identifying the causes of action, which can affect the nature of the domestic remedies available to a claimant complaining of overpaid or prematurely paid ACT. In this case the disputes about the causes of action available to the appellants and about their pleadings shed light on the type of relief to which they may be entitled.

Although, in the light of the court’s conclusions on the substantive questions of construction, it has become strictly unnecessary to do so for the purpose of disposing of this appeal, this court ought, in my view, to deal all the points argued by the parties rather than shelve them.

In ordinary circumstances, I would be firmly opposed to expressing opinions on legal issues that do not need to be decided in order to resolve a case or dispose of an appeal, especially when the issues arise in a developing area of the law, in which it is advisable to tread cautiously. The considerable amount of additional work involved in venturing into these developing areas has had the unfortunate consequence of lengthening the time that it has taken me to finalise this judgment. This has held up the disposal of the appeal, which could have been achieved much earlier on the basis of the lead judgment written by Lloyd LJ on the substantive issues and agreed by Sedley LJ and myself months ago.

This litigation is, however, exceptional in my experience. A huge amount of time and money have been spent by the parties on preparing and presenting full arguments on the pleading and remedies points. The arguments have given the court an overview of the ACT group litigation. The litigation arising from the Hoechst case is by no means confined to the particular appeals before this court or even to ACT. There is a mass of ongoing litigation by hundreds of multi-national groups of companies. Colossal sums of money running into many billions of pounds are involved. Courts at different times and at different levels have decided, and will probably continue to decide, inter-related issues. Many of the more important questions either have been or will be appealed to the House of Lords. Issues of EU law may in due course be referred to the Court of Justice for interpretative rulings. The outcome of further appeals and references may mean that it will become necessary to decide, or at least to have some regard to, domestic remedies issues at some stage in the litigation.

A general idea of the scale of the litigation arising from the ruling in Hoechst and the difficult procedural and other problems to which it has given rise can be gathered from the decision which was handed down by the House of Lords in Autologic Holdings plc v. IRC [2005] UKHL 54, [2005] 3 WLR 339 soon after the end of the hearing of this appeal and in which reference is made to this pending appeal (see paragraph 98). Although that case is concerned with group loss relief, it also involves questions of principle, procedure and remedies arising from the alleged unlawfulness of UK corporation tax legislation restricting fiscal group reliefs or advantages to cases where the parent and relevant subsidiary company in the group are resident in the UK, the extent to which the restriction may be inconsistent with the EC treaty and/or non-discrimination articles in Double Taxation Conventions with other states incorporated into UK law, and the inter-relationship of the ordinary statutory procedures prescribed for the determination of tax disputes between the revenue authorities and the taxpayer, including allegations of contraventions by national rules denying group loss relief. Those proceedings, like these cases, include claims for restitution based on alleged unjust enrichment of the public purse at the expense of the taxpayer and /or damages in respect of payments of ACT allegedly wrongly made under the UK tax regime. The majority in the House of Lords held that the taxpayer had to use the remedies provided by the legislation (paragraph 13) and to follow, where it was available, the statutory procedure in which full effect could be given to, for example, directly enforceable Community rights.

Although I am well aware of the risks involved in expressing views on points of law that do not necessarily arise for decision in the particular hearing, the time has come in this case for the court to take stock of the overall situation and to consider the potential difficulties which, according to the revenue authorities, face the claimant taxpayers, even if they were to succeed on substantive issues, such as the incorporation of the non-discrimination articles into UK law. The pleading points need to be addressed in detail now. Although it is not necessary to deal with all the points on remedies in the same detail, what is said by this court now on the question of remedies generally may be relevant to the future course and conduct of the litigation.

Unlike the parties’ written and oral submissions I will begin at the beginning by taking the pleading points first. It is instructive to look at the way in which the claims were originally advanced.

Considering the size of the sums at stake and the novelty of the questions raised, the original pleadings were uninformative. Claims were advanced on the basis that the claimants were obliged to pay ACT, which would not have been payable, had the parent companies been resident in the United Kingdom. Thus, it was alleged in NEC’s original Claim Form dated 12 January 2001 that-

“2.

