United Biscuits (Pension Trustees) Ltd & Anor v Revenue And Customs

The first Claimant is the trustee of the United Biscuits Pension Fund, a defined benefit occupational pension scheme for employees of United Biscuits (UK) Ltd. The second Claimant is the former trustee of the UB Pension Investment Fund, a collective investment fund in which the assets of the pension scheme were invested from 1989 to 2006. The Claimants sue on behalf of themselves and their predecessors in title as trustees of the pension fund and the investment fund. I will refer to the Claimants and their predecessors in title together as “the Trustees”.

This is the trial of liability in an action in which the Trustees seek to recover from the Defendants (“HMRC”) VAT which they have paid to various investment managers. These managers include both companies authorised to conduct insurance business under the relevant UK Insurance Companies Acts in force from time to time (“Insurers”)and companies not so authorised (“Non-Insurers”) but authorised, nonetheless, by financial regulators to provide the pension fund management services in question. Pension fund management services have been treated by HMRC as exempt supplies when provided by Insurers but as standard rated supplies when provided by Non-Insurers. Until 1 January 2005, the different treatment of supplies by Insurers on the one hand and by Non-Insurers on the other hand was in accordance with UK statute. Following amendment of the UK legislation with effect from that date, the difference in treatment was no longer in accordance with UK statute, but continued to be applied by HMRC.

The Trustees’ case is that the supplies made by Non-Insurers were insurance transactions for the purposes of the 6th Directive and the Principal VAT Directive (as to which see paragraphs 17 and 18 below), and attracted mandatory exemption from VAT. However, UK legislation at all material times failed to provide for the exemption required by the VAT Directives. The Trustees and their predecessors have paid VAT to their suppliers which should not properly have been payable. The Trustees claim that they have a directly effective right to exemption with a consequential right to recover from HMRC the VAT which they should never have been obliged to pay.

HMRC’s primary case is that the supplies by the Non-Insurers were not insurance transactions within the meaning of the VAT Directives and did not attract exemption under those Directives; the supplies were therefore properly standard rated and VAT was correctly paid in accordance with UK law. If that is wrong, their case is that the Trustees have no right to recover directly from HMRC the VAT which was paid to the Non-Insurers.

It is necessary to refer to a few provisions of a number of Directives relating to insurance and to VAT. I start with the insurance Directives.

The First Council Directive (direct insurance other than life assurance) (73/239/EEC) (“the First non-life Directive”). The focus of the First non-life Directive is on the classification of risks in different classes of insurance, as can be seen from the third recital. Article 1 explains that it concerns the activity of direct insurance carried on by insurance undertakings in “the classes of insurance defined in the Annex”. The Annex sets out a classification of risks according to classes of insurance.

Article 2 describes a number of activities to which the Directive does not apply and which are specified in Articles 2(1) and (2). Article 2(1) specifies “The following kinds of insurance” which I can briefly describe as Life Assurance, Annuities, personal injury insurance, insurance forming part of a statutory system of social security and permanent health insurance. Within the meaning of “Life assurance” is included “tontines”. Article 2(2) specifies a number of other “operations” which do not fall within the scope of the Directive.

Article 6 requires each Member State to make the taking-up of the business of direct insurance subject to an official authorization. Article 7 makes provision in relation to such authorization:

“1.

An authorization shall be valid for the entire national territory unless, and in so far as national laws permit, the applicant seeks permission to carry on his business only in a part of the national territory.

2.

Authorization shall be given for a particular class of insurance. The classification by class appears in the Annex. Authorization shall cover the entire class unless the applicant wishes to cover only part of the risks pertaining to such class. The supervisory authorities may restrict an authorization requested for one of the classes to the operations set out in the scheme of operations referred to in Articles 9 and 11.

3.

Each Member State may grant an authorization for two or more of the classes, where its national laws permit such classes to be carried on simultaneously.”

The Directive contains no further definition or exposition of the meaning of “insurance”.

The First Council Directive (direct life insurance) (79/267/EEC) (“the First life Directive”). The second recital explains that “a classification by class of insurance is necessary in order to determine, in particular, the activities subject to compulsory authorization”. Article 1 explains that the Directive concerns “direct insurance” in the form of the activities subsequently defined. Those activities are described in Articles 1(1), (2) and (3).

Article 1(1) relates to the “following kinds of insurance where they are on a contractual basis”. Put briefly, these are:

The opening words of Article 1(2) are as follows:

“The following operations, where they are on a contractual basis, in so far as they are subject to supervision by the administrative authorities responsible for the supervision of private insurance and are authorized in the country concerned”

There then follow 5 sub-paragraphs (a) to (e). Paragraphs (a) and (c) provide as follows:

“(a)

tontines whereby associations of subscribers are set up with a view to jointly capitalizing their contributions and subsequently distributing the assets thus accumulated among the survivors or among the beneficiaries of the deceased;”

“(c)

management of group pension funds, i.e. operations consisting, for the undertaking concerned, in managing the investments, and in particular the assets representing the reserves of bodies that effect payments on death or survival or in the event of discontinuance or curtailment of activity;”

The Annex is headed “Classes of insurance” and sets out the various classes. They include Class V, tontines, and Class VII, management of group pension funds.

The First non-life Directive has been added to by Council Directive (direct insurance other than life insurance) (84/641/EC) but nothing turns on that for present purposes. There have also been subsequent life insurance Directives, ending with the consolidation measure Directive 2002/8/EC of the European Parliament and the Council. It is not necessary to refer to the provisions of those Directives which make no material change to the provisions to which I have referred The legislation was updated and consolidated for both life and non-life insurance by Directive 2009/138/EC (insurance and reinsurance) which makes no substantive change to the provisions which I have already noted (save that there is no definition of tontines corresponding to that in Article 1(2) of the First life Directive, although the same words are to be found in the new Article 2(3)(b)(i)).

References to the First non-life Directive or the First life Directive in this judgment are to be read as including subsequent iterations. There is no material difference in the wording of the successive Directives.

The Sixth Council Directive (value added tax) (77/388/EEC) (“the 6th Directive”). This provides under Title X for exemptions. Article 13(B) provides as follows:

“Without prejudice to other Community provisions, Member States shall exempt the following under conditions which they shall lay down for the purpose of ensuring the correct and straightforward application of the exemptions and of preventing any possible evasion, avoidance or abuse:

(a)

insurance and reinsurance transactions, including related services performed by insurance brokers and insurance agents;

…”

Council Directive 2006/112/EC(common system of value added tax) (the Principal VAT Directive or “the PVD”). Article 135 of the PVD deals with exemptions from VAT. Article 135(1)(a) is in materially the same terms as Article 13(B)(a) of the 6th Directive (with the omission of the opening words of Article 13(B)). It provides that Member States

“shall exempt……insurance and reinsurance transactions, including related services performed by insurance brokers and insurance agents”

In this judgment, I use “Article 135(1)(a)” to refer to both that Article of the PVD and to Article 13(B)(a) of the 6^th Directive.

There is no definition of “insurance” or “insurance transaction” in either the 6th Directive or the PVD.

Mr Scorey and Mr Jones have helpfully included a Schedule to their skeleton argument in which they set out the relevant provisions over time of the UK VAT and insurance legislation. I do not intend to set out the detail in this judgment. It is enough to mention a few provisions.

I can start with the exemption from VAT for insurance found in Item 1 Group 2 Schedule 5 Finance Act 1972 which exempted “Insurance”.

This was revised by the VAT (Insurance) Order 1977 with effect from 1 January 1978 until 31 December 1981. This exempted:

“The provision of insurance and reinsurance by authorised insurers within the meaning of sections 2 to 9 of the Insurance Companies Act 1974”

Sections 2 to 9 Insurance Companies Act 1974 dealt with who was authorised to carry on in Great Britain “insurance business” of a type specified in section 1(1) of the Act. I do not need to set out the classes.

The exemption was therefore subject to two requirements, the first dependent upon the nature of the supply, namely insurance/reinsurance, and the second dependent upon whether the supplier was authorised under the Insurance Companies Act 1974. The UK, it appears from the judgment of the CJEU in Card Protection Plan Ltd v Customs and Excise Comrs (Case C-349/96), [1999] 2 AC 601 (“CPP”)at [34], appears to have considered that the second requirement was authorised by the introductory words to Article 13(B) of the 6th Directive. That was rejected by the Court.

With effect from 1 January 1982, the wording of the VAT exemption was revised (by the VAT (Insurance) Order 1981) to read relevantly:

“The provision of insurance and re-insurance by

(a)

persons permitted, in accordance with section 2 of the Insurance Companies Act 1981, to carry on insurance business…”

So far as concerns the Insurance Companies Act 1981 (“ICA 1981”):

Pension fund management services were, it can be seen, described by ICA 1981 as a class of insurance business, thus reflecting the First life Directive.

The position in relation to VAT remained, in principle, unchanged under successor legislation (including the Value Added Tax Act 1994 (“VATA”)) until 2005, in particular under Item 1 Group 2 Schedule 9 VATA (including its amendment by Finance Act 1997). The VAT insurance exemption was given only to insurers. And pension fund management under the insurance companies Acts continued to be described as a class of insurance business; as a class of insurance, it was expressly deemed to be insurance and therefore exempt. The reforms under the Financial Services and Markets Act 2000 resulted in a radical restructuring of the scheme of regulation but the position remained that VAT exemption was given only to insurers; and pension fund management was treated as a contract of insurance.

As from 1 January 2005, the VAT (Insurance) Order 2004 replaced Items 1 to 3 Group 2 Schedule 9 VATA with the following exemption:

“1.

Insurance transactions and reinsurance transactions”

In summary, the legislation prior to 2005 expressly limited the exemption to insurance supplied by an insurer. The 2005 amendments removed the requirement that the supply had to be by an insurer: this was obviously in response to the decision of the CJEU in CPP. It is to be noted, however, that the requirement was removed simply in relation to “insurance transactions” within the 6th Directive. The amendments had nothing to say about whether pension fund management services were within the scope of insurance transactions.

It is also necessary to mention sections 80 and 80A VATA which deal with credit for and repayment of overstated or overpaid VAT and arrangements for reimbursing customers. Section 80 has been through different versions and section 80A was only introduced by Finance Act 1997. For the purposes of the present proceedings, it is sufficient to refer to the provisions as they now stand. Section 80 provides, so far as material, as follows:

“80.

Credit for, or repayment of, overstated or overpaid VAT

(1)

Where a person –

(a)

has accounted to the Commissioners for VAT for a prescribed accounting period (whenever ended), and

(b)

in doing so, has brought into account as output tax an amount that was not output tax due,

the Commissioners shall be liable to credit the person with that amount.

(2)

The Commissioners shall only be liable to credit ... an amount under this section on a claim being made for the purpose.

(2A) Where –

(a)

as a result of a claim under this section by virtue of subsection (1) or (1A) above an amount falls to be credited to a person, and

(b)

after setting any sums against it under or by virtue of this Act, some or all of that amount remains to his credit,

the Commissioners shall be liable to pay (or repay) to him so much of that amount as so remains.

(3)

It shall be a defence, in relation to a claim under this section by virtue of subsection (1) or (1A) above, that the crediting of an amount would unjustly enrich the claimant.

(4)

The Commissioners shall not be liable on a claim under this section –

(a)

to credit an amount to a person under subsection (1) or (1A) above, or

(b)

to repay an amount to a person under subsection (1B) above, if the claim is made more than 4 years after the relevant date.

(7)

Except as provided by this section (and paragraph 16I of Schedule 3B and paragraph 29 of Schedule 3BA) , the Commissioners shall not be liable to credit or repay any amount accounted for or paid to them by way of VAT that was not VAT due to them.”

Section 80A concerns “reimbursement arrangements”. These are arrangements made for the purposes of a claim under section 80 which are (a) made by any person for the purposes of securing that he is not unjustly enriched and (b) provide for the reimbursement of persons who have for practical purposes borne the whole or any part of the amounts brought into account under, among other matters, section 80(1)(b). Such arrangements will eliminate HMRC’s defence under section 80(3). However, HMRC may make provision by regulation for such arrangements to be disregarded in certain cases. In this way, abuse of the provisions for reimbursement is avoided.

It is to be noted that the person liable to account to HMRC for VAT charged on a supply is the supplier: a customer has no liability to HMRC and his liability to the supplier rests in contract: see [5] of ITC (SC), the important decisionreferred to at paragraph 166 below.

For the purposes of the trial of liability before me, the facts and procedural history can be stated very shortly:

The issues to be resolved by me are as follows:

The issue here is whether the pension fund management services provided by the Non-Insurers fell within the exemption under Article 135(1)(a) as “insurance transactions”. In summary, HMRC say that these services are not insurance transactions: it is a central feature of any such transaction that an element of risk is present. Here, the Non-Insurers undertook no risk and simply provided investment services to the Trustees. The nature of the service was clearly not that of an insurance transaction any more than any other investment service provided other than in the context of the management of group pension funds.

In summary, the Trustees submit that the meaning of “insurance” should be transposed across from the First life Directive under which the services in question are a class of “insurance”. This should be done even though the services were provided by entities which were not authorised insurers. Further or alternatively, a person who is an authorised insurer would be entitled to exemption; the principle of fiscal neutrality requires the same exemption to be given to persons who are not so authorised, such as the Non-Insurers.

Before turning to the rival submissions in more detail, I make some observations in relation to the Directives which I have mentioned.

So far as concerns Article 1 of and the Annex to the First non-life Directive, it is to be noted that all, or nearly all, of the activities listed are within the concept of what an English lawyer would regard as insurance in the sense that each activity carries a risk for the insurer. I say “nearly all” because item 15, “Suretyship”, might not be seen as an ordinary insurance transaction, although it certainly carries an element of risk.

Article 2 (which specifies a number of activities to which the Directive does not apply) draws a distinction between kinds of insurance, dealt with in Article 2(1), and other “operations”, dealt with in Article 2(2). The operations described in Article 2(2) are largely outside the concept of what an English lawyer would regard as insurance. The Directive does not appear to regard them as insurance either: whilst it excludes them from its scope, it does not describe them as a kind of insurance but simply as operations. Given the scope of Article 1, it is not easy, in any case, to see how the activities described in Article 2 might fall within Article 1 unless excluded. But no doubt Article 2 brings further clarity to and emphasis on the types of insurance with which the Directive is concerned. The exclusion of these operations brings clarity about that scope and precludes an argument that, in one or more Member States, such an operation might be regarded as insurance.

In contrast, all of the subparagraphs of Article 2(1) appear to me to constitute “insurance” as the word is understood by English lawyers and no doubt under EU law as well. Subject to one possible exception, the activities described in those subparagraphs are all activities which carry an element of risk for the provider of the relevant product. This includes annuities: annuities clearly carry a risk for the provider which can properly be seen as an insurance risk. The exception is one of the items which appears in paragraph (a). I should set that out in full:

“Life assurance, that is to say, the branch of insurance which comprises, in particular, assurance on survival to a stipulated age only, assurance on death only, assurance on survival to a stipulated age or an earlier death, life assurance with return of premiums, tontines, marriage assurance, and birth assurance”

Mr Scorey submits that a tontine carries no element of risk to the provider: it is simply an investment vehicle under which the last survivor of the group scoops the pool. Accordingly, he suggests that the concept of life assurance (one of the “kinds of insurance” mentioned in the opening words of Article 2(1)) in the Directive includes an item which is not insurance in the sense of carrying a risk. The relevance of risk to each side’s arguments will become apparent later.

