Nec Semi-Conductors Ltd. & Ors v Inland Revenue

These are as follows.

Between 1973 and 1999 the imputation system regulated the taxation treatment of dividends paid by companies resident in the United Kingdom. When such a company paid a dividend the general rule was that it had to pay ACT to the Revenue. There was an exception if the company was a 51 per cent-plus subsidiary of a holding company which was also resident in the United Kingdom. In that case the two companies could make a group income election, and as long as the election was in force the subsidiary did not have to pay ACT. If, however, the United Kingdom company was a subsidiary of a non-resident company the two companies could not make a group income election. The compatibility of this rule with the EC Treaty was challenged in the mid to late 1990s, and in Metallgesellschaft/Hoechst the CJEC held that, as between a United Kingdom subsidiary and a parent company in another member state, the rule was incompatible with the present article 43 of the Treaty (then numbered article 52). Consequential questions were referred back to the English courts to decide, and this led to the making of a group litigation order, a GLO. This is the third case which I have heard pursuant to the GLO. The previous two were Pirelli and DMG. (The Hoechst case was also consequential on an aspect of the Metallgesellschaft/Hoechst decision, but it was not within the ambit of the GLO.) All the cases have involved questions concerned with claims for compensation or restitution by United Kingdom subsidiaries of non-United Kingdom parent companies in respect of ACT which they paid. All the cases until this one have involved dividends flowing from United Kingdom subsidiaries to parent companies established in other member states of the European Community. In this case, however, the parent companies are resident outside the European Community. Restitution for ACT paid is claimed on two alternative grounds.

The first ground relies on the proposition that the relevant United Kingdom law, or more precisely the manner in which it was operated by the Inland Revenue, was in breach of certain provisions of non-discrimination articles in double taxation agreements concluded by the United Kingdom with some overseas nations. For the reasons which I will describe at length below I do not accept the claimants’ case. Although I think that there may have been infringements of the non-discrimination article (infringements which were certainly not perceived to have been such by anyone at the time), in my judgment this aspect of the article never became a part of domestic United Kingdom tax law upon which the taxpayer claimants could rely in proceedings in the national courts.

The second ground on which the claimants say that they are or may be entitled to relief involves reliance on provisions in article 56 of the EC Treaty. The claimants do not suggest that I can myself decide the relevant arguments, but they invite me to refer the relevant questions to the CJEC under article 234 of the EC Treaty. However, in my opinion article 5 7(1) is clear to the effect that the provisions of article 56 on which the claimants wish to rely do not apply to the ACT payments complained of. In the circumstances I will not make a reference to the CJEC. Instead I dismiss the claimants’ second ground for claiming relief, as well as their first.

Repeating what I have said in paragraph 2 above, under the imputation system the general rule under the Taxes Act was that, when a company resident in the United Kingdom paid a dividend to its shareholders, it was liable to pay ACT to the Inland Revenue: s 14. At the times relevant to this case the rate of ACT was 25 per cent of the dividend. The ACT was available to be set off later against the company’s liability for corporation tax (MCT), that is corporation tax on its profits (its taxable income and its chargeable gains) in so far as it had such profits. The effect of the ACT was that, if the company did have taxable profits, it had to pay its corporation tax earlier than it otherwise would have done, and earlier than some of its competitors paid their corporation tax. Further, it was quite common for a company to have paid ACT but not to have a sufficient liability to MCT for it to be able to set off the ACT. In that case the ACT was not just an early payment of corporation tax: it was a true additional tax liability. The general point was that it was disadvantageous to have to pay ACT. However, s 247 enabled a United Kingdom resident company which was a subsidiary of a parent company also resident in the United Kingdom to join with its parent in making a group income election. Once the election was accepted by the Inland Revenue or upheld on appeal the United Kingdom subsidiary could pay dividends to its parent company without being liable to ACT. Thus United Kingdom tax law extended an advantage to subsidiaries of United Kingdom holding companies which it did not extend to subsidiaries of overseas holding companies.

All of this was the background to the Metallgesellschaft/Hoechst case, in which the CJEC held that the rule of United Kingdom law which prohibited a United Kingdom subsidiary and a parent company established in another member state from making a group income election was contrary to article 43 (formerly article 52) of the EC Treaty, the right of freedom of establishment. The CJEC directed that in such cases compensation or restitution was payable by the Inland Revenue, and left consequential questions of United Kingdom law to be determined by the English courts. As I said above, I have already decided three cases of that nature: the Pirelli case, the Hoechst case and the DMG case. Pirelli and DMG are subject to appeal, but the position rests for the time being at my decisions, the references to which are in the table in paragraph 1 of this judgment.

The present case is another one in which restitution is sought by United Kingdom subsidiaries of non-United Kingdom parent companies which paid dividends in the belief, which was certainly encouraged by the clear words of s 247, that they could not make group income elections. The United Kingdom subsidiaries therefore paid ACT, but now they claim that the ACT was unlawful (or more precisely that the rule in s 247 which prevented them and their parent companies making group income elections was unlawful). They say that in consequence they are entitled to restitution. If in the meantime the ACT has been set off against MCT the restitution which they claim is an amount calculated by applying an interest factor over the period beginning with when they paid the ACT and ending with when they set it off against MCT: that period is equivalent to the period for which the ACT rules required them to pay part of their corporation tax liability earlier than they would otherwise have done. If and to the extent that the ACT has not been set off against MCT the subsidiaries claim restitution of the ACT, plus interest.

However, this case, unlike Pirelli, Hoechst and DMG, is not brought by United Kingdom subsidiaries of parent companies established in other member states. Rather it is brought by subsidiaries of parent companies established outside the European Community, and, so far as concerns the part of the case with which I will deal first (the non-discrimination article issue), the subsidiaries rely not on provisions of European law but on so-called non-discrimination articles in double taxation agreements between the United Kingdom and the State where the parent company is based.

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.

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.

