4—A Hierarchy of Reasons for Restitution

4

A Hierarchy of Reasons for Restitution

IN THE PREVIOUS chapter it was argued that once the true nature of the public law unjust factor is understood, as outlined in chapter two, not only does this provide us with the answers to several questions about the scope and operation of that unjust factor, it also means that there is then no reason to continue to separate cases in which the public body is the defendant from those in which it is the claimant. In all cases, if there has been an ultra vires transfer of money which constitutes an event in public law, this can provide the reason for restitution in the private unjust enrichment claim. However, so far the claim has generally been a descriptive one about the capabilities of the public law reason for restitution, rather than a normative one about its proper scope of application. It will be argued here that these descriptive conclusions can have normative force: not only can public law provide a more straightforward reason for restitution in all such cases, the second important implication of the true nature of the public law reason for restitution is that it should do so.

Essentially, public bodies may become involved in unjust enrichment claims in five different sets of circumstances as follows:

(1) A public body can behave in exactly the same manner as would a private individual. In the context of contractual claims, it is suggested that public bodies should be treated the same as private individuals when all they are doing is purchasing paperclips for use in their offices.¹ Perhaps the unjust enrichment equivalent of this is restitution of the price paid for the paperclips following breach of the contract, using, for example, total failure of consideration. This brings us to the second situation.

(2) A public body can behave in a ‘public’ manner, for example by levying tax or raising resources under the private finance initiative, but again, the basis for the transfer may fail. This gives rise to a further distinction between two different scenarios:

(a) A private party and a public body enter a contract, which is terminated for some practical reason, such as a failure of the private party to complete the job given to it by the public body. The validity of the initial contract is not in question.

(b) A private party and a public body enter a contract, which is terminated because it was ultra vires the public body. The swaps cases can fall into this category. Both failure of consideration and the validity of the public body’s actions are relevant.

(3) It is also possible for public bodies acting in a ‘public manner’ as above to become involved in a claim based on mistake of law, for example where a private party misunderstands a taxing statute and pays too much tax, or a levying public body misunderstands a taxing statute and charges too much tax, as in British Steel plc v Customs and Excise Commissioners.2 In such cases, the validity of the particular decision is in question, but the validity of the underlying legislation is not. There are, of course, also cases where the whole concept of the transaction or the legislation is ultra vires, but one of the parties has paid money in the belief that it was intra vires. The swaps cases can fall into this category, as do ‘mistaken Woolwich’³ cases such as, apparently, Deutsche Morgan Grenfell v IRC,⁴ where the trusting private party simply pays the tax believing it to be validly demanded.

(4) Alternatively, it is possible for public bodies behaving in a ‘public’ manner, as above, to become involved in a claim based on a mistake of fact, for example where a tax is charged on red-haired employees and the claimant pays, believing it has such employees when in fact it does not.

(5) Finally, a public body can behave in a ‘public’ manner, and then become involved in a claim in which the unjust factor used by or against it is the public law reason for restitution, involving a combination of public and private law events. Woolwich⁵ falls into this category, as could the Auckland Harbour Board⁶ case and the swaps cases following the re-analysis conducted in chapter three.⁷

It is assumed that in the first case, as in the parallel contract and tort cases, the claim can be viewed from an entirely private law perspective and that the claim need not be based on the public law unjust factor. At the other extreme, of course, is situation (5) in which the public law reason for restitution provides the basis for the claim, which is thus a hybrid of public law and private law reasoning. The claimant must prove that the defendant has been enriched, at the claimant’s expense, and that the defendant cannot plead any of the defences to unjust enrichment (requirements of the private law event of unjust enrichment). The claimant must also prove that the enrichment was unjust (private law requirement), which he, she or it does by proving that a public law event has occurred (public law, the key to this possibility, having been mistranslated in the past as transactional inequality, incapacity and so on).

The Advantages of the Public Law Reason for
Restitution over the Private Law Unjust Factors

First, as has already been suggested in chapter three, this is more desirable in the sense that claims by and against public bodies can thus be brought together and in each case the dual public and private nature of the claim can be addressed directly. Second, as was also seen in chapter three, it is in addition more straightforward in many cases to allow recovery on the basis of the public law illegality itself, rather than requiring the claimant to show a failure of consideration, or mistake, especially when, for example, the existence of a mistake is not wholly uncontroversial,8 or there are other difficulties involved in classifying such a hybrid claim under one of these purely private headings.⁹

However, the argument here is more than just a negative assertion that where more than one unjust factor is available the public law reason for restitution is coincidentally more straightforward in pragmatic terms than the private law reasons for restitution. The point is that only by regarding public law as the reason for restitution does the law gain full access to that crucial event in a manner which allows it to take the truly hybrid approach such claims require. If the action is brought on the basis of mistake or failure of consideration, the public law event may be equally crucial to the claim, but the court will be prevented from seeing the case as anything more than purely private and this is bound in turn to lead to private law mistranslations of concepts, or to the separation of cases which actually entail similar policy concerns in precisely the manner seen in the previous two chapters. It is no coincidence that a round peg fits in a round hole better than does a square peg; the public law unjust factor avoids the problems associated with the private law unjust factors for the simple reason that only it deals with the true reason for restitution. Private law unjust factors may provide a reason for restitution, but as Meier puts it, in most cases they simply end up mirroring the underlying reasons why the transaction was invalid in the first place.10 Thus even if the claimant can frame the action in terms of a private law unjust factor, if the case involves a public law ultra vires event this should take precedence, and only the public law reason for restitution should be used. Certainly this should be the case when the alternative private law reason for restitution rests directly on the public law event (for example, where the ultra vires nature of a contract constitutes a failure of consideration or the subject of a mistake) but it is even arguable that this should be the case when the connection is more remote. For example, let us take the idea suggested above, that a tax is created which will charge companies money based on the number of red-haired employees they have.¹¹ Were a company to believe that it employs some red-haired people when in fact it does not, receipt by the taxing authorities of the resulting payment would be beyond its powers on two fronts: first because as a matter of fact no tax was due even if the rules are applied on their own terms¹² and second because of the patent invalidity of those rules in the first place. The former argument could alternatively be heard as an issue of mistake, but the latter could not. However, if the mistake ground were to be used, neither of the two forms of ultra vires receipt would come to light, whereas both forms would become evident if this was the focus of the investigation. If there is a public law ultra vires event of this nature, it is surely in the public interest that this be discovered as soon as possible and thus it is arguable that even here the company should have to use the public law unjust factor in preference to its alternative ground of mistake. This would also make operation of the hierarchy simpler: wherever the receipt or payment of money by a public body is ultra vires as a matter of public law, the reason for restitution of the money should be the existence of that ultra vires event. Of course, there may well be cases like the example just given where in any event the company would presumably not choose only to reclaim a few counts of tax if it had the option of challenging its liability to pay any of it. But there will also be cases in which the hierarchy proposed does not operate to the claimant’s advantage.