Had the Second Claimant [NEC Corporation] been a UK resident company, the payment of the said dividends would have carried a tax credit pursuant to the advance corporation tax (“ACT”) provisions of the Income and Corporation Taxes Act 1988 (“ICTA”) (namely sections 14, 208, 231 and 247 and Schedule 13).

3.

[Human rights claims pleaded, but no longer pursued].

4.

The First Claimant was obliged to pay advance corporation tax in the circumstances set out in paragraph to [sic] above because its parent entity, the Second Claimant, was a company resident in another nation. Had the Second Claimant been a company resident in the United Kingdom, the credit referred to in paragraph 2 would have applied and no advance corporation tax would have been payable…”

The relief claimed included a declaration that the ACT provisions of the ICTA, as they applied to the claimants, were contrary to the Double Taxation Convention and were illegal. Damages were claimed for breach of that Convention, together with statutory interest and compound or other interest “pursuant to the rules of equity.”

The original Claim Forms did not expressly identify any cause of action or plead any relief by way of restitution or compensation based on allegations that payments of ACT by the appellants/claimants were made either pursuant to an unlawful demand by the respondent or under a mistake of law on the part of the appellants. No allegation that the appellants were unaware that they did not have to make the payments of ACT, or that they were unaware that they could have made a group income election or that, if they had been aware that they could make a group income election, they would have done so, featured in the original pleadings.

Indeed, the original pleadings were totally silent on all of the following matters which are, in my view, reasonably relevant to the claims now advanced: the whole question of group income elections as a means of not having to account to the respondent for ACT on dividends paid; the alleged effects of breaches of the non-discrimination articles in the relevant Double Taxation Conventions on the right of a UK resident subsidiary with a parent resident in the relevant State to make a group income election; and the differential treatment amounting to a breach of Article 56 of the EC Treaty in restricting the parent’s freedom to move capital.

There were some minor variants in the Claim Forms of the different appellants. For example, the claim by Boake Allen dated 6 March 2001, though very similar to that of NEC, mentioned “Restiution [sic]” in paragraph b) of the prayer.

Although someone reading the Claim Forms would gather that the appellants considered that they should be compensated because, as a result of discriminatory legislation, they had been obliged to pay tax which would not have been payable by a subsidiary of UK parent company, neither the causes of action potentially available nor the material facts necessary to support them were pleaded.

The need to reform the pleadings was recognised in the amendments made by the appellants pursuant to the consent order of Park J on 11 July 2003. The appellants were permitted to amend their Details of Claim in accordance with a general pleading entitled “Group Particulars of Claim.” As appears, however, from the 2^nd recital to the Order express limits were placed on the effect of the amendments allowed under the order-

“AND UPON the Claimants having undertaken to accept that any claims added by amendments deemed to have been made by paragraph 2 of this Order and any claims added by amendments made in accordance with paragraph 4 of this Order shall be deemed to have been commenced on the 4^th July 2003 unless the Court subsequently finds that:

(i)

the proposed amendments do not have the effect of adding new claims within the meaning of section 35 of the Limitation Act 1980; or

(ii)

the new claims added by the proposed amendments arise out of the same or substantially the same facts as a claim in respect of which the relevant Claimants have already claimed a remedy

in which case the claims added by the amendment shall be deemed to have been commenced on the same date as the respective original claim.”

The amendments amplified the original pleaded case by alleging that, if the foreign parent had been resident in the UK, the foreign parent and the ACT payers would have made an election under section 247 ICTA and that, as a consequence of such election, no ACT would have been payable by the UK subsidiaries or the UK sub-subsidiaries. The amendments to the prayer for relief specifically and concisely claimed restitution of (and/or compensation or damages for) the ACT payments made by the ACT payers “pursuant to a mistake of law or demands by the defendant and the ACT provisions, those provisions being contrary to Article 56 of the EC Treaty, the Article of the Double Taxation Convention and the Human Rights Convention.”

The respondent argued before Park J and on this appeal that the appellants’ amendments added new claims based on unlawful demands and on mistake of law and that the new claims did not arise out of the same, or substantially the same, facts as a claim in respect of which the appellants had already claimed a remedy. The respondent therefore contends that the claims added by amendment are not deemed to have been commenced on the same date as the respective original claims.

