OPINION OF ADVOCATE GENERAL

FENNELLY

delivered on 12 September 2000 (1)

Joined Cases C-397/98 and C-410/98

Metallgesellschaft Limited, Metallgesellschaft AG and Metallgesellschaft Handel & Beteiligungen AG, The Metal and Commodity Co. Ltd

v

Commissioners of Inland Revenue and HM Attorney General

Hoechst AG and Hoechst UK Ltd

v

Commissioners of Inland Revenue and HM Attorney General

1.

    This reference from the High Court of Justice (England and Wales), Chancery Division (hereinafter the 'national court‘), concerns the extent to which, in the absence of harmonised Community rules, Member State competence in respect of direct corporate taxation is limited by the overriding requirements flowing from the fundamental Treaty principles of free movement. The Court is particularly asked to consider whether the relatively novel notion of fiscal cohesion can be invoked to justify a difference in the tax treatment of certain corporate taxpayers based on the place of residence of their parent companies. If it does not, a consequential question is whether Community law requires that a remedy, either for restitutionary or compensatory damages, be available in national law notwithstanding that the differential treatment resulted merely in the early payment of tax.

I - The legal and factual context

A - The relevant United Kingdom legal provisions

2.

    Under sections 8 and 11 of the Income and Corporation Taxes Act 1988 (hereinafter 'the ICTA 1988‘), corporation tax (hereinafter 'CT‘) is charged on the profits of United-Kingdom-resident companies, as well as on those of non-resident companies which trade in the United Kingdom through branches or agencies, during a given accounting period. (2) For accounting periods ending before 1 October 1993, CT was payable nine months after the end of the accounting period or one month after the issue of the notice of assessment relating to that accounting period, whichever was the later. Since 1 October 1993, it has been payable nine months and one day after the end of the accounting period. (3)

3.

    The present case concerns the payment of advance corporation tax (hereinafter 'ACT‘). (4) It is important to note that the national court has unambiguously found as a matter of United Kingdom law that, under section 4 of the ICTA 1988, '[ACT] is corporation tax and that there is nothing in any other provision of the 1988 Act which calls that into question‘. (5) Under section 14, certain 'qualifying distributions‘, most typically the payment of dividends, entailedthe obligation to pay ACT. Any company resident in the United Kingdom which made such a distribution was liable, on that occasion, to pay ACT on a sum equal to the amount or value of the 'distribution‘ made. Companies were obliged to make quarterly returns showing the amount of any 'distributions‘ made during that period. The resulting ACT was payable within 14 days of the end of that quarterly period. (6)

4.

    In principle, any ACT paid during an ACT accounting period could be set off against the paying company's mainstream CT liability for the CT accounting period in question, or, alternatively, transferred to its subsidiaries, which could then set it off against mainstream CT for which they were liable (sections 239 and 240). CT was not payable until nine months after the expiry of each corporate tax year. However, ACT had to be paid within 14 days of the end of the relevant quarterly ACT period so that, as the national court has found, '[T]he effect of ACT‘, on a company which chose to distribute profits by way of dividend, '[wa]s therefore to advance the date for payment of the CT which would otherwise be due, by a period that w[ould] vary from eight and a half months (in the case of a distribution made on the last day of an accounting period) to one year and five and a half months (where the distribution [wa]s made on the first day of an accounting period)‘. Moreover, where no mainstream CT was payable in respect of the period in question, it is also pointed out in the order for reference that the ACT paid 'c[ould] be set off against profits of subsequent periods, in which case the advance w[ould] have been made for a longer and perhaps indefinite period‘.

5.

    Central to the present case, however, is the exemption from liability to pay ACT which was available where a subsidiary and its parent company made a 'group income election‘ pursuant to section 247 of the ICTA 1988. Such an election was open only to companies one of which owned at least 51% of the other and both of which were resident in the United Kingdom. The effect of exercising such a right of election was that the subsidiary (the paying company) was not required to pay ACT in respect of dividends paid to its parent company, unless it gave notice that it did not wish the group income election to apply in respect of a particular dividend. The request for group income election had to be made to an Inspector of Taxes. If the request was rejected, the requesting company could appeal to the Special or General Commissioners, from whose decision an appeal on a point of law lay (in England and Wales) to the High Court of Justice.

6.

    The alternative claim advanced in the main proceedings concerns the entitlement to a tax credit in respect of ACT paid. Under section 231(1) of the ICTA 1988, the payment of ACT by a subsidiary on dividends distributed to its parent company entitled the latter to a tax credit provided the parent company was resident in the United Kingdom. The amount of the tax credit was equal to the amount of ACT paid by the subsidiary. Such a tax credit could be used by theparent to offset its own liability to pay ACT when it made distributions to its shareholders; i.e. it would be liable to pay ACT only on the excess of those later dividends over those received from its subsidiary. Where a company that was resident in the United Kingdom, but wholly exempt from mainstream CT liability, received a dividend from a subsidiary in respect of which ACT had been paid, it was entitled to a reimbursement of an amount equal to the tax credit.

7.

    Under section 208 of the ICTA 1988, '... [CT was] not chargeable on dividends or other distributions of a company resident in the United Kingdom, nor [were] any such dividends or distributions to be taken into account in computing income for [CT]‘. Non-resident companies, on the other hand, or those which did not trade in the United Kingdom through a branch or agency, although not chargeable to CT, were, in principle, subject to United Kingdom income tax in respect of income having its source in the United Kingdom, which included dividends paid by companies resident in the United Kingdom. However, under section 233(1) of the ICTA 1988, in so far as a non-resident parent company of a United-Kingdom-resident subsidiary was not entitled to a tax credit in respect of such a dividend, it was not subject to a charge to United Kingdom income tax on it either. Conversely, where it was entitled to a tax credit by virtue of a double taxation convention ('DTC‘) in force between the United Kingdom and its country of residence, it was chargeable in the United Kingdom to income tax on dividends received from its United-Kingdom-resident subsidiaries.

8.

    The DTC of 26 November 1964 between the United Kingdom and Germany, as amended on 23 March 1970, grants no right to a tax credit to companies resident in Germany and holding shares in and receiving distributions from companies resident in the United Kingdom. Accordingly, under United Kingdom law, a German parent company is not assessable to tax in respect of dividends received from its subsidiary resident in the United Kingdom and is therefore not entitled to any tax credit. However, a number of DTCs between the United Kingdom and certain EU Member and non-member countries enable parent companies resident in the other country to obtain at least a partial tax credit. (7) Under the DTC between the Netherlands and the United Kingdom, which has been relied upon by the plaintiffs in the main proceedings, the relevant charge to tax in respect of the Netherlands-resident parent companies amounted, at the material time, to 5% of what may conveniently be described as the'grossed-up amount‘ of the dividend, namely the total of half the tax credit plus the dividend. (8)

B - Facts and reference

9.

    The companies involved in Case C-397/98 are Metallgesellschaft Limited, The Metal and Commodity Company Limited, both of which are companies incorporated and resident in the United Kingdom, and Metallgesellschaft AG and Metallgesellschaft Handel & Beteiligungen AG, which are both incorporated and resident in the Federal Republic of Germany (hereinafter 'Metallgesellschaft and Others‘). The companies involved in Case C-410/98 are Hoechst AG, a company incorporated and resident in Germany, and its United-Kingdom-resident subsidiary, Hoechst UK Ltd (hereinafter 'Hoechst‘). Metallgesellschaft Limited, The Metal and Commodity Company Limited and Hoechst UK Ltd (hereinafter 'the United Kingdom subsidiaries‘) each paid, over periods stretching from 1974 to 1995, ACT in respect of dividends paid to their German parent companies. (9)

10.

    In the main proceedings, both sets of companies (hereinafter referred to collectively as 'the plaintiffs‘) brought actions in 1995 before the national court in which they maintained that their United Kingdom subsidiaries had suffered a cash-flow disadvantage in comparison with the subsidiaries of parent companies resident in the United Kingdom, since, unlike the latter, which were permitted to benefit from a group income election, no such option was available to them. They claim principally that that disadvantage constituted discrimination contrary to Articles 6 and 52 of the EC Treaty (now, after amendment, Articles 12 EC and 43 EC). By way of an alternative limb to this claim, the plaintiffs contend that the contested restriction on the entitlement to make a group income election infringed Article 73b of the EC Treaty (now Article 56 EC). Their second and alternative claim is that the parents should be entitled to a tax credit corresponding, at least in part, to the ACT paid by the United Kingdom subsidiaries. By way of a remedy, they are seeking damages or compensation for the loss of the use of the money inrespect of the periods between the payments of ACT made and the time when their mainstream CT, against which those payments were set off, was due.

11.

    The defendants in the main proceedings (the Commissioners of Inland Revenue and the Attorney General) contended that ACT was designed to ensure that the company making the distribution made a payment to match the tax credit or income tax exemption given to the shareholder. If resident subsidiaries of non-resident parents could distribute profits free of ACT, the result would facilitate tax avoidance since neither the parent nor the subsidiary would have to pay ACT, while, conversely, in the case of a United-Kingdom-resident parent company, ACT would be payable once a distribution was made by them outside the group. Thus, any differentiation based on the place of residence of a subsidiary company's parent company was justified. As to the alternative tax-credit claim, the fact that there is no provision for tax credits in the United Kingdom-Germany DTC, while such a provision exists in certain other DTCs, reflects differences between the German tax system and those of the other countries concerned, as well as the result of the overall negotiated arrangements agreed by the parties to the various DTCs. The defendants also deny that any breach of Community law which might have occurred gives rise to an actionable claim for damages. Moreover, they maintain that, as a matter of English law, interest cannot be claimed by way of damages or restitution where, as in the present case, no principal sum is owing.