Reasons for Restitution and Time Limits:
Is a Hierarchy Possible?

In some cases,13 the claimant may have a very good reason for trying to use the ground of mistake in preference to the public law reason for restitution. The Limitation Act 1980 does not clearly set out limitation periods for unjust enrichment actions.¹⁴ Such cases are taken, therefore, to fall within the provisions for contract in section 5. This means that claims in unjust enrichment must be brought within six years from the date on which the defendant became unjustly enriched. However, according to section 32 (1)(c), where ‘the action is for relief from the consequences of a mistake’, the time will not begin to run until ‘the plaintiff has discovered . . . the mistake . . . or could with reasonable diligence have discovered it’, and it was confirmed in Kleinwort Benson Ltd v Lincoln CC that this applied to the recovery of money paid by mistake of law.¹⁵ Recently, a debate has arisen over whether it is only in such cases, where the mistake is the basis of the claim, that the 32(1)(c) extension will apply, or whether, as Edelman argues, it can apply in any case where the relief is from the consequences of a mistake, whether or not that mistake is itself the basis of the claim.¹⁶ Obviously, if Edelman’s argument is correct, then one, perhaps even the most, important objection to the proposed hierarchy is immediately removed, and thus the hierarchy might well then come about naturally on the basis that the public law reason for restitution would just be a simpler route to recovery than having to prove the same ultra vires event plus mistake, failure of consideration etc in any given case. However, the law currently adopts the opposite view,17 and since this is also the approach which causes the most difficulties for the hierarchy proposed here, let us assume that this is correct and proceed on the basis that a claimant could be outside the six year limitation period applicable to the public law unjust factor, but nevertheless able to bring a claim by relying on the ground of mistake as to vires. This is possible even when the same party is responsible both for challenging the validity of the payment and for bringing the action in unjust enrichment, but it is even more likely to arise where an application for judicial review brought by Party A is successful and Party B wishes to rely on this decision to reclaim the money it paid more than six years ago. Indeed, as has already been seen (and will be seen again below when possible defences to the claims are discussed)¹⁸ one of the most important concerns in this area is the possibility that by establishing the public law event once, the first claimant will open the ‘floodgates’ to a host of other claimants basing themselves, apparently, on only the private law event of unjust enrichment, as was the case in Kleinwort Benson v Lincoln itself (with Kleinwort relying on the Hazell v Hammersmith decision¹⁹). This was also an issue of great concern in the more recent case of Deutsche Morgan Grenfell,²⁰ where one of the key questions was whether Deutsche Morgan Grenfell should be allowed to rely on a decision of the European Court of Justice (ECJ) to evade the usual time limits in order to recover money it had paid more than six years previously.²¹ Consideration of Deutsche Morgan Grenfell is therefore vital in making the case for the hierarchy proposed here.

Deutsche Morgan Grenfell v IRC

Deutsche Morgan Grenfell (DMG) had paid tax to the Commissioners of Inland Revenue (IRC) in 1993, 1995 and 1996, in the manner required by the English rules. The rules were that companies whose parent was resident in England, as well as companies whose parent was not, had to pay Advanced Corporation Tax (ACT)22 and Mainstream Corporation Tax (MCT).23 The ACT was payable first, but could then be set off against the MCT paid later. Under the relevant legislation, two companies resident in the UK, one of which held at least 51% of the other could make a group income election. The result of such an election was that the subsidiary could choose not to pay ACT on dividends which it paid to its parent company. However, for companies whose parent was not resident in the UK, this election was not an option, so ACT was payable on all the distributions made by them to their parent companies. Overall, they paid no more than companies whose parent companies were resident in the UK, because the ACT they had paid was later offset against their MCT, but it did mean that in the meantime they had lost interest on the funds used to pay the ACT up front, when subsidiaries with parent companies resident in the UK would have been able to retain the money.

In 2001, two other companies that had suffered under these rules, Metallgesellschaft and Hoechst (M&H), brought a case claiming that the English rules were contrary to EC law. A preliminary reference was made to the ECJ under the procedure laid down in what is now Article 267 of the Treaty on the Functioning of the European Union (TFEU) (ex Art 234 EC) and, in another key landmark case which has since generated a plethora of further litigation,²⁴ the ECJ agreed that:

[I]t is contrary to Article 52 [now Art 49 TFEU ex Art 43 EC] of the Treaty for the tax legislation of a Member State, such as that in issue in the main proceedings, to afford companies resident in that Member State the possibility of benefiting from a taxation regime allowing them to pay dividends to their parent company without having to pay advance corporation tax where their parent company is also resident in that Member State but to deny them that possibility where their parent company has its seat in another Member State.²⁵

And that:

[It] is contrary to Community law for a national court to refuse or reduce a claim brought before it by a resident subsidiary and its non-resident parent company for reimbursement or reparation of the financial loss which they have suffered as a consequence of the advance payment of corporation tax by the subsidiary, on the sole ground that they did not apply to the tax authorities in order to benefit from the taxation regime which would have exempted the subsidiary from making payments in advance and that they therefore did not make use of the legal remedies available to them to challenge the refusals of the tax authorities, by invoking the primacy and direct effect of the provisions of Community law, where upon any view national law denied resident subsidiaries and their non-resident parent companies the benefit of that taxation regime.²⁶

DMG therefore wished to recover the money it had lost by paying the ACT up front.

