Hostname: page-component-74656f9884-2dmjx Total loading time: 0 Render date: 2026-11-08T14:52:17.784Z Has data issue: false hasContentIssue false
  • English
  • Français

IN DEFENCE OF UNJUST ENRICHMENT

Published online by Cambridge University Press:  10 October 2019

Andrew Burrows Open the ORCID record for Andrew Burrows [Opens in a new window]

Abstract

This article seeks to defend the law of unjust enrichment against the recent influential attacks of Robert Stevens (“The Unjust Enrichment Disaster” (2018) 134 LQR 574) and Lionel Smith (“Restitution: A New Start?” in Devonshire and Havelock, The Impact of Equity and Restitution in Commerce (2018), ch. 5). A central argument here put forward is that there is a law of unjust enrichment, embodying a cause of action in unjust enrichment, which unites what Stevens and Smith see as disparate categories. A linked but separate argument is that, within the central area of unjust enrichment, Stevens is incorrect to regard the defendant's acceptance of performance as being necessary to trigger restitution albeit that acceptance may be relevant in establishing that the defendant has been enriched. A further, and more specific, argument is that, with great respect, the overruling, as a matter of principle, of Sempra Metals Ltd. v IRC [2007] UKHL 34, [2008] 1 A.C. 561, by the Supreme Court in Prudential Assurance Ltd. v HMRC [2018] UKSC 39, [2018] 3 WLR 652, seems unfortunate and appears to have been influenced by Stevens's excessively narrow approach to the meaning of “at the expense of”.

Keywords

Unjust EnrichmentRestitutionInterestCause of Action

Information

Type
Articles
Information
The Cambridge Law Journal , Volume 78 , Issue 3 , November 2019 , pp. 521 - 544
Copyright
Copyright © Cambridge Law Journal and Contributors 2019 

I. Introduction: Unjust Enrichment Scepticism, Unifying and Splitting

Despite the best efforts in this jurisdiction of, for example, Goff and JonesFootnote 1 and Peter Birks,Footnote 2 and the authoritative blessing given to the subject by the House of Lords in Lipkin Gorman v Karpnale Ltd.,Footnote 3 there remains significant scepticism about the utility of recognising a law of unjust enrichment. Modern sceptics include Steve Hedley,Footnote 4 Peter Jaffey,Footnote 5 Peter WattsFootnote 6 and Justice Gummow.Footnote 7 More recently, in two important and influential contributions to the debate, former unjust enrichment supporters – my close academic colleagues and friends, Robert StevensFootnote 8 and Lionel SmithFootnote 9 – have joined the sceptics.Footnote 10

Both Stevens and Smith agree that, even putting to one side restitution for wrongs, the law of unjust enrichment as conceived by, for example, Goff and Jones and Birks needs to be split up into distinct categories of law and should not be seen as unified by being based on the underpinning idea of unjust enrichment. So, for example, the cases on restitution for the compulsory discharge of another's liabilityFootnote 11 or on restitution from a public authority that has obtained a payment unlawfully, as in Woolwich Equitable Building Society v IRC,Footnote 12 or on restitution for the receipt of money by the defendant from a third party who has interfered with the claimant's rights, as in Lipkin Gorman v Karpnale Ltd.,Footnote 13 are, they argue, fundamentally different from a mistaken payment case, and from each other, and should not be treated as if they are alike. They also agree that, even within the central cases of restitution exemplified by a payment by C to D caused by C's mistake – where there is a direct transfer of value from C to D (or, as Stevens would prefer to phrase it, a “performance” by C for D) – one needs conduct by D to explain normatively why there is a reason to grant restitution to C.Footnote 14 The impairment of C's consent because of the mistake is insufficient to establish the relevant injustice. All this leads Stevens to argue that there is no law of unjust enrichment, that Birks's four-stage structured approach to a claim (by which one looks at “enrichment”, “at the expense of”, “unjust factors”, and “defences”) has led the House of Lords and Supreme Court into making many errors, that it is better to see the cases as falling within several disparate categories, and that unjust enrichment should therefore be viewed as a “disaster”. Taking a more moderate approach, Lionel Smith argues that we need to start again with only a small area that we can properly call the law of unjust enrichment.

This article seeks to defend the law of unjust enrichment against the attacks of Stevens and, to a lesser extent, Smith. A central argument here put forward is that there is a law of unjust enrichment, embodying a cause of action in unjust enrichment, which unites what Stevens and Smith see as disparate categories. A linked but separate argument is that, within the central area of unjust enrichment, Stevens is incorrect to regard the defendant's acceptance of performance as being necessary to trigger restitution albeit that acceptance may be relevant in establishing that the defendant has been enriched. A further, and more specific, argument is that, with great respect, the overruling, as a matter of principle, of Sempra Metals Ltd. v IRC Footnote 15 by the Supreme Court in Prudential Assurance Ltd. v HMRC Footnote 16 seems unfortunate and appears to have been influenced by Stevens's excessively narrow approach to the meaning of “at the expense of”.

It may be instructive to begin by going back to the late 1970s when I first studied Restitution on the BCL in Oxford in the seminars then being run by Birks and Jack Beatson. That was after the publication of Goff and Jones, The Law of Restitution in 1966 but before Birks's An Introduction to the Law of Restitution which was first published in 1985.Footnote 17 What was happening in those BCL seminars was that Birks and Beatson were trying to bring some sort of order to what was still then largely seen as a subject comprised of disparate categories albeit within the loose umbrella of Goff and Jones's principle against unjust enrichment. One of the great achievements of Birks was that he saw important connections between many of those apparently disparate categories and produced a logical and elegant scheme that, at a practical level, enabled lawyers to address the crucial questions efficiently.

This can be illustrated by pointing to a few of the questions that took up an enormous amount of our time in those seminars, yet which today would be regarded as straightforward; and this exercise can therefore be offered as some proof of just how practically important Birks's approach has been in moving the law forward. So we were puzzled by the question, what is the law on restitution of mistaken payments? Was it correct that there should continue to be a supposed liability rather than a causation test? Was it correct that there should be no restitution for payments made by mistake of law? Was there any connection between mistaken payments and payments made by duress? Was there any connection between payments made by mistake or duress and payments made for a total failure of consideration? Ought one to move to a change of position defence to a mistaken payment or was the law on the defence of estoppel sufficient? Was there any limitation period for restitution of payments made by mistake and, if so, where did one find it? Should there be a special public law ground for restitution from public authorities or were mistake and duress, if properly developed, sufficient? While remaining controversial today, although in my view the answers are clear, we were also puzzled, for example, as to whether there was any connection between mistaken payments and mistakenly rendered services or improvements to land; or between mistaken payments and the compulsory discharge of another's debt? And looming over everything, were there any common principles between the common law of so-called quasi-contract and the law on constructive trusts and tracing in equity (which were being covered by Derek Davies in a separate course of lectures)?

Two important points emerge from this personal historical reflection. The first is that, in the light of the 50 years of progress in understanding the law brought about by the call for, and the recognition of, a law of unjust enrichment, those arguing that the courts need to “start again” or that the whole enterprise has been a disaster, have a heavy burden of proof. Second, for those who criticise Birks as an excessive unifier and profess the merits of being a “splitter”,Footnote 18 one has to recognise the raw state that the law was in – and the difficulty of even agreeing on what were the relevant questions – when Birks started developing and refining his unifying approach.

Every decent common lawyer is to some extent a unifier. We are all concerned with generalising out from a past case to new or hypothetical facts and drawing links between different areas of the law. Why can one say that with such confidence? It rests on two indisputable propositions. The first is that at the heart of principled practical legal reasoning is reasoning by analogy. That is how the common law works. If you do not believe in reasoning by analogy then you are never going to master the common law. The second is that a fundamental principle of justice is that like cases should be treated alike. That underpins the drive towards unification – to finding connections – rather than to splitting.

This is not to deny that there is a serious risk of over-unifying or over-generalisation: in other words, of trying to treat cases alike – or drawing analogies – when they are better viewed as not being alike. Therefore everything hinges on achieving the right balance.

