Opinion of the Advocate-General

Opinion of the Advocate-General

I – Introduction

1. It is hardly surprising that even insuring one’s own life is subject to a tax. At least that is the case in some Member States of the European Union. The fact that other Member States are more lenient in this respect by forgoing such a tax gives rise to a problem in the cross-frontier internal market in life assurance: how are distortions of competition due to the different levels of taxation to be avoided?

2. The reference for a preliminary ruling here at issue concerns a Netherlands assurance undertaking which concluded life assurance contracts with Netherlands policyholders. In the Netherlands, such contracts are not subject to an insurance tax. The situation is, however, different in neighbouring Belgium, to which some policyholders have moved. As with all important assets, they have taken their existing life assurance contracts with them.

3. Are these contracts now to be subject to the Belgian insurance tax simply because the policyholders have changed their place of residence? In principle, European Union (EU) law has an answer to that question, too. However, the EU legislature has not made it all that easy for those who apply the law. It is therefore necessary to determine which of the many provisions of EU law concerning assurance contain the answer.

II – Legal context

A – EU law

4. The decisive legislation in the present case, Directive 2002/83/EC (2) (‘Life Assurance Directive’ or ‘the Directive’), is based on Article 47(2) EC and Article 55 EC, according to which directives are to be issued to facilitate the taking-up and pursuit of activities as self-employed persons with a view to the right of establishment and the freedom to provide services. According to recitals 3 and 5 in the preamble to the Directive, its aim is to promote a single internal market in life assurance. According to recital 55, tax arrangements will also be needed to that end:

‘Some Member States do not subject assurance transactions to any form of indirect taxation, while the majority apply special taxes and other forms of contribution. The structures and rates of such taxes and contributions vary considerably between the Member States in which they are applied. It is desirable to prevent existing differences leading to distortions of competition in assurance services between Member States. Pending subsequent harmonisation, application of the tax systems and other forms of contribution provided for by the Member States in which commitments entered into are likely to remedy that problem and it is for the Member States to make arrangements to ensure that such taxes and contributions are collected.’

5. Pursuant to Article 2(1) of the Life Assurance Directive, the Directive is essentially to be applied to life assurance policies on survival to a stipulated age or on death and to any insurance against incapacity for employment and invalidity associated with such a policy.

6. Title IV of the Life Assurance Directive, which is entitled ‘Provisions relating to right of establishment and freedom to provide services’, contains, in Article 50, the following rules on taxes on premiums:

‘1. Without prejudice to any subsequent harmonisation, every assurance contract shall be subject exclusively to the indirect taxes and parafiscal charges on assurance premiums in the Member State of the commitment; …

2. The law applicable to the contract pursuant to Article 32 shall not affect the fiscal arrangements applicable.

3. Pending future harmonisation, each Member State shall apply to those assurance undertakings which cover commitments situated within its territory its own national provisions for measures to ensure the collection of indirect taxes and parafiscal charges due under paragraph 1.’

7. Article 1(1)(g) of the Life Assurance Directive defines the term ‘Member State of the commitment’ used in Article 50(1) as follows:

‘the Member State where the policyholder has his/her habitual residence or, if the policyholder is a legal person, the Member State where the latter’s establishment, to which the contract relates, is situated’.

8. The term ‘Member State of the commitment’ is also used in other provisions of the Life Assurance Directive. Article 32(1) of the Life Assurance Directive, for example, contains the following provision:

‘The law applicable to contracts relating to the activities referred to in this Directive shall be the law of the Member State of the commitment. However, where the law of that State so allows, the parties may choose the law of another country.’

9. The chapter of the Life Assurance Directive entitled ‘Contract law and conditions of assurance’ similarly sets out in Article 36 the following rules on information for policyholders:

‘1. Before the assurance contract is concluded, at least the information listed in Annex III(A) shall be communicated to the policyholder.

2. The policyholder shall be kept informed throughout the term of the contract of any change concerning the information listed in Annex III(B).

…’

10. Under point (a)(14), Annex III(A) includes in the information to be communicated before the assurance contract is concluded ‘general information on the tax arrangements applicable to the type of policy’. Annex III(B) specifies, under point (b)(2), as information to be communicated during the term of the contract, ‘[a]ll the information listed in points (a)(4) to (a)(12) of A in the event of a change in the policy conditions or amendment of the law applicable to the contract’.

11. Articles 41 and 42 of the Life Assurance Directive lay down certain supervisory obligations in connection with assurance undertakings starting cross-frontier business. Article 41 reads as follows:

‘Any assurance undertaking that intends to carry on business for the first time in one or more Member States under the freedom to provide services shall first inform the competent authorities of the home Member State, indicating the nature of the commitments it proposes to cover.’