The respondent submitted that the claims originally pleaded were based solely on non-compliance with the non-discrimination articles in Double Taxation Conventions resulting in the payment of tax, which would not have been payable in the case of a subsidiary of a UK parent. The appellants had now shifted the basis of their claims to two causes of action, which were not originally pleaded: (1) payment of ACT pursuant to an unlawful demand and (2) payment of ACT under a mistake by the appellants. “Unlawful demand” and “mistake” are or involve making factual allegations. In order for the claims to succeed it is necessary to prove contested facts by evidence. Neither the fact of a demand nor the fact of a mistake, which are central to the new causes of action, were originally pleaded. The respondent accordingly contended that the amendments introduced new causes of action which are not based on the same facts, or on substantially the same facts, as those originally pleaded. They are new causes of action based on new facts and are deemed to have commenced on 4 July 2003.

The claim based on breach of Article 56 is also, according to the respondent’s submissions, a new cause of action. It is alleged that the group income election provisions were a restriction on the freedom of movement of capital and/or payments. A “restriction” is a material fact, which needs to be pleaded and, if disputed, established by evidence. It was not mentioned at all in the original pleadings.

The respondent submitted that Park J erred in his conclusions on the pleadings issues on these points in paragraphs 71(ii) and (iii) of his judgment-

“71.

Because of amendments to pleadings sought to be made by the claimants a point has been raised on whether certain arguments now advanced arise out of the same or substantially the same facts as the arguments originally pleaded. I think that the parties will be content if I simply state my conclusions on this without prolonging the case by explaining the entire procedural and legal background. It is in essence the same as that which I described in paras 22 to 25 of my judgment in the Hoechst case (see the table in para 1 at the beginning of this judgment). My conclusions are as follows.

(i)

The word “without limitation” at a point in the amended pleading are intended to enable the claimants to rely on dividends other than those which were specified in the original pleadings. In that respect the amendments would not arise out of the same or substantially the same facts. [This ruling is not appealed by the claimants].

(ii)

In the amended pleading there is an express allegation that, if the claimants had known that they were entitled to join with their parent companies in group income elections, they would have done so. This was not specifically alleged before. However, in my opinion it does arise out of the same or substantially the same facts. My reason is that, in my view, it is a factual consequence which is so obvious that it would be understood even if not specifically pleaded in so many words.

(iii)

A related point is that the amended pleadings allege that the claimants and their parent companies did not know that they were entitled to make group income elections, and it was by reason of a mistake on their part that they did not know that. This seems to me to arise out of the same or substantially the same facts as those originally pleaded: it is inherently obvious, and in my view it would be pedantic obscurantism to insist that it be separately pleaded.”

Park J accordingly made declaration (2) in his order of 17^th December 2003 that, inter alia, the following claims permitted to be added by the order of 11^th July 2003 arose out of substantially the same facts as a claim in respect of which the claimants claimed a remedy, namely: (i) amendments which added the claim that, if the claimants had known they were entitled to join with their parent companies in group income elections they would have done so; (ii) amendments which added the claim that it was by reason of a mistake on their part that the claimants and their parent companies did not know they were entitled to make income elections; and (iii) amendments to introduce to the claimants’ pleadings claims under Article 56 EC Treaty.

This court is normally reluctant to interfere with case management decisions, such as amendment of pleadings, taken in the exercise of a discretion at trial level, especially in litigation of this magnitude and complexity organised under a group litigation order and supervised by a very experienced judge in the relevant area of law, in this case Park J in revenue law matters. The respondent’s criticisms of the declaration made by Park J are not, however, in the same class of discretionary case management decisions, such as whether the appellants’ proposed amendments should be permitted. Interference with the exercise of discretion would be confined to errors of legal principle and to decisions which, for some other reason, are plainly wrong.

The pleading issue before Park J was concerned with a dispute about the limitations on the effect of the amendments, which could be made under the terms of the general order, namely whether they added new claims and whether the claims arose out of the same or substantially the same facts as those originally pleaded. That is not so much a matter of discretion as one of the construction of the order and its application to the amendments made under the order.