I now come to the First life Directive. It is to be noted that the opening words of Article 1 refer to the activity of “direct insurance … in the form of the activities defined below”. Those activities include not only the kinds of insurance referred to in Article 1(1), but also those referred to in Articles 1(2) and (3). Further, Article 7(2) provides for authorisation to be given for “a particular class of insurance”, the classes appearing in the Annex. Those provisions lend support to the view, advocated by Mr Scorey, that for the purposes of the First life Directive, the activities described in Articles 1(2) (and, in particular, tontines and management of group pension funds) and the corresponding classes in the Annex constitute “direct insurance” and “insurance”.

It is to be noted that Article 1(2) applies to the activities mentioned only where they are carried on by an authorised person: a person who does not provide any of the kinds of insurance specified in Article 1(1) or the operations specified in Article 1(3) is not required to be (indeed cannot be) authorised by Article 6. The activities themselves may or may not amount to insurance as that word is ordinarily to be understood in EU law: as has been pointed out, none of the Directives with which this case is concerned contains a definition of insurance or insurance transaction. But if those activities do amount to insurance as ordinarily understood, it might be thought that there should be compulsory authorisation for them just as much as for the activities mentioned in Articles 1(1) and 1(3). That there is no such requirement is an indication that the activities would not ordinarily be understood in EU law as constituting insurance just as an English lawyer would not regard them as constituting insurance however expansive a meaning of the word one might adopt. Accordingly, the First life Directive can be seen as providing as follows:

As I have already observed, there is no definition of “insurance” or “insurance transaction” in the First non-life Directive (and its later iterations) or in the 6th Directive and the PVD; and there is no clear explicit definition in the First life Directive and its later iterations either. Although different Member States may have different understandings of such concepts as a matter of domestic law, the CJEU has given consideration to what amounts to insurance as a matter of EU law in a number of cases. These show, according to HMRC, that there is, at a European level, an “ordinary” meaning of insurance. This meaning informs the meaning of “insurance” and “insurance transaction” in determining what those words mean in a particular context, such as the First non-life Directive, the First life Directive, the 6th Directive and the PVD.

HMRC rely on the decision of the CJEU in Försäkringsaktiebolaget Skandia (C-240/99) [2001] 1 WLR 1617 (“Skandia”). Before I come to that case, however, I wish to consider the earlier case of CPP. This case is well-known and frequently referred to in relation to the question whether a transaction gives rise to a single supply or a two or more distinct supplies. It is less frequently referred to for what it has to say about the VAT insurance exemption. Taking the facts from the headnote, the plaintiff in that case (“CPP”) was not an insurance company but operated a plan whereby, in return for the payment of a sum of money, customers were indemnified against certain types of costs and expenses which could be incurred in the event of the loss or theft of their credit cards and other possessions such as car keys. The risks indemnified against corresponded to the cover in a block insurance policy taken out by CPP with an insurance company naming CPP’s customers as the insureds. One of the questions referred, as interpreted by the Court, was whether Article 13(B)(a) of the 6th Directive is to be interpreted as meaning that supplies of services such as those described in the plan which CPP provided to its customers constituted insurance transactions or related services of insurance agents.

The Advocate General, at [26]ff of his opinion, considered the Community law notion of insurance. He stated, at [27], that in the absence of a definition of “insurance” or “insurance transaction”, regard should be had to Community legislation regarding insurance, concluding that “for the purposes of the exemption for insurance transactions expressed in Article 13(B)(a) [of the 6th Directive]…, the term “insurance” should be interpreted co-terminously with the scope of the insurance Directives for the time being in force”. And at [34] he restated that conclusion. He then went on to say this:

“The essentials of an insurance transaction are, as generally understood, that one party, the insurer, undertakes to indemnify another, the insured, against the risk of loss (including liability for losses for which the insured may become liable to a third party) in consideration of the payment of a sum of money called a premium: it is the giving of the indemnity that constitutes the insurance and, thus, the supply of the service.”

That passage was expressly referred to and, in effect, adopted by the Court at [17] of its judgment:

“… the essentials of an insurance transaction are, as generally understood, that the insurer undertakes, in return for prior payment of a premium, to provide the insured, in the event of materialisation of the risk covered, with the service agreed when the contract was concluded.”

Reliance is placed by Mr Scorey on the last sentence of [18] which, he submits, supports a read-across from the First life Directive to the 6^th Directive and the PVD of the meaning of “insurance”. Although read in isolation there might be something in that submission, the sentence must be read in its context. I set out the paragraph in full:

“18.

It is not essential that the service the insurer has undertaken to provide in the event of loss consists in the payment of a sum of money, as that service may also take the form of the provision of assistance in cash or in kind of the types listed in the annex to Directive 73/239 as amended by Directive 84/641. There is no reason for the interpretation of the term "insurance" to differ according to whether it appears in the Directive on insurance or in the Sixth Directive.”

The last sentence is concerned with the manner in which the risk referred to in [17] is met when it materialises. It does not assist either way in determining whether or not a transaction which does not involve risk is an insurance transaction.

It is, however, important to note that the reference to the Court for a preliminary ruling was made, and the question was answered, in the context of an underlying insurance transaction which fell within the non-life Directives and not within the life Directives. The non-life Directives all concerned classes of insurance which carried risk; it was in that context that the Advocate General and the Court said what they did in [34] (A-G) and [17] (Court).

The actual decision in CPP was that the services in question were indeed “insurance transactions” notwithstanding that CPP did not provide or undertake to provide insurance cover: it merely promised its customers to do what was necessary for insurance to be provided to them by third parties. That was done through CPP’s block insurance under which its customers were insured. As the Court put it at [17] of its judgment:

“… It procures for those customers, for payment, in its own name and on its own account, to the extent of the services mentioned in the Continental policy, insurance cover by having recourse to an insurer. Consequently, for the purposes of VAT, there is a supply of services between Continental and CPP on the one hand, and between CPP and its customers on the other, and the fact that Continental under the terms of its contract with CPP provides insurance cover directly to CPP’s customers is not material in this respect.”

Skandia concerned exemption under Article 13(B)(a) of the 6th Directive. Försäkringsaktiebolaget Skandia (publ) (“Skandia”) was an insurance company with a subsidiary, referred to as Livbolaget. The relevant background is sufficiently set out in [8] of the Court’s judgment:

“Livbolaget is engaged in the business of life assurance, in particular, in the sector of capital insurance and insurance provision for old-age. Livbolaget and Skandia have studied the possibility of merging (in the broad sense) their insurance activities within a single company. One plan was to transfer Livbolaget's staff and operations to Skandia so that, in effect, Skandia would be conducting all Livbolaget's business, whether this consisted in the sale of insurance, the settlement of claims, the calculation of actuarial forecasts or capital management. In return, Skandia would receive from Livbolaget remuneration at market rates. Skandia would assume no liability in respect of those insurance activities. All risks would devolve wholly upon Livbolaget which would preserve its status of insurer for the purposes of Swedish civil law.”

The Swedish tax authorities decided that Skandia’s commitment to run Livbolaget’s business could not be regarded as the supply of insurance services for the purposes of the relevant Swedish statutory provisions. Later on, in appellate proceedings, the Swedish court referred to the CJEU for a ruling on the question whether a commitment such as Skandia proposed to give to its wholly-owned subsidiary, Livbolaget, to run its business constituted an insurance transaction or transactions within Article 13(B)(a) of the 6th Directive.

The Court answered that such a commitment was not an insurance transaction. It should be borne in mind that what the Court said must be read in the context of the facts and circumstances of the case. Although the reports do not describe the detail of the insurance provided by Livbolaget to its customers, it does appear that it was engaged in the business of life assurance: see the Advocate General’s opinion at [9]. One can assume, therefore, that the underlying business was insurance of a type which fell within Article 1(1) of the First life Directive. But even if that is not a justifiable assumption, there is nothing at all to suggest that any of the business fell within Article 1(2), in particular as a tontine or as management of group pension schemes.

In giving its judgment, the Court reiterated at [23] and [32] that it is settled law that the exemptions provided for by what is now Article 135 PVD constitute independent concepts of Community/EU law whose purpose is to avoid divergences in the application of the VAT system as between one Member State and another and must be placed in the general context of the common system of VAT; and that the terms used to specify the exemptions provided are to be interpreted strictly since they constitute exceptions to the general principle that turnover tax is to be levied on all services supplied for consideration by a taxable person.

Skandia submitted that the interpretation of the term “insurance” should not differ according to whether it appears in the Directives on insurance or in the 6th Directive, relying on CPP at [18]. It relied on Article 8(1)(b) of the First life Directive and successive iterations, which limited its business activities to those referred to in the Directive and operations directly arising therefrom. Skandia’s argument was that, since it could only carry out the activities specified in the Directive, all its transactions must by definition be exempt from VAT as insurance business.

The Court rejected those arguments. It did, however, acknowledge what was said in CPP saying at [30] of its judgment that:

“… it is true that, in paragraph 18 of its judgment in [CPP] the court held that there is no reason for the interpretation of the term “insurance” to differ according to whether it appears in the Directives on insurance or in the Sixth Directive.”

Whilst not casting doubt on that apparently general proposition, the Court went on to consider the type of business which insurance companies could carry out. It was a mistake to maintain that the limitation on the business activities of insurance companies meant that such companies effect only insurance business exempt from VAT under Article 13(B)(a) of the 6th Directive. It was in that context that the Court reiterated the established case law that the exemptions provided by Article 13 are to be interpreted strictly.

The Court then stated (see [33]) that the insurance Directives allow insurance companies to carry out “not only insurance transactions proper” but also “operations arising directly therefrom”, concluding at [34] that:

“…the fact that an insurance company must not engage in business other than insurance business or operations arising directly therefrom does not mean that all the transactions carried out by that company constitute, for tax purposes, insurance transactions in the strict sense, as referred to in article 13(B)(a) of the Sixth Directive.”

The Court then addressed the separate argument that it was unnecessary, in order to bring the case within the VAT exemption, for the transaction to be carried out by a company which has a legal relationship with the end customer, that is to say the insured. It was in that context that the Court turned to consider CPP, referring to [17] of the judgment and affirming in [37]:

“… that the essentials of an insurance transaction are, as generally understood, that the insurer undertakes, in return for prior payment of a premium, to provide the insured, in the event of materialisation of the risk covered, with the service agreed when the contract was concluded.”

That description was then relied on to reach the conclusion in [39] that an insurance transaction necessarily implies the existence of a reciprocal relationship between the provider of the insurance service and the person whose risks are covered by the insurance, namely the insured. One sees, even in the discussion of this issue (ie the need for a contractual relationship), reference to the underlying concept of insurance involving risk on the insurer.

The Court then considered the nature of the relationships found in CPP and the services provided to the insured persons by CPP, stating that a relevant legal relationship had been created between CPP, which offered insurance cover, and the insured, namely the persons whose risks were covered by the insurance. The discussion led to the following conclusions:

“40 However, it is clear that no such legal relationship would exist between Skandia and Livbolaget's clients in the context of the scheme postulated by the two companies in the main proceedings. Skandia would have no contractual relationship with persons insured with Livbolaget and would assume no liability in respect of the insurance business carried out, since all risks would devolve wholly on Livbolaget which would preserve its status of insurer for the purposes of Swedish civil law.

41 According to the definition of insurance transactions set out in paragraph 17 of the judgment in [CPP] and cited in paragraph 37 of this judgment, it appears that the identity of the person supplied with the service is relevant for the purposes of the definition of the type of services covered by article 13(B)(a) of the Sixth Directive and that an insurance transaction necessarily implies the existence of a contractual relationship between the provider of the insurance service and the person whose risks are covered by the insurance, namely the insured.”

There are two significant points arising from that judgment to which I wish to draw particular attention. The first is the contrast drawn in [33] between “insurance transactions proper” and “operations arising directly therefrom”. Mr Scorey suggests that the operations referred to in Article 1(2) are “insurance transactions proper”, lending support to his argument about the meaning of “insurance”. However, it must be noted that the underlying business of Livbolaget was life assurance: see [8] of the Judgment and [9] of the Advocate General’s opinion (and see my comments at paragraph 54 above). This business fell within Article 1(1) of the First life Directive. It was not necessary for the Court to refer to or consider Article 1(2). The only relevant contrast on the facts of the case was between insurance within Article 1(1) – appropriately referred to as “insurance transactions proper” – and operations directly arising from such insurance transactions. The reference in [34] to “insurance transactions in the strict sense” is, in its context, a reference to the “insurance transactions proper” referred to in [33]. Accordingly, these paragraphs of the judgment have nothing to say about the operations referred to in Article 1(2) of the First lie Directive. In my view, Skandia cannot be relied on to support the proposition that “insurance transactions” in the 6th Directive include such operations.

The second point is that the Court was content to adopt the CPP interpretation of “insurance transactions”. It did so in the case before it which was concerned with insurance within the First life Directive. It did so without expressly qualifying what it said by reference to any of the subparagraphs of Article 1(2). There are two possible reasons why this was so. The first is that the Court did not see the operations in those paragraphs as “insurance transactions” at all. The second is that it was concerned only with the kinds of insurance falling within Article 1(1). The first possibility leads inevitably to the conclusion that management of group pension funds does not attract the VAT exemption. The second possibility makes good the view I have expressed in the last sentence of the preceding paragraph of this judgment. I cannot assume against the Trustees that the first possibility is correct. Adopting the second possibility leaves open the argument which Mr Scorey now presents and which I must address.

In December 20105, Advocate General Kokott gave her opinion in Minister Finansów v Aspiro SA (Case C-40/15) [2016] STC 1255 (“Aspiro”). The case concerned the provision by a company, Aspiro, of claims settlement services to an insurer. It was not an insurance company nor an insurance broker nor an insurance agent. At [26] of her Opinion we find the Advocate General describing the assumption of risk, according to the case law, as “the sole constituent of an insurance transaction”.

One issue facing the Court was “whether the settlement of claims, such as that performed by Aspiro, consists of making ‘insurance transactions’ or is to be regarded as ‘related services performed by insurance brokers and insurance agents’”. The Court repeated, at [22], the essentials of an insurance transactions as stated in CPP and Taksatorringen (Case C-8/01) at [39] (a case which in turn relied on Skandia). The Court acknowledged that “insurance transactions” covers not only the transaction carried out by the insurers themselves but also by a party who procures, in the context of a block policy, cover for its customers. It then went on to say this, at [24]:

“However, in the present case, a provider of services such as Aspiro does not itself undertake to ensure that the insured person is covered in respect of a risk and is not connected in any way to the insured person through a contractual relationship.”

And so, even though the claims settlement service was an essential part of the insurance transaction, in that it included determination of liability and the amount of damage, and the decision to pay or refuse compensation to the insured person, the Court held that the service – provided moreover to the insurer – did not constitute an insurance transaction. This conclusion was supported (see [26]) by the need to interpret exemptions strictly. There were, it can be seen, two elements to the conclusion. First, that there was no undertaking by Aspiro to ensure cover for the insured and secondly, that Aspiro was not connected with the insured through a contractual relationship.

In all of these cases, CPP, Skandia, Taksatorringen and Aspiro, there was an underlying transaction which on any reasonable view constituted insurance. Certainly, in every case there was an insured who had paid a premium to someone in order to cover a risk to him which might materialise. It was in that context that the Court answered the different issues with which it was faced. It did not address, and did not need to address, the question whether operations within Article 1(2) of the First life Directive were properly to be described as a kind of insurance or a class of insurance. Nor did it address, or need to address, the argument now addressed to me by Mr Scorey.