It is thus clear that the discriminatory unavailability of the group income election was contrary to Article 49 TFEU (ex Art 43 EC), and the fact that the companies had not tried to make the impossible election or challenge its unavailability without paying the ACT did not block their claims. It is also clear that some form of ‘remedy’ for this unavailability of the election was required at national level in order to comply with the requirements of EC law.27 This gives rise to various important questions: whether the money could be recovered on the basis of unjust enrichment at all, and if so, what should be the correct reason for restitution; or whether the money should instead be recovered as compensation for a loss.

Can the Money Be Recovered as Restitution for Unjust Enrichment at all? Or is Compensation for Loss the Only Route to Recovery?

It is true that none of the previous unjust enrichment authorities dealt with the precise question of loss of interest in this way as opposed to recovery of a principal sum. However, it is possible to draw some evidence from the dicta of Lord Goff and Lord Woolf in Westdeutsche Landesbank Girozentrale v Islington LBC²⁸ that the claim for interest should be seen as one based on unjust enrichment; that full restitution is not achieved unless the correct level of interest is paid. Alison Jones agrees with this view, citing Harbutt’s ‘Plasticine’ Ltd v Wayne Tank and Pump Co Ltd²⁹ in support of her conclusion that interest is not only about compensating a claimant, but is also designed to strip the defendant of any unjust enrichment.³⁰ Certainly, although the independent source of the enrichment (such as payment of interest by a bank) may seem to cloud the issue, put this way both the ‘enrichment’ and ‘at the expense of the plaintiff’requirements can be satisfied. These conclusions have since been confirmed by the decision of the House of Lords in Sempra Metals Ltd v IRC,³¹ another case in the Metallgesellschaft and Hoechst line, which held that the measure of restitution was gain to the defendant rather than loss to the claimant.³²

The more problematic question is whether the early payments of ACT (and thus the loss of interest) could also be regarded as undue, and thus unjust. Obviously the unavailability of the group income election, being a contravention of Article 49, was ultra vires the IRC. On the other hand, the ECJ’s decision did not invalidate the whole ACT/MCT scheme and nor did it refer to the specific payments of ACT as being invalid. Indeed, in Metallgesellschaft and Hoechst the ECJ held that:

It is important to bear in mind in this regard that what is contrary to Community law, in the disputes in the main proceedings, is not the levying of a tax in the United Kingdom on the payment of dividends by a subsidiary to its parent company but the fact that subsidiaries, resident in the United Kingdom, of parent companies having their seat in another Member State were required to pay that tax in advance whereas resident subsidiaries of resident parent companies were able to avoid that requirement.33

Again it reiterated that ‘the breach of Community law arises, not from the payment of the tax itself but from its being levied prematurely’.³⁴ It could thus be argued that only the absence of the election was ultra vires the IRC. On that basis, once the election had not been made, for whatever reason, the ACT would remain payable. It was for this reason that when the issue reached the House of Lords, Lord Scott, drawing on the views of Stevens,³⁵ dissented from the majority decision,³⁶ a conclusion he reiterated in Sempra Metals.³⁷ Key to Lord Scott’s decision was his finding that:

The ECJ did not hold that subjecting resident subsidiaries of non-resident companies to the obligation of paying ACT was contrary to European law. What was unlawful was the discriminatory nature of the group income election provisions . . . The question is what effect this unlawful feature of the ACT tax regime should be held to have had on the obligation of subsidiaries, such as DMG, to pay ACT.³⁸

For Lord Scott the answer to this question was that because

the breach of Community law was the failure of the UK’s ACT tax regime to provide for corporate groups with a parent company resident in a Member State other than the UK the same opportunity of postponing the payment of corporation tax by making a group income election as was available to domestic corporate groups . . . The unlawfulness of the tax regime identified by the ECJ lies not with the imposition of ACT but in the discriminatory nature of the group income election provisions.³⁹

If instead the courts were to conclude that the ACT was unlawful ‘then it would follow that every demand for ACT had been an ultra vires demand, that every payment of ACT had been paid under a mistake of law and that every payer of ACT should be entitled, subject to limitation defences, to a restitutionary remedy’.40 Since ‘there must be many domestic corporate groups which, for one reason or another, did not make a group income election and whose subsidiaries accordingly paid ACT in respect of dividends paid to their corporate shareholders’⁴¹ this conclusion would lead to ‘a remarkable state of affairs’ in which such domestic corporate groups could recover the money on the basis of what Lord Scott called ‘the Woolwich principle’. Thus, while Lord Scott agreed with the majority that ‘English law does now recognise a restitutionary remedy for tax paid under mistake of law’,⁴² money would not be recoverable on this ground unless the mistake had also undermined the legal obligation which required the payment of the money.⁴³ He therefore concluded that DMG could not recover their payments of ACT on the basis of mistake because even if the mistake about the legality of denying the group income election had been a ‘but for’ cause of payment of the ACT, ‘the ACT was paid because there was a legal obligation under a valid statutory provision for the money to be paid’.⁴⁴ Thus, if DMG were to recover at all, for Lord Scott it would have to be via a claim for compensation to recover the loss caused by the breach of EC law.⁴⁵ This would in turn mean that the extended time limit of section 32 of the Limitation Act 1980 would not be open to them, because mistake would not then be an essential element of their claim.⁴⁶

In order to understand this point fully we should imagine three different companies.

