United Kingdom House of Lords Decisions

HOUSE OF LORDS

OPINIONS OF THE LORDS OF APPEAL FOR JUDGMENT

IN THE CAUSE

Deutsche Morgan Grenfell Group Plc (Respondents) v. Her Majesty's Commissioners of Inland Revenue and another (Appellants)

Deutsche Morgan Grenfell Group plc (Appellants) v. Her Majesty's Commissioners of Inland Revenue and another (Respondents)

(Consolidated Appeals)

[2006] UKHL 49

LORD HOFFMANN

My Lords,

1.   On 8 March 2001 the Court of Justice for the European Communities decided that that United Kingdom revenue law, which had since 1973 allowed companies whose parents were resident in the United Kingdom to elect to pay dividends free of advanced corporation tax ("ACT"), discriminated unlawfully against companies with parents resident in other Member States: Metallgesellschaft Ltd v Inland Revenue Commissioners (Joined Cases C-397 and C-410/98) [2001] Ch 620. The exaction of the tax from such companies had been contrary to the EC Treaty and they were entitled to compensation.
2.   The forensic fall-out from this decision has been very considerable. Large numbers of subsidiaries of companies resident in other Member States have lodged claims for compensation or restitution, some raising difficult ancillary points of law. The High Court has made a group litigation order to enable these points to be resolved in an orderly fashion. The main point in this appeal concerns the period of limitation applicable to such claims. But that in turn raises some fundamental questions about the cause of action upon which the claimants rely.
3.   Before coming to these questions, I must briefly enlarge upon the provisions relating to advance corporation tax which the ECJ held to be contrary to Community law. The tax, which was abolished in 1999, was in theory corporation tax payable in advance of the date on which it would otherwise have been payable. A company resident in the United Kingdom pays corporation tax on profits arising in a given accounting period and, generally speaking, the tax is payable nine months after the period ends. But the trigger for the payment of corporation tax was the payment of a dividend. A company which paid a dividend became liable to account to the Inland Revenue for ACT calculated as a proportion of the dividend. This could afterwards be set off against the corporation tax ("mainstream corporation tax" or "MCT") which became chargeable on its profits. The Revenue thereby obtained early payment of the tax and, in cases in which the company's liability for MCT turned out to be less than it had paid as ACT, payment of tax which would not otherwise have fallen due.
4.   The rule that ACT was payable on dividends was however subject to an exception if the dividend was paid to a parent company in the same group. Under section 247 of the Income and Corporation Taxes Act 1988 the company and its parent could jointly make a group income election which gave them the right to be treated for the purposes of ACT as if they were the same company. No ACT would be payable on the distribution by the subsidiary. It would however be payable on any distribution by the parent. The Act confined the right of election to cases in which the parent was resident in the United Kingdom. Otherwise a subsidiary which had elected would not be liable to ACT and the parent, being non-resident, would not be liable either.
5.   In the Metallgesellschaft case the Court of Justice decided that these arrangements infringed the right of establishment guaranteed by article 52 (now 43) of the EC Treaty in that they discriminated against companies resident in other Member States. It held that the companies which had been unlawfully required to pay ACT were entitled to restitution or compensation. The nature of the remedies, the procedures by which they could be enforced and matters like the appropriate limitation periods were said to be matters for domestic law. The only specific qualification imposed by the Court of Justice was that English courts could not apply the rule in the The Pintada (President of India v La Pintada Compania Navigacion SA [1985] AC 104) to deny any recovery of interest to a claimant whose ACT had been set off against MCT before the commencement of proceedings. The claimant was entitled to be compensated for loss of the use of the money between the date on which it was paid and the date when MCT became due.
6.   In these proceedings, commenced on 18 October 2000, Deutsche Morgan Grenfell Group plc ("DMG") claims compensation for having had to pay ACT on three dividends paid to its German parent company between 1993 and 1996: in October 1993, February 1995 and January 1996. (No mention of the 1995 and 1996 ACT payments appeared in the pleadings until an amendment made on 19 August 2002 and there is an issue, to which I shall return later, over whether that latter date should be taken for the purposes of limitation as the commencement date of the proceedings in respect of those dividends.) All the payments were subsequently set off against MCT. The facts are set out more fully in the speech to be delivered by my noble and learned friend Lord Walker of Gestingthorpe, which I have had the privilege of reading in draft.
7.   There is no dispute that if DMG had been entitled to make a group election, it would have done so. There is likewise no dispute that DMG is entitled to compensation for breach of statutory duty (the infringement of article 43) or by way of restitution of tax unlawfully demanded under the principle established in Woolwich Equitable Building Society v Inland Revenue Commissioners [1993] AC 70. But the period of limitation for both of these causes of action runs from the date of payment and DMG wishes to claim in respect of the 1993 payment, which on any view was made more than six years before proceedings were commenced. In addition, if the proceedings in respect of the 1995 and 1996 ACT payments are treated as having been commenced on 19 August 2002, they would also have been more than 6 years earlier. DMG therefore argues that it has an additional cause of action for restitution on the ground that the money was paid by mistake. Section 32(1)(c) of the Limitation Act 1980 provides that where the action is for "relief from the consequences of a mistake", the period of limitation does not begin to run until the claimant has discovered the mistake "or could with reasonable diligence have discovered it." (With effect from 8 September 2003, this provision no longer applies to mistakes of law in tax cases: see section 320 of the Finance Act 2004.) DMG says that it did not discover its mistake until the ECJ gave judgment (after the commencement of proceedings) and no amount of diligence could have enabled it to know in advance what the ECJ was going to say.
8.   The first question, therefore, is whether DMG has a cause of action which can be described as being "for relief from the consequences of a mistake" within the meaning of section 32(1) of the 1980 Act. It claims that it seeks relief against having paid money to the Inland Revenue in the mistaken belief that, since section 247 of the Taxes Act made no provision for a group election by a company with a German parent, it was obliged to pay ACT. In fact, article 43 of the Treaty made this denial of a right of election unlawful and, in consequence, since DMG would have exercised its election, it was not obliged to pay ACT.
9.   Before the decision of the House of Lords in Kleinwort Benson Ltd v Lincoln City Council [1999] 2 AC 349, this mistake would not have given rise to any cause of action because it was a mistake of law. That rule has now been abandoned. Nevertheless, Mr Glick QC for the Inland Revenue submits that while it is now in general true that money paid by mistake can be recovered, whether the mistake is of fact or law, tax is different. There is still no cause of action at common law for the recovery of tax paid under a mistake of law. He says that there are only two remedies for the recovery of tax which was not due. One is the common law remedy to recover tax unlawfully demanded which was established in the Woolwich case. The other is the statutory remedy provided by section 33 of the Taxes Management Act 1973:
33.--(1) If any person who has paid tax charged under an assessment alleges that the assessment was excessive by reason of some error or mistake in a return, he may by notice in writing at any time not later than six years after the end of the year of assessment (or, if the assessment is to corporation tax, the end of the accounting period) in which the assessment was made, make a claim to the Board for relief.

(2) On receiving the claim the Board shall inquire into the matter and shall, subject to the provisions of this section, give by way of repayment such relief …in respect of the error or mistake as is reasonable and just:

Provided that no relief shall be given under this section in respect of an error or mistake as to the basis on which the liability of the claimant ought to have been computed where the return was in fact made on the basis or in accordance with the practice generally prevailing at the time when the return was made.

(3) In determining the claim the Board shall have regard to all the relevant circumstances of the case, and in particular shall consider whether the granting of relief would result in the exclusion from charge to tax of any part of the profits of the claimant, and for this purpose the Board may take into consideration the liability of the claimant and assessments made on him in respect of chargeable periods other than that to which the claim relates.

10.   Whether the claim is under the Woolwich principle or section 33, time runs from when the payment was made. So Mr Glick says that in either case, the claim for interest on the 1993 payment is statute barred. The judge (Park J) rejected the submission. He saw no reason in principle why the right to restitution of payments made by mistake, which had been extended in Kleinwort Benson to include mistakes of law, should not apply to payments of tax. No argument based on section 33 appears to have been advanced to him.
11.   The Court of Appeal (Jonathan Parker, Rix and Buxton LJJ) disagreed. The main reason was their view that Lord Goff of Chieveley, in his speech in Kleinwort Benson, had said that payments of tax under a mistake of law were subject to a separate and distinct regime which provided remedies only under the Woolwich principle and section 33. Buxton LJ also offered some reasons why it would cause difficulties if payment by mistake was accepted as a ground for the recovery of taxes. I will come back to this point later, when I deal with the question of whether DMG did in fact pay the tax by mistake.
12.   First, however, I must deal with the opinion attributed by the Court of Appeal to Lord Goff. Both Jonathan Parker LJ and Buxton LJ subjected his speeches in the Woolwich and Kleinwort Benson case to a detailed analysis which I have read more than once with attention and respect. The chief support for Mr Glick's argument is to be found in the following passage in Kleinwort Benson [1999] 2 AC 349, 382:
"…in our law of restitution, we now find two separate and distinct regimes in respect of the repayment of money paid under a mistake of law. These are (1) cases concerned with repayment of taxes and other similar charges which, when exacted ultra vires, are recoverable as of right at common law on the principle in Woolwich, and otherwise are the subject of statutory regimes regulating recovery; and (2) other cases, which may broadly be described as concerned with repayment of money paid under private transactions, and which are governed by the common law."

13.   There is no doubt that the regimes are different. Both the Woolwich principle and section 33 apply only to the recovery of money paid as taxes or the like. They do not apply to "private transactions". The Woolwich principle is indifferent as to whether the taxpayer paid the tax because he was mistaken or, as in Woolwich, for some other reason. And section 33 has its own rules. So the regime for taxes is certainly different. But the question is whether Lord Goff meant to say that the remedies provided by the two regimes are mutually exclusive. Woolwich and section 33 are available only for "taxes and other similar charges". Does it follow that the common law rule for recovery of payments made by mistake, as applied to private transactions in Kleinwort Benson, does not apply to taxes? That would be going a good deal further. It is one thing to say that the regimes are different and another to say that their remedies are mutually exclusive.
14.   This question is discussed at considerable length in the judgments in the Court of Appeal. It is, I think, neither here nor there for me to say that, as one who (in the end) gave wholehearted concurrence to Lord Goff's speech, I never thought that it had the meaning attributed to it by the Court of Appeal. Once a judgment has been published, its interpretation belongs to posterity and its author and those who agreed with him at the time have no better claim to be able to declare its meaning than anyone else. But to my mind the context in which Lord Goff made the remarks which I have quoted demonstrates conclusively that he could not have meant what the Court of Appeal thought.
15.   Early in his speech ([1999] 2 AC 349, 367) Lord Goff announced that he proposed to address first the question of whether the rule precluding recovery of money paid under a mistake of law should remain part of English law. This had not been much discussed in argument. Counsel for the respondents had not attempted to defend the old rule but had concentrated his fire on the questions of whether someone who acts in accordance with a settled understanding of the law can be said to have made a mistake, or whether, if he has, the rule should be subject to an exception in such a case. Nevertheless, Lord Goff devoted some space to an examination of the history and possible policies of the mistake of law rule and finally concluded (at p 375) that it should be abrogated. That part of his speech contains no hint of an exception for taxes paid under a mistake of law.
16.   Lord Goff then went on to the question of whether it made a difference that the payments were made in accordance with a settled understanding of the law. It is here that the passage which I have quoted appears. He uses the distinction between tax payments and private transactions to argue that the case for a settled law exception is stronger in the case of tax payments ("large numbers of taxpayers may be affected" and "there is an element of public interest") than in the case of private transactions. At the end of this discussion, he leaves the door slightly open for an argument that there is such a defence for tax payments. But he rejects it for private payments.
17.   My Lords, this reasoning is quite inconsistent with the absence of a cause of action for recovery of tax on the grounds of mistake of law. What kind of claim did Lord Goff contemplate that a settled law defence might protect the Revenue against? Surely, a claim to recover tax on the ground that it had been paid under a mistake of law. Lord Goff was not suddenly turning to the Woolwich cause of action and asking whether it should be subject to a defence that the demand for tax, although ultra vires, was in accordance with a settled understanding of the law. The question which he had announced (at p 367) that he intended to answer was "whether…there should be an exception to recovery on the ground of mistake of law…in cases where the money has been paid under a settled understanding of the law which has subsequently been changed by judicial decision." There would be little point in discussing whether a settled understanding of the law should be a defence to a claim for recovery of a tax payment on the grounds of mistake of law if there was no such cause of action.
18.   In my opinion, Lord Goff's speech in Kleinwort Benson does not deny the right to recover tax on the ground that it was paid by mistake. On the contrary, his discussion of a possible settled law defence necessarily entails that he thought that there was such a cause of action. And for the reasons I gave in Kleinwort Benson, I do not think that there is an exception for cases in which there is a settled view of the law.
19.   Mr Glick's alternative submission was that section 33 excluded any common law claim on the grounds of mistake. He said that Parliament, having provided a qualified remedy for one category of mistaken payments of tax (when "the assessment was excessive by reason of some error or mistake in a return"), must be taken to have dealt exhaustively with any kind of mistaken payment of tax and, so far as section 33 did not provide a remedy, must be taken to have intended that no remedy should exist. Mr Glick accepts that section 33 has no application to the present case because ACT was payable without any assessment, but nevertheless submits that section 33 excludes a remedy. In my opinion this goes much too far. Mr Glick advanced a similar argument in the Woolwich case, where section 33 did not apply because there had been no lawful assessment. The House of Lords rejected it. It is true that in Woolwich Mr Glick's argument was more ambitious, in that he was trying to use section 33 to exclude a remedy even when there had been no mistake of any kind. But the question is in the end one of construction. When a special or qualified statutory remedy is provided, it may well be inferred that Parliament intended to exclude any common law remedy which would or might have arisen on the same facts. That was the case in Marcic v Thames Water Utilities Ltd [2003] UKHL 66; [2004] 2 AC 42, upon which Mr Glick relied. But I see no reason to infer that Parliament intended to exclude a common law remedy in all cases of mistake (whether of fact or law) in which the Revenue was unjustly enriched but did not fall within section 33.
20.   The next question is whether the money was paid by mistake. This might seem at first sight to be a simple question but the division of opinion in the Kleinwort Benson case [1999] 2 AC 349 and the academic literature show that it can lead one into deep waters. One might start by asking why it matters. The effect of the decision in Metallgesellschaft [2001] Ch 620 was that the Inland Revenue had not been entitled to the money. Nor could the Revenue have thought that DMG was intending to make an interest-free loan to the British government or that there was any other proper ground on which they had been entitled to retain it. Why, then, is it necessary to investigate the precise state of mind (of which the Revenue would have known nothing) with which DMG made the payment?
21.   The answer, at any rate for the moment, is that unlike civilian systems, English law has no general principle that to retain money paid without any legal basis (such as debt, gift, compromise, etc) is unjust enrichment. In the Woolwich case [1993] AC 70, 172 Lord Goff said that English law might have developed so as to recognise such a general principle - the condictio indebiti of civilian law - but had not done so. In England, the claimant has to prove that the circumstances in which the payment was made come within one of the categories which the law recognizes as sufficient to make retention by the recipient unjust. Lord Goff provided a list in the Woolwich case at pp 164-165 and the decision itself added another. One such category, long recognized, is payment by mistake: see Kelly v Solari (1841) 9 M & W 54. The late Professor Birks argued, in the second edition of his book on Unjust Enrichment (2005), that the trend of recent English decisions meant that, for the purpose of entitling a claimant to recover, the categories were now superfluous. The fact that the money had not been due was, in the absence of some other causa for payment, a sufficient ground for recovery. We have now developed a condictio indebiti. The absence of a basis for the payment is a ground which generalises and subsumes all the separate categories of situation in which a payment of money not due was recoverable.
22.   I do not think it is necessary for us to decide this question about the fundamental basis of enrichment liability because the question before the House is not the fundamental juridical basis of DMG's cause of action but whether the action can be described as being "for relief from the consequences of a mistake" within the meaning of section 32(1)(c) of the 1980 Act. Kleinwort Benson [1999] 2 AC 349 is recent authority for the proposition that an action for restitution of money paid under a void contract can fall within this description. That does not seem to me inconsistent with the existence of the mistake not being essential to the cause of action but merely one example of a case which falls within a more general principle, just as one could have (say, for the purposes of limitation) a category called "clinical negligence" without implying that it is a cause of action different in nature from other kinds of negligence.
23.   I come back, therefore, to the question of whether DMG made a mistake, against the consequences of which the action seeks relief. The first point to make is that the alleged mistake was one of a very special kind. If DMG had known for certain what the Court of Justice was going to say in Metalgesellschaft on 8 March 2001, it is very unlikely that it would have paid ACT. But it had no means of knowing that. It was only in retrospect that it became clear that the ACT could not lawfully have been exacted. Professor Birks said that this was not a mistake at all. It was merely an inability to predict what the Court of Justice was going to say, just as one cannot predict with certainty what the weather is going to be like. And Sir Jack Beatson, writing extra-judicially in the volume to be published in memory of Professor Birks (Unlawful Statutes and Mistake of Law: Is there a Smile on the Face of Schrödinger's Cat? in Mapping the Law (ed Burrows and Rodger) at pp 163-180) describes the majority decision in Kleinwort Benson to treat a similar failure of prediction as a mistake as an "emphatic endorsement…of the declaratory theory of judicial decision-making" and "abstract juridical correctitude". This seems to me, with respect, to muddle two different questions. One is whether judges change the law or merely declare what it has always been. The answer to this question is clear enough. To say that they never change the law is a fiction and to base any practical decision upon such a fiction would indeed be abstract juridical correctitude. But the other question is whether a judicial decision changes the law retrospectively and here the answer is equally clear. It does. It has the immediate practical consequence that the unsuccessful party loses, notwithstanding that, in the nature of things, the relevant events occurred before the court had changed the law: see In re Spectrum Plus Ltd [2005] UKHL 41; [2005] 2 AC 680. There is nothing abstract about this rule. So the main question in the Kleinwort Benson case was whether a person whose understanding of the law (however reasonable and widely shared at the time) is falsified by a subsequent decision of the courts should, for the purposes of the law of unjust enrichment, be treated as having made a mistake. The majority view in Kleinwort Benson was that he should. The effect of the later judgment is that, contrary to his opinion at the time, the money was not owing. It is therefore fair that he should recover it. It may be that this involves extending the concept of a mistake to compensate for the absence of a more general condictio indebiti and perhaps it would make objectors feel better if one said that because the law was now deemed to have been different at the relevant date, he was deemed to have made a mistake. But the reasoning is based upon practical considerations of fairness and not abstract juridical correctitude.
24.   Mr Glick, taking Kleinwort Benson at face value, accepted that the DMG made the 1993 and February 1995 payments by mistake. But he said that the commencement of the Metallgesellschaft litigation in July 1995 must, at the very lowest, have raised a doubt in the relevant minds of DMG. There was evidence that in July 1995 DMG had discussed the implications of the Metalgesellschaft challenge and considered making a "protective" group election in case Metallgesellschaft won its case. (Whether such an election would have done them any good is another matter: see Park J at [2003] 4 All ER 645, 657.) So the Revenue say that the 1996 payment was made when a state of doubt existed and that this was not a mistake.
25.   There is some authority for the view that a state of doubt does not amount to a mistake: see Burrows The Law of Restitution (2nd ed 2002) pp 139-140. In the Kleinwort Benson case ([1999] 2 AC 349, 410) my noble and learned friend Lord Hope of Craighead said:
"A state of doubt is different from that of mistake. A person who pays when in doubt takes the risk that he may be wrong - and that is so whether the issue is one of fact or one of law."