There is a danger in academic life of putting forward a theory and then spending the rest of one's career trying to defend that theory rather than accepting valid criticisms of it. Yet even with that in mind, it is my continuing view that, in achieving the right balance between splitting and unifying, there should be no dispute that there is a law of unjust enrichment or, put another way, that there is a cause of action in unjust enrichment. What we should be concerning ourselves with – where, in other words, there can be reasonable disagreement – is the width of that area of law or the scope of that cause of action. Birks went too far, for example, in extending his notion of “at the expense of” to include “but for” causation.Footnote 19 In other words, it was a mistake – and has occasionally led the courts into error – to have a starting vision where “at the expense of” means nothing more than a “but for” causal connection between the claimant's “loss” and the defendant's gain. But it is submitted that Stevens and, to a lesser extent, Smith are going too far in the opposite direction and risk throwing away the great strides forward achieved by Birks's work.Footnote 20

My essential argument here therefore is that, although the modern law of unjust enrichment can be seen as a battle between unifying and splitting, the really important question is how big is the law of unjust enrichment and not whether it exists at all. Even if it were correct that Birks went too far and swept too much into the law of unjust enrichment, the correct reaction to that should be to recognise the limits of the law of unjust enrichment and not to abandon it altogether.

II. A Response to Lionel Smith's Attack on Unjust Enrichment as a Cause of Action

Smith attacks the modern approach to unjust enrichment, as exemplified by the work of Birks, before sketching in bare outline a possible “middle way”Footnote 21 which sees some of the subject united around the idea of “unintended benefits” (including a transfer of rights) which have been requested or accepted.Footnote 22 This section is a reply to Smith's central argument that unjust enrichment is not a (single) cause of action and that Birks and others following him have gone wrong in thinking that it is.Footnote 23

One can state a cause of action at a higher or a lower level of generality. A cause of action is a series of facts (or as Birks preferred to express it, an event) which triggers a legal right/remedy (or, as Birks preferred to express it, a legal response).Footnote 24 Take the law of tort. Let us say that B has punched A. The relevant cause of action can be stated at a high level by saying that A has a cause of action against B in tort for damages. Or more specifically one can say that A has a cause of action against B for the tort of trespass to the person for damages. Or more specific still, one can say that A has a cause of action against B for the tort of battery for damages. One could also simply say that A has a cause of action against B for damages because B hit A. All these are correct statements and just differ in their specificity. Much the same applies to the law of unjust enrichment. One can say that, if A pays B £1,000 by mistake, A has a cause of action against B in unjust enrichment for restitution of £1,000. Or more specifically one can say that A has a cause of action against B for restitution of £1,000 because the money was paid by mistake. There is no difficulty in recognising that causes of action can be expressed in general or specific terms.

Just as the different torts have different characteristics but are united by the underpinning principle that they are common law wrongs (other than breach of contract) triggering compensation for loss caused so the different examples of unjust enrichment have different characteristics but are united by the underpinning principle of unjust enrichment. Not all examples of unjust enrichment have to have identical characteristics. Just as with torts, there can be differences. For example, just as contributory negligence is a defence to most torts but not all,Footnote 25 so change of position is a defence to most examples of restitution for unjust enrichment but not all.Footnote 26 What one is looking for is a rational explanation for differences. In so far as there is no rational reason for differences, they should be eliminated.

It is practically helpful – and certainly not something to be criticised – to break-down the cause of action of unjust enrichment into its constituent elements. This simply provides the conceptual structure of the cause of action. Enrichment, at the expense of, unjust factors, and defences, is nothing more than the equivalent in unjust enrichment of the conceptual structure of, say, the tort of negligence which one can divide into duty, breach of duty, causal damage and defences. Smith writes that, in the Birksian scheme, “the three-part test [enrichment, at the claimant's expense, unjust factor] was treated as applicable to the facts. It was treated as a cause of action”.Footnote 27 It is not entirely clear what is here being criticised. In so far as it is being suggested that unjust enrichment was seen as a cause of action that could be applied without reference to the existing law and precedent, then I disagree. While incremental principled development is the life-blood of the common law, Birks did not suggest that the elements of the cause of action in unjust enrichment are free-standing of the black letter law or can be used to ignore the present law or the need to respect precedent. In an important and welcome judgment, Lord Reed has made clear in Investment Trust Companies v HMRC Footnote 28 that that is the role of the four-part structured enquiry. In line with that, this is what I wrote on this issue, prior to that case, and continue to believe to be accurate:

The recognition that there is a claim in English law for restitution of an unjust (non-wrongful) enrichment does not mean that our law has moved to a discretionary system under which the judges simply ask themselves, “is this enrichment unjust?” On the contrary, the approach of the courts, in true common law style, is still very much an incremental one, whereby what is meant by unjust enrichment is dependent on legal principle as discerned from past cases …. In seeking to link the black letter law laid down in the cases to the cause of action of unjust enrichment, the most rational and helpful approach is to analyse any claim for restitution of an unjust enrichment in terms of four distinct questions: (1) has the defendant been benefited (i.e., enriched)? (2) was the enrichment at the claimant's expense? (3) was the enrichment unjust? (4) are there any defences? … These four elements – “benefit”, “at the claimant's expense”, “unjust factors”, and “defences” – constitute the fundamental conceptual structure of an unjust enrichment claim.Footnote 29

It may be correct that the straightforward language of “unjust enrichment at the claimant's expense” gives the impression that the law is more wide-ranging and loose than it in reality is; and no doubt claims will be made in which it may be in the interests of litigants to attempt to push forward the boundaries of the subject. But like any other area of the common law, there are significant dangers if the courts seek to go beyond incremental development of the law.Footnote 30 In this respect the overruling of TFL Management Services Ltd. v Lloyds Bank plc. Footnote 31 by the Supreme Court in Investment Trust Companies v HMRC Footnote 32 may be regarded as a clear signal that the courts should not be willing to allow claimants seeking restitution for unjust enrichment to overcome too easily attempts to throw out the claim at the striking out or (reverse) summary judgment stage.

Smith has argued that those believing in unjust enrichment have assumed that there is a single cause of action with identical characteristics.Footnote 33 It is not entirely clear what is meant by this. In my view, if one does believe in there being a law of unjust enrichment, however small, one can helpfully talk about unjust enrichment as a cause of action and one can structurally divide the issues between enrichment, at the expense of, unjust factors and defences. However, that does not equate to saying that one has to accept that the same answer will be given to each of those four elements whichever example of unjust enrichment one is talking about. Nor does it equate to saying that there is a single test for answering each of the four elements. In particular, there are many unjust factors, and while it is true that Birks in his final work saw them as united by the single master unjust factor of “absence of basis”,Footnote 34 that more civilian approach has not generally found favour among commentators or courts in this jurisdiction. But even if that were the correct approach to the unjust question, it is not clear that this has any relevance to whether unjust enrichment is a cause of action or not.

Another way of expressing what I am here arguing is as follows.Footnote 35 The classification/categorisation we are here concerned with is of causes of action. Unjust enrichment is a cause of action but, not least in order to stop duplication with other established causes of action in English law,Footnote 36 one excludes restitution of gains where the cause of action is a wrong (including breach of a contractual promise to make restitution).Footnote 37 At a very high level of generality, one can say that, behind that cause of action, there is a single normative principle against unjust enrichment (from which one is excluding restitution for wrongs).Footnote 38 But that is not to deny that there are more specific normative manifestations of that underlying principle and it may well be, therefore, that the subject is more normatively disparate than contract. The unjust factor triggering restitution of a payment made by mistake is significantly different from the unjust factor triggering restitution of a payment made to a public authority acting ultra vires but they are helpfully viewed alongside each other because they can both be straightforwardly treated as examples of an unjust enrichment at the claimant's expense triggering restitution.Footnote 39

Put another way still, there are significant differences in the law of tort between the intentional torts (for example, deceit) and strict liability torts (such as conversion). What the law of tort treats as a wrong may therefore be said to involve significantly different normative considerations. That does not undermine the utility of recognising tort as a classification of causes of action tied together by the single high level normative principle of there being a wrong (triggering compensation for causal loss). The same can be said of unjust enrichment. The single high level normative principle of unjust enrichment ties together what may be regarded as more specific causes of action which have significantly different normative explanations.