I will now set out the text of the critical sub-paragraphs of the double taxation agreements, and then I will set out the legislative provisions which bring the contents of such agreements, or some of their contents, into force as parts of United Kingdom domestic law. Article 25(3) of the United Kingdom/Japan double taxation agreement is as follows. (I have interpolated in square brackets references to the United Kingdom and to Japan as appropriate.)

The statutory provision which governs the introduction of double taxation agreements into domestic law is now s 788 of the Taxes Act 1988. It traces its origins back to the Finance Act 1945, which was enacted in anticipation of double taxation agreements expected to be concluded between the United Kingdom and other States. I believe that an agreement between the United Kingdom and the United States was the first such agreement between this country and another sovereign State. The United Kingdom also concluded agreements with many British dependent territories. The critical subsection of the present s 788 is subsection (3), but I need also to set out subsection (1).

To complete the citations of the legislative background to double taxation agreements in United Kingdom tax law I should also refer to the Orders in Council which are made as contemplated by s 788(l), In the case of Japan the Order is the Double Taxation Relief (Taxes on Income) (Japan) Order 1970. It was made by Her Majesty the Queen. After recitals the operative paragraph is paragraph 2. It is worded as one would expect given the terms of s 788. It provides:

The facts of the test cases are all agreed, and I can deal with them very briefly. In the case of the NEC group the United Kingdom subsidiary was a wholly owned subsidiary of the Japanese parent company. The agreed statement of facts gives particulars of four dividends paid between 1994 and 1996 and of the associated ACT payments, which were made on dates in 1995 and 1996. I do not need to set out the specific dates or amounts, since nothing turns on those details. Most of the ACT paid was set off against MCT, but a substantial part of it was not. It is agreed that, if s 247 had permitted a group income election to be made, the companies would have made one.

In the case of the Bush Boake Allen group the details are obviously different, but the underlying principles are the same. There were several United Kingdom subsidiaries, not just one, and the parent company was (and is) in the United States, not in Japan. The agreed statement of facts gives details of ACT payments made between 1996 and 1999. As with the NEC group most of the ACT was set off against MCT, but not all of it. There is the same agreement that, if s 247 had permitted a group income election to be made, the companies would have made one. One extra feature of the Bush Boake Allen group is that under the United Kingdom/USA double taxation agreement of 1980 (which replaced an earlier one of 1945 or 1946) the United States parent company received from the United Kingdom Revenue tax credit payments similar to those which were at the core of the arguments in the Pirelli case. There was nothing similar in the case of the NEC group. The explanation is that the tax credit payments were an aspect of the imputation system which commenced in this country in 1973, before the 1980 double taxation agreement with the United States but after the double taxation agreement with Japan.

The argument for the claimant companies seems to me to have three main strands to it. The first is an argument about the interpretation of the critical sub-paragraph in the non-discrimination article; the second is an argument about the interpretation of s 788 of the United Kingdom statute; the third is an argument about the remedy to which the claimants are entitled if they are correct on their first two arguments.

For convenience I will repeat here the relevant parts of the critical sub-paragraph in the article, but substituting references to ‘the United Kingdom’ and to ‘Japan’ for the references to a ‘Contracting State’.

I move to the second strand in the claimants’ argument. They say that the particular effect of the non-discrimination article on which they rely is made part of United Kingdom domestic law by s 788 and the Order in Council. In this connection they base themselves principally on s.788(3)(a), by which provisions in a double taxation agreement (provided that it has been brought into effect domestically by an Order in Council) have effect ‘in so far as they provide for relief from ... corporation tax in respect of income or chargeable gains’. Their case is that the provisions of the non-discrimination article, in so far as they meant that the United Kingdom subsidiaries concerned could make group income elections or otherwise pay dividends without being liable to ACT, provided for relief from corporation tax in respect of income or chargeable gains. I will examine the argument in detail later, but it may be helpful for me to mention now that this is the central point at which, in my view, the case for the claimants breaks down.

The third strand in the claimants’ argument is that, if they are right so far, they are entitled at common law to restitution from the Revenue, on the authority of either or both of the Woolwich or Kleinwort Benson cases (for references to those cases see the table in paragraph 1 above). As I have said earlier they argue that, in so far as the ACT has been set off against MCT, the restitution is an amount calculated by reference to interest over the period between the payment of the ACT and the setting-off of it against MCT; in so far as the ACT has not been set off against MCT, they claim restitution of the full amount of ACT not so set off, with interest. They also argue that it is proper for them to pursue their claim by an action in the High Court, rather than following the more usual route for resolving tax disputes of taking an appeal to the Special Commissioners: in the special circumstances of this case they say that the route of statutory tax appeals would be of no use to them even if their arguments on the points of principle are correct.

Under the various sub-headings which follow I will evaluate the foregoing arguments advanced by the claimants.

Did the paragraph of the article have the effect of prohibiting the United Kingdom from denying to United Kingdom subsidiaries of parent companies in, for example, Japan and the United States the right to make group income elections? For convenience I set out again an abbreviated paraphrase of the article:

I begin with the background observation that, if the paragraph did have the effect contended for by the claimants, it was certainly not an effect which was intended. In a sense that is an obvious and trite point, because the wording of the paragraph is taken from the OECD model Convention, which . was certainly not formulated with the United Kingdom’s imputation system in mind. To the same effect the United Kingdom/Japan double taxation agreement was agreed in 1969, four years before the imputation system commenced.

However, the United Kingdom/US double taxation agreement was negotiated and agreed after the commencement of the system, and it is, I think, generally accepted and understood that a major reason why the United Kingdom and the United States entered into negotiations for a new agreement in place of the one which had operated from 1946 onwards was because the imputation system made a renegotiation desirable. Thus the international tax aspects of this country’s imputation system were in the forefront of the negotiators’ minds. It certainly did not occur to the United Kingdom’s negotiators that the standard wording of the non-discrimination article in the OECD model would mean that United Kingdom subsidiaries could pay dividends to United States parents under group income elections and without ACT. If that thought had occurred to the United Kingdom negotiators I am sure that they would have insisted on some provision to prevent the effect arising.

