Test Claimants in the FII Group Litigation v Commissioners of Inland Revenue, The Commissioners for her Majesty s Revenue & Customs

Transcription

1 Opinion of Advocate General Jääskinen, 19 July Case C-35/11 Test Claimants in the FII Group Litigation v Commissioners of Inland Revenue, The Commissioners for her Majesty s Revenue & Customs Table of contents I Introduction II Background to the present preliminary reference III On the context of the order for reference IV Question 1 a. Introduction b. A duty to give a credit corresponding to the statutory rate of the source State c. Both effective and nominal rate d. Statutory rate e. The existence of a restriction and its justification f. Conclusion V Question 2 VI Question 3 A The question and observations received VII Question 4 VIII Question 5 IX Conclusion I Introduction 1. On 12 December 2006 the Court gave a preliminary ruling in Case C-446/04 Test Claimants in the FII Group Litigation ( the first FII judgment ), 2 in which it answered, among others, the question 3 whether United Kingdom corporation tax laws that treated share dividends differently by reference to whether they were issued by United Kingdom-resident companies or by non-resident companies were compatible with certain provisions of EU law. 2. The national proceedings are still pending before the Chancery Division of the High Court of Justice of England and Wales ( the High Court ), which decided to stay the proceedings again and refer five further questions to the Court for preliminary ruling. Some of these seek elaboration of the answers provided by the Court in the first FII judgment while others seek guidance on fresh issues that have arisen in the domestic proceedings. 3. The mitigation of corporate economic double taxation (that is, taxation of the same income twice in the hands of two different taxpayers) is an area of major economic importance to cross-border activity and of direct concern to a vast number of companies in the EU. This reference for a preliminary ruling illustrates problems arising from the interaction of internal market law with national and international tax law, an issue not without controversy. 4 II Background to the present preliminary reference 4. The essence of the dispute is as follows. The principal aim and effect of the United Kingdom legislation in force during the pertinent period ( ), 5 was to provide a measure of relief for shareholders from economic double taxation. Two different systems were applied: 1. Original language: French. 2. [2006] ECR I As this case builds upon the first FII case it is assumed that the reader has already read both the opinion of Advocate General Geelhoed and the judgment. 3. The order for reference covered two subject areas, and comprised a total of nine questions. The first five questions concerned the substantive British tax laws at issue. The last four questions concerned remedies and temporal issues. 4. A concise overview of direct taxation based restrictions to the fundamental freedoms is given e.g. by Metzler, V., The relevance of the Fundamental Freedoms for Direct Taxation, in Lang, M. et al. (eds), Introduction to European Tax Law on Direct Taxation, Linde, Vienna, 2008, p. 35. On the debate in general, see e.g. Kingston, S., A light in the darkness: recent developments in the ECJ s direct tax jurisprudence, Common Market Law Review, 2007, pp , Graetz, M. Warren, A., Dividend Taxation in Europe: When the ECJ makes tax policy, Common Market Law Review, 2007, pp , and Snell, J., Non-discriminatory Tax Obstacles in Community Law, International and Comparative Law Quarterly 2007, p. 339.

2 nationally-sourced dividends were subject to an exemption method whereas foreign-sourced dividends were subject to an imputation (or credit) method. Under the exemption method, resident companies which received dividends from other resident companies were simply exempt from paying tax on the dividends on the assumption that corporation tax had already been levied against the company that issued them. However, under the imputation method, dividends issued by non-resident companies, that is foreign-sourced dividends, attracted only a tax credit in the hands of the United Kingdom company receiving them The proceedings before the national court are based on litigation between Test Claimants in the Franked Investment Income (FII) Group Litigation ( the Test Claimants ) and Commissioners of Inland Revenue and the Commissioners for Her Majesty s Revenue and Customs (both jointly referred to as HMRC ). 6. In answering the first of the nine questions referred in the first FII judgment, the Court of Justice held that Articles 49 and 63 TFEU 7 did not preclude legislation of a Member State which on the one hand exempted from corporation tax dividends which a resident company received from another resident company and on the other hand imposed corporation tax on dividends which a resident company received from a non-resident company (and in which the resident company held at least 10% of the voting rights) while at the same time granting a tax credit in the latter case for the tax actually paid by the company making the distribution in the Member State in which it was resident. 8 However, this was subject to the proviso that: the rate of tax applied to foreign sourced dividends is no higher than the tax rate applied to nationally sourced dividends and that the tax credit is at least equal to the amount paid in the Member State of the company making the distribution, up to the limit of the amount of the tax charged in the Member State of the company making the distribution, up to the limit of the amount of the tax charged in the Member State of the company receiving the distribution This statement forms the essence of the order for reference in the present case. This is so because the Court added, at paragraph 56 of its judgment, that: it is for the national court to determine whether the tax rates are indeed the same and whether different levels of taxation occur only in certain cases by reason of a change to the tax base as a result of certain exceptional reliefs. 8. This approach, accepting in principle the simultaneous application of two different systems to domestic and foreign dividend income, has been applied to date in cases such as Haribo Lakritzen Hans Riegel and Österreichische Salinen and Accor, 10 both of which related to the issue of mitigation of economic double taxation of foreign dividend income in the context of corporate taxation. 9. After analysis of the first FII judgment, the High Court decided to stay the national proceedings again and to refer, by order of 15 December 2010, the following five questions to the Court for a preliminary ruling: Do the references to tax rates and different levels of taxation at paragraph 56 of the [first FII judgment]: a. refer solely to statutory or nominal rates of tax; or b. refer to the effective rates of tax paid as well as the statutory or nominal rates of tax; or c. do the phrases referred to have some different meaning and, if so, what? 2. Does it make any difference to the Court s answer to Questions 2 and 4 of the [first FII reference] if: a. foreign corporation tax is not (or not wholly) paid by the non-resident company paying the dividend to the resident company, but that dividend is paid from profits comprising dividends paid by its direct or indirect subsidiary resident in a Member State and which were paid out of profits on which tax has been paid in that State; and/or b. advance corporation tax ( ACT ) is not paid by the resident company which receives the dividend from a non-resident company, but is paid by its direct or indirect resident parent company upon the further distribution of the profits of the recipient company that directly or indirectly comprise the dividend? 