Having examined in some detail the Directives and the relevant case-law, I return to Mr Scorey’s argument, which can now be put more clearly in context. In essence, I understand his argument to be as follows. Management of group pension funds is “insurance” within the meaning of “direct insurance” in the opening words of Article 1 of the First life Directive and “a class of insurance” within Article 7(2) and the Annex. The case law of the CJEU mandates an approach under which the meaning of “insurance transactions” in the VAT Directives should be coterminous with the meaning of “insurance” in the insurance Directives. Management of group pension funds falls within the insurance Directives and therefore attracts VAT exemption.

The interpretation of “insurance” given in CPP and followed in subsequent cases provides, in his submission, no obstacle to this conclusion. My summary of his reasons for saying that is as follows:

Dealing with that last point first, I do not think that there is anything in it. Under the CPP interpretation there needs to be prior payment of a premium, a risk undertaken by the insurer and provision of the service agreed in the event of the materialisation of the risk covered. In my view, annuity business falls within that interpretation. Under a typical simple annuity, the insurer – the annuity provider – receives a payment or payments from its customer – the annuitant – in return for which it contracts to make a series of payments to the annuitant from a certain age until death. The payment to the insurer is properly to be seen as the premium and the commitment to make the annuity flow and subsequent provision of the annuity flow are properly to be seen as the provision of the service. The annuity flow is provided in the event of the materialisation of the risk covered, namely that the annuitant will be alive on the date for due payment of the instalment of the annuity. Even if Mr Scorey is correct that, as a strict matter of language, an annuity does not fall within the CPP description of insurance, it is beyond doubt, in my view, that the risk which the annuity provider undertakes in return for payment from the annuitant is nonetheless properly to be seen as insurance under EU law and, in particular, as insurance for the purposes of the First life Directive and as falling within the meaning of “insurance transactions” in the 6th Directive and the PVD.

As to the remainder of Mr Scorey’s argument as summarised above and which he developed in detail, I do not accept it.

First, it is only the express inclusion of management of group pension funds in the First life Directive which gives rise to the argument. If Article 1(2)(c) of the First life Directive had not been included, I do not consider that there could be the slightest argument that such activities amounted to insurance or insurance transactions. There is nothing to suggest that there would be any doubt at an EU level – for instance in the interpretation of the 6th Directive or the PVD – about that; it has not been suggested that in any Member State such activities would be regarded as insurance.

Secondly, to make a point which I have already identified, if management of group pension funds were insurance as ordinarily understood, it is at least curious that it was not included within Article 1(1) of the First life Directive. The explanation for that is as follows. This activity, on its own, was not to attract mandatory regulation of the service provider as an insurer: a provider who did not provide life insurance as ordinarily understood or the other services specified in Article 1(1) did not require regulation. But an insurer which, in addition to its main insurance business, wished to provide a service of managing group pension funds, was to be entitled to do so and that activity would fall to be regulated along with its main business. The clear inference, it seems to me, is that the activity of management of pensions funds did not, of itself, amount to insurance.

Mr Scorey also contends that management of group pension funds is the relevant activity which constitutes insurance: the words in Article 1(2) of the First life Directive “in so far as they are subject to supervision etc” do not circumscribe or form part of the definition of the activity: rather, the relevant insurance activity is carrying out management of group pension funds, the quoted words going only to the need for regulation.

I do not agree with this contention either. The First life Directive tells us nothing about whether these activities carried out by a non-insurer constitute insurance or not. It would not, in Community/European law at a general level, be a proper use of language to describe the activities as “insurance” or “insurance transactions”. That is not to say that it is an impermissible drafting approach to treat them as such or to include those activities in a portmanteau description as “insurance” as was done in the First life Directive. In my view, the management of group pension funds becomes “insurance” for the purposes of the First life Directive if but only if it is carried out by a person who is authorised within the contemplation of Article 1(2). If a person is not an authorised insurer, he does not, simply by carrying out these activities, carry on insurance or effect insurance transactions.

Thirdly, Mr Scorey’s argument places more weight than it can bear on the inclusion of tontines in Article 2(1) of the First non-life Directive. In that Directive, tontines appear as part of the definition of Life assurance. In the First life Directive, however, tontines are removed from that definition (and thus from a “kind of insurance” within Article 1(1)): they are not seen as life assurance. They fall within the Directive under the umbrella of insurance only as “operations” and then only if carried out by an insurer.

It is, in any case, not clear what the First non-life Directive is referring to when it uses the word “tontines”. Although there is a definition of “tontines” in the First life Directive, there is no definition in the First non-life Directive. It is not clear that “tontines” has the same meaning in both.

The historic concept of a tontine in English law is effectively a group annuity with the share of a person who dies being reallocated among the remaining contributors. Capital is never paid back so that the provider is exposed to risk in the same way as the provider of an ordinary annuity. One sees this reflected in the definitions found in legal dictionaries such as Jowitt’s Dictionary of English Law 4th ed. (which I include for its interesting historical references, including the adoption of this sort of tontine in France):

“A financial arrangement under which subscribers to a loan each receive for life an annuity, which increases as other subscribers die, so that ultimately the last survivor receives for whatever is left of his life an annuity amounting in the aggregate to the total of the original annuities. Lorenzo Tonti, an Italian, invented this kind of security in the 17th century, when the governments of Europe had some difficulty in raising money in consequence of the wars of Louis XIV, who first adopted the plan in France. The system was used on several occasions during the 18th century by the British Government. The last Act passed for the purpose was the statute 1789, 29 Geo. 3, c.41, entitled an Act for raising a certain sum of money, by way of annuities, to be attended with the benefit of survivorship, in classes.”

The definition of “tontine” in Article 1(2)(a) of the First life Directive is significantly different. That definition is concerned with operations where an association is set up not only with a view to capitalising the members’ contributions but also with a view to subsequently distributing the accumulated assets to the survivor or among the beneficiaries of the deceased. Such arrangements may or may not carry risk for the provider.

It is not at all clear to me whether the historic type of tontine which I have described falls within the express definition in the First life Directive: as a matter of language, it does not appear to do so. Nor is it clear precisely what is covered by the word “tontines” in the definition of Life assurance in the First non-life Directive. It would be surprising if it did not include the historic type of tontine since that is the one sort of tontine which does carry risk for the provider unless, perhaps, this type of tontine falls independently within the annuities class. The definition of Life assurance in the First non-life Directive lists a number of types of business, each of which (apart from tontines in the sense defined in the First life Directive) carries an element of risk and each of which can sensibly be described as insurance (whether at an EU level or even in English law). It would be entirely unsafe to conclude that, because tontines are mentioned, something which does not carry risk to the provider can be within a concept of insurance to be carried across to the VAT Directives.

It follows from the above discussion, in my judgment, that the management of group pension funds does not fall within the exemption provided by Article 135(1)(a).

Support for that conclusion can be gained from the timing of the relevant Directives. The First non-life Directive came into effect in July 1973, the 6th Directive in 1977 and the First life Directive in 1979. There is no support for the conclusion that, immediately after July 1973, the activity of management of group pension funds would have been seen as a type of insurance within Article 2 of the First non-life Directive. It certainly did not fall within any of the subparagraphs of that Article and nor is there anything in the case law of the CJEU which supports such a conclusion. It is true that there is no case law to the contrary either. But nor is there case law to the effect that managing a private portfolio is not an insurance activity when manifestly it is not: it does not need an authority for that to be clear.

The same can be said for tontines in the sense which they came to be subsequently defined in the First life-Directive. I do not think that it could sensibly have been suggested in, say, 1978, that a tontine in that sense fell within the meaning of insurance transactions in the 6th Directive.

That was the position in relation to both management of group pension funds and tontines when the 6th Directive was promulgated in May 1977. Whatever UK law may have provided, Article 13(B)(a) of the 6th Directive did not exempt the management of group pension funds at any time before the First life Directive was promulgated on 5 March 1979. The position prior to 5 March 1979 was that exemption was not available under Article 13(B)(a) in respect of such services; that was, in my view, acte clair.

It is unlikely that the European legislator intended, by implementing the First life Directive, to alter the scope of the VAT exemption under Article 13(B)(a) of the 6th Directive. Exemptions are an exception from the general order and are strictly interpreted. The influence of the insurance Directives on the meaning of “insurance transaction” in the 6th Directive is but one factor in the overall interpretation exercise. It is true that the First non-life Directive tells us very little about the meaning of insurance in a life context. The case law which mandates to a greater or lesser extent a carry-across of meaning from the First non-life Directive is not necessarily conclusive that the risk element which those cases establish in the case of non-life business is applicable to life business.

But what Mr Scorey proposes in the present case is far more radical: it is to treat something which was not insurance and not exempt prior to the First life Directive as becoming insurance and therefore exempt as a result of that Directive. I do not consider that the principle of interpretation which requires regard to be paid to the meaning of insurance in the insurance Directives leads to the conclusion that the management of group pension funds falls within “insurance transactions” in the 6th Directive and the PVD. My reasons for reaching that conclusion are explained above. The most important reason is that such management is not regarded by the insurance Directives as insurance when carried out by a Non-Insurer: indeed, the position is quite the reverse since it is to be inferred from the opening words of Article 1(2) of the First life Directive that the activity becomes insurance for the purposes of that Directive (if it becomes insurance at all) only where it is carried out by an Insurer. The VAT exemption, at least in so far as concerns non-insurers, is not therefore available.

I do not consider that the principle of fiscal neutrality can be relied on to cast any doubt on this conclusion at an EU level. According to that principle, supplies of goods or services which are identical or similar from the point of view of the consumer, and are therefore in competition with each other, must be taxed in the same way: see for instance, Rank Group plc v HMRC (Joined cases C-259/10 and C-260/10) [2012] STC 23, at [32]. Indeed, far from casting doubt, considerations of fiscal neutrality support my conclusion. To achieve consistency between the treatment, at an EU level, of such supplies by Insurers and Non-Insurers, the preferable course is surely to deny exemption to a supply which can only be considered to be insurance at all because of the treatment afforded to such supplies by the First life Directive and which, absent that treatment, could not be regarded as insurance transactions in the first place.

If this approach is correct, as I think it is, then it follows that management of group pensions funds is not an VAT-exempt activity, whether carried out by Insurers or Non-Insurers.

I should add that, in my judgment, the Trustees cannot rely on the principle of fiscal neutrality in order to obtain exemption on supplies by Non-Insurers, which were standard rated under UK law, on the basis that similar supplies by Insurers were exempt under UK law (even assuming that the actual supplies were similar, which is a matter of contention between the parties).

In the first place, the scope of the VAT exemption is governed exclusively (and exhaustively) by EU law. The exemption is not only one which a Member State must provide but is also one which a Member State cannot extend. The position is, or may be, depending on the facts, different where a Member State is left with an element of discretion as to the implementation of a Directive. This can be seen from, for instance, the exemptions at issue in Rank(concerning Article 13B(f) of the 6th Directive and Article 135(1)(i) PVD: (see the judgment at [37] to [51])) and JP Morgan Fleming Claverhouse Investment Trust plc and The Association of Investment Trust Companies v HMRC (“Claverhouse”) [2007] ECR I-5517(concerning Article 13B(d)(6) of the 6th Directive and Article 135(1)(g) PVD: see the judgment at [38] to [47]). I do not propose to lengthen this judgment yet further by citation of those paragraphs.

In the second place, the Trustees’ complaint is, necessarily, that the Insurers were given a different treatment from that of Non-Insurers by being treated as exempt, a treatment which was unlawful as a matter of EU law. However, if a Member State wrongly affords exemption to one supplier, as a matter of domestic law or practice, another supplier cannot invoke the principle of fiscal neutrality to obtain the same treatment. The Trustees cannot, therefore, have or assert any EU law right (by any route) to have supplies to them by Non-Insurers treated in a manner that is itself unlawful under EU law.

In the draft of this judgment sent to the parties for correction in the usual way, I stated that these conclusions made it unnecessary to consider whether, on the facts, the supplies made by the Insurers and the Non-Insurers were the same or similar (I will refer to this question as the “same or similar” issue), which, as I have said, is a matter of contention between the parties. I stated that I did not propose to address the issue further. The Trustees have urged me to reconsider that decision. They express concern that, if such findings were needed later on, then the only course would be for the case to be remitted in order that they could be made, which they perceive as having obvious drawbacks. I sought HMRC’s view on the Trustees’ request; and I asked the Trustees to identify precisely what findings of fact they wished me to make. Each side has made written representations.

HMRC’s position is that issue is irrelevant in the light of my conclusions on Issue 1 and that it is not necessary to consider the merits of the Trustees’ case on the “same or similar” issue. The Trustees maintain their request that I make findings of fact.

It is common ground that the “same or similar” issue is an aspect of EU law relating to fiscal neutrality, which is to be tested from the viewpoint of the hypothetical typical consumer; and that what is required is an objective assessment of whether the two different supplies are the same or similar.

The Trustees do not ask me (contrary to my initial impression of what I was being asked to do) to carry out such an assessment but they do ask that the primary facts relating to the exercise be found. Somewhat surprisingly, the submission on behalf of the Trustees seeks to introduce two additional documents to bolster their case. The apparent justification is that these documents were not included in the trial bundle because a distinction now drawn by HMRC between active investment (carried out on their case only by Non-Insurers) and passive investment (carried out on their case only by Insurers) was not mentioned until Mr Macnab’s closing oral submissions, notwithstanding that the documents had been provided to HMRC in disclosure.

HMRC object to the late introduction of these documents and reject the justification for their late introduction. They correctly contend that the “same or similar” issue was and remains an issue on the pleadings. The burden is and always has been on the Trustees to prove their case. HMRC were entitled to make their submissions on the evidence which was actually before the Court.

I accept that, in a case where there has been a conflict of evidence and where the performance of a witness in cross-examination is important in the assessment of the reliability of the evidence, a judge ought ordinarily to make findings in order to avoid the need to remit the matter should the judgment be overturned on appeal. In the present case, however, it seems to me that an appellate court would be in just as good a position as I am to deal with the “same or similar” issue. The only witness evidence came from Mrs Furst. Her witness statement is clear. The cross-examination was not extensive and did not take matters much further. There is no issue of credibility, only of the relevance of her evidence (including her answers in cross-examination) to the underlying issue. Otherwise, each side relies primarily on the documentary material. It seems to me, in these circumstances, that there is little point in my making findings of fact unless I go on to decide the “same or similar” issue, especially given that there is no hard line in the present case between the suggested primary facts and the conclusion on the “same or similar” issue.

Essentially, the “same or similar” issue would turn on the submissions on the documentary evidence and the relevance of Mrs Furst’s evidence, about her own subjective perceptions, to the ascertainment of the objective assessment from the view point of the hypothetical typical consumer. An appellate court will be in as good a position as I am to deal with the issue. Indeed, an appellate court addressing the “same or similar” issue would be presented with all of the arguments which I have already heard (and doubtless further arguments as well) and might well wish that it was able to make the relevant findings of fact itself in the light of such developed arguments. Further, I do not think that the “same or similar” issue can arise at all: this is because I consider it to be clear beyond argument that a person cannot rely on the principle of fiscal neutrality to obtain an exemption which the 6^th Directive and the PVD do not provide for. Thus, the “same or similar” issue cannot arise if I am right Issue 1. But if I were wrong on Issue 1, then the Trustees do not need to rely on the principle of fiscal neutrality in the first place: they would have a directly effective right to rely on the exemption. In these circumstances, I decline to follow the Trustees’ request and decline to make any of the findings sought.

My answer to Issue 1 is No: as a matter of EU law, the pension fund management services provided by Non-Insurers were not exempt from VAT during the relevant period.