Company A has its parent in the UK and is thus able to make an election if it wants to. It chooses for its own reasons not to do so and thus pays ACT. Under the ECJ’s ruling, this receipt of ACT by the Revenue is perfectly intra vires, because the ECJ did not invalidate the whole scheme of elections, ACT and MCT.⁴⁷

Company B is in the situation of DMG and M&H. It has its parent abroad and therefore cannot make the election, but would have preferred to do so and thus does not want to pay ACT. Lord Scott’s view is that since the demand for and receipt of the ACT was not ultra vires in the case of Company A, it cannot be ultra vires in the case of Company B either. With respect to Lord Scott, however, it is submitted that this does not follow and that, unlike Company A, this time the collection of the ACT is invalid and ultra vires as a matter of EU law. It is true that, as Stevens notes,48 the ECJ did not ‘say that the tax was not due’. He is also right to point out that the decision on whether UK legislation is compatible with EU law is one for the national courts.⁴⁹ However, the ECJ made it clear that what was contrary to Community law was the fact that only domestically-resident companies could ‘benefit’ from the ACT/MCT system,⁵⁰ and the ability to make the group income election is of no ‘benefit’ at all to the company without the resulting ability to defer payments of ACT. So the ECJ’s references to ‘benefits’ must surely refer not only to the ability to make the election but also to the ability to avoid paying the tax until later as a result of that election. It was this whole, combined process of ‘affording’ only resident companies ‘the possibility of benefiting’ that was invalid as a matter of EC law. Payment of ACT by a company which could not have made the election thus becomes invalid precisely because that company, contrary to EC law, has had no option to act differently, and therefore it does follow that because the denial of an election was unlawful, any ACT paid as a result was not due.

As ever, it is important to recognise the public law nature of this finding of ultra vires, which explains why it may have a broader scope than, for example, the setting aside of a transaction in purely private law, to which Lord Scott kept comparing it. Thus, Lord Scott argued that it would not be enough for the buyer of a horse to assert that but for his mistake about the horse’s breeding he would not have entered into the contract to buy it and so would not have paid the money.⁵¹ This may well be the case, but there is nothing to suggest that what is necessary for the setting aside of a contract in private law must be the same as what is necessary for a finding that something was beyond the powers of a public body, nor that the extent of those two invalidities must be the same.⁵² Indeed, when the different policy considerations underlying the two sets of rules are considered, it would not be wholly surprising to find that the rules for setting aside contracts were the more stringent of the two. When it is recalled that, in some instances, public actions may be rendered ultra vires even though the illegality tainting the action has had no causative effect whatsoever,⁵³ the suggestion that an illegality which has had such a causative effect may have rendered a further action (here the charging of ACT) invalid, does not seem particularly strange. Indeed, in Fleming (trading as Bodycraft) and Condé Nast Publications v Revenue and Customs Commissioners,54 the House of Lords, including Lord Scott himself, unanimously held that the invalidity of the relevant legislation in that case left taxpayers with an ‘accrued right’ even if they would not have taken advantage of it had it been available to them, let alone if, like Company B, they would have done so. Lord Neuberger’s view was that the legislation in question was intended to be for the benefit of anyone who could take advantage of it, whether or not they in fact did so.⁵⁵ It is submitted that this is perfectly in keeping with the fact that what was being discussed was the extent, under EC law, of the Customs Commissioners’ powers in public law to enact rules. Indeed, as will be seen in the discussion of Company C below, the question is not whether the public law invalidity is wide enough to cover Company B, but whether it is so wide that it covers Company C as well. To ask whether receipt of a tax can be set aside as if it were a transaction in private law is therefore, with respect, neither here nor there. The real question is whether receipt of the ACT was beyond the powers of the taxing authorities and, for the reasons given here, it was.

Stevens’ concern is that this is a ‘radical conclusion as it means that the effect of European Community Law is not merely to render legislation of the United Kingdom unlawful but also that it is, to an extent, a nullity’.⁵⁶ However, this is not quite the way in which EU and national law interrelate. In Ministero Delle Finanze v IN.CO.GE.’90 Srl,⁵⁷ the ECJ rejected the Commission’s argument that a national rule imposing unlawful charges should be treated as invalid rather than inapplicable. There is therefore no question of a decision of the ECJ making national law a nullity; all it can ever do is to identify a potential incompatibility. This is then applied to the facts of the case by the national court, and if the incompatibility is found by that court to exist on the facts, it renders the national law inapplicable during that country’s membership of the EU.⁵⁸ Here, given that the denial of the benefit of the ACT/MCT scheme was incompatible with EC law, the form of this denial (namely charging companies the ACT upfront following a refusal of the group income election option) was, as a matter of UK law ultra vires the Revenue on the basis that it contravened the will of Parliament as expressed in the European Communities Act 1972.⁵⁹

The more difficult question, as noted above, is precisely how broad the scope of this finding of illegality should be. It is relatively easy to argue that, as in the case of DMG itself (ie Company B) the receipt of ACT was ultra vires where the company had wanted instead to make a group income election and was denied the ability to do so in breach of EC law. The harder case concerns Company C. This company has its parent abroad and so would not have been able to make the election, like Company B, but it would in any event have chosen not to do so, like Company A. Stevens’ arguments imply that if we are to regard the ACT as undue in the case of Company B, we must also regard it as undue in the case of Company C.60

If the invalidity only extends to the denial of the election then it is not ultra vires the Revenue to receive money from a company which would not have made the election anyway because the payment is nothing to do with the ultra vires refusal of the election. Should the company nevertheless be allowed to use the fortuitous application of EU law to escape from what is essentially a bad bargain with the Revenue in such a case? There are good reasons why recovery in such cases should not be allowed; as Henderson J puts it in Europcar UK Ltd v The Commissioners for HM Revenue and Customs,⁶¹ another of the Metallgesellschaft & Hoechst⁶² progeny, ‘if the election would not have been made, even if it was available, the claimant would have nothing to complain about’.⁶³ The question is therefore whether recovery is indeed automatic in the case of Company C on the argument made here. From the point of view of the private law elements of the claim, it is difficult to argue that the ultra vires denial of the group income election caused Company C’s subsequent payment of ACT as the unjust enrichment elements of the hybrid claim would presumably require.⁶⁴

