I. Introduction
Ever since Professor Peter Birks first published his seminal article on “Restitution and Resulting Trusts”,Footnote 1 there has been a heated debate amongst restitution and property lawyers about the existence and form of so-called “proprietary restitution”. It concerns the extent to which the defendant's unjust enrichment at the claimant's expense can and will generate a right or other interest in rem (i.e. a right or interest relating to a specific asset which is prima facie binding on the whole world, as opposed to a claim exigible merely against a particular debtor).Footnote 2 Take a simple example: A pays £100 to B in the mistaken belief that it is owing. There is no doubt that A has a personal claim against B to recover the value of B's enrichment, but the question arises whether A can also choose to assert a proprietary interest in the money he has handed over. He will typically wish to do this to obtain priority over B's general creditors in the event of B's insolvency, because the cash and/or its traceable proceeds will then not form part of B's estate. If the law does allow A to skip the queue of B's creditors in this way, a further question arises about the exact nature of A's proprietary interest in the money paid over, and whether this interest in any way depends on A's mistake having been induced by B, fraudulently or otherwise.
The debate between advocates of proprietary restitution and its opponents has hitherto focused on the resulting trust as a means of carrying the defendant's enrichment back to the claimant.Footnote 3 However, it has also come to be recognised that so-called “powers in rem”, which are merely embryonic or potential rights in rem, have a role to play in the reversal of unjust enrichments. Indeed, we will see that proponents of the resulting trust model of proprietary restitution now generally concede that powers in rem displace vested equitable interests under a trust in cases where the creation of an immediate interest would conflict with the nature of the parties' relationship, especially in the contractual context.Footnote 4 This paper seeks to suggest that, with respect to impaired consent transfer cases of the kind exemplified above, where the transferor's consent is defective in some legally relevant way (albeit not so defective as to be completely disregarded), the identification of the resulting trust as both central case and principal vehicle of proprietary restitution may have been misconceived. Conveyances intended to be outright but made under the impression of a mistake, or those induced by duress or undue influence, should never be automatically reversed in equity. Looking at the issue from the angle of property law, and taking account of the concerns expressed in various judicial decisions, it will instead be argued that proprietary restitution after such impaired consent transfers should always take the form of a power (which has to be exercised before any property right revests), even where the transferor has a fully-fledged personal claim to restitution springing up – without more – from the moment of the transferee's receipt.
We will begin by outlining the three main rival approaches towards proprietary restitution in impaired consent transfer cases (the so-called “abstractionist” view, the “immediate interest model” and the “power model”) and then briefly examine the extent to which each of them can derive support from the case law.Footnote 5 The bulk of the paper is devoted to demonstrating that the power model, which is already well-established in the contractual context, has important conceptual and practical advantages over the immediate interest model which warrant its invocation even in non-contractual situations. Moreover, the power model is sufficiently sensitive to the reasons for restitutionFootnote 6 to allow for fine-tuning at the margins and for the judicial development of morally significant distinctions which neither of the two other models can accommodate. For exposition purposes, the present paper will focus on impaired consent transfers of (title to) movable objects, but its central thesis could, with appropriate modifications, also extend to conveyances of interests in landFootnote 7 and even to transfers of pure intangibles.Footnote 8
II. Three Main Approaches towards Proprietary Restitution
A. No Proprietary Restitution: The “Abstractionist” View