12.

    The national court points out that it is common ground that:

-    under United Kingdom law, group income elections can be made only where both the parent company and its subsidiary are resident in the United Kingdom;

-    the plaintiffs never made a group income election but could have been expected to do so if they had appreciated that Community law required the right of election to be available where the parent company was non-resident;

-    if they had tried to exercise such a right, their application would have been rejected by an Inspector of Taxes, since the parent companies were not resident in the United Kingdom, but that such a rejection could have been challenged;

-    before any such challenge was finally determined, the plaintiffs would still have been obliged, on pain of financial penalties including possible statutory penalties (if they were deemed to have acted negligently and without reasonable excuse in failing to provide such an account of dividends paid), to pay the ACT relating to all the dividends which they had paid;

-    under the law of the United Kingdom there would have been no right to the repayment of such ACT even if the action had been successful. (10)

13.

    The following questions have been referred to the Court:

'(1)    In the circumstances set out in the order for reference, is it consistent with Community law and, in particular, with Articles 6, 52, 58 and/or 73b ^[(11)^] of the EC Treaty for the legislation of a Member State to permit a group income election (allowing distributions to be paid by a subsidiary to its parent without accounting for advance corporation tax) only when both the subsidiary and parent are resident in that Member State?

(2)    If the answer to Question 1 is ”no”, do the abovementioned provisions of the EC Treaty give rise to a restitutionary right to a resident subsidiary of a parent company resident in another Member State and/or the said parent to claim a sum of money by way of interest on the ACT which the subsidiary paid on the basis that the national laws did not allow it to make a group income election, or can such a sum only be claimed, if at all, by wayof an action for damages pursuant to the principles laid down by the Court of Justice in Joined Cases C-46/93 and C-48/93 Brasserie du Pêcheur SA v Germany and R v Secretary of State for Transport, ex parte Factortame and Others ^[(12)^] and Case C-66/95 R v The Queen v Secretary of State for Social Security, ex parte Eunice Sutton, ^[(13)^] and in either case is the national court obliged to grant a remedy even if under national law interest cannot be awarded (whether directly or by way of restitution or damages) on principal sums no longer owing to the plaintiff?

(3)    In the circumstances set out in the order for reference, is it consistent with the abovementioned provisions of the EC Treaty for the authorities of one Member State to deny any tax credit to a company resident in another Member State when it grants such credit to resident companies and to companies resident in certain other Member States by virtue of the terms of its double taxation conventions with those other Member States?

(4)    If the answer to Question 3 above is ”no”, is and was the first Member State at all material times obliged to make a tax credit available to such company on the same terms as to resident companies or as to companies resident in Member States with provision for such credits in their double taxation conventions?

(5)    Is a Member State entitled to plead in answer to such a claim for restitution, tax credit or damages, that the plaintiffs are not entitled to recover, or that the plaintiffs' claim should be reduced, on the grounds that, despite the terms of the national Statute which prevented them from doing so as a matter of national law they ought to have made a group income election, or claimed a tax credit and have appealed to the Commissioners and, if necessary, the courts against the decision of the inspector of taxes refusing the election or claim, relying upon the primacy and direct effect of the provisions of Community law?‘

II - Observations

14.

    Written and oral observations were submitted by the plaintiffs, the United Kingdom of Great Britain and Northern Ireland, the Kingdom of the Netherlands and the Commission. In addition, the Republic of Finland submitted written observations, while the Republic of Austria and the French Republic, as well as the Federal Republic of Germany, presented oral observations. The observations of the plaintiffs, the United Kingdom, Finland and the Commission treat of thevarious issues raised by the questions referred. The Netherlands has submitted observations in respect of the first and third questions concerning the group income election and the tax credits, while Austria, France and Germany limited their oral observations to the third question.

III - Overview

15.

    The plaintiffs' fundamental claim is that the United Kingdom subsidiaries' exclusion from the possibility of making a group income election, on the basis that their parent companies were resident in Germany, was incompatible with the freedom of establishment guaranteed by Article 52 of the EC Treaty (now, after amendment, Article 43 EC) and Article 58 of the EC Treaty (now Article 48 EC). (14) In support of their principal claim, it is also clear that the plaintiffs rely principally on Article 52 of the EC Treaty and not on the Treaty provisions concerning the free movement of capital. It is appropriate, therefore, first to consider the compatibility with Community law of limiting the availability of a relief such as group income relief to companies resident in the United Kingdom.

IV - Question 1 and group income relief

A - Synopsis of the observations

16.

    The plaintiffs submit that Member States must exercise their fiscal sovereignty in respect of their domestic systems of direct taxation consistently with the fundamental principles of Community law. Denying the possibility of making a group income election deters non-United Kingdom companies from establishing subsidiaries in the United Kingdom and is incompatible with the Treaty-guaranteed freedom of establishment. By placing the subsidiaries of foreign companies on the same footing as those of companies established in the United Kingdom for the purposes of their respective CT liabilities, the United Kingdom has acknowledged that there is no objective difference between them which could justify a difference in treatment with regard to a tax advantage like the group income election. (15) It is immaterial that it is possible to avoid ACT by setting up branches or agencies rather than subsidiaries, since Articles 52 and 58 of the Treaty require that traders be free to choose the appropriate legal form in which to exercise the right of establishment. (16) The difference in treatment is not justified on grounds of preventing tax avoidance, since the only effect of granting the right to make agroup income election would be to postpone, until the time when mainstream CT liability arises, but not remove the ultimate tax liability of the subsidiary. The fact that a non-resident company is not required to pay ACT when it later pays a dividend, because it is not subject to United Kingdom CT, whereas a similar payment by a United-Kingdom-resident company would be so subject, does not justify the difference in treatment: first, there is no tax avoidance since the former will be subject to the tax legislation of the State in which it is established; (17) secondly, the possible diminution in the tax revenue of one Member State is neither a ground listed in Article 56 of the EC Treaty (now, after amendment, Article 46 EC) nor a matter of overriding general interest capable of justifying unequal treatment contrary to Article 52. (18)

17.

    Furthermore, the denial of the group income election is not justified in the interests of fiscal cohesion. In Bachmann v Belgium and Commission v Belgium, (19) there was a 'direct link‘ between the tax deductibility of contributions and the taxation of sums payable by insurers under sickness and invalidity as well as under old-age and life assurance policies. Under Belgian tax law, the loss of revenue resulting from the deductions allowed against taxable income was compensated by the taxation of sums payable under such policies in respect of the same taxpayer. There is no such link between the exercise of a group income election by a subsidiary in respect of a dividend paid to its parent company and the payment of ACT by the latter on the subsequent distribution of a similar amount outside a group income election. The subsidiary remains liable to United Kingdom CT. The plaintiffs draw attention to the different provisions in force in Ireland in respect of non-resident parent companies. (20) They submit that the United Kingdom's blanket refusal of the election to groups with foreign-resident parent companies was disproportionate.

18.

    The United Kingdom submits that the difference of treatment in respect of group income elections is justified by the need to preserve the cohesion of its tax system. The principle is that there should be a charge to CT on company profits while their shareholders should be subject to income tax whenever profits aredistributed to them by way of dividend. As the straightforward application of that principle would result in the double taxation of the same profits, i.e. once in the hands of the company and again in the hands of the shareholder, the partial imputation system introduced in 1973 and reflected in the ICTA 1988 was adopted. It mitigates such double taxation by exempting corporate shareholders resident in the United Kingdom from CT on the dividends they receive. As a company may have distributable profits and make distributions without making a taxable profit, to ensure that the tax exemption for the dividend in the hands of the corporate shareholder is matched by a charge to tax, the company paying the dividend is required to account for the ACT. This ensures that, before any relief or mitigation is afforded to the shareholder, there is a liability to tax on the paying company. (21) There is thus, in its view, 'a clear and direct link between the tax exemption accorded to the dividend in the hands of the shareholder, and the matching charge to ACT‘, while '[T]he logic of allowing a group income election for dividends paid between a subsidiary and a parent company is that such distributions amount in effect to internal transfers within a single economic entity (even though between two corporate entities)‘. (22) On the other hand, the exemption from liability to United Kingdom tax of the receipt by a non-resident company of dividends from a company resident in the United Kingdom is matched by a charge to ACT.

19.