26.   This was a very compressed remark in the course of a discussion of other matters and I do not think that Lord Hope could have meant that a state of doubt was actually inconsistent with making a mistake. Contestants in quiz shows may have doubts about the answer ("it sounds like Haydn, but then it may be Mozart") but if they then give the wrong answer, they have made a mistake. The real point is whether the person who made the payment took the risk that he might be wrong. If he did, then he cannot recover the money. Speaking for myself, I think that there is a parallel here with the question of whether a common mistake vitiates a contract. As Steyn J said in Associated Japanese Bank (International) Ltd v Credit du Nord SA [1989] 1 WLR 255, 268:
"Logically, before one can turn to the rules as to mistake…one must first determine whether the contract itself, by express or implied condition precedent or otherwise, provides who bears the risk of the relevant mistake. It is at this hurdle that many pleas of mistake will either fail or prove to have been unnecessary."

27.   Likewise, the circumstances in which a payment is made may show that the person who made the payment took the risk that, if the question was fully litigated, it might turn out that he did not owe the money. Payment under a compromise is an obvious example: see Brennan v Bolt Burdon [2004] EWCA Civ 1017; [2005] QB 303. I would not regard the fact that the person making the payment had doubts about his liability as conclusive of the question of whether he took the risk, particularly if the existence of these doubts was unknown to the receiving party. It would be strange if a party whose lawyer had raised a doubt on the question but who decided nevertheless that he had better pay should be in a worse position than a party who had no doubts because he had never taken any advice, particularly if the receiving party had no idea that there was any difference in the circumstances in which the two payments had been made. It would be more rational if the question of whether a party should be treated as having taken the risk depended upon the objective circumstances surrounding the payment as they could reasonably have been known to both parties, including of course the extent to which the law was known to be in doubt.
28.   These thoughts may be said to support the view of Professor Birks that English law should be less concerned with whether the person who paid the money made a mistake (involving an inquiry into his subjective state of mind) than with whether there was a valid causa for the payment, such as a debt, compromise or gift, these being matters of objective inquiry into the circumstances of the payment as they would have been known to both parties. But they do not arise in this case because both the judge and Jonathan Parker LJ in the Court of Appeal decided on the facts that, even if a state of doubt was inconsistent with a mistake, DMG had been mistaken. Mr Thomason, the Head of Taxation at DMG, gave evidence about his state of mind when it paid the ACT. He said:
"At all times prior to the determination of the European Court in the Hoechst case, I believed that the United Kingdom statute denying the ability to make a group income election was the law and that I was bound to act in accordance with this law. …It did not occur to me that I could ignore the law as it stood for the simple reason that the law is the law. Just because another taxpayer challenged the law that did not mean that I could or should ignore it."

29.   Park J accepted this evidence as showing that, whether or not a state of doubt was consistent with making a mistake, DMG, in the person of Mr Thomason, was not in a state of doubt. Jonathan Parker LJ agreed. He said ([2006] 2 WLR 103, 179, para 235):
"I accept Mr Rabinowitz's submission … that mere knowledge that the statutory provisions in question are under challenge is not to be equated with a state of doubt as to the validity of those provisions. In any event, as I have already pointed out, Mr Thomason's evidence was that he was in no doubt that the ACT was payable, whatever the decision in the Metallgesellschaft case."

30.   Buxton and Rix LJJ, on the other hand, doubted whether this could be reconciled with the observations of Lord Hope which I have cited above. But I do not understand why this should be so. The judge attributed Mr Thomason's state of mind to DMG and found as a fact that he was not in doubt. He thought that DMG had to pay. Someone with a more sophisticated approach to the law might have had doubts - might even have thought that Metallgesellschaft had a good case and that the ECJ ruling would apply retrospectively - but not Mr Thomason. I would therefore agree with the judge and Jonathan Parker LJ that on any view of the law on this point, he made a mistake.
31.   If DMG made a mistake about the law, when could they "with reasonable diligence" have discovered it? On this question it is important to bear in mind the special nature of the mistake, namely that it was deemed to have been made because of the retrospective operation of a later decision of the Court of Justice. The "reasonable diligence" proviso depends upon the true state of affairs being there to be discovered. In this case, however, the true state of affairs was not discoverable until the Court of Justice pronounced its judgment. One might make guesses or predictions, especially after the opinion of the Advocate General. This gave DMG sufficient confidence to issue proceedings. But they could not have discovered the truth because the truth did not yet exist. In my opinion, therefore, the mistake was not reasonably discoverable until after the judgment had been delivered.
32.   Two footnotes on the question of mistake. First, Park J took a rather sophisticated view of the nature of the mistake. He said that the mistake was not about whether ACT was payable. DMG had not made an election and therefore it was payable. The mistake was about whether DMG should have been allowed to elect. But I agree with the Court of Appeal that the mistake was about whether DMG was liable for ACT. The election provisions were purely machinery, which DMG would undoubtedly have used, by which it could enforce its right to exemption from liability.
33.   Secondly, Buxton LJ said that the problems about what counted as a mistake, some of which I have discussed above, showed the "ineptitude" of extending the law of payment under a mistake of law to payment of taxes. It is true that there may be anomalies in the different limitation treatment of claimants who paid under a mistake and those who paid without a mistake but pursuant to an ultra vires demand (Woolwich) or a void contract (Kleinwort Benson): see [2006] 2 WLR 103, 190 and 192. But these anomalies flow from any recognition of payment by mistake as a cause of action and not from a distinction between payments of tax and private payments. In either case, there is the possibility of alternative causes of action, with one producing more favourable treatment under the Limitation Act than the other.
34.   It follows that in my opinion section 32(1)(c) postponed the commencement of the limitation period in respect of all three ACT payments until 8 March 2001. That makes it unnecessary to decide whether the proceedings in respect of the last two payments were begun when the proceedings were issued or when they were amended. I will say only that on this point I agree with the majority of the Court of Appeal. I would allow the appeal and restore the judgment of Park J.

LORD HOPE OF CRAIGHEAD

My Lords,

35.   I have had the great advantage of reading in draft the speeches of my noble and learned friends Lord Hoffmann and Lord Walker of Gestingthorpe, in which the background to this appeal is so comprehensively and helpfully set out. With that advantage, which I gratefully acknowledge, I can proceed directly to the three of the various issues in dispute which Lord Walker has summarised in para 117 on which I wish to comment. These are (1) the cause of action" issue; (2) the "mistake" issue; and (3) the "discovery" issue.
36.   As Lord Walker points out (para 96), the focus of the present appeal is on limitation of actions. DMG seeks to avoid the six year time limit by taking advantage of the extended limitation period that is available in England and Wales (but not in Scotland: see the Prescription and Limitation (Scotland) Act 1973, section 6 and Schedule I, para 1(b)) under section 32(1)(c) of the Limitation Act 1980 which provides that, where the action is for relief from the consequences of a mistake, the period of limitation shall not begin to run until the claimant has discovered the mistake or could with reasonable diligence have discovered it. This provision no longer applies in relation to a mistake of law relating to a taxation matter under the care and management of the Commissioners for Revenue and Customs: Finance Act 2004, section 320 (read together with section 5 of the Commissioners for Revenue and Customs Act 2005). That section, which applies only to actions brought on or after 8 September 2003, was enacted in response to the judgment of Park J in this case which was given on 18 July 2003: [2003] 4 All ER 645. The application of this provision to claims existing but not yet made by 8 September 2003 is being challenged under Community law: Aegis Group plc v IRC [2005] EWHC 1468 (Ch); [2006] STC 23. It is enough, however, for the purposes of this case to say that it is not in dispute that the extended limitation period is available to DMG if the issues which are raised in this appeal are decided in its favour.
37.   The fact that the benefit of the extended period is being removed from future cases does not deprive the question whether English law recognises a restitutionary claim for tax paid under a mistake of law of all its interest. But the fact that the limitation period is now the same, at least for all new claims, whichever of the various available avenues for the recovery of money from the Revenue is chosen makes it unlikely that this issue will require to be revisited in cases to which it is a party. Tax paid in response to an unlawful demand will be recoverable under the Woolwich principle, subject to the ordinary time limit of six years: Woolwich Equitable Building Society v Inland Revenue Commissioners [1993] AC 70. Where the revenue's assessment to tax was excessive by reason of some error or mistake made by the taxpayer in a return, a statutory remedy for repayment is available. The leading examples are to be found in section 33 of the Taxes Management Act 1970 as to income tax and capital gains tax and the Finance Act 1998, Schedule 18, para 51 as to corporation tax. The remedy available under these provisions must be sought not later than six years after the end of the relevant accounting period: see TMA 1970, section 33(1). Cases not covered by one or other of these remedies in which it will be necessary to resort to a remedy under the principle of unjust enrichment are likely to be rare.

    The cause of action issue

38.   Lord Goff of Chieveley made it clear in Kleinwort Benson Limited v Lincoln City Council [1999] 2 AC 349, 371H-372A (his speech in that case, so rich in scholarship, was the last which he prepared before his retirement as Senior Law Lord) that he was under no illusions about the difficulties which the House faced in formulating satisfactory limits to the right to recover money paid under a mistake of law. He observed that there was more sense in the old mistake of law rule than its more strident critics had been prepared to admit. But its rejection by the common law world was not, as he said at pp 372F-373D, due to a wish to depart from the policy underlying the rule but rather to an acknowledgement, due essentially to the work of scholars, that it could best be achieved by recognising a general right of recovery subject to specific defences to cater for the fears which formerly appeared to require a blanket exclusion of recovery. It was his acceptance of this general right of recovery of money paid under a mistake, whether of fact or law, that lay at the centre of the discussion that then followed. As he put it at p 373C, a blanket rule of non-recovery, irrespective of the justice of the case, could not survive in a rubric of the law based on the principle of unjust enrichment. Instead it was for the law to evolve appropriate defences which could, together with the acknowledged defence of change of position, provide protection where appropriate for recipients of money paid under a mistake of law in those cases in which justice or policy does not require them to refund the money.
39.   This then is the background against which the argument for the Revenue on this issue must be examined. On the one hand there is the proposition which lay at the heart of Lord Goff's analysis. (An indication of the immense care he took over his speech, which was so evident to those of your Lordships who had the privilege of sitting with him on that case, can be gathered from the fact that the original version of it lacked some of the section headings that appear in the revised version that was later published in the Appeal Cases: see [1998] 3 WLR 1095 at pp 1119G and 1121A-B; [1999] 2 AC 349 at pp 379F and 381B). As he put it at p 385A-B, money paid under a mistake of law is recoverable on the ground that its receipt by the defendant will, prima facie, lead to his unjust enrichment, just as receipt of money paid under a mistake of fact will do so. Then there is the principle recognised by Professor Peter Birks, to whom Lord Walker has so fittingly paid tribute. With characteristic simplicity he declared that, unless displaced by statute, causes of action good against private citizens are no less good against public bodies: see his essay (in the volume Essays on Restitution (1990), edited by Professor P D Finn) entitled Restitution from the Executive, at p 174. That was why he acknowledged that, if in Woolwich Equitable Building Society v Inland Revenue Commissioners [1993] AC 70 the building society had made a mistake of fact, it would undoubtedly have entitled the society to restitution of the money paid to the Revenue in consequence of its mistake, just as it plainly would have been had the transaction been with a private citizen.
40.   The question is whether an exception to the general rule which Lord Goff identified should now be recognised in the case of payments made under a mistake of law to the Revenue. How would Lord Goff himself have reacted to it? I think that there is no doubt what his initial reaction would have been. We can see what he made of the argument that the Revenue was in a special position in the Woolwich case at pp 171-172. The Revenue had made an unlawful demand for tax but it was asserting that it was under no obligation to pay back the money. That position seemed to him, as matter of common justice, to be untenable - a position made worse by the fact that it involved the Revenue having the benefit of a massive interest-free loan as the fruit of its unlawful action:
"Common justice seems to require that tax to be repaid, unless special circumstances or some principle of policy require otherwise; prima facie, the taxpayer should be entitled to repayment as of right."

    We can also see, from the way in which he dealt with the suggested defence of honest receipt in Kleinwort Benson [1999] 2 AC 349, pp 384-385, how determined he was to preserve the purity and simplicity of the principle which he had described earlier:

"In my opinion, it would be most unwise for the common law, having recognised the right to recover money paid under a mistake of law on the ground of unjust enrichment, immediately to proceed to the recognition of so wide a defence as this which would exclude the right of recovery in a very large proportion of cases. The proper course is surely to identify particular sets of circumstances which, as a matter of principle or policy, may lead to the conclusion that recovery should not be allowed."

    I think that Mr Rabinowitz QC for DMG was right therefore to take as the starting point for his argument the general right to recover, and then to ask on what grounds either of policy or principle the Revenue can claim that an exception should be made in its case.

41.   I should add, before I complete this introduction, that I believe that we are in a very different field from that which Lord Rodger of Earlsferry was contemplating when in Commissioners of Customs and Excise v Barclays Bank plc [2006] UKHL 28; [2006] 3 WLR 1, 19, para 51 he reminded us of the philosopher's advice to "Seek simplicity and distrust it." The proposition that money paid under a mistake is recoverable is based on the principle that, prima facie, its receipt by the defendant will lead to his unjust enrichment. There is no reason to distrust a proposition based on such an elementary principle just because it is simple. Now that the common law world has recognised that there is a general right of recovery whether the mistake is of fact or law, it should be careful not to disturb its purity and its simplicity unless there is a clear basis on grounds of principle or policy for doing so.
42.   The Revenue submit that a common law claim for restitution based on mistake of law is not available to a party in relation to an overpayment of tax to the Revenue. They maintain that common law claims for restitution arising from a tax measure which is ultra vires or otherwise unlawful can only be made on the principle in Woolwich or by analogy with that principle, and that they are subject to a limitation period of six years. They also point to the existence of statutory remedies. To bring the case within the Woolwich principle by analogy they submit that, while ACT cannot be said to have been an unlawful tax, the payments that were made in this case were the result of the unlawful exclusion of certain subsidiaries in section 247(1) of the ICTA 1988. It follows that a right to a remedy arises in respect of that unlawfulness. As for the statutory remedies, they submit that there is a general principle that a common law remedy will be excluded where Parliament has enacted a statutory scheme inconsistent with the remedy. That scheme is revealed by the various provisions that have been enacted which, subject to certain conditions, permit the recovery of tax paid by reason of an error or a mistake in an assessment. The authority which they cite in support of this proposition is Marcic v Thames Water Utilities Ltd [2004] 2 AC 42.
43.   Two issues of principle lie at heart of this argument. The first is whether the remedy in restitution that is available for payments made under a mistake recognised in Kleinwort Benson Limited v Lincoln City Council [1999] 2 AC 349 is subject to an exception in the case of taxes paid under a mistake of law to the Revenue. The second is whether, if they are not subject to any such exception, it is open to a litigant to choose whichever of two or more concurrent remedies best suits his interests. The argument that, where there is a statutory scheme which occupies the same ground as the common law remedy, the common law remedy must be held to be excluded by it, cannot be said to raise an issue of principle as the effect of the scheme must depend on the words of the statute. Nevertheless it raises an issue of policy that needs to be dealt with.
44.   The submission that the restitutionary remedy for payments made under a mistake is subject to an exception in favour of the Revenue where the mistake was one of law runs into difficulty as soon as it is articulated. It seeks to build in two exceptions, not just one, into the generality of the remedy that was recognised in Kleinwort Benson. The first exception would involve treating payments made under a mistake of fact differently from payments made under a mistake of law. The second would involve treating the Revenue differently from all other public authorities which receive payments made under a mistake of law. If this argument were to succeed it would have a significant impact on the law's taxonomy. English law has been moving step by step towards a principled statement of the law of restitution. The carving out of exceptions which are not clearly based on principle would risk reversing this process.