Some authors believe that proprietary restitution does not or should not exist: the event unjust enrichment never generates property rights of any kind. Prominent amongst them is Mr. William Swadling, who has argued that every (purported) transfer falls into one of two categories. It is either completely void or fully valid.Footnote 9 In the first category of case, the transferor's intent to pass title is so severely impaired as to be completely non-existent in the eyes of the law.Footnote 10 Typically, this occurs where the transferor is labouring under some fundamental mistake when giving up possession of the goods in question, e.g. as to the identity of the transferee,Footnote 11 and it may also be true where an extreme (physical) form of duress is being applied to him.Footnote 12 Swadling does not regard this as an unjust enrichment situation.Footnote 13 The second category encompasses all other cases. Here title passes, and it always does so irretrievably. Although lesser impairments of consent, such as the liability mistake in our example, may in appropriate circumstances give rise to a personal claim to restitution for the transferee's unjust enrichment, the conveyance as such can never be impeached, whether at law or in equity. There is no space for a tertium quid in the form of a resulting trustFootnote 14 or based on the notion of a voidable transfer.Footnote 15 Swadling justifies his view primarily by reference to a “principle of abstraction” operating in English law.Footnote 16 Since the validity of a conveyance by delivery does not depend on the existence or validity of the obligatory contract in pursuance of which it is made, Swadling contends, most impairments of consent do not prevent a fully valid and unimpeachableFootnote 17 conveyance occurring. He bolsters this “abstractionist” argument by outlining various policy objections to proprietary restitution, in particular that it leaves (certain) enrichment creditors with what he regards as an unwarranted windfall in the event of the transferee-enrichee's insolvency.Footnote 18
B. Immediately Vested Interest under a Trust
Those who believe that the event “unjust enrichment” is capable of generating a proprietary response have hitherto maintained that – with the possible exception of cases in which the right to restitution depends upon e.g. a prior rescission of contractFootnote 19 – proprietary restitution takes the form of an immediately vested interest under a trust. In our mistaken payment case, B would therefore hold the money on trust for A from the moment of receipt. Since the equitable interest thus “jumps back” to A, the trust can be described as a “resulting” trust. And since the trust is created non-consensually, by operation of law, it can also be described as “constructive”. Two principal justifications have been advanced for the creation of a trust to reverse unjust enrichment. They are, to some extent, related.
1. “Equity regards as done that which ought to be done”
The first invokes the maxim that “equity regards as done that which ought to be done”.Footnote 20 It rests on the assumption that where A has bestowed a valuable benefit on B in circumstances which make B's receipt “unjust”, B comes under a duty to surrender his enrichment in specie whenever possible (rather than just its value). If the transfer consisted in the conveyance of a chattel, then this chattel must be re-conveyed.Footnote 21 Equity specifically enforces A's personal claim to restitution, so the reasoning goes, and B therefore holds whatever rights he has obtained (in our case title to the money)Footnote 22 on trust for A.
2. The a fortiori argument from “presumed resulting trusts”
The second explanation for the immediate interest model of proprietary restitution is most prominently espoused by Professors Peter Birks and Robert Chambers.Footnote 23 It draws on the case law concerning so-called “presumed resulting trusts”. These are trusts which typically arise when a gratuitous transfer is made outside a relationship of advancement. Birks and Chambers have argued that what is being presumed in these cases is that the transferor did not intend to benefit the transferee. A gratuitous transfer, they say, is an “apparent gift”, and “equity is suspicious of gifts”. If, however, equity imposes a resulting trust where a non-beneficial transfer is merely presumed, then such a trust must a fortiori arise where it is positively proven,Footnote 24e.g. because the transferor's consent to the transfer was vitiated by mistake. Hence, they conclude, an immediately vested equitable interest should be regarded as a generalised proprietary response to unjust enrichment in all cases of impaired consent transfers.
C. The Power Model of Proprietary Restitution
The power model of proprietary restitution contended for here shares its conceptual underpinnings with the law of rescission. Its starting point is not the fortification of the transferor's personal claim to restitution, as the first justification of the immediate interest model would seem to suggest. Instead, it understands proprietary restitution as an integral part and direct consequence of the way in which conveyances are effected under English law. Powers in rem are parallel to – but not necessarily based on – rescission in the contractual context.