    The United Kingdom, supported by Finland and the Netherlands, submits that the difference in treatment between groups with resident parent companies and those with non-resident parents regarding such group income elections is objectively justified because the positions of the two respective groups are not comparable; where the parent is resident, the exemption from ACT on the occasion of a distribution by the subsidiary (which matches the tax exemption of the dividend in the hands of the parent company) is itself matched by a charge to ACT on the occasion of making a distribution by the parent company, whereas, in the case of non-resident parents, the waiver of ACT on the occasion of making a distribution by the subsidiary is matched by no corresponding payment. Whilst acknowledging that a different system could be applied, the United Kingdom denies that the restriction on the availability of the election is disproportionate. In Bachmann it would have been possible for the Belgian tax rules to have allowed nationals of other Member States to deduct life assurance contributions paid in other Member States, notwithstanding the fact that no tax would be paid inBelgium on the sums paid out in due course by the insurers. (23) Yet the Court upheld Belgium's right to formulate its own tax system. The legislative choice made by the United Kingdom falls within the legitimate range of choices allowed to Member States by Community law.

20.

    The Commission submits that there is no justification for the difference in treatment. The mere desirability of ensuring that the profits of the United Kingdom subsidiary of a non-resident company should bear a certain minimum amount of United Kingdom tax does not justify requiring ACT to be paid by such subsidiaries earlier than the normal date for the payment of mainstream CT. ACT is an advance payment of CT, but non-resident parent companies are not liable to United Kingdom CT. A group income election does not enable a subsidiary to escape its United Kingdom mainstream CT liability and allowing it to subsidiaries of non-resident parents would not therefore assist tax avoidance. There is no loss of tax revenue for the United Kingdom tax authorities in respect of the subsidiary's profits, since the deferral of ACT inherent in the group election relieves the subsidiary only of its obligation to make advance payments of CT. The mere economic advantage for the United Kingdom of receiving such CT payments in advance so as to compensate for the fact that no ACT will later be payable by non-resident parents in respect of dividends received from United Kingdom subsidiaries cannot justify such discrimination. (24)

B - Analysis

21.

    The Court has consistently emphasised that 'although direct taxation falls within their competence, the Member States must none the less exercise that competence consistently with Community law‘. (25) The issues raised in this case essentially concern whether the United Kingdom respected the limits imposed by Community law on its fiscal sovereignty in the direct taxation field when it restricted group income elections in respect of ACT to United-Kingdom-resident companies.

(i) Tax revenues

22.

    I agree with the plaintiffs' submission that extending the right to make a group income election to United Kingdom subsidiaries of non-resident parents would not facilitate tax avoidance or evasion. It is clear that ACT is nothing otherthan an advance payment of mainstream CT. The exercise of a group income election by subsidiaries of non-resident parents would merely enable them to enjoy the same cash-flow advantage as that enjoyed by subsidiaries of United-Kingdom-resident companies. In the case of both types of subsidiary, mainstream CT liability would, in due course, arise in the same manner on their respective corporate profits. In my opinion, permitting subsidiaries of non-resident companies to make group income elections would not facilitate the avoidance of ACT on the part of their parents. As a non-resident parent company is not liable to United Kingdom CT, it should not be liable to pay ACT either. The situations of the resident and non-resident parent are not objectively comparable. The former will be obliged to pay ACT if and only if it makes a qualifying distribution for the very reason that it is liable to pay mainstream United Kingdom CT on its own profits, while the latter is not liable to pay United Kingdom CT but will be subject to the tax laws of its country of residence. The relevance of this factor was recognised by the Court in ICI v Colmer (HMIT), which, moreover, also unambiguously rejected the argument advanced by the United Kingdom in that case that a diminution in tax revenue could justify the discriminatory treatment of non-resident subsidiaries. (26)

23.

    Furthermore, allowing a subsidiary with a non-resident parent to make an election would not be to grant any 'exemption‘ from ACT. It is a fallacy to speak in terms of an 'exemption‘. (27) The true position is that United Kingdom CT is due by the subsidiary of the non-resident parent but the latter is not subject to CT on its own profits in the United Kingdom. The fact that parent companies resident in the United Kingdom may, in certain cases, have to pay ACT on the excess of their own dividends over those received from their subsidiaries cannot justify imposing an obligation on United Kingdom subsidiaries of non-resident companies always to pay ACT whenever they opt to pay dividends to their parents. As is mentioned in the previous paragraph, this flows from their objectively different situation, to wit from their fiscal residence in the United Kingdom.

24.

    The United Kingdom considers that the effect of a group income election on dividend payments made within the group is to transform them into mere 'internal transfers‘ whose effect is to transfer the subsidiary's liability, whenever a dividend is paid outside the group, to pay ACT to the parent company. Since a non-resident parent would not be liable to pay ACT even though, under a group income election, its subsidiary's liability to ACT would have been transferred to it, it is said that its situation is distinguishable from that of the resident subsidiary and parent companies. The Netherlands expresses this in another way by invoking the principle of territoriality, which, it asserts, would be infringed if one Member State were required to treat a company established in another Member State and not doing business in the former State as part of a fiscal group in that State. It shouldonly be possible, in its view, to transfer profits and losses between those members of a corporate group which are resident in the same Member State.

25.

    These submissions effectively amount to a contention that the difference in treatment of subsidiaries of non-resident companies is justified by the fact that they are not subject to United Kingdom CT liability on their profits. This difference in the respective fiscal situations of the parents does not provide a justification, in my view, for denying to the subsidiaries duly established in the United Kingdom, who are so subject, a tax advantage available to comparable subsidiaries of United Kingdom parents.

(ii) Fiscal cohesion

26.

    The United Kingdom relies principally on Bachmann and Commission v Belgium to justify its refusal to allow group income elections by subsidiaries of non-resident companies, citing the need to preserve the fiscal cohesion of its tax system. In that case, a Belgian tax rule differentiated between contributions made to Belgian insurance companies and contributions made to other non-resident insurance companies. Only those contributions which were made to resident insurance companies were tax deductible. The loss of revenue thereby resulting for Belgian tax revenues was offset by subjecting the capital sums or surrender values of the policies in question to tax, a tax which was not payable where there had been no deductions of contributions. Belgium justified its differential treatment of insurance contributions by reference to the need to ensure the coherence of its fiscal system. In particular, it could not be sure that tax on capital sums could be collected from non-resident insurance companies.

27.

    The Court accepted this argument. It found that 'the cohesion of such a system ... presupposes that, in the event of a State being obliged to allow the deduction of life assurance contributions paid in another Member State, it should be able to tax sums payable by insurers‘. (28) It was not satisfied that an undertaking by an insurer to pay the tax in question could 'constitute an adequate safeguard‘, (29) since it would probably have to be supported by the deposit of a guarantee, the overall effects of which 'would involve the insurer in additional expense‘ (30) which would be passed on and probably remove the incentive for migrant workers to retain their existing policies on moving to Belgium. Although it recognised that bilateral treaties allocating fiscal competence between Member States or harmonised Community direct taxation rules could remove Belgium's concern, it concluded that, as Community law then stood, 'it [was] not possible toensure the cohesion of such a tax system by means of measures which are less restrictive than those at issue in the main proceedings ...‘. (31)

28.

    The Court did not in Bachmann define the notion of fiscal cohesion and that remains the only case in which a Member State has successfully invoked it to defend a national provision otherwise incompatible with one of the fundamental Treaty provisions. Wielockx v Inspecteur der Direckte Belastingen (32) was a case which, at first sight, was reasonably comparable with Bachmann. In the Netherlands, resident but not non-resident taxpayers could deduct payments to a pension reserve. Payments made on the liquidation of the reserve, or made periodically from it, were treated as income and subject to tax. The Netherlands relied, inter alia, on the Netherlands-Belgium DTC, under which such income was taxable only in the State of residence, to claim that the fiscal cohesion of its system would be jeopardised if Belgian residents like Mr Wielockx could deduct payments into the reserve from their Netherlands tax liability. The Court stated that the fiscal cohesion discussed in Bachmann required 'a correlation between the sums which are deducted from the taxable income and the sums which are subject to tax‘. (33) It noted that the effect of the DTC was that the State of residence would tax all pensions received by its residents regardless of where the contributions were paid and, conversely, waive the right to tax pensions received abroad even where it had treated the contributions made thereto on its territory as being tax deductible. The Court ruled that '[F]iscal cohesion ha[d] not therefore been established in relation to one and the same person by a strict correlation between the deductibility of contributions and the taxation of pensions but [wa]s shifted to another level, that of the reciprocity of the rules applicable in the Contracting States‘. (34) Where fiscal cohesion is secured by a DTC with another Member State, 'that principle may not be invoked to justify the refusal of a deduction such as that at issue‘. (35)

29.

    A similar strict approach was applied a year later to a different tax situation in Asscher v Staatssecretaris van Financiën. (36) The Netherlands tax authorities sought to justify applying a higher initial (first band) rate of income tax to non-resident than to resident taxpayers. A less favourable rate of taxation for non-residents could not be justified by pointing to the fact that social security contributions were no longer deductible in the Netherlands, which was notnecessarily the case in other Member States. The Court held that there was 'no direct link between the application of a higher rate to the income of certain non-residents who receive less than 90% of their worldwide income in the Netherlands and the fact that no social security contributions [were] levied on the income of such non-residents from sources in the Netherlands‘. (37) In ICI v Colmer (HMIT), the Court rejected the United Kingdom's submission that fiscal cohesion required that consortium relief, whereby the members of a consortium could transfer losses incurred by subsidiaries of a holding company owned by them for relief against their own profits, be limited to cases where the majority of the subsidiaries in question were United Kingdom residents. (38) It emerges clearly from those cases that a mere threat to fiscal revenues of a Member State does not qualify for consideration as fiscal cohesion in the sense recognised by the Court.