3. In the circumstances described in Question 2(b) above, does the company paying the ACT have a claim for the repayment of the tax unduly levied (San Giorgio 12 ) or only a claim for damages (Brasserie du Pêcheur and Factortame 13 )? 5. The original system of advance corporation tax ( ACT ) was operated since It was amended as of 1 July 1994 when the treatment for foreign income dividend ( FID ) was introduced. For a more detailed description of the national legislation, and the national proceedings, see first FII judgment, paragraphs 6 to 30, and opinion of Advocate General Geelhoed, points 2 to Credit was given for any withholding tax on the dividend and, under certain conditions, for the underlying tax paid by the non-resident companies on their profits in their country of residence. 7. For the sake of clarity, the FEU Treaty is referred to throughout. 8. Paragraph 73 of the first FII judgment. 9. Paragraph 73 of the first FII judgment; see also paragraph 57 of the first FII judgment. This passage seems to contain a lapsus linguae: the judgment speaks of the tax rate applied to nationally-sourced dividends. However, the High Court notes in its judgment that nationally-sourced dividends are exempt from tax. The existence of this error in the first FII judgment excludes in my view the literal interpretation of that judgment. 10. Joined Cases C-436/08 and C-437/08 Haribo Lakritzen Hans Riegel and Österreichische Salinen [2011] ECR I-0000, paragraph 86, and Case C-310/09 Accor [2011] ECR I-0000, paragraph For the sake of clarity I wish to add that the original decision by the High Court of 27 November 2008 to make a second reference to the Court only contained Questions 2, 3 and 5 (see [2008] EWHC 2893 (Ch)). That decision was in part appealed, and the Court of Appeal, by its judgment of 23 February 2010 (see [2010] EWCA Civ 103), added Question 1, and the Supreme Court, on further appeal, by its Order of 8 November 2010, added Question 4. The questions included in the request for preliminary ruling were set out in full in the aforementioned order of 15 December 2010 by the High Court. In this case the questions submitted have gone through a thorough national procedure and they reflect a careful and detailed consideration as to the issues on which the national court seeks assistance from the Court. 12. Case 199/82 San Giorgio [1983] ECR Joined Cases C-46/93 and C-48/93 Brasserie du Pêcheur and Factortame [1996] ECR I-1029.

3 4. Where the national legislation in question does not apply exclusively to situations in which the parent company exercises decisive influence over the dividend paying company, can a resident company rely upon Article 63 TFEU in respect of dividends received from a subsidiary over which it exercises decisive influence and which is resident in a third country? 5. Does the Court s answer to Question 3 of the [first FII reference] also apply where the non-resident subsidiaries to which no surrender could be made are not subject to tax in the Member State of the parent company? 10. Written observations have been presented on behalf of the Test Claimants, the United Kingdom Government, the German Government (on Questions 1 and 4), the French Government (on Questions 1 and 4), Ireland (on Question 1), the Netherlands Government (on Question 4), as well as by the European Commission. A hearing was held on 7 February It was attended by the Test Claimants, the United Kingdom Government, the German Government, Ireland and the European Commission. III On the context of the order for reference 11. The following diagram aims to clarify the corporate structures underlying the preliminary questions: United Kingdom A B C D F E G Other Member States of the European Union Non-member countries (third countries) 12. The diagram depicts three groups of countries; the United Kingdom, other Member States of the European Union and non-member countries (third countries). It contains seven companies, in hierarchical order, marked with characters A to F. Company A is the ultimate parent company resident in the United Kingdom. Companies B and C are subsidiaries of A and resident in the United Kingdom. Companies D and E are subsidiaries of C resident in another EU Member State. Companies F and G are subsidiaries of C resident in a non-member country Question 1 concerns the comparison between the concepts of tax rates and different levels of taxation. It relates to comparison between taxation of dividends sourced in the United Kingdom (companies B and/or C) and those sourced in other Member States of the European Union (company D) and non-member countries (company F). 14. Questions 2(a) and 2(b) of the present preliminary ruling are a follow-up on the replies of the Court to Questions 2 and 4 of the first FII judgment. They concern a situation in which company D resident in another Member State of the European Union pays dividend to its parent company C resident in the United Kingdom. 15. On this point, the first FII judgment was based on two assumptions. First, that company D had paid company tax in its Member State of residence. Second, that company C had paid corporation tax in the United Kingdom in form of ACT. 16. Against this background, Question 2(a) seeks to clarify whether it would make a difference to the answers of the Court if company D paying dividend had not itself paid (any/full) company tax in its Member State of residence, but that the tax was paid by a lower level company E, in that Member State or in another Member State. 17. Question 2(b) asks whether it would make a difference if company C did not pay the United Kingdom company tax in the form of ACT itself, but that tax was paid higher up in the corporate chain (by company B or A) as a result of group income election rules. 18. Question 3 concerns the question whether ACT, paid by A or B in the United Kingdom, can be recovered by means of an action for recovery of unlawful tax or by a damages action for breach of EU law. 19. Question 4 relates to dividends paid from third countries into a United Kingdom corporate structure. In essence, it addresses the applicability of Article 63 TFEU to a situation in which company F, resident in a third country, pays dividends to C, resident in the United Kingdom, and where C can exercise decisive influence over F. 20. Finally, Question 5 relates to companies D and F and to whether ACT paid by companies A, B or C in the United Kingdom could be surrendered in their favour in a situation where companies D and F are not subject to corporation tax in the United Kingdom. 14. In certain international tax situations companies C, D and F can act as so-called water s edge companies that are used to channel the distributions from/to other companies in the group.