    (a) the debatable passage

45.   The Revenue's argument on grounds of principle is based on a passage in Lord Goff's speech in Kleinwort Benson which Lord Walker has referred to (para 29(5)) as the debatable passage. In this passage Lord Goff drew a distinction between, on the one hand, "payments of taxes and other similar charges and, on the other hand, payments made under ordinary private transactions": p 381G-H. Elaborating on this distinction at p 382B-D, he said that under our law of restitution there were now to be found "two separate and distinct regimes" in respect of the repayment of money paid under a mistake of law. These were (1) cases concerned with repayment of "taxes and other similar charges" exacted ultra vires, recoverable as of right at common law under the Woolwich principle, and (2) other cases, which might broadly be described as concerned with the repayment of money paid under private transactions, governed by the common law. At p 382D-E he went on to say that a case might be made in favour of a principle that, in cases concerned with taxes, payments made in accordance with a prevailing practice or a settled understanding of the law should be irrecoverable.
46.   There are a number of points about this passage that have to be taken into account in assessing its significance. The first is that it appears in a section of Lord Goff's speech where, having concluded at p 375 that English law should recognise that there was a general right to recover money paid under a mistake whether of fact or law, he was considering whether it would be appropriate to develop a defence of settled understanding of the law on the lines proposed by the Law Commission as a corollary to the newly developed right of recovery: p 381B-C. He was, in effect clearing the ground for a further examination of this point. Cases where taxes and other similar charges exacted ultra vires were recoverable as of right under the Woolwich principle could be left on one side because there was no need in those cases to invoke a mistake of law by the payer: p 381H. The phrase "taxes and other similar charges" lacks the precision that would be needed if it was intended to define the extent of an exception to the general right of recovery. But Lord Goff's point was simply that there was no room in the case of an ultra vires demand for a defence that it was made on a settled understanding of the law. The only context in which a limit to recovery on that ground needed to be considered was where repayment of money was sought as having been paid under a mistake of law under the newly recognised common law principle.
47.   In the Court of Appeal [2006] 2 WLR 103, 173-174, para 205 Jonathan Parker LJ said that the first of the two regimes which Lord Goff identified in this passage was a comprehensive and complete regime in relation to overpayments of tax made under a mistake of law, and that it was not limited to cases where the payment of tax was made pursuant to an unlawful demand - in other words that it was not limited to Woolwich cases. I do not think that this can be right. Not only does it read far more into the passage than the words used justify. It also fails to take account of the fact that in the Woolwich case there was no error of law by the taxpayer. So the House was not called upon to consider the effect of a mistake of law in that case at all. Lord Goff's use of the phrase "or other similar charges" is perfectly intelligible if is understood as referring to the case of an ultra vires demand. All statutory charges which are the subject of an ultra vires demand fall easily within this category. It begs many questions if it was intended to identify a category that was to be excluded from the general common law right of recovery for payments made under a mistake: see, for example, Sir Jack Beatson's comments in Chapter 9, Unlawful Statutes and Mistake of Law, p 174 - 175 in Burrows and Rodger (Eds), Mapping the Law (Oxford 2006).
48.   It has to be acknowledged, of course, that at p 382C-D Lord Goff indicated that these cases were limited to private transactions, and that he repeated this point at p 382E-F when he said that the case he was dealing with was concerned with payments made under private transactions. But it can, I think, be inferred that the reason for this is to be found in his reference at p 381H-382A to the various statutory provisions which regulate the repayment of overpaid tax. This is why, when he came at p 382C-D to identify the first regime more precisely, he included in it not only those cases where the payment was recoverable under the Woolwich principle but also those cases which were the subject of statutory regimes regulating recovery. Here too it was unnecessary to consider a defence of common understanding, as it was open to the statutory regime which regulated recovery to deny relief where, for example, the return was made on the basis of or in accordance with prevailing practice: see, for example, the proviso to section 33(2) of TMA 1970. When he was describing these two separate and distinct regimes Lord Goff did not contemplate the possibility that there was a third category: cases concerned with the repayment of money paid under a mistake of law to a public authority which was not covered by any statutory regime regulating recovery and which, although recoverable as of right on the principle in Woolwich because it had been exacted by a demand ultra vires, was also within the scope of the newly articulated common law principle. It is into this third category that this case falls.
49.   I think that it is safe to assume that if he had appreciated that there was this third category Lord Goff would have treated it in the same way as the second. In other words he would have recognised that, as the common law remedy was available, the question whether a settled understanding defence should be available was relevant here too. The element of public interest which is lacking in the case of private transactions would have to be taken into account in considering whether there was room for such a defence. But it would have been wholly inconsistent with the general principle which he had identified for Lord Goff to conclude that there was no cause of action on the ground of unjust enrichment at common law for payments made under a mistake of law in the case of the third category just because it was also within the scope of the Woolwich principle. I respectfully agree with Sir Jack Beatson's comment in his chapter on Unlawful Statutes and Mistake of Law at p 173 that, if Lord Goff had thought that the general right of recovery did not apply to payments of taxes and general charges, he would surely have said so in his discussion of that general right.

    (b) concurrent remedies

50.   The question then is whether DMG must be denied a remedy on the ground that the payments were made under a mistake because a remedy under the Woolwich principle is available. Lord Goff treated the two categories which he identified in the debatable passage as providing separate and distinct remedies. This might be taken to suggest that, if a remedy was available to DMG by analogy with the Woolwich principle, it should not be allowed to pursue a remedy on the common law ground of unjust enrichment. But in his discussion at p 387 of the question whether in the context of void transactions failure of consideration should be allowed to trump mistake of law as a ground for recovery of benefits conferred in consequence of that mistake, he pointed out that an equally strong argument might perhaps be made in favour of mistake of law trumping failure of consideration. The same point could be made in this case. The Revenue argue that primacy should be given to a ground of recovery based on the Woolwich analogy because, if this is given, the six year limitation period will apply. DMG, on the other hand, wishes to take the benefit of section 32(1)(c) of the 1980 Act, which is why it suits it to base their claim on the general right to recover money paid under a mistake.
51.   There is no obvious way of deciding which of these two approaches must be adopted if only one can be allowed. The question however is whether a claimant is under an obligation to select the remedy that will best suit his opponent. In his note on this case, Restitution in Respect of Mistakenly Paid Tax (2005) 121 LQR 540, 542, Professor Andrew Burrows said:
"The starting point for a principled analysis is that, in general terms now that this concurrent liability has been accepted, a claimant ought to be free to choose between causes of action and that it would be odd for one cause of action, offering an advantage to a claimant, to be knocked out by a wider cause of action which does not offer that advantage."

    It would indeed be odd, and I can think of no principle that could justify such a strange result. The answer to this point is to be found in an observation by Lord Goff in a case where the question was whether a contract legislated exclusively for the parties, with the result that a parallel duty of care was excluded by it. He said that there is no sound basis for a rule which automatically restricts the claimant to either a tortious or a contractual remedy and that there could be no objection to his taking advantage of the remedy which was most advantageous to him: Henderson v Merrett Syndicates Ltd [1995] 2 AC 145, 193. We are in a different field, but I think that his reasoning is just as compelling in this context: see his reference in Kleinwort Benson [1999] 2 AC 349, 387 at p 387 to the usual preference of English law to allow either of two alternative remedies to be available, leaving any possible conflict to be resolved by election at a late stage.

    (c) the statutory regime

52.   The issue here is whether DMG's claim under the Kleinwort Benson principle is excluded on grounds of policy. The policy which the Revenue invoke is that, where there is a statutory regime for the recovery of payments made under a mistake, a common law claim cannot exist in parallel with it. The argument is that the statutory regime, of which section 33 of TMA 1970 provides the leading example, excludes recovery on the ground of mistake at common law whether the mistake is of fact or law, and whether or not the statutory regime applies to the payment that is in question. DMG, for its part, accepts that there can be no recovery at common law where the claim falls within the ambit of the statutory regime. But it submits, first, that section 33 has no application to this case and, secondly, that Parliament cannot be taken to have intended that restitution should be barred by the statutory regime where it does not provide a remedy because the payment was not made under an excessive assessment.
53.   Mr Glick QC for the Revenue said that section 33 was an exhaustive provision which covered the whole field of recovery for payments made under a mistake by the taxpayer. It did so both in respect of the mistakes for which it provided expressly and also, by necessary implication, in respect of those situations for which Parliament had deliberately chosen not to legislate. I understood him to submit that ACT fell within section 33 because it was a form of corporation tax which is charged on profits of companies and is recoverable under an assessment. Although he accepted that it was possible to envisage a case where the mistake did not fall within the terms of section 33, he said that the gap if it did exist was at best a very narrow one.
54.   The problem with fitting the payment of ACT into the regime provided for by section 33 lies in the way this tax was collected. The system that was laid down for its collection in para 1 of Schedule 13 to ICTA 1980 was for a return to be made for each of the company's accounting periods of the franked payments made during that period which was to be accompanied by the amount of the ACT, if any, payable by it in respect of those payments. It was not tax charged under an assessment, which is what section 33(1) of TMA 1970 contemplates.
55.   In support of his argument that, even if section 33 of TMA 1970 did not apply, Parliament had enacted a statutory scheme which was inconsistent with the common law remedy, Mr Glick relied on the judgment of your Lordships' House in Marcic v Thames Water Utilities Ltd [2004] 2 AC 42. I do not think that that case is in point here. Mr Marcic's claim in nuisance was held to be inconsistent with the statutory scheme. His argument was that Thames Water ought to have built new sewers to prevent flooding of his property. But, as Lord Nicholls of Birkenhead pointed out, this ignored the statutory limitations on the enforcement of sewerage undertakers' drainage obligations: para 35. An important purpose of the statutory scheme was that individual householders should not be able to launch proceedings in respect of a failure to build sufficient sewers. That would supplant the regulatory role of the industry's regulator, whose role was to decide whether to make an enforcement order when questions of flooding arose. Section 33 has none of the features of a statutory scheme of that kind.
56.   For all these reasons I would hold that the general right to recover payments made under a mistake of law on the Kleinwort Benson principle extends to the payment of taxes made to the Revenue on the mistaken belief that they were due and payable, and that DMG is entitled to take advantage of section 32(1)(c) of the Limitation Act 1980 by basing its claim for restitution on that principle.

    The mistake issue

57.   The availability of a cause of action under the Kleinwort Benson principle is of no help to DMG unless it can show that it made the payments of ACT under a mistake. The Revenue maintain that it did not make the payments under any mistake. They say that the tax was due and payable when DMG paid it because it had not made a group income election in respect of the relevant dividends.
58.   There is no doubt that the only way that a company resident in the United Kingdom could avoid liability under section 14(1) of ICTA 1988 to ACT on qualifying distributions made to its shareholders was by making an election jointly with the receiving company under section 247(1) of the Act, a group income election, that section 247(1) was to apply to the dividends received from the paying company. So long as a group income election was in force the election dividends, as section 247(1) described them, were excluded from section 14(1). But if no group income election was in force ACT was due and payable. So, if the correct approach is to look only at the system laid down by the statute, it is plain that because there was no election there was no mistake.
59.   But this approach overlooks the principle on which the claim for restitution that was recognised in Kleinwort Benson is founded, which is unjust enrichment. As Lord Goff put it at [1999] 2 AC 349, 385, it is unjust for the defendant to retain the money paid under a mistake. The essence of the principle is that it is unjust for a person to retain a benefit which he has received at the expense of another which that person did not intend him to receive because it was made under a mistake that it was due. The claimant must prove that he acted under a mistake. But the stage when he made his mistake does not matter, so long as it can be said that if he had known of the true state of the facts or of the law at the time of the payment he would not have made it. A wrong turning half way along the journey is just as capable of being treated as a relevant mistake as one that is made on the doorstep at the point of arrival.
60.   Robert Goff J, as he then was, said in Barclays Bank Ltd v W J Simms Son & Cooke (Southern) Ltd [1980] QB 677, 694, after a careful review of the leading authorities about payments made under a mistake of fact, that it is sufficient to ground recovery that the claimant's mistake should have caused him to pay the money to the payee. As Professor Burrows, The Law of Restitution, 2nd ed (2002), p 136 puts it, the type of mistake does not matter. It is purely its effect on the payer that counts. In Kleinwort Benson at p 379 Lord Goff said that it was plain that the money in that case was paid over under a mistake:
"The payer believed, when he paid the money, that he was bound in law to pay it. He is now told that, on the law as held to be applicable at the date of the payment, he was not bound to pay it. Plainly, therefore, he paid the money under a mistake of law, and accordingly, subject to any applicable defences, he is entitled to recover it."

61.   Mr Peter Thomason, DMG's Head of Taxation in London, gave evidence about his state of mind at the time when the payments of ACT were made. He made it clear in his witness statement that the ACT was paid because the relevant provisions of ICTA 1988 required it to be paid and because he believed that the UK statute denying the ability to make a group income election was the law and that he was bound to act in accordance with it. As Park J records in his judgment at para 27, he was cross-examined on his witness statement. But the judge did not believe that this passage in his evidence was challenged or affected by his answers on other points. In his opinion the mistake that DMG made when the Act was paid was that it did not realise that it could have made a valid group election with the non-resident companies. In para 29 he repeated a point that he made in para 11 when he was summarising the evidence. He said that he had no doubt that if DMG had submitted elections the Revenue would have pointed to the clear terms of the statute and rejected them, and that DMG would have been liable to pay the ACT and would have paid it.
62.   Park J acknowledged in para 25 of his judgment that DMG's mistake was not directly a mistake about whether there was a liability to pay ACT. As he put it, it was directly a mistake about whether group income elections could be made. The liabilities to pay ACT arose as secondary consequences of that primary mistake. In the Court of Appeal Jonathan Parker LJ said that he could not agree with this analysis and that DMG's mistake lay not in its belief that a group 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: paras 231-232. I think, with respect, that Park J's analysis was the correct way of looking at what happened in this case. It was the mistaken belief that group relief could not be claimed that led inevitably to the liability to pay ACT which, absent a valid claim to group relief, DMG was not in a position to dispute. That was where the mistake was made, of which the payment of ACT was a secondary consequence. But, as Park J was right to recognise, if the mistake about the availability of group income relief had not existed, the ACT would not have been paid. There was an unbroken causative link between the mistake and the payment. It follows that the payments were made under a mistake. The mistake was, of course, a mistake of law. But under the Kleinwort Benson principle a cause of action at common law for their recovery is available.