In order to deal with the respondent’s submissions on the amendments to the claims it was necessary for Park J to compare the claims and the factual allegations in the original pleadings with the claims and factual allegations in the amended pleadings. The questions for decision were whether new claims were added by the amendments and, if so, whether they arose out of the same facts or substantially the same facts. See Savings and Investment Bank v. Fincken [2001] EWCA Civ 1639 at para 30 and Paragon Finance plc v. DB Thakerar & Co [1999] 1 All ER 400.

Mr Cavender, who argued this part of the case for the appellants, submitted that the basic facts were pleaded in the original pleadings, which should be viewed in the context that this whole matter is at the cutting edge of a whole range of different issues. The court should, he said, take “a commercial view” of the pleadings. He cited Lloyds Bank v. Rogers (The Times March 24 1997) in support. On that approach he contended that the basic factual elements of the claim were pleaded from the start. It was inherent in the matters pleaded that the appellants were mistaken in making the payments of ACT and in not making a group election. The amendments only added further information about pleaded claims, which the respondent could have asked the claimants to supply on the original pleaded case.

While it is good sense not to be pernickety about pleadings, the basic requirement that material facts should be pleaded is there for a good reason - so that the other side can respond to the pleaded case by way of admission or denial of facts, thereby defining the issues for decision for the benefit of the parties and the court. Proper pleading of the material facts is essential for the orderly progress of the case and for its sound determination. The definition of the issues has an impact on such important matters as disclosure of relevant documents and the relevant oral evidence to be adduced at trial. In my view, the fact that the nature of the grievance may be obvious to the respondent or that the respondent can ask for further information to be supplied by the claimant are not normally valid excuses for a claimant’s failure to formulate and serve a properly pleaded case setting out the material facts in support of the cause of action. If the pleading has to be amended, it is reasonable that the party, who has not complied with well known pleading requirements, should suffer the consequences with regard to such matters as limitation.

On the correct approach to the construction and application of the terms of the group order the judge would, in my view, have been bound to conclude that the amendments added new claims, not just details about pleaded causes of action, and that the new claims did not arise out of the same, or substantially the same, facts as a claim in respect of which the relevant claimants had already sought a remedy.

Material facts relating to the overpayment or premature payment of ACT pursuant to an alleged unlawful demand or under a mistake were pleaded for the first time in the proposed amendments. On the appellants’ own case as to the relevant law, the fact of an unlawful demand or the fact of a payment under a mistake are material facts in the claims for restitution based on unjust enrichment. The amendments introduced the new facts of demand and mistake necessary to establish a claim for restitution based on unjust enrichment. The new facts alleged would be relied on as the reasons why there was “unjust” enrichment of the respondent and why the appellants were entitled to restitution of the benefits conferred on the respondent in consequence of payments of ACT unlawfully demanded or mistakenly made.

Before the amendments were made the claims were for compensation for non-compliance with the relevant discrimination provisions of the Double Taxation Conventions, the loss being the payment of ACT when it was not payable. The amendments were not just new instances of those particular claims already raised or for a new remedy arising out of the pleaded facts or substantially the same facts. If the claims were based on unlawful demands for payment and upon mistaken payments of ACT, the respondent should have been notified in the pleading that there was an alleged unlawful demand and what it was and that there was an alleged mistake and what it was.

It follows that, on the proper construction of the consent order, the appellants’ claims are to be treated as having been made when the amendments were made and that they are not deemed to have been made at the time of the original claim forms commencing the proceedings. I would set aside the relevant declarations made by Park J in relation to the amendment of the pleadings and make contrary declarations in their place.

On the assumption that the appellants are entitled to succeed on the substantive issue of non-compliance with non-discrimination convention obligations incorporated into domestic law, the respondent submitted that the effect of section 788 ICTA and the statutory instrument implementing the relevant Double Taxation Convention would have been that a UK subsidiary with a parent resident in a country covered by the Convention was entitled to make a group income election under section 247 jointly with the parent.

The respondent contended that it would follow that the appellants were not obliged to pay the ACT in the first place. Like subsidiaries of a UK parent, the UK resident subsidiary of a parent company covered by the Double Taxation Convention could have made a group income election under section 247 ICTA by way of a claim to the Board under section 788(6) ICTA. The election would be made jointly with the parent resident in the country covered by the Convention.