The question then arises whether that answer is acte clair and, if not, whether a reference is needed. Mr Scorey submits, of course, that it is not acte clair. He points out that the UK legislation expressly limited exemption in respect of insurance transactions to Insurers until 2005, when the requirement to be an insurer was removed following the decision of the CJEU in CPP. Exemption from VAT was in fact afforded to Insurers in respect of pension management services both before and after the change in legislation. It appears that this exemption was afforded pursuant to the insurance exemption on the basis that the services did, contrary to my conclusion, constitute insurance transactions.

HMRC’s most recent statement of their position is found in their Revenue and Customs Brief 3 (2017) “VAT - treatment of pension fund management services” published on 5 October 2017, shortly before the hearing in this case. That states, under the heading “Background”, HMRC policy as allowing all pension fund management services provided by insurance companies to be exempt: “This treatment arises from the UK’s original application of the insurance exemption to all of an insurer’s regulated insurance activities, including management of pension funds”. The Brief then mentions the decision of the CJEU in CPP, which, according to the Brief, “makes it clear that the EU insurance exemption applies only to the underwriting of risk and doesn’t apply to other supplies made by insurers”. Before the decision in CPP, HMRC’s policy to exempt insurers was purportedly justified by the regulatory regime under which insurance companies operated. Revenue and Customs Brief 3 (2017) explains that this policy was continued after CPP: “UK policy continued to allow insurers to exempt their supplies of pension fund management services”.

The Brief does not explain, and the evidence before me does not establish, whether that was a policy continued notwithstanding a view then held that CPP made it clear that the exemption applies only to the underwriting of risk or whether it is simply that HMRC now consider that that conclusion is clear. HMRC recognised as long ago as 2002, in their draft consultation paper “Consultation on the VAT Treatment of Pension Fund Management”, that the decisions in CPP and Skandia sat uneasily with the then policy of regarding pension fund management as insurance when it is provided by an insurer, but that is a slightly different point from the identification of what can properly be regarded as insurance in the first place. Whether the position concerning the underwriting of risk is now acte clair or not cannot, of course, depend on what HMRC did or did not think. But there is a point to be made that, if the taxing authority itself did not appreciate the true position, then the matter cannot be said to be clear. Quite apart from that, Mr Scorey says that the position in other Member States is not known (not known in the sense that it has not been looked into and that no evidence has been put before the court). It may be that Member States generally would benefit from clarification by the CJEU.

If the matter is acte clair, I do not, strictly, need to go on to consider Issues 2 to 4. The answer on Issue 1 would be conclusive against the Trustees. If the matter is not acte clair, it does not follow that a reference has to be made. A reference is required only if I need an answer in order to be able to determine the case before me. The case before me is whether the Trustees are entitled to VAT exemption on the provision of the relevant services to it by Non-Insureds. In order to answer that question, it may not be necessary to know for certain the answer to Issue 1. If I were to decide Issue 2 in favour of HMRC, then that answer would be conclusive against the Trustees, making a reference unnecessary. Further, even if I were to decide that Issue 1 is acte clair, there would remain the possibility that the Court of Appeal would take a different view, in which case my decision on Issues 2 to 4, which have been fully argued, would be helpful. I therefore propose to deal with Issue 2 to 4 and to leave the question of reference hanging for the moment. My conclusion can be found at paragraph 245 below.

Issue 2 is whether it is “impossible or excessively difficult” as a matter of EU law for the Trustees to obtain reimbursement of the wrongly charged VAT from the Non-Insurers, the suppliers of pension fund management services. This issue arises only if I am wrong on Issue 1. There are two main limbs to this Issue. The first is whether the Trustees have any claim at all against the Non-Insurers; the second is whether, if they do have a claim, it would be impossible or excessively difficult to obtain reimbursement. In considering the first limb, it is necessary to examine whether the Non-Insurers themselves have any claim against HMRC, since that will have an impact on whether the Trustees have a claim against the Non-Insurers and whether they have a right directly against HMRC.

I propose to deal with the first limb as both a matter of EU law and English law before turning to the second limb.

If I am wrong on Issue 1, the Non-Insurers’ supplies to the Trustees should not have been subject to VAT but should have been exempt. VAT has been wrongly levied by UK law. It is said by the Trustees that they have a right under EU law to recover from HMRC the VAT which they should not have had to pay. They contend that their right is to recover the whole of the amount which they paid from HMRC. HMRC contend that there is no right in the Trustees to recover the whole of the VAT from HMRC; the Trustees’ primary claim is against the Non-Insurers who received the money from the Trustees with at best a claim against HMRC for the balance. Claims are, in any event, subject to time limits.

There is a considerable amount of EU authority about the recovery of taxes and duties improperly levied by a Member State. I can start with the well-known decision in Case C-199/92 Amministrazione delle Finanze dello Stato v SpA San Giorgio [1983] ECR 3595, (“San Giorgio”). The case demonstrates the broad proposition that a Member State is, in principle, required to provide for repayment of charges levied in breach of EU law (in that case health inspection charges imposed by Italy in breach of EU law on imported and exported meat), this being a consequence and complement of the rights conferred on individuals by the provisions of EU law prohibiting such charges. Since VAT is a tax governed by EU law, the San Giorgioprinciple applies as much to VAT levied in breach of EU law as it does to other taxes and charges levied in breach of EU law. This is common ground.

In the light of the submissions made, it is necessary to say something more about San Giorgio. In the first place, the claim was being asserted by the person, SpA San Giorgio, which had paid the unlawful levy. It had brought an action claiming recovery from the State Finance Administration. Italian legislation precluded recovery where the charge in question had been passed on in any way whatsoever to other persons, except in case of substantive error. The charge was presumed to have been passed on whenever the relevant goods had been transferred in the absence of documentary proof to the contrary. The question referred to the Court, as reformulated by it in the usual way, was in summary whether a Member State could make repayment of charges levied in breach of Community law conditional upon proof that those charges had not been passed on to other persons:

The following points which I wish to note are made in the judgment:

To like effect are subsequent authorities including Case C-147/01 Weber’s Wine World [2003] ECR I-11365, Joined Cases C-192/95 to C-218/95 Société Comateb [1997] ECR I-165 (“Comateb”) and Case C-398/09 Lady & Kid A/S and others v Skatteministeriet [2012] STC 854 (“Lady and Kid”).

As regards the rules and conditions which a Member State may lay down, as a matter of domestic law, in respect of a claim for repayment of charges levied in breach of EU law, there is no objection, in principle, to a limitation period being imposed on such claims, provided it satisfies the principles of equivalence and effectiveness. It is established that the current limitation period in section 80(4) VATA is not objectionable from an EU law perspective.

The position of the person on whom the charge is levied in breach of EU law is clear when the charge has not been passed on. But what of the position when that person has in fact passed on the charge? National law is permitted – although not obliged – to prevent reimbursement. One question then is whether the person to whom the charge has been passed on can recover the charge himself; another question is which (if either) of the person on whom the charge is levied and the person to whom that charge has been passed on can recover the charge from the State in cases where the former has a right under national law to recover from the latter.

Some guidance can be found in three decision of the CJEU: Case C-35/05 Reemtsma Cigarettenfabriken GmbH [2007] ECR 1-2425 (“Reemtsma”), Case C-94/10, Danfoss A/S, Sauer Danfoss ApS [2013] STC 1651 (“Danfoss”) ; and Case C-564/15 Tibor Farkas v Nemzeti Adó-és Vámhivatal Dél-alfödi Regionális Adó Főigazgatósága (“Farkas”).

In Reemtsma, the taxpayer was a German company, a cigarette manufacturer, with no permanent establishment in Italy. It was provided with marketing services by an Italian company. VAT was charged in Italy by the supplier. The tax was not due and was invoiced in error. The correct position was that Reemtsma was liable for VAT in Germany under the reverse charge. The taxpayer paid the relevant VAT to the supplier. The supplier accounted for it to the Italian authorities. The German company sought reimbursement of the wrongly paid tax from the Italian authorities. Under Italian domestic law, the supplier was entitled to reimbursement of wrongly paid tax, but a recipient established in another Member State could only seek reimbursement from the supplier: see [36] of the judgment and [79] of the opinion of Advocate General Sharpston. The authorities refused to reimburse the taxpayer on the grounds that the invoices related to services which were not subject to VAT. The reference to the CJEU asked two questions, the first of which is not relevant in the present case. The second question was whether it was sufficient to comply with EU law that the recipient of services was entitled to request reimbursement of the VAT from the supplier who incorrectly invoiced the tax and who could, in turn, seek reimbursement from the tax authorities or whether such a recipient had to be able to bring a claim directly against those authorities.

In considering that question, the Court asked (i) whether the recipient of a service may be considered, generally speaking, to be the person liable for payment of VAT for the purposes of the tax authorities of the Member State where the services are supplied and (ii) whether the common system of VAT and the principles of neutrality, effectiveness and non-discrimination preclude national legislation such as that at issue in that case, which does not entitle the recipient of services to reimbursement of VAT by the tax authorities where that tax was not due, but was nevertheless paid by that recipient to the tax authorities of the Member State where the services were supplied.

The answer to first of those questions was that, except in cases expressly provided for in Article 2(1) of the 6th Directive (which is not relevant in the case before me), only the supplier must be considered to be liable for the payment of VAT for the purposes of the tax authorities of the Member State where the services were provided. In other words, the recipient of a supply is not liable to account to the tax authorities for the VAT leviable in respect of the supply. It follows from that that the obligation of the recipient of the supply to pay to his supplier VAT on the supply must be found elsewhere; in England, the obligation will arise as a matter of contract between the supplier and the customer.

As to the second question, Reemtsma argued that the principle of effectiveness meant that national legislation should not present an obstacle to the exercise of a right to reimbursement of sums paid in contravention of the applicable EU rules. The Commission argued that the scheme laid down in the Italian legislation was acceptable: Members States are free to choose the procedure which they judge appropriate for guaranteeing reimbursement provided that the principle of effectiveness is respected.

The CJEU pointed out that, subject to observing the principles of effectiveness and equivalence, it was for the domestic legal system of each Member State to lay down the conditions for applying for repayment of wrongly paid tax. The position under Italian law (in which it was the supplier who had paid VAT to the tax authorities who could seek reimbursement from them, with the recipient of the services having a civil law claim for recovery against the supplier) was consistent with those principles: seeat [34] - [42]. It would only be if reimbursement of VAT from the supplier became impossible or excessively difficult (for instance if the supplier was insolvent) that EU law may require the recipient of services to have a direct claim against the tax authorities: see [41].

Reemtsma, it is to be noted, was a case where the domestic legislation had properly implemented the VAT Directives insofar as the tax charge was concerned, the dispute being about the method by which reimbursement of tax, invoiced and paid in error, was to be effected. The error was made by the supplier and the German company in giving effect to the domestic legislation. No issue was raised or considered by the Court in relation to the effect of the passing-on of the VAT by the supplier to the German company in the context of the supplier’s claim against the Italian tax authorities. The Court proceeded on the basis that Italian law did provide a right for the supplier to recover the tax from the authorities: see again [36] of the judgment and [79] of the opinion of Advocate General Sharpston.

In Danfoss, a Danish law charged excise duty on the supply of mineral oils, in breach of EU law. Danfoss purchased lubricant oils from various Danish oil companies. Those companies paid the duty to the Danish exchequer and then passed on the total amount of the duty to Danfoss. Danfoss, in turn, sold some of the oils to another company, Sauer-Danfoss, including in the sale price duty on the oils. Danfoss and Sauer-Danfoss claimed a direct right of reimbursement under EU law from the tax authorities (the oil companies not having done so themselves). The Danish authorities refused reimbursement on the ground that the right to make a claim accrued only to the person who was directly taxable (the oil companies) and not to subsequent links in the chain (Danfoss and Sauer-Danfoss) who were not required to pay the duty themselves and did not pay any amounts to the exchequer which they could claim back.

Again, the CJEU held that, it being for the domestic legal system of each Member State to lay down the conditions under which claims for wrongly levied charges may be subject (provided they observe the principles of equivalence and effectiveness), a Member State may, in principle, oppose a claim against the tax authorities for reimbursement of wrongly paid duty by a consumer to whom the duty has been passed on, provided the consumer - who, in the final analysis, bears the burden of that duty - is able, under national law, to bring a civil law action for recovery against the taxable person for recovery of the sums unduly paid: see at [19] to [29]. It is worth noting, in the light of some the arguments raised before me, that at [26], the Court said this:

“…it has been held that, if the final consumer is able, on the basis of national law, to obtain reimbursement through the taxable person of the amount of the charge passed on to him, that taxable person must in turn be able to obtain reimbursement from the national authorities (see Comateb, para 24) …”

Accordingly, if tax passed on is recoverable under domestic law from the supplier as the taxable person, the supplier is in turn able to recover from the tax authority. It would not be open to domestic law to allow a claim by the consumer against the supplier and yet to deny a claim by the supplier against the tax authority. This means that, in Reemtsma, Italian law, which allowed a claim by the German company against the Italian supplier, could not have denied a claim by that supplier against the tax authority and in Danfoss, Danish law could only deny a claim by Danfoss against the tax authority if it provided it with a claim against the oil companies.

I would add, however, that logically the tax authority ought to be able to oppose, in whole or in part, a claim which a supplier would have in principle against the tax authority if it can be shown that the supplier would nonetheless be unjustly enriched if full reimbursement were made.

The Court reiterated, however, that if reimbursement from the taxable person were to prove impossible or excessively difficult (as in the case of insolvency), the principle of effectiveness would require the purchaser to have a claim for reimbursement against the tax authorities themselves.

Mr Macnab makes the point that there is no suggestion in Danfossthat the fact that the suppliers had not claimed reimbursement of the unlawful charges (and had thus not asserted their directly effective EU law rights) had any bearing on the CJEU’s analysis. At no point in Danfoss did the CJEU (or the Advocate General) suggest that the end consumer could have no claim against his supplier (and therefore had to be permitted a direct claim against the Member State) because the end consumer had no right against the supplier unless and until the supplier asserted his own EU law right to reimbursement of the unlawful charge against the Member State. That may be so, but whether or not the end customer does have such a claim is a matter of domestic law which can be invoked to give effect to that person’s directly effective rights. This is an aspect to which I will return.

In Farkas, Mr Farkas purchased a mobile hangar from an insolvent limited liability company (the seller) with an outstanding tax liability. The seller issued an invoice, which included VAT relating to that transaction, in accordance with the rules applicable to the ordinary tax system. When Mr Farkas paid the selling price, he included the VAT indicated by the seller, who paid that tax to the Hungarian tax authority.

Mr Farkas then deducted the output VAT recorded in that invoice. The Hungarian tax authority then carried out checks on the refunds requested by Mr Farkas in the VAT declarations. It concluded that the transaction should have been subjected to the reverse charge mechanism. Under that mechanism Mr Farkas was obliged to pay that VAT directly to the Treasury. The tax authority therefore requested that payment and, in addition, fined Mr Farkas in the amount of 50% of the VAT due.

 Mr Farkas claimed that the Hungarian tax authority had deprived him of his right to deduct VAT as a result of a formal defect, that is to say, the invoice in question had been issued in accordance with the ordinary tax system, rather than the reverse charge system, and thus infringed EU law. He submitted that the decision making him liable for the tax difference was unjustified because the seller in question paid the VAT in question to the Treasury. The referring court asked whether such decisions of the tax authorities complied with the PVD.