Further, as regards the United States negotiators I can feel almost as confident that the argument which is now placed before me did not occur to them either. In 1977 the US Treasury Department published a Technical Explanation of the new double taxation agreement which had been signed in December 1975. (The agreement did not commence to operate until 1980, but, if my memory serves me correctly, that was because of difficulties concerning taxation in some States of the USA, and had nothing to do with the subject matter of this case.) I believe that the Technical Explanation is accorded significant weight in the courts of the United States, and I imagine (though I do not know) that the United States negotiators contributed to it. The explanation of article 24(5) itself casts no light on the argument now advanced by the claimants, but some of the contents of the explanation of article 23 are informative. That article was in many respects the most important new article in the agreement, because it regulated the extent to which United States recipients of dividends from United Kingdom companies could receive and benefit from tax credits available against the United Kingdom Revenue. (In paragraph 12 above I set out provisions of s 788 of the United Kingdom Act, including s 788(3)(d). That paragraph had been introduced by the Finance Act 1972 specifically to give domestic effect in United Kingdom law to anticipated articles in future double taxation agreements along the lines of article 23 of the 1975 double taxation agreement with the United States.) The detailed explanations of article 23 do not matter, but what is significant is that they contain a number of worked examples in which a United. Kingdom subsidiary is assumed to have paid a dividend to a US parent company. In all of the examples it is assumed that ACT would be paid. If the United States negotiators had thought that the non-discrimination article would permit group income elections to be made, the examples would not have been expressed as they were.

Thus, if the non-discrimination article does have the effect contended for by the claimants in the present case, that is a fortuitous and unintended windfall. However, I do not think that that in itself means that the interpretation of the article contended for is wrong. Indeed, I have concluded that the claimants are right on this part of the case. A non-discrimination article by its nature is unlikely to be directed at one or more specific provisions of a Contracting State’s tax legislation as respects which the discrimination was already present to the minds of the negotiators. Rather it is in the nature of a general precautionary provision, meaning that if, over time in the course of operation of the double taxation agreement, it emerges that some provision of the tax law of one of the contracting states (whether tax law already existing at the time of the agreement or tax law introduced thereafter) is discriminatory within the terms of the article, the discrimination ought not to be permitted. In this case, notwithstanding that neither the United Kingdom nor the United States negotiators appear to have thought of the possibility that the inability of United Kingdom subsidiaries to pay dividends to United States parents without having to pay ACT might infringe the article, now that the possibility had been raised I think that I must make up my own mind upon it.

It seems to me that there are three questions raised by the wording of the relevant sub-paragraph of the article, which I set out in an order different from that in which they appear in the sub-paragraph itself. First, what is meant by ‘other similar enterprises of [the United Kingdom]’? (In that question ‘the United Kingdom’ replaces the actual words ‘that first-mentioned State’.) Secondly, given that the position under United Kingdom statute law was that the United Kingdom subsidiary of a Japanese or US parent company could not make a group income election and, if it paid dividends to its parent, had to pay ACT, was the position thereby created ‘taxation or any requirement connected therewith’? Thirdly, if it was, was it ‘other or more burdensome’ than the taxation and connected requirements to which the ‘other similar enterprises’ of the United Kingdom were or might be subject.

So the first question is: what is meant by ‘other similar enterprises’ of the United Kingdom? Paraphrasing the question: what is the comparator with which the taxation of the actual United Kingdom subsidiary is to be compared? In my judgment, in agreement with Mr Aaronson QC for the claimants, the comparator is a United Kingdom subsidiary of a United Kingdom parent, as opposed to the actual case of the United Kingdom subsidiary of, to take the NEC case, a Japanese parent. I accept that the sub-article could have been more explicit on the point, but in my view the interpretation which I favour is clearly correct. In theory there could be four possibilities to examine:

I do not accept Mr Glick’s argument. I agree that the differences between the actual parent company and the hypothetical parent company which Mr Glick describes would have existed, but I do not think that they are relevant to the exercise with which I am concerned. I am concerned with whether the actual Japanese-owned company (eg NEC Semi-Conductors Ltd) and another hypothetical United Kingdom company are ‘similar’. I agree that it is necessary to assume that, if the actual company is a subsidiary of another company, then so is the hypothetical comparator company; but I do not agree that the exercise has to be taken to the lengths of examining whether the hypothetical parent company, though resident in the United Kingdom, would have a tax treatment as respects dividends received and paid which would be the same as if it was not resident in the United Kingdom. In my judgment the actual company (e.g. NEC Semi-conductors Ltd) and a hypothetical other United Kingdom resident company which is a subsidiary are fairly comparable with each other in this context, and I do not accept that they cease to be comparable because there would be differences between the United Kingdom tax treatments of their respective actual and hypothetical parent companies. I also believe that Mr Glick’s submission would be difficult to reconcile with the OBCD commentary on the sub-article, which states that the provision ‘relates to the taxation only of the enterprises and not of the persons owning or controlling their capital’. Therefore I do not accept the submission that there did not exist any such comparator company as was visualised by article 25(3) of the United Kingdom/Japan double taxation agreement (and of the equivalent sub-article in other double taxation agreements).

I return to what I believe to be the correct interpretation. The context is a United Kingdom subsidiary controlled by a parent company resident in the other Contracting State (eg Japan) which pays dividends to its parent. In that context I believe that ‘other similar enterprises of [the United Kingdom]’ means other United Kingdom subsidiaries controlled by parent companies resident, not in the other Contracting State, but in the same Contracting State (ie in the United Kingdom) which pay dividends to their parents. In my view it is obvious from the whole scheme and purpose of the non-discrimination article that that is what the particular sub-article is getting at. It is clearly the view taken by the United States Treasury. The technical explanation (to which I referred in paragraph 24 above) reads as follows (with my italicising of the vital words).

in all the circumstances I agree with Mr Aaronson that in the present case the comparator ‘similar enterprise’ postulated by the relevant paragraph of the non-discrimination article does exist, and that it is a United Kingdom subsidiary company owned by a United Kingdom parent.