If the Board had refused to accept the election, the appellants could have appealed against that to the General Commissioners or to the Special Commissioners. Until the appeal was determined in their favour, any dividend was subject to ACT, but that was not discriminatory or unlawful, as the same was also true in the case of a UK subsidiary with a UK resident parent. The right to make a group income election was available in the case of the appellants and of UK resident parent companies alike. The appellants were in the same position as a wholly resident UK group. If they had made such an election, the subsidiaries could have paid dividends to their parent companies free of ACT. If they paid the dividends without making a group income election, the ACT would have been due and payable and there would be no basis for seeking restitution of the ACT paid by them.

The appellants do not accept that this analysis of the statutory regime precludes them from making restitutionary claims in respect of the ACT paid by them. Their case is that the payments of ACT made by them were unlawful and that they are recoverable as having been made either pursuant to an unlawful demand or under a mistake of law. I will summarise the rival arguments in respect of each cause of action in turn.

In Woolwich Equitable Building Society v. IRC [1993] AC 70 the majority in the House of Lords held that money in the form of taxes or other levies paid by a citizen pursuant to an ultra vires demand of a public authority is prima facie recoverable by the citizen as of right. The principle applies whether or not the citizen has paid the money under a mistake and, if there has been a mistake, whether it is one of fact or of law: see Lord Goff at p177F-H.

The claim for the recovery of tax paid in response to an unlawful demand is relied on by the appellants in the amended pleadings on the basis that the non-discrimination articles in the Double Taxation Conventions were incorporated into English law.

Lord Browne-Wilkinson described the unlawful tax demand situation at p197C-D as “the paradigm of a case of unjust enrichment”-

“In the present case, the concept of unjust enrichment suggests that the plaintiffs should have a remedy. The revenue demanded and received payment of the sum by way of tax alleged to be due under regulations subsequently held by your Lordships’ House to be ultra vires. The payment was made under protest. Yet the revenue maintains that it was under no legal obligation to repay the wrongly extracted tax and in consequence is not liable to pay interest on the sum held by it between the date it received the money and the date of the order of Nolan J. If the revenue is right, it will be enriched by the interest on money to which it had no right during that period.”

The same principle was applied by the Court of Appeal in British Steel plc v. Customs & Excise Commissioners [1997] 2 All ER 366 to a case of demand based on mistaken view of lawful regulations or a mistaken view of the facts.

The respondent submitted that a Woolwich claim based on an unlawful demand is not available to the appellants, as there was no demand, let alone any unlawful demand, for payment of the ACT, whether by way of assessment or otherwise. The appellants are not therefore entitled to recover the ACT paid by them.

ACT was not collected through the machinery of demand. The procedure was that the appellants made returns of the ACT due for the relevant quarters under schedule 13 ICTA and paid the ACT. In the absence of a group income election, they paid ACT which was lawfully due in accordance with their returns in respect of the dividends to which they related.

The appellants’ response was that the requirement of a “demand” in Woolwich should not be approached in an unduly narrow or technical way. The court should look at the justice of the situation in which the ACT was paid. ACT was not a gratuitous voluntary payment. It was paid by the appellants in response to the statutory regime. The demand to pay the ACT was to be found in the legislation itself. Through the combined effect of sections 14 and 247 of ICTA the legislation provided the necessary demand to the paying company making a qualifying distribution. Where a qualifying distribution is made the paying company “shall be liable to pay” an amount of corporation tax (section 14). According to the terms of the legislation, a joint election to avoid paying such ACT could only be made where both the paying company and the receiving company were resident in the UK. Companies in the position of the appellants were not in a position to make the election. If such companies had made an election, the Board would have refused to accept it, whereas they would not have refused to accept such an election from eligible UK companies. Thus, in practice, companies in the position of the Appellants would not have been given the same treatment as UK companies, even though they would have been entitled to it.

This analysis was disputed by the respondent on the ground that the legislation only “demanded” the ACT when the paying and receiving companies did not make the election, which the legislation entitled them to make. There was nothing unlawful in that. The appellants were in the same position as a wholly resident UK group if they were correct in their submissions on the effect of the incorporation of the non-discrimination provisions of the Conventions so that they became part of the tax code taking precedence over inconsistent provisions in the code. The effect of section 788 and the statutory instrument was to disapply the requirement in section 247 that the receiving parent company be resident in the United Kingdom. That meant that nothing discriminatory was left in the statutory regime. A UK resident subsidiary with a parent company resident in a country covered by the Convention would have been entitled to make a group income election jointly with its parent under section 247. There was no unlawful discrimination. If no election was made, the “demand” for ACT, even assuming that there was one, would have been lawful.

