Hostname: page-component-5db58dd55d-smskv Total loading time: 0 Render date: 2026-06-03T05:30:09.677Z Has data issue: false hasContentIssue false
  • English
  • Français

COUNTERFACTUALS IN UNJUST ENRICHMENT

Published online by Cambridge University Press:  14 July 2025

Francis Cardell-Oliver
Francis Cardell-Oliver*
Affiliation:
Barrister, Essex Court Chambers, London.
*
Address for Correspondence: Essex Court Chambers, 24 Lincoln’s Inn Fields, London, WC2A 3EG, UK. Email: fcardell-oliver@essexcourt.net.

Abstract

This article is concerned with the question whether a defendant in an unjust enrichment action can reduce or eliminate its liability by establishing that it could have obtained the enrichment (or part of it) from the claimant in a way that would not have given rise to liability. The answer in principle ought to be no. In arguing for that conclusion, I consider the meaning of “enrichment” and “loss”, the nature of the change of position defence and the basis of liability for unjust enrichment in cases involving ultra vires charges by public authorities and the taking of money without consent by private defendants.

Keywords

restitutionunjust enrichmentcounterfactualspublic authoritieslack of consent

Information

Type
Articles
Information
The Cambridge Law Journal , Volume 84 , Issue 2 , July 2025 , pp. 253 - 278
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2025. Published by Cambridge University Press on behalf of The Faculty of Law, University of Cambridge

I. Introduction

The existence of a body of English private law concerned with reversing unjust enrichment is now well established as a matter of stare decisis.Footnote 1 That body of law, however, remains the subject of intense academic controversy. There is debate as to its existence as a unified subject and as to large parts of its content and boundaries. For some decades, the leading scholars focused their energy on identifying what Peter Birks called a “stable set of large questions”Footnote 2 by drawing out unifying principles from the wide range of cases in which restitution had been awarded but which could not be explained by the presence of a contract or a wrong. The questions on which Birks settled (since endorsed by the courts) were: (1) was the defendant enriched; (2) was the enrichment at the claimant’s expense; (3) was the enrichment unjust; and (4) are there any defences?Footnote 3 These questions are, perhaps necessarily, expressed at a high level of abstraction. There were always sceptics who doubted their coherence and utility.Footnote 4 In recent years, criticism of the Birksian analysis has been amplified by the important work of Robert Stevens and Lionel Smith, who argue that Birks’s four questions are overgeneralisations and that what is often regarded as the law of unjust enrichment should be disaggregated into several distinct areas of law, perhaps with different names.Footnote 5

This article addresses one relatively underexplored corner of this wider controversy: when, if ever, are counterfactuals relevant to a claim in unjust enrichment? A counterfactual is (literally) a proposition contrary to fact. By themselves such propositions are nonsensical but they become more interesting when expressed in conditional form: if X had occurred then Y would or could have happened (where X did not in fact occur). Counterfactual propositions of this conditional kind are most familiar to lawyers from the law of wrongs (i.e. torts and breaches of contractual and equitable duties). In that context we routinely ask what would have happened if (hypothetically) the defendant had acted with reasonable care, had performed their contractual duty, etc. Counterfactual conditionals might, however, also be relevant in unjust enrichment cases. In the following section (Section II) I outline three situations in which counterfactual arguments have been made (with varying degrees of success) in unjust enrichment cases. At a high level of abstraction, the counterfactual arguments made in the cases take a common form: the defendant has obtained something from the claimant (typically a payment of money) in circumstances which are recognised as giving rise to a right to restitution, but the defendant could have obtained the same or a lesser amount of money in a way that would not have given rise to a duty to make restitution. The question I am interested in is: whatever the state of the authorities, should the defendant be able to deploy counterfactuals of this kind in order to reduce or eliminate what would otherwise be their liability to make restitution? In other words, why, if at all, should counterfactuals have normative significance in the context of an unjust enrichment claim?

Logically there are three possible answers: (1) counterfactuals might negative an element of the claimant’s claim (in other words they might be a denial); (2) they might engage a recognised defence (the only plausible candidate, as we will see, is change of position); or (3) there might be additional reasons why, in justice, the courts should recognise a new defence in which counterfactuals are an element. Commentators and courts who have considered this issue have seen numerous ways of fitting counterfactuals into each of these categories.Footnote 6 This is, perhaps, unsurprising if one starts with the four abstract questions posed by Birks. The four enquiries are sufficiently protean that counterfactuals might fit into any or all of them. The requirement of an “enrichment” at the claimant’s “expense” might require the claimant to show that the defendant is better off, or the claimant worse off, than they would have been had the payment not been made. Even if there is an enrichment at the claimant’s expense, perhaps counterfactuals might go to whether it is just to require the defendant to return that enrichment or might provide additional reasons defeating a prima facie right to restitution (i.e. a defence).

Sections III–V below work through the three logical possibilities: denial, recognised defence and novel defence, in that order. My central thesis is negative: whichever category one turns to, there is no persuasive reason for attributing normative force to counterfactual propositions about what the defendant could have done differently. That negative thesis is important in itself. Its corollary is that several decisions of the Court of Appeal, House of Lords and Privy Council are either wrongly decidedFootnote 7 or wrongly reasoned.Footnote 8 The argument that leads to it also has broader implications, however, because it sheds light on each of the Birksian questions: on the meaning of “enrichment”, “at the expense of” and “unjust”, and on the nature of change of position. In particular, I defend the following propositions:

  1. (1) Where a claimant pays a defendant money, we should not ask whether the defendant is better off than they would have been had they not received the money. The subject matter of the claim – the “enrichment” to be reversed – is the payment, not the difference it made to the defendant’s life;

  2. (2) Similarly, the requirement that the payment be at the claimant’s “expense” does not require us to ask if the claimant is worse off than they would have been had they not made the payment;

  3. (3) The principle of subjective devaluation, if it should continue to be recognised at all, should be limited to cases where A confers an unsolicited, non-money benefit on B. An examination of the basic rationale for the principle shows that that rationale cannot sensibly apply to cases involving payments of money;

  4. (4) At least in the three situations I deal with, what makes the enrichment “unjust” is the fact that the money was, on the facts as they actually occurred, obtained ultra vires or without the claimant’s consent. The fact that the defendant could have obtained money from the claimant in a way that would not have given rise to a duty to make restitution does not negative that injustice;

  5. (5) Although the change of position defence does involve some counterfactual analysis, the counterfactuals involved are both more various and more nuanced than is often assumed. There are at least two different forms of the defence governed by different rules and involving different counterfactuals which are employed for different reasons. Neither form of the change of position defence invites an enquiry into whether the defendant could have obtained the same enrichment lawfully.

II. Three Examples

The balance of my argument focuses on three situations, taken from the case law, in which counterfactual arguments have been advanced by a defendant in an attempt to reduce or eliminate their liability to make restitution of an unjust enrichment.

Example 1: A public authority (D), empowered by Parliament to impose some form of compulsory charge (a tax, rate, fee, levy, etc.), errs in the exercise of that power. D’s decision is quashed. In the meantime, taxpayer C has paid up. Had D acted lawfully, it could have imposed a charge of a lower amount.

On these facts, C has a right to restitution on the ground recognised in Woolwich Equitable Building Society v Inland Revenue Commissioners.Footnote 9 But what is the quantum of their claim? Should C recover (1) the whole of the amount it paid; or (2) only the difference between the amount it paid and the amount it would have had to pay if D had acted lawfully?