It may be that this debate can be helpfully thought of in terms of whether the cause of action in unjust enrichment is more like tort than contract. Perhaps Smith sees contract as the model that he has in mind whereby the making of the contract, its breach and causal non-remote loss may be thought of as comprising a purer, tighter and more unified set of rules than those in tort. However, it is hard to see why, in contrast to contract, a more disparate set of rules cannot form a practically useful and normatively justified cause of action; that is well-illustrated by the law of tort and that, in my view, is how we should view unjust enrichment. This also appears to be the view of the editors of Goff and Jones, The Law of Unjust Enrichment: “the law of unjust enrichment more closely resembles the law of torts (recognising a variety of reasons why a defendant must compensate a claimant for harm) than it does the law of contract (embodying a single principle that expectations engendered by binding promises must be fulfilled).”Footnote 40

III. A Response to Stevens's “Acceptance of Performance” Thesis

What has been argued in the last section in relation to Smith's attack also serves to defend unjust enrichment generally against Stevens's view, shared with Smith,Footnote 41 that supporters of unjust enrichment have unacceptably linked together what are, in reality, disparate categories of law.

This section turns to focus on Stevens's “acceptance of performance” thesis which is central to his analysis of specific aspects of the law (albeit that his criticisms of the reasoning and decisions of the House of Lords or Supreme Court are not reliant on just this thesis).Footnote 42 In Stevens's view, one needs a performance by the claimant for the defendant which is accepted by the defendant.Footnote 43 So according to Stevens it is unhelpful to separate out “enrichment” from “at the expense of” in a mistaken payment case because the important concept is performance by the claimant which is accepted by the defendant. Put another way, it is unhelpful to distinguish “enrichment” from “at the expense of” because performance covers both elements at one and the same time.

Indeed, according to Stevens – and this is another of his radical and important claims – one should not really be talking about enrichment at all, even in the central cases, because it is irrelevant whether the performance is beneficial or not to the defendant.Footnote 44 We are not concerned with a benefit in any normal sense of the word because there need be no consequential factual increase in the defendant's wealth. Moreover, what restitution is concerned with is the payment of the objective value of the performance not reversing enrichment. The implication is that the law would be better described as something like “the law of non-contractual accepted performance” not the law of unjust enrichment. Therefore when C pays D by mistake, C is performing for D and there has to be an acceptance by D not only because there is no such concept as a payment without the payee's acceptance but, in any event, normatively because there has to have been conduct by D to explain why there should be liability on D to pay back the money. It is the performance by C and its acceptance by D that meets the normative imperative – in line with the work of, for example, Ernest WeinribFootnote 45 – of correlativity between claimant and defendant.

The idea of performance is very close to my preferred approach to “at the expense of” across most (but not all) of unjust enrichment which is to talk of a direct conferral of a benefit by C on D.Footnote 46 Although the difference between us may, on the face of it, seem quite small, there are a number of problems with Stevens's “acceptance of performance” thesis.

A. The Language of “Performance”

The language of “performance” is very odd in the vast majority of restitution cases where one is talking about a (non-contractual) payment by C to D. The language of performance is fine where there has been a request in advance or even, if no request, where services are being rendered. However, it is a very odd use of the English language in a straightforward payment case to talk of performance and valuing performance. Take the insurance company's payment of money by mistake (thinking that there was a contractual obligation to do so) to the widow in the leading case of Kelly v Solari.Footnote 47 One would not naturally describe that as a performance by the insurance company to the widow or that restitution was there concerned with the objective valuation of that performance. Non-contractual payment and repayment are not helpfully described as performance and restitution for the value of that performance.

B. “Performance” not Enrichment

It is very hard to accept Stevens's criticismFootnote 48 that the law is not concerned with enrichment or benefit in any normal sense of the word. Say one were to ask, what do the following have in common: payments to D, the rendering of services to D, the transfer of goods or land to D, giving D the opportunity to use money, goods or land, or discharging a liability of D? It is surely an obvious and natural answer that those are all examples of benefits to D. And it is those situations that the cases on the law of unjust enrichment have been principally concerned with. True it is that D may validly argue that what is objectively a benefit may, in the light of D's particular preferences or circumstances, not be a benefit (or at least not a benefit measured by the market value) to this particular defendant. That was recognised long ago by Goff and Jones and Birks and underpinned Birks's notion of “subjective devaluation” which he saw as designed to respect the defendant's freedom of choice.Footnote 49 It is also true that there has been a debate about how one identifies the benefit where services have been rendered and there has been some confusion as to whether it is the services or the end product of those services that counts. The correct approach is that it depends on what is the direct benefit conferred. If C is a builder, building on D's land, C is directly providing services (including, commonly, materials) and that is the benefit that is being valued not the enhancement in value, if any, of D's land by reason of the building. That is the equivalent of saying that in a payment case it is the payment that is the benefit and not any additional profit from using the payment. It follows that in a services case, it may not matter that there is no end product of the services or that the end product does not enhance D's wealth such as the examples of painting coal white or constructing a hideous folly on D's land that reduces the land's value. In those examples, it is the services (including, commonly, the materials) that one is concerned to value, starting with the objective market value of those services which then may need to be reduced to respect the defendant's freedom of choice: so it is a saving of expense not a positive factual increase in the defendant's wealth that is in issue.

C. Does Acceptance Here Have Normative Force?

It is very hard to see why, normatively, acceptance in the weak sense in which Stevens is using it should be regarded as so important. If C owes D £1,000 in cash and pays over £1,100, mistakenly including £100, and D takes that £1,100 cash without checking it, it is hard to see what the importance of D's acceptance – which itself is mistaken – is meant to be. In other words, why should a mistaken acceptance of this kind count as being normatively crucial? Somewhat similarly, say £10,000 is incorrectly paid into D's bank account (because the payer had the incorrect bank account details and was intending to pay someone else). D, the customer, knows nothing about it. Yet, from the date of the payment into the account, D is bound to (has a liability) to make restitution of the £10,000. One initial problem for Stevens is to explain how in this situation C can be said to be performing for D.Footnote 50 Even if there is performance, is there any acceptance here by D? Stevens says that, although D has not herself done anything to accept, D's bank has done something to accept the payment and the bank is acting as D's agent to receive payments into D's bank account. Even if that is correct, it is hard to see the force of the normative work that acceptance is here supposed to be doing. Certainly acceptance may be normatively important where one is seeking to establish that the defendant is enriched, as in the case of services,Footnote 51 but that is not normally in issue in respect of payments.

Indeed the insistence on “acceptance” appears to be equivalent to saying that “free acceptance” by the defendant is necessary whatever the type of benefit. The concept of “free acceptance” was primarily developed by Goff and Jones and Birks as one method for overcoming “subjective devaluation”.Footnote 52 It rested on the idea that the defendant could not validly complain that she was not benefited by the objective market value of services or goods if she had had the opportunity to reject the services or goods, knowing that the claimant was expecting payment, but had chosen not to do so. The courts in recent cases appear to have recognised that free acceptance is a test of benefit.Footnote 53 However, on the Stevens approach, acceptance is normatively relevant not because it helps to establish that the defendant has been benefited but in order to establish conduct of the defendant as a correlative necessity to the conduct of the claimant.Footnote 54 The essential objection to Stevens is that, if it is clear that the defendant has been benefited, the normative work of the defendant's acceptance has been exhausted: the acceptance of the defendant has nothing to do with matching the conduct of the claimant.

More generally, one can understand that, where one is concerned with compensation for loss, one needs wrongful conduct of the defendant in order to pin the claimant's loss on the defendant. Therefore, in the realm of wrongs the correlativity between the claimant who has suffered the infringement of a right and the defendant who has infringed that right seems obvious. However, in the law of unjust enrichment, where one precisely is not concerned with wrongs, the equivalent inextricable link between claimant and defendant is to be found in the enrichment of the defendant at the claimant's expense.Footnote 55

D. Mistaken Improvement of Land or Goods

One of the ways to test the force of “the acceptance of performance” thesis is to ask what the law is, or ought to be, where C mistakenly improves D's land or goods in a situation where D has not had the opportunity to accept or reject (for example, because she does not know that the land or goods belong to her until after the improvements have been carried out). This may be described as the Greenwood v Bennett Footnote 56 situation. In that case, Harper had repaired a car that he had bought thinking that he had acquired good title to it. In fact the car belonged to the owners of a garage managed by Bennett. In interpleader proceedings taken out by the police to determine the parties’ rights, Harper claimed that he ought to be paid £226 for the improvements he had made to the car. The Court of Appeal allowed his claim. The majority, Phillimore and Cairns L.JJ., did so on the basis that the situation should be treated as if Bennett had brought a tort action (in what was then the tort of detinue and is now the tort of conversion) and that the court could require Bennett to reimburse Harper as a condition of an order for delivery up of the car.Footnote 57 But importantly Lord Denning reasoned that, irrespective of being able to impose a condition on delivery up, the mistaken improver, Harper, had a claim for the value of his work based on unjust enrichment.