The appellants submitted, in the alternative, that they have a restitutionary cause of action based on a mistake of law. They were acting under a mistake of law when they made the payments of ACT. Their mistake was that they believed that section 247 was lawful and that ACT was payable when, on the correct interpretation of the provisions, the ACT was not payable. As a result of the mistake they believed that they were not entitled, as subsidiaries of foreign resident parent companies, to avoid the obligation to pay ACT on distributions due pursuant to section 14 ICTA by making an election under section 247 ICTA. The Revenue believed the same (if that be relevant).

Kleinwort Benson Ltd v. Lincoln City Council [1999] 2 AC 349 was cited as establishing that payment under a mistake of law was a permissible ground for a claim for recovery. That was not, the respondent noted, a case of a claim for recovery of overpaid tax.

The majority of the House of Lords held, in the context of a private law transaction (a claim for restitution of money paid to a local authority under an ultra vires interest rate swap agreement), that the principle of unjust enrichment applied so as to make money paid under a mistake of law or fact recoverable, subject to defences available in the law of restitution.

In this case the respondent contended that, even if the appellants succeeded on the substantive issues, the claim for unjust enrichment as a result of overpayment of tax under a mistake of law must be rejected. The recent decision of this court in Deutsche Morgan Grenfell Group Plc v. CIR [2005] STC 329 (DMG) was cited as binding authority for denying restitutionary claims for recovery of tax payments made under a mistake of law.

In DMG the legislation was unlawful on the face of it, the relevant ACT statutory regime having been held by the Court of Justice to be discriminatory contrary to Article 52 (now Article 43) of the EC Treaty. The UK subsidiary of a German parent company did not make an election. The ACT in question had been set off against MTC, but some of that had taken place more than six years before the commencement of the proceedings. DMG contended that the mistake of law was no bar to a claim for restitution since the decision in Kleinwort and that the mistake was not discoverable with reasonable diligence until the decision of the Court of Justice that the UK’s ACT regime was contrary to the discrimination provisions of the EC Treaty: section 32 Limitation Act 1980.

The Revenue’s position was that Kleinwort did not affect claims for mistaken payment of tax. Such claims could only be made either under relevant statutory provisions or in reliance on the Woolwich principle, which was concerned with a case of an unlawful demand for payment of tax by a public authority. The claim based on Woolwich was statute barred in those cases where more than six years had passed since the making of the payments.

The Court of Appeal held that, on a proper analysis of the decisions at the highest level, restitution of overpaid tax is only possible at common law under the Woolwich principle i.e. recoverability at common law is confined to the case of payment pursuant to an unlawful demand. The Woolwich claim was statute barred. Neither the Kleinwort claim for mistake of law nor the statutory claim, which depended on the payments having been made under an assessment, were available.

The leading judgment in DMG was given by Jonathan Parker LJ. Following a detailed survey of the authorities the following conclusion was drawn-

“208.

In my judgment, therefore, on a true analysis of Lord Goff’s speeches in Woolwich and Kleinwort Benson, a claimant who makes a payment to the Revenue under a mistake of law is not entitled to a restitutionary remedy in respect of that payment otherwise than under the Woolwich principle (where the demand is unlawful) or under the relevant statutory regime (where the demand is lawful).

Rix LJ agreed. He said-

“261.

… Where a tax payment has been exacted by an unlawful ultra vires demand, as in this case, the applicable cause of action is the Woolwich cause of action ... as of right as a matter of high constitutional principle, in which mistake of law, whether it occurs or not (and in Woolwich it did not occur), is irrelevant. Otherwise mistake of law in the context of payment of taxes is only relevant to the extent that it has a possible role to play within the applicable statutory provisions. The rules of recovery in private law transactions, however, in as much as they are based on mistake of fact or law, are governed by different principles and fall under a “separate and distinct” regime.”

The claim was for compensation for the period between the date of payment of ACT and the date on which they set it off against MCT.

