In the same case at p. 393 Brennan J. said:

  "When a defendant receives a payment which he has no right to receive and which the plaintiff has paid to him by mistake, the injustice of the defendant's enrichment does not depend on the nature of the mistake that caused the payment to be made. Whether the plaintiff made a mistake of law or a mistake of fact, the defendant, having no right to receive the payment, is unjustly enriched by its receipt. Then should the distinction between the two categories of mistake make any difference to a finding of unjust enrichment?"

In answer to his own question he said that he agreed with the majority in rejecting the distinction as critical to the question whether the defendant has been unjustly enriched. He went on to say that he did not see the distinction between mistakes of fact and mistakes of law as being immaterial to the question whether the payment was recoverable. This is a qualification to which I must return later. For the present it is sufficient to note his agreement with the majority on the absence of any good ground for the distinction. Similar views were expressed in Willis Faber and in Morgan Guaranty after a review of the civilian authorities. The unanimity which has been expressed on this point, in which common law jurisdictions are at one with those whose roots lie in the civil law, gives powerful recognition to the unifying effect of the principle of unjust enrichment. I do not think that the law of England, having accepted the principle, can continue to resist its consequences.

    The third consideration relates to the practical effects of the distinction. I do not need to elaborate on this point. As Mr. Southwell Q.C. explained, the rule is subject to numerous exceptions and qualifications. I wish to mention only a few examples. Relief may be given where the error of law is one as to the construction of a private contract between the two parties which affects them only: Cooper v. Phibbs (1867) L.R. 2 H.L. 149; Earl Beauchamp v. Winn (1873) L.R. 6 H.L. 223; see also British Hydro- Carbon Chemicals Ltd. and British Transport Commission, Petitioners 1961 S.L.T. 280, per Lord Kilbrandon. And there is the exception which was described by Lord Denning in Kiriri Cotton Co. Ltd. v. Dewani [1960] A.C. 192, 204 which applies where there is "something more" in addition to the mistake of law such as something in the defendant's conduct which shows that he was the one who was primarily responsible for the mistake. Experience has shown that in practice the rule has tended to lead to unjust results and to a desire to avoid the consequences. It is unsatisfactory that the law should have had to resort to exceptions and qualifications, the subsequent application of which to other cases can give rise to difficulty.

Was There a Mistake?

    Subject to any defences that may arise from the circumstances, a claim for restitution of money paid under a mistake raises three questions: (1) was there a mistake? (2) did the mistake cause the payment? and (3) did the payee have a right to receive the sum which was paid to him?

    The first question arises because the mistake provides the cause of action for recovery of the money had and received by the payee. Unless the payer can prove that he acted under a mistake, he cannot maintain an action for money had and received on this ground. The second question arises because it will not be enough for the payer to prove that he made a mistake. He must prove that he would not have made the payment had he known of his mistake at the time when it was made. If the payer would have made the payment even if he had known of his mistake, the sum paid is not recoverable on the ground of that mistake. The third question arises because the payee cannot be said to have been unjustly enriched if he was entitled to receive the sum paid to him. The payer may have been mistaken as to the grounds on which the sum was due to the payee, but his mistake will not provide a ground for its recovery if the payee can show that he was entitled to it on some other ground.

    In the present case the second and third questions do not appear to present any difficulty. But the first question raises an issue of very real importance. The answer which is given to it will have significant implications for the future development of the law of restitution on the ground of unjust enrichment.

    In my opinion the proper starting point for an examination of this issue is the principle on which the claim for restitution of these payments is founded, which is that of unjust enrichment. The essence of this principle is that it is unjust for a person to retain a benefit which he has received at the expense of another, without any legal ground to justify its retention, which that other person did not intend him to receive. This has been the basis for the law of unjust enrichment as it has developed both in the civilian aken advice from a qualified legal practitioner who was y civilian and partly common law. On the whole, now that the common law systems see their law of restitution as being based upon this principle, one would expect them to apply it, broadly speaking, in the same way and to reach results which, broadly speaking, were similar: Zweigert and Kotz, An Introduction to Comparative Law, 2nd ed. (1987), vol.II, pp. 262-263, 267.

    What, then, is the function of mistake in the field of restitution on the ground of unjust enrichment? The answer, one may say, is that its function is to show that the benefit which has been received was an unintended benefit. A declaration of intention to confer the benefit, even if unenforceable, will be enough to justify the retention of the enrichment. A mistake, on the other hand, will be enough to justify the restitutionary remedy, on the ground that a benefit which cannot be legally justified should not be retained where it was a mistaken--and thus unintended--benefit.