12. Article 42 of the Life Assurance Directive reads as follows:

‘1. Within one month of the notification provided for in Article 41, the competent authorities of the home Member State shall communicate to the Member State or Member States within the territory of which the assurance undertaking intends to carry on business …

…

3. The assurance undertaking may start business on the certified date on which it is informed of the communication provided for in the first subparagraph of paragraph 1.’

13. Besides the Life Assurance Directive, EU law includes other directives concerning other segments of the insurance market. Article 46(2) of Directive 92/49/EEC, (3) which, according to Article 2(2) thereof in conjunction with Article 2(1)(a) of Directive 73/239/EEC, (4) does not apply to life assurance, provides:

‘Without prejudice to any subsequent harmonisation, every insurance contract concluded by way of provision of services shall be subject exclusively to the indirect taxes and parafiscal charges on insurance premiums in the Member State in which the risk is situated within the meaning of Article 2(d) of Article 88/357/EEC …’

14. The ‘Member State in which the risk is situated’ is defined by Article 2(d) of Directive 88/357/EEC, (5) as a function of the type of insurance, as follows:

‘– the Member State in which the property is situated, where the insurance relates either to buildings or to buildings and their contents, …

– the Member State of registration, where the insurance relates to vehicles …

– the Member State where the policyholder took out the policy in the case of policies … covering travel or holiday risks, …

– the Member State where the policyholder has his habitual residence or, if the policyholder is a legal person, the Member State where the latter’s establishment, to which the contract relates, is situated, in all cases not explicitly covered by the foregoing indents’.

15. All the aforementioned directives will be replaced by Directive 2009/138/EC (6) with effect from 1 November 2012. Article 157 of that directive, which concerns taxes on premiums, reads as follows:

‘1. Without prejudice to any subsequent harmonisation, every insurance contract shall be subject exclusively to the indirect taxes and parafiscal charges on insurance premiums in the Member State in which the risk is situated or the Member State of the commitment.

…’

B – Belgian law

16. In Belgium, an annual insurance tax was levied in the years of relevance to the main proceedings. Pursuant to Article 173 of the Wetboek diverse rechten en taksen (Code on miscellaneous levies and taxes; ‘WDRT’), that tax is levied on insurance contracts if the insured risk is situated in Belgium. According to that provision, the risk is deemed to be situated in Belgium if the policyholder is habitually resident or, as a legal person, has his registered office, to which the contract relates, in Belgium.

17. Article 173(3) of the WDRT provides for a special tax rate of 1.1% of the insurance premiums in the case of life assurance.

III – The main proceedings and the questions referred

18. The applicant in the main proceedings is the Netherlands company RVS Levensverzekeringen NV, which offers life assurance policies (‘the taxpayer’). The main proceedings concern the insurance tax owed by the taxpayer in Belgium for the years 2006 and 2007.

19. The taxpayer has concluded insurance contracts with a number of persons who were resident in the Netherlands at the time of the conclusion of the contracts, but resident in Belgium in 2006 and 2007. The Belgian tax administration holds that, in view of those insurance contracts, the taxpayer is liable for Belgian insurance tax amounting to EUR 16 542 for the years 2006 and 2007.

20. Having initially paid that amount to the Belgian tax administration, the taxpayer has brought an action before the referring court for its repayment. The taxpayer argues that it is liable to insurance tax in Belgium only if the policyholders were resident in Belgium when the contracts were concluded. The Belgian tax administration, on the other hand, takes the view that the determining factor is the place of residence at the time when the insurance premium is paid.

21. It is against that background that the Rechtbank van eerste aanleg te Brussel (Court of First Instance, Brussels) has referred the following questions to the Court of Justice:

‘(1) Does Article 50 of [the Life Assurance Directive], which in paragraph 1 provides that, without prejudice to any subsequent harmonisation, every assurance contract shall be subject exclusively to the indirect taxes and parafiscal charges on assurance premiums in the Member State of the commitment, and which in paragraph 3 provides that, pending future harmonisation, each Member State shall apply to those assurance undertakings which cover commitments situated within its territory its own national provisions for measures to ensure the collection of indirect taxes and parafiscal charges due under paragraph 1, preclude a national rule as laid down in Article 173 and Article 175(3) of the [WDRT], which provides that insurance transactions (including life assurance policies) are subject to an annual tax, when the risk is situated in Belgium, in particular if the policyholder has his/her habitual residence in Belgium, or, if the policyholder is a legal person, if the establishment of that legal person, to which the contract relates, is situated in Belgium, without any account being taken of the place of residence of the policyholder at the time that the contract was concluded?

(2) Do the Community principles concerning the elimination, between the Member States of the Community, of obstacles to the freedom of movement of persons and the freedom to provide services arising out of Articles 49 and 56 of the Treaty on the Functioning of the European Union preclude a national rule as laid down in Article 173 and Article 175(3) of the [WDRT], which provides that insurance transactions (including life assurance policies) are subject to an annual tax, when the risk is situated in Belgium, in particular if the policyholder has his/her habitual residence in Belgium, or, if the policyholder is a legal person, if the establishment of that legal person, to which the contract relates, is situated in Belgium, without any account being taken of the place of residence of the policyholder at the time that the contract was concluded?’