Buxton LJ agreed. He said-

“291.

In Woolwich Lord Goff, with the concurrence of the majority of the House, crafted a carefully designed and limited remedy to address the specific problem of ultra vires demands. He recognised that it might be necessary to limit the ambit of recovery of right against public funds; for instance by the introduction of specific time limits. He emphasised that the remedy turned on the nature of the demand, and not on any mistake of the payer. Quite apart from the language used in Kleinwort, it is wholly implausible that Lord Goff in that case envisaged that the same problem could be addressed by a remedy with much more extensive time limits and formulated entirely in terms of mistake by the payer. It is not open to the claimants in this case to adopt the remedy in terms of mistake of law that was introduced for private law cases in Kleinwort.”

The ratio of DMG, for which leave has been granted to appeal to the House of Lords, is, of course, binding on this court. In my judgment it has been clearly held at this level of decision that tax can only be recovered under statutory provisions or where there has been an unlawful demand.

It is right to note, however, that the judgments of this court in DMG have been criticised on a number of grounds: that the passages in the speech of Lord Goff relied on in Kleinwort as holding that mistake of law was unavailable in the case of tax payments were read out of the context of his speech as a whole; that they were obiter; that insufficient attention was paid to the other majority speeches in Kleinwort; and that it was contrary to principle and policy to exclude mistakenly paid tax from the ruling in Kleinwort that mistake of law was no bar to restitution. It has been argued that the true effect of Woolwich was to extend, and not to limit, the causes of action available by recognising a cause of action for unjust enrichment based on an unlawful demand made ultra vires. This was added to the existing causes of action for payments made under a mistake or under duress. As held in Kleinwort mistake of law was not a bar to restitution. The court in DMG was accordingly wrong in confining the cause of action available in the case of tax payments to cases of unlawful demands. A mistake of law does found a cause of action for restitution of overpaid tax, as well as for payments in private law.

An even more fundamental criticism can be advanced against the whole enterprise of searching for a mistake of law on the part of the taxpayer on which to found the restitutionary claim. It can be argued that the true foundation of the restitutionary claim in such cases lies in the absence of a basis for the payment rather than in the making of a mistaken payment.

In my judgment, unless and until the House of Lords overrules DMG, the position is that the claims in respect of overpaid or prematurely paid ACT can only succeed on the basis of Woolwich by establishing an unlawful demand for payment of ACT. For the reasons advanced by the respondent there was no unlawful demand by a public authority in relation to the payment of ACT.

On the above analysis of the appellants’ cause of action this point only arises if the appellants establish that they made payments of ACT in respect of dividends paid to non-EU parents pursuant to unlawful demands and that they are entitled to restitution from the respondent in respect of the ACT wrongly paid pursuant to unlawful demands.

The respondent submitted that the appellants’ claim for restitution on the basis of Woolwich would be restricted by the application of the Pintada principle. Even if the appellants have a Woolwich claim based on unlawful demands for payment of ACT, it is only a claim for the time value of money. The respondent contended that this should be equated to a claim for interest and that the appellants are not entitled to any interest, simple or compound, on the sums claimed at common law, in equity or under statute (i.e. section 35A of the Supreme Court Act 1981), as there was no principal sum outstanding to the appellants: President of India v. Pintada [1985] 1 AC 104 (Pintada). There was no principal sum outstanding, as the appellants set off ACT payments against MCT.

The appellants’ response was that the application of the Pintada principle to claims of unjust enrichment based on Woolwich would produce a very surprising result, when compared with the case of a breach of an article of Community law, such as Article 56 of the EC Treaty, where a claim for full restitution, including loss of the time value of money, would be available to the appellants: see the comments of Lord Goff in Woolwich at p.177E regarding the relationship between remedies under domestic law and under Community law and his surprise at the prospect of the remedy under Community law being more generous than the remedy under UK law; Hoechst at paragraph 88, which characterised the principal sum in an action for restitution as “the amount of interest which would have been generated by the sum, the use of which was lost as a result of the premature levy of the tax”; and the judgment of Chadwick LJ in the Community context in Sempra Metals v. IRC [2005] EWCA (Civ) 389 at paragraphs 23 and 46.