It is submitted that, in line with Lord Denning's judgment in that case, C should be entitled to restitution provided it can be established that D is enriched. That appears also to be the law in some other major jurisdictions.Footnote 58 However, for Stevens, there can be no possible direct claim for restitution in this situation because there is no acceptanceFootnote 59 (and it may be that there are also issues as to whether there is a “performance for D” where C improves D's property thinking it belongs to C).Footnote 60 If Stevens is incorrect about that, his whole theory collapses.

E. Change of Position

Applying Stevens's thesis – that we are not concerned with unjust enrichment – it is not clear what role is being played by the change of position defence.Footnote 61 That defence, as best understood, is an enrichment-based defence and is applicable only to claims for unjust enrichment. Prima facie it applies to all claims for unjust enrichment although there may be particular reasons why it does not apply which have to be rationally explained (for example, where the Woolwich principle appliesFootnote 62 or where the defendant has been acting in bad faith).Footnote 63 Back in the BCL seminars in the late 1970s we used to debate whether one could properly talk of there being an English law of unjust enrichment unless and until a change of position defence had been recognised. We looked forward to the day when that would be so. That day came with Lipkin Gorman v Karpnale Ltd. Footnote 64 However, if one is rejecting unjust enrichment, it is not clear how the change of position defence can be justified or how it is meant to work. If one believes in a law of non-contractual accepted performance, why should change of position be a defence?

F. Contract?

Stevens's notion of “an accepted performance” looks as if it is the law of contract. The claimant offers performance on the basis that the defendant will pay the value of that performance and, by accepting performance, the defendant is impliedly promising that it will pay that value. We appear to have offer, acceptance and consideration. What may be happening therefore is that in a search for the supposed normative basis of the subject, Stevens has supplied it by turning to contract as the normative basis with D's promissory conduct being crucial. One can query therefore whether a law of “non-contractual accepted performance” makes sense and whether, in reality, Stevens is here really talking about contract.

G. The Narrow Meaning of “at the Expense of”

1. Three hypothetical examples

What is attractive about Stevens's thesis is that it chimes well with the recent authoritative clarification of the limits of unjust enrichment through the narrowing of the possible meaning of “at the expense of”. Let us take three simple hypothetical examples concerned with the meaning of “at the expense of”:

  1. (1) C pays £1,000 to X by mistake. X, as a result, pays D £750 as a gift. Is C entitled to restitution of £750 from D? This is an example of what can be described as “the third party issue”.

  2. (2) C mistakenly cuts down trees on her land which enhances the view that her neighbour, D, has from his land thereby increasing the market value of D's land by £5,000. Is C entitled to restitution from D of the market value of the work (let us say £3,500) in cutting down the trees? This is an example of what can be described as “the incidental benefit issue”.

  3. (3) C pays £10,000 to D by mistake. D uses that £10,000 over a three-month period to make a profit of £6,000. Applying the market rate of interest for three months on the £10,000 would yield, let us say, £500. Is C entitled to restitution of £16,000 from D? This is an example of what can be described as “the consequential benefit issue”.

After Investment Trust Companies v HMRC Footnote 65 and Prudential Assurance Co. v HMRC Footnote 66 it is clear that the answers to each of those three questions is “no”. There is no entitlement to restitution of £750 in example (a) or £3,500 in example (b) or £16,000 (as opposed to £10,000 or £10,500) in example (c). This is because in each of those situations, D's enrichment was not “at the expense of C” in the relevant sense. The best explanation for that turns on questions of policy that are analogous to those on legal causation and remoteness in the law of tort and, in my view, one needs, in general, a direct conferral of a benefit by C on D. A conferral by a third party,Footnote 67 incidental benefitsFootnote 68 and consequential benefits do not generally satisfy the “at the expense of” requirement because they do not constitute a direct conferral of a benefit by C on D.Footnote 69 True it is that a strength of Stevens's thesis is that it gives a clear and straightforward answer for why there should be no restitution without recourse to “directness” policy considerations. According to Stevens, there is no claim to restitution by C from D in those three situations because there was no relevant performance by C to D. In (a), C's performance was for X not for D. In (b), C was not performing for anyone other than herself. And in (c), the only performance was C's payment of £10,000 to D. But one can equally well explain all that by reference to “at the expense of” requiring there to be, in these types of cases, a direct conferral of a benefit by C on D.

Moreover, while the Stevens “performance thesis” may produce neat and correct answers to those hypothetical examples, it is, arguably, too narrow and inflexible and gives incorrect answers in other situations. An important example of this is that, applying his “performance” approach, Stevens argued that the decision of the House of Lords in Sempra Metals Ltd. v IRC,Footnote 70 to award compound interest as restitution of an unjust enrichment, was incorrect as a matter of principle. In Stevens's words: “The relevant performance is the payment of the capital sum. The next day, and every day thereafter, there is no fresh performance capable of supporting a claim.”Footnote 71

Although in Prudential the Supreme Court, apparently relying on Stevens's article (albeit unpublished at the time of the hearing),Footnote 72 has overruled Sempra Metals, this is highly controversial and, with respect to the Supreme Court, seems incorrect as a matter of principle.

2. Was Prudential correct to overrule Sempra Metals?

The Supreme Court indicated that there were several reasons for regarding the decision on restitution in Sempra Metals as problematic.Footnote 73 These included first, the disruption to public finances that was being caused by that decision taken together with Kleinwort Benson Ltd. v Lincoln City Council Footnote 74 (allowing restitution for payments, including taxes, made by mistake of law) and, second, the clash with statutes on the repayment of overpaid taxes which have allowed simple interest only (such as in sections 78 and 80 of the Value Added Tax Act 1994). However, the essential reasoning of the Supreme Court in Prudential was that Sempra Metals was incorrectly decided as a matter of principle because the opportunity to use the money was not “at the expense of the claimant” in the narrow sense required by the law of unjust enrichment.

In examining this, it is noteworthy that the overruling of Sempra Metals was not what was argued for by HMRC in Prudential.Footnote 75 HMRC instead put forward arguments that would have denied the disputed compound interest claimed while leaving Sempra Metals intact.Footnote 76

Let us first consider the Sempra Metals award itself. Say C pays D £20,000 by mistake. Five years later C discovers the mistake. C is indisputably entitled to restitution of £20,000 (subject to defences). According to Sempra Metals C is also entitled to interest (which may be compound, depending on the facts) on the £20,000 in the law of unjust enrichment (subject to any clash with section 35A of the Senior Courts Act 1981 which allows simple interest only).Footnote 77 This was applying the reasoning that C transferred to D not merely the £20,000 but also the opportunity to use that £20,000. It is no doubt true that there was only one performance – the payment of the capital sum of money – but the reasoning of the majority of the House of Lords, led by Lord Nicholls, was that more than one benefit was being transferred. Lord Nicholls for the majority said the following:

The benefits transferred by Sempra to the Inland Revenue comprised, in short, (1) the amounts of tax paid to the Inland Revenue and … (2) the opportunity for the Inland Revenue, or the Government of which the Inland Revenue is a department, to use this money for the period of prematurity. The Inland Revenue was enriched by the latter head in addition to the former. … Restitution, if it is to be complete, must encompass both heads. Restitution by the revenue requires (1) repayment of the amounts of tax paid prematurely (this claim became spent once set off occurred) and (2) payment for having the use of the money for the period of prematurity.Footnote 78

In Prudential, the Supreme Court disagreed with that reasoning of Lord Nicholls. According to the Supreme Court, the only benefit directly transferred was the capital sum. But surely, in line with Lord Nicholls's reasoning, there were indeed two separate benefits directly provided or transferred, namely the capital sum and the opportunity to use that capital sum. To focus on the former, to the exclusion of the latter, leaves a recipient directly and unjustly enriched.Footnote 79

One can test the force of Sempra Metals, as against Prudential, by switching from money to a car example. C transfers a car to D mistakenly believing that D is entitled to the car under the terms of her employment. C discovers the mistake after one year and informs D that she must return the car which she does. Surely – and in line with Sempra Metals – C is entitled prima facie (i.e. assuming no problems about “subjective devaluation” or a change of position defence) to a restitutionary remedy in the law of unjust enrichment for the use value of the car during that year because otherwise D would have had the beneficial use of the car, which has been mistakenly transferred directly by C, for nothing.Footnote 80 However, the decision in Prudential would mean that there would be no entitlement to an award for the one year's use of the car in the law of unjust enrichment because there has been only one direct transfer of value (or on Stevens's approach, only one performance), namely the transfer of the title to the car.