In my view that deals with the only difficult point on the question of construction of the paragraph, but I need to cover the other two questions which I identified in paragraph 26 above. One of those questions focused on the treatment under the United Kingdom statutory provisions of a United Kingdom subsidiary which paid dividends to its non-resident parent company. The subsidiary had to pay ACT in consequence of paying the dividends, and it could not make a group income election with its parent to remove the ACT liability. Was that a treatment covered by the description ‘taxation or any requirement connected therewith’? In my judgment it was. It could be regarded as ‘taxation’ in that the result was that ACT (which is certainly taxation of some sort) had to be paid. It could be regarded as a requirement connected with taxation in the sense that, in order to make a group income election, the United Kingdom subsidiary needed to have a United Kingdom parent company — a requirement with which it could not comply. I only add that I do not think that Mr Glick effectively contested the proposition that the facts of the present case did exhibit taxation or a requirement connected with it.

The other question which I identified in paragraph 26 above was whether the taxation or connected requirement to which the actual United Kingdom subsidiaries (eg NEC Semi-conductors Ltd) were subjected was ‘other or more burdensome’ than the taxation or connected requirements to which the hypothetical United Kingdom comparator company would have been subjected. Given my conclusion in earlier paragraphs that the hypothetical comparator is a United Kingdom subsidiary of a United Kingdom parent company, the answer is clear: yes, the taxation and connected requirements were other than the taxation and connected requirements which would have applied to the comparator; further, they were more burdensome.

For the foregoing reasons I accept the first part of Mr Aaronson’s submissions on behalf of the claimants. In my judgment the relevant rules in the Taxes Act were, so far as their impact on the payment of dividends by the claimant United Kingdom subsidiaries was concerned, contrary to article 25(3) of the United Kingdom/Japan double taxation agreement, and were also contrary to the identically worded article 24(5) of the United Kingdom/United States double taxation agreement. I will only add this. The respect in which the rules were contrary to the double taxation agreements was only that they did not permit the subsidiaries to join with their non-resident parent companies in making group income elections. I do not accept one submission which Mr Aaronson made, namely that by virtue of the double taxation agreements the subsidiaries should simply have been allowed to pay dividends to their parents without ACT and without first making ACT elections and observing all the procedural requirements in connection therewith which had to be observed by United Kingdom subsidiaries which made elections with their United Kingdom parent companies.

This is in my view the critical question in this case. In my view the answer to it is: no. I will enlarge on my reasons in the paragraphs which follow, but it might be helpful if I state here the two key stages in my reasoning. They are as follows.

In this and a number of following paragraphs I enlarge on the proposition summarised in (i) above. A taxpayer who claims that some aspect of his United Kingdom tax treatment is contrary to a provision of a double taxation agreement must first succeed in showing that the treatment complained of is indeed contrary to the agreement (as I have accepted that the claimants have shown in this case): but that only gets the taxpayer part of the way home. If a resident company wishes to be able to rely on an infringement of the agreement in direct proceedings against the Inland Revenue, it also needs to establish that the particular effect of the agreement has been incorporated into United Kingdom domestic law. If it has not been so incorporated, then the actions of the Inland Revenue in operating the tax system in a way which conflicts with the agreement may be a breach of international law between the United Kingdom and the other State. But that does not give to the aggrieved taxpayer rights for which it can claim protection in the national courts of this country.

The vital point in this connection is that s 788(3) provides that, once the ‘arrangements’ in a double taxation agreement have become the subject matter of an Order in Council declaring it expedient that they shall take effect, then they shall ‘notwithstanding anything in any enactment, have effect in relation to income tax and corporation tax in so far as they provide’ for any of the various matters spelt out in paragraphs (a) to (d) of the subsection. If the infringement of the terms of the double taxation agreement is within one of the matters spelt out in those paragraphs, the taxpayer can rely on it in domestic proceedings. Usually the domestic proceedings will take the form of a conventional tax appeal which begins before the Special or General Commissioners, but exceptionally they may take the form of a civil action in the national courts. If, however, the infringement of the double taxation agreement is not within any of the matters listed in paragraphs (a) to (d) of s 788(3), then that aspect of the agreement is not part of the domestic law of this country. The taxpayer cannot rely on it in a conventional tax appeal or in a civil action against the Inland Revenue.

Although Mr Aaronson was cautious about accepting it, there is in my opinion no doubt that the words ‘in so far as they provide for’ are limiting words. Mr Aaronson’s considered position on this was that the words did have a limiting effect, but only a small one. He said that it would have been different if they had read ‘only in so far as they provide for’, etc. I do not agree. In my opinion the addition of the word ‘only’ would not have made any difference.

Mr Aaronson urged that I should give a wide interpretation to the various matters listed in paragraphs (a) to (d), in order that as many provisions of double taxation agreements as possible should be covered by s 788(3) and thereby made parts of domestic law. I do not see why I should do that. I intend to give a natural interpretation, neither wide nor narrow, to those of the matters listed in paragraphs (a) to (d) which are argued to cover the particular application of the non-discrimination article in point in this case. The matter principally relied on by the claimants is the reference in paragraph (a) to providing for ‘relief from ... corporation tax in respect of income or chargeable gains’. If, as is the case, I do not think that the natural interpretation of those words includes relief from ACT, I do not think that I should give an unnatural and extended interpretation to the words in order, through the non-discrimination article, to bring treaty relief from ACT within the scope of domestic law. I do not accept the submission that the structure of s 788(3) (or its statutory predecessors going back to the Finance Act 1945) evinces an intention on the part of the draftsman to list everything which might realistically be expected to be covered by a double taxation agreement. If the draftsman’s intention had been to bring into domestic law everything contained in a double taxation agreement he could have provided that ‘the arrangements shall, notwithstanding anything in any enactment, have effect’. And he could have left it at that. To my mind the structure of the subsection shows clearly that the draftsman did not want to secure that everything in a double taxation agreement should become part of domestic law. That is why he continued with the words ‘in relation to income tax and corporation tax in so far as they provide for’ the listed matters.