IV – Legal assessment

A – Interpretation of the questions referred

22. Before beginning to answer the questions referred, I consider it necessary to determine their scope.

23. As regards the first question referred, the European Commission has rightly pointed out that it is worded too generally, given the facts of the case as described in the main proceedings, and should therefore be interpreted restrictively. The current proceedings do not, indeed, give cause to consider the compliance of the Belgian tax arrangement with EU law as comprehensively as the wording of the question referred would require.

24. For one thing, the taxation of insurance policies held by legal persons is clearly irrelevant in the main proceedings. For another, the only question in the present case is whether Article 50 of the Life Assurance Directive precludes a specific national arrangement for life assurance contracts rather than for any type of insurance contract. Consequently, the question referred concerns only the interpretation of Article 50 of the Life Assurance Directive as it relates to life assurance contracts and only in respect of policyholders who are natural persons.

25. In the second question referred, the referring court seeks additional clarification as to whether certain fundamental freedoms preclude the Belgian arrangement.

26. In the present case, however, it should be borne in mind that Article 50 of the Life Assurance Directive itself sets out a conclusive provision of secondary law by which the admissibility of the Belgian arrangement under EU law is judged. That is the target of the first question referred. Against that background, the Belgian arrangement must be measured not against the fundamental freedoms laid down in primary law, but against the secondary law of Article 50 of the Life Assurance Directive. In the interpretation of that provision, however, the importance of the fundamental freedoms must be taken into account in order to preclude an interpretation of Article 50 which is inconsistent with the fundamental freedoms. The aspects of primary law to which the second question refers must therefore be considered in the answer to the first question.

27. Nor, then, is it necessary to decide whether the second question is in fact admissible as such. As the Commission has rightly argued, in particular, it is not entirely clear from the reference for a preliminary ruling which fundamental freedoms the referring court wishes to have interpreted.

28. In summary, the foregoing indicates that the questions referred can be taken to mean that the referring court wants to know whether Article 50 of the Life Assurance Directive – with due regard for the fundamental freedoms – precludes a national arrangement whereby life assurance contracts are subject to an annual tax if, in the given year, the policyholder is habitually resident in the Member State concerned, irrespective of his place of residence at the time when the contract was concluded.

B – Answer to the questions referred

29. For this question to be answered, it needs to be established how Article 50 of the Life Assurance Directive shares taxation powers amongst Member States with respect to tax on life assurance contracts.

30. Pursuant to Article 50(1) of the Life Assurance Directive, the ‘Member State of the commitment’ has that power to tax. Insurance contracts are accordingly subject only to the indirect taxes levied on insurance premiums in the Member State of the commitment. ‘Member State of the commitment’ is defined in Article 1(1)(g) of the Directive as the Member State in which the policyholder has his or her habitual residence. The present proceedings have shown that those provisions are essentially open to two interpretations.

31. On the one hand, a ‘static’ interpretation is possible, the ‘Member State of the commitment’ being specified on one occasion when the contract is concluded. The Member State in which the policyholder has his or her habitual residence at the time when the contract is concluded then has the power of taxation throughout the term of the contract. That interpretation is advocated by the taxpayer and the Estonian Government.

32. On the other hand, there is the possibility of a ‘dynamic’ interpretation, whereby the ‘Member State of the commitment’ may change with the passage of time. The Member State in which the policyholder has his or her habitual residence at the time when the tax on the assurance premium is collected would thus have the power to tax. The place of residence might change from one premium payment to the next. Such a dynamic interpretation is supported by the Belgian and Austrian Governments and by the Commission.

33. In the following, I will show that, (1) although the rules laid down by the EU legislature for the sharing of the powers to levy insurance tax are not clear, (2) the context and (3) the purpose of the arrangement lead, with due regard for the fundamental freedoms, to a static interpretation of Article 50 of the Life Assurance Directive.

1.  Interpretation of the wording

34. Contrary to the arguments of some participants in the proceedings, the wording does not provide an unambiguous answer to the question posed in the present case. That is true not only of the version in the language of the proceedings, Dutch, but also of the German-, English- and French-language versions of Article 50(1) and the supplementary Article 1(1)(g) of the Life Assurance Directive. Their wording makes it clear only that the tax is to be levied where the policyholder has his or her habitual residence. The decisive question here – at what time – is not, however, answered by the wording in the aforementioned language versions.

35. Nor is any information on the relevant time to be derived from the designation as Member State ‘of the commitment’, since that ‘commitment’ can be taken to mean not only the entry into an obligation but also the existence of an obligation. Consequently, it cannot be unequivocally inferred from that wording either that the Member State is to be designated on one occasion, at the time of the conclusion of the contract, or that it must be designated anew throughout the existence of the contract.