Or, to take another similar example, say C is induced to buy a car from D for £5,000 by D's innocent misrepresentation as to the car's mileage. After six months, C discovers the truth and immediately rescinds the contract by returning the car and D returns the £5,000 to C. But that would leave each party unjustly enriched by the opportunity to use, during the six-month period, the money and the car. Just as C should be entitled to interest on the £5,000 payment so D should be entitled to payment (a quantum valebat) for C's six-month use of the car. This is straightforward restitution and counter-restitution of the benefits each has conferred on the other. But applying the performance thesis, neither party would be entitled to interest or use value because each is only making one performance – that is, C is paying D and D is transferring title to the car to C. The fact that C is also conferring on D (i.e. is also transferring to D) the opportunity to use the money and that D is also conferring on C (i.e. is also transferring to C) the opportunity to use the car is being treated as irrelevant. That seems needlessly narrow.

Although these matters are not straightforward, it would seem that, apparently influenced by Stevens's “performance” approach, Prudential was incorrect to overrule Sempra Metals in so far as it did so as a matter of principle.Footnote 81

For the purposes of completeness, it should be stressed that, in Prudential, there was a valid argument of principle for denying that part of the interest claimed which constituted interest on top of the Sempra Metals award of interest.Footnote 82 That is because that interest could not be said to have been directly conferred by the claimant on the defendant. If C pays D £1,000 tax which was unlawfully levied or mistakenly paid, the opportunity to use that £1,000 tax obviously ends when the £1,000 tax is repaid or is offset against lawful tax. So that while C should be entitled as restitution of unjust enrichment to interest on the £1,000 until repaid or offset, because that is the opportunity to use transferred by C to D, C is not entitled, as restitution of an unjust enrichment, to interest on top of the value of the opportunity to use that was transferred. True it is that D may gain by not repaying the interest on the £1,000 tax payment. But that further gain has not been transferred by C to D and, therefore, is not at the expense of C in the relevant sense.Footnote 83 That gain is derived from D not paying back to C the debt comprising the interest on the tax payment. It is indistinguishable from any other failure to pay a debt.Footnote 84

IV. The Positive Case as to What Constitutes the Law of Unjust Enrichment

It is all too easy to criticise the work of others. This final section therefore deals with the positive case, very briefly summarised, as to what the law of unjust enrichment comprises.Footnote 85

Having taken out wrongful enrichment, the unifying underpinning of this area of the law is unjust enrichment at the claimant's expense. There is no great mystery to this. To take the central case of a mistaken payment, the payee has to repay the mistaken payment precisely because the payee has been unjustly enriched at the payor's expense.Footnote 86 At a lower level of specificity, D has to repay because, albeit without committing a wrong, D has received a benefit that has been non-consensually conferred on D by C. More specifically still, C has not truly consented in conferring the benefit on D because the mistake impaired C's consent.

The vast bulk of the subject is directly analogous to a claim for restitution of a mistaken payment because it shares the following four features:

  1. (1) the defendant has received a benefit (whether a payment or services or goods or land or the opportunity to use money, goods or land) which has been directly conferred on the defendant by the claimant; and

  2. (2) that enrichment at the expense of the claimant was unjust because the claimant's consent to the conferral of the benefit was impaired, whether by, for example, duress or undue influence, or was conditional (i.e. conferred for a consideration) and that condition (that consideration) has failed; or because, irrespective of impaired or conditional consent, there is a policy reason (Birks called this “policy-motivated restitution”)Footnote 87 to reverse the enrichment as, for example, where it was conferred on (or by) a public authority ultra vires or where it was conferred in circumstances of necessity; and

  3. (3) restitution reverses that enrichment by a personal monetary award measured by the value to the defendant of the benefit conferred by the claimant; and

  4. (4) good faith change of position is prima facie a defence open to the defendant although there may be a good reason why in relation to the particular unjust factor (especially where based on a particular policy) the defence does not apply (as, for example, in a Woolwich claim).Footnote 88

The (non-contractual) law on restitution for discharge of another's obligationFootnote 89 follows a very similar pattern and is therefore also helpfully viewed as within the law of unjust enrichment. For Stevens there is here no connection with mistaken payments because there is no performance by C for D: rather the performance is for X albeit that the liability discharged is that of D. But this area has always been seen as central to unjust enrichment and rightly so. It belongs within Birks's “policy-motivated restitution”. There is a clear policy reason for restitution which is to ensure that the defendant does not undeservedly escape from an obligation. On the face of it, a major difference between this and the usual case of unjust enrichment is that the conferral of the benefit on the defendant by the claimant may be incidental (incidental benefits do not normally count).Footnote 90 However, while the benefit may be incidental in the sense that the discharge of D's liability is secondary to C's conduct, it would be absurd to treat the benefit as incidental for the following reason. The question of when another's liability is discharged by C's payment or services rendered to X is a difficult question of law. Having decided that there is such a discharge, it would contradict that legal rule then to say that the benefit is incidental. One might say either that this area epitomises a valid exception to the need for there to be a direct conferral of a benefit or that, because of the automatic connection in law, there is a direct conferral of a benefit on D when C pays X discharging D's liability.

More controversially, the law of unjust enrichment may also be said to embrace restitution for receiving C's (traceable) asset from a third party without C's consent/authority (i.e. the asset has been wrongfully “taken” from C by X).Footnote 91 A major difference here from the usual unjust enrichment case is that there is no relevant conferral of a benefit on D by C. Rather the enrichment is obtained by D receiving C's (traced) asset from X and it does not matter that the benefit has been conferred by X not C. What this area essentially shares with the bulk of unjust enrichment set out above is that: first, the cause of action does not involve a wrong by D; second, although there is no conferral of a benefit on D by C, there is a strong link, here through tracing, between C's “loss” and D's gain; and, third, the cause of action only accrues as and when D is enriched. The classic examples in the cases are where (personal) restitution is awarded against a defendant who is not a bona fide purchaser for value without notice and has received from a third party wrongdoer an asset that the claimant can trace into the defendant's “hands”. The leading example in the case law is Lipkin Gorman v Karpnale Ltd.Footnote 92

Although this too is controversial, instead of a personal monetary award, as in (3) above, one can argue that the law of unjust enrichment sometimes reverses the enrichment by “proprietary restitution”.Footnote 93 This is particularly crucial where the defendant is insolvent so that the normal personal monetary award is unlikely to be satisfied. Proprietary restitution may be effected by giving the claimant a lien/charge (including by subrogation to a lien/charge) over a relevant asset of the defendant which secures the monetary award measured by the value of the defendant's enrichment. But proprietary restitution may also be effected by revesting rights (through rescission) or imposing a (constructive) trust of rights. In those situations, the enrichment comprises rights that the defendant holds and, in order to effect restitution, there is no need to ascertain the value of those rights.

Although Stevens concludes his “disaster” article by taking issue with the decision or reasoning in nearly all the cases on restitution that have reached the House of Lords or the Supreme Court this century, it is my view that, while some of them require an extended notion of “at the expense of” as an exception to the normal “direct conferral” approach,Footnote 94 nearly all the decisions and reasoning that Stevens criticises can be justified. Once one has boxed oneself in by a narrow and inflexible notion of “accepted performance”,Footnote 95 it follows that some of those decisions and/or the reasoning in them will appear as problematic. With respect, that is a problem with the theory not with the decisions or reasoning.

V. Conclusion

This article has been concerned to defend the law of unjust enrichment against a powerful new wave of scepticism. There are worrying signs that this new wave may carry the courts with it. The huge improvement in the law that we have seen in this jurisdiction over the last 30–40 years is under threat. It is therefore important to reiterate the importance of recognising the law of unjust enrichment and its limits. Above all, the difficulties with an overextended approach to the meaning of “at the expense of” do not provide a good reason for “jumping ship” or setting sail on different waters. Unjust enrichment may not yet be a glorious cruise-liner but it is certainly not a disastrous shipwreck.

Footnotes

*

Q.C., F.B.A., D.C.L., Professor of the Law of England in the University of Oxford and Fellow of All Souls College.