It is also worth pointing out that, whereas the generality of provisions in double taxation agreements are stated in the agreements to apply, so far as United Kingdom tax law is concerned, only to income tax, corporation tax and capital gains tax, the provisions of non-discrimination articles are different and are commonly expressed to apply to all kinds of taxes. In the United Kingdom/Japan double taxation agreement paragraph (5) of article 25 (the non-discrimination article) reads: ‘In this article the term ‘taxation’ means taxes of every kind and description.’ In the case of the double taxation agreement with the United States a similar result is produced by article 2(4): ‘For the purposes of Article 24 (non-discrimination) this Convention shall also apply to taxes of every kind and description imposed by each Contracting State, or by its political subdivisions or local authorities.’ Section 788(3) provides that double taxation agreement provisions shall have effect in domestic law in relation to income tax and corporation tax (and then only in so far as they provide for the matters listed in paragraphs (a) to (d)). So there can be provisions in the non-discrimination article of an agreement which prohibit discrimination in relation to a wide range of taxes going beyond income tax, corporation tax and capital gains tax; but it is certain that any such provisions are not brought into domestic effect by s 788(3).

The proposition that not everything in a double taxation agreement is brought into domestic law by s 788(3) is confirmed by the decision of the Divisional court (Nolan LJ and Henry J) in R v IRC ex parte Commerzbank AG [1991] STC 271. The taxpayer, a German bank with a branch in the United Kingdom, had already won a tax case against the Inland Revenue, with the effect that it received a substantial repayment of corporation tax. It claimed also to be paid a repayment supplement (a payment in the nature of interest from the Inland Revenue), but was faced with the problem that the statutory provisions required the taxpayer to be a United Kingdom resident before a repayment supplement could be paid. In the case before the Divisional Court one of its arguments was that the denial to it of the supplement was a breach of the non-discrimination article in the United Kingdom/Germany double taxation agreement. The court agreed that that appeared to be so, but nevertheless held that the article did not give to the bank rights on which it could rely in the United Kingdom courts in order to secure a repayment supplement. One of the reasons was that the payment of interest or repayment supplement was not within any of the matters set out in paragraphs (a) to (d) of s 788(3). Therefore, even accepting that the denial of the supplement was a breach of the non-discrimination article, that particular application of the article had not become a part of United Kingdom domestic law. That is the only point in the protracted Commerzbank litigation which was relevant to the present case, but for completeness I mention that the Divisional Court referred to the CJEC the question of whether the denial of the repayment supplement to the bank infringed the bank’s rights under Community law. In the CJEC it was held that it did (see Case C-330/91, reported among other places at [1993] STC 605), so the bank got its repayment supplement in the end. However, it did not get it through reliance on the non-discrimination article in the United Kingdom/Germany double taxation agreement and on s 788(3), and that is the important point for present purposes.

I will return later and consider several other matters to which Mr Aaronson has referred in seeking to persuade me to give to s 788(3) a wider meaning than I believe that it can bear, but for the moment I wish to move on to the next vital stage in the analysis. This is proposition (ii) in paragraph 34 above. I have accepted that the non-discrimination article, as a matter of interpretation of the double taxation agreement, meant that United Kingdom subsidiaries of parent companies in, for example, Japan and the United States ought to have been permitted to make group income elections with their parent companies, and then pursuant to the elections to pay dividends without being liable to pay ACT. Was that an effect which was covered by any of the matters set out in paragraphs (a) to (d) of s 788(3)? If the answer is no, the claims in this case cannot succeed. In my judgment the answer is indeed: no.

In my opinion the question turns on paragraph (a), but I will first mention paragraph (b), since a submission was addressed to me in reliance on that paragraph. It gives effect to treaty provisions for charging the income from sources ... in the United Kingdom to persons not resident in the United Kingdom’. Mr Aaronson suggested that those words can embrace the particular application of the non-discrimination article which the claimants seek to rely on in this case. In my opinion they cannot. Paragraph (b) is clearly directed at the United Kingdom tax treatment of a non-resident which receives income from a United Kingdom source. The complaint which the claimants advance here is concerned with the United Kingdom tax treatment of a resident which pays income (a dividend) to a non-resident recipient. The paragraph is of no assistance to the claimants.

So I concentrate on paragraph (a). That paragraph gives effect in domestic law to provisions in double taxation agreements in so far as they provide for relief from income tax, or from corporation tax in respect of income or chargeable gains’. I have held that the paragraph of the non-discrimination article relied on by the claimants meant that a United Kingdom subsidiary of a parent company resident in the other Contracting State ought to have been permitted to make a group income election, so as not to be liable to pay ACT in consequence of paying dividends. I accept that in that respect the paragraph provided for relief from corporation tax. However, I do not accept that it provided for relief from corporation tax in respect of income or from corporation tax in respect of chargeable gains. In my opinion ACT, advance corporation tax, is a species of corporation tax, but it is a different species of corporation tax. It is not corporation tax in respect of income or chargeable gains.

Corporation tax was introduced in 1965. ACT was not introduced until some years later. The legislation was contained in the Finance Act 1972 and took effect from 6 April 1973. Reverting to the introduction of corporation tax, the Finance Act 1965 provided that corporation tax was to be charged on the ‘profits’ of companies, and ‘profits’ meant income and chargeable gains. Those provisions are still in force, but as a result of intervening consolidations they are now in s 6(1) and 6(4)(a) of the Taxes Act 1988. Corporation tax in respect of income plainly was, and in my view still is, the part of the company’s corporation tax liability on its profits referable to the income content of the profits. Corporation tax in respect of chargeable gains similarly was, and in my view still is, the part of the company’s corporation tax liability on its profits referable to the chargeable gains content of the profits. That is what, before 6 April 1973 when ACT began to be charged by reference to dividends, was meant by the words now in s 788(3)(a): ‘corporation tax in respect of income or chargeable gains’. It was the corporation tax which a company was liable to pay on the income and the chargeable gains which were comprised in its taxable profits. And in my opinion the words did not acquire an enlarged meaning from 6 April 1973 onwards so as to include ACT.