    The discovery issue

63.   The next question is when, for the purposes of section 32(1)(c) of the 1980 Act, DMG discovered its mistake or could with reasonable diligence have done so. The Revenue submit that the relevant mistake was discovered in relation to the 1993 and 1995 ACT payments in or about July 1995 when DMG learned that the provisions of section 247 of ICTA 1988 were the subject of serious legal challenge on the basis of EC law in the Hoechst case and might not be lawful. They submit that the 1996 payment was not made under a mistake as, when this payment was made, DMG was aware that the provisions of section 247 were the subject of serious challenge and might not be lawful.
64.   In support of these arguments reference is made to what I said in Kleinwort Benson [1999] 2 AC 349 about the state of mind of the payer who claims to have made a payment under a mistake. At p 409-410 I said that cases of mistake could vary from complete ignorance to a state of ample knowledge but a misapplication of what was known to the facts - from sheer ignorance to positive but incorrect belief, as Mason CJ said in David Securities Pty Ltd v Commonwealth Bank of Australia (1992) 175 CLR 353, 374. But I also said that a state of doubt was different from a mistake, and that a person who pays when in doubt takes the risk that he may be wrong. I ended this passage at p 411C-D by saying that the critical question was whether the payer would have made the payment if he had known what he is now being told was the law.
65.   These propositions are capable of further refinement: see Professor McKendrick, 'Mistake of Law' - Time for a Change, in The Limits of Restitutionary Claims: A Comparative Analysis (ed Swadling, 1997), pp 232-233; Graham Virgo, The Principles of the Law of Restitution (1999), p 161; Professor Burrows, the Law of Restitution, 2nd ed (2002), p 140. The difficult question is what degree of doubt is compatible with a mistake claim, as Professor Burrows points out. I see the issue as being essentially one of causation. What was the effect of the mistake on the payer? But the basic principle is, of course, that of unjust enrichment. At what point can it be said that the payee has been unjustly enriched? The answer to these questions will depend on the facts of the case. One can leave on one side cases where there is another ground on which the payee was entitled to be paid: frustra petis quod mox restiturus es. As for the rest, the payer's reason for making the payment despite his doubt will have a part to play in resolving the issue as to whether the payer, who would not have made the payment had he known the true state of the facts or the law at the time of the payment, should bear the risk or can recover on the ground that he was mistaken.
66.   The argument that DMG simply took the risk that it might have been wrong when it made the payments was considered and rejected by Park J. He held in para 30 of his judgment that DMG did not discover its mistake in 1995 when it learned about the argument that Hoechst was advancing. It did not do so until the decision in Hoechst was released by the European Court on 8 March 2001. In the Court of Appeal Jonathan Parker LJ said in para 234 that he agreed with the judge's conclusion on this point. But Rix LJ and Buxton LJ found it difficult to reconcile the judge's conclusion that DMG was still labouring under a mistake of law with what I said in Kleinwort Benson. Buxton LJ, recalling that that case was about the rights of the opposite party in a private law transaction, said in para 282 that the considerations which applied in such a case could only be applied very imperfectly to a citizen faced with an ultra vires demand. He said that persons who pay in response to a demand by the Revenue are in a quite different position from persons who pay under a private transaction, as the demand for tax is implicitly backed by the coercive powers of the state. He compared the Woolwich case, where the society was certain that the Revenue's demand was ultra vires but paid none the less and was restricted to a claim within the six year period, with the present case where DMG claims relief because of payment under a mistake. Although it was aware of doubts raised by other persons and of pending litigation, it paid because it thought, as it turned out incorrectly, that the law was the law. This result, he said, was reached through machinery developed to deal with the quite different issue that arises where a citizen makes a private law payment under a mistake that is his own mistake.
67.   I recognise the force of Buxton LJ's concern. But I see it as a criticism primarily of the uneven way the limitation rules operate, rather than of the conclusion which the judge drew in the light of the evidence. In so far as it amounts to a suggestion that the law ought to distinguish between payments made under a mistake of law in private transactions and those made in response to a demand by a public authority made in the mistaken belief that the law under which is was made was the law, I think that the answer to it must be found in a return to first principles. The fundamental point, as Lord Goff recognised in Kleinwort Benson at [1999] 2 AC 349, 373, is that a blanket rule of non-recovery cannot sensibly survive in a rubric of the law based on the principle of unjust enrichment. The enrichment of a public authority because a payment was made to it in response to a demand in the mistaken belief that the law under which the demand was made was the law is no less unjust than an enrichment arising from a mistake of law in a private transaction. Unevenness in the operation of the limitation rules is a matter for Parliament.
68.   The matter has, of course, been addressed so far as the Revenue is concerned by section 320 of the Finance Act 2004. But there are still some loose ends. As Lord Goff mentioned in his concluding remarks in Kleinwort Benson, this is a matter for consideration by the Law Commission with a view to finding a solution that can be made the subject of legislation by Parliament. As Professor Burrows has indicated, if there is perceived to be a problem, the right way forward is not to distort the common law of restitution by artificially limiting the scope of Kleinwort Benson but for the legislature to stop time running indefinitely in mistake cases: (2005) 121 LQR 540, 543. I would respectfully endorse his reference in this passage to mistake cases generally. There seems to be no good reason why cases of mistake of fact, the nature of which and the occasion for the discovery of which can vary greatly, should not be treated in the same way for limitation purposes as mistakes of law. Relief from the general six year time limit could be given in either case where the mistake was induced by the payee's words or conduct.
69.   What then of the evidence? Mr Thomason said in his witness statement that he learned of Hoechst's challenge to the ACT rules in around July 1995. Whilst he monitored developments, it was not clear what the outcome would be. His position was that at all times prior to the determination of the European Court in the Hoechst case the law was to be found in the UK statute. In para 45 he said that, so far as he was concerned, there were no good grounds prior to the date when the European court delivered its judgment for doubting the validity of the statutory payments requiring payment of ACT. In cross-examination he said that he did not feeling anything near certainty that Hoechst's challenge would succeed.
70.   It may be said that there was enough in Hoechst's challenge to raise a doubt in Mr Thomason's mind as to whether or not the law with which he was faced would survive it. But the situation with which he was faced is quite different from that where a payer who is in a state of doubt about the state of the law in a private transaction decides to pay nevertheless and take the risk that he may be wrong. In such a case the causative link between the mistake and the payment may be broken because the evidence shows that the payer was content to pay irrespective of whether the sum is due or not. In this case DMG was in no position to do anything else but pay the ACT. It had not claimed group income relief because, wrongly as it turned out, it believed that group income relief was not available.
71.   It is plain, as the judge recognised, that if DMG had submitted a claim for group income relief under section 247(1) the Revenue would have pointed to the clear terms of the statute and rejected it. It has never been suggested that they would have conceded in a question with DMG the point which they were resisting so strongly in their litigation with Hoechst. DMG's mistaken belief that a group income election was not available was not shown to be wrong until the issue which Hoechst had raised was determined by the European Court on 8 March 2001. The issue, which was one of law, was not capable of being resolved except by litigation. Until the determination was made the mistake could not have been "discovered" in the sense referred to in section 32(1) of the 1980 Act. In this situation I would hold, in agreement with Park J and Jonathan Parker LJ, that DMG was not in a state of doubt, when it paid the ACT. It was not then obvious that the payments might not be due. They were made in accordance with the law as it was then understood to be. There was nothing yet to be discovered to the contrary which could have been revealed by the exercise of reasonable diligence.

    Conclusion

72.   For these reasons and those given by Lord Hoffmann and Lord Walker, with which I agree, I would allow the appeal and make the order that Lord Walker proposes.

LORD SCOTT OF FOSCOTE

My Lords,

73.   This appeal is another, but assuredly not the last, of the appeals to your Lordships' House in the litigation that has arisen as a result of the judgment of the European Court of Justice (the "ECJ"), given on 8 March 2001, in the Metallgesellschaft Ltd and Hoechst AG conjoined cases, but which I will, for convenience refer to as the "Hoechst case", reported in [2001] Ch 620. The ECJ held, first, that it was contrary to Article 52 of the Treaty (now Article 43 EC) for the tax legislation of the United Kingdom to establish a tax regime under which group income elections (enabling dividends to be paid by UK subsidiaries to their parent companies without incurring an obligation to pay advance corporation tax) could be made only where the parent company was resident in the UK, and, accordingly, which denied the possibility of benefiting from a group income election where the parent company had its seat in another Member State. Secondly, the ECJ held that where a UK subsidiary had been obliged to pay ACT in respect of dividends paid to its parent company with a seat in another Member State, even though UK subsidiaries with parent companies resident in the UK were entitled in similar circumstances to make group income elections and thereby avoid that obligation, European law (i.e. Article 43 EC) required that the subsidiary and its non-resident parent
"should have an effective legal remedy in order to obtain reimbursement or reparation of the financial loss which they have sustained and from which the authorities of the [UK] have benefited as a result of the advance payment of tax by the [UK subsidiary]" (ruling 2).

74.   The issue in this particular appeal is whether and, if so, how section 32 (1) (c) of the Limitation Act 1980 applies to the cause (or causes) of action on which the appellant company ("DMG") is suing the Revenue. The factual background to DMG's action and the manner in which the action was dealt with in the courts below is comprehensively set out in the opinion of my noble and learned friend Lord Walker of Gestingthorpe. I have had the great advantage of reading his opinion in draft and gratefully adopt and need not repeat all that he has said. But as I have the misfortune to differ from him, and a majority of your Lordships, as to the right conclusion to be reached in this appeal I must refer to some of the background in order to explain the reasons for my dissent.
75.   In paragraph 117 of his opinion Lord Walker has helpfully summarised six issues in dispute between the parties. The issues which have led to my dissent are those Lord Walker has numbered (2) and (5). Issue (2) asks the question whether DMG made its payments of ACT under a mistake of law. Issue (5) asks whether DMG can bring its cause of action within section 32 (1) (c) of the 1980 Act. As to issue (5) I would respectfully venture an amendment to Lord Walker's description of the issue. Lord Walker has suggested that issue (5) arises if DMG fails on issue (1), that is to say, fails to satisfy your Lordships that English law recognises a restitutionary claim for tax paid under a mistake of law. I agree that issue (5) arises in that event, but it would arise also if DMG were to succeed on issue (1) but fail on issue (2). If that were to be the case DMG's cause of action could not be a restitutionary claim for tax paid under a mistake of law (the recognition of which by English law could not therefore assist DMG) but would have to be either a compensation claim or, perhaps, a claim based upon unjust enrichment, the Revenue having obtained payment of an amount of corporation tax earlier than would have happened if the group of which DMG formed part had been able to make a group income election. Whichever way the claim were put the question would arise whether DMG's action could be described, for section 32 (1) (c) purposes, as an "action for relief from the consequences of a mistake".
76.   My Lords, these issues, and the identification of the nature of DMG's action against the Revenue, require, in my opinion, as an essential preliminary, a careful analysis of the ECJ judgment in order to be clear as to what it was about the UK tax regime that was ruled to be a breach of European law. Park J said that
"the mistake of law which DMG made was not that it paid ACT which was not payable: the ACT was as a matter of law payable when DMG paid it, and the decision of the [ECJ] in Metallgesellschaft/Hoechst does not mean that it was not payable" (para 22).

But, in the Court of Appeal, Jonathan Parker LJ disagreed. He said, at para 231

" … the Court of Justice, at para 76, held that the relevant statutory regime was contrary to article 52, and consequently unlawful under Community law. It follows that although under the terms of that regime ACT was due and payable, 'the true state of the law' (to use the judge's expression) was that the regime gave rise to no obligation to pay. Thus the payments were made pursuant to an unlawful demand."

If the Lord Justice's analysis of the Hoechst judgment is correct, it must follow that his conclusion that DMG was under no legal obligation to pay ACT too is correct. The critical question is whether that conclusion is correct.

    The Hoechst judgment

77.   The unlawfulness found by the ECJ in the UK's ACT tax regime was the unjustified discrimination against corporate groups the parent companies of which were resident in Member States other than the UK. The discrimination consisted of withholding from those corporate groups the option of making group income elections and thereby avoiding the obligation of a UK resident subsidiary of paying ACT to the Revenue when paying dividends to its non-resident parent company. This discrimination was held to be incompatible with Article 52 because it would tend to discourage companies resident in other Member States from establishing subsidiaries in the UK. It thus constituted an unlawful restriction on the freedom of establishment guaranteed by Article 52 (see para 36 of the ECJ judgment).
78.   The ECJ was answering two questions which had been referred to it by the Chancery Division of the High Court. The first was whether it was consistent with Community law
"for the legislation of [the UK] to permit a group income election …. only where both the subsidiary and parent are resident in [the UK]" (para 33(1)).

The ECJ gave to this question the answer "no". It did so because

"The abolition [by Article 52] of restrictions on freedom of establishment also applies to restrictions on the setting up of … subsidiaries by nationals of any Member State established in the territory of another Member State" (para 41)

At paragraph 54 the ECJ explained that

"to afford resident subsidiaries of non-resident companies the possibility of making a group income election would do no more than allow them to retain the sums which would otherwise be payable by way of [ACT] until such time as Mainstream Corporation Tax [MCT] falls due. They would thus enjoy the same cashflow advantage as resident subsidiaries of resident parent companies…."

And, at paragraph 55, that

" … the fact that a non-resident parent company will, unlike a resident parent company, not be subject to [ACT] when it in turn pays out dividends, cannot justify denying the resident subsidiary of the non-resident parent the possibility of exemption from payment of [ACT] when paying dividends to the parent."

And, in paragraph 76, the ECJ answered the first question by declaring that it was

" … contrary to Article 52 of the Treaty for the tax legislation of [the UK] … to afford companies resident in [the UK] the possibility of benefiting from a taxation regime allowing them to pay dividends to their parent company without having to pay [ACT] where their parent company is also resident in [the UK], but to deny them that possibility where their parent company has its seat in another Member State."

The reference by the ECJ in each cited passage to "the possibility" of making a group income election, a "possibility" that the tax regime denied to corporate groups with a non-resident parent company, identifies the element in the tax regime that the ECJ found to be inconsistent with European law. 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 which failed to give corporate groups with a non-resident parent company the same possibility of enabling resident subsidiaries to avoid the payment of ACT as was open to corporate groups with a resident parent company. 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 in respect of dividends paid to their non-resident parent companies. This is a question to which I must return.

79.   The second question for the ECJ to answer concerned remedies. The Chancery Division had asked, if the answer to the first question was 'no', about the remedies which national courts should grant. In paragraph 77 of its judgment the ECJ explained the question. Was the subsidiary and/or its parent company
" … entitled to obtain a sum equal to the interest accrued on the advance payments made by the subsidiary from the date of those payments until the date on which the tax became chargeable [i.e. until MCT became payable], even when national law prohibits the payment of interest on a principal sum which is not due."

The ECJ recorded that the Chancery Division had framed that question in two hypotheses -

" … in the first alternative, where the claim by the subsidiary and/or parent company is made in an action for restitution of taxes levied in breach of Community law and, in the second, where the claim is made in an action for compensation for damage resulting from the breach of Community law."

In paragraph 81, however, the ECJ stressed that it was not for the ECJ but for the parties and the national law of the forum to specify the nature and basis of the remedy or remedies being sought. Community law required no more than that

" … resident subsidiaries and their non-resident parent companies should have an effective legal remedy in order to obtain reimbursement or reparation of the financial loss which they have sustained and from which the authorities of the Member State concerned have benefited as a result of the advance payment of tax by the subsidiaries" (para 96)

80.   The Hoechst judgment contains passages which cast light upon the nature of the breach of Community law for which the appropriate remedy must be provided. Thus:
"It is important to bear in mind in this regard that what is contrary to Community law … 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" (para 83);

and, to the same effect,

" … the breach of Community law arises, not from the payment of the tax itself but from its being levied prematurely" (para 87);

and

" … the principal sum due is none other than the amount of interest which would have been generated by the sum, use of which was lost as a result of the premature levy of the tax" (para 88)

81.   These passages show that 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. Does it follow that section 14 (1) of the Income and Corporation Taxes Act 1988 (originally section 84 (1) of the Finance Act 1972), which imposed the obligation on UK companies to pay ACT in respect of dividends paid to their shareholders, was unlawful, or, to put the question another way, imposed an unlawful tax? In my opinion, it does not. The effect of the ECJ judgment in the Hoechst case on the validity of the section 14 (1) charge to tax is a matter for the domestic courts of the UK. (See Thoburn v. Sunderland City Council [2002] EWHC 195 (Admin); [2003] QB 151 paras 68 and 69) If, under domestic law, the right conclusion is that section 14 (1) imposed an unlawful tax, 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. 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. If ACT was an unlawful tax, the tax would have been paid under a mistake of law and, in any event, would be recoverable under the principle established in the Woolwich case [1993] AC 70. This would be a remarkable state of affairs. There is nothing, however, in the ECJ's Hoechst judgment that requires this conclusion and no reason at all why the domestic courts should reach it.
82.   In my opinion, the proposition that the Hoechst judgment requires the conclusion that section 14 (1) imposed an unlawful tax cannot be sustained. 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. So how should the domestic law deal with that unlawfulness? It would clearly be an inappropriate response for the domestic courts to hold that the benefits afforded by the group income election provisions to domestic corporate groups were unlawful and ultra vires. Nor was that suggested by the ECJ. The simple answer is that the UK courts must provide a remedy to the non-domestic corporate groups for the failure of the statutory provisions to extend the same benefits to them. The ECJ made it clear that it was no concern of Community law how, as a matter of domestic law, that remedy was to be categorised provided the remedy was an adequate and sufficient one (para 96). However the remedy cannot, in my opinion, be based on the premise that the payments of ACT were a response to unlawful demands or were payments of an unlawful tax or were payments that were not due. None of these premises is required to be made by the ECJ's judgment in the Hoechst case and none should be constructed by UK domestic law.
83.   The first issue referred to by Lord Walker, the issue to which more time was devoted by counsel than to all the other issues combined, was whether English law recognised a restitutionary claim for tax paid under a mistake of law. On this issue I am in full agreement with, and there is nothing I wish to add to, the opinions given by my noble and learned friends. I agree that the effect of the speeches in this House in the Woolwich case [1993] AC 70 is that tax paid in response to an unlawful demand can be recovered by the taxpayer regardless of compulsion and regardless of whether the taxpayer made the payment in consequence of a mistake; and I agree that the effect of the speeches in the Kleinwort Benson case [1999] 2 AC 349 was that the rule barring a restitutionary remedy for money paid, or property transferred, under a mistake of law was overturned. Accordingly, I agree that the answer to issue (1) is that English law does now recognise a restitutionary remedy for tax paid under a mistake of law.
84.   But the answer to issue (1) does not mean that money paid, or property transferred, under a mistake of law is necessarily recoverable, just as money paid, or property transferred, under a mistake of fact is not necessarily recoverable. It surely all depends on the part played by the mistake, whether of fact or law, in the sequence of events that has led to the payment or transfer. If A and B enter into a contract with one another for the sale by A to B of a horse and B pays the price and takes delivery of the horse, B cannot, absent some causative misrepresentation on A's part, claim his money back, proffering the horse in return, on the ground that he was mistaken as to the horse's breeding. The money once paid would be irrecoverable unless there were some ground for invalidating the contract. It cannot be enough for the buyer to assert and prove that but for his mistake about the horse's breeding he would not have entered into the contract and so would not have paid the money. In Barclays Bank Ltd v. W J Simms Son & Cooke (Southern) Ltd [1980] QB 677 at 695 Robert Goff J (as he then was) said that
"if a person pays money to another under a mistake of fact which causes him to make the payment, he is prima facie entitled to recover it as money paid under a mistake of fact" (emphasis added).

He then described three types of circumstance which might lead to the claim failing. The first of these was that the payor intended the payee to have the money in any event, whether the believed fact was true or false. The second was that

" … the payment is made for good consideration, in particular if the money is paid to discharge, and does discharge, a debt owed to the payee … by the payer or by a third party by whom he is authorised to discharge the debt…"

The third was that the payee had changed his position in good faith. There is, I would respectfully suggest, a fundamental difference between the first and the third of these types of circumstances on the one hand and the second of them on the other. The first and third involve circumstances which have had no causative effect on the payment of the money. In the first, the money would have been paid in any event and the third involves circumstances which have nothing at all to do with the reasons why the payer paid the money or with any misapprehensions under which the payer may have been labouring. Neither of these types of case invalidates Robert Goff J's general proposition that if a mistake of fact causes a payment to be made that would not have been paid but for the mistake, the payer will have a cause of action for its recovery. They are not true exceptions. The second, however, does invalidate that proposition. If a contract has been entered into that would not have been entered into but for a mistake, but the contract is then completed by a payment of the price for the goods or services that the payee has supplied, the payment cannot be recovered unless the contract can be set aside. The proposition seems such an obviously correct one that it may seem pointless to ask why it is that it is correct. But I think the question does need to be asked for the answer casts, in my opinion, valuable light on the nature of the restitutionary remedy for the recovery of money paid under a mistake.