Central to the generalised power model of proprietary restitution is the notion of a voidable transfer.Footnote 25 Just as a contracting party can, under certain circumstances, revoke his consent to the bargain he has struck, so a transferor can, under certain circumstances, revoke his consent to a conveyance he has made. All conveyances, howsoever effected, depend for their validity on the transferor's intention to pass title to the transferee.Footnote 26 In the absence of such an intention, the delivery of a chattel, for example, amounts to no more than a voluntary bailment. Swadling is right to point out that only few defects are serious enough to nullify the transferor's intention altogether. They prevent title passing. However, the power model of proprietary restitution diverges from Swadling's view in holding that where title does pass, it need not do so unimpeachably. Less fundamental defects of consent may allow the transferor to avoid the conveyance and thereby to “pull back” his former title.Footnote 27 Such powers are literally “restitutionary”. Yet not all impairments of consent warrant this response. Historically, the common law has limited its intervention to cases in which consent was procured by fraud or duress, leaving equity to deal with other instances where a defect of consent was regarded as relevant.Footnote 28 A revocation of consent which is recognised only in equity does not lead to a revesting of title. Instead, it leads to the creation of a trust.Footnote 29 This trust is “resulting” in pattern and “constructive” in the sense of arising irrespective of the transferee-trustee's consent. In a second step, the transferor-beneficiary can then – of course – call for the retransfer of title under the rule in Saunders v. Vautier.Footnote 30
Until the transferor has chosen to exercise his (legal or equitable) right to avoid the conveyance, the transferee holds title to the object in question outright. The transferor has no vested proprietary interest.Footnote 31 However, because he has the option of pulling back title or bringing a trust into existence, and because this option is available against any third party who subsequently acquires the objectFootnote 32 without qualifying as a bona fide purchaser, he can be described as having a (legal or equitable) power in rem.Footnote 33 Powers in rem are merely potential rights in rem. They are best conceived of as a sort of inchoate or embryonic property right. Under the power model of proprietary restitution, therefore, transfers induced by deliberate deception and illegitimate pressure amounting to duress generate a legal power in rem, while transfers made under the impression of an impairment of consent which equity recognises, e.g. undue influence or an innocently induced mistake,Footnote 34 generate an equitable power in rem.
III. The Relevant Case Law
Before looking at the advantages of adopting a generalised power model of proprietary restitution, it will be convenient to briefly set out the most important cases touching on our topic. For the benefit of the ensuing discussion, however, their summary must be prefaced with a word of warning. Each of the decisions presents a slight variation on the example given above. We need to be alert to these differences. Some cases were decided in the contractual context, while others were not. Some involved fraud, others undue influence or even spontaneous mistakes. Different modes of conveyance will have been used. More fundamentally, almost all of the cases concerned the payment of money. We know (or must assume) that these payments were not made in cash. Payments are nowadays typically mediated via the banking system. As such, they can be distinguished from our paradigm of a direct transfer between A and B. It is only because the relevant judgments appear to treat payments by way of cheque or bank transfer on the same footing as cash transactions that we are justified in considering them at all and gleaning from them some guidance for our present inquiry.
A. Caldwell
Car & Universal Finance Co Ltd v. Caldwell Footnote 35 is the only case not involving money and the best illustration of the power model. Caldwell sold and delivered his Jaguar car to a rogue calling himself Norris. The cheque he accepted in payment was dishonoured. As soon as he became aware of the fraud, Caldwell informed the police and the Automobile Association and asked for their help with finding the vehicle. This happened before the Finance Company, an innocent purchaser for value, bought the Jaguar further down a chain of transactions (from an intermediary who had notice of the fraud). In interpleader proceedings, the Court of Appeal held that although Caldwell had been unable to communicate his intention to avoid the contract to Norris, the fact that he had done everything he could to regain possession of the Jaguar was sufficient to effect rescission and a consequent revesting of title. In other words, Caldwell had validly exercised a common law power in rem before this was barred by the intervention of third party rights.Footnote 36