At [47] to [49] of its judgment, the Court stated as follows:

“47 Moreover, it must be stated that the right to deduct can be exercised only in respect of taxes actually due, that is to say, the taxes corresponding to a transaction subject to VAT or paid in so far as they were due (see, to that effect, judgment of 6 February 2014, Fatorie, C-424/12, EU:C:2014:50, paragraph 39). As it happens, the VAT paid by Mr Farkas to the seller of the mobile hangar at issue in the main proceedings was not due.

48 Thus, since the VAT was not due and the payment was made in breach of a substantive requirement of the reverse charge regime, Mr Farkas cannot claim the right to deduct that VAT.

50.

However, Mr Farkas may claim reimbursement of the tax unduly paid to the seller of the mobile hangar in accordance with national law (see, to that effect, judgment of 6 February 2014, Fatorie, C-424/12, EU:C:2014:50, paragraph 42).”

What was said in [50] (and similarly with [42] of Fatorie) must be taken to be a statement of what Mr Farkas was able to do if the local law actually provided for it. It was not a statement of a right which was conferred by EU law on Mr Farkas since EU law did not of itself provide for a right of reimbursement from the sellers in Farkas.

The Court repeated the conclusions in Reemtsma that a system in which, first, the seller of the property who has paid the VAT to the tax authority in error may seek to be reimbursed and, secondly, the purchaser of that property may bring a civil law action against that seller for recovery of the sums paid but not due, observes the principles of neutrality and effectiveness. Such a system enables the purchaser who bore the tax invoiced in error to obtain reimbursement of the sums unduly paid. In addition, the detailed procedural rules designed to ensure the protection of the rights which individuals acquire under EU law are a matter for the domestic legal order of each Member State, under the principle of the procedural autonomy of the Member States.

The principle of effectiveness had to be respected so that, once again, if reimbursement from the seller was impossible or excessively difficult, there may be a direct claim against the tax authority. This would be so, the Court thought, in the case of insolvency of the seller. I would add that it would be so, too, if national law did not provide a remedy for the purchaser against the seller.

Farkas, like Reemtsma, was a case where the domestic legislation had properly implemented the VAT Directives insofar as the tax charge was concerned, the dispute being about the method by which reimbursement of tax, invoiced and paid in error, was to be effected. I make the same observation in relation to Farkas as I made in relation to Reemtsma.

I come next to Case C-207/87 Weissgerber v Finanzamt Neustadt an der Weinstrasse[1988] ECR 4433, [1991] STC 589 (“Weissgerber”) which, by the time of the hearing, had come to form a central plank in Mr Scorey’s argument. I take the following, with minor changes, from the headnote in the STC report. The case concerned Article 13(B)(d)(i) of the 6th Directive, which provided exemption from VAT for “the granting and the negotiation of credit and the management of credit by the person granting it”. In Case 8/81 Becker v Finanzamt Münster-Innenstadt [1982] ECR 53 (“Becker”), the Court held that as from 1 January 1979 a credit negotiator could rely on Article 13(B)(d)(1) if he had not passed on the tax to persons following him in the chain of supply. Later, in Case 7083, Kloppenburg v Finanzamt Leer [1984] ECR 1075 (“Kloppenburg”),the Court reached the same conclusion as regards transactions carried out between 1 January and 30 June 1978. The taxpayer, Herr Weissgerber, was an insurance agent and finance negotiator who had received commissions during 1978 and 1979 from German banks in return for introducing and vouching for the solvency of clients seeking credit. The commissions were paid into his account by the banks. The credit notes sent by the banks to him did not show any amount of VAT. A VAT assessment was made by the German tax authorities for 1978 and 1979 in which were included those commissions, in accordance with the taxpayer's tax returns for those years, completed during 1980. However, when the taxpayer became aware of the judgments in Becker and Kloppenburg, he requested that his assessments to value added tax for the first half of 1978 and for 1979 be amended. The German tax authorities upheld the original assessments on the ground that value added tax had been passed on covertly. The taxpayer brought two joined actions against the notices charging value added tax for 1978 and 1979.

The Finanzgericht (Finance Court) Rheinland-Pfalz referred the following questions to the CJEU:

“1.

In relation to transactions carried out between 1 January 1978 and 30 June 1978 and transactions carried out in 1979, may the provision contained in Article 13(B)(d)(1) of [the 6th Directive] concerning the exemption from turnover tax of transactions consisting of the negotiation of credit be relied upon, in the absence of the implementation of that directive, by a credit negotiator if he did not pass that tax on to the persons receiving his services?

2.

If Question 1 is answered in the affirmative, must the credit negotiator pay turnover tax if he “covertly” passed on the tax, or only if he “overtly” passed it on?

3.

If turnover tax is also payable where the tax is passed on covertly, is it sufficient, for there to have been a covert passing on of turnover tax, that the credit negotiator, in agreeing his commission, expected that out of it he would have to pay turnover tax?”

Becker was at the forefront of Herr Weissgerber’s case. In Becker, the central issue was whether Article 13(B)(d)(1) of the 6th Directive could be relied on by individuals, in other words whether it was of direct effect. “Direct effect” is itself a shorthand. An explanation of what direct effect entails can be found in [17] to [25] of the Court’s judgment (although the phrase itself is used only once, in [14]). A summary is found at [25]:

“Thus, wherever the provisions of a directive appear, as far as their subject matter is concerned, to be unconditional and sufficiently precise, those provisions may, in the absence of implementing measures adopted within the prescribed period, be relied upon as against any national provision which is incompatible with the directive or in so far as the provisions define rights which individuals are able to assert against the State.”

That was repeated almost verbatim more recently in Claverhouse at [58], referring to that paragraph and to Case C-141/00 Kügler [2002] ECR 1-6833, paragraph 51; Joined Cases C-465/00, C-138/01 and C-139/01 Österreichischer Rundfunk and Others [2003] ECR 1-4989, paragraph 98; and Joined Cases C-453/02 and C-462/02 Linneweber and Akritidis [2005] ECR I-1131, paragraph 33. In Claverhouse the fourth question did use the phrase “direct effect” (the question being whether Article 13(B)(d)(6) of the 6^th Directive was of direct effect). The answer was that it was of direct effect; there can be no doubt at all that the exemption with which I am concerned, Article 135(1)(a), is also of direct effect at least to the same extent as Article 13(B)(d)(6).

In the section of its judgment under the heading “The system of value-added tax” starting at [41] the Court recorded a number of arguments raised by the Finanzamt and the German Government, based on the chain of taxation which derives from the mechanism of the right of deduction. The first matter raised by the Finanzamt and dealt within [42] was that the exemption might be disadvantageous to the very person entitled to it.

As to that the Court pointed out, at [45], that by availing themselves of an exemption, persons entitled thereto necessarily waive the right to claim a deduction in respect of input tax; and that, having been exempted from the tax, they are unable to pass on any charge whatsoever to the person following them in the chain of supply, with the result that the rights of third parties in principle cannot be affected. The arguments as to a disruption of the normal pattern of carrying forward the charge to value-added tax were therefore unfounded where a taxpayer has expressed his intention to avail himself of the exemption conferred by the directive and moreover bears the consequences of his choice.

Particular reliance was placed by the Finanzamt on the disruption caused by the fact that exemption might be claimed a posteriori to the detriment of taxpayers in the chain of transactions.

As to that the Court observed, at [46], that the objection was not relevant to a taxpayer who had claimed the benefit of the exemption when submitting his tax return and who had consequently refrained from invoicing the tax to the recipients of his services, with the result that third parties were not affected. That is the explanation for the wording of the dispositif which is careful to restrict the decision to situations where the tax was not passed on and which is focused on the inability of the State to claim that it had failed to implement the Directive.

It is to be noted, however, that the Court did not rule that a taxpayer would never be able to rely on the exemption where he had passed on the tax. The position where tax was passed on was simply not addressed in Becker, whether as an aspect of direct effect or as a defence to a San Giorgio claim.

In Weissgerber, the tax authorities argued that the VAT had been passed on, albeit covertly, taking the view that, if the tax had been passed on, Herr Weissgerber would not be entitled to the exemption under Article 13(B)(d)(1) of the 6th Directive because it would not be of direct effect. In other words, the tax authorities must have read Becker as deciding that, where tax is passed on, the disruption to the VAT system precluded the existence of directly effective rights enforceable by Herr Weissgerber. By covertly, I understand it to be meant that there was no tax invoice to the banks but that the amount paid reflected the possibility that VAT might be due. Herr Weissgerber argued that, notwithstanding the covert passing on, he had refrained from passing on the tax for the purposes of the qualification in Becker. The following questions were referred:

“1.

In relation to transactions carried out between 1 January 1978 and 30 June 1978 and transactions carried out in 1979, may the provision contained in Article 13(B)(d)(1) of [the 6th Directive] concerning the exemption from turnover tax of transactions consisting of the negotiation of credit be relied upon, in the absence of the implementation of that directive, by a credit negotiator if he did not pass that tax on to the persons receiving his services?

2.

If Question 1 is answered in the affirmative, must the credit negotiator pay turnover tax if he “covertly” passed on the tax, or only if he “overtly” passed it on?

3.

If turnover tax is also payable where the tax is passed on covertly, is it sufficient, for there to have been a covert passing on of turnover tax, that the credit negotiator, in agreeing his commission, expected that out of it he would have to pay turnover tax?”

The Court considered that Question 1 did not raise any novel question and simply followed Becker. The Court saw Questions 2 and 3 as essentially seeking a fuller explanation of the condition laid down in Becker and Kloppenburg and answered those two questions together. It considered (see at [15]) that the purpose of the condition was to prevent a claim for exemption made a posteriori by a credit negotiator from having an adverse effect on a recipient of his services who had already deducted the tax in question as input tax. Such a consequence could arise only if the trader claiming exemption had passed on the tax in accordance with the formalities prescribed by the 6th Directive and “if the recipient of the services is himself subject to VAT”, those quoted words being reflected in the dispositf in the phrase “so as to entitle that person [the recipient of his services] to deduct the input tax”.

Those last words are important. Where the recipient of the services is not entitled to deduct input tax, that is to say where he is not subject to VAT, the Weissgerber qualification, whatever its scope, simply does not apply.

The (composite) answer to the questions referred to the court was therefore that, in the absence of implementation of the 6th Directive, a credit negotiator could rely on the tax exemption contained in Article 13(B)(d)(i) if he did not pass value added tax on to the person receiving his services so as to entitle that person to deduct the input tax. Although in answering Questions 1 and 2 the Court was concerned with passing on, it was not concerned with, and did not address, any question of unjust enrichment. It did not therefore address the defence which passing on might afford, typically where a person asserts directly effective rights by making a money claim against the tax authority. Nonetheless, the decision is an example of the sort of considerations that apply in relation to passing on and may have relevance to an unjust enrichment defence.

Although the qualification in Becker and Kloppenburg was described as a condition in [15] of the judgment in Weissgerber, it was not a condition in the sense understood by an English lawyer. Rather, it was a restriction on the scope of the ruling: the Court was saying that its ruling was not conclusive of the position where there was passing on. And so the Court in Weissgerber was able to say that the purpose of that condition (ie restriction) was to prevent a claim having adverse effects on other traders. What the Court decided in Weissgerber was that a “covert” passing on would not risk this consequence since only an “overt” passing on would entitle another trader to deduct the input tax. What the Court did not say was that, in all cases, passing on in a way which would lead to a right to deduct would necessarily preclude reliance by the taxpayer on the exemption.

There is disagreement between the parties about the impact of Weissgerber. There are two analyses:

In my judgement, this is an irrelevant distinction for present purposes. The question in any particular case is whether the supplier is able to recover the wrongly paid tax from the tax authority and the facts of each case have to be looked at carefully. In a case where there is no passing on, or only covert passing on, Becker and Weissgerber show that the relevant Article is capable of having direct effect; but neither of those decisions (or other decision along the same lines) say that, because there is no overt passing on, a defence of unjust enrichment (as understood in EU law) would necessarily fail. What Becker and Weissgerber showis that where there is no passing on or only covert passing on, the relevant Article has direct effect.

Nor do either of those decisions demonstrate that where there is actual passing on (ie where the recipient of the supply is subject to VAT so as to be entitled to deduct the input tax), the relevant Article is necessarily not of direct effect. Whether the supplier can claim against the tax authority will depend on what happens down the chain of supply. As the Court explained in [15] of its Judgment in Weissgerber, the purposes of the condition, to repeat, is “to prevent a claim for exemption…. from having an adverse effect on other traders who have already deducted the tax in question as input tax”. That consequence could only arise if the supplier had “passed on the tax in accordance with the formalities prescribed” and if the recipient “is himself subject to value added tax”.

It is clear, therefore, that where the recipient of the supply is not subject to VAT, the condition is not applicable even if the recipient has been charged VAT on the supply. This remains the case even if on the facts it can also be shown that the price for the supply would have been less if the exemption had been available, although there may then be a defence of unjust enrichment available as between the supplier and the tax authority if domestic law provides it.

However, just as a person who is not subject to VAT is not adversely effected in a relevant way by an a posteriori claim for exemption, so too the recipient of the supply who is subject to VAT is not adversely affected in a relevant way if he himself seeks to assert his directly effective rights. By asserting such rights, he has to acknowledge that he is not entitled to claim the tax as input tax: see for instance Becker at [45]. Accordingly, the condition stated in Weissgerber is not, in my judgment, applicable where the recipient of the supply asserts his own directly effective rights. There is then no relevant adverse effect on the recipient within the contemplation of the condition in Weissgerber. This is so whether the recipient of the services has been able to claim the whole of the tax as input tax or whether (as in the present case) he is only able to claim a part of the tax as input tax pursuant to a partial exemption method. This is not to say that the recipient does not have directly effective rights: he clearly does, and domestic law must provide him with a remedy.

The point can be put in a slightly different way. Article 135(1)(a) is clearly capable of having direct effect, as is the case where VAT is not passed on. Although, where the recipient of the services is subject to VAT, there could (see Weissgerber) be an adverse effect on the recipient if the Article were to have effect as between the supplier and the taxing authority, that adverse effect is something which can be waived, as between the recipient and the taxing authority. In other words, the recipient is not to be denied directly effective rights on the basis of disruption to the VAT system. If the recipient choses to enforce his directly effective rights, then the very basis on which the supplier is deprived of his own directly effective right falls away.

Depending on the facts of a particular case, it may be necessary to consider the position of persons further down a chain of supply. In the present case, however, there is no suggestion that any person to whom the Trustees have made any supply would be affected, adversely or otherwise, if the Non-Insurers supplies to the Trustees were to be treated as exempt.

There is one final area to address, for the sake of completeness, on this aspect of the case. The Trustees are registered for VAT but are entitled only to partial exemption. That means that a proportion of the VAT which they have paid to the Non-Insurers is deductible as input tax, but not all of it. One thing is clear, I think, which is that Article 135(1)(a) cannot be of direct effect so far as concerns part of the consideration paid for a supply but not of direct effect so far as concerns another part of that consideration: there is a single supply which cannot be divided up in a way so as to be both exempt and non-exempt.

If I am wrong in concluding, for the reasons which I have, that the Non-Insurers do have a claim against HMRC in the present case, then Mr Scorey’s approach would result in their having no directly effective right at all in relation to their supplies to the Trustees. Indeed, that is precisely his submission based on Weissgerber. For my part however, I can see no reason why, as a matter principle, the proportion of the tax payable to the Non-Insurers which is not deductible as input tax by the Trustees should be treated any differently from the way in which the whole of the tax would be treated if the Trustees were not within the VAT system in the first place: so far as concerns that proportion of the tax payable, there would be no adverse consequence to the Trustees in granting the exemption any more than there would be an adverse consequence to them if they were not subject to VAT at all. Accordingly, even if Mr Scorey’s approach is correct, there is a powerful argument for saying that it should be restricted to cases where the whole of the tax paid is deductible as input tax by the recipient of the services. It is not necessary to give a concluded view on this point, which is not, in any case, acte clair and would require a reference to be made if an answer were necessary.