30.

    National rules designed to alleviate double imposition of tax on the same or similar economic activity have led Member States to provide certain tax advantages which are generally limited to resident individuals or companies. The pursuit of such a policy, which is clearly legitimate and desirable in itself, underlies three of the more recent cases concerning fiscal cohesion. The problem is that, in withholding the benefits in question from non-residents, Member States refuse to take account of the foreign taxes which they pay.

31.

    Eurowings Luftverkehrs v Finanzamt Dortmund-Unna (39) concerned certain German trade tax rules which treated traders leasing assets from non-resident lessors less favourably than those leasing from residents. The trade tax was calculated by adding back half the rental value of the assets for the purpose of calculating leasing income only if the lessor did not pay German trade tax. The cohesion argument was that the rules in question were designed to avoid only double payment of German tax, i.e. not the duplication of German tax and that of another Member State. The Court held there was merely an 'indirect link‘ between a fiscal advantage accorded to a German lessee of a German-established lessor and the unfavourable tax treatment of such lessors in the form of their liability to pay tax on their rental income. (40) At issue in Verkooijen was a partial exemption from personal income tax, conferred in respect of share dividends, provided the companies paying the dividends were established in theNetherlands. (41) The Netherlands and other intervening Member States supported the limitation of this advantage to the case of dividend income from resident companies. The double taxation which the rules were designed to avoid was CT and income tax on the same profits or income, and this would not arise if the former were paid in another Member State. The Court held that there was no 'direct link‘ but, rather, 'two separate taxes levied on different taxpayers‘. (42) Baars also concerned Netherlands tax law, in this case wealth tax. (43) A taxpayer was entitled to certain exemptions in respect of 'substantial holdings‘ in companies provided they were established in the Netherlands. The exclusion of holdings of companies established in other Member States raised very similar issues regarding the claimed objective of precluding double taxation as between the imposition of taxation on the company and a personal wealth tax on the holder of its shares as those which arose in Verkooijen. The Court, rejecting the defence, held that it was 'irrelevant ... that companies established in the Netherlands are subject to CT in the Netherlands and that companies established in another Member State are not‘. (44)

32.

    The cases on fiscal cohesion have arisen in the context of all of the Treaty freedoms: Bachmann and Asscher concerned the free movement of persons; ICI and Baars concerned the freedom of establishment; Eurowings concerned a recipient of services; while Verkooijen concerned the free movement of capital. In all cases, save Bachmann, the Court held that the national rules in question could not be justified by any notion of fiscal cohesion. The last three cases concerned the aim of avoiding double taxation, which was restricted to mean only two national taxes. The Court stated on each occasion that there was either no 'direct link‘ between the tax differentiation in question and the proclaimed object of the system, or that there was no or an insufficient correspondence between the different taxpayers and the taxes at issue. It is clear that a mere diminution in the tax revenues of the host Member State cannot justify a refusal to extend a particular benefit to non-resident companies. That Member State must take account of the liability of such non-residents to pay comparable taxes in their Member State of residence. Thus, it would seem that the true scope for fiscal cohesion as a justification for the differential treatment of non-residents would concern only situations in which there is a real and substantial risk that extending equal treatment in respect of a particular benefit would potentially facilitate tax evasion in both the host Member State and the Member State of residence of theclaimant non-resident taxpayer. This may well have been the real concern underlying the now unique judgment in Bachmann. (45)

33.

    What is clear, in any event, is that for the defence to succeed there must be a direct and, from the point of view of the application of the particular tax in question, fundamental organic link between the application of that tax and the exemption or relief therefrom, which, though made available to the resident taxpayer, is denied to his non-resident counterpart. In my view, such a strict correlation is lacking in the present case.

34.

    The argument advanced by the United Kingdom centres on the theory that an element of the recipient shareholder's income tax is imputed to the charge to CT of the company paying the dividend. The United Kingdom contends that the element in question is reflected in that part of the dividend-paying company's CT liability that was payable in advance by way of ACT. The United Kingdom revenue authorities maintain that they require the advance use of the revenues generated by way of ACT on the payment of dividends so as to compensate them for any later reimbursements of the income tax treated as imputed to those ACT payments which they have to make in favour of certain individual shareholders, who, although receiving such dividends, are not, for one reason or another, liable to pay any United Kingdom income tax. (46)

35.

    I do not accept that this argument justifies the unfavourable treatment of non-resident taxpayers. It is based on the misconception that ACT may, somehow, be regarded as a separate tax from mainstream CT. Since there is no question regarding the liability of the subsidiaries of both resident and non-resident parent companies to pay United Kingdom CT, the grant to one but not to the other of a significant tax advantage cannot be justified by a difference in the CT liability of the parent companies to which the dividends are paid. In other words, the objectively different CT positions of resident and non-resident parents cannot justify the imposition of an effectively higher CT burden only on the subsidiaries of the latter.

36.

    Moreover, in the case of individual shareholders of those parent companies who are resident in other Member States and who are, thus, subject to the tax laws of those States, there is nothing on the case-file to suggest that the United Kingdom authorities have ever been obliged to make any income tax reimbursements. There would, in the case of non-resident parent companies who make downstream dividends to individual shareholders of profits distributed to them by their United-Kingdom-resident subsidiaries, be at most a remote link between allowing a group income election regarding the subsidiaries' obligation to pay ACT and possible claims, by the parents' own shareholders, for the reimbursement of the (United Kingdom) income tax element imputed to the dividends paid by those subsidiaries. This is a fortiori the case since dividends paid by non-resident parent companies do not carry a United Kingdom tax credit merely because they were themselves funded from dividends received from United-Kingdom-resident subsidiaries. There is therefore no real and substantial risk to the cohesion of the United Kingdom tax system capable of justifying the differential treatment at issue.

37.

    If the Court were to disagree with this recommendation, the outright refusal in the United Kingdom rules to extend the exemption from ACT to subsidiaries of non-resident companies would, at all events, appear to be disproportionate. I do not accept the United Kingdom's submission to the effect that, once legitimate concerns regarding fiscal cohesion underlie a differential fiscal treatment of non-residents, the Member State concerned is not obliged to take into account the fact that there may be less restrictive means of achieving the desired coherence. Thus, I do not agree with the contention that, in answer to a specific claim of discrimination contrary to Article 52 of the Treaty made in respect of its rules, the United Kingdom was not obliged to consider the appropriateness of the less restrictive rules regarding non-resident parents applied by another Member State (to wit Ireland) operating a very similar system of ACT.

38.

    In any event, as the Commission submits, the objective underlying ACT could just as easily have been achieved by the imposition of a general requirement, on some or all companies, to pay a certain proportion of their CT liability in advance. Indeed, as the plaintiffs point out, this, indeed, is the system which has recently been introduced in the United Kingdom, at least in respect of largercompanies, by sections 30 and 31 of the Finance Act 1998. The refusal to permit subsidiaries of parent companies resident in other Member States to make a group income election for ACT purposes thus clearly went beyond what might conceivably have been capable of justification on the ground of preserving the fiscal cohesion of the system established by the ICTA 1988.

39.

    For all of the above reasons, I am satisfied that a restriction on the availability of a tax advantage, such as the exemption from the obligation to make advance payments of CT inherent in a group income election of the sort at issue in the main proceedings, is incompatible with Article 52 of the EC Treaty.

40.

    In the circumstances, I am of the view that it is not necessary to consider whether the unfavourable treatment of non-resident companies regarding ACT hindered direct investment in the United Kingdom by companies resident in other Member States and so restricted the free movement of capital. As the Court pointed out unequivocally in Bachmann, in respect of the former Article 67 (later Article 73b) of the EC Treaty (now Article 56 EC), 'it does not prohibit restrictions which do not relate to the movement of capital but which result indirectly from restrictions on other fundamental freedoms‘. (47) I agree with the view expressed by Advocate General Tesauro in his Opinion in Safir v Skattemyndigheten i Dalarnas Län (48) that, where the free movement of capital and other fundamental freedom provisions of the Treaty are potentially infringed by a particular national rule, the Court should consider the former provision only 'if the measure at issue directly restricts the transfer of capital, rendering it impossible or more difficult, for example by subjecting it to mandatory authorisation ...‘. (49) This view was implicitly endorsed by the Court, which, having found that the impugned tax on life assurance contracts entered into with non-Swedish established providers of such insurance was incompatible with Article 59 of the EC Treaty (now, after amendment, Article 49 EC), held that 'it [wa]s not necessary to determine whether such legislation [wa]s also incompatible with Articles 6, 73b and 73d of the Treaty‘. (50) In my opinion, as a restriction like that at issue in the main proceedings is incompatible with the freedom of establishment, it is unnecessary to consider whether it also constitutes a restriction on direct foreign investment in the United Kingdom.

V - Question 2 and appropriate remedy

41.