B. Re Goldcorp
Re Goldcorp Exchange Ltd (In Receivership),Footnote 37 by contrast, seems to point the other way. A company had defrauded some of its customers into purchasing gold bullion for future delivery on a non-allocated basis. When the company was wound up, the Privy Council held that these customers did not have any proprietary entitlement to the company's assets such as would have afforded them priority over a bank's crystallised floating charge. They had neither acquired rights in the bullion held by the company,Footnote 38 nor – and this is crucial for our purposes – could they assert a proprietary interest in the money paid as purchase price. Lord Mustill stated that “[u]pon payment by the customers the purchase moneys became, and rescission or no rescission remained, the unencumbered property of the company”.Footnote 39
C. Allcard v. Skinner
Allcard v. Skinner Footnote 40 is the most prominent of a number of cases dealing with “rescission” of gifts. A young woman had made substantial gifts of money and railway stock to the Sisters of the Poor, a protestant religious order, when entering the Sisterhood. Over five years after leaving it in order to join the Roman Catholic Church, she sought to recover what remained of the gifts on the basis that they had been effected under the undue influence of the Mother Superior. The Court of Appeal held that she could indeed have revoked the gifts upon leaving the Sisterhood, but by a majority found that she had since delayed too long.Footnote 41 Drawing an analogy with contract law, Lindley L.J. said that undue influence rendered a gift voidable in equity, but that if the donor did not call it into question within a reasonable time after the influence under which it was made ceased, this would be treated as a kind of affirmation or ratification.Footnote 42 He also stated obiter that the gifts would have been voidable at common law had fraud or duress been in play.Footnote 43Allcard v. Skinner thus clearly adopts a power model in the gift context.
D. Banque Belge
In Banque Belge pour l'Etranger v. Hambrouck,Footnote 44 a rogue named Hambrouck had tortiously obtained a number of cheques purportedly drawn by his employer on the claimant bank, Banque Belge. He paid these into his account at Farrow's Bank and thereby defrauded Banque Belge into parting with a sum of about £6,000. Hambrouck subsequently gave part of the money to his mistress, a certain Mlle. Spanoghe, for no or immoral consideration. She deposited it in her account at the London Joint City and Midland Bank. When Hambrouck's frauds were discovered, Mlle. Spanoghe's account was still £315 in credit. In interpleader proceedings, the Court of Appeal held that Banque Belge was entitled to recover this amount. The credit on Mlle. Spanoghe's account represented the traceable proceeds of Hambrouck's frauds, and Mlle. Spanoghe was not a bona fide purchaser for value. Two of the three judges expressly premised their conclusion on the fact that Hambrouck had obtained a “voidable title” to the money in question, which Banque Belge had elected to avoid by bringing the action.Footnote 45
E. Chase Manhattan
Chase Manhattan Bank N.A. v. Israel-British Bank (London) Ltd Footnote 46 is the strongest case supporting the immediate interest model of proprietary restitution. Due to a clerical error, the claimant bank had made a double payment of US$ 2 million to the defendant bank. When the latter became insolvent, the question arose whether the claimants could assert any proprietary interest in the traceable proceeds of the second payment. Goulding J. held that, as a matter of New York law (by reference to which the substantive dispute fell to be decided), a mistaken payer “retained” an equitable proprietary interest in the overpaid amount. From the moment of receipt, the recipient held the money on “constructive trust” for him. Goulding J. added obiter that the position would be the same in English law.Footnote 47
F. Westdeutsche
Doubt was cast on Goulding J.'s dictum by the House of Lords in Westdeutsche Landesbank Girozentrale v. Islington London Borough Council.Footnote 48 The defendant local authority had entered into an interest rate swap agreement with the claimant bank. Agreements of this kind were later held to be ultra vires and void.Footnote 49 It had already been decided that the council was under an obligation to repay to the bank the net balance between the (fixed-rate) payment it had received up front and its own (floating-rate) interest payments.Footnote 50 Their Lordships were only asked to consider whether this restitutionary liability attracted simple or compound interest. As the law then stood,Footnote 51 compound interest was available under the courts' equitable jurisdiction, but not for common law claims. Counsel for the claimant bank argued that the bank's liability mistake had initially prevented any property right in the money paid up-front passing to the council, and that when the bank's legal title was destroyed through mixing, it retained an equitable interest, thus creating a trust in its favour.Footnote 52