4 IV Question By Question 1 the High Court seeks clarification of the meaning of tax rates and different levels of taxation mentioned at paragraph 56 of the first FII judgment. 22. In the litigation leading to the first FII judgment, the Court was asked whether it was contrary to Articles 49 and 63 TFEU for a Member State to apply rules which exempted from corporation tax dividends received by a resident company from other resident companies while providing for the taxation of dividends received from companies resident in other Member States (after giving double taxation relief for any withholding tax on the dividend and, under certain conditions, for the underlying tax paid by the non-resident companies on their profits in their country of residence). 23. When the case returned to the High Court, the parties were divided on the correct interpretation of the first FII judgment, and in particular paragraphs 54 to 56 thereof. 24. The Test Claimants argued that the task of the national court referred to at paragraph 56 of the first FII judgment 15 was to examine whether the exemption of domestic distributed profits could result in a lower effective tax burden than that following the granting of indirect tax credit on foreign-sourced distributed profits. They presented to the High Court expert evidence according to which the effective level of taxation on the profits of resident companies was lower than the statutory rate in the majority of cases. Therefore this did not occur only in highly exceptional circumstances, as had been argued by the United Kingdom Government in the first FII case. 16 This finding was not as such contested by HMRC. Rather, their position was that the national court was merely required to verify that different statutory rates of tax occurred only in highly exceptional circumstances and not to examine the effective levels of taxation. 25. The High Court agreed with the Test Claimants interpretation of the judgment. On appeal, the Court of Appeal was divided on this issue. Two of the judges favoured HMRC s view, whereas the third judge agreed with the High Court s finding. In view of the disagreement, the Court of Appeal decided to refer the matter back to this Court for a preliminary ruling on the interpretation to be given to the relevant paragraphs of the first FII judgment. This decision was appealed to the Supreme Court which referred the case back to the High Court for the purposes of making a preliminary reference. 26. In their observations before the Court, the German Government, Ireland and the United Kingdom Government propose that the references to tax rates and different levels of taxation at paragraph 56 of the first FII judgment refer solely to statutory or nominal rates of tax. The Test Claimants suggest that these expressions refer to the effective rates of tax paid as well as statutory or nominal rates of tax. 17 The Commission suggests that the Member State must calculate the tax credit on the basis of the nominal rate of tax applicable in the source State. a. Introduction 27. It is useful to briefly recall the different approaches adopted by the Advocate General and the Court in the first FII judgment. 28. Advocate General Geelhoed considered in the first FII case that application of two different systems for relieving economic double taxation of dividends could in principle comply with the Treaty. However, after detailed analysis, he concluded that applying two systems, one to nationally-sourced dividends and another to foreign-sourced dividends, was inevitably discriminatory and incompatible with the Treaty. 29. The Advocate General argued that this was so because the application of a credit system by the [United Kingdom] for the relief of double economic taxation on foreign-source dividends can in certain cases have less favourable effects than the pure exemption system applied to nationally-source dividends. While, under an exemption system, the benefits of underlying corporation tax exemptions and allowances may be passed on to the parent company receiving the dividends, under a credit system these benefits cannot be passed on as the tax borne by the dividends is topped up to the standard [United Kingdom] corporation rate. In such cases, the effect of this could be seen as the application by the [United Kingdom] of a different (lower) tax rate to domestic-source dividends than to foreign-source dividends At this juncture it is necessary to make two observations. Firstly, the aim of applying an imputation system to foreign-sourced dividends is, of course, to achieve the effect described by Advocate General Geelhoed, in other words, to eliminate the effect in the residence State taxation of a lower effective tax rate in the source State. This is achieved by taxing the difference between the effective rate in the source State and the rate 19 applicable to the foreign-sourced dividends in the residence State in the latter State. 31. Secondly, the opinion seems to suggest that Advocate General Geelhoed did not disagree with the United Kingdom and the Commission to the extent that they claimed that both systems lead to economic double taxation being relieved Paragraph 56 of the first FII judgment is cited above in point 7 of this opinion. 16. Paragraph This is in essence also the position of the French Government, which, however, draws different conclusions of that position, see footnote 36 below. 18. Point 50 of the Opinion. 19. The High Court notes in its judgment of 27 November 2008 (cited above in footnote 11, point 51 of the judgment) that the United Kingdom parent company will not necessarily pay corporation tax at the statutory rate on its foreign-sourced dividends, because it may well have reliefs of its own available to it. In other words, the effective rate on foreign-sourced dividends may also be lower than the statutory rate and the aggregate tax burden will not always be topped up to the standard United Kingdom rate as Advocate General Geelhoed had claimed in his opinion (point 50 of the opinion, cited above in footnote 2). 20. See point 48 read together with point 51 of the Opinion.