36. Finally, the wording of Article 50(1) of the Life Assurance Directive is equally ambivalent in that it refers both to the collection of the tax on assurance premiums and to the assurance contracts which are subject to the tax. It thus remains unclear whether the decisive link according to the Directive is the conclusion of the contract or the payment of the premiums.

2.  Systematic interpretation

37. However, there are numerous provisions surrounding Article 50(1) of the Life Assurance Directive which may give some indication of how it is to be interpreted.

a) Collection of the insurance tax

38. In this context, reference must first be made to Article 50(3) of the Life Assurance Directive. According to that provision, the collection of the insurance tax is to be ensured on the basis of national provisions. The Member State applies those provisions to assurance undertakings which ‘cover commitments situated within its territory’. From that, the taxpayer in particular infers that the time of the conclusion of the contract determines the attribution of the power to tax.

39. However, the Commission has rightly pointed out that the wording chosen for Article 50(3) does not necessarily refer to the conclusion of the assurance contract. The view taken depends on to what the phrase ‘situated within its territory’ refers.

40. On the one hand, it may refer to ‘cover’ and so emphasise the place where the contract is concluded. On the other hand, it is also possible for that phrase, and thus the designation of the place, to be related to ‘the commitments’. The decisive factor would then in each case be where the commitments covered exist. That possible interpretation is reflected even more clearly in the English-language version, which refers to the covering of the commitments situated in a Member State. (7) The place where a commitment is situated may be, for example, the habitual residence of the policyholder, but in any event it is not the place where the contract is concluded.

41. The rule laid down in Article 50(3) thus favours neither the static nor the dynamic interpretation of Article 50(1) of the Life Assurance Directive.

b) The term ‘Member State of the commitment’ in other provisions

42. The use of the phrase ‘Member State of the commitment’ in other provisions of the Life Assurance Directive may be relevant to the interpretation of Article 50(1), since Article 1(1)(g) of the Directive sets out a uniform definition of that term for all the provisions of the Directive.

43. However, that uniform definition is already guaranteed by the fact that in every provision it means the Member State in which the policyholder has his or her habitual residence. The decisive time is not, on the other hand, specified in Article 1(1)(g) of the Directive. If, then, the relevant time does not form part of the definition, it may be determined differently in any provision which refers to the ‘Member State of the commitment.’

44. That is also true of the rule laid down in the first sentence of Article 32(1) of the Life Assurance Directive, which has been cited by the taxpayer. For the determination of the law to be applied to the assurance contract, that provision refers to the ‘Member State of the commitment’. I indeed share the taxpayer’s view that a dynamic interpretation of that provision under which the law applicable changes whenever the policyholder changes his place of residence cannot have been the intention of the EU legislature. In the context of the first sentence of Article 32(1) of the Life Assurance Directive, the ‘Member State of the commitment’ is therefore likely to be the Member State in which the policyholder is habitually resident at the time of the conclusion of the contract.

45. In view of the foregoing, that does not, however, mean that the same time is relevant in the context of the provision to be considered here, Article 50(1) of the Life Assurance Directive. Furthermore, Article 50(2) of the Life Assurance Directive provides that the fiscal arrangements applicable are not affected by the law applicable to the assurance contract pursuant to Article 32 of the Directive. Although that rule is primarily intended to cover the law chosen by way of derogation by the parties pursuant to the second sentence of Article 32(1), Article 50(2) also emphasises the independence of the applicable law from the applicable fiscal arrangements.

c) Information in the assurance contract

46. However, the Life Assurance Directive contains elsewhere further rules on the assurance contract which may give some indication of how the questions referred should be answered. Thus Article 36 provides that certain information is to be communicated at the beginning and during the term of the contract.

47. Pursuant to Article 36(1) of in conjunction with Annex III(A)(a)(14) to the Life Assurance Directive, ‘general information on the tax arrangements applicable to the type of policy’ is to be communicated to the policyholder before the conclusion of the contract. What tax arrangements are applicable in this respect depends on the ‘Member State of the commitment’ within the meaning of Article 50(1) of the Life Assurance Directive.

48. During the term of the contract certain information is similarly to be communicated to the policyholder in accordance with Article 36(2) of the Life Assurance Directive. Pursuant to Annex III(B)(b)(2) to the Life Assurance Directive, that includes the updating of wide-ranging information on the rights and obligations of the parties to the contract which was to have been communicated at the time of the conclusion of the contract, in the event of an amendment of the law applicable to the contract. The reference to the information to be communicated at the time of the conclusion of the contract does not, however, extend to information on the applicable tax arrangements.

49. From that rule, it can be seen that the EU legislature clearly did not expect the applicable tax arrangements to change during the term of the contract. It cannot be assumed that, while the applicable tax arrangements are so important that they figure among the information which must be communicated before the conclusion of the contract, a communication on any change to the applicable tax arrangements during the term of the contract is superfluous.