I will now seek to demonstrate that ACT, though designated as corporation tax by the legislation, cannot sensibly be regarded as corporation tax in respect of income or chargeable gains. The charging provision became (on consolidation) s 14(1) of the Taxes Act 1988:

In this connection it is worth reflecting on how the draftsman has expressed s 14(4), especially the second part of it:

In s 788(3) itself there is another statutory hint to the effect that not everything which is called ‘corporation tax’ has to be corporation tax in respect of income or chargeable gains, although this may have been unintentional and I would not place much weight on it. If I set out words taken from the opening part of the subsection and of paragraph (a), but omitting references to income tax, I find the following:

For the foregoing reasons I do not accept the test claimants’ case that they can seek restitution at common law for the denial to them of the right to make group income elections and to pay dividends to their Japanese or US parents without also paying ACT to the Inland Revenue. They could only have a claim for such restitution if they had a right in domestic law to make the group income elections. In my judgment they did not have that right because the particular provision of the non-discrimination article on which they rely did not become part of the domestic tax law of this country. Essentially what I have said so far gives the whole of my reasoning, but there are a number of associated matters on which I wish to make some observations, doing so in many respects in deference to arguments which were presented to me but which I have not found to be persuasive.

Mr Aaronson drew my attention to a Memorandum from the Inland Revenue submitted to a Select Committee of the House of Commons in 1994. I will not quote from the memorandum here. It certainly expresses the view that non-discrimination articles are important parts of double taxation agreements, but I do not detect in it any express statement one way or the other about whether their provisions are or are not incorporated into domestic law. There are two points which it is worth recalling in this connection.

Mr Aaronson sought to rely on articles 26 and 27 of the Vienna Convention on the Law of Treaties as supporting his submission that I should strive to conclude that the non-discrimination article took domestic effect. I will not prolong this judgment by examining this particular point at length. I do not think that the articles touch the issue with which I am concerned. They are concerned with the obligations in international law between the States which are the parties to international treaties, not with issues of how private parties may or may not be able to rely against a State on the contents of a treaty. It is necessary to distinguish between two distinct questions: (1) How should the words of a treaty be interpreted? (2) Is the effect of the treaty, as properly interpreted, incorporated into domestic law so that private parties can rely upon it against others, including against the State? The articles of the Vienna Convention are material to question (1), not to question (2). In this case, furthermore, I have already held that the words of the treaty (the double taxation agreement, and in particular the crucial paragraph of the non-discrimination article) should be interpreted so as to have the meaning for which Mr Aaronson contends. The question which is determinative is question (2), which resolves itself into a question of the interpretation of s 788(3) of the United Kingdom statute. The articles of the Convention are not directed at all towards a question of that nature, and in my opinion they give no assistance upon it.

Mr Aaronson also urged me to pay careful heed to a passage in the judgment of Diplock LJ in Salomon v Commissioners of Customs & Excise [1967] 2 QB 116, especially at 143. I have read the passage carefully, but in my opinion it is of no assistance for much the same reason as that which I have explained in connection with the articles of the Vienna Convention. Diplock LJ is analysing the approach to the interpretation of a United Kingdom statute which seeks to enact the effect of an international treaty, and to do so, not in the actual words of the treaty itself, but rather in terms which the Parliamentary draftsman has chosen in order to reproduce what he understands the treaty to mean. The present case is not concerned with that sort of thing. Apart from anything else, in so far as the contents of a double taxation agreement become part of United Kingdom law, they do so in the actual words of the agreement (the treaty) itself. There is no question of the United Kingdom draftsman seeking to reproduce in his own chosen wording what he conceives to be the effect of the wording of the treaty. I repeat that the crucial question in this case is not what the non-discrimination article of the double taxation agreement means, but rather whether, given what it means, s 788(3) has introduced that meaning into domestic tax law. That is a question of interpretation of s 788(3), and Diplock LJ’s remarks, important ‘though they are in their context, are not relevant to it. Generally on this matter I agree with Mr Glick’s submissions that s 788 is a general enabling provision which traces back to 1945; it contemplates no particular treaty; and it is not appropriate to attempt to construe the breadth (or narrowness) of the enabling provision by reference to the detailed contents of subsequent treaties.

Mr Aaronson also sought to derive support from the Human Rights Act 1998 and from the European Convention on Human Rights for his contention that the non-discrimination article, in the particular application of it which he relies on here, is incorporated into domestic law. I do not accept his arguments in this respect either. He points out, no doubt correctly, that a right under national law to a tax benefit can be a ‘possession’ protected by article 1 of the First Protocol to the Convention. However, in my view that article cannot be prayed in aid in order to establish the existence of such a right under national law. If the right exists, then I would accept that under article 1 the taxpayer is entitled to peaceful possession of it and is not to be arbitrarily deprived of it. In this case, however, the question is whether the right which the claimants assert does or does not exist under national law (ie under domestic United Kingdom tax law). If, absent the Human Rights Act and the Convention, it would not exist under national law, then, present the Human Rights Act and the Convention, it still does not exist under national law. The Act and the Convention may create some rights which would not exist without them, but a right in United Kingdom subsidiaries of overseas parent companies to make group income elections is not one of them. (For completeness I mention that there is any event a major argument that the Human Rights Act could not operate retrospectively to create rights taken to have subsisted under United Kingdom law at times before the Convention became part of the law of England and Wales.)