    The second question referred by the national court raises two distinct issues, the first of which itself contains two alternatives. It asks first what remedy should be available to taxpayers like the plaintiffs if they are correct in contending that subsidiaries of non-resident parent companies were discriminated against in being denied the advantage involved in making a group income election; are they entitled to a restitutionary claim or only to a compensatory claim for damages for breach of Community law? Secondly, on the assumption that such taxpayers are, in principle, entitled to a remedy, is that remedy affected by the fact their claim extends only to a claim for loss of a cash-flow advantage, namely for interest, in circumstances where the monies of whose use they were deprived were later set off against their proper mainstream CT liabilities and, in particular, does such a matter depend on the applicable national procedural rules?

42.

    The plaintiffs contend principally that their claim amounts to a restitutionary claim. They rely on the Court's well-established case-law that Member States which have levied taxes in contravention of directly effective provisions of Community law must repay them; in their view, this is an adjunct of the directly effective nature of the Community-law rights which have thereby been infringed. (51) Although they recognise that the Court has not yet had to consider a claim based entirely on interest, to allow such a claim would constitute a logical extension of that case-law; it would be futile to recognise the validity of their right to rely directly upon the right of establishment but not then to accord them any remedy. If their claim cannot be classified as a restitutionary claim, they assert, in the alternative, that they have a right to bring a compensatory claim for breach of Community law based on the principles laid down in the Francovich line of case-law, notwithstanding that they are seeking interest for the temporary loss of the use of monies paid by way of ACT. (52) They maintain, in this respect, that the Court recognised in Marshall II (53) that full compensation for the loss and damage sustained as a result of an infringement of Community law could not leave out of account the effect of factors such as the effluxion of time, and that the award of interest may, in some cases, be an essential component of compensation. Theyseek to distinguish Sutton, where, in the context of a claim for interest on arrears of social security benefit, it was held that there was no right under Community law to interest in connection with a claim for restitution, on the basis that the payment of interest was found not to be an essential component of the right in issue. (54)

43.

    The United Kingdom, supported in substance by Finland, submits that the plaintiffs' claim is in substance that its revenue authorities have incurred non-contractual liability to them; the claim has therefore no connection with restitutionary claims for recovery of sums paid but not due or with related claims to interest on such sums. While the right to reparation is founded directly on Community law, it is for the State concerned to make good the consequences of the damage caused in accordance with the relevant national law of liability. In particular, the United Kingdom relies on Fromme v BALM in support of the view that the question whether interest is payable in connection with charges levied contrary to Community law is a matter for national law. (55) It also relies on Sutton; there are similarities between a claim for interest on money paid late, allegedly contrary to Community law, as in that case and a claim to interest on money levied early, again allegedly contrary to Community law, as in the present case, since both actions turn on the consequences of the claimant's being deprived of the use of a sum of money for a certain period. (56)

44.

    The Commission submits that a claim such as that brought by the plaintiffs is restitutionary in nature. The early use of the money constituted a financial benefit obtained unlawfully by the Member State whose value may be quantified. The precise manner in which it is to be quantified is a matter for the national court alone but any national rules applied may not render ineffective the plaintiffs' right under Community law. In the alternative, the Commission submits that the same result should be available through the application of the Francovich and Brasserie du Pêcheur and Factortame case-law.

45.

    The Court has consistently held that Member States must reimburse taxes levied in breach of Community law and that the right to such a reimbursement is a consequence of, and a complement to, the rights conferred on individuals by the directly effective provisions of Community law. (57) In its more recent case-law, the Court has added that Member States are 'required in principle to repay chargeslevied in breach of Community law‘. (58) The notion underlying this principle is that a Member State must not profit and an individual who has been required to pay the unlawful charge must not suffer loss as a result of the imposition of the charge. However, the Court has also recognised that, in the absence of harmonised Community law rules governing actions for recovery of sums unduly paid, 'it is for the domestic legal system of each Member State to designate the courts and tribunals having jurisdiction and to lay down the detailed procedural rules governing actions for safeguarding rights which individuals derive from Community law, provided, first, that such rules are not less favourable than those governing similar domestic actions (principle of equivalence) and, second, that they do not render virtually impossible or excessively difficult the exercise of rights conferred by Community law (principle of effectiveness)‘. (59)

46.

    The United Kingdom submits that among the procedural matters governed by national law is the question of interest. In its view, since in English law no action for interest in respect of the loss of the use of monies which were ultimately later set off against the paying company's CT liability would lie, to deny a remedy in the main proceedings would not infringe the principle of non-discrimination. However, if the national court agrees with the United Kingdom's interpretation of the applicability of the rule upheld by the House of Lords in President of India to the plaintiffs' claim, the effect of applying the principle of national procedural autonomy in respect of interest would be to deny a remedy to taxpayers like the plaintiffs who suffered a cash-flow disadvantage by virtue of being obliged to pay ACT. (60) In my view, such a result would run counter to the principle of effectiveness that lies at the heart of the Court's case-law in respect of the recovery of unduly paid taxes.

47.

    While the Court has not yet had occasion to consider a claim consisting entirely of the loss of the benefit of certain monies, I am satisfied that, in principle, the mere fact that such a loss is the only loss suffered as a result of a temporary violation of Community law is not in itself a reason for refusing to recognise the legitimacy of the claim. It would be anomalous if a claim, valued by the plaintiffs at the oral hearing as possibly amounting to some GBP 8 000 000, could not be made merely because the loss at issue concerned the temporary use of money whose payment was later properly demanded by the Member State concerned. In my view, since it is without question that a Member State may, in principle, be required to pay interest on a capital sum unlawfully levied in contravention of Community law, albeit in accordance with the applicable national legal provisions, it follows, as a logical extension, that where the entire claim at issue concerns thepayment of interest, such interest must in principle be payable. Any other view would, quite simply, entail denying to the taxpayer concerned the opportunity of exercising the rights enjoyed under the directly effective provisions of Community law.

48.

    I would draw support for this view from Marshall II and Brasserie du Pêcheur and Factortame. In Marshall II, while the Court was interpreting a provision of a Community directive, its application of the principle of effectiveness in respect of financial compensation is nevertheless instructive. It held that: 'Where financial compensation is the measure adopted in order to achieve [an effective remedy for wrongful sex discrimination], it must be adequate, in that it must enable the loss and damage actually sustained ... to be made good in full in accordance with the applicable national rules‘. (61) Moreover, in Brasserie du Pêcheur and Factortame, it is also noteworthy that the Court specifically observed that: '[T]otal exclusion of loss of profit as a head of damage for which reparation may be awarded in the case of a breach of Community law cannot be accepted. Especially in the context of economic or commercial litigation, such a total exclusion of loss of profit would be such as to make reparation of damage practically impossible‘. The same principle, to my mind, applies in respect of a claim to interest based on the loss of the use of money.

49.

    I do not think that this view is undermined by the Court's case-law concerning interest. The Court first considered the question of interest in Roquette v Commission. (62) In that case, the applicant claimed interest on certain payments (monetary compensatory amounts) it had been obliged to make to the relevant national authority acting, as agent for the Commission, under a Commission regulation that was subsequently found to be invalid. (63) Advocate General Trabucchi advised the Court that 'the payment of the interest on a capital sum unduly paid is strictly dependent upon the right to repayment of the principal itself ... . An application for interest is subject to the same criteria as those laid down by the case-law of the Court in respect of the claim for repayment of the capital on which the interest is based. An application for interest must, therefore, be made in accordance with the same procedure as that applicable to recovery of the capital sum‘. (64) The Court accepted this advice. It ruled that 'in the absence of provisions of Community law on this point, it is currently for the national authorities, in the case of the reimbursement of dues improperly collected, to settle all ancillary questions relating to such reimbursement, such as the payment ofinterest‘. (65) In Express Dairy Foods v Intervention Board for Agricultural Produce, which also concerned an action for the recovery of monetary compensatory amounts paid pursuant to a Community regulation which had been declared invalid, the Court held that, in the absence of harmonised Community rules, it was for national courts 'to settle all ancillary questions relating to such reimbursement, such as the payment of interest, by applying their domestic rules regarding the rate of interest and the date from which interest must be calculated‘. (66) That case-law concerned capital sums paid pursuant to invalid Community measures but which had been reimbursed to the claimants. The question of whether interest should also be paid was, as the Court made clear, 'ancillary‘. In the present case, as the plaintiffs stressed at the hearing, the claim for interest represents the entirety of the claim. Although in Fromme v BALM, cited by the United Kingdom, the Court also classified as ancillary the nature of claims to interest, that case may also be distinguished from the present case. (67) It concerned a claim by the German authorities for interest to be paid by the undertaking Fromme in respect of certain premiums for the denaturing of cereals which, it was common ground, had been wrongly paid to it by those authorities. The Court held that the Member States' obligation under the relevant Community legislation was 'to take the measures necessary to recover sums lost as a result of irregularities or negligence‘; (68) it was therefore for national law to regulate 'ancillary questions‘ such as that concerning the payment of interest subject to the requirement that the obligations imposed on undertakings having received payments based on Community law 'not be more stringent‘ than those imposed on undertakings having received payments based on national law. (69) In my view, this case-law provides no support for the view defended by the United Kingdom; that, even in a case which only concerns interest, the matter falls to be regulated by national law alone, regardless of whether the relevant rules preclude such purely interest-based claims.

50.