There is a line of authority supporting option (2).Footnote 10 The point was first addressed in detail in R. (Hemming (t/a Simply Pleasure Ltd.) and others) v Westminster City Council,Footnote 11 which involved annual licence fees for sex shops in the City of Westminster. The fees were to be determined annually by the respondent council but it made various public law errors with the result that its determinations going back several years were void. It was ordered to redetermine lawful fees for each of those years retrospectively and then refund to the claimants the difference between the lawful fees and the amounts actually charged. The result, therefore, can be readily explained without reference to counterfactuals. Nonetheless, in the course of its reasoning the Court of Appeal held that, irrespective of whether it had determined a lawful fee retrospectively, the council was only ever unjustly enriched by the difference between what was paid and what might have been demanded if the power had been exercised lawfully.Footnote 12 This was said to recognise the “economic reality” of the situation.Footnote 13

There is some intuitive attraction to this: why should the fact that D has made what might be a minor and honest mistake of law mean that C makes no contribution at all to public revenues and gets its licence over many years for free? However, the reasoning leads to a conundrum: once one starts looking for counterfactual lawful ways of recovering money, where does the hypothesising stop? Could a public authority hypothesise new legislation conferring new and different powers on it? The Court of Appeal had to address that issue in Vodafone Ltd. v Office of Communications,Footnote 14 in which it was argued (at least at first instance) that there was no limit, in principle, to the counterfactual hypotheses available: even changes to primary legislation could be hypothesised if the evidence supported such an argument.Footnote 15 The Court of Appeal rejected that suggestion, on the basis that a counterfactual analysis of what the defendant would have done was simply irrelevant.Footnote 16 Nonetheless, in deference to Hemming and some prior authorities (none of which the Court of Appeal was in a position to overrule),Footnote 17 the court recognised that it was appropriate to deduct the amount that could lawfully have been charged under existing legislation.Footnote 18 The defendant’s counterfactual in Vodafone involved the making of new regulations, that is, new secondary legislation, and was therefore impermissible.Footnote 19

Where this appears to leave the law is that it is now permissible to look at counterfactual things the defendant would have done so long as that does not involve positing hypothetical legislation (including regulations). Despite the Court of Appeal’s assertion to the contrary, this does not avoid counterfactuals but simply places a limit on the kinds of counterfactual that can be posed. This limit, however, seems arbitrary: if counterfactuals are relevant, why should the fact that the power must be exercised by making regulations, as opposed to promulgating a non-legislative instrument, make all the difference?Footnote 20 The better answer, as I argue below and as the Court of Appeal may well have sensed in Vodafone, is that counterfactuals should not be relevant at all.

Example 2: C, a retailer, supplies goods to end consumers. The revenue (D) charges C £100 in VAT (at a rate of 20 per cent) on the supply of those goods. Had VAT not been charged, C would have charged its customers the same price and would therefore have increased its profits by 20 per cent. The increased profits would in turn have led to C having £25 worth of additional income tax liability. It turns out that D had no power to charge VAT on the relevant supplies.

Again, C has a claim against D for restitution of £100 on the ground recognised in the Woolwich case. The question is: can D reduce its liability by setting off the £25 C would have had to pay in income tax but for the VAT liability? An argument to that effect was run in Littlewoods Retail Ltd. and others v Revenue and Customs Commissioners.Footnote 21 As Henderson J. recognised, there is some “initial attraction” to the argument:Footnote 22 why should C be placed in a tax position it could never have occupied in real life, whereby it has effectively received part of its profits tax free? However, Henderson J. ultimately held, albeit in obiter, that such a counterfactual argument is not available in English law.Footnote 23 For present purposes, three aspects of the reasoning bear emphasis. First, Henderson J. felt an intuitive unease with asking, in the context of an unjust enrichment claim, whether the revenue was benefited or the taxpayer had suffered a loss, relative to a counterfactual world in which the payments on account of VAT had not been made.Footnote 24 Second, he was open to the possibility that a change of position defence might be available.Footnote 25 Third, however, he gave short shrift to a suggestion that the counterfactual argument offended constitutional principle because it effectively allowed the revenue to levy tax in circumstances where doing so was not authorised by Parliament.Footnote 26 As we shall see, the intuitions underlying the first and second points were well founded, but the third point dismisses too readily the public law issues that face counterfactual arguments in taxation cases.

Example 3: D, a director of company C, makes unauthorised payments to himself out of corporate funds. Had D asked, C’s board would probably have authorised the payments, which merely reflected the going market rate for directors of D’s calibre. Alternatively, had the board refused, D would have resigned and the board would have had to authorise payments of the same amount to secure a replacement director on the market.

Can D avoid an obligation to repay the amounts he has misappropriated by appealing to either of these counterfactual propositions about things which could have happened, but did not? An affirmative answer was urged by the defendant director (Mr. Ferster) in Interactive Technology Corporation Ltd. v Ferster.Footnote 27 The analysis here is complicated by the fact that the company had three possible claims: (1) a claim that Mr. Ferster compensate it for its losses suffered by reason of his breach of fiduciary duty; (2) a claim for an account of profits, requiring Mr. Ferster to disgorge the gains he had obtained from misuse of his fiduciary position; and (3) a claim in unjust enrichment for restitution of the amounts paid over simply because they were paid without the company’s authority.Footnote 28 The Court of Appeal held, in obiter, that it was “difficult” to see how Mr. Ferster’s counterfactual argument “could prevail as an answer to a claim … for payment of sums dishonestly taken as ‘remuneration’ without authority”.Footnote 29 This appears to be a reference to claims of both type (2) and type (3) above. If so, the Court of Appeal’s first impression was (at least in relation to a claim of type (3)) correct.Footnote 30

III. Counterfactuals and the Elements of an Unjust Enrichment Claim

As I have indicated already, there are, logically, three ways in which counterfactual arguments of the kind discussed above could take on normative significance: as a denial of an element of the claimant’s claim (enrichment, “at the expense of”, unjust factor); as an element in a recognised defence; or as an element of a new defence. This and the following two sections work through those options in that order.

A. Enrichment at the Claimant’s Expense

At first glance the answer to the first two Birksian enquiries (is the claimant enriched? Is the enrichment at the defendant’s expense?) is straightforward. The cases we are concerned with all involve a payment directly from C to D. This is often regarded as the paradigm example of an enrichment of D at C’s expense.Footnote 31 There are, however, three possible bases on which D might seek to deny part or all of the apparent enrichment at C’s expense.

First, several leading commentators and at least two members of the Supreme Court have suggested that the story does not end with the payment. We have to go on to ask if, and to what extent, the payment has improved D’s net position compared to a world in which they did not receive the payment.Footnote 32 For example: A pays B £100 and C (seeing B’s new-found wealth) abstains from making B a gift of £25 they would otherwise have made. Is B “enriched” by £100 or by £75? This line of thinking cannot explain Examples 1 or 3. But for the payments there, D would simply have been poorer by the amount of the payment. There is no reason to think that the mere fact of non-payment by itself would have inspired the public authority to remake its decision lawfully or the director to seek board approval. However, it might explain Example 2: but for the payment of £100 on account of VAT, D would have received from C £25 by way of income tax, so the net improvement in D’s position is only £75, not £100.Footnote 33

Second, for an enrichment to be at C’s expense, C must suffer some form of “subtraction” or “loss” as a result of making the payment.Footnote 34 Again, taking that requirement at face value, this “loss” could be measured against a counterfactual baseline that might help explain Example 2: but for the £100 payment C would still have made a £25 payment, so their net loss is only £75.Footnote 35

Third, one might appeal to the principle of subjective devaluation. A defendant is entitled to reduce the measure of their enrichment by showing that they placed a lesser value on the thing received than its objective market value.Footnote 36 Whether one can subjectively devalue money is controversial,Footnote 37 but arguably D in our examples, now knowing that they could have obtained some or all of the money lawfully, might be entitled to say “I didn’t want or value C’s money” (at least, not in the circumstances in which it was paid).Footnote 38

None of these possible analyses is right. Each depends on an overbroad understanding of the concepts of enrichment, loss and subjective devaluation.

1. No enrichment

As to the first, the conclusion in the example of the counterfactual gift is right but the explanation is wrong. As we shall see, the correct explanation of what might be called “prevention of benefit” cases is that they may (in some circumstances) give rise to a defence (a form of change of position). Using counterfactuals to identify and measure “enrichment”, on the other hand, leads to untenable results.