As far as the reach of the public law through the public law unjust factor is concerned, no answer to this question was given by the ECJ and it must therefore turn on the extent of the invalidity as a matter of national public law. It may be the case that any demand for ACT which follows a denial of the group income election is ultra vires, but this will not necessarily be so. Although, as noted above, in some cases a public body’s action can be set aside even when its ultra vires nature had no practical effect on the outcome,65 conversely not all technical invalidities in a public body’s actions will lead to the decision being overturned if the court considers that this would be inappropriate,⁶⁶ and in particular, a flaw in the public body’s decision will not always be regarded as materially vitiating that decision if it would have made no difference to the outcome of the case.⁶⁷ Obviously the situation is slightly different where the question is not what effect the flaw had on the reasoning of the public body, but rather what effect it had on the reasoning of the taxpayer. Nevertheless, the materiality cases suggest that by analogy the ultra vires refusal of the election may not necessarily have tainted the collection of ACT from a non-elector like Company C. A second, interconnected, circumstance in which the decision will not be struck down occurs where there has been no ‘prejudice’ as a result of the public law wrong⁶⁸ and here again it is difficult to see what prejudice Company C might have suffered. Finally, and perhaps most relevantly, there are certain circumstances in which a party can waive its rights, for example, to a fair hearing. Any such waiver ‘must be clear and unequivocal and made with full knowledge of all of the facts relevant to the decision whether to waive or not’.⁶⁹ Of course the decision made by Company C would have to comply with this requirement, but as long as it had been aware of the discrimination and had made the decision to pay ACT purely for its own reasons, the requirement would presumably be met. Thus, taking these three sets of rules together, it is at least possible to argue by analogy that the receipt of ACT from Company C would not be tainted with the ultra vires refusal of the election to that company and thus it would not be possible to argue that there had been a public law event capable of providing the reason for restitution in that case.

On the other hand, support for the opposite point of view comes from the decision of the House of Lords in Fleming and Condé Nast.70 This case concerned the introduction, without any transitional period, of a retrospective time limit for the reclaiming of VAT in a manner which was contrary to EC law,⁷¹ and the Customs Commissioners argued that only those who could and would have made claims during a transitional period had one been available should be entitled to complain that it was not. However, as noted above, this argument was unanimously rejected by their Lordships,⁷² Lord Neuberger holding that it was ‘wrong in principle and inconvenient in practice’,⁷³ while Lord Walker similarly held that

it would be contrary to legal certainty and administratively unworkable, for the extent of disapplication to depend not only on the duration of the transitional period but also on an hypothetical question to be answered by reference to the circumstances and states of mind of the particular taxpayers.⁷⁴

Following this logic, therefore, Company C would also be entitled to bring a claim on the basis that it would be ‘unworkable’ to try to distinguish between Companies B and C on the facts of a given case.

The conclusion must therefore be that the Company C dilemma simply provides us with yet another example of something which may be novel to private lawyers, but which is wholly familiar to the public law which ought also to be operating in this context, and which already contains, as outlined above, a set of tools to deal with this problem. On the one hand, allowing Company C to recover provides a more efficient result in the sense that if Company B would have made the election and therefore will have a good claim at a future point⁷⁵ it is better that the Revenue knows this as soon as possible, even if the case is brought by Company C. Otherwise the Revenue will continue to act in an ultra vires manner, opening up the possibility of further challenges, in the period before Company B claims. Also, as noted above, in principle, the whole point of imposing limits on the powers of public authorities is that correct application of the relevant rules (whether they concern time limits or the rules for levying tax in the first place) is ‘intended to be for the benefit of anyone who could take advantage’ of them,⁷⁶ whether or not they in fact do so, or, more broadly, whether or not the invalidity makes a practical difference to the outcome of the case.⁷⁷ However, in public law, these principles are also counterbalanced by the doctrines such as materiality and waiver discussed above. Again, therefore, it is regrettable that private lawyers are left reinventing the public law wheel, when an understanding that these cases involve both public and private law would have allowed them to draw on the experience of a sphere of law already used to having to balance such competing concerns. Taking a hybrid approach which involved the use of public law in such cases would not avoid the dilemma, but it would provide a well established set of tools for resolving it and would allow the courts to develop their reasoning on the extent of public law invalidity in a manner which would increase both understanding and certainty in future cases.

Of course, as will be discussed further below, it may be necessary for the causal link between the ultra vires breach of EU law and the loss suffered by the claimant to be closer in the case of an unjust enrichment action than it need be for a Francovich and Brasserie/Factortame III 78 claim for state liability in damages of the kind favoured by Stevens and Lord Scott.⁷⁹ But however strict the causation rules for unjust enrichment claims may be, Company B and DMG can be regarded as falling within them, first for the reasons given above and second because there is nothing in the more recent case law on this question to suggest otherwise.⁸⁰

Conversely, while on the question of causation an action in state liability for damages may be slightly wider than a cause of action in unjust enrichment, an action for damages still requires the existence of an ultra vires event, and it should be borne in mind that in addition, a Francovich and Brasserie/Factortame III action also requires the ultra vires action to have been a sufficiently serious breach of EU law which amounts to a manifest and grave disregard of the limits of the member state’s discretion.⁸¹ In the case of Deutsche Morgan Grenfell itself this could be made out fairly readily given the overt nature of the discrimination, but this need not always be the case. Indeed, in Test Claimants in the FII Group Litigation v Commissioners of the Inland Revenue⁸² (FII) AG Geelhoed pointed out that not only could actions in unjust enrichment arise in circumstances where the ‘sufficiently serious breach’ condition would not be fulfilled, but also, for precisely this reason, the Member State’s obligation to provide an effective remedy for a breach of EU law might require it to provide ‘reimbursement’ instead.83 It therefore appears that to adopt the solutions proposed by Stevens and Lord Scott would be to make life potentially harder for claimants such as DMG, with no corresponding benefits either to them or to the clarity of the law in general. This does not mean that the claim cannot be classified as a wrong, merely that it need not be; that, following the FII decision, classification as a wrong and exclusion of the unjust enrichment cause of action might be a breach of the EU principle of effectiveness, and that conversely the claim can, without undue danger, be more simply classified as an unjust enrichment. And, it should not be forgotten that it was the unjust enrichment conception of the facts that was preferred by AG Fennelly in Metallgesellschaft and Hoechst, the ECJ decision which gave rise, in the first place, to the issue in Deutsche Morgan Grenfell and other related litigation.⁸⁴ Any problems, therefore, arise not from the adoption of unjust enrichment as opposed to a cause of action based on wrongdoing. The problem, as usual, arises from a failure to understand that the approach to the claims must be a hybrid one, incorporating both public and private law. Indeed, once it is thus assumed that, at least on the facts of Deutsche Morgan Grenfell itself (Company B), the ACT can be regarded as ultra vires and undue, it then becomes possible to address this key question directly; the choice of unjust factor to be used for the claim and in particular the relationship between the public law reason for restitution and the private law concept of mistake.