5 32. The Court, however, concluded in the first FII judgment that the application of two different systems to relieve economic double taxation of nationally-sourced and of foreign-sourced dividends could be compatible with the Treaty, provided that certain conditions were fulfilled. 21 The Court is now requested to clarify its ruling. 33. In my view, the proposal of the Advocate General would have been better in line with the case-law of the Court regarding direct taxation restrictions on fundamental freedoms. Clearly the simultaneous application of two different methods to relieve economic double taxation of nationally-sourced dividends and foreign-sourced dividends inevitably deviates from capital export neutrality. 22 The two methods aim at different results with respect to the possibility of passing on to the shareholder the underlying corporation tax exemptions and advantages. The imputation method seeks to exclude the passing on whereas the exemption method aims at this, provided that no supplementary taxation exists to top up the taxation of the distributed dividends to statutory rates Moreover, as this lack of neutrality is created by the rules applicable in the Member State of residence of the shareholder, it is not as such a direct result of the differences in the tax legislations of different Member States However, having said that, the approach chosen by the Court in the first FII judgment has been applied in subsequent cases. 25 That being so, and for reasons of legal certainty, I am not suggesting that the Court should depart from this line of case-law, which has certainly been relied upon by courts, undertakings and tax administrations in the Member States. Nonetheless, this line of case-law cannot be upheld unless the Court accepts that the application of the above described mixed asymmetric system leads to less favourable treatment of foreignsourced dividends. This follows from the difference concerning the possibility to pass on to the shareholder the tax allowances applicable to the underlying corporate profits However, should the Court decide to reconsider this relatively recently established case-law, the most appropriate solution would be to adopt the approach proposed by Advocate General Geelhoed in the first FII reference. 37. I will now proceed to a discussion of the three alternative interpretations set out by the High Court in Question 1 of the preliminary reference. b. A duty to give a credit corresponding to the statutory rate of the source State 38. The High Court asks, as the third alternative interpretation, whether the references to tax rates and different levels of taxation at paragraph 56 of the first FII judgment have a meaning that differs from statutory tax rates or effective tax rates and, if so, what. 39. Only the Commission s proposal adopts this approach. The Commission suggests that the answer to Question 1 should be that the Member State must ensure that the tax credit is equivalent to the relief granted in respect of [nationally-sourced] dividends, by calculating the credit on the basis of the nominal rate of tax applicable in the State from which the dividends originate. 40. According to the Commission, this proposal seeks to ensure formal equality of treatment and ease of application, while achieving a fair result. On the one hand, this is achieved without systematic favouring of foreign-sourced dividends originating from source States with low tax rates. On the other hand, there would be no need for systematic re-calculation of the tax position of a foreign company making the dividend distribution, simulating the tax it would have paid were it resident in the United Kingdom. This method would, according to the Commission, correspond more faithfully to exemption of nationally-sourced dividends. 41. Despite the simplicity and elegance of the Commission s proposal, I do not believe the Court should adopt it. I have four reasons for this. 42. Firstly, the Commission s proposal is not connected to the first FII judgment, nor to the arguments of the parties in the context of the first FII reference, unless it is understood as an explanation of the words in the same way in the answer to the first preliminary question in the first FII judgment. The Commission s solution has emerged as an independent alternative, separate from the arguments presented in the main proceedings. 43. Secondly, adopting the Commission s solution would mean that there was only one EU law compatible alternative of applying the imputation method for mitigating economic double taxation of foreign-sourced dividends under a fiscal system that exempts nationallysourced dividends. This would amount to judicial harmonisation of fiscal provisions concerning an issue falling within the national competence, despite the fact that the method, as admitted by the Commission, does not ensure substantive equal treatment in all cases but is proposed because of its practicability. Weighing the degree of equality desired and administrative practicability is, by its nature, a legislative and not a judicial task Paragraph 56 of the first FII judgment, cited above in point 7 of this opinion. 22. Capital export neutrality can be characterised as the situation where the investors are subject to the same level of taxes on capital income regardless of the country in which the income is earned. Conversely, capital import neutrality refers to the situation where investments within a country are subject to the same level of taxes regardless of whether they are made by domestic or foreign investor. The credit method illustrates the former principle, while the exemption method illustrates the latter. See Larking, B., IBFD International Tax Glossary. 5th ed., Amsterdam, IBFD This is my reading of the principle underpinning e.g. Case C-35/98 Verkooijen [2000] ECR I-4071; Case C-168/01 Bosal [2003] ECR I-9409; Case C-315/02 Lenz [2004] ECR I-7063; Case C-319/02 Manninen [2004] ECR I-7477; Case C-446/03 Marks & Spencer [2005] ECR I-10837; and Case C-196/04 Cadbury Schweppes and Cadbury Schweppes Overseas [2006] ECR I See opinions of Advocate General Geelhoed in Case C-374/04 Test Claimants in Class IV of the ACT Group Litigation [2006] ECR I-11673, points 31 to 54, and in first FII case, point 38; Case C-513/04 Kerckhaert and Morres [2006] ECR I-10967, paragraphs 20 and 22, and opinion of Advocate General Geelhoed, point See above point 8 and footnote See below under heading (e) (point 58 et seq.). 27. It should be added that this solution has not been included in Council Directive 90/435/EEC of 23 July 1990 on the common system of taxation applicable in the case of parent companies and subsidiaries of different Member States (OJ 1990 L 225, p. 6).