50. Article 36 of in conjunction with Annex III to the Life Assurance Directive thus contains a clear indication of the need for a static interpretation of Article 50(1) of the Life Assurance Directive.

d) Requirements regarding the supervision of activities

51. In the context of systematic interpretation, the Estonian Government has also referred to the duty of notification to which the assurance undertaking is subject pursuant to Article 41 of the Life Assurance Directive. According to that provision, an official notification is necessary, it being, pursuant to Article 42(1) and (3) of the Directive, one of the requirements for an activity in another Member State.

52. As I see it, however, the Estonian Government is wrong to assume that assurance undertakings could not comply with that notification requirement if the habitual place of residence of the policyholder at the time of the payment of the premium was decisive for taxation, since the interpretation of the term ‘carry[ing] on business’ as used in Article 41 of the Life Assurance Directive must be distinguished from the question of the attribution of the power to tax.

53. As the Commission, too, has emphasised, the question in the present case is not whether, as a result of a policyholder’s emigration to another Member State, an assurance undertaking carries on business there within the meaning Article 41 of the Life Assurance Directive. The sole question is what tax implications emigration has under Article 50 of the Life Assurance Directive. Nor is there any apparent compelling reason for giving the same answer to both questions.

e) Other directives on the assurance market

54. Leaving the framework of the Life Assurance Directive, both the Belgian Government and the Commission have made a comparison with the provisions of directives applicable in EU law to insurance other than life assurance.

55. Article 46(2) of Directive 92/49, for example, provides for a tax rule which matches that laid down in Article 50(1) of the Life Assurance Directive. The only difference is that the former provision refers not to the ‘Member State of the commitment’ but to the ‘Member State in which the risk is situated within the meaning of Article 2(d) of Directive 88/357/EEC’.

56. The definition of the situation of the risk in Article 2(d) of Directive 88/357 differs as a function of the type of insurance contract. While the fourth indent of that provision contains a catch-all definition which is identical to the definition of the ‘Member State of the commitment’ in the Life Assurance Directive, the third indent considers the Member State in which the policyholder concluded the contract to be decisive in the case of travel insurance contracts.

57. I cannot, however, share the conclusion drawn by the Belgian Government from that comparison, namely that the conclusion of the contract is not therefore relevant to the application of Article 50(1) of the Life Assurance Directive. Although it is true that it can be inferred from the differing definitions given in the third and fourth indents of Article 2(d) of Directive 88/357 that the EU legislature did not intend the ‘Member State of the commitment’ to be the place where the contract was concluded, the question in the present case is not whether the place where the contract is concluded is decisive in the determination of the applicable tax arrangements under Article 50(1) of the Life Assurance Directive, but whether the habitual place of residence of the policyholder at the time when the contract is concluded is decisive. As the habitual place of residence is, as a rule, the same as the permanent residence of the policyholder, the Member State in which the contract is concluded and the Member State in which the policyholder has his habitual residence at that time may well be different.

58. Nor can Directive 2009/138, partly cited above, which will replace the various insurance directives in EU law in the future, shed any further light on the matter which might help to answer the questions referred.

59. Article 157 of that directive considers the ‘Member State in which the risk is situated or the Member State of the commitment’ to be decisive for the applicable fiscal arrangements. From that, it can be inferred, at best, that the EU legislature envisaged that a distinction was to be made between the term ‘Member State of the commitment,’ as used in Article 50(1) of the Life Assurance Directive, and the Member State in which the risk is situated.

60. It should be beyond question, however, that the place where the risk is situated is irrelevant to the application of Article 50(1) of the Life Assurance Directive, since, given the subject-matter of the contract, that place would have to be determined by reference to the place of residence of the person whose life was assured and who need not necessarily be the policyholder. As, however, that place is not decisive for either the static or the dynamic interpretation, the distinction made by Article 157 of Directive 2009/138, for instance, between the Member State of the commitment and the Member State where the risk is situated is irrelevant to the present interpretation problem.

3.  Teleological interpretation

61. Now that it has been established that, from a systematic viewpoint, Article 36 of in conjunction with Annex III to the Life Assurance Directive argues for a static interpretation of Article 50(1) of the Directive, (8) I will turn to the purpose of that latter provision.

62. Its purpose has been specified by the EU legislature in recital 55 to the Life Assurance Directive. On that basis, Article 50 of the Life Assurance Directive is meant to prevent distortions of competition in assurance services between Member States. Distortions of competition may result from differences between the Member States in the collection of indirect taxes on assurance contracts. That is to be countered by the uniform application of the tax system which applies in the Member State ‘in which commitments [are] entered into’.