    It may be helpful to mention the material we were given to illustrate its function in the civilian systems. The details vary as between the major civil codes. But in simple terms, the law looks for the absence of a legal justification for the enrichment: Zweigert and Kotz, p. 232. If the payer paid in the mistaken belief that he was under a duty to pay, it is prima facie unjust that the payee should be allowed to retain what he received. But the burden of proving that the payer knew that there was no duty, and was not mistaken, is on the recipient: Englard, International Encyclopedia of Comparative Law, (1991) vol. X, pp. 8-9, para. 5.13. Mistake in this context means lack of knowledge, and it makes no difference whether this is of fact or of law: Englard, p. 18, para. 5.30. As for the concept of enrichment, a person is enriched when he receives a payment which the payer was not bound by any obligation to make to him. The payee is entitled to retain the payment if it was made to him voluntarily, as in the case of a gift. The enrichment is unjust if the person who made the payment did not do so voluntarily and there was no obligation to confer the benefit: Zweigert and Kotz, p. 261.

    The approach of the common law is to look for an unjust factor, something which makes it unjust to allow the payee to retain the benefit: Birks, An Introduction to the Law of Restitution, 2nd ed. (1989), pp. 140 et seq. It is the mistake by the payer which, as in the case of failure of consideration and compulsion, renders the enrichment of the payee unjust. The common law accepts that the payee is enriched where the sum was not due to be paid to him, but it requires the payer to show that this was unjust. Whereas in civilian systems proof of knowledge that there was no legal obligation to pay is a defence which may be invoked by the payee, under the common law it is for the payer to show that he paid under a mistake. My impression is that the common law tends to place more emphasis on the need for proof of a mistake. But the underlying principle in both systems is that of unjust enrichment. The purpose of the principle is to provide a remedy for recovery of the enrichment where no legal ground exists to justify its retention. But does it matter whether the mistake is one of fact or one of law?

    To answer this question one must have in mind both the state of mind of the payer and the state of the facts or the law about which there is said to have been a mistake. The state of mind of the payer must be related to the time when the payment was made. So also must the state of the facts or the law. That is the time as at which it must be determined whether the payment was or was not legally justified. I agree with Brennan J.'s observation in David Securities that the right to recover the amount paid by mistake accrues at the moment when the sum is received by the payee: see also Baker v. Courage & Co. [1910] 1 K.B. 56. The point of the inquiry is to show that, had the payer known the true state of the facts or the law at that time, he would not have made the payment to the payee.

    The inquiry will not be a difficult one, where the mistake is said to have been one of fact, if the facts have not changed since the date of the payment and the payer is able to show that he paid due to a misunderstanding of them, to incorrect information or to ignorance. In such a case the requirements for recovery will normally be satisfied. Nor is it difficult to deal with the case where the facts have changed. In such a case proof that the alleged state of the facts at the time did not emerge until afterwards will usually be sufficient to show that there was, at the time of payment, no mistake. The case may be more difficult where the mistake is said to have been a mistake of law. But I do not think that there is any essential difference in principle. A question of law may be as capable of being answered as precisely and with as much certainty as a question of fact, or it may be- -as are some questions of fact--a matter of opinion.

     Nor is there any essential difference as between fact and law in regard to the payer's state of mind. This may vary from one of complete ignorance to a state of ample knowledge but a misapplication of what is known to the facts. The mistake may have been caused by a failure to take advice, by omitting to examine the available information or by misunderstanding the information which has been obtained. Or it may have been due to a failure to predict correctly how the court would determine issues which were unresolved at the time of the payment, or even to foresee that there was an issue which would have to be resolved by the court. As Mason C.J. said in David Securities, p. 374 the concept of mistake includes cases of sheer ignorance as well as of positive but incorrect belief.

    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. As for mistake, this may arise where there is no suggestion that the law has changed since the payment was made. If it can be demonstrated by reference to statute or to case law that the law was overlooked or was applied wrongly, the position will be the same as that where the mistake was one as to the state of the facts. It is very unusual for a statute to provide for the law to be changed retrospectively, but this is not unknown: see the War Damage Act 1965. If the law is changed retrospectively by statute, so that a payment which was legally due when it was paid has now become undue, the correct analysis will be that there was no mistake at the time when it was made. The enrichment will have been due to the fact that the law was changed retrospectively by the statute.