Also in the context of the Convention Mr Aaronson drew my attention to the recent decision of the European court of Human Rights in SA Dangeville v France, reported at, among other places, [2003] STC 771. It is an interesting and potentially important case, and it may show a trend towards greater involvement of the Convention jurisprudence in tax law and procedure. Nevertheless, I cannot see anything in it which is of assistance on the question which I have to decide in this case, and I will not take time to discuss it further.

I have nothing more to say on the issue of whether the non-discrimination article, so far as it has the effect sought to be relied on by the claimants in this case, is or is not incorporated into United Kingdom domestic law. In my view it is not, and therefore the claim for restitution for infringement of the article fails in any event.

If I had taken the view that the article was incorporated into United Kingdom domestic law a number of issues concerning remedies, the quantum of restitution, and limitation of actions would have arisen. They were fully argued before me. I am not going to go into them at length, given that on the basis of what I have decided so far they do not arise, but I do think that I should describe what they were and in some respects make a few comments about them.

I should first record one point which Mr Glick made clear on behalf of the Revenue. In most cases where a taxpayer contends that the tax treatment which the Revenue seeks to apply to it is in breach of its rights under a double taxation agreement the taxpayer’s remedy is to appeal against a tax assessment or against the refusal of a claim and to have the issue decided within the recognised statutory machinery for resolution of appeals, beginning with a hearing before the Special or General Commissioners. (In cases of a sophistication such as those covered by the present GLO it would almost certainly be the Special Commissioners.) The Revenue’s position is that, if a taxpayer could sensibly have raised its argument challenging the tax treatment by way of appeal to the Commissioners, it cannot instead bring a claim for compensation or restitution in the High court: if it attempted to do that the Revenue would say that the High Court should strike the claim out on the ground that the proper procedure was for the taxpayer to pursue its statutory remedy by way of appeal. The Revenue has not pressed that argument in this particular case. I am not sure exactly why not, but two possible explanations occur to me, either of which seems satisfactory.

Nevertheless I think that the Inland Revenue would wish me to record that, although it has not taken a jurisdictional point in the present case, it should not be thought to be accepting tacitly that claims for compensation or restitution of taxes which a taxpayer says it ought not to have been required to pay can always be brought by way of a common law action. The Revenue position is that, if the argument could properly (and I would be inclined to add realistically) have been raised on a conventional tax appeal, it cannot be raised alternatively by a common law claim. The professional advisers of the taxpayers in this case, who are also involved in several other GLOs in which major questions of Community law or of the effect of non-discrimination articles are raised, do not take the same view as the Revenue, and this is likely to be a controversial issue in future cases.

A second point to make on remedies is that, if I had agreed that the denial of the ability to make group income elections amounted to a breach of the non-discrimination article in the double taxation agreements, the position as to remedies would not have been as simple as it was in the cases which followed Metallgesellschaft/Hoechst in the CJEC. In Metallgesellschaft/Hoechst it was part of the CJEC’s decision that national law had to give an effective remedy even if otherwise it would not have done. In the present case there is nothing comparable to the CJEC’s decision, and I have to consider whether a remedy exists under the common law. The claimants say that such a remedy does exist. They claim, not damages, but common law restitution either of tax paid pursuant to an unlawful demand (on the analogy of the Woolwich case — see the table in paragraph 1 of this judgment), or of payments made under a mistake of law (as in the Kleinwort Benson case — also identified in the table in paragraph 1 above — and in the DMG case). The Inland Revenue denies that a Woolwich-style claim could exist, because, as it correctly points out, the ACT payments in respect of which the claimants seek restitution were on any view lawfully payable and were not unlawfully demanded or paid. That is because of the point to which I referred in paragraph 56(1) above, namely that, unless and until a group income election has in fact been made and has been accepted by the inspector or upheld on an appeal against his refusal of it, ACT was lawfully payable on all dividends. I will not try to give an obiter decision on whether the Revenue is right that therefore a Woolwich-type restitution claim cannot succeed. I will, however, say that I can see that there could be substance in the point which the Revenue makes.

As respects a claim for restitution based on mistake of law, I think that the Revenue accepts that, if my decision in DMG is upheld on appeal (and if the effect of the non-discrimination articles was incorporated into domestic law), the claimants would have a restitutionary claim of that sort. Its real point here is that it hopes to establish in the House of Lords that my decision in DMG is wrong, in which case it would deny that, even if the relevant effect of the non-discrimination articles was incorporated into domestic law, the claimants in this case would have any sort of restitutionary remedy for the alleged infringements.

I add for completeness that, as my judgment in DMG explains, the ability to claim restitution on the ground of mistake of law is particularly important to the law of limitation of actions. A test case raising the limitation issues is included in the test cases before me on the present appeal. It involves the United Kingdom subsidiary of the US-owned Acushnet group. It was agreed, however, that the decision would at present be to the same effect as my decision in DMG, and the matter was not argued. I say no more about it here, but obviously the issue could become important once the outcome of an appeal in DMG is known.

Moving on from the basis of any remedy to which the claimants might be entitled to the quantum of it, there are three points which I will briefly make.

There is one other point on this part of the case which I should mention. I do not give any decision about it, but it concerns an argument which could be significant in later cases and as respects which I would not wish my judgment to be said to imply that there can be no force in the argument. The Inland Revenue contends that, if the claimants were entitled to any reliefs under domestic law arising from the double taxation agreements concerned, they had formally to claim them: see s 788(6). They have not done that, and (although I do not recall this particular point being gone into) they might by now be barred by time limits from making valid claims. I do not base any part of my decision on this point, and I express no view on whether it is a good point or not. It is, however, plainly a serious point which may have to be considered fully in some other case.

I have nothing further to say about the aspects of this case which arise in connection with the claimants’ arguments upon the non-discrimination articles of the relevant double taxation agreements.

I now come to the second part of the case which I mentioned in the overview at the beginning of this judgment, whereby the claimants wish me to refer certain arguments of Community law to the CJEC, but I am not prepared to do that.