6 44. Thirdly, the solution is not capital export neutral if the Member State of residence of the dividend recipient has effective tax rates close to the statutory rate, and the source State combines high statutory rates with low effective rates. Put differently, the recipient s Member State would be obliged to grant tax credit corresponding to the difference between the effective and statutory tax rate on underlying profits in the source State, in other words, to grant tax credit for an unpaid foreign tax. 28 Economically the solution comes close to an obligation to give socalled tax sparing credit, used in double-taxation treaties between industrialised and developing countries, as it also seeks to pass on the reliefs and tax incentives of the source State to the taxation in the residence country Fourthly, the solution is, in my mind, intellectually incoherent. As the Commission itself observes, the idea of applying the imputation method to foreign-sourced dividends while exempting nationally-sourced dividends may be particularly useful for taking into account differences between levels of taxation in the source State and in the residence State. According to the Commission, it is legitimate for a Member State to aim at ensuring that income, including dividends, received by its resident companies is taxed at the rate laid down in its own legislation. 46. However, if this view were accepted, it would be inconsistent to require, first, that the recipient s Member State does not tax the foreign-sourced dividends with regard to the difference between the effective and statutory tax rate in the source State, while allowing, second, that the residence State taxes the difference between (the lower) statutory rate in the source State and the statutory rate of the residence State. It does not seem logical that only the effect of the allowances and exemptions applicable in the source State (which create the difference between the effective rate and the statutory rate in that State) should be passed on to the residence State taxation of the shareholder, but not the effect of lower foreign statutory rates in a situation where the residence State exempts nationally-sourced dividends. c. Both effective and nominal rate 47. The Test Claimants support an interpretation of the first FII judgment that both statutory and effective rates should be taken into account in determining whether there is a difference in the level of taxation of foreign-sourced dividends and corporate taxation of profits underlying domestic dividends, the latter being as such exempted. Adoption of this interpretation would entail that if such a difference exists (or is found to exist by the national court) more frequently than only in exceptional situations, the foreign-sourced dividends would be discriminated against resulting in a restriction of the freedom of establishment. 48. The concept of statutory or nominal tax rates is sufficiently clear for the purposes of this preliminary reference. It refers to the percentage of tax that a certain amount of taxable income must bear according to the applicable legal rules. In the context of the present case there are two statutory rates, namely the United Kingdom corporation tax rate applicable to foreign-sourced dividends, and the United Kingdom corporation tax rate applied in the taxation of underlying profits of distributing United Kingdom companies. As nationally-sourced dividends are exempted, no statutory rate applies to them. 49. The concept of effective rate is much more ambiguous. 30 It may refer to the actual level of taxation of a given income or tax subject, but it can also refer to a statistical measure developed for assessing the tax burdens charging certain activity The concept of effective tax rate used by the High Court, and supported by the Test Claimants, refers to the proportional amount of tax actually paid on accounting profits. It seems to be undisputed between the parties, and accepted by the High Court that this effective rate may be lower than the statutory rate because of reliefs and allowances that reduce the tax burden of a subsidiary resident in the United Kingdom. It is also accepted that this occurs frequently and not only in highly exceptional circumstances. 51. Applying this notion of effective tax rate would when comparing the tax burdens on foreign-sourced dividends and nationallysourced dividends lead to considerable theoretical and practical problems. The effective tax rate is different for each company and for each financial year, depending on the allowances and exemptions affecting the definition of the tax base (such as carry-on losses or group relief). 52. In this regard, Ireland points out rightly that for tax purposes, profits refer to accounting profits as adjusted by any requirement imposed by applicable laws. It is therefore highly unlikely that, in a particular case, the amount of accounting profits would coincide with the amount of the profits for tax purposes. Ireland claims that in delivering the first FII judgment the Court was aware, on the basis of the submissions made to it, that where there was a difference between accounting profits and profits for tax purposes, which is almost invariably the case, the effective tax rate would differ from the statutory tax rate. Indeed it was because of the likelihood that the statutory rate and effective rate would differ that Advocate General Geelhoed concluded that Articles 49 and 63 TFEU precluded the simultaneous use of exemption and imputation systems In conclusion, effective tax rates calculated on the basis of corporation tax actually paid on accounting profits would only exceptionally be equal to statutory or nominal rates applied to taxable profits. Moreover, this comparison cannot be carried out reasonably without full knowledge of the fiscally relevant properties of the companies to be compared and their activities. 28. I take note that in its written observations the Commission advises a Member State applying such a measure to insert a safeguard clause limiting its scope to dividends distributed by a company which is subject to the normal system of taxation in the source State. 29. On tax sparing credits see e.g. Viherkenttä, T., Tax incentives in developing countries and international taxation, Deventer, Kluwer 1991, p and 206; and Terra, B., Wattel, P., European Tax Law, 6th ed., Alphen an den Rijn, Wolters Kluwer, 2012, p A recent allusion to tax sparing credit was made in Case C-157/10 Banco Bilbao Vizcaya Argentaria [2011] ECR I-0000, paragraph Effective rate of tax has been characterised as [t]he taxpayers actual tax liability (or a reasonable estimate thereof) expressed as a percentage of a pre-tax income base rather than as a percentage of taxable income, i.e. tax rates which take into account not only the statutory tax rate, but other aspects of the tax system which determine the amount of tax paid. The effective rate of tax indicates the real, economic tax burden as opposed to the relationship between the tax liability and the profits, etc., as artificially adjusted for tax purposes. See Larking, cited above in footnote 22, p On issues relating to effective tax rates, see for example Nicodème, G., Computing effective corporate tax rates: comparisons and results, European Commission, Economic paper, Number 153 June 2001, available at The majority in the Court of Appeal (see Annex 3 to the Judgment of Court of Appeal of 23 February 2010, cited above in footnote 11) submits that assuming the Court had meant to refer to effective rates in the first FII judgment entails that it had misunderstood the arguments of the Test Claimants, explanations of the United Kingdom Government and the opinion of Advocate General Geelhoed.