I would like to thank Charles Mitchell, Helen Scott, Stephen Smith, Lionel Smith and two anonymous reviewers for their helpful comments on earlier versions of this article.

References

1 The first edition of Goff, R. and Jones, G., The Law of Restitution, 1st ed. (London) was published in 1966Google Scholar. Under C. Mitchell, P. Mitchell and S. Watterson, the title of the eighth edition in 2011 was changed to The Law of Unjust Enrichment and the material on restitution for wrongs was excluded. The latest (ninth) edition was published in London in 2016.

2 Birks, P., An Introduction to the Law of Restitution, revised ed. (Oxford 1989)Google Scholar; The Law of Unjust Enrichment, 2nd ed. (Oxford 2005).

3 Lipkin Gorman v Karpnale Ltd. [1991] 2 A.C. 548.

4 See e.g. Hedley, S., “Unjust Enrichment as the Basis of Restitution – an Overworked Concept” (1985) 5 LS 56Google Scholar; “Unjust Enrichment” [1995] C.L.J. 578; “Restitution: Contract's Twin?” in F. Rose (ed.), Failure of Contact (Oxford 1997), 247–74; Restitution: Its Division and Ordering (London 2001); A Critical Introduction to Restitution (London 2001); “Unjust Enrichment – a Middle Course?” (2002) 2 O.U.C.L.J. 181; “Implied Contract and Restitution” [2004] C.L.J. 435; “Farewell to Unjustified Enrichment: A Common Law Response” (2016) 20 Edin.L.R. 326. For some of my replies, see e.g. Burrows, A., Understanding the Law of Obligations (Oxford 1998) 99–108Google Scholar; The Law of Restitution, 3rd ed. (Oxford 2011), 34.

5 Jaffey, P., The Nature and Scope of Restitution (Oxford 2000)Google Scholar; Private Law and Property Claims (Oxford 2007). For criticism, see Burrows, Law of Restitution, p. 33.

6 Watts, P., “Restitution – a Property Principle and a Services Principle” [1995] R.L.R. 49Google Scholar; “‘Unjust Enrichment’– the Potion that Induces Well-meaning Sloppiness of Thought” (2016) 69 C.L.P. 289. The latter words are those of Scrutton L.J. in Holt v Markham [1923] 1 K.B. 504, 513, who was referring to the history of the action for money had and received and, in particular, Lord Mansfield's approach which may be seen as the basis for the modern law of unjust enrichment.

7 Formerly of the High Court of Australia. See e.g. his judgment in Roxborough v Rothmans of Pall Mall Australia Ltd. (2001) 208 C.L.R. 516. For my response, see , Burrows, “The Australian Law of Restitution: Has the High Court Lost Its Way?” in Bant, E. and Harding, M. (eds.), Exploring Private Law (Cambridge 2010), ch. 3Google Scholar.

8 Stevens, R., “The Unjust Enrichment Disaster” (2018) 134 L.Q.R. 574Google Scholar.

9 Smith, L., “Restitution: A New Start?” in Devonshire, P. and Havelock, R. (eds.), The Impact of Equity and Restitution in Commerce (Oxford 2018), ch. 5Google Scholar.

10 There has also been an interesting attack on the unity of unjustified enrichment in civil law by Jansen, N., “Farewell to Unjustified Enrichment” (2016) 20 Edin.L.R. 123CrossRefGoogle Scholar. For a reply to that article, see MacQueen, H., “The Sophistication of Unjustified Enrichment: A Response to Nils Jansen” (2016) 20 Edin.L.R. 312CrossRefGoogle Scholar.

11 Burrows, Law of Restitution, ch. 17.

12 Woolwich Equitable Building Society v IRC [1993] A.C. 70.

13 Lipkin Gorman [1991] 2 A.C. 548.

14 Stevens, “Disaster”, pp. 577–78, 581–82; Smith, “A New Start?”, p. 98: “I now think that when explaining why the defendant is liable, we must always consider, and can never ignore, the defendant's role in the events that generate liability.”

15 Sempra Metals Ltd. v IRC [2007] UKHL 34, [2008] 1 A.C. 561.

16 Prudential Assurance Ltd. v HMRC [2018] UKSC 39, [2018] 3 W.L.R. 652.

17 There was a revised paperback edition published in 1989.

18 In co-teaching with Stevens on the Oxford BCL, he has often used this terminology.

19 For an in-depth examination of this, see Burrows, “‘At the Expense of the Claimant’: A Fresh Look” [2017] R.L.R. 167. For similar reasons I think it was also a mistake for Birks to include the idea of “interceptive subtraction”: see Birks, An Introduction, pp. 133–39.

20 It may be that what we are seeing in relation to unjust enrichment is the almost inevitable cycle that occurs with general theories. A general theory, which seeks to explain large areas, initially makes huge strides forward before it comes under attack as being too wide-ranging. So if one looks, e.g., at the German law of unjustified enrichment, one sees in the early twentieth century the development of a general theory of unjustified enrichment. Subsequently in the 1950s the very influential work of Wilburg and von Caemmerer cut back on that general theory by arguing that the law was better understood as confined to four specific categories which were seen as loosely linked together by the idea of unjustified enrichment: performance, infringement of rights, discharge of another's liability, improvement of another's property. But in Germany, almost inevitably, the pressure is on to expand those four categories. They are a good solid base but reasoning outwards from them is almost inevitable. In particular, one can helpfully think of three not four categories: performance, infringement of rights, and enrichment by neither performance nor infringement of rights. That last category would sweep up e.g. compulsory discharge of another's debt and mistaken improvements of land but would extend to other cases as well.

21 Smith, “A New Start?”, p. 112.

22 Ibid., at pp. 106–12.

23 Ibid., at pp. 88–100.

24 “[A cause of action is] every fact which it would be necessary for the plaintiff to prove, if traversed, in order to support his right to the judgment of the Court”: Read v Brown (1888) 22 Q.B.D. 128, 131, per Lord Esher. See also Letang v Cooper [1965] 1 Q.B. 232, 242–43. While a cause of action isolates particular facts, it is the legal consequence of those facts that matters.

25 E.g. contributory negligence is not a defence to the tort of deceit (Standard Chartered Bank v Pakistan National Shipping Corp. (No 2) [2002] UKHL 43, [2003] 1 A.C. 959) or the tort of trespass to the person (Pritchard v Co-operative Group Ltd. [2011] EWCA Civ 329, [2012] Q.B. 320). Smith himself refers to “justification” being a defence to a few torts, such as the tort of inducing breach of contract, but not most torts.

26 Henderson J. accepted that change of position does not apply to a “Woolwich claim” for restitution from a public authority: Test Claimants in the FII Group Litigation v HMRC (No 2) [2014] EWHC 4302 (Ch), [2015] S.T.C. 1471, at [309]–[315]. For another example, see Skandinaviska Enskilda Banken AB (Publ) v Conway [2019] UKPC 36 (in the context of a liquidator's claim for restitution from a preferred creditor, it was held that the preferred creditor could not invoke a change of position defence because this would undermine the statutory regime dealing with voidable preferences).

27 Smith, “A New Start?”, p. 93.

28 Investment Trust Companies v HMRC [2017] UKSC 29, [2017] 2 W.L.R 1200, at [39]–[42]. See also Lord Reed's similar formulation in relation to the duty of care in tort in Robinson v Chief Constable of West Yorkshire [2018] UKSC 11, [2018] A.C. 736, at [21]–[29]: it is noteworthy that, of course, no-one would suggest that this clarification casts doubt on the existence of the tort of negligence as a cause of action.

29 Burrows, Law of Restitution, pp. 26–27.

30 The classic example of the dangers of rejecting incremental development is in relation to the tort of negligence where Anns v Merton LBC [1978] A.C. 728 – with its approach of setting out a principled prima facie liability which could then be cut-back for policy reasons – was subsequently rejected for having pushed the law too far forward and away from existing precedent.

31 TFL Management Services Ltd. v Lloyds Bank plc. [2013] EWCA Civ 1415, [2014] 1 W.L.R. 2006. In this case, A had gone to court to try to establish that X owed A money. In fact the legal proceedings established that X owed the money to B. The question was whether A had a claim for restitution in the law of unjust enrichment against B for enriching B by the expense of those legal proceedings. The Court of Appeal refused to strike out the claim. But, as made clear by the Supreme Court in Investment Trust Companies [2017] UKSC 29, [2017] 2 W.L.R. 1200, at [56]–[57], the claim should have been struck out because the benefit was an incidental benefit.