    Nor do I consider Sutton to be of assistance to the United Kingdom. This case concerned a claim for the payment of interest on an amount awarded by way of arrears of a particular social security benefit which had initially been denied for reasons which amounted to sex discrimination contrary to Directive 79/7/EEC. (70) The Court held that the right provided by Article 6 of the Directive for victims of such discrimination was 'to obtain the benefits to which they would have been entitled in the absence of discrimination‘, but that 'the payment of interest onarrears‘ did not constitute 'an essential component of the right as so defined‘. (71) The Court thus followed the advice of Advocate General Léger, who, noting that the benefit claimed had been paid to Ms Sutton, observed that 'the discrimination [had] already been removed in conformity with the rules of national law and the national system can be regarded as having ensured the effectiveness of the principle in practice‘. (72) Thus, in the absence of Community rules, he was satisfied that the question whether the claimant also had a right to interest should be left to national law. The situation is different, in my view, in a case such as the present. Not only is the claim for interest essential, it is the only claim made by the plaintiffs. This arises from the fact that the breach of Community law consists entirely in the temporary removal of sums of money from the resources of certain companies. If Community law were not to require that national legal provisions which would preclude such claims be set aside, the result would be wholly to negate the exercise of a right based on a fundamental principle of Community law. Such a consequence would undermine the effectiveness of the right of establishment by rendering 'impossible in practice‘ the exercise of the right. (73)

51.

    I would reject the submission of the United Kingdom that the plaintiffs' claims in the main proceedings cannot be regarded as restitutionary in nature, solely because, not having sought to exercise a group income election, their action should, at best, be treated as akin to an action for damages against the United Kingdom for the loss they suffered as a result of being left in uncertainty as to their Community-law rights. The ACT payments made by the plaintiffs were made on the basis of national legislation which allowed them no choice. Since such legislation is not compatible, in my view, with Community law, they should, in principle, be entitled to seek restitution for those payments.

52.

    I believe that it is more correct and more logical to treat the plaintiffs' claim as restitutionary rather than as a compensatory claim for damages. ACT was, on the basis of my foregoing analysis, exacted from them in contravention of Community law and, therefore, unlawfully. In the period between payment of ACT and its being taken into account in respect of the CT liability of the subsidiaries, it should have been repaid to the plaintiffs by the United Kingdom. If it had been possible to bring legal proceedings during this period, the plaintiffs would, in my view, have been entitled to interest. It is neither logical nor just to deprive them of this entitlement merely because, in the meantime, the liability of the United Kingdom to repay the principal sum has been discharged. In a practical sense, also, the claim for interest is closer to a restitutionary rather than a compensatory claim. The underlying sums are known and indisputable. All that is necessary isfor the national court to establish an appropriate interest rate for the relevant period.

53.

    If, however, the Court were to disagree with my view that a restitutionary claim to interest for the loss of the use of money should be available in circumstances such as those involved in the main proceedings, it would have to consider the plaintiffs' alternative claim that there should be a right of action for compensatory damages for such a loss. While it is true that in that case-law the damages allegedly suffered by the claimants have normally been unliquidated in amount, I see no reason, in principle, why it should not be possible to claim compensatory damages for a loss that is quantifiable, as in the present case, provided the relevant conditions are satisfied. The United Kingdom relies on the fact that, while the three basic conditions for potential Member State liability are set out in the relevant case-law, 'the national law on liability provides the framework within which the State must make reparation for the consequences of the loss and damage ...‘. (74) This framework, in its view, includes the question of interest. However, the Court has been equally clear that the rules in question must be non-discriminatory and 'must not be so framed as to make it virtually impossible or excessively difficult to obtain reparation‘. (75) This latter principle requires, in my view, that an action seeking compensation in the form of interest in respect of the cash-flow disadvantage occasioned by the loss of the use of money be, in principle, permissible where that is the only loss suffered as a result of a Member State's breach of Community law.

54.

    The three conditions which must be satisfied before a State may be held liable for a breach of Community law have been confirmed consistently by the Court since its initial decision in Francovich. They are that: 'the rule of law infringed must be intended to confer rights on individuals; the breach must be sufficiently serious; and there must be a direct causal link between the breach of the obligation resting on the State and the damage sustained by the injured parties‘. (76) As no questions have been referred in respect of the interpretation of these conditions, and as it is for the national court in each concrete case ultimately to determine whether they are satisfied, I do not propose to consider them in detail. However, as it may be of some assistance to the national court, I shall briefly consider the Commission's submission that those three conditions are met in the present case.

55.

    First, it is beyond doubt that Article 52 of the Treaty creates rights for individuals and that a breach of that provision would therefore satisfy the firstcondition. (77) Secondly, it seems to me manifest that, in principle, there is a direct causal link between the statutory exclusion of the group income election to subsidiaries whose parents were not resident in the United Kingdom and the loss suffered by the plaintiffs. As regards the nature of the breach, I agree with the Commission that the national court may have some doubts as to whether the breach of Community law by the United Kingdom constituted such a sufficiently serious breach of Community law as to justify imposing liability on it. The United Kingdom submits that any breach of Community law was excusable and that any damage caused was involuntary. (78)

56.

    Since there can be no question in the present case of the Community institutions having contributed to the infringement of Community law at issue, the question for the national court would be whether, in the exercise of its legislative powers, the United Kingdom 'manifestly and gravely disregarded the limits on the exercise of its powers‘. (79) The issue is whether the clarity and precision of Article 52 of the EC Treaty were such that the breach may be regarded as sufficiently serious. This has to be viewed in the light of the widespread use of residence as a criterion for direct taxation purposes coupled with the state of development of the relevant case-law (80) at the material time. (81) This will concern the limits which affect the use by Member States of that criterion where it is detrimental to the interests of residents from other Member States. In short, was the refusal to allow the group income election, viewed objectively, 'excusable or inexcusable‘? (82) Although it is clear that measures which discriminate directly on grounds of nationality and which are not justified on the basis of one of the exceptions set out in the Treaty itself would fall to be considered as 'sufficientlyserious‘, (83) the present case concerns indirect discrimination. (84) Indirect discrimination should, in general, be regarded as 'sufficiently serious‘. As the Court declared as long ago as 1986 in respect of direct taxation, '[A]cceptance of the proposition that the Member State in which a company seeks to establish itself may freely apply to it a different treatment solely by reason of the fact that its registered office is situated in another Member State would thus deprive [Articles 52 and 58 of the EC Treaty] of all meaning‘. (85) As regards the possible defence of fiscal cohesion, the Court's recognition in Bachmann that such indirect discrimination may be capable of justification on grounds connected with preserving fiscal cohesion does not, in itself, render 'excusable‘ the breach of Community law. To classify a breach of Article 52 of the Treaty such as that involved in the present case as 'excusable‘, the national court must be satisfied not only that the United Kingdom authorities genuinely believed that refusing to extend the benefit of the group exemption in question to groups whose parent company was non-resident was strictly necessary, but also, viewed objectively in the light of Bachmann and the principle of strict interpretation of exceptions to fundamental Treaty rules like the freedom of establishment, that this belief was reasonable. The national court should also bear in mind the importance of ensuring the effectiveness of rights derived from Community law, particularly fundamental Treaty-based rights.

VI - Questions 3 and 4 concerning tax credits

57.

    In view of the recommendation that I have made in respect of the principal claim brought in the main proceedings, I do not consider it necessary to consider the extremely complex issues raised by the alternative claim in respect of the possible entitlement of the German parent companies, by analogy with the DTC applicable between the United Kingdom and the Netherlands, to a partial tax credit in respect of ACT paid by United Kingdom subsidiaries.

VII - Question 5 and the alleged laches of the plaintiffs

58.

    Since I consider that the Court should rule that the denial of the option to make a group income election to subsidiaries whose parent companies were resident in other Member States constituted unlawful discrimination contrary to Article 52 of the EC Treaty, and that the mere fact that the alleged resulting loss suffered by such subsidiaries concerned the time value of the use of the monies paid by way of ACT does not preclude their claim, it is necessary to consider briefly whether the alleged omission of the plaintiffs, over an extended period of time, to challenge that denial, on the basis of the relevant national statutory appeal mechanism, or, indeed, by way of an earlier direct judicial review application than that actually brought in the main proceedings, may be invoked by the defendant Member State to defeat or reduce the damages sought subsequently by them in a claim based on its incompatibility with Community law. It is true that it has been accepted by the Court that a failure to show 'reasonable diligence‘ in order to avoid loss or damage or to reduce its extent and particularly to avail 'in time of all the legal remedies available‘, may, if similar rules would be applied in purely national-law cases, be taken into account by the national court to reduce, and perhaps in extreme cases, eliminate Member State liability. (86) In my opinion, it should not be permissible, save in the most extreme of cases, for a Member State, whose legislation created a difference in treatment to the detriment of non-residents that admitted of no exceptions and which would have required them, on pain of penalties, to continue paying the tax in question even if its compatibility with Community law had been called into question, to rely upon a taxpayer's failure to use a statutory remedy - one which, moreover, was not, in its own terms, applicable to it - for the purpose of making such a Community-law claim, or to rely upon the direct effect and supremacy of Article 52 of the EC Treaty, as an excuse for seeking to limit a subsequent claim for damages based on the incompatibility of that legislation with Community law.