7 54. Therefore, in my view, the combination of nominal and effective rates is, from the outset, not meaningful. Such a system would be difficult or even impossible to apply objectively. d. Statutory rate 55. The third option for interpreting paragraph 56 of the first FII judgment entails applying statutory or nominal tax rates. According to this alternative the Court referred to statutory tax rates for the purposes of assessing the impact of the simultaneous application of imputation and exemption methods. 56. In view of the discussion of the parties and the Court s rejection of the solution proposed by the Advocate General, this seems the most plausible interpretation of the first FII judgment. The question left for the national court would thus consist of examining whether it is true that only in exceptional cases lower nominal rates than the standard statutory rate would be used in corporate taxation of profits, which underlies the tax regime concerning nationally-sourced dividends in the United Kingdom. 57. Although I intend to propose answering Question 1 in the sense that the Court meant statutory or nominal rates, I will continue on the subject and consider issues that, in my opinion, necessarily follow from such an answer. e. The existence of a restriction and its justification 58. As I have already mentioned, I am of the opinion that Advocate General Geelhoed was correct when he considered that combining exemption of nationally-sourced dividends with credit for foreign-sourced dividends inevitably leads to less favourable treatment of foreign-sourced dividends. 33 This conclusion seems valid in the case of the United Kingdom independently of whether the comparison is based solely on statutory rates or on a combination of statutory and effective rates. 59. In fact, if the comparison is based on statutory rates, the less favourable treatment of foreign-sourced dividends is a systemic consequence of the differences between the two methods in regard to the possibility to pass on the tax advantages applicable to underlying corporation tax. However, if the comparison is based on a combination of statutory and effective rates, the less favourable treatment of foreignsourced dividends is a factual finding on how the United Kingdom system actually works, and as such not disputed in the main proceedings. 60. Therefore, in order to properly assist the national court, and to avoid a third preliminary reference in the main proceedings, the Court should in my opinion address the issue of whether the situation described above amounts to a restriction of freedom of establishment, and if it does, whether such restriction can be objectively justified. 61. If the Court gives the answer that paragraph 56 of the first FII judgment refers to statutory or nominal rates, and if the statutory rates are the same (except for exceptional situations), the issue of less favourable treatment of foreign-sourced dividends as a systemic consequence of the application of two different rules in comparable situations remains, and its nature either as a non-restriction or as a restriction that can or cannot be justified. In the same vein, if the Court opts for a combined application of nominal and effective rates, the national court will need guidance as to how effective rates are to be calculated. The national court will also need guidance as to the question of whether any difference between effective rates always constitutes a restriction or whether some margin can be allowed before a difference amounts to a restriction. The issue of justification is also relevant in this situation. 62. The restriction, if it can be considered to be one, is not created because a part of the foreign-sourced dividends were subject to economic double taxation, of which the nationally-sourced dividends would be spared. 34 The restriction is created because a part of the profits underlying the nationally-sourced dividends are not taxed at all, due to the fact that the effective corporation tax rate of the distributing company is lower than the statutory rate, and the exemption of dividends passes this relief on to the shareholders. Therefore the valid comparison is not between economic double taxation and single taxation, but between single taxation and partial zero taxation. From the point of view of relieving economic double taxation, the imputation and exemption methods are equally effective systems. 63. The next question is whether there is a restriction to the freedom of establishment, and if so, whether such a restriction can be justified. As I have already noted, my reading of the case-law preceding the first FII judgment is that the applicable national rules in the United Kingdom created a restriction as regards cross-border situations and that this restriction could not be justified. 64. However, on the basis of the first FII judgment and the subsequent case-law there is now also a possible alternative conclusion. 65. With regard to portfolio dividends, the objective underpinning the use of the imputation method, which is to top up the taxation of foreign-sourced dividends to the national level of taxation, was expressly accepted by the Court in Haribo Lakritzen Hans Riegel and Österreichische Salinen. The Court held that [a]pplication of the imputation method to dividends from non-resident companies makes it possible to ensure that foreign-sourced and nationally-sourced portfolio dividends bear the same tax burden, in particular where the State from which the dividends come applies, in the context of corporation tax, a lower tax rate than that applicable in the Member State where the company receiving the dividends is established. In such a case, exempting dividends from non-resident companies would give taxpayers that have invested in foreign holdings an advantage compared with those having invested in domestic holdings From this it could be concluded that the Member State of the recipient does not have to pass on the tax advantages provided in the tax legislation of the source State to the recipients of dividends, but can legitimately eliminate the effect of such advantages in its domestic taxation. In other words, although a Member State aiming at the elimination of economic double taxation at national level must take into account taxes paid abroad, that Member State is not obliged to recognise tax advantages in foreign source States. 67. However, as Advocate General Kokott observed in her opinion in Haribo Lakritzen Hans Riegel and Österreichische Salinen, if a Member State has chosen to aim at preventing double taxation of corporate profits by exempting nationally sourced dividends from corpo- 33. Point 50 of the opinion, cited above in point In the first FII judgment a restriction of free movement of capital concerning portfolio investments was established as no tax credit was available, thus leading to double economic taxation. 35. Cited in paragraph 89 of the judgment.