85.   The reason, it seems to me, why the proposition is correct is that the mistake does not necessarily undermine the legal obligation which required the payment of the money or for the discharge of which the money was paid. If the mistake does enable the contract to be set aside then, subject to a change of position defence, the money should be recoverable. If the contract was void from the outset (as in the "swaps" cases) or had been avoided before the payment was made, the money should be recoverable. But if the legal obligation under which the money was paid cannot be, or has not been, invalidated, then, in my opinion, whether or not it can be shown that "but for" the mistake in question the money would not have been paid, a restitutionary remedy for the recovery of the money would not be available.
86.   This approach to the recovery of money paid under a mistake should, in my opinion, apply whether the mistake be of fact or of law. In the Woolwich case [1993] AC 70 the tax paid by the Woolwich was recoverable because it had been paid on account of a non-existent legal obligation. In the Kleinwort Benson [1999] 2 AC 349 case the payments of which recovery was sought by the bank had been paid under contracts that were void ab initio. In neither case could it have been suggested that the payments had been made with any animus donandi.
87.   There are, I think, some problems about voluntary payments made as gifts but that would not have been made but for some causative mistake, whether of fact or law, e.g. a gift of £1000 by A to B where B is believed by A to be impecunious but is in fact a person of substantial wealth and where A would not have made the gift if he had known that to be so. My present opinion is that unless there were some other reason, such as a misrepresentation by B, to enable the gift to be set aside, the mistake made by A would not suffice, notwithstanding that the payment had not been made pursuant to any legal obligation and that but for the mistake it would not have been made. But the availability of a restitutionary remedy to recover gifts which would not have been made but for some mistake of fact or of law does not need to be pursued on this appeal and can be left for another day.
88.   In my opinion the important question for the purposes of this appeal is whether the payments by DMG of the ACT were made in discharge of a legal obligation to pay. Park J thought they were. Jonathan Parker LJ thought not. For the reasons I have already expressed I agree with Park J. The mistake made by DMG and its parent company in Germany was in not realising that the failure of the ACT tax regime to allow the group to make a group income election was unlawful and could successfully have been challenged. If the group had purported to make a group income election the Revenue would certainly have rejected it. That rejection we now know, post Hoechst, would have been unlawful under Community law but even so, in my opinion, the lawfulness of section 14 (1) would not have been undermined. What might, perhaps, have been undermined would have been the Revenue's ability to recover the ACT in an action against DMG since any claim for payment might have been met by an equitable set-off or cross claim based upon the Revenue's breach of Community law in rejecting the group income election.
89.   In the events that actually happened DMG paid the ACT that was due under section 14 (1). DMG and its German parent company did make a mistake of law in that they did not realise that they could successfully challenge the failure of the ACT tax regime to allow them, too, to make a group income election. It is possible to argue that this mistake of law was, in a "but for" sense, a cause of their payment of the ACT. The argument has difficulties but even if it were right it would not, in my opinion, justify the conclusion that the ACT was paid under a mistake of law in the sense necessary for a restitutionary remedy. The ACT was paid because there was a legal obligation under a valid statutory provision for the money to be paid. DMG's remedy was, in my opinion, not a restitutionary one for the re-payment of money paid that was not due, or for the repayment of money paid under a mistake, but was, and is, a claim for compensation to recover the loss caused by the breach of Community law.
90.   Before leaving the important issue of whether, in the light of Hoechst, DMG was under a legal obligation to pay the ACT, I should record the assistance I have had from the article "Justified Enrichment" by Robert Stevens (2005) 5 OUCLJ 141 and in particular from the section entitled "Was the Money Due". I am in respectful agreement with the conclusions there expressed.

DMG's cause of action and section 32 (1) (c) of the 1980 Act

91.   It is common ground that DMG has a cause of action in tort for compensation for the loss caused by the breach of Community law found by the ECJ in the Hoechst case to be inherent in the ACT tax regime. It is also common ground that for Limitation Act purposes time began to run when each payment of ACT was made. The details of the payments are set out in paragraph 111 of Lord Walker's opinion. Section 32 (1) (c) applies where an action is brought "for relief from the consequences of a mistake". In Phillips-Higgins v. Harper [1954] 1 QB 411 Pearson J expressed the view at p 419 that section 26 (c) of the Limitation Act 1939 (the statutory predecessor of section 32 (1) (c) of the 1980 Act)
" … applies only where the mistake is an essential ingredient of the cause of action …"

If it is right, as I think it is, that when the ACT was paid by DMG the ACT was due, DMG's cause of action is an action for compensation for tort. It is not, in my opinion, an action for restitution based on the payment of money under a mistake. An allegation of a mistake but for which the ACT would not have been paid is not an essential ingredient in DMG's cause of action.

92.   I would accordingly dismiss this appeal albeit for reasons different to those of the Court of Appeal.

The other issues

93.   In each of the other issues I am in respectful agreement with the opinions of my noble and learned friends, and accordingly, if it were relevant to do so, would dismiss the Revenue's cross-appeal.

LORD WALKER OF GESTINGTHORPE

My Lords,

    Introduction

94.   Article 43 (formerly 52) of the EU Treaty, headed 'Right of Establishment' provides as follows:
"Within the framework of the provisions set out below, restrictions on the freedom of establishment of nationals of a Member State in the territory of another Member State shall be prohibited. Such prohibition shall also apply to restrictions on the setting-up of agencies, branches or subsidiaries by nationals of any Member State established in the territory of any Member State.

Freedom of establishment shall include the right to take up and pursue activities as self-employed persons and to set up and manage undertakings, in particular companies or firms within the meaning of the second paragraph of Article 48, under the conditions laid down for its own nationals by the law of the country where such establishment is effected, subject to the provisions of the chapter relating to capital."

95.   Several different statutory provisions relating to United Kingdom corporation tax (in particular, as it applies to multi-national groups or consortia of companies) have been successfully challenged, or are at present under challenge, as infringing Article 43. The challenges have been so numerous that litigation is proceeding under several different group litigation orders ("GLOs") made in the Chancery Division, and some appeals have already reached this House: Pirelli Cable Holding NV v Inland Revenue Commissioners [2006] UKHL 4; [2006] 1 WLR 400 and (as to procedural matters) Autologic Holdings plc v Inland Revenue Commissioners [2005] UKHL 54; [2006] 1 AC 118. At the end of 2005 the Court of Justice of the European Communities (the "ECJ") gave judgment on a reference made by the Chancery Division in litigation of this type (though not proceeding under a GLO): Marks & Spencer plc v Halsey (Case C-446/03), [2006] Ch 184.
96.   The particular focus of the present appeal is on limitation of actions - in practical terms, the length of time during which the premature payment of advance corporation tax ("ACT") by the appellant company Deutsche Morgan Grenfell Group plc ("DMG") gives rise to a remedy enforceable in the national court. DMG's claim has been selected (under a GLO made on 26 November 2001) as a suitable test case for the determination of what are defined (in a further order made on 12 March 2002) as "EU issues (i) Limitation (A) and (B)." DMG was successful before Park J (whose judgment given on 18 July 2003 is reported at [2003] 4 AER 645, [2003] STC 1017) but on 4 February 2005 the Court of Appeal allowed the appeal of the Commissioners of Inland Revenue ([2006] 2 WLR 103, [2005] STC 329). Her Majesty's Attorney General is also a party to the proceedings but I shall refer to the respondents simply as "the Revenue".
97.   The issues summarised above may appear fairly narrow and very technical, and on one possible view this appeal could be disposed of as a relatively short point of construction on section 32(1)(c) of the Limitation Act 1980. But it has produced some complex and sophisticated arguments from counsel. The decisions of Park J and the Court of Appeal have also led to some stimulating and far-reaching comments from legal scholars, including Professor Andrew Burrows, "Restitution in respect of Mistakenly Paid Tax", (2005) 121 LQR 540, Professor Steve Hedley, "Tax Wrongly Paid - Basis of Recovery - Limitation" [2005] CLJ 296, Robert Stevens, "Justified Enrichment" (2005) 5 OUCLJ 141, Graham Virgo, "Deutsche Morgan Grenfell: the right to restitution of tax paid by mistake rejected" [2005] BTR 281, Sir Jack Beatson, Chapter 9 ("Unlawful Statutes and Mistake of Law") in Burrows and Rodger (Eds), Mapping the Law (to be published shortly) and two articles by Dr James Edelman, "Limitation Periods and the Theory of Unjust Enrichment" (2005) 68 MLR 848 and "The Meaning of 'Unjust' in the English Law of Unjust Enrichment" (2006) 3 ERPL 309. I shall return to some of these at a later stage in this opinion, but I wish to acknowledge at once my debt to these commentators and to other academic writers, including of course the most recent work of the late Professor Peter Birks.
98.   I shall begin with a brief summary of the relevant statutory provisions relating to ACT, followed by a rather fuller summary of the decision of the ECJ in the important joined cases of Metallgesellschaft Ltd and others v Inland Revenue Commissioners and another and Hoechst AG and another v Inland Revenue Commissioners and another (Joined Cases C-397 and C-410/98) [2001] Ch 620 ("Hoechst"). Then I will come to the facts of the test case, the pleadings and the course of the litigation on its way to your Lordships' House. Then it will be necessary to look in some detail at the very important decisions of this House in Woolwich Equitable Building Society v Inland Revenue Commissioners [1993] AC 70 ("Woolwich") and Kleinwort Benson Limited v Lincoln City Council [1999] 2 AC 349 ("Kleinwort Benson"), both majority decisions in which the leading speech was given by Lord Goff of Chieveley.
ACT
99.   The relevant statutory provisions are very clearly described in the judgment of Park J (paras 6 and 7) and a brief summary will therefore suffice. Corporation tax was first introduced by the Finance Act 1965. The structure of the tax was radically altered by the Finance Act 1972 with the introduction of ACT, which was in force from 1973 until its abolition (probably hastened by the oncoming wave of litigation based on Article 43) in 1999. From 1988 until 1999 the relevant provisions were to be found in a consolidating statute, the Income and Corporation Taxes Act 1988 ("ICTA"). The political and economic background to these changes has been described by Professor John Tiley, Revenue Law 4th ed (2000), pp 758-762.
100.   When a company paid a dividend to its shareholders it was required to pay to the Revenue (within 14 days from the end of the quarter in which the dividend was paid) an amount of ACT equal (for most of the relevant period) to 25% of the dividend (ICTA section 14). The ACT paid in this way was later set off against the company's liability for mainstream corporation tax ("MCT") if (as happened in this case) the company made sufficient taxable profits to be liable to pay at least as much MCT as it had paid ACT. The payment of MCT would thus be accelerated by a period of between 8½ and 17½ months (depending on the timing of the dividend payment). A subsidiary company paying a dividend to its parent company could however avoid liability to pay ACT if both companies made a group income election ("GIE") under ICTA section 247. The GIE had to be made at least three months in advance of payment of the dividend. This relief was however available only if both companies were for tax purposes resident in the United Kingdom. So a dividend paid by a British subsidiary to a German parent company was not (on the face of the United Kingdom corporation tax legislation) able to make a GIE and so obtain relief under section 247. In Hoechst [2001] Ch 620 the ECJ declared that part of the corporation tax legislation to be unlawful under EU law. The judgment of the ECJ also deals with the consequences of the infringement in terms of remedies in the national courts.

Hoechst

101.   The claimants in these two joined cases were two or more companies in groups which had a German parent company and one or more United Kingdom subsidiaries. In actions against the Revenue (in which the Attorney General was also made a defendant) the claimants sought damages or compensation for loss (during the period until they would have been set off against MCT) of sums paid in ACT by the English subsidiaries in respect of dividends paid to the German parent companies. The complaint was that these losses were caused by the discriminatory and unlawful character (under EU law) of the ACT regime under which GEIs were not available to these multinational groups, but would have been available if all the relevant members of the group had been United Kingdom companies.
102.   Claims were brought in 1995 in respect of ACT paid (in one case) between 1974 and 1995 and (in the other case) between 1989 and 1994. In 1998 the Chancery Division made references to the ECJ for preliminary rulings. Advocate General Fennelly delivered his opinion on 12 September 2000 and the ECJ gave its judgement on 8 March 2001. The latter date has come to be seen as being of considerable importance in these proceedings, so far as they turn on mistake of law.
103.   On the first referred question (paras 35-76 of its judgment) the ECJ concluded (para 76)
"that it is contrary to article 52 [now 43] 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"

In reaching that conclusion the ECJ rejected the argument (put forward by the United Kingdom and some other member states) that the difference in tax treatment was objectively justified by the differing circumstances of national and multinational groups, and in order to preserve the coherence of the United Kingdom's tax system.

104.   The second question (set out in para 33(2) of the ECJ's judgment) was whether EU law gave rise to a restitutionary right to claim a sum of money by way of interest, or whether the claimant was limited to an action for damages, and
"in either case is the national court obliged to grant a remedy even if under national law interest cannot be awarded (whether directly or by way of restitution or damages) on principal sums which are no longer owing to the claimants?"

The rule of national law alluded to was that upheld by this House in President of India v La Pintada Compania Navigacion SA [1985] AC 104, mentioned in para 12 of the opinion of the Advocate General.

105.   The whole of the ECJ's answer to the second question (paras 77-96) is set out in full in the judgment of Jonathan Parker LJ (para 45) and it is unnecessary to repeat it in full. The crucial passages (in the text as slightly edited in the Law Reports, [2001] Ch 620, 663-664), are as follows:
"81.  It must be stressed that it is not for the Court of Justice to assign a legal classification to the actions brought by the claimants before the national court. In the circumstances, it is for the claimants to specify the nature and basis of their actions (whether they are actions for restitution or actions for compensation for damage), subject to the supervision of the national court. …

83.  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.

84.  According to well-established case law, the right to a refund of charges levied in a member state in breach of rules of Community law is the consequence and complement of the rights conferred on individuals by Community provisions as interpreted by the court: Amministrazione delle Finanze dello Stato v Sp A San Giorgio (Case 199/82) [1983] ECR 3595, 3612, para 12 [etc]… The member state is therefore required in principle to repay charges levied in breach of Community law…

85.  In the absence of Community rules on the restitution of national charges that have been improperly levied, it is for the domestic legal system of each member state to designate the courts and tribunals having jurisdiction to lay down the detailed procedural rules governing actions for safeguarding rights which individuals derive from Community law, provided, first, that such rules are not less favourable than those governing similar domestic actions (principle of equivalence) and, secondly, that they do not render practically impossible or excessively difficult the exercise of rights conferred by Community law (principle of effectiveness)…

86.  It is likewise for national law to settle all ancillary questions relating to the reimbursement of charges improperly levied, such as the payment of interest, including the rate of interest and the date from which it must be calculated…

87.  In the main proceedings, however, the claim for payment of interest covering the cost of loss of the use of sums paid by way of [ACT] is not ancillary, but is the very objective sought by the claimants' actions in the main proceedings. In such circumstances, where the breach of Community law arises, not from the payment of the tax itself but from its being levied prematurely, the award of interest represents the 'reimbursement' of that which was improperly paid and would appear to be essential in restoring the equal treatment guaranteed by article 52 [now 43] of the Treaty.

88.  The national court has said that it is in dispute whether English law provides for restitution in respect of damage arising from loss of the use of sums of money where no principal sum is due. It must be stressed that in an action for restitution the principal sum due is none other than the amount of interest which would have been generated by the sum, use of which was lost as a result of the premature levy of the tax."

106.   The fifth question, though phrased in abstract terms, asked in effect whether the United Kingdom could rely on the defence that the two groups of companies did not seek to make GIE's and (para 98)
"did not therefore 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 on any view national law denied resident subsidiaries and their non-resident parent companies the benefit of that taxation regime".

In rejecting this argument the ECJ invoked the principle of effectiveness (para 106):

"The exercise of rights conferred on private persons by directly applicable provisions of Community law would, however, be rendered impossible or excessively difficult if their claims for restitution or compensation based on Community law were rejected or reduced solely because the persons concerned had not applied for a tax advantage which national law denied them, with a view to challenging the refusal of the tax authorities by means of the legal remedies provided for that purpose, invoking the primacy and direct effect of Community law"

107.   The Revenue's position before your Lordships is to concede that the ECJ expects DMG to be entitled to a restitutionary remedy, but only (as Mr Glick QC put it) by analogy with the remedy in Woolwich [1993] AC 70 (and, crucially, subject to the usual 6-year limitation period). The Revenue contend that DMG is trying to obtain a concurrent domestic remedy in circumstances where, if the case were purely domestic, it would get nothing. This is of course only a brief summary of its argument on an important point to which it will be necessary to return.

The facts

108.   As just noted, DMG is seeking a concurrent domestic remedy based on mistake of law, because it wishes to rely on section 32(1)(c) of the Limitation Act 1980. Section 32(1) provides that (subject to immaterial exceptions)
"…where in the case of any action for which a period of limitation is prescribed by this Act, either -

a)  the action is based upon the fraud of the defendant; or

b)  any fact relevant to the plaintiff's right of action has been deliberately concealed from him by the defendant; or

c)  the action is for relief from the consequences of a mistake;

the period of limitation shall not begin to run until the plaintiff has discovered the fraud, concealment or mistake (as the case may be) or could with reasonable diligence have discovered it."

It was therefore necessary for Park J to make findings of fact relevant to this issue, and he heard oral evidence from Mr Peter Thomason, DMG's Head of Taxation, who was cross-examined by Mr Glick. Mr Glick took your Lordships through quite a lot of his oral evidence but before coming to the judge's findings on the issue of mistake I should record the undisputed facts about DMG, the group of which it came to form part, and the relevant payments of ATC.