63. First of all, it is impossible to infer from that wording the relevant Member State as I have already explained with regard to the interpretation of Article 50(3) of the Life Assurance Directive. (9) That assessment is not fundamentally challenged by the fact that the English-language wording of recital 55 differs from that of Article 50(3) of the Life Assurance Directive. (10) Not even an interpretation of the wording of recital 55 as meaning that the Member State in which the contract is concluded determines which tax system is applied would argue for either the static or the dynamic interpretation of Article 50(1) of the Life Assurance Directive. As I have already pointed out, (11) the place at which the contract is concluded must be distinguished from the place where the policyholder is habitually resident at the time when the contract is concluded.

64. Moreover, the primary objective in teleological interpretation should not be to interpret the wording of a recital, but to take account of the sense and purpose of the provision to be interpreted here, Article 50(1) of the Life Assurance Directive, as determined from the clear sense of the recital and from other sources.

a) Prevention of distortions of competition

65. Article 50(1) of the Life Assurance Directive is undoubtedly intended, according to recital 55 thereto, to prevent distortions of competition in assurance services.

66. The taxpayer has argued on this issue that there is no longer a competitive situation where a policyholder has concluded an assurance contract in the Netherlands and then moved his domicile to Belgium. A static interpretation would therefore be consistent with the objective of Article 50(1) of the Life Assurance Directive.

67. As I see it, although that may be true of competition for the conclusion of the contract, it is not true of competition for the change of service provider in the case of an existing assurance contract. The latter competition is distorted if only the habitual residence of the policyholder at the time of the conclusion of the contract is to be taken into account. After emigration, the policyholder would, for example, enjoy the privilege of non-taxation if, at the time of concluding the contract, he had his habitual place of residence in a Member State which did not provide for the taxation of assurance contracts. If, however, an assurance tax was collected at this new place of residence, it would have to be levied on the conclusion of a new assurance contract. That disadvantage might deter the policyholder from the outset from changing his assurance provider.

68. Admittedly, recital 55 to the Life Assurance Directive emphasises that the applicable tax arrangements should prevent distortions of competition ‘pending subsequent harmonisation’. That appears to imply an acknowledgement that the applicable tax arrangements do not yet entirely prevent distortions of competition. An interpretation of Article 50 of the Life Assurance Directive which does not preclude all distortions of competition would therefore still be consistent with the purpose of the arrangement.

69. That view is also endorsed by the fact that, in the discussion at the hearing of a written question from the Court of Justice, the Commission stated that, as far as it knew, questions relating to the mobility of the policyholder and its effects on the applicable tax arrangements had not been considered either in the context of the evolution of the Life Assurance Directive or of the directives preceding it. It is not therefore clear that the EU legislature intended Article 50 of the Life Assurance Directive to be the vehicle for preventing distortions of competition during a contractual relationship.

70. Given the nature of risk in the life assurance sphere, competition between current contracts and new contracts concluded with a different service provider is in any case likely to be limited, since both the risk of survival until a stipulated age and the risk of death grow with the increasing age of the policyholder. For this reason, the premiums on newly concluded life assurance contracts are likely to be structurally higher than those on existing contracts.

71. The goal of preventing distortions of competition thus tends to argue for a dynamic interpretation of Article 50(1) of the Life Assurance Directive, though not compellingly so.

b) Avoiding double taxation and non-taxation

72. As the objective of the arrangement, the Court of Justice has also referred, in the context of comparable tax rules laid down in Directive 88/357 and the fourth indent of Article 2(d) of that directive, (12) not only to the avoidance of distortions of competition but also to the prevention of double taxation and non-taxation. In that respect, it has placed the emphasis on objectively verifiable characteristics. (13)

73. First of all, I am not persuaded in the present context by the Commission’s view that there is a particular risk of double taxation where taxation depends on the policyholder’s residence at the time of the conclusion of the contract. That risk exists only if different Member States construe the tax arrangement differently. In principle, then, both double taxation and non-taxation are prevented by the uniform interpretation and application of Article 50(1) of the Life Assurance Directive in the European Union, regardless of the decisive criterion in that interpretation.

74. It must be said, however, that the objective verifiability for which the Court calls is better guaranteed for all concerned by a static interpretation. The habitual place of residence of the policyholder then needs to be established only once, at the beginning of the contract. As a result, situations in which different Member States differ in their views on where a given policyholder has his habitual residence in the year concerned can be reduced to a minimum. It is indeed situations of that nature which would lead to double taxation or non-taxation.

75. The objective associated with Article 50(1) of the Life Assurance Directive of preventing double taxation and non-taxation thus points to a static interpretation.

c) Taking the fundamental freedoms into account

76. In the context of teleological interpretation, account also needs to be taken of the fundamental freedoms, in the light of which secondary law must be interpreted. (14) The very heading of Title IV of the Life Assurance Directive, of which the provision to be interpreted here, Article 50(1), forms part, points in particular to the objective of guaranteeing the freedom of establishment and the freedom to provide services.

77. In taking the fundamental freedoms into account, the aim is not, in the present context, to examine a specific infringement of one of those freedoms, but rather to determine which interpretation better safeguards the fundamental freedoms in all the cases covered by Article 50(1) of the Life Assurance Directive.