8 ration tax, it can be assumed that the desired level of taxation is already guaranteed through the levying of corporation tax on the distributing company. Because in specific cases this internal link between exemption at shareholder level and taxation at company level may be partially or entirely absent, an examination of whether discrimination occurs must be based not on an analysis of specific cases, but on an overview of the system Advocate General Kokott further notes that the close link between exemption applicable to nationally-sourced dividends and taxation at company level underlying an exemption system cannot be negated by common methods of reducing the tax burden such as the offsetting of losses and group relief. A tax system can be deemed not to seek to eliminate economic double taxation only if an overview analysis of the system shows that there is merely an apparent link between exemption and advance payment of tax, or such a link is clearly non-existent Thus the exemption method, when applied in corporate group taxation, rests on the principle that at the systemic level corporation tax levied on underlying profits is sufficient. In other words, the national legislator chooses to avoid a situation where the effects of tax advantages enjoyed by one company in the group would be eliminated in the taxation of companies at a higher level in the group. 70. Hence, according to this approach, in the absence of EU harmonisation, Member States would neither be obliged to recognise the economic effect of the tax policy choices of the source State in their tax treatment of foreign-sourced dividends, nor be bound to tax nationallysourced dividends distributed from profits which have been subject to corporation tax in accordance with the applicable tax law provisions. Rather, Member States would be entitled to apply their tax policies regarding statutory rates and tax bases both in relation to foreignsourced dividends and nationally-sourced dividends. 38 In consequence, the lack of capital export neutrality and the corresponding disincentive to the freedom of establishment would not amount to a forbidden restriction, provided that the same nominal tax rates were applied. 71. However, such asymmetric taxation is not an inevitable consequence of split tax jurisdictions within the EU. Rather, it follows from the tax policy choices of the Member State of the parent company. In fact this policy choice consists of the adoption of two tax policy elements that, in themselves, are justified under EU law, but where the simultaneous application of them both leads to a difference in treatment. 72. In conclusion, the application of an asymmetric mixed system tends to lead to less favourable treatment of foreign-sourced dividends, independently of whether the effective rates or the statutory rates are considered. The approach sketched above would see this difference of treatment as a consequence of the combined application of two legitimate principles of tax policy, and as such either as not amounting to a restriction or as a justified one. Admittedly this would lead to a more flexible application of internal market principles in this field of direct taxation than generally. f. Conclusion 73. In the light of the foregoing considerations, the answer to Question 1 should be that the references to tax rates and different levels of taxation at paragraph 56 of the first FII judgment refer to statutory or nominal rates of tax. For the reasons explained above, this answer leaves the issue of restriction and its justification open. This issue could be addressed either by returning to the answer suggested by Advocate General Geelhoed at point 56 of his opinion in the first FII case which forms my secondary proposal, or by simply admitting the acceptability of economic consequences of the asymmetric mixed system in EU law as it stands. V Question Question 2 seeks clarification of the Court s reply to Questions 2 and 4 of the first FII reference, concerning the United Kingdom s advance corporate tax and foreign income dividend regimes In response to Question 2 in the first FII reference, the Court held that Articles 49 and 63 TFEU precluded legislation of a Member State which allowed a resident company receiving dividends from another resident company to deduct from its own ACT liability the ACT paid by the dividend paying company, when no such deduction was permitted in the case of a resident company receiving dividends from a non-resident company in respect of the corporation tax on the distributed profits paid by the dividend-paying company in its Member State of residence. 76. The High Court notes that the Court s response focused on ACT paid by a resident company in direct receipt of a foreign-sourced dividend, in cases where corporation tax was paid by the dividend-paying non-resident company ( the water s edge company, or company D in the above diagram 40 ). In practice, however, very often the water s edge company did not pay any tax in its State of residence on the profits out of which the dividend was paid to its parent resident in the United Kingdom (company C in the diagram) because of the widespread use by international groups of intermediate holding companies that paid little or no tax on their profits. 36. See points 33, 34 and 39 of the Opinion. Actually, the answer proposed by the French Government in the current proceedings requires in essence that the national court examines, on the basis of effective tax rates charged to the distributing United Kingdom companies and the receiving United Kingdom companies, whether the exemption system applied does not in reality aim at mitigating economic double taxation or chain taxation but at making it possible for the recipient companies to benefit from tax exemptions of the distributing company that are not exceptional. 37. See Advocate General Kokott s Opinion in Haribo Lakritzen Hans Riegel and Österreichische Salinen, cited above in footnote 10, point The Commission points out rightly that the asymmetric system leads to different treatment of a similar relief granted in the source State and in the residence State. However, it is also possible that the residence State has a tax system where the differences between effective and statutory levels of corporate taxation result solely from generous possibilities to benefit at the group level from losses suffered by any company belonging to the group whereas the source State has a policy where significant tax advantages are granted on the basis of industrial and regional policy considerations. 39. See footnote 5 above. 40. See point 11 above.