109.   DMG (formerly Morgan Grenfell Group Plc) was established in 1971 as a holding company for Morgan Grenfell & Co Limited, a long-established merchant bank. DMG was listed on the London Stock Exchange in 1986. Deutsche Bank AG ("DBAG"), a well-known German bank, had acquired a 5% stake in DMG in 1984 and in 1989 it made a successful takeover bid. DBAG became the direct holder of about 14% of the shares in DMG and the remainder (about 86%) were held through another United Kingdom subsidiary of DBAG, DB Investments (GB) Limited ("DBI"). At first DMG continued to operate in a fairly independent way, despite the takeover, but in October 1994 DBAG decided to adopt a more integrated approach and to consolidate all its investment banking opportunities in London. But the integration was not fully achieved, as far as the tax departments were concerned, until May 1996. Until then Mr Thomason has responsibility for the tax affairs of DMG but not for those of DBI (or, of course, DBAG).
110.   I gratefully adopt Park J's summary of the history of the dividend payments:
"10.  According to the terms of ICTA, section 247, group income elections could not have been made between DMG and DBAG or between DBI and DBAG because DBAG was not a company resident in the United Kingdom. A group income election could have been made between DMG and DBI (because both companies were resident in the United Kingdom), but no such election was made. The reason was that, although an election would have enabled dividends to flow from DMG to DBI without ACT, onward dividends from DBI to DBAG would have had to be paid subject to a liability on the part of DBI to pay ACT: or at least that was how the matter appeared on the terms of the domestic United Kingdom legislation. In the circumstances DMG paid ACT by reference to all its dividends - the 86% of them paid to DBI as well as the 14% of them paid to DBAG - and DBI did not pay ACT when it paid onward dividends to DBAG. DBI did not have to pay ACT on those onward dividends because of the rules about franked investment income in ICTA, sections 238 and 241.

11.  DMG has pleaded in this case, and the Revenue admit, that, if section 247 had permitted group income elections to be made between a United Kingdom subsidiary and a parent company in another member state, elections would have been made between DMG and DBI, between DMG and DBAG and between DBI and DBAG. The corollary is that, because section 247 appeared clearly not to permit group income elections to which a parent company in another member state was a party, the companies did not attempt to make group income elections. There is no doubt that, if they had attempted to make them, the Revenue would have rejected the elections and pointed to the clear terms of the United Kingdom statute in justification of the rejection."

111.   The amounts of ACT paid by DMG are now agreed to have been as follows (rounding pence):

Date of Payment	On dividends to DBAG	On dividends to DBI
14 October 1993	£887,076	£5,472,956
15 February 1995	£41,843	£258,157
14 January 1996	£487,471	£3,007,529

These figures have however emerged by a surprisingly laboured process. There has been more than one amendment of the particulars of claim. Further discussion of this point is best deferred until I come to the last issue in the appeal, the pleading point.

112.   The judge accepted that at the time of the ACT payments in October 1993 and February 1995 DMG (primarily in the person of Mr Thomason) knew nothing about the argument which was ultimately successful in Hoechst. But by July 1995 DMG knew about the Hoechst litigation, which received a good deal of publicity in the financial press. There was some discussion at DMG about the possibility of a 'protective GIE' (Mr Thomason was cross-examined about this) but in fact nothing was done before a dividend was paid on 17 July 1995 (giving rise to the ACT payment in January 1996).
113.   The judge's main findings of fact on this point are set out in para 27 of his judgment:
"Before I move on I wish to record the evidence which Mr Thomason (the Head of Taxation at DMG) gave about his state of mind when he learned of the Hoechst argument. In his witness statement he said this:

'At all times prior to the determination of the European Court in the Hoechst case, I believed that the United Kingdom statute denying the ability to make a group income election was the law and I was bound to act in accordance with this law…It did not occur to me that I could ignore the law as it stood for the simple reason that the law is the law. Just because another taxpayer challenged the law that did not mean that I could or should ignore it.'

Mr Thomason was cross-examined, but I do not believe that the foregoing passage was challenged or affected by his answers on other points. He added the general point (obvious but plainly relevant) that it was not clear in 1995 what would be the outcome of the Hoechst case (or, as it turned out, of the Metallgesellschaft/Hoechst conjoined cases). He also said that there were more arguments being advanced by Hoechst in 1995 than the one which eventually succeeded in the [ECJ]."

114.   The judge concluded (paras 30-31) that all three relevant payments of ACT were made under a mistake of law and that DMG did not discover its mistake until the decision of the ECJ on 8 March 2001, after DMG had commenced proceedings against the Revenue (it did so shortly after delivery of the opinion of Advocate General Fennelly).

The Pleadings

115.   DMG issued its claim form on 18 October 2000, describing the claim as "for damages for breach of European Community law". In its original particulars of claim (served on 9 February 2001) it pleaded that ICTA sections 14 and 247 and Schedule 13 ("the statutory provisions") were contrary to EU law. It pleaded (paras 9 and 11) that from time to time DMG paid dividends to DBAG and DBI (which passed them on to DBAG). Para 9 (but not para 11 in its original form) pleaded that ACT was paid in respect of these dividends "pursuant to the unlawful statutory provisions and to unlawful demands made by the First and/or Second Defendant and further and/or alternatively under a mistake of law as to the validity of the statutory provisons". Para 11 was later amended to bring it into line with para 9. By para 13 DMG claimed to be entitled to "restitution of, and further and/or alternatively compensation for, and further and/or alternatively compensation for the loss of use of, monies paid on account of ACT pursuant to unlawful demands by the First and/or Second Defendant and/or under a mistake of law and pursuant to the statutory provisions in respect of the DBAG dividends" (later amended to both sets of dividends), compensation for loss of use of the monies to be calculated "from the dates of payment until set-off against mainstream corporation tax or other repayment or utilisation".
116.   Paras 9 and 11 each ended with an averment that the amounts of ACT "include, but are not limited to, payments of ACT made by the Claimant as specified in the [First or Second] Schedule hereto." Both Schedules (giving dates and amounts of ACT payments) were substantially amended on various later dates. The pleading issue is whether these amendments added or substituted a new cause of action outside the category permitted by section 35(5)(a) of the Limitation Act 1980 ("if the new cause of action arises out of the same facts or substantially the same facts as are already in issue on any claim previously made in the original action.")

    The issues

117.   Before your Lordships, as in the courts below, there have been six issues in dispute. Drawing on both the agreed statement of facts and issues and the competing version in para 5 of DMG's printed case (but finding the latter rather more helpful) I would summarise them as follows.

    (1)  The "cause of action" issue: does English law recognise a restitutionary claim for tax paid under a mistake of law?

    (2)  The "mistake" issue (arising if DMG succeeds on issue (1)): did DMG make the payments of ACT under a mistake of law?

    (3)  The "discovery" issue (arising if DMG succeeds on issues (1) and (2)): when did DMG discover (or could with reasonable diligence have discovered) its mistake?

    (4)  The "settled law" issue (arising if DMG succeeds on issues (1) to (3)): does English law recognise a defence of "settled law" to a claim for restitution of tax paid under a mistake of law?

    (5)  The "scope of section 32(1)(c)" issue (arising if DMG fails on issue (1), and raised for the first time in this House): can DMG bring itself within section 32(1)(c) even if mistake of law is not an essential element of its cause of action?

    (6)  The "pleading" issue (raised by the Revenue's cross-appeal and linked to the limitation issues): this is the point on section 35 of the Limitation Act 1980 already identified.

118.   It will be apparent that it may not be strictly necessary for your Lordships to resolve each of these issues in order to dispose of the appeal. But as this is a test case under the GLO, and the facts of other cases will differ, it seems desirable to consider all the issues. On DMG's printed case there is a seventh issue, the "expiration" issue, arising only if DMG cannot rely on section 32(1)(c), and if the Revenue succeeds on its cross-appeal. This is a truly arcane point raised (but then left undecided) by Park J in paras 39 and 40 of his judgment, and not discussed at all in the Court of Appeal. Like Park J I would leave it for another day unless there is some practical need to answer it.

    Woolwich and Kleinwort Benson

119.   The speeches delivered in this House in Woolwich [1993] AC 70 and Kleinwort Benson [1999] 2 AC 349 have attracted a great deal of academic interest and commentary. The speeches (especially those of Lord Goff) were also subjected to very close scrutiny in the courts below. Buxton LJ warned himself (para 265) against treating Lord Goff's speeches as if they were statutes, but the fact is that they were in the courts below subjected to the same sort of minute examination as would be appropriate for a statutory (or even, if I may say so without disrespect, a scriptural) text. Your Lordships have been invited to undertake the same sort of exercise.
120.   That invitation cannot be wholly disregarded. But before getting too caught up in the exegesis of Lord Goff's speeches, and at the danger of stating the obvious, I would by way of preliminary make some simple observations about the issues in the two cases, and about all the speeches which were delivered (in each case there were quite full speeches from every member of the Appellate Committee, with two dissenting speeches).
121.   To my mind the salient points about Woolwich [1993] AC 70 were as follows:

    (1)  It was decided in 1992 when the general rule barring recovery of money paid under a mistake of law still held the field.

    (2)  In any case there was no mistake on the part of the building society, which had from the first challenged the validity of the transitional provisions in the relevant regulations. It seems to have received confident professional advice and to have acted confidently on the advice. It paid the tax under protest, which was a reasonable thing to do since thousands of small investors might have been alarmed by the news that it was said to be in default.

    (3)  The claim (as originally formulated) was for the repayment of a principal sum of almost £57m with interest. After the issue of the writ the principal sum was repaid (with interest from the date of the writ). So only pre-writ interest was eventually at stake, but there had been a principal sum claimed to be due when the writ was issued.

    (4)  The majority (Lord Goff, Lord Browne-Wilkinson and Lord Slynn of Hadley) upheld the simple principle that tax paid in response to an unlawful demand could be recovered, regardless of compulsion, and that this result was not ousted (or "trumped") by any concurrent mistake of law (Lord Goff, at p 177 F-H, Lord Browne-Wilkinson, at p 198 C-H, agreeing with Lord Goff but not expressly mentioning mistake of law; Lord Slynn, at pp 204 H-205 A).

    (5)  The minority (Lord Keith of Kinkel and Lord Jauncey of Tullichettle) considered that there could be no recovery in the absence of compulsion or duress, and that there was no relevant distinction between payment of tax in response to a demand under an ultra vires regulation and payment in response to a demand under a misconstrued statute. If the statutory provisions for repayment of overpaid taxes were inadequate that was a matter for Parliament.

    (6)  The case had no European element at all, though Lord Goff did refer (at p 177 C-E) to the San Giorgio case [1983] ECR 3593, remarking that it would be strange if the right of the citizen to recover overpaid charges were to be more restricted under domestic law than it is under European law.

122.   The salient points about Kleinwort Benson [1999] 2 AC 349 were as follows:

    (1)  It was not about tax but about interest rate swaps.

    (2)  There were arguably concurrent restitutionary claims, that is a claim based on money paid under a void contract and (if and only if the House abrogated the long-standing bar on recovery of money paid under a mistake of law) a claim on that ground.

    (3)  The significance of section 32(1)(c) of the Limitation Act 1980 was clearly recognised; that was the whole point of the appellant bank seeking to overturn the rule barring recovery for mistake of law. The majority decided to overturn the rule despite explicitly acknowledging that an indefinitely extended limitation period might produce unsatisfactory results (Lord Goff at p 389 A-C; Lord Hoffmann, at p 401 D-F; Lord Hope of Craighead, at p 417 G-H). The minority also noted this point (Lord Browne-Wilkinson at p 359 H; Lord Lloyd of Berwick at p 390 A).

    (4)  The House divided as to whether a mistake of law included a mis-prediction of the result of litigation (in that case, of the outcome of Hazell v Hammersmith and Fulham London Borough Council [1992] 2 AC 1, decided by this House in 1991 after a challenge made in 1989 by a local authority auditor). This issue brought the House into the difficult area of the declaratory nature of judgments. The majority held that mistake of law should be broadly interpreted. I will not try to note all the relevant passages in their complex reasoning, but I note in particular Lord Goff at p 379 G; Lord Hoffmann at p 400 D-E and Lord Hope at p 411 C-D. The minority held that this was unrealistic and amounted to falsification of history (Lord Browne-Wilkinson at p 358 A-B and Lord Lloyd at p 394 A).

    (5)  "Settled law" was therefore for the minority not a defence to a prima facie restitutionary claim, but a reason why there was no claim at all (Lord Browne-Wilkinson at pp 360A-362G and Lord Lloyd at p 394 B-H). The majority discussed "settled law" as a defence, as had recently been proposed by the Law Commission in its Report (No 227, 1994) on "Restitution: Mistakes of Law and Ultra Vires Public Authority Receipts and Payments" ('the 1994 report') para 5. The majority rejected the proposed defence (Lord Goff at pp 381 B-383C; Lord Hoffmann at p 401 B-D; Lord Hope at pp 414A-415B). The passage in Lord Goff's speech which I have just mentioned has led to much discussion and I shall refer to it as the debatable passage.

    (6)  There was a further issue as to "closed swaps" which is not relevant to this appeal.

    (7)  The case had no European element at all.

123.   The decision in Kleinwort Benson has attracted some academic criticism (notably from Professor - as he then was - Jack Beatson, who was the chief architect of the Law Commission report). It has also attracted some approbation (notably from Professor Burrows). It has not been argued that your Lordships should depart from the decision.

    The judgments below

124.   Park J addressed the first issue in a section of his judgment (paras 14-19) headed 'Does English law recognise a claim in restitution to recover taxes paid under a mistake of law?' He quoted at length what I have called the debatable passage in Lord Goff's speech in Kleinwort Benson. He concluded that it did not bear the meaning for which the Revenue contended, for a number of reasons (set out in para 18 of his judgment). In particular he noted that none of the other speeches in Kleinwort Benson supported the view that the payment of unlawfully exacted tax was to be an exception to the new principle permitting recovery of money paid under a mistake of law.
125.   As to the second and third issues, Park J held, after a careful analysis (paras 20-32), that DMG did make all three payments of ACT under a mistake of law, and that the mistake was not discovered until the decision of the ECJ in Hoechst on 8 March 2001. He had already, at the Revenue's request, para 18 (iv), left open the fourth issue (as to a possible settled law defence). The fifth issue was not raised before him. On the sixth issue (as to the pleadings) he decided (para 38) that the amendments did not plead a new cause of action. Instead they merely gave further particulars of a cause of action which had already been pleaded.
126.   In the Court of Appeal (Buxton, Rix and Jonathan Parker LJJ) the cause of action issue was most fully covered in the long and careful judgment of Jonathan Parker LJ (paras 146-160, summarising the arguments, and paras 192-227, the Lord Justice's reasoning and conclusions). Buxton LJ set out his reasons separately (paras 265-291). On this issue Rix LJ agreed with both (paras 250 and 261).
127.   The reasoning on the cause of action issue in the judgments of Jonathan Parker LJ and Buxton LJ is complex and it is not easy to summarise it briefly. But the crucial passage in Jonathan Parker LJ's judgment seems to me to be para 195. He had just quoted from Lord Goff's speech in Woolwich [1993] AC 70, 176 approving the dissenting judgment of Wilson J in Air Canada v British Columbia (1989) 59 DLR (4th) 161, 169. The Lord Justice continued (para 195):
"It seems to me that the above passage, notwithstanding that it is strictly obiter, is of direct significance in the context of the cause of action issue. Lord Goff is plainly referring to cases (of which the Air Canada case was one) in which there is an unlawful demand and a mistake of law by the taxpayer; indeed, if it were otherwise his reference to the mistake of law rule would not be explicable. His conclusion (as I read it) is that in such cases the mistake of law rule has no application since the taxpayer's cause of action is founded not on his mistake but on the unlawful nature of the demand (in effect, on the revenue's mistake): in other words, that the Woolwich cause of action effectively subsumes any cause of action which might otherwise exist for mistake of law."

Jonathan Parker LJ went on to a very detailed discussion of the debatable passage in Lord Goff's speech in Kleinwort Benson, and to a consideration of Park J's reasons for not giving that passage the significance which the Revenue sought to attach to it.

128.   Buxton LJ took a rather different approach to the cause of action issue (paras 265-278). But he too concentrated on analysing the speeches of Lord Goff in Woolwich and Kleinwort Benson. He saw the development of the law in Woolwich (para 270) as "not any deduction from or development of the existing rules of restitution" and (para 271) as "a new remedy, perhaps drawing upon, but certainly not directly applying, ordinary principles of restitution." He pointed out that in Woolwich Lord Goff contemplated ([1993] AC 70, 174C-177C) that strict rules of limitation might be needed for cases of unlawfully exacted tax but that on DMG's interpretation of his speech Lord Goff must have "none the less looked with equanimity in the Kleinwort Benson case upon the prospect of claims in respect of such demands that, in the present case, were not raised until nine or ten years after demand made." On the same theme the Lord Justice referred (para 283) to the fact that "Woolwich is restricted to a claim with a six-year limitation period because, as Lord Goff pointed out in the Kleinwort Benson case, it had made no mistake."
129.   On the second and third issues Jonathan Parker LJ agreed with the judge (paras 229-236). Buxton LJ took a different line (paras 279-284) developing his earlier reasoning and expressing the view (para 279):
"that the considerations that lead to recovery in a case of an ultra vires demand can only with the greatest difficulty be fitted into the rules governing recovery on grounds of mistake."

Rix LJ (para 262) agreed with Buxton LJ's observations.