The point may be tested, without raising the potential change of position defence, by considering cases where the difference between D’s position in real life and in a world where they had never received the payment is a positive value that exceeds the amount of the payment itself: for example, D receives £100 and uses it to make an investment they would not otherwise have made, doubling their money. The measure of restitution should remain £100, not £200. If the defendant has acted wrongfully (e.g. by exceeding public law limits on its powers or taking C’s money without C’s consent, in our examples) that might be thought to provide a reason for stripping them of their gain and placing them in the position that they would have been in but for the payment. But restitution, like all private law remedies, is bilateral. We need to explain both why the defendant should be stripped of their gain and why it should be re-allocated to the claimant.Footnote 39 Cases where reallocation of D’s gains to C is justified are rare. Where, for instance, D has breached a duty of loyalty owed to C, there is a plausible argument that giving full effect to that duty requires the reallocation of D’s profits to C.Footnote 40 Likewise where D has breached a monopoly right of C’s, such as a patent.Footnote 41 But no such rationale applies in Examples 1 and 2. In Example 3 there is, of course, a breach of a duty of loyalty and C has a parallel, alternative claim to an account of the profits D has made from that breach. The unjust enrichment claim, however, is (for reasons discussed more fully below) simply premised on the fact that the payment was not authorised, irrespective of D’s status as a fiduciary. The fact that there was an unauthorised payment, in and of itself, does not justify an order that D repay to C anything more than the amount of the payment. The different claims focus on different aspects of the same facts and give rise to different remedies.

2. No loss

Much, I suspect, of the intuitive attraction of the counterfactual arguments in our examples arises from the sense that, if C gets back the full payment, they are making an unjustifiable windfall. In Example 2, for instance, it was inevitable that C would have had to pay one or the other tax and it seems unrealistic to award them a remedy that places them in a more advantageous tax position than they could ever have occupied in real life. This line of thinking, however, confuses matters by introducing a counterfactual concept of loss, familiar from the law of wrongs, into a context where it is inapposite.Footnote 42

The law of wrongs is concerned with primary duties: duties to keep promises, not to disclose secrets, to drive one’s car reasonably carefully, etc. When we breach such a duty, then – for the same reasons we have the primary duty in the first place – we come under a secondary duty to do the next best thing, to make the world as close as possible to how it would have been if we had not breached the primary duty.Footnote 43 Any theory of remedies for wrongs thus requires (1) a counterfactual enquiry which tells us what the world would have been like but for the breach (and hence in what respects it is now different); and (2) some conception of the purpose of the primary duty (which tells us how best to make up for that difference). (To the latter, English law largely gives a very simple answer: it proceeds on the assumption that the purpose of almost all primary duties is to improve, or not to worsen, the financial position of the person to whom the duty is owed and hence that almost all breaches can be remedied by paying money.Footnote 44)

Unjust enrichment is doing something very different. There are wrongs in our three examples (that is, breaches of duties, whether imposed by private or public law), but the object of the unjust enrichment claim is not to make good or cancel out all the consequences of those wrongs. (There is a parallel claim with that objective in Example 3, though not Examples 1–2.Footnote 45) The best understanding of the reason for restitution – the “unjust factor” – in these cases is that both public authorities and private defendants are not permitted to take money from others unless they have their consent or (in the case of a public authority) they have statutory authority to do soFootnote 46 and have made any necessary administrative decisions in accordance with the administrative law requirements reflected in the grounds of judicial review (which are essentially all aspects of the rule of law).Footnote 47 If a payment is made and those conditions are not met, the payment should be reversed. Questions about the consequences of the payment – whether it made anyone worse or better off and in what sense – are irrelevant.

The only way that a counterfactual could become relevant to this kind of claim would be if one could show that the exceptions to the prohibition on the revenue or private defendants taking money applied not just in cases of actual authority or consent but also in every case where (and to the extent that) authority or consent could, hypothetically, have existed or been obtained but did or was not. I defer discussion of that possibility to the section below on unjust factors. Suffice it to say for now that this is not a position that the law should (nor, for the most part, does) adopt.

3. Subjective devaluation

Subjective devaluation is about personal autonomy. If C performs services for or supplies goods to D without D requesting or accepting them and then claims their market value, there is an obvious risk of the law riding roughshod over D’s right to make decisions about how to spend their money by effectively forcing them to buy C’s goods or services at market rates. Two solutions are possible: either one incorporates D’s choice into the preconditions of liability (they must actually have requested or accepted the benefit),Footnote 48 or you need a liability-limiting device, which asks what (if anything) D would have chosen to pay for the goods/services if they had been given the choice.Footnote 49 D can always say “I didn’t want it” (or at least not at the price now being asked). English law currently follows the latter approach.Footnote 50

Subjective devaluation is therefore a solution to a problem created by a broad understanding of the criteria for liability in unjust enrichment. D’s complaint, at its highest, in our three examples is very different. It is not that D is being forced by an external actor (C, calling in aid the coercive powers of the courts) to spend money on non-money benefits it did not choose, but rather that it made certain decisions on bad information. It chose to obtain C’s money via an unlawful route when it could (and would, had it realised its error at the time) have chosen a lawful route which would have allowed it to keep at least some of the money. (Of course, this issue does not arise if D realised all along that what they were doing was unlawful.) Whether that raises a problem of autonomy at all is controversial.Footnote 51 If it does, it is a problem of a very different kind from that addressed by subjective devaluation. It is not a problem created by the liability criteria for unjust enrichment. Nor is it C’s fault. Merely observing that there is a potential autonomy problem does not (without more) explain why the law should save D from it by denying C a remedy.

B. The Unjust Factor

1. Causation

The ground for restitution in Examples 1 and 2 has its origins in the Woolwich case but finds its modern formulation in the Supreme Court’s first judgment in the Franked Investment Income (FII) group litigation. There must be a payment, to a public authority, “in response to (and sufficiently causally connected with) an apparent statutory requirement to pay tax which … is not lawfully due”.Footnote 52 This appears to open the door to a “but for” test of causation which might be thought to explain these examples: “but for” the unlawful charge, there would have been a lawful charge, in which case the unlawful charge did not cause the payment (at least to the extent of the lawful charge).Footnote 53

This might appear, on its face, to be a straightforward way to fit counterfactual reasoning into the elements of a Woolwich claim. This sort of causal analysis is, however, fallacious because it overlooks the problem of pre-emption. The only cause of the actual payment was that the taxpayer was responding to the unlawful decision (in Example 1) or the unlawful purported obligation to pay VAT (in Example 2). They plainly were not responding to a hypothetical lawful decision that had not been made or paying tax on taxable income that they never earned. Those hypothetical events could have resulted in the same outcome but they did not because they were pre-empted by the actual cause (the unlawful charges).Footnote 54 Whether counterfactual theories of causation can cope with cases of pre-emption like this is a controversial question,Footnote 55 but if they cannot that is a problem for counterfactual theories (and a good reason not to use counterfactuals as a heuristic for causation in such cases), not a good reason to doubt that it was the real charge that led to the real payment.

2. The reason for restitution

I have suggested already that the reason D in each of our examples should be obliged to repay the money received from C is that there is a general prohibition on the revenue charging taxes (or other compulsory exactions) without lawful authority and on private defendants taking money from others without their consent. Absent consent or authorisation, there is therefore a good reason for requiring any payment made to be returned.

That conclusion does not necessarily depend upon acceptance of the much broader thesis that all payments which are made in the absence of a legally recognised “basis” (e.g. with the consent of the payer or in discharge of an obligation to pay) should be returned.Footnote 56 The absence of basis thesis would be critical to the outcome in a case where, for example, X pays what Y alleges to be a debt owed, but the debt is not (objectively) due and X believes that it is probably not due. Is the fact that the payment was not due enough to justify restitution or need X show something more (e.g. that they paid under a mistake, which on these facts they cannot)?Footnote 57 Even if the broader thesis is wrong and X is not entitled to restitution, there are sound reasons of policy for singling out those who take money without authority or consent for different treatment. In the case of public authorities, an exaction which is not backed by legislation or which is the result of an invalid administrative decision is not just not legally obligatory: it is offensive to the separation of the powersFootnote 58 and the rule of law,Footnote 59 respectively. In a case where one private party takes another’s money without consent, the payment is not simply unjustified, it involves a deliberate interference by the first party with the second’s property.