The Reason for Restitution and the Decision of Park J at First Instance

At first instance, Park J’s conclusions directly contradicted the possibility of the public law reason for restitution finding a place at the top of the hierarchy, relying on Lord Goff’s dicta in Kleinwort Benson v Lincoln⁸⁵ to hold that there was no principle forbidding the recovery of overpaid taxes on the basis of mistake. This was particularly surprising given that in 1995 DMG had learned that M&H were pursuing their claim before the ECJ. It was therefore at least arguable that from 1995 onwards DMG knew that the tax might infringe EC law requirements and hence that it was no longer mistaken about the law when it made its third payment in 1996. Of course whether DMG was mistaken or not depends on whether the mistake concerned the lawfulness of denying the election or the lawfulness of the subsequent collection of ACT. Park J’s view was that the claimant’s mistake concerned the unavailability of the election, which meant that the claimant could assert that it would not have paid ACT but for this mistake. However, if this is so then it might be thought difficult to argue, contrary to the protests of DMG’s Head of Taxation, that this mistake was not affected by knowledge of M&H’s action. The precise operation of such mistakes will be discussed in further detail below,86 but it will be noted there that it is not entirely clear how far a claim grounded on mistake should be undermined by the fact that the claimant suspected the true state of affairs. Academic opinion on the matter is divided,⁸⁷ but it is at least arguable that on the facts DMG’s knowledge was too great to allow it to recover for mistake if the relevant mistake does indeed concern the unavailability of the group income election. It is always difficult to apply to large companies concepts more usually applied to individuals⁸⁸ and it may therefore be that Park J was simply avoiding in a pragmatic way the difficulties of establishing whether DMG could have been said to be ‘mistaken’ in the usual sense of the word. On the other hand, DMG issued its claim form after the delivery of the Advocate General’s opinion, which points to discovery at that point if not sooner. Also, Park J sought to shore up his decision by pointing to the fact that even if DMG had tried to take the appropriate action the Revenue would have rejected this attempt and DMG would still have been liable to pay the tax according to the IRC.⁸⁹ If DMG’s inaction is simply evidence of their mistake then it is hard to see why the Revenue’s action is relevant. The IRC’s continued mistake would hardly make up for an action indicating a lack of mistake on the part of DMG.⁹⁰

The key to Park J’s decision to allow recovery must therefore be found instead in his conclusion that the ACT was, as a matter of law, payable when DMG paid it.⁹¹ Thus for Park J, DMG’s claim was not affected by DMG’s knowledge of the Metallgesellschaft and Hoechst claim because by the time it discovered its mistake on the election it was already locked into a system which later led to a valid ACT payment. If the above conclusion that the ECJ invalidated receipt of the ACT, as well as the unavailability of the group income election is accepted, however, Park J’s reasoning will not work. On the other hand, since the precise status of the ACT collection was not wholly clear, DMG could possibly be said to have remained mistaken about that, even after discovery of the Metallgesellschaft and Hoechst action.

[A]lthough under [the national] regime ACT was due and payable, the true state of the law . . . was that the regime gave rise to no obligation to pay. Thus the payments were made pursuant to an unlawful demand . . . Thus DMG’s mistake lay not in its belief that a group income election was not available, but rather in its belief that the ACT was payable when, on the true state of the law, it was not. That was plainly a mistake of law.92

Given the relative simplicity of this aspect of their decision, the Court of Appeal’s full judgment in Deutsche Morgan Grenfell was surprising for two reasons. First, and most significantly, although the same outcome in the appeal could have been achieved solely on this issue of mistake, all three judges chose what Jonathan Parker LJ called ‘the cause of action issue’ (the relationship between Woolwich and Kleinwort Benson v Lincoln) as their central ground of decision instead. Had this survived the appeal to the House of Lords it would have made Deutsche Morgan Grenfell extremely important, because in deciding this question the Court of Appeal did two vital and interlinked things. The first was to reiterate that there is something different about ‘the Woolwich cause of action’ that distinguishes it from all other private law reasons for restitution.⁹³ The second was to establish a hierarchy of the kind suggested here, at least between actions for mistake of law and Woolwich, such that actions for return of tax which could have been brought using Woolwich, or as it has been reinterpreted here, the public law reason for restitution, could not be brought using mistake of law. In other words, as Jonathan Parker LJ put it, ‘the Woolwich cause of action effectively subsumes any cause of action which might otherwise exist for mistake of law’.⁹⁴ This support for at least a limited form of hierarchy of reasons for restitution should not be underestimated, particularly in the light of Park J’s decision at first instance and the fact that in all other circumstances claimants had been thought free to pick and choose between the reasons for restitution on which they rely. As such, the decision of the Court of Appeal in Deutsche Morgan Grenfell had the potential to be vitally important.

Jonathan Parker LJ’s judgment exhibited this interpretation technique to the greatest extent. He took the view that ‘it is clear from [Lord Goff’s] comprehensive treatment of the subject matter [in those two cases] that he cannot have intended to leave over for future consideration an issue as important as [this]’.95 This led Jonathan Parker LJ to conclude that the answer must therefore be found by scrutinising Lord Goff’s dicta in each of them.