78. While there is no discernible effect on the assurance undertaking’s freedom of establishment within the scope of Article 50(1) of the Life Assurance Directive, a closer look needs to be taken at the relevance to the interpretation of that provision of the assurance undertaking’s freedom to provide services and the policyholder’s freedom of movement.

i) Freedom to provide services

79. First of all, the issue is what influence the freedom to provide services granted by the Treaty may have on the interpretation of Article 50(1) of the Life Assurance Directive.

80. According to the Court’s settled case-law, national measures which prohibit, impede or render less attractive the exercise of the freedom to provide services are to be regarded as constituting restrictions on that freedom. (15) As the present case concerns not the assessment of a national measure, but the consideration of the freedom to provide services in the interpretation of secondary law, it needs to be determined which interpretation least restricts the freedom to provide services in that respect.

81. The taxpayer has argued in this context that a dynamic interpretation would restrict the freedom to provide services, since Netherlands assurance undertakings would have to pay the tax when a policyholder changed his place of residence, without being able to obtain reimbursement from the policyholder under the contract. Passing the tax on to the policyholder would consequently no longer be possible.

82. To counter that argument, however, it must be said that passing the tax on in the event of a change of residence is merely a question of the structure of the contract. The taxpayer appears not to have assumed a dynamic interpretation when structuring its contracts and has not therefore agreed an adjustment of the consideration in the event of the policyholder’s change of residence. In the case of a dynamic interpretation, the taxpayer would thus be confronted with a problem resulting not from that interpretation itself, but solely from its incorrect assessment of the applicable legal situation. It must be admitted, nevertheless, that that problem can arise only in the case of a dynamic interpretation.

83. In another respect, however, a dynamic interpretation clearly has a greater adverse effect on the freedom to provide services than a static interpretation.

84. The Estonian Government has rightly referred to the difficulties that would arise for assurance undertakings, even if the tax was passed on, as a result of a dynamic interpretation of Article 50(1) of the Life Assurance Directive. In that event, to fulfil their fiscal obligations, assurance undertakings would have constantly to verify the current habitual places of residence of their policyholders and, above all, the tax rules currently applicable at each such place. The undertakings would thus be forced to comply with many different tax rules laid down by the Member States solely because, whether they liked it or not, their contractual partners had moved their places of residence to other Member States.

85. In a static interpretation, on the other hand, the tax rules of another Member State need to be taken into account only if an assurance undertaking consciously decides to conclude a life assurance contract with a person who has his habitual place of residence in another Member State. Above all, only those tax rules are to be applied throughout the term of the contract.

86. A dynamic interpretation thus imposes a greater restriction on the cross-frontier services of an assurance undertaking, since it makes the contract subject to a different tax regime if the policyholder changes his place of residence.

87. A legal consequence of that nature might also deter assurance undertakings from continuing to provide assurance services after the policyholder changes his place of residence. To avoid the expense associated with the dynamic interpretation of Article 50(1) of the Life Assurance Directive in complying with the tax rules of another Member State, they might be at pains to conclude contracts only on condition that policyholders maintain a given place of residence. Regardless of whether such conduct by an assurance undertaking would be compatible with EU law, it is at least not to be feared in the case of a static interpretation.

88. In effect, then, the guarantee of the freedom to provide services tends to argue for a static interpretation.

ii) Freedom of movement for the policyholder

89. I am, on the other hand, unable to identify a direct restriction of the policyholder’s various rights to freedom of movement due to a dynamic interpretation of Article 50(1) of the Life Assurance Directive, as some of the participants in the proceedings have claimed.

90. Although a dynamic interpretation might, depending on the structure of the contract, result in the policyholder having to pay more for his life assurance contract at his new place of residence because of a higher insurance tax, the Court has already held in settled case-law that the Treaty offers no guarantee to a citizen of the European Union that transferring his activities to a Member State other than that in which he previously resided will be neutral as regards taxation. (16)

91. On the other hand, the policyholder’s freedom of movement might be affected indirectly in the event of a dynamic interpretation if – as previously explained (17) – his assurance undertaking refused to uphold the contract after a cross-frontier change of residence. As there would be no fear of such adverse consequences for the policyholder in the case of a static interpretation, the guarantee of his freedom of movement again argues for that interpretation.

d) Appropriateness of the allocation of taxation powers

92. Finally, the appropriateness of the allocation of taxation powers amongst Member States is not – contrary to the argument presented by the Austrian Government – an objective to be borne in mind in the interpretation of Article 50(1) of the Life Assurance Directive. Whether a certain allocation of the Member States’ taxation powers should be regarded as appropriate after a static rather than a dynamic interpretation could be determined only by reference to the purpose of collecting an insurance tax. That purpose might suggest which Member State should be entitled to the tax, especially if it could be determined which of the two contracting parties to a life assurance contract is intended to carry the tax burden.