59.

    This conclusion reflects the important principle that a Member State must not be allowed to profit from its own wrong. It may not, therefore, insist on the application of its rules against taxpayers and then, when those rules are found to be contrary to Community law, deny an obligation to make reparation for the loss it caused on the basis that those rules were not immediately challenged. In my view, in cases such as the present case, where claimants are essentially faced with an unambiguous national legislative rule, on the one hand, and the possible right to oppose the application against them of that rule on the basis of Community law, on the other hand, and where neither the rule in question nor any similar rule of another Member State has previously been considered by this Court, a delay on the part of the claimant in challenging the national rule in question should only be taken into account by the competent national court when considering the possiblelimits affecting the claim before it flowing from national limitation periods or from other comparable rules regarding laches that would also apply to similar claims based purely on national law.

VIII - Conclusion

60.

    In the light of the foregoing, I propose that the Court answer the first, second and fifth questions referred by the High Court of Justice (England and Wales), Chancery Division, as follows:

(1)    It is contrary to Article 52 of the EC Treaty (now, after amendment, Article 43 EC) for the legislation of a Member State to permit a tax advantage such as a group income election (allowing distributions to be paid by a subsidiary to its parent without the subsidiary being required to make advance payments of corporation tax in respect of the profits it has earned in that Member State) only when both the subsidiary and parent are resident in that Member State;

(2)    Where a subsidiary whose parent is not resident in such a Member State has been required to make advance payments of its corporation tax, while in similar circumstances subsidiaries of resident parents were permitted to avoid that requirement by making a group income election, the directly effective right granted by Article 52 of the EC Treaty requires that an effective remedy be available, in principle, to such companies to seek restitution to it of the financial benefit acquired by the Member State authorities concerned as a result of receiving early payment of the taxes of such subsidiaries. The mere fact that any such claim would only be for interest in respect of the financial loss incurred on the loss of the use of the monies paid cannot, in itself, constitute a reason for precluding the taxpayer's right to pursue such a claim. It is for national law to regulate all ancillary matters, such as the limitation period and applicable rates of interest applying to such claims. However, such rules must be no more restrictive than those applicable to similar or comparable claims based purely on national law and must not operate to render virtually impossible the exercise of the right conferred by Community law;

(3)    A Member State may not plead, in answer to a claim for such restitutionary damages, that it should be disallowed or reduced on the grounds that, despite the national legal rules which prevented them from doing so as a matter of national law, the taxpayers concerned ought to have claimed the relevant tax advantage by making use of any statutory remedies available to them and/or relying upon the primacy and direct effect of the relevant provisions of Community law.

1: Original language: English.

2: -    An accounting period, under section 2, is generally a period of 12 months.

3: -    For convenience, I shall hereinafter adopt the expression 'mainstream CT‘, also used by the national court, to describe the normal obligation to pay CT which arises only following this year-long accounting period.

4: -    ACT was abolished by section 31 of the Finance Act 1998 with effect from 6 April 1999.

5: -    See the judgment of Neuberger J. of 2 October 1998 annexed to the order for reference, p. 38. The payment of CT following the end of the relevant tax year will be referred to as 'mainstream‘ CT, so as to distinguish it, for the sake of convenience, from the advance payments of the same tax due by way of ACT.

6: -    See Schedule 13, paragraphs 1 and 3 to the ICTA 1988.

7: -    The national court, whilst observing that the 'arrangements vary‘ under such DTCs, states that 'the general pattern is to grant the tax credit in whole or in part and to make a reduced charge to tax on the aggregate of the amount of the dividend and the amount of the tax credit‘, with the result that '[T]he net effect is to give a partial refund of the ACT‘.

8: -    On the example used by the Hoechst plaintiffs in their written and oral observations (which has not been disputed by the United Kingdom) and which related to a distribution of GBP 43 000 000 paid by Hoechst UK Ltd to its German parent on 16 January 1989, ACT of GBP 14 333 333 was paid. Thus, if Hoechst AG had been a United-Kingdom-resident parent company, it would have benefited from a full tax credit of GBP 14 333 333, while if it had been Dutch, it would have been entitled to a payment of half that tax credit less the 5% charge on the grossed-up amount, viz. to GBP 4 658 333.

9: -    The date of the facts in the main proceedings is not stated in the orders for reference. However, in response to a written question from the Court, it was agreed between the Metallgesellschaft and Others and Hoechst, on the one hand, and the United Kingdom, on the other, that the disputed payments of ACT in Case C-397/98 were made between 16 January 1989 and 26 April 1994, while those involved in Case C-410/98 were made between 16 April 1974 and 13 October 1995.

10: -    In their written observations, Metallgesellschaft and Others observe that the defendants' submission is based on what they describe as a much criticised rule upheld and confirmed (albeit reluctantly) by the House of Lords in President of India v La Pintada Compania Navigacion S.A. [1985] 1 A.C. 104 (hereinafter 'President of India‘). The rule itself dates back at least to London, Chatham and Dover Railway Co. v South Eastern Railway Co. [1893] A.C. 429, where the House of Lords, as a matter of English common law, held that, 'in the absence of any agreement or statutory provisions for the payment of interest, a court had no power to award interest, simple or compound, by way of damages for the detention (i.e. late payment) of a debt‘; per Lord Brandon in President of India ([1985] 1 A.C. 104, p. 115). Lord Brandon (with whose speech the other Law Lords concurred) later expressed approval for the view that the rule now only applies to claims for interest in respect of debts paid late but before any legal proceedings for their recovery have been brought ([1985] 1 A.C. 104, pp. 127 to 129). Metallgesellschaft and Others contest the view that this rule covers a claim such as that involved in the main proceedings. Since that is a matter for the national court alone to decide, I shall assume that the rule, even as now restricted, applies to a claim such as that involved in these proceedings.

11: -    The national court asks about Article 73b of the EC Treaty (now Article 56 EC) only in respect of periods after 1 January 1995. It observes that, before that date, the relevant provision concerning the free movement of capital was Article 67 of the EC Treaty (repealed by the Treaty of Amsterdam), as implemented by relevant directives; see Council Directive 60/921 First Directive for the implementation of Article 67 of the Treaty (OJ, English Special Edition, Series I (1959-1960), p. 49), as amended by Council Directive 63/21/EEC Second Council Directive of 18 December 1962 adding to and amending the First Directive for the implementation of Article 67 of the Treaty (OJ, English Special Edition, Series I (1963-1964), p. 5) for periods before 1 July 1990 and by Council Directive 88/361/EEC of 24 June 1988 for the implementation of Article 67 of the Treaty (OJ 1988 L 178, p. 5) for the period between 1 July 1990 and 31 December 1994. As pointed out in footnote 8 above, it is common ground that periods involved in the main proceedings concern, respectively, 16 January 1989 to 26 April 1994 and 16 April 1974 to 13 October 1995.

12: -    [1996] ECR I-1029 (hereinafter 'Brasserie du Pêcheur and Factortame‘).

13: -    [1997] ECR I-2163 (hereinafter 'Sutton‘).

14: -    Counsel for the plaintiffs stressed at the oral hearing that the alternative claims were 'wholly subsidiary‘ to this principal claim.

15: -    Reference is made in particular to Case 270/83 Commission v France [1986] ECR 273 (hereinafter 'Avoir fiscal‘), paragraph 20.

16: -    Ibid., paragraph 22.

17: -    Case C-264/96 ICI v Colmer (HMIT) [1998] ECR I-4695, paragraphs 25 and 26.

18: -    Reference is made to Avoir fiscal, paragraph 25; Case C-175/88 Biehl [1990] ECR I-1779, paragraph 16; Case C-288/89 Collectieve Antennevoorziening Gouda [1991] ECR I-4007, paragraph 11; Case C-484/93 Svensson and Gustavsson v Ministre du Logement et de l'Urbanisme [1995] ECR I-3955, paragraph 15, and ICI v Colmer (HMIT), ibid., paragraph 28.

19: -    See Case C-204/90 [1992] ECR I-249 (hereinafter 'Bachmann‘) and Case C-300/90 [1992] ECR I-305.

20: -    In Ireland, under section 46 of the Finance Act 1983, a group income election was permitted once the Irish subsidiary was at least 75% owned by the foreign parent and provided there was a DTC in force between Ireland and the State of residence of the parent company.

21: -    Once ACT has been charged to match the exemption from tax which is accorded to dividends in the hands of a corporate shareholder, the United Kingdom observes that any further charge to ACT made when the corporate shareholder distributes to its own shareholders the dividends it has received would give rise to another type of double taxation. This is why a corporate shareholder resident in the United Kingdom and receiving a dividend from another United-Kingdom-resident company is entitled to a tax credit.

22: -    Paragraph 41 of its written observations (emphasis in original).

23: -    Reference is made to paragraph 23 in particular.

24: -     See Case C-35/98 Staatssecretaris van Financiën v Verkooijen [2000] ECR I-0000 (hereinafter 'Verkooijen‘), paragraph 48 and the case-law cited there.

25: -    See Verkooijen, ibid., paragraph 32, and the case-law cited there.

26: -    Op. cit., paragraphs 26 and 28.

27: -    See paragraph 18 above.