My conclusion on this part of the discussion is that EU law does not result in 135(1)(a) being deprived of direct effect even where the tax is overtly passed on or result necessarily in a supplier, such as the Non-Insurers in the present case, being unable to assert a right against the taxing authority to recover any tax which it has paid but which, under EU law, should not have been levied. EU law does not, therefore, preclude domestic provisions which provide the recipient of a supply with a remedy against the supplier who in turn has a remedy against the taxing authority in a case where the tax has been passed on by the supplier to the recipient. In my judgement, Weissgerber does not have the consequence that the Non-Insurers can have no claim against HMRC for the VAT which they charged the Trustees. I regard that as acte clair.

But even if I were wrong in thinking that the matter is acte clair, the Trustees face another problem in their reliance of Weissgerber. The Court subsequently said in Danfoss at [26] (see paragraph 122above) that, if the final consumer can recover from the taxable person, the taxable person can in turn recover from the tax authority. There is no authority for the proposition that, where the customer to whom the VAT has been passed on has a claim against the taxable person, a claim by the taxable person against the tax authority would nonetheless be prevented as a matter of EU law. If that were so, the result would be that the taxable person would have to bear the burden of the unlawful tax with no remedy against the tax authority. But that result cannot be contemplated and, since EU law is clear that the method by which a Member State gives effect to a person’s directly effective rights is a matter for the rules and procedures of the Member State, the correct way to avoid that result is to recognise that, as a matter of EU law, there is no objection to giving the customer a claim against the taxable person with the taxable person having a remedy against the Member State. Such a system would be a sufficient vindication of everyone’s rights and claims under EU law: the person to whom the tax has been passed on does not have to be given a right to claim against the Members State.

Accordingly, in the present case, it is purely a matter of UK law (i) what, if any, remedy the Non-Insurers have against HMRC and (ii) what, if any, remedy the Trustees have against the Non-Insurers and HMRC and how those rights might differ depending on whether the Non-Insurers assert any claims against HMRC. It is those aspects of the case to which I now turn.

Although VAT statute law is UK-wide, the present case involves consideration of common law rights under the law of England. It is that position with which I am concerned, the relationships between the Trustees and the Non-Insurers and between each of those parties and HMRC being a matter of English law.

I have set out the relevant statutory provisions above (see paragraphs 20ff above). Under those provisions it is clear that pension management services supplied to the Trustees by Non-Insurers were not exempt supplies.

It is also clear that if the Non-Insurers have directly effective rights to treat their supplies as exempt which they can and do assert, the output tax for which they accounted in respect of their transactions with the Trustees should not have been brought into account within the meaning of section 80(1) VATA. HMRC are accordingly liable to credit the Non-Insurers with the relevant amounts. That is subject to a defence of unjust enrichment under section 80(3).

However, the Non-Insurers do not have to assert any directly effective rights which they have. So far as concerns periods in respect of which the UK statutory provisions provided for no exemption in relation to the relevant supplies, the Non-Insurers, as between themselves and HMRC, are entitled to rely on those provisions and to treat the supplies to the Trustees as standard rated. For periods after 1 January 2005, those provisions followed the wording of the 6^th Directive and the PVD and would fall to be interpreted consistently with those Directives: it would follow that if the provision of pension management services constituted insurance transactions (contrary to my decision on Issue 1) then the Non-Insurers’ supplies would be exempt under UK law and the Non-Insurers would have no right to assert that the supplies were standard rated. For periods before 1 January 2005, the position may be the same applying a Marleasing approach to construction, as to which see paragraphs 235ff below although, prima facie, such supplies were not exempt.

The Trustees were not liable to account to HMRC for any VAT in respect of the supplies to themselves. At the European level this can be seen from Reemtsma at [30] – [33] to which I have already referred. In English law, there is ordinarily nothing in the UK statutory provisions which casts an obligation on a person to account for VAT to HMRC on supplies made to him. Certainly, if, in the present case, the provision of pension fund management services by Non-Insurers to the Trustees are properly standard rated (which I consider to be the case in the light of my answer to Issue 1), there would have been no obligation on the Trustees to account to HMRC for the VAT even if the Non-Insurers had failed to do so (for instance if, counter-factually, the Non-Insurers had become insolvent).

The obligation (if any) of a customer to pay VAT in respect of the supplies to him will, in England at least, ordinarily arise out the contract between the parties. A contract may be totally silent about VAT, in which case the customer simply pays the contract price and is not liable to his supplier to pay any additional amount by way of VAT. In that case, the price received by the supplier falls to be treated, as between the supplier and HMRC, as a VAT-inclusive price and he must account for VAT accordingly. In contrast, the contract might provide that the price is exclusive of VAT and for VAT (at the standard rate in the case of standard-rated supplies) to be paid in addition.

At this stage, I introduce thelitigation between certain investment trust companies and HMRC (“the ITC litigation”) which has taken a prominent part in these proceedings. This ended with the decision of the Supreme Court in Investment Trust Companies (in Liquidation) v HMRC [2017] UKSC 29, [2017] STC 985, [2017] 3 All ER 113, [2017] 2 WLR 1200 (“ITC (SC)”); for earlier proceedings in the litigation see the decision of the Court of Appeal at [2015] EWCA (Civ) 82, [2015] STC 1280 and the decisions of Henderson J at [2012] EWHC 458 (Ch), [2012] STC 1150 (“ITC 1”) and [2013] EWHC 665 (Ch), [2013] STC 1129.

For present purposes, the facts can be very shortly stated:

In ITC 1, Henderson J held, among other things:

At [81] of ITC 1, Henderson J referred to a number of admissions which HMRC had made in their pleadings. The first was that, insofar as any VAT may have been paid by an ITC to a Manager in circumstances where VAT was not properly charged and due as a matter of law, the VAT was paid by mistake. In [82], he considered that this admission left room for an argument that, despite the unlawfulness of the VAT as established in Claverhouse, it was nevertheless still due as a matter of contract between the ITCs and the Managers, with the consequence that they were obliged to pay it in any event, and the payments were not caused by the mistake. Such an argument was deployed, on behalf of HMRC rather than the taxpayer, by Mr Swift QC. But, as Henderson J explained,

“the objections all but evaporated in the course of the hearing, and when he came to address the court Mr Swift expressly accepted that the contracts between the claimants and the Managers did not, on their true construction, oblige the claimants to pay VAT to the Managers if it was not legally due. In other words, he accepted that, on the hypothesis that the unlawfulness of the VAT charges had been known to the investment trusts and the Managers, the investment trusts would not have been under any contractual obligation to pay the charges, had the Managers in fact sought to impose them.”

I do not know precisely how far Mr Swift’s acceptance went. But as recorded by Henderson J, his acceptance was in a context where the tax was not “legally due” which must mean legally due from the Managers to HMRC. In the ITC litigation it was known, by the time the proceedings had been commenced, that the Managers were themselves asserting directly effective rights and thus asserting that the VAT was not legally due since domestic law would have to recognise the directly effective rights. In contrast, if the Managers had stood by their rights (and obligations) under the UK statute, HMRC could not have objected and, as between them, the VAT would have been legally due.

However, it is apparent from the last sentence of the citation above that Henderson J was focusing on the position which would have obtained if, at the time when the VAT was in fact paid, it had been known that the VAT charge was “unlawful”. He clearly saw the question whether VAT was “applicable” as being the same as the question whether the VAT was legally due, drawing no distinction between the case where the taxpayer (the Managers) did assert their directly effective rights and the case where the taxpayer did not assert those rights. Indeed, such a distinction could not have been drawn as of the time of actual payment since, at that time, the parties were proceeding on the basis that VAT was (lawfully) due and the Managers had not had cause to consider whether or not to assert their directly effective rights.

Since there was no live issue about the construction of the contracts between the ITCs and the Managers, Henderson J did not set out the relevant provisions or discuss the rival arguments. He did, nonetheless, express his agreement with the conclusion that, as a matter of construction of the agreements, there was no contractual obligation on the ITCs to pay the VAT charges to the Managers, because the VAT was not lawfully due on the services supplied by the Managers. That appeared to him to be the natural construction of an obligation to pay VAT “if due”, or words to similar effect. He considered that it would require very clear words to impose a liability to pay VAT if it was not in fact due. By “due” in this context he can only have meant that which would have been due if EU law had been properly transposed into domestic legislation. That is entirely consistent with what he had said in the last sentence of the citation above, in accordance with my explanation of what he was saying.

Thus, the position, as he saw it, was that the claimants paid the VAT charged on the invoices in the mistaken belief that it was legally due, but when they were in fact under no obligation to do so. Since there was an operative mistake, and since a mistake of law is capable of giving rise to a restitutionary claim, Henderson J was able to find a remedy available to the ITCs against the Managers.

In the Supreme Court, Lord Reed stated at [4] and [70] that the contracts provided for fees to be paid plus VAT “if applicable” rather than “if due”, but I do not think that anything turns on that difference.

A similar argument has been run by Mr Scorey in the present case. I consider that the argument is open to him and disagree with Mr Macnab that the point has been conclusively determined in the ITC litigation. Before Henderson J, the argument evaporated. The judges at all levels appear to have been content to proceed on the basis that the relevant VAT was not “due” or “applicable”. There has been no actual decision to that effect although Henderson J did express his (obiter) agreement with Mr Swift’s acceptance that it was not. Even if there had been an actual decision to that effect, that would not strictly be conclusive on the meaning of the contracts in the present case, although I am bound to say that I can detect no material difference.

That said, I agree with Henderson J for the reasons which he gave. In my judgment, whether VAT is “applicable” or not cannot depend on whether it has actually been paid or not. Henderson J was correct to ask himself the question what the position would have been if the unlawfulness of the VAT charges had been known and had the Managers sought to impose the charges.

I would put the point more forcefully. Suppose that in the present case the Trustees had failed to pay VAT to the Non-Insurers and were now being sued for that VAT under the relevant asset management contracts. It is now known that the supplies by Non-Insurers were exempt supplies in accordance with EU law and that the UK has failed properly to transpose the mandatory exemption. Could the Non-Insurers now recover VAT under the contracts? In my view, the answer is that they could not. The answer, as between the Trustees and the Non-Insurers, cannot depend on whether or not the Non-Insurers seek to enforce their directly effective rights. If it did so depend, then the VAT might be “applicable” today at a time when (let it be assumed – the position is not clear) the Non-Insurers have not asserted their directly effective rights but not “applicable” tomorrow when they do assert those rights. That is not a sensible construction of the contracts.

The choice, in all cases, is between the VAT being “applicable” or not and that choice cannot depend on when the choice is made. Given that binary choice, I conclude that the VAT is not “applicable” within the meaning of the contracts.

It is clear that the 6th Directive and the PVD do not have horizontal effect, that is to say that persons (whether individuals or corporations) cannot assert the direct effect of a Directive as between themselves. Accordingly, the Trustees and the Non-Insurers cannot rely on any horizontal effect. I do not consider that my conclusion on construction of the management contracts is, in effect, to give horizontal effect to the 6th Directive and the PVD by the back door. It is simply a recognition that the contracts between the Trustees and the Non-Insurers reflect that there is an obligation only to pay VAT which is properly due.

Both parties appealed against Henderson’ J’s substantive decision. In the light of the later decision of the Supreme Court which took a markedly different approach, allowing HMRC’s appeal and dismissing the ITCs’ cross-appeals, I do not think that I need to spend any time on the Court of Appeal’s decision or judgments. It is to ITC (SC) to which I now turn.

Lord Reed gave the only judgment, with which the other Justices agreed. It is to be noted that the appeal proceeded on the basis that there had been a relevant mistake capable of supporting a restitutionary claim by the ITCs against the Managers. Thus, at the outset, Lord Reed refers to supplies which “were mistakenly believed to be taxable by the customer who paid an amount charged in respect of the tax”: see [1] of the judgment.

In [2], Lord Reed identified the principal issues for decision:

In [10] to [14] of his judgment, Lord Reed summarised the facts which I have already covered at paragraph 167 above. It is to be noted that it was common ground between the Managers and HMRC that, in principle, claims under section 80 VATA could be made by the Managers. It is not common ground before me, although Mr Macnab says it should be since there is no serious argument that it is wrong whereas Mr Scorey argues that it is wrong and since it was common ground and not argued it is open to me to differ.

On the basis of that common ground, Lord Reed proceeded to consider the common law claims of customers, who pay undue VAT charged by their suppliers, against HMRC based on unjust enrichment. He did so by reference to the four questions asked by Lord Steyn in Banque Financière de la Cité v Parc (Battersea) Ltd [1999] 1 AC 221, 227: (a) Has the defendant been benefited, in the sense of being enriched? (b) Was the enrichment at the claimant's expense? (c) Was the enrichment unjust? (d) Are there any defences? The caveat expressed in [41] should however, be noted:

“… Lord Steyn's four questions are no more than broad headings for ease of exposition. They are intended to ensure a structured approach to the analysis of unjust enrichment, by identifying the essential elements in broad terms. If they are not separately considered and answered, there is a risk that courts will resort to an unstructured approach driven by perceptions of fairness, with consequent uncertainty and unpredictability. At the same time, the questions are not themselves legal tests, but are signposts towards areas of inquiry involving a number of distinct legal requirements. In particular, the words “at the expense of” do not express a legal test; and a test cannot be derived by exegesis of those words, as if they were the words of a statute.”

As a result of the common ground mentioned in the preceding paragraph of this judgment, the Managers had already received a refund of the £75 (plus interest). What was in dispute was the £25 which they did not receive. It was held that HMRC had not been enriched by the amount that the managers had deducted as input tax ie the £25. Lord Reed recorded the argument of the ITCs at [27] but rejected it. He referred in [29] to Becker and to Joined cases C—1433/13, C-154/13 and C-160/13, VDP Dental Laboratory [2015] STC 1133. He concluded that it followed from those authorities that the Managers could not both claim reimbursement of the output tax which they had paid to HMRC, under section 80 of the VATA, on the basis that their supplies were exempt from VAT, and simultaneously assert an entitlement to retain the amounts which they had deducted as input tax, on the basis that their supplies were taxable. He concluded at [30] and [31] as follows:

“30.

The commissioners were not, therefore, enriched by the managers' retention of the notional £25, and the managers have, in principle, no defence to a claim by the lead claimants for the restitution of that amount. That conclusion is as one would expect. The lead claimants' claim to restitution against the commissioners proceeds on the basis that the supplies which they received from the managers were exempt from VAT. That being so, it would be surprising if they could present that claim, in relation to the measure of restitution, on a basis which was predicated on the supplies being taxable.

31.

It follows that the commissioners' enrichment was only to the extent of the notional £75.”

A change of position defence to a claim by the ITCs would clearly not be available to the Managers since they would have obtained the economic benefit of the £25 by deducting it as input tax thus decreasing their liability to account for VAT to HMRC.

Lord Reed turned at [32]ff to the question whether the enrichment of HMRC was at the expense of the ITCs. In economic terms, HMRC were enriched at the expense of the ITCs. The mistake led to the payment of an amount equal to the Managers’ output tax (£100) by the ITCs to the Managers and to the payment of the amount by which the output tax exceeded the input tax (£75) by the Managers to HMRC. The net result was that the ITCs were worse off by £100 and HMRC were better off by £75.