    What then is the position where the fact that the payment was not legally due at the time when it was made was only revealed later by subsequent case law? In posing this question I am not dealing with the situation where a judgment of the court that a sum is due has become final and been acted upon, but is afterwards overruled by a higher court in a different case. The law of unjust enrichment does not disturb transactions of that kind. Where the payment is made because the court has held that the sum is due to be paid to the payee, the obligation to pay is to be found in the order which has been made by the court. I am dealing with the case where the payment was made on the understanding that the law on the point was settled and that understanding was shown by subsequent case law to have been wrong.

    The answer to this question may be said to depend upon whether the decision in question has changed the law or has merely declared what the law always was. óKóóK We were reminded of Lord Reid's observation that to say that the judges never change the law is a fairy tale: (1972-73) J.S.P.T.L. 22. Experience has shown that the judges do from time to time change the law, in order to adapt it to changed social conditions or in response to other factors which show that the law has become out of date. But it would be equally wrong to say that the judges never declare the law. It may simply be that there was a gap which needed to be filled, or that there was a defect in thinking which needed to be revealed so that a point could be clarified. And to overturn an established line of authority is one thing. It is quite another where there was no previous decision on a point which no-one had sought to bring before the court previously. It may be said that a view of the law can be regarded as settled even where there is no case law at all on the subject, because all those interested in it have acted on a common understanding of what the law requires. But I would find it difficult to accept that a judge who said that that common understanding was wrong, and that the law was different from what everyone previously had thought it was, had changed the law. It would seem to be more accurate to say that, as it was for the judge to say what the law was, he was merely declaring what the law was and that he was not changing it.

    On the whole it seems to me to be preferable to avoid being drawn into a discussion as to whether a particular decision changed the law or whether it was merely declaratory. It would not be possible to lay down any hard and fast rules on this point. Each case would have to be decided on what may in the end be a matter of opinion, about which there may be room for a good deal of dispute. It is better to face up to the fact that every decision as to the law by a judge operates retrospectively, and to concentrate instead on the question--which I would regard as the critical question--whether the payer would have made the payment if he had known what he is now being told was the law. It is the state of the law at the time of the payment which will determine whether or not the payment was or was not legally due to be paid, and it is the state of mind of the payer at the time of payment which will determine whether he paid under a mistake. But there seems to me to be no reason in principle why the law of unjust enrichment should insist that that mistake must be capable of being demonstrated at the same time as the time when the payment was made. A mistake of fact may take some time to discover. If there is a dispute about this, the question whether there was a mistake may remain in doubt until the issue has been resolved by a judge. Why should this not be so where the mistake is one of law?

    In the present case we have no evidence about the state of the law at the time of the payments other than what can be derived from the agreed facts. But the background, as it can be discovered from the judgment of Hobhouse J. in the Westdeutsche case, is reasonably clear. He said at p. 931E that the effect of the statutory provisions of which the relevant bank had previously been unaware was subsequently "declared" by the Divisional Court and the House of Lords in the Hazell case. His choice of language seems to me to have been entirely appropriate. There had been no previous judicial decision on the point until the practice in the money markets was challenged for the first time in that case by the district auditor. Nor is it suggested that an opinion had been expressed about it which could be regarded as authoritative in the sense that it was binding on all parties including the auditor. If it were necessary to decide this point, I do not think that it would be right to say that the decision in the Hazell case "changed" the law. What it did was to clarify a point which had been overlooked and was in need of determination by the court. But the situation seems to me to be no different in principle from one where the facts are shown, as a result of inquiries which at the time of the payment were overlooked or not thought to be necessary, to have been different from what they had been thought to be at the time of the payment by the payer. Prima facie the Bank is entitled to restitution on the ground of mistake.

The Defences

    The question is whether the removal of the mistake of law rule requires, on grounds of public policy, that there should nevertheless be defences in mistake of law cases which are not available in mistake cases generally. It is appropriate as a first step however to recognise that the defences which are available generally already cover much of the ground where to allow recovery would lead to injustice. At this early stage it may be unwise to assume, until the matter has been tested on a case by case basis, that there are significant gaps in mistake of law cases which still need to be filled. Despite the careful study which has been given to this subject by the two Law Commissions, I would be inclined to proceed cautiously at this stage.