It is certainly true, as Buxton LJ puts it at the beginning of his judgment, that Lord Goff’s

speeches are no ordinary judgments. In both cases the leading authority in a particular area of the law set himself to explain developments in that law that were acknowledged to be of general significance and which, before those cases, had been feared to be beyond the reach of judicial as opposed to legislative intervention.⁹⁶

Indeed, the proper relationship between development by the common law as opposed to statute was the issue that split the majority and the minority in Woolwich. However, Buxton LJ also notes that he has warned himself ‘against treating Lord Goff’s speeches in Woolwich . . . and in Kleinwort as if they were statutes’⁹⁷ and it is not difficult to see why. On the one hand, the advantage of development by the common law is that courts are able to proceed on an evolutionary and principled basis with the substance of particular issues before them. On the other, statutes may result in the courts’ room for manoeuvre being limited, but have the weight of Parliamentary democracy behind them. Approaching this important question by analysing the judgment, even of so influential a figure as Lord Goff, as if it contained the answer to a further common law issue, could be seen as containing the worst of both these worlds. Development on a principled, evolutionary basis is excluded, and yet the text removing this flexibility is not supported by the weight of democratic legitimacy.

In addition, the greater the weight placed on interpretation, the more easily the judgment can be weakened simply by presenting an alternative interpretation. Beginning with Woolwich, Jonathan Parker LJ took the following passage from Lord Goff’s judgment:

[I]f there is to be a right of recovery in respect of taxes exacted by the revenue, it is irrelevant to consider whether the old rule barring recovery of money paid under mistake of law should be abolished, for that rule can have no application where the remedy arises not from error on the part of the taxpayer, but from the unlawful nature of the demand by the revenue.98

This was open to two possible interpretations. The first was indeed that chosen by Jonathan Parker LJ, who concluded that Lord Goff would not have allowed mistake of law as an alternative to the ‘Woolwich cause of action’.⁹⁹ It is submitted that while this result was absolutely correct as a matter of principle, it cannot be derived from the dicta of Lord Goff. The second, and arguably more persuasive interpretation of this passage arises because, as Jonathan Parker LJ himself noted,¹⁰⁰ Woolwich was not mistaken, so even if Lord Goff had wanted to do away with the mistake of law rule in that case, it would not have done Woolwich any good and thus another reason for restitution had to be found. Nothing in the passage can therefore be taken to indicate what Lord Goff thought should happen where the taxpayer was mistaken, precisely because he was considering a case in which it was not.

Furthermore, it is even less likely that Lord Goff in Woolwich was dealing with the relationship between that action and one based on mistake of law given that, as Jonathan Parker LJ again notes himself, once the ‘Woolwich cause of action’ was established, taxpayers could recover as of right ‘without the need to invoke a mistake of law’¹⁰¹ (emphasis added). This pushes the question of mistake of law further away from the subject matter at issue in Woolwich, because now fewer claimants need use mistake. The only circumstances in which they would want to go to the added effort of doing so are those that arose in Deutsche Morgan Grenfell itself, where mistake is necessary to extend the limitation period. Given how far this matter is from the context of Lord Goff’s dicta in Woolwich, it must be at least arguable that he was not dealing with it there.

Perhaps the oddest thing about Jonathan Parker LJ’s choice of interpretation is that he himself reached the opposite conclusion in considering Lord Goff’s treatment of Willis Faber Enthoven (Pty) Ltd v Receiver of Revenue,¹⁰² Air Canada v British Columbia¹⁰³ and Commissioner of State Revenue v Royal Insurance Australia Ltd¹⁰⁴ together. All three cases concerned recovery of overpayments of tax for mistake and at first instance Park J thought Lord Goff’s treatment of them indicated that Lord Goff had no objection to a similar result forming part of English law.105 Jonathan Parker LJ, however, disagreed with Park J’s interpretation, precisely because the taxpayers in Woolwich were not mistaken and thus he did not believe Lord Goff’s dicta in that context should be seen as deciding anything about cases in which they were.¹⁰⁶

Moving on to consider Lord Goff’s judgment in Kleinwort Benson v Lincoln, Jonathan Parker LJ reached the same conclusion; that Lord Goff did not intend mistake of law as an alternative to the right to recovery established in Woolwich. Again, it is certainly true (as noted above) that Lord Goff thought ‘private law transactions’, such as those involving swaps, could not be dealt with using the ‘Woolwich cause of action’,¹⁰⁷ but this does not necessarily mean, conversely, that he thought that mistake of law could not be used to deal with taxes. He was not being asked to answer that question so, just as with his dicta in Woolwich, we cannot necessarily assume that he did.

It would be unfair to suggest that Jonathan Parker LJ’s judgment was entirely based on interpretation and not principle. In particular, he was concerned that if the common law were to allow recovery for mistake alongside the statutory regime, this would undermine that regime, whereas the ‘ Woolwich cause of action’ operates around the regime, preserving it. However, if it is accepted that nothing Lord Goff says in Woolwich is intended to deal with a situation where the taxpayer was mistaken, and conversely that nothing he says in Kleinwort Benson v Lincoln is intended to deal with the situation where the payment is of tax, then this more principled objection also falls. If Lord Goff has not considered the relationship between Woolwich and mistake of law, it is unsurprising that there is no discussion of how the statutory rights of recovery should fit within that relationship. It is submitted that Jonathan Parker LJ’s principal error is to regard ‘cases where a payment is made to the revenue which engages the relevant statutory regime regulating recovery for mistake of law’ as being cases of ‘lawful’ demands.¹⁰⁸ If instead it is considered that it is ultra vires the Revenue to receive any more tax than Parliament has given it the right to receive, whether those overpayments result from mistaken and unlawful demands by the IRC, or spontaneously mistaken overpayments by taxpayers, then none of them are lawful. The only difference is that some such unlawful receipts are governed by legislation while others are governed by the common law. Thus it is simply a matter of recognising that statutory rights must be exhausted before common law rights can be used (whatever the reason for restitution) and that the common law cannot be used to undermine the statutory framework. This may well mean that the time limits available at common law should not be used to undermine the time limits available in the statutory framework,¹⁰⁹ but that is a question of applicable defences and will be dealt with as such in chapter five. This defences ‘cart’ should not come before the cause of action ‘horse’.