In each case, the reason for restitution focuses on what actually happened in the real world. A counterfactual might become relevant if we thought that restitution ceased to be justified not just in cases of actual authority or consent but also in cases where the public authority could have got to the same outcome if it had exercised a different power (Example 2) or exercised the same power lawfully (Example 1) or where the board would have authorised the payments if asked (Example 3). There is some intuitive appeal in such an approach. There is arguably a difference between a case where a public authority demands money with no plausible statutory basis at all and one in which there is a power to charge and all the authority has done is to rely on the wrong power (Example 2) or make an error in the exercise of an existing, valid power (Example 1, e.g. erring on a difficult question of statutory interpretation).Footnote 60 Perhaps the first scenario is enough to justify restitution but the second is not. Likewise, in Example 3 the gist of the director’s conduct might be characterised as “failing to ask first”, which is arguably different from procuring payments to which the board would never have consented. On closer examination, however, neither argument is persuasive.

3. Public authorities

In order to address this line of thinking as applied to Examples 1 and 2, it is necessary to turn to public law, specifically administrative law, which lies at the heart of the Woolwich ground. The attractions and difficulties of counterfactual analysis in the context of administrative decisions are familiar to public lawyers. This is therefore a powerful example of an area where scholars and judges approaching unjust enrichment from a private law perspective can avoid reinventing the wheel by drawing on the significant body of pre-existing public law learning.Footnote 61

Some rules of administrative law prohibit outcomes: for instance, a decision maker can never achieve an outcome which infringes certain protected human rights,Footnote 62 or which is so irrational that it is outside the range of decisions any reasonable decision maker could have made.Footnote 63 Most, however, are concerned with the process by which administrative decisions are made (in which term I include the reasoning for a decision). A decision may be quashed because the process actually used was improper even if it can be shown that a proper process could and would have led to the same outcome. The clearest example of this is apprehended bias: the substantive quality of the decision-making (and hence the correctness of the outcome) is never in question, but the impropriety of the process nonetheless ordinarily requires that the decision be quashed.Footnote 64

There are good reasons for thinking that public authorities should not be able to exert power over citizens (including exacting payments of money) without complying with proper process, even if the same outcome could have been achieved lawfully. Administrative law, in other words, serves ends other than simply keeping decision-makers within the universe of possible lawful outcomes. Even within that universe, proper process ensures that decision-makers are, and are seen to be, applying the law (rather than engaging in arbitrary exercises of raw political power) and are doing so accurately, impartially, rationally, consistently and fairly. Those are all aspects of the rule of law.Footnote 65 Compliance with the rule of law allows citizens to plan their lives within a predictable framework of rules; it fosters confidence in and respect for official decision-making even when the outcomes are unpalatable to those affected by them; and it respects the status of citizens as autonomous agents who are entitled to participate in the making of, and at least to try to influence the outcome of, decisions affecting their lives.Footnote 66

Example 2 involves a more unusual example of a procedural error: the revenue has relied on a non-existent power to charge VAT when it could have allowed the amounts collected qua VAT to accrue as income and then taxed them as such. I described this earlier as relying on the wrong taxing power, but the problem is not merely one of nomenclature, of the sort that might arise if I am served with a parking ticket quoting the wrong by-law. The error lies not in identifying the wrong source for the right power, but in purporting to exercise a qualitatively different (and non-existent) power. That too raises a rule of law problem: the ultimate economic outcome may be the same, but the rule of law requires that it be reached via a lawful route rather than by an extralegal one.

The rule of law is not an absolute, nor is it the only virtue at which the law should aim.Footnote 67 Trade-offs may be appropriate. One possible such trade-off is the rule of administrative law now reflected in section 31(2A) of the Senior Courts Act 1981: the High Court must refuse to grant relief on an application for judicial review (including awarding restitution)Footnote 68 if it appears to the court to be highly likely that the outcome would have been substantially the same if the conduct complained of had not occurred (that is, had the decision maker adhered to a lawful rather than an unlawful process). Section 31(2A) may apply to a Woolwich claim – effectively duplicating the result currently required by the common law – where (1) the applicant seeks restitution in the course of or following an application for judicial review; and (2) the counterfactual lawful charge would have been substantially the same as the actual charge. If the rationale for section 31(2A) were sound, the same rationale might justify reducing the quantum of restitution in cases where a lawful process would have led to a lower charge substantially different from the actual charge.

It is not easy to pin down just what that rationale is. A few candidates are apparent, but none of them justifies a counterfactual approach to unjust enrichment.

One possible rationale for a counterfactual test in public law is that it shows that the individual concerned has not been made any worse off by the decision to act unlawfully rather than lawfully. This might conceivably be relevant if one thought (as some do) that at least one purpose of at least some administrative law duties is to benefit individuals or to avoid making them worse off.Footnote 69 At their highest, however, such arguments simply seek to add to the list of reasons for enforcing the grounds of review. The absence of counterfactual worse-off-ness is thus a neutral factor: it negates one factor in favour of quashing some procedurally flawed decisions, but one still needs additional reasons to overcome the broader rule of law considerations summarised above.

A second justification for the section 31(2A) test (and possibly the argument that motivated its enactment)Footnote 70 really trades on a predictive rather than a counterfactual proposition. The consequence of quashing an administrative decision is often simply that the matter is remitted to the original decision maker to be re-decided. If everyone can see that the outcome is highly likely to be the same, then there is an argument that the rule of law values vindicated by affording the decision no legal effect are outweighed by the broader negative consequences of inefficiency and delay (e.g. remaking the decision uses up limited public funds, delaying planning approvals deters investment, etc.).Footnote 71 This rationale does not apply if the decision cannot be remade. For instance, in Example 1 the authority may have no power to determine a charge retrospectively, nunc pro tunc.Footnote 72 In Example 2, the revenue cannot go back and change the historical fact that the VAT payments did not form part of the taxpayer’s taxable income. It is in precisely these cases that the counterfactual argument is critical. In other cases, such as Hemming, where the charge can be remade, there may be some limited inefficiency if the defendant is obliged to repay the amounts charged in full before recharging a portion of that amount, but as Hemming itself illustrates, this issue can be ameliorated by framing the relief appropriately: it is not an argument for any broader form of counterfactual analysis.

A third possible justification can be illustrated with an example. Suppose Parliament has enacted legislation which contemplates a “user pays” scheme of regulation or services: the scheme contemplates that all licensees or recipients of services (collectively, “users”) will contribute to the costs of the scheme, though the level of contribution and how it is distributed among different users is left to the discretion of the public authority administering the scheme. If the public authority uses the wrong procedure to determine the annual charge for the first three years of the scheme and cannot now remake those charges, the effect of awarding restitution of the full amount of the unlawful charges is to exempt users in the first three years from making any financial contribution to the scheme. The nature of public finances means that that loss of revenue can only be made up by reducing the services provided, levying extra revenue on future users or obtaining money from some other source (e.g. additional central government funding or borrowing, either of which will also ultimately be paid for by some other taxpayer). Arguably this is distributively unjust and the distributive injustice outweighs the rule of law considerations in favour of treating the decision as wholly ineffective and awarding restitution of the whole amount paid over.

The example, though intuitively attractive, rests on two bad premises. The first is that cases in which there is a counterfactual lawful route to the same outcome necessarily involve less serious violations of the rule of law than those where there is not. That cannot be correct. Compare (1) a case where a public authority makes a decision the outcome of which could have been reached lawfully, but the decision is in fact attended by a series of outrageous procedural defects (actual bias, bad faith, improper purposes, etc.); and (2) a case where a public authority takes a view, reasonably open to it, of the meaning of its empowering statute, but a court disagrees and finds that it had no power to act as it did (and there is no alternative source of power that might have supported the same conduct). Whether case (2) involves any violation of the rule of law at all is controversial,Footnote 73 but at the very least, to suggest that cases of this kind are always, necessarily more offensive to the rule of law than the most extreme case of type (1) is to overstate the position.

The second premise which underpins the distributive justice argument set out above is that, within the class of cases where there is a counterfactual lawful route to the same outcome, concerns regarding distributive injustice should always trump rule of law concerns. Such an absolute proposition cannot be right because we need to know the extent of the distributive injustice before reaching a conclusion. We regularly accept that at least some level of diversion of public funds from service-provision to fighting litigation, remaking decisions and making financial reparation to individuals, is acceptable. If matters were otherwise, the Government could never be sued or compelled by a court to do anything with any level of cost attached to it.Footnote 74 The argument must, in truth, be that at some point the amount of redistribution becomes unjust.