The claimants wish to argue that, entirely independently of the non-discrimination articles in double taxation agreements, they may be entitled to compensation or restitution under European Community law on a basis which is analogous, though not identical, to the basis on which the CJEC held that the . claimants in Metallgesellschaft/Hoechst were entitled to compensation or restitution. Those claimants were so entitled because the CJEC considered that the ACT provisions operated to restrict the freedom of establishment to which nationals of Member States were entitled under article 43 of the EC Treaty. Article 43 prohibited provisions in United Kingdom law in so far as they operated as restrictions on the freedom of, for example, German companies to establish subsidiaries in this country. The CJEC considered that the ACT provisions were capable of operating as restrictions of that kind. Article 43 did not prohibit provisions in United Kingdom law in so far as they operated as restrictions on the freedom of, for example, Japanese or United States corporations to establish subsidiaries in this country. Therefore the claimants in the present test cases cannot claim compensation or restitution under Community law on the ground that it was a breach of article 43 for United Kingdom law to require them to pay ACT in consequence of paying dividends to their parent companies. But they argue that they can claim compensation or restitution under Community law on the ground that the application of ACT to the dividends which they paid was a breach, not of article 43, but of article 56.

Article 56 has two sub-articles, each of which is argued by the claimants in the present case to have applied. It is as follows. I italicise the critical words, and I also set out article 57(1), which is also crucial to the present issue.

At first sight it may seem curious to regard ACT as a restriction on movements of capital. However, the CJEC in Case C-35/98 Staatsecretaris van Financien v Verkooijen, decided that a Netherlands provision which affected the taxation of dividends in such a way as to encourage Netherlands residents to invest in shares in Netherlands companies rather than in shares in companies established in other States was a restriction on movements of capital. This was for the purposes of a Council Directive which had been in force at the time. It was not for the purposes of article 56(1), which was not in force at the time. However, there appears to be no material difference between the earlier directive and the present article 56(1). The payment of the dividend in Verkooijen was not itself a movement of capital, but the Netherlands tax rules affecting dividends might be expected to have some effect on the placing of investments, and thus on the capital movements connected therewith. Before me both parties accepted, and indeed both positively asserted, that, if the United Kingdom ACT provisions, which did apply to dividends paid by United Kingdom subsidiaries to non-resident parents but did not apply to dividends paid by United Kingdom subsidiaries to resident parents, were restrictions at all (something which the Inland Revenue does not concede), then they were restrictions on capital movements.

The claimants say that therefore the ACT provisions were restrictions within the scope of the prohibition in article 56(1). The claimants say further that the ACT provisions were also restrictions on ‘payments’ within the prohibition in article 56(2). They wish me to refer to the CJEC under article 234 of the EC Treaty the question of whether as a consequence they, the United Kingdom subsidiaries, have rights to compensation or restitution analogous to the rights of the claimant companies in Metallgesellschaft/Hoechst. The Inland Revenue, the defendant, accepts that, if the issue rested solely on article 56(1) and (2), the claimants’ arguments would be sufficiently maintainable for them to be appropriately referred to the CJEC. It says, however, that, when article 57(1) is taken into consideration, it is clear that the ACT provisions in their application to dividends paid by the United Kingdom to third countries outside the European Community cannot have been in breach of article 56.

I agree with the Inland Revenue. The starting point for the claimants’ argument is that the ACT provisions were both restrictions on capital movements in the sense of article 56(1) and restrictions on payments in the sense of article 56(2). Article 57(1) preserves from prohibition under article 56 the application to third countries (like Japan and the United States) of, among other things, restrictions which existed on 31 December 1993 under national law in respect of the movement of capital to or from third countries involving direct investment. The ACT provisions were provisions of national law which existed on 31 December 1993, and an area where they applied was to the payment of dividends on direct investments by third country shareholders in United Kingdom companies. As the Verkooijen case demonstrated, tax provisions of that type were restrictions on movement of capital. It appears to me clearly to follow that the ACT provisions are excluded from article 56. Further, on the plain wording of article 57(1), they are excluded from the whole of article 56, that is from both article 56(1) and from article 56(2). The claimants wish to contend that article 57(1) is an exclusion from article 56(1) but not from article 56(2). But article 57(1) begins ‘The provisions of Article 56 shall be without prejudice to ...’. It does not begin ‘The provisions of Article 56(1) shall be without prejudice to ...’. It seems plain to me that any provision which is within article 57(1) is taken out of any part of article 56 if it would otherwise have been within it. Of course, if it would otherwise have been only within article 56(1) and not within article 56(2) in the first place, it can only be taken out of article 56(1). But if the ACT provisions are within both article 56(1) and article 56(2) (as the claimants themselves say that they are), then they are taken out of both article 56(1) and article 56(2).

In the circumstances I reject the claimants’ argument based on article 56, and I do not think it appropriate to refer the article 56 question to the CJEC. I do not think that there is sufficient doubt about it for me to do that. I have carefully studied the judgment of Chadwick LJ in Trinity Mirror plc v Customs and Excise Commissioners [2001] STC 192 at 211 to 213, and the citations therein from relevant authorities. Applying the principles there set out it appears to me that this is not a case where I should make a reference under article 234. I avoid using the expression ‘acte claire’, which seems to me principally appropriate to be considered in cases where the issue is whether or not a reference should be made by a court of final appeal, which in this country usually means the House of Lords. However, I will say this. If I thought that the there was room for serious argument against the Inland Revenue’s answer based on article 57(1) I would make a reference. Since I do not think that, I make no reference and I decide the issue myself.

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 paragraphs 22 to 25 of my judgment in the Hoechst case (see the table in paragraph 1 at the beginning of this judgment). My conclusions are as follows.

I hope, that I have now covered all of the matters which arose in this substantial case. I have nothing further to add. The overall result is that the claims are dismissed, and so is the request that I refer a question for the ruling of the CJEC under article 234 of the EC Treaty.