28: -    Bachmann, loc. cit., footnote 18 above, paragraph 23, Commission v Belgium, paragraph 16.

29: -    Ibid., paragraphs 24 and 17 respectively.

30: -    Paragraphs 25 and 18 respectively

31: -    Bachmann, loc. cit., paragraph 27, Commission v Belgium, mutatis mutandis, paragraph 20. For convenience, all references concerning these cases will hereinafter be to Bachmann only.

32: -    Case C-80/94 [1995] ECR I-2493 (hereinafter 'Wielockx‘).

33: -    Wielockx, paragraph 23.

34: -    Paragraph 24.

35: -    Paragraph 25.

36: -    Case C-107/94 [1996] ECR I-3089 (hereinafter 'Asscher‘).

37: -    Paragraph 59. The Court noted that the non-application of the Netherlands social security system to non-residents like Mr Asscher was probably justified under Council Regulation (EEC) No 1408/71 of 14 June 1971 on the application of social security schemes to employed persons and their families moving within the Community (OJ, English Special Edition, Series I (1971) (II), p. 416).

38: -    Op. cit., footnote 16 above, paragraph 29.

39: -    Case C-294/97 [1999] ECR I-7447 (hereinafter 'Eurowings‘).

40: -    Paragraph 42.

41: -    Op. cit., footnote 23 above.

42: -    Paragraph 58.

43: -    Case C-251/98 Baars [2000] ECR I-0000.

44: -    Ibid., paragraph 40.

45: -    Much of the academic criticism of Bachmann has centred on the fact that the Court ignored Belgium's DTC with Germany, which might well have permitted Belgium to tax the capital payments made to migrant workers like Mr Bachmann, at least if they remained resident in Belgium, and the fact that there were probably less restrictive means by which the taxation of such payments could have been secured, such as by imposing obligations on insurers who wished to have their policies qualify for deduction. See, inter alia, Knobbe-Keuk, 'Restrictions on the Fundamental Freedoms Enshrined in the EC Treaty by Discriminatory Tax Provisions‘ (1994) EC Tax Review 74; Hatzopoulas, 'Fiscalité directe des États membres et ”libertés personnelles” reconnues par le traité CE‘ (1995) Rev Marché Unique Eur 121, pp. 143 to 152; Quaghebeur, 'A Bridge over Muddled Waters - Coherence in the Case Law of the Court of Justice of the European Communities relating to Discrimination against Non-resident Taxpayers‘ (1995/1996) EC Tax Review 109; Farmer, 'EC Law and Direct Taxation - Some Thoughts on Recent Issues‘ (1995/1996) EC Tax Review 101; Wattel, 'The EC Court's Attempts to Reconcile the Treaty Freedoms with International Tax Law‘ (1996) 22 CML Rev 223; Vanistendael, 'The Consequences of Schumacher and Wielockx: Two Steps Forward in the Tax Procession of Echternach‘ (1996) 33 CML Rev 255; Lang, 'The Binding Effect of the EC Fundamental Freedoms on Tax Treaties‘ and Schuch, 'Will EC Law Transform Tax Treaties into Most-Favoured Nation Clauses?‘, both in Gassner, Lang and Lechner, eds., Tax Treaties and EC Law (London, The Hague) (1997).

46: -    As corporate shareholders are not liable to pay CT on dividends, no element of tax otherwise due by the German parents in the present case may be imputed to the ACT payments made by their United Kingdom subsidiaries.

47: -    Paragraph 34. The Court, thus, accepted the recommendation of Advocate General Mischo that the transfer of capital necessary to pay contributions to non-Belgian insurers was not subject to any restraint and that any causal nexus between the non-deductibility of the premiums at issue and the free movement of capital was too remote to constitute a restriction on the latter.

48: -    Case C-118/96 [1998] ECR I-1897.

49: -    Paragraph 17 (emphasis in original).

50: -    Paragraph 35.

51: -    Case 199/82 Amministrazione delle Finanze dello Stato v San Giorgio [1983] ECR 3595 (hereinafter 'San Giorgio‘), paragraph 12, and Joined Cases C-192/95 to 218/95 Comateb and Others v Directeur Général de Douanes et Droits Indirects [1997] ECR I-165 (hereinafter 'Comateb and Others‘), paragraph 20 are cited by Metallgesellschaft and Others, while Hoechst also cites Case C-242/95 GT-Link v DSB [1997] ECR I-4449, paragraphs 58 to 60, and Joined Cases C-10/97 to C-22/97 Ministero de Finanze v IN.CO.GE. '90 and Others [1998] ECR I-6307, paragraphs 23, 24 and 29.

52: -    Reference is made to Joined Cases C-6/90 and C-9/90 Francovich and Others [1991] ECR I-5357 (hereinafter 'Francovich‘) and Brasserie du Pêcheur and Factortame, loc. cit., footnote 11 above.

53: -    Case C-271/91 Marshall v Southampton and South West Hampshire Area Health Authority [1993] ECR I-4367 (hereinafter 'Marshall II‘).

54: -    Sutton, loc. cit., footnote 12 above.

55: -    Case 54/81 [1982] ECR 1449.

56: -    It refers, in particular, to paragraph 31 of the judgment in Sutton.

57: -    See, inter alia, San Giorgio, loc. cit., footnote 50 above.

58: -    Comateb and Others, loc. cit., footnote 50 above, paragraph 20, Case C-343/96 Dilexport v Amministrazione delle Finanze dello Stato [1999] ECR I-579 (hereinafter 'Dilexport‘), paragraph 23.

59: -    Dilexport, ibid., paragraph 25.

60: -    See footnote 9 above.

61: -    Loc. cit., footnote 52 above, paragraph 26.

62: -    Case 26/74 [1976] ECR 677.

63: -    The applicant had obtained, in parallel proceedings brought before the national courts, an order for the reimbursement of the capital sum of the amounts actually paid; see paragraphs 3 to 7 of the judgment for an account of those proceedings.

64: -    Ibid., [1976] ECR 677, p. 691.

65: -    Paragraph 12.

66: -    Case 130/79 [1980] ECR 1887, paragraph 17.

67: -    Loc. cit., footnote 54 above.

68: -    Paragraph 5.

69: -    Paragraph 6.

70: -    Council Directive 79/7/EEC of 19 December 1978 on the progressive implementation of the principle of equal treatment for men and women in matters of social security, OJ 1979 L 6, p. 24.

71: -    Paragraph 25.

72: -    Paragraph 62 (emphasis in original).

73: -    See paragraph 172 of the Opinion of Advocate General Jacobs in Case C-90/94 Haahr Petroleum v Åbenrå Havn and Others [1997] ECR I-4085.

74: -    See, for example, Francovich, paragraph 41, Brasserie du Pêcheur and Factortame, paragraph 83 and Sutton, paragraph 33.

75: -    See, for example, Francovich, paragraph 43, Brasserie du Pêcheur and Factortame, paragraph 83 and Sutton, paragraph 33.

76: -    See, for example, Sutton, paragraph 32.

77: -    See Brasserie du Pêcheur and Factortame, paragraph 54.

78: -    It relies particularly upon Brasserie du Pêcheur and Factortame, paragraph 56.

79: -    Case C-392/93 The Queen v H.M. Treasury, ex parte British Telecommunications [1996] ECR I-1631, paragraph 42. This test was confirmed recently in Case C-140/97 Rechberger and Others [1999] ECR I-3499, paragraph 50. See also in this respect Brasserie du Pêcheur and Factortame, paragraph 56.

80: -    In Brasserie du Pêcheur and Factortame, paragraph 57, the Court held that a breach of Community law would be 'sufficiently serious‘ if it continued 'despite a judgment finding the infringement in question to be established, or a preliminary ruling or settled case-law of the Court on the matter from which it [wa]s clear that the conduct in question constitute[d] an infringement‘.

81: -    For Metallgesellschaft and Others, it comprises developments up to 26 April 1994, while for Hoechst it extends to 13 October 1995; see footnote 8 above. It is, therefore, only in respect of Hoechst's claim that the Court's post-Bachmann case-law, beginning with Wielockx, loc. cit., footnote 31 above, which was decided on 1 August 1995, would be relevant.

82: -    Brasserie du Pêcheur and Factortame, paragraph 56.

83: -    Ibid., paragraph 62.

84: -    Since Avoir fiscal in 1986, the Court has consistently held that 'the seat of a company, in the sense of its registered office, central administration or principal place of business, serves the function of nationality for natural persons and that to treat non-resident corporate taxpayers less favourably because of their foreign places of residence may, in the absence of objective justification, constitute indirect discrimination on grounds of nationality‘; op. cit., footnote 14 above, paragraph 18 of the judgment. This view has been confirmed, inter alia, in Case C-330/91 The Queen v IRC, ex parte Commerzbank, paragraph 15 and ICI v Colmer (HMIT), paragraph 23.

85: -    See Avoir fiscal, paragraph 18. See also Commerzbank, ibid., paragraphs 18 to 19 and ICI v Colmer (HMIT), paragraphs 23 to 24.

86: -    Brasserie du Pêcheur and Factortame, paragraph 84. See also paragraph 104, in particular, of the Opinion of Advocate General Tesauro in those cases.