32 Investment Trust Companies [2017] UKSC 29, [2017] 2 W.L.R. 1200.

33 Smith, “A New Start?”, p. 91. See also Stevens, “Disaster”, pp. 576–77.

34 Birks, Unjust Enrichment.

35 Yet another way of expressing my argument is as follows. Unjust enrichment is concerned with all gain-based claims (claims to restitution) other than those which are excluded because they rest on other well-established causes of action (e.g. restitution for wrongs).

36 It is an interesting question whether the exclusion of restitution for wrongs (including breach of a contract to make restitution) is principled or pragmatic. Say e.g. in Attorney-General v Blake [2001] 1 A.C. 268 the royalties were received by Blake more than six years after the breach of contract. Could the Crown have successfully argued that the limitation period should run from when the ill-gotten gains were first received by Blake, rather than from the date of the breach of contract – on the basis that Blake's enrichment was essential to an unjust enrichment cause of action?

37 This can also explain why delivery up of a chattel which one owns is not regarded as restitution of an unjust enrichment, i.e. it is a remedy for the tort of conversion. But, in any event, the vindication of pre-existing proprietary rights differs from unjust enrichment because, in contrast to unjust enrichment, it does not create a new right to restitution.

38 That there is an underlying normative principle against unjust enrichment may be traced back to Roman law times (the second century AD): see e.g. Pomponius, Digest 12.6.14, “According to the law of nature, it is equitable that no person be enriched to the detriment of another”; and Digest 50.17.206, “According to the law of nature, it is equitable that no person shall be wrongly enriched to the detriment of another”. In the seventeenth century, Grotius in his Introduction to Roman-Dutch Law, at 3.30.3, wrote: “equity does not permit that one person should be enriched at another's expense.”

39 Contrast J. Edelman and E. Bant, Unjust Enrichment, 2nd ed. (Oxford), ch. 13 who argue that Birks's policy-based restitution should not be viewed as part of the law of unjust enrichment because the policy reason for restitution is independent of considerations between the parties.

40 Mitchell et al., The Law of Unjust Enrichment, 9th ed. See further Mitchell, C., “Other Reasons for Restitution” in Bant, E., Barker, K. and Degeling, S. (eds.), Unjust Enrichment and Restitution Handbook, (Cheltenham 2020), ch. 20Google Scholar. See also Smith, S., “Unjust Enrichment: Nearer to Tort than Contract” in Chambers, R., Mitchell, C. and Penner, J. (eds.), Philosophical Foundations of the Law of Unjust Enrichment (Oxford 2009), pp. 181–208CrossRefGoogle Scholar.

41 See above text to notes 11–13 above.

42 In particular he has a separate criticism of Deutsche Morgan Grenfell v IRC [2006] UKHL 49, [2007] 1 A.C. 558 and International Energy Group v Zurich Insurance [2015] UKSC 33, [2016] A.C. 509 for infringing the rule – which he sees as a rigid rule without exceptions – that one cannot have restitution of a benefit that was legally owed (under a valid contractual or statutory obligation) by the claimant to the defendant. My view is that there can be exceptions to that rule and that Deutsche Morgan is a justified exception because the very unlawfulness in question was the non-provision of an option for group relief which the claimant would have exercised had it been available.

43 Stevens, “Disaster”, p. 581.

44 Ibid., at p. 580.

45 E.g. Weinrib, E., “The Normative Structure of Unjust Enrichment” in Rickett, C. and Grantham, R. (eds.), Structure and Justification in Private Law (Oxford 2008), 21Google Scholar; “Correctively Unjust Enrichment” in Chambers et al., Philosophical Foundations, pp. 81, 86–93.

46 Burrows, “At the Expense of the Claimant”. One may similarly talk of a “direct transfer of value from C to D”.

47 Kelly v Solari (1841) 11 L.J. Ex. 10.

48 Stevens, “Disaster”, pp. 579–80. See above text to note 44.

49 There was some criticism of the language of “subjective devaluation” in Benedetti v Sawiris [2013] UKSC 50, [2014] A.C. 938 but that did not cast doubt on the essential point, underlying Birks's terminology, that one needs to respect the freedom of choice of the defendant or, put another way, one needs to ensure that the objective benefit is a benefit to this particular defendant.

50 This is insisted on by Stevens as essential in understanding what is meant by “performance”: see Stevens, “Disaster”, p. 581. In order for the performance here to be “for D”, Stevens would presumably have to say that the payor's performance is for the person with those bank account details. But that would seem to undermine the reason for saying that the performance must be “for” D. In reality the mistake means that the performance is “for” X not D.

51 Hence the idea, accepted in the recent case law, that “free acceptance” is a test of benefit (see note 53 below). That a defendant can easily “subjectively devalue” services is best captured in the well-known statement by Pollock C.B. in Taylor v Laird (1856) 25 L.J. Ex. 329, 332: “One cleans another's shoes; what can the other do but put them on?”

52 Goff and Jones, The Law of Restitution, 1st ed., pp. 30–31; Birks, An Introduction, p. 104; Birks, P., “In Defence of Free Acceptance” in Burrows, A. (ed.), Essays on the Law of Restitution (Oxford 1991), 105, 144–45Google Scholar. Birks also initially saw “free acceptance” as constituting a separate unjust factor.

53 Cressman v Coys of Kensington (Sales) Ltd. [2004] EWCA Civ 47, [2004] 1 W.L.R. 2775; Chief Constable of the Greater Manchester Police v Wigan Athletic A.F.C. Ltd. [2008] EWCA Civ 1449, [2009] 1 W.L.R. 1580, and Benedetti v Sawiris [2010] EWCA Civ 1427 (reversed without dealing with this point by the Supreme Court in Benedetti [2013] UKSC 50, [2014] A.C. 938). See also C. Mitchell, P. Mitchell and S. Watterson, Goff and Jones on the Law of Unjust Enrichment, 9th ed. (2016), paras. 4-45 to 4-50.

54 See similarly e.g. Weinrib, E., “The Structure of Unjustness” (2012) 92 B.U.L. Rev. 1067Google Scholar. As Weinrib makes clear, at p. 1072, his acceptance condition “has nothing to do with establishing the enrichment; [instead, acceptance] makes the unjustness of the [payee's] retention of the benefit correlative to the unjustness of the [payor's] gratuitous but non-donative transfer”. For criticism of Weinrib's emphasis on acceptance, including his “deeming” of an acceptance’ to deal with incontrovertible benefits, see e.g. D. Klimchuk, “The Normative Foundations of Unjust Enrichment” in Chambers et al., Philosophical Foundations, pp. 81, 86–93.

55 If one takes Stephen Smith's view – see e.g. Rights, Wrongs, and Injustices: The Structure of Remedial Law (forthcoming, OUP) – that conduct by the defendant is necessary to establish a substantive duty owed by the defendant, one might say that the defendant has no (substantive) duty to the claimant to make restitution but is merely subject to a liability, and the claimant has the power to obtain, a court order against the defendant. In other words, it would appear that, normatively, Smith would regard there as being no requirement for conduct of the defendant because one is not trying to explain a (substantive) duty owed by the defendant to the claimant but merely a liability.

56 Greenwood v Bennett [1973] Q.B. 195. For discussion see Burrows, Law of Restitution, pp. 237–39.

57 See now s. 6(1) of the Torts (Interference with Goods) Act 1977.

58 See e.g. Du Plessis, J., The South African Law of Unjustified Enrichment (Cape Town 2012), ch. 9Google Scholar; American Law Institute, Restatement of the Law Third, Restitution and Unjust Enrichment (Philadelphia 2011), pp. 111–36Google Scholar.

59 Stevens, “Disaster”, pp. 583–84.

60 Ibid., at p. 581.

61 Ibid., at p. 587.

62 See note 26 above.

63 The judicial formulations of the defence stress that it is not open to a defendant who has changed its position in bad faith: see e.g. Lipkin Gorman [1991] 2 A.C. 548, 580.