That leaves open the possibility that the courts might be entitled to refuse to order restitution by public authorities in cases involving a sufficiently large quantum. This was, at one stage, an approach endorsed by the Canadian courts.Footnote 75 There are good reasons for doubting its correctness: such an approach would require the courts to assess questions of distributive justice which are essentially political in nature and ill-suited for judicial determination.Footnote 76 In any event, and critically for my purposes, the argument has by this stage ceased to be an argument for counterfactual analysis at all. Once one accepts that (1) cases where the counterfactual argument is open might involve more serious violations of the rule of law than cases where it is not; and (2) the degree of distributive injustice might be greater in a case where a counterfactual analysis is not open than in a case where it is, then the fact that a counterfactual lawful route to the same outcome was open, by itself, tells you nothing. What is really needed is a direct focus on the underlying competing factors: the rule of law and distributive justice. Counterfactual analysis tells you nothing about either.

4. Private defendants

Example 3 only arises in circumstances where (1) at some point some iteration of the company’s board could have been persuaded to authorise, or retrospectively to ratify, the unauthorised payments; but (2) when faced in reality with a choice between ratifying or suing to recover the payments, the board has chosen the latter course. It is worth noting the logical limits of the argument. The counterfactual possibility of consent need not necessarily exist at the time the payments are made: the argument “I could have obtained these payments lawfully” holds good provided that there is any point in history at which the then board could have been persuaded to reach a binding agreement to make payments in the future or to ratify payments already made.

The rule that money or other assets, taken without consent, are recoverable by their original owner by (inter alia) a claim for unjust enrichment, is a basic corollary of the original owner’s right to exercise autonomy over the disposition of their own assets. Autonomy, however, includes the capacity to change one’s mind. That in turn requires that one not be bound to a choice unless and until one has actually made it. Suppose I am faced with a choice between X and Y. Once I have actually made a decision and communicated it there may well be good reasons, in the interests of finality or certainty, for preventing me from withdrawing that decision. Suppose, however, that after having been inclined at various points towards X or Y, I settle on one or the other. The mere fact that, had you asked at the right time, you could have persuaded me to X or Y should not disentitle me from insisting on my actual decision of Y or X (respectively) once it is made. The exercise of autonomy consists in the actual decision and mere hypothetical decisions cannot trump it. Likewise in Example 3: having failed to ask at the right time, D should be stuck with whatever actual decision C makes when it discovers the payment.

IV. Counterfactuals as a (Recognised) Defence

Counterfactual arguments might still have normative significance for an unjust enrichment claim if they constituted a defence, offering additional reasons why the claim should fail despite the presence of good prima facie reasons for restitution. There are two possibilities here: either the counterfactual argument fits within one of the existing recognised defences to such a claim or it should give rise to a new, sui generis defence. The only existing defence that might involve a relevant counterfactual enquiry is change of position.

A. Change of Position

It should be noted at the outset that, as a matter of (first instance) authority, change of position is never available to a Woolwich defendant, so for that reason it might not help to explain Examples 1 and 2.Footnote 77 This threshold objection can be put to one side for two reasons. First, there is no such objection in Example 3, so detailed consideration of change of position is warranted in any event. Second, the rationale given for excluding change of position in Woolwich claims is that the defence is “outweighed” by the constitutional principle against taxation without parliamentary authorisation which provides the ground for restitution.Footnote 78 The weighing metaphor is, however, unhelpful unless we have first understood exactly what policy or interests the change of position defence serves. Only then can we ask whether such policies or interests are engaged in the context of public administration and whether they are commensurable with and outweighed by the policies (discussed in the previous section) in favour of awarding full restitution.

It is often said that change of position ensures that defendants are not made “worse off” by having to make restitution of a benefit.Footnote 79 It is commonly assumed that worse-off-ness in this context, like “loss” in tort and contract, is measured against a counterfactual baseline: if the defendant is required to make restitution, will they be worse off than they would have been, at the same point in time, in a counterfactual world where they had never received (or had never been led to expect that they would receive) the enrichment?Footnote 80 There are two broad situations in which the defendant might be worse off in this counterfactual sense: (1) where they have been deprived of the very thing received from the claimant (e.g. where it has been stolen or destroyed: “loss of benefit” cases); and (2) where the defendant still has the thing or right constituting the enrichment but, relying on the assumption that the enrichment was theirs to keep, they have made dispositions of other assets, foregone opportunities or made other changes to their life which they would not otherwise have made (“reliance” cases).Footnote 81 In both cases, the change in the defendant’s circumstances is counterfactually dependent on the receipt of the enrichment (i.e. the change would not have happened “but for” the receipt),Footnote 82 and it is generally assumed that this betokens causation: the enrichment must be a “but for cause” of the change of position.Footnote 83 This is incorrect. There are counterfactuals at work, but not all of them are causal, not all of them will ground a defence without proof of additional matters and some counterfactual worse-off-ness should probably give rise to no defence at all.

1. Loss of benefit

If C pays D £100 in cash, D is mugged and the cash stolen, and C then demands restitution, D will be worse off financially than they would have been if (counterfactually) C had never come into the story. The relationship between C’s payment and D’s loss is not causal but analytical. One cannot lose money one never had.Footnote 84 The best explanation of these cases is that the combined impact of the payment and restitution claim (both of which are initiated by C) on D’s life is to leave D (who has been a passive actor throughout) worse off than they would have been absent C’s intervention and the court should prioritise protecting D from that counterfactual worse-off-ness over reversing C’s payment in full. That being so, it seems to be important that C, not D, is responsible (though not necessarily in a moral sense) for the two interventions in D’s life (the payment and the restitution claim). That suggests that there should be no defence if D demanded or induced the payment in some way. However, unlike the reliance cases (discussed below), the mere fact that the defendant knew they were not entitled to retain the payment at the time the benefit was lost should not deprive them of a defence.

The prevention of benefit cases discussed earlier share the critical features of a loss of benefit case. If C pays D and then demands the money back, but in the meantime receipt of C’s payment has prevented D from receiving a benefit from an alternative source, the result is that D (who has been throughout a passive object of C’s actions) is financially worse off than they would have been in a counterfactual world where C had never intervened. There is no apparent reason to treat these cases differently from the loss of benefit cases. Both should, and probably do, give rise to a change of position defence.Footnote 85

There are further, and more controversial, examples which look analogous to these two cases. Suppose C mistakenly pays D £100 in cash. D, realising there has been a mistake, makes a trip she would not otherwise have made to the bank to deposit the money for safekeeping until she can arrange repayment, but on the way is hit by a bus and incurs more than £100 in medical bills.Footnote 86 D is worse off overall than they would have been absent C’s payment and subsequent claim. One might be inclined to deny D a defence on the ground that being hit by buses is just one of the unfortunate vicissitudes of life, and the connection between C’s conduct and D’s injury is too freakish to justify shifting that risk from D to C.Footnote 87 We need not reach a final verdict here but the example may suggest that the test outlined above (is D a faultless and passive actor who will be worse off after C’s payment and restitution claim than they would have been without them?) needs to be further cut back.

In any event, the broader test suggested above does not explain any of our examples. For the reasons given earlier,Footnote 88 it cannot explain Examples 1 and 3: but for the payment actually made, D would simply have been worse off by the amount of that payment. There is no reason to think that mere non-receipt of the payment would, by itself, have inspired them to remake the relevant administrative decision (Example 1) or to seek board consent (Example 3). Example 2 does fit the pattern of a prevention of benefit case: but for the VAT payment, D would have received £25 more by way of income tax. However, D is not a mere passive object of C’s conduct in making the VAT payment. Although VAT is, in the first instance, self-assessed, it is hardly self-motivated: the revenue is responsible for the promulgation of tax legislation and guidance which requires payment and often provides a detailed explanation of the circumstances in which payment is expected. Except in the incredibly unlikely event that the taxpayer pays more than the legislation and guidance suggest should be due, the revenue is therefore responsible for inducing the payment and should be disentitled from relying on a prevention of benefit defence.