    The initial requirements already mentioned which the plaintiff must satisfy will do much to sort out those case which deserve a remedy and those which do not. He must show that he acted under a mistake which caused him to pay a sum which the payee was not legally entitled to receive. A payment made in the knowledge that there was a ground to contest liability will be irrecoverable: see Kelly v. Solari (1841) 9 M. & W. 54, 58 per Lord Abinger C.B. Then there are the defences of undoubted general application, as well as that of estoppel which requires no elaboration. The first is that of change of position which was recognised in Lipkin Gorman v. Karpnale Ltd. [1991] 2 A.C. 548. One of the examples which were given of its application by Lord Goff was where the plaintiff pays to the defendant under a mistake of fact and the defendant, while acting in good faith, pays the money or part of it to a charity. I think that it would be equally unjust to require the defendant to make restitution in such a case where the plaintiff pays the money under a mistake of law. The nature of the mistake makes no difference to the defendant who is acting in good faith. Mason C.J. seems to have been viewing this defence as applicable generally when he said in David Securities at p. 385 that a defence of change of position was necessary to ensure that enrichment of the recipient was prevented only where this would be unjust in a case where the defendant had acted to his or her detriment on the faith of the receipt.

    Then there is the defence that the money was paid as, or as part of, a compromise. Brennan J. in the same case said at p. 395 that, where a claim is satisfied by accord and satisfaction, a payment made in satisfaction is made in discharge of an obligation created by the accord: it is unaffected by any mistake as to the validity of the compromise. That must be so, irrespective whether the mistake is as to the facts or the law regarding its validity. In the Ontario Hydro case Dickson J. said at p. 380 that there was a head of public policy which recognised that there was a need to preserve the validity of compromises freely entered into with advice. I think that it is possible to find a more principled basis for the defence, as Brennan J. has suggested. But my main point is that it is available irrespective of the nature of the mistake.

    It has been suggested that it should be a defence that the money was paid in settlement of an honest claim: Goff and Jones, The Law of Restitution, 3rd ed. (1986), pp. 118-119. In the Ontario Hydro case at p. 364 Dickson J. accepted that a payment made in these circumstances would be irrecoverable, even if later events indicated that the payer was foolish to have acceded to the request for payment. It is not clear from his brief comment whether he saw this defence as one which would be relevant in mistake of law cases only. But in his summary of Dickson J.'s analysis in the Air Canada case at p. 191 La Forest J. treats this defence as one which would be applicable to any case of enrichment at the plaintiff's expense. In David Securities Mason C.J. in the majority judgment at p. 374 uses the expression "honest claim" to distinguish a voluntary payment which is made irrespective of the validity or invalidity of the obligation from the case where the payment is made under compulsion or undue influence. It is clear that the majority saw such factual circumstances, if relevant, as applicable to mistake of fact cases as well as to mistake of law cases. This was one reason why they regarded the mistake of law rule as broader and, as they put it, more preclusive than was necessary.

    In the Westdeutsche case Hobhouse J. said at p. 934 that the principle of voluntary payments could not be applied unless there was a conscious appreciation by the payer that the contracts were or might be void, and that on the evidence in the Islington case there clearly was no voluntary assumption of risk in any respect that was relevant. It is not clear, as there has been no evidence, whether there was a voluntary assumption of risk in any of the cases which are before us in these appeals. So I would not be prepared to say that it was a defence which in these cases was available. It is sufficient for my purpose that, while the precise limits of it have still to be clarified, it is a defence which applies generally irrespective of the nature of the mistake.

    Two defences have however been suggested which are designed specifically for cases where the mistake was a mistake of law. I take first the defence which was formulated in David Securities by Brennan J. in his dissenting opinion at pp. 398-399. He said that it should be a defence to a claim for money paid or property transferred under a mistake of law that the defendant honestly believed when he learned of the payment or transfer that he was entitled to receive and to retain the money or the property. I regret that, while I have derived much assistance from his judgment, I am unable to agree with him on this point. I have some difficulty in seeing why this defence, if there is merit in it as a means of preventing recovery where this would be unjust, should be confined to mistake of law cases. If an honest belief on the part of the payee can overcome the fact that it is prima facie unjust that the payer should not be able to recover what he paid under a mistake, why should this not be so in all cases? The reason, I think, is that in mistake of fact cases such a defence has never been recognised. To admit it now in such cases would be to run counter to many authorities. The defence seems to me to be based on expediency not on principle, and in any event to be too wide. But there are other objections. It does not sit easily with the defence of change of position. Indeed, in mistake of law cases, that defence would become unnecessary. The element of good faith would seem to be enough even though the defendant had not acted on the faith of the receipt. I think that this shows that it is lacking in principle. It would also tend to perpetuate the distinction between mistakes of fact and mistakes of law, which is itself undesirable. The Law Commissions have not supported it. I would not favour the adoption of the defence as part of English law.