93. The purpose of collecting an insurance tax is, however, irrelevant to the allocation of taxation powers amongst Member States pursuant to Article 50(1) of the Life Assurance Directive. Seen from the perspective of EU law, therefore, it is also irrelevant in principle who carries the burden of an insurance tax. For one thing, that follows from the fact that the Life Assurance Directive is based on Article 47(2) EC and Article 55 EC, which are devoted solely to the achievement of the freedom of establishment and the freedom to provide services. For another, the justification of the allocation of taxation powers for which Article 50(1) of the Life Assurance Directive provides is limited in recital 55 thereto to the prevention of distortions of competition.

94. EU law thus governs the applicability of national taxes solely from the viewpoint of distortions of competition affecting the internal market. In other words, it does not seek to levy an insurance tax or to encumber a given contracting party with the tax. Rather, the existence of national insurance taxes is accepted as given. Their collection – and thus their purpose – continues to be a matter for the Member States alone, as Article 135(1)(a) and Article 401 of Directive 2006/112/EC make clear. (18)

e) Outcome of the teleological interpretation

95. From the discussion of the sense and purpose of Article 50(1) of the Life Assurance Directive, it can thus be inferred that both the objective of preventing double taxation and non-taxation and the guarantee of the freedom to provide services and also the policyholder’s freedom of movement argue for a static interpretation. Given that, the fact that a dynamic interpretation might prevent distortions of competition more effectively than a static interpretation is of less importance.

V – Conclusion

96. As both the context and the sense and purpose of Article 50(1) of the Life Assurance Directive, with due regard for the fundamental freedoms, thus argue for a static interpretation, I propose that the Court of Justice should answer the questions referred by the Rechtbank van eerste aanleg te Brussel as follows:

Article 50(1) of Directive 2002/83/EC of the European Parliament and of the Council of 5 November 2002 concerning life assurance should be interpreted as meaning that the ‘Member State of the commitment’ is to be determined, in the case of natural persons, at the time when the life assurance contract is concluded. It therefore precludes a national arrangement under which life assurance contracts are subject to an annual tax if in the year concerned the policyholder has his habitual place of residence in the Member State but the policyholder’s place of residence at the time of the conclusion of the contract is not taken into account.

(1) .

(2)  – Directive of the European Parliament and of the Council of 5 November 2002 concerning life assurance (OJ 2002 L 345, p. 1).

(3)  – Council Directive of 18 June 1992 on the coordination of laws, regulations and administrative provisions relating to direct insurance other than life assurance and amending Directives 73/239/EEC and 88/357/EEC (Third Non-life Insurance Directive) (OJ 1992 L 228, p. 1).

(4)  – First Council Directive on the coordination of laws, regulations and administrative provisions relating to the taking-up and pursuit of the business of direct insurance other than life assurance (OJ 1973 L 228, p. 3).

(5)  – Second Council Directive of 22 June 1988 on the coordination of laws, regulations and administrative provisions relating to direct insurance other than life assurance and laying down provisions to facilitate the effective exercise of freedom to provide services and amending Directive 73/239/EEC (OJ 1988 L 172, p. 1).

(6)  – Directive of the European Parliament and of the Council of 25 November 2009 on the taking-up and pursuit of the business of Insurance and Reinsurance (Solvency II) (OJ 2009 L 335, p. 1); see recital 1 to and Article 310(1) of the directive.

(7)  – ‘… [E]ach Member State shall apply to those assurance undertakings which cover commitments situated within its territory …’

(8)  – See point 46 et seq. above.

(9)  – See point 38 et seq. above.

(10)  – ‘… application of the tax systems … provided for by the Member States in which commitments entered into …’.

(11)  – See point 57 above.

(12)  – See point 55 et seq. above.

(13)  – See Case C-191/99 Kvaerner [2001] ECR I-4447, paragraph 51.

(14)  – See, to that effect, Case 246/80 Broekmeulen [1981] ECR 2311, paragraph 20; Case C-101/01 Lindqvist [2003] ECR I-12971, paragraph 87; Case C-305/05 Ordre des barreaux francophones et germanophone and Others [2007] ECR I-5305, paragraph 28; and the Opinion of Advocate General Alber in Case C-435/00 Geha Naftiliaki and Others [2002] ECR I-10615, point 40.

(15)  – See, for example, Case C-9/11 Waypoint Aviation [2011] ECR I-9697, paragraph 22 and the case-law cited.

(16)  – See, for example, Case C-387/01 Weigel [2004] ECR I-4981, paragraph 55; Case C-403/03 Schempp [2005] ECR I-6421, paragraph 45; and Case C-392/05 Alevizos [2007] ECR I-3505, paragraph 76.

(17)  – See point 87 above.

(18)  – Council Directive of 28 November 2006 on the common system of value added tax (OJ 2006 L 347, p. 1).