As noted above, although Jonathan Parker LJ relied to the greatest extent on interpreting the judgments of Lord Goff, he was not alone in doing so. Rix LJ’s conclusion was relatively short, simply expressing agreement with the conclusions of his colleagues, but Buxton LJ, as the quotes above indicate, also thought that Lord Goff’s dicta were central to the decision in Deutsche Morgan Grenfell.110 However, his judgment is arguably to be preferred to that of Jonathan Parker LJ, because it relies less on an examination of what Lord Goff said and more on consideration of why in principle Lord Goff might have been coming to those conclusions. For example, Buxton LJ accepts that Lord Goff may not have seen Woolwich as an appropriate vehicle for removal of the mistake of law rule, but explains that this was not just because there was no mistake in Woolwich, rather it was because such mistakes are in any case inevitably irrelevant when the reason for recovery was the ‘unlawful nature of the action on the part of the revenue’.¹¹¹ Similarly, he did examine the context and detail of Lord Goff’s dicta on the separateness of public and ‘private’ law transactions in Kleinwort Benson v Lincoln, but concluded that he was

unable to accept the argument of the respondents that . . . [that] crucial part of Lord Goff’s judgment is in some way limited in its generality because it occurs during a discussion of the settled law defence . . . precisely because Lord Goff needed to set out the general structure of the law before identifying the different effect of the arguments about the settled law defence.¹¹² (emphasis added)

This kind of examination comes much closer to the approach I have been advocating and furthermore, having examined Lord Goff’s reasons, Buxton LJ added some of his own. Far from being ‘superogatory’ as he feared, it is this that strengthens Buxton LJ’s judgment and makes it the most important of the three, containing several important points.

One of these points is his consideration of the first characteristic he believes to be peculiar to the Woolwich cause of action. This is that ‘although stress is laid upon the coercive power of the state, that in itself does not satisfy the requirements for restitutionary recovery on the ground of compulsion’.113 This is extremely important. I have argued that the significance of coercive power is indeed not that it provides the reason for restitution, but that it indicates the presence of the public law, or ultra vires event, which does provide the reason.¹¹⁴ The coercive powers trigger the application of public law, which then provides the relevant ultra vires action, which in turn forms the reason for restitution. Once it is thus understood that the role of the coercive powers is at a stage removed from the reason for restitution it is possible to liberate the ‘ Woolwich cause of action’ from the confines of fact situations in which the public body is the defendant and to make it a general ‘public law event’ reason for restitution, applicable whenever the transfer of money was caused by such an event, even if the public body is the claimant.¹¹⁵

Of course, this suggestion is directly contrary to the dicta of Lord Goff in Kleinwort Benson v Lincoln which were so important to the Court of Appeal in Deutsche Morgan Grenfell. The swaps cases provide a classic example of an ultra vires event generating a right to recover where the public body could be the claimant instead of the defendant and yet, as noted in chapter three, in Kleinwort Benson v Lincoln Lord Goff specifically designated them to be ‘private law’ cases to which the ‘ Woolwich’ cause of action would be inapplicable. This is precisely why a full understanding of the so-called ‘Woolwich’ or ‘public law event’ cause of action cannot be derived simply by examining Lord Goff’s previous judgments. I have argued in chapter three that Lord Goff’s statement in Kleinwort Benson v Lincoln was itself upside-down.¹¹⁶ As Buxton LJ himself points out, it is necessary to set out the general structure of the law before identifying the application of defences and yet this is exactly the opposite of what Lord Goff did in Kleinwort Benson v Lincoln. It is true that Buxton LJ also suggests that in the swaps cases ‘the element of public interest [present in Woolwich] is lacking’, but it is difficult to see why this is so, given the parallels between the swaps cases and Woolwich. And once he moves away from considering Lord Goff’s dicta, there is another aspect to Buxton LJ’s judgment which supports the inclusion of the swaps cases in the Woolwich, or public law, reason for restitution. As he puts it, it is ‘inept’ to try to include Woolwich-type cases in the definition of mistake. His concern is the difficulty of accommodating DMG’s knowledge of the Metallgesellschaft and Hoechst decision, but it is equally difficult to accommodate the misprediction/mistake dilemma in Kleinwort Benson v Lincoln. In both cases, the attempts to do so are beside the point. The real question is not whether cases such as Deutsche Morgan Grenfell can be fitted into the category of mistake of law, but whether we really want them to belong there anyway. This is what explains why the Court of Appeal chose to deal with the ‘cause of action’ issue rather than simply challenging Park J’s definition of mistake and leaving it at that. The ‘vice’, as Buxton LJ puts it,117 of both Deutsche Morgan Grenfell and Kleinwort Benson v Lincoln is the ultra vires action by the public authority. By using the public law reason for restitution that vice can be addressed directly and the specifically tailored¹¹⁸ defences would then apply to all such cases, following the requirements of their specific reason for restitution, rather than vice versa. Buxton LJ was right to suggest that such an unjust factor would stand ‘outside the main stream of restitution as understood in a private law context’.¹¹⁹

All this connects to a further point. Throughout his judgment, Jonathan Parker LJ referred to Woolwich as being concerned with taxes, or payments to the revenue specifically. Buxton LJ’s language, however, was wider, since throughout his judgment he refers to ‘ultra vires payments’. As I have shown in chapter three,¹²⁰ although Woolwich itself was only concerned with taxes, dicta at all levels in that case vary in terms of their generality and it is not possible to discern, just by looking at that case, the extent of the payments or receipts to which it might extend. However, by expanding the point above, the full scope of application of Woolwich, or public law, reason for restitution becomes clear. It need not be limited, as Jonathan Parker LJ suggests, just to taxes, but is in fact applicable to any payment or receipt which is the result of an action held to be ultra vires as a matter of public law.

The Court of Appeal’s judgment in Deutsche Morgan Grenfell thus represented an important step in support of the establishment of the hierarchy I proposed, albeit that it was a step based more on the interpretation of Lord Goff’s dicta than on principled reasoning of the kind outlined here. However, the decision also proved to be a high point from which the law has since receded.

The Reason for Restitution and the Decision of the House of Lords

In July 2006, Deutsche Morgan Grenfell finally reached the House of Lords, which gave judgment in October,¹²¹ allowing the appeal from the Court of Appeal, by a 4:1 majority¹²² and essentially restoring the judgment of Park J at first instance by a majority of 3:2.¹²³