2. Reliance

One obvious difference between reliance and loss of benefit cases is that the former involve decisions that are within the defendant’s control and the defence is (and should be) conditioned on the defendant’s good faith. They must not have known the facts entitling the claimant to restitution at the time they decided to change position.Footnote 89

The sort of worse-off-ness involved here is also different. Wealth has not simply disappeared from the defendant’s hands without a trace. Rather, they have made choices to acquire, or to expend (or forego) money doing, things they value: a good meal, a holiday, the freedom of early retirement, the satisfaction of gift-giving.Footnote 90 If they are required to make restitution they will still have all those advantages. If they are worse off, it is in a different and more subtle sense from the loss or prevention of benefit cases. The point here is that the defendant made their choices on the basis of background assumptions about their available resources and restitution will alter those assumptions, leaving the defendant in a very different overall position, one they did not choose. The critical features are (1) D has a good faith belief that they are entitled to keep the benefit (or will be so entitled when it arrives); (2) they have relied on that belief in changing their position (which may involve expenditure or foregoing a benefit);Footnote 91 and (3) the change of position is irreversible; the choice made cannot now be undone.Footnote 92 If those criteria are made out, respect for the defendant’s autonomy seems to provide a good reason to refrain from pulling the rug out from under them by falsifying ex post facto the honest assumptions on which they have ordered their life.

The counterfactual here, if any be needed, arises from element (2) and (unlike the loss or prevention of benefit cases) probably is causal. Claims about reliance are claims about a person’s reasons for acting and the relationship between reasons and action is best understood as causal.Footnote 93 A “but for” test is therefore available as a heuristic for causation (though it is not the only possible test of reliance/causation).Footnote 94 The test would be: but for D’s belief that they were (or would be) entitled to keep the benefit, would they have acted in the relevant way (disposing of assets, foregoing rights, etc.)?

This cannot explain our examples. In each example it might be said that D has changed their position by choosing, in good faith, to obtain C’s money in a way that turns out to be unlawful and thereafter failing to fix their error (by promulgating a lawful charge, refunding the unlawful VAT payments or obtaining board consent, respectively); in doing so they have foregone a valuable right (the right to extract money from C lawfully); and they should not now be left worse off for having foregone that right. The decision to act unlawfully, however, is not caused by a belief that D is (or will be) entitled to keep C’s payment. If D acts in good faith (that is, subjectively believing that their unlawful actions are lawful) then the decision to obtain the payment in a particular way may be accompanied by a belief that D will be, or is, entitled to keep the payment, but the mere coexistence of the decision and the belief does not make the latter a cause of the former.

A counterfactual, “but for” test of causation may obscure all of this. One might argue that, (1) but for D’s good faith belief that they were entitled to keep the payment, they would have realised that they would have to make restitution; and that entails that (2) D would have realised that what they were doing was unlawful; and (3) D would instead have chosen to act lawfully. (Step (3) does not necessarily follow from (2) because D might not know without more in just what respect the decision is unlawful, but let us overlook that problem for now.) This just demonstrates another potential weakness of counterfactual tests of causation.Footnote 95 The problem can be analysed technically as a case of epiphenomena (effects with a common cause). If factory hooters sound in two cities at 5 p.m. each weekday then it may well be the case that if the hooters were not sounding in London, they would not be sounding in Manchester; but the London hooter does not cause the Manchester one, rather both are common effects of its being 5 p.m. on a weekday.Footnote 96 In our case, the common starting point is that D believes that θ-ing is a lawful way to obtain payment from C. Two distinct conclusions follow from that premise. If:

  1. (1) D wishes to obtain C’s money; and

  2. (2) D wishes to act lawfully; and

  3. (3) D believes that θ-ing is a lawful way to obtain C’s money; then

  4. (4) D will θ.

And, if:

  1. (5) Money obtained lawfully need not be repaid; and

  2. (6) D knows (5); and

  3. (7) D believes that θ-ing is a lawful way to obtain C’s money; then

  4. (8) D will believe that, provided they θ, they will be entitled to keep C’s money.

The belief that θ-ing is lawful (steps (3) and (7)) is thus a cause of both (4) D θ-ing and (8) D’s belief that they will be entitled to keep C’s money. That does not make (8) a cause of (4), which is what is required to make out reliance and establish a change of position defence. The fact that counterfactuals may lead us astray here is (again) a reason to be sceptical of counterfactuals, not a reason to change our intuitions about causation.

V. Counterfactuals as a New Defence

The remaining possibility is that there might be reasons for denying a remedy or limiting the amount of restitution in our three examples which are not reflected in the elements of an unjust enrichment claim or in any of the existing defences and which might therefore justify the recognition of a new, sui generis defence. That might be one way of thinking about the arguments addressed above when considering the unjust factor at play. Perhaps tellingly, however, none of the judges or commentators who have considered the issue so far have attempted explicitly to articulate the grounds of any such new defence. For my own part, I cannot think of any reasons for denying or limiting liability which are not already addressed above. This third logical category appears, therefore, to be an empty one.

VI. Conclusion

It follows that counterfactual arguments about what the defendant could or would have done differently are neither denials of a claim to restitution, nor do or should they constitute a defence. The cases holding otherwise in the context of Woolwich claims are wrongly decided or, in the case of Hemming and Vodafone, wrongly reasoned. Explaining why has required a close analysis of the role of counterfactuals in unjust enrichment, with the following conclusions:

  1. (1) Enrichment, in the sense of counterfactual betterment, is irrelevant. The subject matter of the claim is the payment itself, not the difference it has made to the defendant’s life. Talking about defendants being “enriched”, rather than receiving payments, may obscure this;

  2. (2) Loss, in the familiar sense of whether (and to what extent) the claimant has been made worse off by the defendant’s unlawful conduct, is also irrelevant. This is a critical distinction between claims for restitution and claims for compensation. The only “loss” we are (and should be) interested in is that constituted by the making of the payment. Again, speaking of a “loss” incurred in providing an “enrichment”, rather than simply a payment, may obscure this;

  3. (3) We should be wary of attempts to extend the principle of subjective devaluation beyond cases involving unsolicited non-money benefits. Any such extension would involve the development of a new principle with a different rationale;

  4. (4) The unjust factor in the case of unlawful exactions by public authorities and the unauthorised taking of money by private defendants is concerned only with the presence or absence of lawful authority or consent in the real world. Hypothetical lawful conduct and hypothetical consent are irrelevant. In the case of public authorities, this point can only be properly appreciated if one approaches the matter with regard to the scope and nature of judicial review and its underlying rationale. It is therefore a good example of a situation where, as Rebecca Williams has argued, a Woolwich claim cannot be properly understood through purely private law spectacles;Footnote 97

  5. (5) Change of position does not involve a straightforward test of counterfactual worse-off-ness. On closer analysis it may be better understood as a collection of analytically and normatively diverse defences, involving different counterfactuals, with different content and different purposes. None of those counterfactuals helps to explain our three examples;

  6. (6) Lastly, there is no good reason to introduce a new, sui generis defence to an unjust enrichment claim based on the counterfactual proposition that the defendant could have obtained the benefit lawfully.

Footnotes

I am grateful to Helen Scott, Robert Stevens, Stephen Watterson and Rebecca Williams for comments on an earlier iteration of this article and to the two anonymous reviewers for several helpful suggestions. All errors remain my own.

References

1 E.g. Benedetti and another v Sawiris and others [2013] UKSC 50, [2014] A.C. 938; Investment Trust Companies v Revenue and Customs Commissioners [2017] UKSC 29, [2018] A.C. 275; Prudential Assurance Co. Ltd. v Revenue and Customs Commissioners [2018] UKSC 39, [2019] A.C. 929; Test Claimants in the Franked Investment Group Litigation v Revenue and Customs Commissioners [2021] UKSC 31, [2021] 1 W.L.R. 4354 (hereafter, FII Claimants).

2 P. Birks, An Introduction to the Law of Restitution (Oxford 1985), 7.

3 Ibid.; Banque Financière de la Cité v Parc (Battersea) Ltd. and Others [1999] 1 A.C. 221, 227B (H.L.) (Lord Steyn); Investment Trust Companies v Revenue and Customs [2017] UKSC 29, at [41] (Lord Reed).