130.   Only Jonathan Parker LJ dealt with the fourth issue (settled law) and he did so briefly (para 237), agreeing with the view of Lord Hoffmann in Kleinwort Benson that it was a matter for Parliament. The Court of Appeal did not consider the fifth issue. On the sixth issue (pleadings) the majority agreed with the judge, but Buxton LJ dissented (see Jonathan Parker LJ at paras 238-248, Rix LJ at paras 252-260 and Buxton LJ at paras 292-297).

    The cause of action issue

131.   My Lords, it is impossible to read through the judgments (totalling almost 300 paragraphs) in the Court of Appeal without some surprise at the amount of time which the court devoted to what a single Law Lord (however eminent, and without doubt Lord Goff is pre-eminent in this field) said about mistake of law in a case (Woolwich) in which the claimant was never under a mistake (and so the mistake of law rule was not an issue) and about unlawfully exacted taxes in a case (Kleinwort Benson) which had nothing to do with tax. I fear that this may have led to some basic matters of principle being overlooked, or at least being insufficiently considered.
132.   These matters of principle (which have very properly been relied on and developed before your Lordships) include the following: (1) the constitutional principle of equality; (2) the need for coherence in the development of the English law of unjust enrichment; (3) English law's general readiness to permit a claimant to chose between concurrent claims; (4) the consequences of treating the Woolwich principle as exhaustive and exclusive, especially as regards (i) mistake of fact (ii) limitation of actions and (iii) legal certainty of any proposed exception; and (5) the European element (which was not present in either Woolwich or Kleinwort Benson). I must develop these points briefly.
133.   Under the rule of law, the Crown (that is the executive government in its various emanations) is in general subject to the same common law obligations as ordinary citizens. This is recognised by section 21(1) of the Crown Proceedings Act 1947: see the observations of Vaisey J in Sebel Products Limited v Commissioners of Customs and Excise [1949] Ch 409, 413 (though the notion of implied contract applied in that case must now be obsolete). There are exceptions to this principle but they are generally the subject of express enactment. The Revenue's case requires a non-statutory exception whose extent is both far-reaching and uncertain.
134.   Buxton LJ saw Woolwich as a major departure in the common law, rather than as an incremental development of the English law of unjust enrichment. I would respectfully disagree. Woolwich was certainly a very important development but the majority saw it as a development which was within the proper province of the court, and not an exercise in exorbitant judicial legislation. The Law Commission's thorough examination of Woolwich in its 1994 Report (Part VI paras 6.1-6.46) seems to have accepted this, while noting some uncertainties as to how far the principle would extend.
135.   When Parliament enacts a special regime providing special rights and remedies, that regime may (but does not always) supersede and displace common law rights and remedies (or more general statutory rights and remedies). Whether it has that effect is a question of statutory construction: Marcic v Thames Water Utilities Limited [2003] UKHL 66; [2004] 2 AC 42, 56-58, paras 29-36 and Autologic Holdings plc v Inland Revenue Commissioners [2006] 1 AC 118 (which Mr Rabinowitz QC for DMG put forward as a procedural analogue to the present case). Where section 33 of the Taxes Management Act 1970 ("TMA") applies it does no doubt displace any common law remedy for tax paid under a mistake. But in Woolwich tax was demanded under a regulation which was void. There was therefore no valid assessment and the statutory regime was simply not engaged: see Lord Goff in Woolwich [1993] AC 70, 169. Similarly, the ECJ has decided in Hoechst that the ACT regime is unlawful under EU law so far as it discriminates between national and multi-national groups of companies. The Revenue accepts that neither section 33 of TMA (as it stood at the relevant time) nor any other statutory provision applies to the situation in which DMG finds itself. The gap in the statutory provisions cannot provide the Revenue with a defence, both because of Woolwich and, in the context of EU law, because of the principle of effectiveness. The appropriate remedy for DMG is, as the Revenue concedes, restitutionary in nature. It is not for the repayment of a principal sum unlawfully exacted from the taxpayer, but for its unlawful exaction before it became due. The ECJ has made it clear that the domestic court must provide an appropriate remedy, despite the principle reaffirmed by this House in President of India v La Pintada Compania Navigacion SA [1985] AC 104. Your Lordships heard no argument as to whether and how the Pintada principle applies to the law of unjust enrichment; I note that in his leading speech in the case, Lord Brandon covered (at pp 115-116) the position in terms of damages under common law, Admiralty law, equity and statute law, but did not refer to restitution or unjust enrichment.
136.   I would leave open the question whether DMG has to rely on any special principle of EU law in order to obtain a remedy in respect of an exaction which is unlawful only in being premature. In other respects, DMG's claim relies on ordinary domestic principles. English law generally permits a claimant to chose between concurrent causes of action and concurrent remedies as best suits his interests. Lord Goff has been one of the most important influences in developing that approach: see especially Henderson v Merrett Syndicates Limited [1995] 2 AC 145, 193-194:
"My own belief is that, in the present context, the common law is not antipathetic to concurrent liability, and that there is no sound basis for a rule which automatically restricts the claimant to either a tortious or a contractual remedy. The result may be untidy; but, given that the tortious duty is imposed by the general law, and the contractual duty is attributable to the will of the parties, I do not find it objectionable that the claimant may be entitled to take advantage of the remedy which is most advantageous to him, subject only to ascertaining whether the tortious duty is so inconsistent with the applicable contract that, in accordance with ordinary principle, the parties must be taken to have agreed that the tortious remedy is to be limited or excluded."

137.   Lord Goff reaffirmed the same approach, in the context of unjust enrichment, in Kleinwort Benson [1999] 2 AC 349, 387:
"However an equally strong argument may perhaps be made in favour of mistake of law trumping failure of consideration, though either approach is antagonistic to the usual preference of English law to allow either of two alternative remedies to be available, leaving any possible conflict to be resolved by election at a later stage."

The Law Commission took the same view in the 1994 Report (para 10 footnote 1). The claimant bank was allowed to put its case on mistake of law, rather than on payment under a void contract, precisely in order to rely on an extended limitation period under section 32(1)(c) of the Limitation Act 1980. None of the members of the Court of Appeal cited this passage from Lord Goff's speech or even, so far as I can see, paid any regard to it.

138.   I have so far been considering what appear to me to be the main positive arguments in favour of allowing the appeal, and their cumulative effect appears to me to be formidable. That effect is buttressed by considering some of the difficulties involved in the Revenue's case. In the first place, if the Woolwich principle is the only ground for recovery of wrongly paid tax, it would exclude not only an alternative cause of action based on mistake of law, but also one based on mistake of fact. That would involve a departure from a well-established principle from which public authorities are not excepted (see for instance Meadows v Grand Junction Waterworks Company (1905) 21 TLR 538) and I can see no good reason for doing so.
139.   I do not think that the departure can be justified simply by pointing to the possibility that the claimant may, by relying on mistake, obtain a longer limitation period. Parliament has enacted that where there is an action for relief from the consequences of a mistake, time should not run so long as the mistake remains undiscovered. If that is thought to be too generous in some cases, Parliament can change it (and has done so by section 320 of the Finance Act 2004). But the notion that it is right to show some indulgence to a claimant labouring under an undiscovered mistake is an entirely natural one, and it is not to my mind made unnatural simply because the claim is for unlawfully exacted tax. Taxpayers naturally assume that Parliament has got it right. I am puzzled by Buxton LJ's reference (para 283) to the building society being
"restricted to a claim with a six-year limitation period because, as Lord Goff pointed out in the Kleinwort Benson case it had made no mistake"

    with the suggestion that this was somehow unfair. In Woolwich the building society was fortunate in having expert legal advice that the relevant regulation was invalid, and the resources and determination to take on the Revenue promptly. It did not need any extended limitation period. Many other taxpayers would not be in the same position.

140.   If the Woolwich principle is to be an exhaustive and exclusive regime for unlawfully exacted taxes, set apart from the general law of unjust enrichment, legal certainty would require the limits of the exception to be ascertainable with a fair degree of precision. In the debatable passage in his speech in Kleinwort Benson Lord Goff referred twice to "taxes and other similar charges," which is imprecise. Mr Glick suggested that the similarity was to be found in the existence of a special statutory regime regulating recovery. This suggested test was not fully explored in argument but it seems unlikely to be a satisfactory means of setting a clear dividing line within a spectrum which stretches from central government taxes and duties through rates, community charge, drainage rates and charges, special levies and licence fees imposed by statute on different industrial and commercial activities, and charges made by statutory undertakers (as to the last category see South of Scotland Electricity Board v British Oxygen Co Limited (No 2) [1959] 1 WLR 587). This point was touched on, inconclusively, by the Privy Council in Waikato Regional Airport v Attorney-General of New Zealand [2003] UKPC 50 at para 80 (discriminatory charges levied on New Zealand airports in respect of official biosecurity services).
141.   My Lords, all these considerations lead me to the conclusion that the judge was right, and the Court of Appeal was wrong, in relation to the cause of action issue. That is not to say that I see no difficulty in reconciling this conclusion with the debatable passage in Lord Goff's speech in Kleinwort Benson. Undoubtedly that passage represented part of Lord Goff's mature reflections on the development of the English law of unjust enrichment. It is not easy to reconcile the passage with the rest of Lord Goff's reasoning, but I think that the explanation is provided in the opinions of my noble and learned friends Lord Hoffmann and Lord Hope of Craighead, both of whom sat with Lord Goff in Kleinwort Benson (and whose opinions I have had the great advantage of reading in draft). But whatever the difficulties of the passage, there are to my mind two clear beacons in Lord Goff's speech, one before and one after the debatable passage, which are clearer and more reliable guides. The first is his affirmation of a general right to recover money paid under a mistake of law ([1999] 2 AC 349, 375):
"I would therefore conclude on issue (1) that the mistake of law rule should no longer be maintained as part of English law, and that English law should now recognise that there is a general right to recover money paid under a mistake, whether of fact or law, subject to the defences available in the law of restitution."

The other passage, at p 387, is that referring to "the usual preference of English law to allow either of two alternative remedies" which I have already cited.

142.   I reach this conclusion without any heavy reliance on principles of EU law, except perhaps in relation to the Pintada principle (which I regard as uncertain in relation to the law of unjust enrichment). But the European context confirms my conclusion. The domestic court must give DMG an equivalent and effective remedy, and that would not be achieved, in my opinion, if recovery were limited so as to exclude an alternative concurrent remedy which would be available in a dispute between private citizens.

    The mistake issue

143.   I can set out my views on this issue more briefly. I agree with the judge's conclusions, and I largely agree with his reasoning, though I respectfully think that he was rather over-analytical in his approach. I agree with the observation of Neuberger J in Nurdin & Peacock plc v D D B Ramsden & Co Ltd [1999]1 WLR 1249, 1272:
"For the issue of recoverability to turn upon a nice analysis as to the precise nature of the mistake of law appears to me to be almost as undesirable as it is for recoverability to turn upon whether the mistake made by the payer was one of fact or law."

The straightforward test of causation put forward by Robert Goff J (as he then was), after a full survey of authority, in Barclays Bank Limited v W J Simms Son & Cook (Southern) Limited [1980] QB 677, has stood the test of time. DMG paid the ACT because it mistakenly thought that it had to. The fact that there was a procedural requirement for a GIE does not alter the substance of its mistake, since (as Park J expressly found, para 11) any attempt at making a GIE would undoubtedly have been rejected in this case.

144.   I think the judge and Jonathan Parker LJ were correct in their views that the mistake was not discovered until the ECJ gave judgment in Hoechst [2001] Ch 620. Perusal of the report in that case suggests that the United Kingdom government tenaciously defended the ACT regime on every available ground. At no time before the judgment did the government concede that the ACT regime was (in discriminating between national and multi-national groups) contrary to EU law and unlawful. It was the judgment that first turned recognition of the possibility of a mistake into knowledge that there had indeed been a mistake. I agree with the view of Lightman J in First Roodhill Leasing Limited v Gillingham Operating Company Limited (5 July 2001) para 22 that there may be cases
"where a party may be held to have discovered a mistake without there being an authoritative pronouncement directly on point on the facts of that case by a court, let alone an appellate court."

It all depends on the facts. But in this case it is, in my opinion, clear that the judgment of the ECJ on 8 March 2001 was the decisive moment.

    Settled law

145.   Like the judge, I agree with the view expressed in Kleinwort Benson by my noble and learned friend Lord Hoffmann, that the fashioning of a settled law defence would be a legislative act and is therefore a matter for Parliament. Parliament has indeed incorporated its own version of the defence, in terms of "the practice generally prevailing", in section 33(2A)(a) of TMA 1970, re-enacting the proviso to section 33(2). As Lord Goff recognised in Woolwich and Kleinwort Benson, there are strong policy arguments in favour of a general common law right of recovery of wrongly exacted tax, and also strong countervailing arguments in favour of some restrictions on that right of recovery, especially as the defence of change of position could seldom, if ever, apply to wrongfully exacted tax. But the balancing of these high policy arguments is essentially a matter for the legislature.

    The "scope of section 32(1)(c)" issue

146.   This issue arises only if DMG fails on the first (cause of action) issue. In my opinion it does not arise on this appeal. The rule that in order to come within section 32(1) a mistake must be an essential ingredient of the claimant's cause of action rests on a surprisingly uncertain basis, that is a view expressed by Pearson J in Phillips-Higgins v Harper [1954] 1QB 411, 419. Nevertheless it has been generally accepted (with some dissentient academic voices raised against it) for over fifty years.
147.   The Law Commission has now completed and published its review of the law of limitation of actions (2001, Law Com. 270) and the government has accepted its general recommendations with a view to legislation as soon as time permits. In those circumstances your Lordships need not, in my opinion, reconsider the now nearly traditional view of the scope of section 32(1)(c), although there are persuasive arguments for its reinterpretation (see Dr James Edelman, "Limitation Periods and the Theory of Unjust Enrichment" (2005) 68 MLR 848 and Professor Andrew Burrows, "Restitution in respect of Mistakenly Paid Tax" (2005) 121 LQR 540).

    The pleading issue

148.   This point was fully considered in the courts below. I agree with the reasoning and conclusions of Park J and the majority of the Court of Appeal. I would only echo the observations of Millett LJ in Paragon Finance plc v DB Thakerar & Co (a firm) [1999] 1 All ER 400, 405:
"The pleading of unnecessary allegations or the addition of further instances or better particulars do not amount to a distinct cause of action. The selection of the material facts to define the cause of action must be made at the lightest level of abstraction".

The formula used in DMG's particulars of claim ("Such amounts include, but are not limited to, payments of ACT…specified in the [First or Second] Schedule") was imprecise and unsatisfactory, but the defect could have been cured by a request for particulars.

149.   For these reasons I would allow the appeal and restore the order of Park J.

The foundations of unjust enrichment

150.   I do however wish to add some brief observations on the views of the late Professor Peter Birks, who died on 6 July 2004, after the first-instance decision of Park J but before the decision of the Court of Appeal.
151.   Professor Birks was (in the words of Professor Burrows in his introduction to the second edition of Birks' Unjust Enrichment, published posthumously in 2005) one of the most influential academic lawyers of our time. After Lord Goff and Professor Gareth Jones (the first edition of whose joint work on The Law of Restitution was published in 1966) Professor Birks was one of the foremost, and many would say the very foremost, of the scholars working on developing the law of restitution. In the revised paperback edition of his Introduction to The Law of Restitution (1989) he explained his views as to the essential function in the law of unjust enrichment of "unjust factors" (such as mistake, undue influence, and compulsion). Most if not all other leading textbooks adopt the same or similar terminology, for instance Burrows, The Law of Restitution, second edition (2002) pp 41 ff; Goff and Jones, The Law of Restitution, sixth edition (2002) para 1-053; Hedley and Halliwell (gen. eds.) The Law of Restitution (2002) para 1.20; Tettenborn, Law of Restitution in England and Ireland, third edition (2002) pp 13 ff; Virgo, The Principles of the Law of Restitution (1999) pp 119 ff.
152.   But in the preface to the first (2003) edition of Unjust Enrichment Birks, with characteristic intellectual courage, explained that he had changed his views in some fundamental respects (he wrote to Sir Jack Beatson that he was "back to square one"). In particular, he switched from the concept of "unjust factors" to the civilian concept of "lack of basis" (or "lack of cause"). As he said in his preface to the first edition, summarising the points on which his views had changed:
"Thirdly, and perhaps most importantly, the onward march of case law and academic analysis, especially comparative analysis, has both compelled and convinced me that it was a misjudgement to insist that the common law, late coming, had by good luck hit upon a better way of answering the question whether an enrichment was unjust than, with their much longer experience, the civilian jurisdictions had achieved."

153.   Birks ascribed his change of view partly to the swaps litigation, and especially to Kleinwort Benson. In his own contribution, the first chapter, to Birks and Rose (Eds), Lessons of the Swaps Litigation, (2000) he wrote of the decision of this House in Kleinwort Benson (p14),
"The disagreement between the majority and the minority turns on whether one can be retrospectively mistaken in any sense relevant to restitutionary relief. The majority holds that one can, and in so holding it appears to have approved a notion of operative mistake of law which is not only broader than operative mistake of fact but broader than the rationale underlying relief for mistake. Without saying so expressly, it has moved English law towards a civilian condictio indebiti. 'Condictio indebiti' means 'Claim in respect of something not due'."

    This can be contrasted with the well-known statement by Lord Goff in Woolwich [1993] AC 70, 172,

"The first [objection] is to be found in the structure of our law of restitution, as it developed during the 19th and early 20th centuries. That law might have developed so as to recognise a condictio indebiti - an action for the recovery of money on the ground that it was not due. But it did not do so. Instead, as we have seen, there developed common law actions for the recovery of money paid under a mistake of fact, and under certain forms of compulsion."