64 Ibid.

65 Investment Trust Companies [2017] UKSC 29, [2017] 2 W.L.R. 1200.

66 Prudential Assurance Ltd. [2018] UKSC 39, [2018] 3 W.L.R. 652.

67 Where there is a contract between X and C, under which C is required to carry out services for the benefit of D in return for payment by X, it is X, not C, who is directly conferring the benefit on D: see MacDonald Dickens & Macklin v Costello [2011] EWCA Civ 930, [2011] 3 W.L.R. 1341; Lumbers v W Cook Builders Pty Ltd. [2008] HCA 27, (2008) 232 C.L.R. 635. An important policy behind this is that the law on unjust enrichment ought not to undermine the risks, including the risk of insolvency, which have been undertaken in the contract between C and X. See Burrows, Law of Restitution, pp. 74–75; Burrows, A., A Restatement of the English Law of Unjust Enrichment (Oxford 2012), 52–54Google Scholar.

68 An incidental benefit is one that is a secondary consequence from the claimant's actions. E.g., A heats A's flat and the rising heat, heats B's flat; A drains her land and also drains B's land; A destroys his stamp which enhances the value of B's stamp; A goes to court to try to establish that X owes A money but in fact it is established that X owes B money. It is incorrect to think that the key to understanding an incidental benefit is whether A is acting in self-interest because in almost all cases of restitution of unjust enrichment the claimant was acting in self-interest. It is true that in three of the above hypothetical examples A was acting in self-interest but the reason for denying restitution is that the benefit conferred was a secondary consequence of A's (self-interested) actions. In the stamp example, A was not acting in self-interest but contrary to his own interests.

69 Burrows, “At the Expense of the Claimant”. See also A. Burrows, “Unjust Enrichment and Restitution” in The Oxford Handbook of New Private Law (forthcoming, OUP). For reference to a concept of “directness”, see e.g. Tettenborn, A., Law of Restitution in England and Ireland, 3rd ed. (London 2002), 30–31Google Scholar; Dawson, J.P., Unjust Enrichment: A Comparative Analysis (Boston 1951), 120–27Google Scholar.

70 Sempra Metals Ltd. [2007] UKHL 34, [2008] 1 A.C. 561. The question as to whether Prudential was correct to overrule Sempra Metals was left open for Australian law (while pointing out some of the issues that would need to be considered) by Justice Edelman in Northern Territory v Griffiths [2019] HCA 7, at [339].

71 Stevens, “Disaster”, p. 597. Stevens asks what the use value transferred is on day one. But what is transferred is the continuing use value (the continuing opportunity to use) which increases according to the period of use in question. One day's use is likely to be minimal. One year's use is likely to be significant.

72 A few days after the judgment was handed down, the electronic version of para. 71 of the judgment on the Supreme Court website was amended to read: “This analysis has a number of questionable features (discussed in Stevens, R., “The Unjust Enrichment Disaster” (2018) 134 L.Q.R.Google Scholar (October issue, forthcoming)), which can be illustrated by an example” (emphasis added). But that amendment does not appear in the electronic version of the judgment now on the Supreme Court website, nor has it made it through to the official reports.

73 Prudential Assurance Ltd. [2018] UKSC 39, [2018] 3 W.L.R. 652, at [55]–[67]. As was made clear, at [44], nothing said in the Supreme Court casts doubt on the obiter dicta in Sempra Metals supporting the availability of compensatory damages (as opposed to restitution) for (compound) interest. For an excellent case-note on Prudential, see Wilmot-Smith, F., “A Prudent Decision” (2019) 135 L.Q.R. 195Google Scholar.

74 Kleinwort Benson Ltd. v Lincoln City Council [1999] 2 A.C. 349.

75 It should also be noted that: (1) there was no express reference by the Supreme Court to the 1966 Practice Statement [1966] 1 W.L.R. 1234 (albeit that the Supreme Court's judgment refers to reasons that would meet the vague threshold for the highest court overruling itself laid down in that Practice Statement); (2) the panel of five was not enlarged as is the usual convention if the highest court is being asked to overrule itself. As regards (1), see also Wilmot-Smith, “A Prudent Decision”, pp. 199–200.

76 I should make clear that I was one of the counsel acting on behalf of HMRC.

77 It is important to appreciate that in Sempra Metals itself, there was no possibility of a claim for (simple) interest under s. 35A of the Senior Courts Act 1981 because that section requires there to have been a claim for “a debt or damages” to which the interest can be added. The restitutionary award of compound interest sought was for the tax payment being premature so that the interest sought was itself the principal sum/debt. In that sense, there was no clash between the award of compound interest and s. 35A. For discussion, see Burrows, Law of Restitution, pp. 24–25.

78 Sempra Metals Ltd. [2007] UKHL 34, [2008] 1 A.C. 561, at [102].

79 For similar criticism, see C. Mitchell, “End of the Road for the Overpaid Tax Litigation?” (2019) Supreme Court Year Book 1, 16–17. Moreover, one has to recognise that there is now a significant lacuna in our law on interest. This is because if a mistaken payment is returned before any proceedings have been commenced for restitution there is no right even to simple interest under the Senior Courts Act 1981, s. 35A. So if C mistakenly pays D £50,000 and this is discovered after a year and D, without C commencing any legal action, pays back £50,000 to C, C has no legal entitlement to one year's interest, simple or compound, whether at common law or under statute. Cf. the judgment of the Supreme Court in Prudential Assurance Ltd. [2018] UKSC 39, [2018] 3 W.L.R. 652, at [78].

80 Note that if title had not passed and C is given specific restitution (or compensatory damages for the value of the car) in the tort of conversion, C would presumably also be entitled to compensatory damages for the loss of use of the car for that year.

81 This is not to deny that there may have been other good reasons as set out by the Supreme Court (see the text to note 73 above) – such as the policy behind statutes or the desire not to undermine public finances – that did justify overruling Sempra Metals.

82 This was referred to in the case as the category (c) interest; and it was the argument of principle set out in this paragraph, and directed to the category (c) interest, that was put forward to the Supreme Court by HMRC in the Prudential case. This argument of principle has no bearing on the award of interest in Sempra Metals.

83 Let us assume, to continue with the above example of the £1,000 tax payment, that up to the time of the repayment, or offsetting against lawful tax, of the £1,000 tax payment, the value of the opportunity to use that tax payment – the interest (including compound interest) – amounts to £200. Applying Sempra Metals, C would be entitled as restitution of an unjust enrichment to £200. But C would not be entitled to further interest on the £200 because that further interest does not constitute the conferral (or transfer) of an opportunity to use the £1,000 tax payment. It is simply a gain from not paying back the £200.

84 It is precisely because the “at the expense of” requirement is not satisfied that no-one would seriously suggest that there is a claim by a creditor for restitution of unjust enrichment to remove a gain that a debtor has made by not paying its contractual debt on time (even if there were an unjust factor such as that the creditor has mistakenly failed to enforce the debt on time).

85 For a full treatment, see, e.g., Burrows, Law of Restitution, Parts 1–3.

86 Cf. Stevens, “Disaster”, pp. 577–78; Wilmot-Smith, F., “Should the Payee Pay?” (2017) 37 O.J.L.S. 844Google Scholar (denying that there is any convincing reason for requiring restitution of a mistaken payment but one might say that his arguments fail to give due weight to the combination of the payor's impaired consent and the payee's enrichment at the payor's expense).

87 Birks, An Introduction, ch. IX.

88 See note 26 above.

89 See e.g. Burrows, Law of Restitution, ch. 17.

90 See above text to note 68.

91 See Burrows, Law of Restitution, ch. 16 (but note that that chapter arguably confused matters by starting with two-party hypothetical conferral cases and two-party cases where D was a tortfeasor).

92 Lipkin Gorman [1991] 2 A.C. 548. That was a decision at common law. The equivalent in equity may be said to be Ministry of Health v Simpson (Re Diplock) [1951] A.C. 521.

93 See Burrows, Law of Restitution, ch. 8.

94 In Investment Trust Companies v HMRC [2017] UKSC 29, [2018] A.C. 275, Lord Reed set out various three-party situations which might be classed as exceptions to the “direct conferral” general rule. These included agency and tracing but also, and most difficult of all, “co-ordinated transactions” which he relied on to explain the subrogation cases of Banque Financiere de la Cite v Parc (Battersea) Ltd. [1999] A.C. 221 and Meneloau v Bank of Cyprus UK Ltd. [2015] UKSC 66, [2016] A.C. 176.

95 Or by a “bright-line” rule that there can never be restitution where the benefit was legally owed (under a valid contractual or statutory obligation) by the claimant to the defendant. See note 42 above.