But no payment was made by the ITCs to HMRC. Nor were the Managers simply a conduit for the VAT. Nor was the payment by the ITCs to the Managers the cause of the payment of VAT by the Managers to HMRC. As Lord Reed had already explained (and as I have explained too), the Managers were liable to account to HMRC for output tax once they had made the supplies whether or not they received payment from the ITCs and the ITCs themselves had no direct liability to HMRC.

At [34] and [35], Lord Reed considered the approach which had been taken by Henderson J to the exceptions to the general rule that a defendant was legally enriched at the expense of the person from whom the benefit in question was directly received (the general rule, if it applied, restricting the ITCs to a claim against the Managers with no claim against HMRC). In accordance with his approach, Henderson J had decided that the case was exceptional, as explained by Lord Reed at [35], concluding that, in the context of VAT, the final consumer who paid the tax had a sufficient economic connection with HMRC to be able to say that they had been enriched at his expense when the tax ought never to have been imposed in the first place. This was a conclusion with which the Court of Appeal agreed.

Whilst eschewing, in [38], any attempt at a definitive statement of the circumstances in which the enrichment of a defendant can be said to be at the expense of a claimant, Lord Reed considered that the Supreme Court had a responsibility to establish more precise criteria. This he proceeded to do.

So far as relevant to the present case, I would particularly refer to [42] and [43]:

“42.

The structured approach provided by the four questions [Lord Steyn’s questions] does not, therefore, dispense with the necessity for a careful legal analysis of individual cases. In carrying out that analysis, it is important to have at the forefront of one's mind the purpose of the law of unjust enrichment … it is designed to correct normatively defective transfers of value, usually by restoring the parties to their pre-transfer positions. It reflects an Aristotelian conception of justice as the restoration of a balance or equilibrium which has been disrupted. That is why restitution is usually the appropriate remedy.

43.

The nature of the various legal requirements indicated by the “at the expense of” question follows from that principle of corrective justice. They are designed to ensure that there has been a transfer of value, of a kind which may have been normatively defective: that is to say, defective in a way which is recognised by the law of unjust enrichment (for example, because of a failure of the basis on which the benefit was conferred). The expression “transfer of value” is, however, also too general to serve as a legal test. More precisely, it means in the first place that the defendant has received a benefit from the claimant. But that is not in itself enough. The reversal of unjust enrichment, usually by a restitutionary remedy, is premised on the claimant's also having suffered a loss through his provision of the benefit.”

At [59] and [60], Lord Reed considered economic and commercial reality, stating that the “at the expense of” requirement was not satisfied by a connection between the parties’ respective benefit and loss merely as a matter of economic or commercial reality and identifying in [59] a number of problems with such an approach. And at [60] he stated that a more fundamental difficulty arose from the fact that the purpose of restitution is not to compensate for loss, but to reverse the defective transfer. Looking to see who suffered the economic loss is not, in principle, the correct way of identifying the appropriate claimant.

At this point, I interrupt Lord Reed to refer to the judgment of Underhill LJ in Test Claimants in the FII Group Litigation v Revenue and Customs Comms [2016] EWCA Civ 1180, [2017] STC 696. At [201] he identified an inherent tension between the English law remedies in restitution and the San Giorgio rights to recover overpayments of tax. The latter is not a right which takes any account of the level of enrichment of the tax authority receiving the overpayment of tax. In contrast, the former considers the level of actual enrichment of the recipient, which is irrelevant to both the existence and extent of the latter. A related point, identified at [202], is that if the domestic law remedy is not adequate to vindicate EU law rights, the remedy must be moulded or reformed in order to so. And thus:

“the exercise that the domestic court is engaged upon is to ‘mould’ a restitutionary remedy, which focuses on focuses on the enrichment of the defendant, in order to provide full compensation for overpaid tax. That might be regarded more accurately as a total transformation than any kind of moulding process.”

but, as he observed at [335], that is what the court has to do.

Returning to Lord Reed, he came to “at the expense of” in the case before him at [67] to [74]. A considerable part of this section was devoted to considering whether there were only personal obligations or whether there were agency or trust considerations to take into account. I do not need to refer to those any further as it is not contended that agency or trust considerations arise in the present case. I can start at [71] where Lord Reed said this:

“71.

Returning, then, to the question whether the unjust enrichment of the commissioners was at the expense of the lead claimants, and focusing on whether there was a transfer of value from the lead claimants to the commissioners, the answer is in the negative. There was a transfer of value, comprising the notional £100, from the lead claimants to the managers, under the contract between them. It was defective, because it was made in performance of a contractual obligation which was mistakenly believed to be owed. There was a subsequent transfer of value, comprising the notional £75, from the managers to the commissioners. It was also defective, because it was made in compliance with a statutory obligation which was inapplicable because it was incompatible with European Union law. These two transfers cannot be collapsed into a single transfer of value from the lead claimants to the commissioners.”

He went on in [72] to explain why that was so (first, there was no agency, secondly, there was no possibility of tracing and thirdly, it was impossible to disregard the fact that there were separate transactions – one between the ITCs and the Managers and another between the Managers and HMRC). Further the “economic and commercial reality” approach was to be rejected for reasons already given.

It is true that the payment of the £75 was described in [71] as defective for the reason given. But even if, as Mr Scorey would argue in the present case, the payment from the Non-Insurers to HMRC was not defective in a relevant way because it was made in compliance with domestic law which the Non-Insurers (let it be assumed for the purposes of the argument) have not challenged, the case for collapsing the two transfers into single transfer is not strengthened in any way. Each of Lord Reed’s points applies with equal force.

Lord Reed’s conclusion at [73] was therefore that the ITCs did not in principle have any right to restitution against HMRC. They did, in contrast, have a right to restitution against the Managers. That right was for the entire amount ie the full £100. The Managers did not have a change of position defence in respect of the £75 since that change of position was reversible under section 80 VATA (as he went on to explain and to which I too will come). Nor did they have a change of position defence in respect of the £25 which they retained.

Lord Reed then went on to consider section 80 VATA and the arguments which had been raised about it: see [75] to [88]. At [77] he recorded some important common ground:

Lord Reed then turned to the first issue in dispute which was whether the effect of section 80 is to exclude a common law claim by the ITCs, assuming, contrary to his earlier conclusion, that such a claim might otherwise be brought. In his discussion of section 80 and 80A (and the Regulations made under section 80A) he explains as follows in [80] and [81]:

“81.

… [The statutory provisions] create a scheme which enables consumers who have been wrongly charged VAT to obtain reimbursement. The consumers are able to recover the VAT which they were wrongly charged, to the extent that it was remitted by the supplier to the commissioners, through the medium of the supplier's claim under section 80.

82.

Although the consumers' remedy is indirect, it can generally be expected to be effective: if the supplier is otherwise reluctant to make a claim, the consumers have a direct claim against him, as explained below. Subject to the question of time bar, these arrangements therefore remove any need there might otherwise be, in most circumstances, for the consumer to have a direct remedy against the commissioners. It will be necessary at a later point to return to the question whether there may nevertheless be some circumstances in which a direct remedy is required by European Union law.”

In [88], Lord Reed expressed his conclusions that section 80 bars claims by consumers who ultimately bear the burden of VAT. It nevertheless enables them to be reimbursed, subject to a limitation period designed to avoid the disruption of public finances. That ability to be reimbursed is clearly a reference to the scheme which he had described in [81]. I should deal briefly with his reasoning on [83].

In [89] to [92], Lord Reed considered the compatibility of his conclusions with EU law. He referred to San Giorgio, Reemtsma and Danfoss, establishing that a system under which only the supplier to entitled to seek reimbursement of VAT from the tax authorities, and the consumer can seek restitution from the supplier, meets the requirements of EU law, subject to the caveat in Reemtsma where reimbursement becomes impossible or excessively difficult.

Having discussed the general principles, Lord Reed turned at [93] and [94] to consider the case before him:

“93.

In the present case, the lead claimants had a common law right to restitution of the amounts mistakenly paid to the managers, whose enforcement was neither impossible nor excessively difficult. The managers had a statutory right to recover the notional £75 from the commissioners, under arrangements which ensured that it was passed on to the lead claimants. The managers retained the remaining £25 and were not insolvent. They were therefore in a position to refund it to the lead claimants. The only amounts which the lead claimants could not recover were the amounts which they had paid during the “dead periods”, to the extent that those amounts had been paid by the managers to the commissioners: that is to say, the notional £75 whose recovery from the commissioners was time-barred under section 80(4) of the 1994 Act. Although a claim by the lead claimants against the managers in respect of the dead periods would not have been time-barred, because of the more generous limitation period allowed by section 32(1)(c) of the Limitation Act 1980 , the managers would have a defence of change of position, since the amounts which they paid to the commissioners during those periods were irrecoverable. The inability of the lead claimants to recover those sums is not, however, incompatible with European Union law: as explained earlier, it is conceded that the three year limitation period imposed by section 80(4) of the 1994 Act is compatible with European Union law.”

94.

In these circumstances, the inability of the lead claimants to pursue a direct claim for restitution against the commissioners is not incompatible with European Union law. That follows from the application of well-established principles of European Union law. There is therefore no need for any reference to the Court of Justice. Nor is it necessary or appropriate to consider what the position would be in a hypothetical case where the supplier was insolvent: the court has heard no submissions, and has no information before it, as to how reimbursement arrangements under section 80 might operate in that situation.”

Accordingly, he allowed HMRC’s appeal.

Mr Macnab contends that ITC (SC) is conclusive on Issue 2 in favour of HMRC. Mr Scorey submits that the decision turned on propositions which were common ground but which he contends were wrong (namely that the Managers had a directly effective EU law right against HMRC and that the ITCs had a commonlaw claim against the Managers); and he submits that his argument based on Weissgerber was not considered and remains open.

The first question, then, is what, if any, claims the Non-Insurers could assert against HMRC. Although they are not compelled by law to assert their directly effective rights it is clear that, if they do so, they have a claim in principle under section 80, although that claim will be subject to available defences. Once the directly effective rights are asserted, the position will be that the Non-Insurers would have accounted for output tax which was not due so that HMRC would be liable to credit them with that amount under section 80(1). HMRC would then be liable to pay to the Non-Insurers that amount but with a deduction under section 80(2A) of an amount equal to the input tax which they had deducted. Lord Reed explains this aspect of section 80 in [13] of his judgment. In the illustrative example, the £75 would be repaid to the Managers but the £25 which they retained would not.

Mr Scorey contends, however, that the Non-Insurers have no effective claim against HMRC even if they assert their directly effective rights. He argues that, as a matter of EU law, HMRC would have a defence, in that the Non-Insurers would be unjustly enriched were they to receive reimbursement of the VAT since the tax had been passed on to the Trustees. Weissgerber, he argues, establishes that a taxpayer who has passed on the tax so that the recipient of the services can deduct that as input tax cannot rely the exemption at issue (Article 13(B)(d)(i) of the 6th Directive in Weissgerber and Article 135(1)(a) in the present case). This point was not argued in the ITC litigation and remains open.

I have already addressed Weissgerber at some length. For reasons already given, I do not consider that it assists the Trustees in their case. In my view, both the Trustees and the Non-Insurers are entitled (although neither of them have to do so) to rely on the direct effect of the exemption, so that UK law and procedures must provide them with remedies which vindicate their directly enforceable rights.

The position, then, is that the Non-Insurers have valid claims in principle against HMRC under section 80. The remedy which UK legislation gives the Non-Insurers is crediting the amount of output tax accounted for and repayment of the amount specified in section 80(2A), £75 in the illustrative example. However, the claim in respect of the output tax accounted for within section 80(1) (and therefore of the amount of repayment under section 80(2A)) is subject to the unjust enrichment defence under section 80(3). There will be no enrichment, let alone unjust enrichment, if the supplier makes arrangements under which any reimbursement obtained from HMRC is in turn reimbursed to the customer. This is recognised by section 80A which makes provision for the relevant reimbursement arrangement to be disregarded for the purposes of section 80(3) unless it complies with applicable regulations.

Let it be supposed for the moment that the customer has a claim against the supplier to recover the VAT under a mistake of law. The supplier faced with such a claim has it in his own hands, subject to any limitation issues to which I will come, to obtain a remedy against HMRC if he wishes in order to obtain an indemnity or partial indemnity against the customer’s claim. This can be done by (i) invoking his directly effective rights and (ii) making arrangements to pay the amounts recovered from HMRC to the customer. This would eliminate such defence as HMRC would otherwise have under section 80(3) to the supplier’s claim (subject to compliance with any applicable regulations under section 80A). The supplier does not have to take this course: he can decline to invoke his directly effective rights or decline to make such arrangements. He would then find himself unable to recover from HMRC. That, however, would not provide him with a defence to a claim by the customer.

I turn now to consider the Trustees’ claims, if any, against the Non-Insurers. It is important for HMRC, if they are to succeed in resisting the Trustees’ claim, that the Trustees have, in principle, (that is to say, ignoring defences which HMRC might have, including change of position and limitation defences), claims against the Non-Insurers. If there were, in principle, no such claim, the Trustees would, in order to vindicate their directly effective rights, have to be provided with a remedy against HMRC. This would be a claim to be reimbursed for the VAT which they have paid to the Non-Insurers, subject to giving credit for any input tax which they (and perhaps the employers) have already recovered and subject to any defences which HMRC might have. As to those defences, section 80(7) VATA is relied on by HMRC as defence to any common law claim (although absent any other remedy, that section may have to be disapplied, vis à vis the Trustees, in whole or in part).

The starting point in deciding whether the Trustees do have a claim against the Non-Insurers is the contractual arrangements between them. I have already dealt with the proper interpretation of the words “if applicable” in the relevant contracts and have concluded that, in the present case, VAT was not “applicable”. Accordingly, the Trustees have paid VAT to the Non-Insurers when the contracts did not require them to do so. Such payments were made because VATA required it to be paid: but although UK statute did so require, the parties were acting under a mistake of law in thinking that the contracts required such payments to be made. It follows that, in principle, the Trustees have a restitutionary claim for reimbursement.

I reject Mr Scorey’s submissions that the payments of VAT by the Trustees were not made under any operative mistake. He makes, in essence two points. The first is that the VAT was, indeed, “applicable” within the meaning of the management contracts; I have already dealt with and rejected that submission.

The second of Mr Scorey’s points is that there was no mistake. The payments were made in accordance with domestic law as it then stood. Unless and until the Non-Insurers invoke their directly effective rights, it cannot be said that there is any mistake at all. Even once they have been invoked, if they are, that does not affect the position at the time when the payments were made, namely that they were due under domestic law so that there was no mistake. This second argument is not, it seems to me, anything other than another way of putting the first argument. The answer to it is that the mistake was an understanding that the contract obliged the Trustees to pay the VAT when, had the law been properly known and understood, it would have been appreciated that the contract did not require it to be paid. In other words, as a matter of domestic law – that is to say, the true construction of the contracts – VAT was not payable by the Trustees. The mistake was to pay the VAT when it was not “applicable”. The fact that the mistake arose because of a mistake as to the requirements of EU law does not mean that the payments were not made under a relevant mistake in the first place. In any case, this point only arises in relation to the periods before 1 January 2005: after that date, the UK legislation reflected the wording of the 6^th Directive and the PVD and the mistake (if there was one) was as to the meaning of the domestic provisions.