9 77. When the case came back to the High Court, HMRC claimed that the Court s reply to Question 2 of the first FII reference only covered the case where the water s edge company itself had paid corporation tax in its State of residence. The Test Claimants, by contrast, considered that the Court s judgment applied equally where the dividend was paid out of profits comprising dividends paid by a lower-tier subsidiary resident in another Member State out of profits on which corporation tax was paid in that State (company E in the diagram). 78. The same issue arises in relation to the Court s reply to Question 4 of the first FII reference, where the Court held that Articles 49 TFEU and 63 TFEU precluded legislation of a Member State which, while exempting from ACT resident companies paying dividends to their shareholders which had their origin in nationally-sourced dividends, allowed resident parent companies distributing foreign-sourced dividends to their shareholders to elect to be taxed under the FID regime. The FID regime, firstly, allowed them to recover the ACT paid but required them first to pay it and seek a repayment and, secondly, entailed the loss of the tax credit for shareholders which attached to dividends distributed out of nationally-sourced dividends. 79. The Test Claimants and the Commission suggest that the Court s answer to Questions 2 and 4 in the first FII judgment should apply in cases described in Questions 2(a) and 2(b). However, the United Kingdom Government proposes that the judgment should be interpreted as meaning that Articles 49 TFEU and 63 TFEU are not infringed in either case. 80. At first glance I see no reason why the change in the subsidiary paying the tax (companies D or E in the above diagram) should lead to a different interpretation than the one given by the Court in the first FII judgment. Indeed, the legal principle applied by the Court in the relevant paragraphs of the first FII judgment was that of non-discrimination between foreign-sourced and nationally-sourced dividends in the application of the objective of preventing the imposition of a series of charges to tax which the United Kingdom legislation sought to avoid In essence, the national court seeks guidance on whether a similar obligation existed for the Member States already on the basis of the Treaty provisions independently of the situations covered by the provisions of Directive 90/ as it seems clear that Directive 90/435 and the amended provisions in particular are not applicable, given their material and temporal scope. 82. On this question I would subscribe to the Commission s analysis. It notes that under the ACT scheme a resident company could distribute dividends to its shareholders without paying ACT to the extent that those dividends stemmed from dividends paid to it by a resident subsidiary. No such exemption from ACT was available in the case of dividends funded by dividends paid by a foreign subsidiary. The payment of ACT in connection with such distributions resulted at least in a cash flow penalty in comparison with distributions funded by domestic dividends. In many cases, it resulted in an additional tax charge on foreign income which did not and could not exist in relation to domestic income. This additional charge constituted economic double taxation. 83. It is important to recall that the ACT formed an advance payment of corporation tax. Therefore any payment of ACT on distributions which included foreign dividends was justified solely to the extent that the foreign dividends stemmed from profits taxed at a rate lower than that applicable in the United Kingdom. 84. In the purely domestic scenario, ACT will be paid once, either by the United Kingdom subsidiary on the distribution of its profits or by the parent company on the ultimate distribution to individual shareholders. The payment of ACT will later be offset against the corporation tax liability of one of these companies. In the cross-border scenario there is no ground for the payment of ACT because there is no corporation tax liability in the United Kingdom (except to cover a difference in rates between the United Kingdom and the source State). 85. As the Court stated in paragraph 87 of the first FII judgment, a company receiving foreign-sourced dividends is, seen in the light of the objective of preventing the imposition of a series of charges to tax which the legislation at issue in the main proceedings seeks to avoid, in a comparable situation to that of a company receiving nationally-sourced dividends, even though only the latter receives dividends on which ACT has been paid. To my mind, this is so irrespective of the fact that it receives those dividends through an intermediate subsidiary. 86. These reasons are also valid for the answer to Question 2(b). A resident company which receives dividends from a foreign company should not need to pay ACT because it has no mainstream corporation tax liability in respect of those dividends (subject to any compensating payment as already mentioned). Equally, its parent company to which it distributes its profits including those dividends has no mainstream corporation tax liability in respect of the portion of the profits corresponding to those dividends, and there is thus no basis at all for requiring it to pay ACT. 87. In view of the above elements, I propose that the reply to Question 2 should be that the two scenarios presented under (a) and (b) make no difference to the Court s answer to Questions 2 and 4 of the first FII reference. VI Question 3 A The question and observations received 88. By Question 3 the national court wishes to explore the consequences possibly following from the answer to Question 2(b). Namely, if a parent company in the United Kingdom which received foreign dividends indirectly through a resident intermediate subsidiary has been unlawfully obliged to pay ACT, is it entitled to repayment of the tax unduly levied or solely to compensation in accordance with the conditions laid down in Brasserie du Pêcheur and Factortame See first FII judgment, cited above in footnote 2, paragraph I should mention, for the sake of completeness, that this question is to some extent regulated by Article 4(1) of Directive 90/435. In fact, the original version of Article 4(1), second indent, of Directive 90/435 referred to the tax paid by the subsidiary which relates to those profits. In 2003, however, the Commission proposed that this part would be amended to read tax paid by the subsidiary and any lower-tier subsidiary which relates to those profits, see COM(2003) 462, points 17 to 19. The Council adopted this amendment in Directive 2003/123/EC, but added a proviso reading subject to the condition that at each tier a company and its lower-tier subsidiary meet the requirements provided for in Articles 2 and 3 (see Council Directive 2003/123/EC of 22 December 2003 (OJ 2004 L 7, p. 41). 43. Cited above in footnote 13.