    The law of Scotland has, by contrast, closely followed the civilian model: see the Lord President (Rodger) in Shilliday v Smith [1998] SC 725, 727-728 and generally Professor Robin Evans-Jones, Unjustified Enrichment: Vol.One, Enrichment by Deliberate Conferral: Condictio (2003).

154.   In explaining his changes of view Professor Birks also acknowledged the influence of Dr Sonja Meier, a lecturer in law at the University of Regensberg. Her views are accessible in English in her contributions to Lessons of the Swaps Litigation (chapter 6, Restitution After Executed Void Contracts) and to Johnston and Zimmermann, Unjustified Enrichment: Key Issues in Comparison (2002) (chapter 2, Unjust Factors and Legal Grounds). On this point English law may be at something of a crossroads (a metaphor first used, I think, by Waller LJ in Guinness Mahon & Co Ltd v Kensington and Chelsea London Borough Council [1999] QB 215, 233, and since picked up by others including Dr Krebs, Restitution at the Crossroads, A Comparative Study (2001)). The choice as to the way forward which restitution scholars identify is between continuing to view unjust enrichment as depending on the presence of one or more of a variety of "unjust factors" and adopting the single test of "absence of basis".
155.   My Lords, the House is being invited (much more pressingly, it must be said, by scholars than by counsel for the parties) to make a choice at a very high level of abstraction. Most scholars would take the view (though Professor Birks himself would not, I suspect, have agreed, since he regarded taxonomy as very important) that the choice is one which will rarely make much if any practical difference to the outcome of any particular case before the court. For several reasons I doubt whether this is the right time for your Lordships to decide whether to rebase the whole law of unjust enrichment on a highly abstract principle which (although familiar to civilians and to Scottish lawyers, and discussed in the speech of my noble and learned friend Lord Hope of Craighead in Kleinwort Benson [1999] 2 AC 349, pp 408-409) would represent a distinct departure from established doctrine.
156.   It is of the nature of the common law to develop slowly, and attempts at dramatic simplification may turn out to have been premature and indeed mistaken. As Lord Rodger of Earlsferry put it in Commissioners of Customs and Excise v Barclays Bank plc [2006] UKHL 28; [2006] 3 WLR 1, 19, para 51:
"Part of the function of appeal courts is to try to assist judges and practitioners by boiling down a mass of case law and distilling some shorter statement of the applicable law. The temptation to try to identify some compact underlying rule which can then be applied to solve all future cases is obvious. [Counsel for the appellants] submitted that in this area the House had identified such a rule in the need to find that the defendant had voluntarily assumed responsibility. But the unhappy experience with the rule so elegantly formulated by Lord Wilberforce in Anns v Merton London Borough Council [1978] AC 728, 751-752, suggests that appellate judges should follow the philosopher's advice to 'Seek simplicity, and distrust it.'"

    Other members of the House showed a similar disinclination to wide generalisation: see Lord Bingham of Cornhill at para 8, Lord Hoffmann at para 36 and Lord Mance at para 83. Commissioners of Customs & Excise v Barclays Bank plc shows that more than forty years on from Hedley Byrne & Co Ltd v Heller & Partners Ltd [1964] AC 465, the true foundations of the law of tortious liability for negligent mis-statement are still open to debate.

157.   By contrast the English law of unjust enrichment has in the space of a decade seen four very important developments, all informed by the learning of Lord Goff: Lipkin Gorman v Karpnale Ltd [1991] 2 AC 548 in 1991, Woolwich [1993] AC 70 in 1992, Westdeutsche Landesbank Girozentrale v Islington London Borough Council [1996] AC 669 in 1996, and Kleinwort Benson [1999] 2AC 349 in 1998. The change in the views of Professor Birks is a recent development (which sadly he could not pursue further) and it has not yet been fully considered by other legal scholars. There is, it seems to me, much to be said for a period of reappraisal.
158.   Nevertheless I would add that my tentative inclination is to welcome any tendency of the English law of unjust enrichment to align itself more closely with Scottish law, and so to civilian roots. I see attractions in the suggestion made by Professor Birks in Unjust Enrichment (2nd edition, p.116, under the heading 'The Pyramid: a Limited Reconciliation'):
"A pyramid can be constructed in which, at the base, the particular unjust factors such as mistake, pressure and undue influence become reasons why, higher up, there is no basis for the defendant's acquisition, which is then the master reason why, higher up still, the enrichment is unjust and must be surrendered."

I would be glad to see the law developing on those lines. The recognition of "no basis" as a single unifying principle would preserve what Lord Hope refers to as the purity of the principle on which unjust enrichment is founded, without in any way removing (as this case illustrates) the need for careful analysis of the content of particular "unjust factors" such as mistake.

LORD BROWN OF EATON-UNDER-HEYWOOD

My Lords,

159.   DMG's claim in these proceedings (issued on 18 October 2000) is for compensation in respect of three payments of ACT made respectively on 14 October 1993, 15 February 1995 and 14 January 1996. These payments could and would have been avoided had the UK tax regime not breached article 43 (formerly 52) of the EU Treaty by denying DMG (as a multinational rather than exclusively UK group) the right to make a group income election. So much, the majority of your Lordships would hold, was established by the ECJ's decision in Hoechst on 8 March 2001. The claim is not for the capital sums paid: those were subsequently set off against the company's liability for MCT. Rather it is for compensation for the loss of use of the monies prior to such setoffs, i.e. for having made accelerated payments.
160.   The Revenue do not dispute DMG's entitlement to such compensation in principle. On the contrary, they readily accept that the company has a perfectly good remedy, either in tort for breach of statutory duty or in restitution by analogy with the principle established in Woolwich—see Woolwich Equitable Buildings Society v Inland Revenue Commissioners [1993] AC 70. But under these causes of action DMG would be confined to a six year limitation period (i.e. six years from the date of the respective payments) and on this basis would fail in respect of the October 1993 payment (and fail too in respect of the later payments were the Revenue to succeed in their cross appeal on the pleading issue, contending as they do that such claims were only brought by later amendment rather than when the writ was first issued). That is why DMG assert their claim as one for restitution based on a mistake of law—precisely so as to benefit from section 32 (1)(c) of the Limitation Act 1980, just as the claimants succeeded in doing in the landmark case of Kleinwort Benson Ltd v Lincoln City Council [1999] 2 AC 349.
161.   All this is fully and lucidly explained in the speech of my noble and learned friend, Lord Walker of Gestingthorpe, as too are the various issues which arise for decision on the appeal. With almost all of Lord Walker's speech (and the speeches of my noble and learned friends, Lord Hoffmann and Lord Hope of Craighead) I find myself in full agreement and, in common with them, I too would allow DMG's appeal and restore Park J's order. On one issue, however, the "discovery" issue (linked in one respect, as it is, with the "mistake" issue), I have the misfortune to take a different view, indeed a view which puts me in a minority of one. As it happens, because of the inter-relationship of this issue with other issues arising on the appeal, my disagreement has in fact no effect whatsoever on the overall outcome of the proceedings. But it would not, I think, be right to ignore it entirely: the point could well be of decisive importance in other actions, perhaps indeed actions awaiting the outcome of this very appeal. Let me explain.
162.   As I understand it, your Lordships have concluded, first, that DMG was acting under a mistake of law when making the three relevant payments of ACT (it is on this issue that my noble and learned friend, Lord Scott of Foscote dissents on the basis that DMG was in any event under a legal obligation to make the payments and so should not be regarded within the Kleinwort Benson principle as having made them under a mistake of law); and, secondly, that that mistake was not discovered until the decision of the ECJ in the Hoechst case was handed down on 8 March 2001. The contrary view which I take is that DMG ceased to be acting under any relevant mistake of law in July 1995 when they first became aware of the Hoechst proceedings and recognised that there was a serious legal challenge to the legality of the UK's ACT regime under EC law. If I am correct in this view it would follow that DMG, although paying the October 1993 and February 1995 payments under a mistake of law, discovered that mistake in July 1995 so as to set time running in respect of those particular payments as from that date; and that the January 1996 payment was not made under a mistake of law at all. (It is because, however, as already stated, the claim was issued on 18 October 2000, within six years of July 1995, and because DMG have the benefit of undisputed alternative grounds of claim in respect of the January 1996 payment, that none of this matters given your Lordships' rejection, with which I agree, of the Revenue's cross appeal.)
163.   Let me turn then to this narrow though potentially important issue upon which I diffidently disagree with your Lordships. First I should for convenience set out again section 32(1) of the 1980 Act:
" … where in the case of any action for which a period of limitation is prescribed by this Act, either—

(a)  the action is based upon the fraud of the defendant; or

(b)  any fact relevant to the plaintiff's right of action has been deliberately concealed from him by the defendant; or

(c)  the action is for relief from the consequences of a mistake;

the period of limitation shall not begin to run until the plaintiff has discovered the fraud, concealment or mistake (as the case may be) or could with reasonable diligence have discovered it."

164.   It is, I acknowledge, difficult to articulate the precise touchstone by which to determine whether a payment ought properly to be held to have been made under a mistake of law (or, indeed, answering essentially the same question from a different standpoint, whether such a mistake should be held to have been discovered—or able with reasonable diligence to have been discovered). Section 32 itself affords little help on the question—unsurprisingly, perhaps, given that the provision (based as it was on section 26 of the Limitation Act 1939) was framed at a time when no plaintiff could have hoped to pray in aid a mistake of law.
165.   For my part, however, I would hold that as soon as a paying party recognises that a worthwhile claim arises that he should not after all have made the payment and accordingly is entitled to recover it (or, as here, to compensation for the loss of its use), he has "discovered" the mistake within the meaning of section 32; and, by the same token, I would hold that if he makes any further payments thereafter, they are not to be regarded as payments made under a mistake of law.
166.   Where, I would respectfully ask, is there any injustice in this? No one is suggesting, let me repeat, that the monies are not recoverable or that the payee should remain unjustly enriched. All that is required is that the payer does not sit upon what ex-hypothesi he recognises to be a worthwhile legal argument for more than six years. Provided he acts within that (surely ample) time, he can pursue his claim (whether in respect of past payments or, indeed, payments he may choose to continue making) under whatever may be the appropriate cause of action: restitution for mistake of law in respect of past payments made when he had no reason to question his liability to make them, total failure of consideration, or a claim based on the Woolwich principle.
167.   Once a plaintiff recognises that he has a worthwhile case on the facts to pursue a claim in fraud or to extend the limitation period for a particular claim because of the defendant's deliberate concealment of a fact relevant to his cause of action, time surely then starts to run against him under section 32: he could not successfully argue that time starts running only when the court eventually comes to reject the defendant's denial of wrongdoing and to find fraud (or, as the case may be, deliberate concealment) established.
168.   Nor do I find it easy to reconcile the approach taken by your Lordships with other provisions of the 1980 Act, most notably section 14A(9): "Knowledge that any acts or omissions did or did not, as a matter of law, involve negligence is irrelevant for the purposes of subsection (5) above." (Section 14A(5) determines the starting date for reckoning the period of limitation based on the claimant's date of knowledge.) As Lord Nicholls of Birkenhead recently observed in Haward v Fawcetts [2006] 1 WLR 682, 686F, para 12: "A claimant need not know he has a worthwhile cause of action." On your Lordships' view, however, such a claimant could well argue that he was labouring under a mistake of law so as to extend time under section 32(1).
169.   Woolwich was decided on the explicit basis that the building society was in no way mistaken when making its payments under the disputed regulations—see particularly the speech of Lord Goff at p 173D and that of Lord Slynn at p 201A-B. Rather the payments were made because, as Lord Goff put it at p 171G-H:
"[The taxpayer] is faced with the revenue, armed with the coercive power of the state, including what is in practice a power to charge interest which is penal in its effect. In addition, being a reputable society which alone among building societies is challenging the lawfulness of the demand, it understandably fears damage to its reputation if it does not pay. So it decides to pay first, asserting that it will challenge the lawfulness of the demand in litigation."

    I can see no real distinction between that case and this (this case, that is, after July 1995). True, as Lord Goff observed (at p 171G), Woolwich was 'convinced that the demand [was] unlawful' whereas here Mr Thomason, DMG's Head of Taxation, believed that the company was in law bound to make the payments (precisely, indeed, as Lord Scott would hold to be so). But I fail to see why the question whether monies are paid under a mistake of law should turn on the degree of conviction or optimism which the parties hold upon the legal issue dividing them. Were the claimants in Hoechst (who issued their proceedings against the Revenue in 1995) nonetheless to be regarded as having made all subsequent payments under a mistake of law? Surely not. Even DMG itself, it will have been noted, brought its claim in October 2000. Is it nevertheless to be said that their original mistake remained undiscovered until the ECJ's actual decision in Hoechst some five months later?

170.   In Kleinwort Benson, as I understand that case, nothing turned upon the particular state of mind of the payer as to whether the payments for which restitution was sought were made under a mistake. As Lord Hope noted (at p 403G), the only issue on that part of the case was whether the bank's mistake was one of law. It was certainly not critical to the decision there that the mistake of law was assumed, at least by Lord Lloyd, to have been discovered only when the House of Lords finally held in Hazell v Hammersmith & Fulham LBC [1992] 2 AC 1 that the swap agreements were void. As Lord Hoffmann observed (at p 401F-H), the decision (in Kleinwort Benson) "leaves open what may be difficult evidential questions over whether a person making a payment has made a mistake or not. … There is room for a spectrum of states of mind between genuine belief in validity, founding a claim based on mistake, and a clear acceptance of the risk that they are not. But these questions are not presently before your Lordships."
171.   Lord Hope too left open for another day cases where payments are made in a state of doubt about the law. The Revenue on the present appeal understandably place some reliance on what Lord Hope said at p 410B-C:
"Cases where the payer was aware that there was an issue of law which was relevant but, being in doubt as to what the law was, paid without waiting to resolve that doubt may be left on one side. A state of doubt is different from that of mistake. A person who pays when in doubt takes the risk that he may be wrong—and that is so whether the issue is one of fact or one of law."

172.   On the present appeal, however, Lord Hope concludes his judgment on "the discovery issue" (paragraphs 63-71 of his speech) with the view that, when DMG paid the ACT, "[i]t was not then obvious that the payments might not be due." I confess to some difficulty with that conclusion. Surely, when DMG learned in July 1995 that there was a serious legal challenge to the legality of the ACT regime, it must then have been obvious to them that these payments might not after all be due. Of course they could not be sure and of course nothing short of a final judgment from the ECJ would have persuaded the Revenue to accept any claim by DMG here for group income relief. But it does not seem to me to follow that DMG paid under a mistake of law—any more than Woolwich would be regarded as having paid under such a mistake simply because the Revenue in that case were insisting on the validity of the contested regulations.
173.   I have the same difficulty with paragraph 144 of Lord Walker's opinion. Again, I see no good reason why the Revenue's tenacious defence of their position and their refusal to concede its unlawfulness means that DMG's mistake must be treated as undiscovered prior to the Hoechst judgment. The passage quoted by Lord Walker from Lightman J's judgment in First Roodhill Leasing Ltd v Gillingham Operating Company Ltd (5 July 2001) para 22, continues:
"For this purpose it cannot be necessary that the party knows of the mistake as a certainty. There are gradations of knowledge. It may well be sufficient to constitute the necessary discovery when the claimant has good reason to believe that a mistake has been made (consider Earl Beatty v IRC [1953] 1 WLR 1090) or has been given 'a line' on this question (see G L Baker v Medway [1958] 1 WLR 1216 at 1224)."

174.   To much the same effect is Maurice Kay LJ's judgment in Brennan v Bolt Burdon [2004] EWCA Civ 1017; [2005] QB 303, 315:
"This [the plaintiff's extreme difficulty in obtaining permission to appeal and 'small chance' of persuading the Court of Appeal], it seems to me, falls short of the unequivocal but mistaken view of the law which underlay the Kleinwort Benson case [1999] 2 AC 349. As Lord Hope observed, at p 410B, the House of Lords was not dealing with the case where there is doubt as to the law—'a state of doubt is different from that of mistake. An appeal might have been correctly perceived as an uphill struggle but not as an inherently insuperable one—as subsequent events were to prove."

175.   Lord Hoffmann suggests (at paragraph 26) that: "The real point is whether the person who made the payment took the risk that he might be wrong. If he did, then he cannot recover the money." But my thesis is not that, if someone pays money knowing that he may not be under any liability to do so, he cannot recover it. Rather it is that he cannot recover it as money paid under a mistake of law so as to benefit from the longer limitation period available under section 32. Certainly he can recover the money provided only that he sues in time and has some other cause of action, such as total failure of consideration. Clearly the quiz contestant who, in doubt whether Haydn or Mozart wrote the eine kleine nachtmusik answers Haydn, made a mistake. Suppose, however, that, making that mistake, he had paid out money legally due only if Haydn had been the correct answer. To my mind he, no less than the quiz contestant, took the risk that he might be wrong: he could not recover his payment as money paid under a mistake of law (or fact) although, provided he sued within six years, he could well recover it on another basis.
176.   The precise point at which a party may be said to be, or to cease being, under a mistake of law is, I acknowledge, by no means easy to formulate. Just when a party comes to recognise he has "a worthwhile claim" (the touchstone I have suggested in paragraph 163 above) will not always be obvious. Essentially, however, I am in broad agreement with Lightman J's and Maurice Kay LJ's approach in the cases mentioned above, as indeed I am with the views of the majority of the Court of Appeal on this issue in the present case—see Rix LJ's judgment at para 262 and Buxton LJ's judgment at paras 281-283.
177.   For the reason already given, however, even if that view were shared by a majority of this Committee, it would avail the Revenue nothing. The appeal must in any event be allowed.