In my judgment, therefore, the Trustees have, in principle, a right in restitution against the Non-Insurers to recover the VAT which they have paid. In the illustrative example, that is the £100, not just the £75. Having rejected Mr Scorey’s submissions on the meaning “if applicable” in the contracts, the decision in ITC(SC) leads inevitably to this conclusion.

This right is not dependent in any way on whether the Non-Insurers have any right against HMRC. Against that conclusion, it might be argued that, unless and until the Non-Insurers do assert their directly effective rights, the Trustees can have no claim against them. I would not agree with such an argument. It amounts, essentially, to the same as the arguments to the effect that the VAT was “applicable” and that there was no mistake because domestic law required the VAT to be levied by the Non-Insurers on the supplies to the Trustees. As between the Non-Insurers and the Trustees there was, I consider, a material mistake, whatever the rights as between the Non-Insurers and HMRC may be before the Non-Insurers assert their directly effective rights. This is not to give horizontal effect to the exemption as between the Trustees and the Non-Insurers. A mistake about the law giving rise to a right to restitution remains a mistake even where it arises out of an incorrect transposition of a directive into domestic law. The remedy of restitution is given to correct the consequences of that mistake as between the Trustees and the Non-Insurers: it is not to allow the Trustees to assert a directly effective right against the Non-Insurers.

In any case, an argument on the part of the Non-Insurers that the Trustees could have a claim against them only once they, the Non-Insurers, had themselves asserted their directly effective rights would have to be based on the proposition that they have no right to obtain reimbursement from HMRC until they had done so. The answer to that argument is that it is open to the Non-Insurers in fact to assert their directly effective rights and to make arrangements to reimburse to the Trustees any amounts they recover from HMRC. That they may choose not to do so is not an answer to the Trustees’ prima facie restitutionary claim.

In the light of this discussion, the answer to the first limb of Issue 2 identified at paragraph 105 above is, in my judgment, that the Trustees have a restitutionary claim against the Non-Insurers for the whole of the VAT which was paid.

I turn now to the second limb of Issue 2, namely whether it would be impossible or excessively difficult to obtain reimbursement. I address this question ignoring limitation issues for the moment. A claim by the Trustees against the Non-Insurers would not be straightforward. This, however, would be largely, although not exclusively, due to the underlying merits of the claim rather than to any procedural complexities.

In the ITC litigation, the VAT liability issue had already been determined by Claverhouse. In contrast, in the present case, the VAT status of the supplies is in dispute (that is Issue 1). Mr Scorey relies on this distinction, suggesting that this is one factor to be taken into account in considering whether a claim by the Trustees against the Non-Insurers would have been impossible or excessively difficult. Mr Macnab submits that this distinction is irrelevant: the fact that the merits of the Trustees’ underlying rights under EU law are debatable has no bearing on the compatibility with EU law of the UK’s national procedural legal system for giving effect to any EU law rights which they may have. In that context Mr Macnab has referred to Leeds City Council v RCC[2016] STC 2256 (“Leeds”) where Lewison LJ said this at [42]:

“42.

So far as the second point is concerned, article 4.5 [6VD] may indeed be difficult to understand or to apply. I express no view one way or the other. But the fact that a piece of European legislation is difficult to understand or apply cannot justify an extension of the limitation period. If the meaning of a piece of European legislation is unclear it can be referred to the CJEU which sometimes manages to clarify its meaning. If and in so far as there was a perceived problem it arose because of uncertainties about the law, and had nothing to do with any shortcomings in domestic procedure for claims for repayment of VAT.”

What Lewison LJ said there was in the different context of limitation. The judge saw the obscurities of EU law as providing no justification for any extension of a limitation period. It is a different issue whether a similar obscurity would make a claim by the Trustees impossible or excessively difficult. However, in order for any claim by the Trustees to succeed, the difficulty of EU law would have to be grappled with and resolved in some tribunal or court. The difficulties are what they are: but they are no more difficult of resolution in an action between the Trustees and the Non-Insurers than they are in an action (the present action) between the Trustees and HMRC. In other words, the claim which the Trustees assert against HMRC is no less “impossible or excessively difficult” than would be a claim against the Non-Insurers. Accordingly, there is in my view no force in the “impossible or excessively difficult” argument. I consider that the approach of Lewison J in a limitation context is equally applicable in the present context.

In any case, although HMRC accept that there may be (actual) exceptional circumstances where EU law requires a departure from the “ordinary” system under domestic law for giving effect to the rights that persons may derive from EU law. Mr Macnab submits that such circumstances must, however, be unrelated to the merits of the claim itself.

In that context, he refers to the insolvency of the supplier (the instance given by the Court in Reemtsma, Danfoss and Farkas). He also refers to [82] to [92] of the Opinion of Advocate General Sharpston in Reemtsma, placing particular reliance on [86] where she stated that the system of remedies in that case (under Italian law) was sufficient unless it “has, as a result of material circumstances unrelated to the merits of the claim [a footnote refers to insolvency of the supplier] failed to produce the normal outcome”. He also refers to a number of other cases: Littlewoods (No 2) [2014] STC 1761 at [286], Henderson J; Littlewoods (CA) [2015] STC 2014 at [93], Arden LJ; Birmingham Hippodrome Theatre Trust Ltd v HMRC [2014] EWCA Civ 684, [2014] STC 2222 at [41], Lewison LJ: “Moreover, the principle [of effectiveness] is concerned with the enforcement of rights under EU law. It does not identify what those rights are”.

Without necessarily accepting Mr Macnab’s submission that the circumstances which would render a claim impossible or excessively difficult cannot ever relate to the merits of the claim, I do not consider that the difficulties of interpretation of the 6th Directive and the PVD or HMRC’s own interpretation of the legislation (even assuming in favour of the Trustees, and contrary to HMRC’s position, that HMRC have consistently treated the management of group pension funds by an Insurer as insurance business falling within the “insurance transactions” exemption) give rise to circumstances rendering claims against the Non-Insurers as impossible or excessively difficult.

There are three other points. The first is that there are no special factors unrelated to the merits of the claim on which the Trustees rely to support the contention that a claim against the Non-Insurers would be impossible or excessively difficult.

The second point is that it was only after the decision in Kleinwort Benson Ltd v Lincoln City Council [1999] 2 AC 349 that English law clearly recognised a restitutionary claim based on a mistake of law. It is not necessary to consider what might have been the position prior to that decision so far as concerns the Trustees’ remedy to enforce their directly effective rights. This is because the claims in the present case were all commenced after that time. The relevant time at which to identify the available remedies must, at the earliest, be when the proceedings were commenced, by which time a claim by the Trustees against the Non-Insurers based on a mistake of law was in principle available.

The third point relates to limitation. The limitation period in relation to the Trustees’ claim against the Non-Insurers is not entirely clear, and is not a matter which has been the subject matter of submissions. As will be seen when I come to Issue 4, I consider that the limitation period relevant to a claim by the Trustees against HMRC (were I to be wrong on both Issues 1 and 2) would be 4 years and not 6 years. Similar arguments to those which lead to my conclusion on Issue 4 could be deployed in relation to the Trustees’ claims against the Non-Insurers.

However, it is at the very least possible that the Trustees’ claim against the Non-Insurers is subject to the relevant period under the Limitation Act 1980 (6 years under section 5 or an extended period under section 32(1)(c)) rather than the 4-year period under section 80(4) whereas the Non-Insurers’ claim against HMRC is subject to the shorter 4-year period. There is therefore the possibility that a supplier in the position of the Non-Insurers might find himself subject to a claim brought by a customer within the limitation period applicable to that claim but after the period of 4-years available to the supplier in relation to his own claim against HMRC. The answer to that may well be that the supplier would have a change of position defence. But whether or not he does, the customer (the Trustees in the present case) has no argument relating to this aspect of limitation which makes his claim impossible or excessively difficult.

In my judgment, the answer to the second limb of Issue 2 is that it would not be impossible or excessively difficult for the Trustees to obtain reimbursement from the Non-Insurers.

Accordingly, the answer to Issue 2 is that it is not “impossible or excessively difficult” as a matter of EU law for the Trustees to obtain reimbursement of the wrongly charged VAT from the Non-Insurer suppliers of pension fund management services.

Issues 3 and 4 do not arise. There is a limit to the extent to which I should go in answering questions which are not necessary for my decision. I shall therefore say very little about them.

If I am wrong on Issues 1 and 2, what remedy do the Trustees have? Clearly, they then have a remedy against HMRC, since that is the only way in which their directly effective rights can be vindicated. It would not, I think, be possible to adapt or mould section 80 so as to give the Trustees a direct right against HMRC under section 80(1). They must instead be given a right akin to a claim in restitution, with section 80(7) being disapplied to enable that right to be exercised.

The answer to Issue 3 would be that the Trustees are to be given a restitutionary remedy with section 80(7) being disapplied but subject to the time-limit issue, which is Issue 4 to which I now turn.

The time-limit in relation to a direct claim against HMRC would prima facie, be 6 years under section 5 Limitation Act 1980 but subject to postponement under section 32(1)(c). The time limits for claims under section 80 is 4 years: see section 80(4). That time limit, it is common ground, is compatible with EU law in the context of the ordinary application of section 80. Mr Macnab submits that the 4-year time limit should be applied to any claim which the Trustees have against HMRC. There is some sense in the result, since it produces a consistency between the time limit for claims, whomsoever they can properly be made against. Just as the 4-year time limit in relation to claims against HMRC under section 80 is compatible with EU, so too it would be compatible in relation to common law claims against HMRC as a matter of EU law. There could have been no objection to the enactment of an express time-limit to that effect.

To illustrate the sense of that result, consider a claim by a customer against his supplier, where VAT has been incorrectly paid. Suppose that the claim would in principle be a simple claim but on the facts is virtually impossible because the supplier has gone into insolvency with only a small dividend to each creditor. Absent insolvency, the customer would claim from the supplier. If he did not do so within the 4-year time-limit provided for under section 80(4), it is well-arguable that the supplier would have a change of position defence to the claim, namely that he had not made a claim against HMRC and his claim had become time-barred and that it would be wrong in principle to allow the customer to obtain a remedy against HMRC when, but for the insolvency, his claim would have been against the supplier, and unenforceable because of the change of position defence.

Can that result be achieved? It might, on one view, be capable of being achieved as a matter of interpretation of section 80(7). I do not consider that this result can be achieved as a matter of interpretation, even adopting the highly muscular approach which the English Courts have developed in applying the Marleasing principle. To interpret section 80(7) as not covering a common law claim by a third party against HMRC would go against the grain of the provision, whose purpose is precisely to exclude all such claims.

In order to give effect to the Trustees’ directly effective rights, section 80(7) therefore has to be disapplied. The question then is the extent to which it has to be disapplied. Since the purpose of any disapplication is to effect compliance with EU law, it is arguable that the disapplication should not go further than is necessary to achieve that purpose. Since, within its own provisions, section 80 lays down a time-limit of 4 years for bringing claims, there is a strong argument for saying that the disapplication should not go further than allowing claims brought within the same 4-year period.

In this context, the well-known decision in Fleming provides some help. The issue there concerned the shortening of an existing time limit for making claims without providing for any transitional period. It was recognised that, although it was the duty of the United Kingdom courts to disapply the relevant regulation in that case to the extent necessary to enable claims based on accrued rights to be brought within an appropriate period, it was open to the domestic court in a suitable case to makes its own assessment of what, in accordance with Community law, was an adequate transitional period. It was held by the House of Lords that the English Court could not, on the facts of the appeals, incorporate a reasonable transitional period where none had been provided for.

Although the situation in Fleming was very different from that in the present case, I consider that it supports the proposition that, although it is the duty of this Court to disapply section 80(7), such disapplication would be only to the extent necessary to enable claims to be brought within an appropriate period. In contrast with the position in Fleming where the facts were such that the Court could not incorporate a reasonable transitional period into the relevant provision, the facts of the present case are very different, in that section 80 already provides a certain time limit. In Fleming, the courts would have had to select a reasonable time limit but, until it did so, there would be uncertainty and taxpayers would not know where they stood. It was not appropriate for the court to reach its own decision on what would be a reasonable time for making claims: that was a matter for Parliament or possibly HMRC by means of an announcement disseminated to all taxpayers. In the present case, however, Parliament has already shown clearly the period which it considers appropriate for making claims against HMRC to recover overpaid VAT. It would entirely appropriate and certain to allow recovery by persons with a directly effective right who do not themselves have a right to reimbursement under section 80 (such as the Trustees) to be subject to the same time limit as those who can rely on section 80 (such as the Non-Insurers).

My view is that section 80(7) can be disapplied in this way so as to limit any common law claim by the Trustees against HMRC to claims brought within the same period of 4 years as applies to a claim under section 80 by the Non-Insurers. Of course, this issue arises only if I am wrong on both Issues 1 and 2. It is not a matter of decision, but only the expression of an obiter view which may be of assistance to the parties or to an appellate court should matters proceed that far.

The answer to Issue 4 is that the Trustees’ claims, if any, against HMRC would be time-barred after the 4-year period provided for in section 80(4).

At the outset of the hearing, Mr Scorey raised a pleading point against HMRC. He said that it was a necessary part of HMRC’s case that the treatment which they had afforded to Insurers (namely exemption on supplies of pension fund management services) was unlawful, in that it did not comply with the requirements of EU law. Such a case had not been pleaded. If HMRC seek to rely on the point, they should plead it. Mr Macnab contends that it has never been any part of HMRC’s case that Insurers were entitled to the exemption. HMRC’s position is as follows:

Nonetheless, it is the case that HMRC continue to allow Insurers to rely on the exemption in relation to the provision of pension fund management services, and will do so until a change of practice which is anticipated to take place in 2018.

I think Mr Macnab is correct to say that HMRC’s pleadings rely only on the proposition that Non-Insurers are not entitled to the exemption and that they do not assert that Insurers are entitled to the exemption, although the arguments which lead to the conclusion that Non-Insurers are not entitled to the exemption may also lead to the conclusion that Insurers are not entitled to it either. There is nothing in the pleadings to preclude HMRC from arguing that Non-Insurers are not entitled to the exemption as a matter of EU law. And, since I do not consider that the principle of fiscal neutrality requires Non-Insurers to be granted an exemption to which they are not entitled simply because it is given, whether lawfully or not, to Insurers, HMRC do not need positively to contend that the exemption is not available to Insurers.

In any event, if it were necessary for HMRC to amend, I see no reason why they should not be granted permission to do so. The arguments of substance before me would have been precisely the same as they actually were and the Trustees have not been prejudiced in any way. It is entirely academic, therefore, whether an amendment is in fact needed as a matter of formality.

There is no need to make a reference to the CJEU. In my view, the answer to Issue 1 is acte clair and no reference is required, although, if that is wrong, then, of course, in order to know the answer to Issue 1, it would be necessary to make a reference. However, my decision on Issue 2 results in the Trustees having no claim against HMRC even if the activities in question are “insurance transactions”. The Trustees’ claim (for the full £100 in the illustrative example) is against the Non-Insurers who, in turn, have a claim (for the £75 in the illustrative example) subject, in each case, to any limitation defence available to the Non-Insurers or HMRC (in the latter case under section 80(4)). The remedies to give effect to directly effective rights are matters of domestic law. I consider that there is no question which requires a reference in order to answer Issue 2. Since the action can be disposed of on the basis of Issue 2 without a reference, there is no call for a reference on Issue 1. If this matter goes further, and if the Court of Appeal considers that I am wrong on Issue 2, it can make a reference in relation to Issue 1 if, contrary to my view, it considers the answer is not acte clair.

The Trustees’ action is dismissed.