4 E.g. S. Hedley, A Critical Introduction to Restitution (London 2001); P.G. Watts, “‘Unjust Enrichment’ – the Potion that Induces Well-Meaning Sloppiness of Thought” (2016) 69 C.L.P. 289.

5 R. Stevens, The Laws of Restitution (Oxford 2023); L. Smith, “Restitution: A New Start?” in P. Devonshire and R. Havelock (eds.), The Impact of Equity and Restitution in Commerce (Oxford 2018), ch. 5, 91; cf. A. Burrows, “In Defence of Unjust Enrichment” [2019] C.L.J. 521.

6 R. (Hemming (t/a Simply Pleasure Ltd.) and others) v Westminster City Council [2013] EWCA Civ 591, [2013] P.T.S.R. 1377, at [121] (Beatson L.J.); Vodafone Ltd. and others v Office of Communications [2019] EWHC 1234 (Comm), [2020] Q.B. 200, at [5], [90]–[91], [97]–[100]; C. Mitchell, “Counterfactual Arguments against Woolwich Liability” in A. Dyson, J. Goudkamp and F. Wilmot-Smith (eds.), Defences in Unjust Enrichment (Oxford 2016), ch. 12, 279, 299–306; G. Virgo, “Causation and Remoteness within the Law of Unjust Enrichment” in S. Degeling and J. Edelman (eds.), Unjust Enrichment in Commercial Law (Sydney 2008), 147.

7 South of Scotland Electricity Board and Another v British Oxygen Co. Ltd.; South of Scotland Electricity Board v British Oxygen Gases Ltd. [1959] 1 W.L.R. 587 (H.L.); Waikato Regional Airport Ltd. and Others v A-G of New Zealand [2003] UKPC 50, [2004] 3 N.Z.L.R. 1.

8 R. (Hemming) v Westminster CC [2013] EWCA Civ 591; Vodafone Ltd. v Office of Communications [2020] EWCA Civ 183, [2020] Q.B. 857.

9 [1993] A.C. 70 (H.L.).

10 South of Scotland Electricity Board v British Oxygen [1959] 1 W.L.R. 587, 596 (Viscount Kilmuir L.C.), 606–07 (Lord Merriman), 608–09 (Lord Reid); Waikato Regional Airport v A-G [2003] UKPC 50, at [84]; R. (Hemming) v Westminster CC [2013] EWCA Civ 591, at [121], [127]–[128] (Beatson L.J.).

11 [2013] EWCA Civ 591.

12 Ibid., at [121], [127]–[128] (Beatson L.J.).

13 Ibid., at [110], [129]; cf. Investment Trust Companies v Revenue and Customs [2017] UKSC 29, at [59] (Lord Reed).

14 [2020] EWCA Civ 183, [2020] Q.B. 857.

15 Vodafone v Office of Communications [2019] EWHC 1234 (Comm), at [64] (Adrian Beltrami Q.C.).

16 Vodafone v Office of Communications [2020] EWCA Civ 183, at [92], [95], [100] (Sir Geoffrey Vos C.).

17 Young v Bristol Aeroplane Co. Ltd. [1944] K.B. 718, 729 (C.A.) (Lord Greene M.R.).

18 Vodafone v Office of Communications [2020] EWCA Civ 183, at [92], [94], [100], [105] (Sir Geoffrey Vos C.).

19 Ibid., at [92]–[93].

20 O.F. Sherman, “Counterfactual Arguments in Unjust Enrichment” (2019) 135 L.Q.R. 561, 563–64.

21 [2014] EWHC 868 (Ch), [2014] S.T.C. 1761.

22 Ibid., at [434].

23 Ibid.

24 Ibid., at [434(5)]; see also [427], [433].

25 Ibid., at [434(6)].

26 Ibid., at [429].

27 [2018] EWCA Civ 1594.

28 Criterion Properties plc v Stratford UK Properties L.L.C. and others [2004] UKHL 28, [2004] 1 W.L.R. 1846, at [4] (Lord Nicholls); Zenith Logistics Services (UK) Ltd. and others v Keates and others [2022] EWHC 1496 (Comm), at [158]–[159] (H.H.J. Keyser Q.C.)

29 [2018] EWCA Civ 1594, at [30] (David Richards L.J.)

30 Counterfactuals of this kind have also been held to be irrelevant to claims of type (2): Recovery Partners GP Ltd. and another v Rukhadze and others [2025] UKSC 10, [2025] 2 W.L.R. 529, at [5], [40], [58] (Lord Briggs, Lords Reed, Hodge and Richards agreeing), [177]–[181] (Lord Legatt), [270], [288] (Lord Burrows), [307] (Lady Rose).

31 P. Birks, Unjust Enrichment, 2nd ed. (Oxford 2005), ch. 1, 49, 73–75; L. Smith, “Restitution: The Heart of Corrective Justice” (2001) 79 Texas Law Review 2115, 2141.

32 A. Burrows, The Law of Restitution, 3rd ed. (Oxford 2011), 50; J. Edelman and E. Bant, Unjust Enrichment, 2nd ed. (Oxford 2016), 357; FII Claimants [2021] UKSC 31, at [170] (Lord Reed and Lord Hodge).

33 Mitchell, “Counterfactual Arguments”, 299–300.

34 Investment Trust Companies v Revenue and Customs [2017] UKSC 29, at [52] (Lord Reed).

35 Mitchell, “Counterfactual Arguments”, 302–04.

36 Benedetti v Sawiris [2013] UKSC 50.

37 B.P. Exploration Co. (Libya) Ltd. v Hunt (No. 2) [1979] 1 W.L.R. 783, 799F (Q.B.) (Goff J.); Birks, Introduction, 109–10; Burrows, Law of Restitution, 47–48.

38 Mitchell, “Counterfactual Arguments”, 300–01.

39 R. Stevens, “Private Law and the Form of Reasons” in A. Robertson and J. Goudkamp (eds.), Form and Substance in the Law of Obligations (Oxford 2019), ch. 6, 119, 124; F. Wilmot-Smith, “Should the Payee Pay?” (2017) 37 O.J.L.S. 844, 845.

40 L. Smith, “Fiduciary Relationships: Ensuring the Loyal Exercise of Judgement on Behalf of Another” (2014) 130 L.Q.R. 608, 628–30; Recovery Partners GP Ltd. v Rukhadze [2025] UKSC 10, [106] (Lord Leggatt), [278]–[286] (Lord Burrows).

41 Stevens, Laws of Restitution, 306.

42 As Lord Reed noted in Investment Trust Companies v Revenue and Customs [2017] UKSC 29, at [45].

43 J. Gardner, “What Is Tort Law For? Part 1. The Place of Corrective Justice” (2011) 30 Law and Philosophy 1, 28ff.; cf. S.A. Smith, Rights, Wrongs, and Injustices: The Structure of Remedial Law (Oxford 2019).

44 Duties to avoid personal injury are obvious exceptions that immediately throw up the difficult question: why do we protect bodily/mental integrity and can money sensibly respond to those reasons?

45 Three Rivers District Council and others v Governor and Company of the Bank of England (No. 3) [2001] UKHL 16, [2003] 2 A.C. 1, 229E–F (Lord Hobhouse); X (Minors) v Bedfordshire County Council [1995] 2 A.C. 633, 730G (H.L.) (Lord Browne-Wilkinson).

46 Bill of Rights 1689, art. 4.

47 R. (Cart) v Upper Tribunal (Public Law Project and another intervening) [2011] UKSC 28, [2012] 1 A.C. 663, at [37] (Baroness Hale); J. Raz, “The Law’s Own Virtue” (2019) 39 O.J.L.S. 1, 8.

48 Smith, “Restitution: A New Start?”, 112; Stevens, Laws of Restitution, 46–48.

49 Birks, Introduction, 109–10.

50 Benedetti v Sawiris [2013] UKSC 50.

51 The distinction I am drawing tracks a broader problem for theories of personal autonomy. If B acts because A tricks or coerces them – e.g. a court orders (coerces) B to pay money in exchange for goods or services B does not want – the decision to act is, in an obvious and important sense, not B’s. Take out A, however, and it becomes much harder to determine in what circumstances choices made by B should no longer count as being really, authentically B’s choices simply because of some unilateral mistake, desire, influence, etc.: S. Buss and A. Westlund, “Personal Autonomy” in E.N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Stanford 2018), available at https://plato.stanford.edu/archives/spr2018/entries/personal-autonomy (last accessed 18 December 2024).