The appellants emphasised that they are not claiming interest on a contractual debt or on damages for tort or breach of contract. Their cause of action is neither tortious nor contractual. They are not seeking compensation for the loss of the use of the money. The Pintada principle does not, they argued, apply to an unjust enrichment claim. The cause of action is restitutionary. It is based on the unjust enrichment of the respondent and the absence of any justification for retaining the benefits unjustly gained. They seek an order preventing the unjust enrichment of the respondent and making that cause of action effective by requiring the respondent to disgorge the time benefit value of money, which it has enjoyed as a result of an unlawful demand. The claim represents the restitutionary value of money of which the appellants were wrongfully deprived. It is for the time value of money paid in respect of tax at a time when it should not have been paid. It is based on the fact that the respondent has been unjustly enriched by having their money paid as ACT for the relevant period of time, during which there was no justification for retaining it. It was equivalent to a “massive interest free loan” for the period between the date of payment of ACT and the date of set off, for which the respondent should make restitution for the time value of the money to which the appellants were entitled as of right.

The appellants submitted that the judge erred in holding that, if there was a restitutionary common law claim, it was unnecessary for him to decide whether the claim for payment of interest would be limited by the Pintada principle. Pintada was irrelevant, being a case on payments of debts at common law.

It is important to bear in mind that the difficulty with an interest claim by the appellants arises in those cases where the appellants had set off their ACT payments against MCT before they issued proceedings. At that time they had no claim for the return of the ACT they had paid. The appellants’ claim is for restitution of the “time value” of money in respect of tax paid at a time when it should not have been paid. Such a claim is, they argued, within the underlying rationale of Woolwich of providing a remedy for unjust enrichment.

The respondent pointed out that the appellants paid ACT because, like the respondent itself, they believed that the legislation compelled them to pay it. The respondent had not wrongfully made a demand on the appellants to pay the ACT.

Further, no such restitutionary claim for a benefit wrongfully obtained was recognised or even mentioned by the House of Lords in Woolwich or in Westdeutsche Landesbank v. Islington LBC [1996] AC 669.

The appellants’ general comment on those authorities was that this branch of the law is constantly developing and evolving to meet new challenges. In particular Mr Cavender observed that in Westdeutsche the focus was on a different issue, namely whether, on the balancing to which they were entitled in the context of a void commercial transaction, the claimants were entitled to simple interest or to compound interest. As to Woolwich, section 35A of the 1981 Act was not mentioned and there was no issue as to whether the payment of the time value of money was under restitutionary principles or under statute. The issue identified by Lord Goff was whether the building society had a right to re-payment of the principal sum or whether it was being gratuitously returned to the building society. Lord Browne–Wilkinson (at p197D) regarded repayment of interest earned on the money retained as a paradigm of a case of unjust enrichment.

In my judgment, in the absence of clear binding authority on this point, it falls to be decided as a matter of principle. On that approach the appellants seeking a restitutionary remedy should succeed in obtaining the relief sought, if they establish that the ACT was paid to the respondent pursuant to an unlawful demand.

The matter is not settled, as the respondent contends, by La Pintada (see above) as the appellants do not claim an entitlement to interest, as creditors, on a debt or on damages by way of compensation for loss of the use of the money demanded and retained by the respondent.

Nor is the matter settled by the fact that there were no pronouncements on the point by the House of Lords in Woolwich or in Westdeutsche. As the focus in those cases was on different issues the point was not directly addressed.

Although the Woolwich claim is not based on EC law, as in Hoechst and Sempra, it would, in my view, be surprising, as Lord Goff noted in Woolwich, if the extent of the restitutionary relief available turned on whether the cause of action for restitution was based on Community law or common law and that the relief under Community law was more generous than the relief available at common law. To allow full restitution would also be more consistent with the underlying principle that an unjust enrichment claim is not a claim for compensation for loss, but for recovery of a benefit unjustly gained and retained by the person enriched at the expense of the claimant.

The appeal is dismissed for the reasons given by Lloyd LJ. The respondent’s cross appeal relating to the amendment of the pleadings is allowed for the reasons given above. The question whether there should be a reference to the European Court of Justice on points arising from Articles 56 and 57 should be deferred until it is known whether there is to be an appeal to the House of Lords.