52 Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners [2012] UKSC 19, [2012] 2 A.C. 337, at [79] (Lord Walker).

53 This was the focus of the appellant’s arguments in Vodafone v Office of Communications [2020] Q.B. 857, 861G, 863H (C.A.).

54 H.L.A. Hart and T. Honoré, Causation in the Law, 2nd ed. (Oxford 1985), 207, 249–53.

55 L.A. Paul and N. Hall, Causation: A User’s Guide (Oxford 2013), 74–92.

56 Birks, Unjust Enrichment, chs. 5–6; Stevens, Laws of Restitution, ch. 4.

57 Marine Trade S.A. v Pioneer Freight Futures Co. Ltd. BVI and another [2009] EWHC 2656 (Comm), [2009] 2 C.L.C. 657, at [76] (Flaux J.); Jazztel Plc v Revenue and Customs Commissioners [2017] EWHC 677 (Ch), [2017] 1 W.L.R. 3869, at [30] (Marcus Smith J.).

58 Bill of Rights 1689, art. 4; Woolwich Equitable Building Society v IRC [1993] A.C. 70, 172E (H.L.) (Lord Goff).

59 R. (Cart) v Upper Tribunal [2011] UKSC 28, at [37] (Baroness Hale); Raz, “Law’s Own Virtue”, 8.

60 E.g. EE Ltd. v Office of Communications [2017] EWCA Civ 1873, [2018] 1 W.L.R. 1868, the precursor to Vodafone v Office of Communications [2019] EWHC 1234 (Comm).

61 See generally R. Williams, Unjust Enrichment and Public Law (Oxford 2010), 36–39.

62 Human Rights Act 1998, s. 6.

63 Associated Provincial Picture Houses Ltd. v Wednesbury Corporation [1948] 1 K.B. 223 (C.A.).

64 Dimes v The Proprietors of the Grand Junction Canal (1852) 10 E.R. 301, 315 (Lord Campbell).

65 Raz, “Law’s Own Virtue”, 8; T. Bingham, The Rule of Law (London 2011), ch. 6.

66 Raz, “Law’s Own Virtue”, 8; Bingham, Rule of Law, ch. 6; P. Cane, Administrative Law, 5th ed. (Oxford 2011), 405–09; C. Crummey, “Why Fair Procedures Always Make a Difference” (2020) 83 M.L.R. 1221; T.R.S. Allan, “Procedural Fairness and the Duty of Respect” (1998) 18 O.J.L.S. 497; J. Waldron, “How Law Protects Dignity” [2012] C.L.J. 200.

67 L.L. Fuller, The Morality of Law, revised ed. (New Haven 1969), chs. 1–2; J. Raz, “The Rule of Law and Its Virtue” in J. Raz, The Authority of Law, 2nd ed. (Oxford 2009), ch. 11.

68 See Senior Courts Act 1981, s. 31(4).

69 T. Cornford, Towards a Public Law of Tort (Aldershot 2008), 16–30.

70 Ministry of Justice, Judicial Review – Proposals for Further Reform: The Government Response, Cm. 8811 (2014), [10]–[11].

71 See e.g. R. (Michael) v Governor of Whitemoor Prison and another [2020] EWCA Civ 29, [2020] 1 W.L.R. 2524, at [51] (Lord Burnett C.J.); A v Kirklees Metropolitan Council and Dorsey [2001] EWCA Civ 582, [2001] E.L.R. 657, at [25] (Sedley L.J.).

72 Waikato Regional Airport v A-G [2003] UKPC 50, at [81]; Vodafone v Office of Communications [2019] EWHC 1234 (Comm), at [27(b)].

73 Cf. R. (Cart) v Upper Tribunal [2009] EWHC 3052 (Admin), [2011] Q.B. 120, at [36]–[38] (Laws L.J.); P. Craig, Administrative Law, 9th ed. (London 2021), [16-040].

74 P. Cane, “Damages in Public Law” (1999) 9 Otago Law Review 489, 495–96; R. Stevens, Torts and Rights (Oxford 2007), 235.

75 Air Canada v British Columbia [1989] 1 S.C.R. 1161 (Supreme Court of Canada); cf. Kingstreet Investments Ltd. v New Brunswick (Finance) [2007] 1 S.C.R. 3 (Supreme Court of Canada).

76 R. v Cambridge Health Authority, ex parte B. [1995] 1 W.L.R. 898, 906F (C.A.) (Sir Thomas Bingham M.R.); R. v Secretary of State for the Environment, ex parte Hammersmith and Fulham London Borough Council and Others [1991] 1 A.C. 521, 593F–H (H.L.) (Lord Bridge); Nottinghamshire County Council v Secretary of State for the Environment [1986] A.C. 240, 247C, F–G (H.L.) (Lord Scarman).

77 Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners [2014] EWHC 4302 (Ch), [2015] S.T.C. 1471, at [309]–[315] (Henderson J.).

78 Burrows, Law of Restitution, 550; E. Bant, “Change of Position as a Defence to Restitution of Unlawfully Exacted Tax” [2012] L.M.C.L.Q. 122, 140.

79 A. Burrows, “Good Consideration in the Law of Unjust Enrichment” (2013) 129 L.Q.R. 329, 330; Stevens, Laws of Restitution, 356; Birks, Unjust Enrichment, 223.

80 A. Burrows, A Restatement of the English Law of Unjust Enrichment (Oxford 2012), 16 (section 23(1)).

81 E. Bant, The Change of Position Defence (Oxford 2009), 145; G. Virgo, The Principles of the Law of Restitution, 4th ed. (Oxford 2024), 757.

82 A “no-payment” counterfactual does not work in cases of anticipatory reliance: A. Ratan, “The Unity of Pre-Receipt and Post-Receipt Detriment” in A. Dyson, J. Goudkamp and F. Wilmot-Smith (eds.), Defences in Unjust Enrichment (Oxford and Portland 2016), ch. 5, 87. My own solution to this problem is outlined in the discussion of reliance below.

83 Virgo, Principles of the Law of Restitution, 757; Burrows, Law of Restitution, 531; Scottish Equitable plc v Derby [2001] EWCA Civ 369, [2001] 3 All E.R. 818, at [30]–[31] (Robert Walker L.J.).

84 Hart and Honoré, Causation in the Law, 114–15.

85 TRA Global Pty Ltd. v Kebakoska [2011] VSC 480, (2011) 209 I.R. 453 (Supreme Court of Victoria).

86 L.D. Smith, The Law of Tracing (Oxford 1997), 36–37.

87 A similar example is discussed in the causation literature: D speeds early on in a car journey with the result that later, while driving safely, he is in just the wrong place at the wrong time to collide with C. Does the speeding cause the collision?: Hart and Honoré, Causation in the Law, 121–22; M.S. Moore, Causation and Responsibility: An Essay in Law, Morals, and Metaphysics (Oxford 2009), 85, 88, 277; J. Edelman, “Unnecessary Causation” (2015) 89 Australian Law Journal 20, 23.

88 Text to notes 31–32 above.

89 Lipkin Gorman (a firm) v Karpnale Ltd. [1991] 2 A.C. 548, 580C (H.L.) (Lord Goff).

90 On the last, see H.W. Tang, “Restitution for Mistaken Gifts” (2004) 20 Journal of Contract Law 1.

91 Commerzbank A.G. v Price-Jones [2003] EWCA Civ 1663, at [39] (Mummery L.J.).

92 Bant, Change of Position Defence, 134–38.

93 D. Davidson, “Actions, Reasons, and Causes” (1963) 60 Journal of Philosophy 685; Campbell v Griffin [2001] EWCA Civ 990, at [19] (Robert Walker L.J.).

94 Bant, Change of Position Defence, 35–36, 148, 163.

95 See generally Moore, Causation and Responsibility, chs. 16–17.

96 J.L. Mackie, The Cement of the Universe: A Study of Causation (Oxford 1980), 83–84.

97 Williams, Unjust Enrichment, 36–39.