The House of Lords unanimously rejected this argument and by a majority gave judgment for the defendants.Footnote 53 A liability mistake did not negative the intention to pass title, nor did the mere receipt of money paid under a void contract turn the recipient into a trustee for the payer.Footnote 54 The suggestion that a person who had held property outright could “retain” an equitable interest in it after making an impaired consent transfer was fallacious since it was based on a misunderstanding of the relationship between law and equity.Footnote 55 And in so far as a new equitable interest “obtained” by the transferor was at issue, an extension of the traditional categories in which a resulting trust arose was not warranted.Footnote 56 Both Lord Goff and Lord Browne-Wilkinson thought it inappropriate in policy terms to afford mistaken payers priority in the event of the recipient's insolvency.Footnote 57 Moreover, Lord Browne-Wilkinson identified two practical reasons militating against the imposition of a trust from the moment of receipt. The first concern was rooted in the rule that a vested equitable interest, though subject to bona fide purchase of a legal interest for value without notice, takes precedence over all subsequently created equitable interests.Footnote 58 If the recipient of a mistaken payment held the money received on trust, then a person who advanced a loan secured by an equitable charge over the transferee-enrichee's assets could be defeated by the enrichment creditor, even though he may have had no reason to suspect that the assets in question were held on trust.Footnote 59 Lord Browne-Wilkinson's second concern was with the position of the transferee-enrichee himself, in particular with the dangers of imposing upon him the (potentially)Footnote 60 onerous duties of trusteeship without his consent or even knowledge.Footnote 61 He said: “Since the equitable jurisdiction to enforce trusts depends on the conscience of the holder of the legal interest being affected, he cannot be a trustee of the property if and so long as he is ignorant of the facts alleged to affect his conscience”.Footnote 62 The actual outcome in Chase Manhattan could, in his Lordship's opinion, nevertheless be justified, given that the defendant bank there found out about the claimant bank's mistake within two days of receiving the payment.Footnote 63
IV. Arguments in Favour of a Generalised Power Model
A. Power Model Well-Established in the Contractual Context
Let us begin by considering the question of proprietary restitution in the contractual context. There are two reasons why this is a useful starting point for any argument in favour of a generalised power model. One is that English law, as it stands, is already committed to the power model in those cases where proprietary restitution goes hand in hand with rescission of contract. The other is that most proponents of the immediate interest model now accept that powers in rem trump vested equitable interests whenever the transferor's right to restitution of the transferee's enrichment depends on his first setting aside the contract by virtue or in pursuance of which the conveyance was made.
1. English law does not operate “abstractionism” of the kind advocated by Swadling
As we have seen, the decision in Car & Universal Finance Co Ltd v. Caldwell depended on the defrauded transferor having exercised a common law power in rem.Footnote 64Caldwell is representative of a whole line of cases in which the courts have held that rescission of a fraudulently induced contract of sale leads to a revesting of title in the seller.Footnote 65 The concept of a voidable title is, moreover, presupposed by section 23 of the Sale of Goods Act 1979.Footnote 66 It is, of course, possible to regard Caldwell-type cases as special, as no more than a progeny of the rule that contracts for the sale of goods are capable of acting as vehicles of conveyance.Footnote 67 The seller's power in rem would then be an incidental (and accidental) corollary of rescission in the sale context, without any wider significance for cases in which title passes by other means. Swadling takes this view and describes the Caldwell right to rescind and thereby to revest title as “anomalous”.Footnote 68 Yet such a conclusion would be rash. There are many indications that powers in rem also exist where title does not pass solo consensu by virtue of the contract but where it requires a separate act of conveyance.Footnote 69 Money payments are the obvious example.Footnote 70 Title to notes and coins cannot be transferred by consent alone, even where the cash has been specifically set aside for a particular creditor. In Moynes v. Coopper, a mistaken payment case, Lord Goddard C.J. observed in general terms that “[w]here a transfer of property is obtained by fraud, there is no doubt but that the property does pass, subject, however, to the right of the defrauded party on discovering the fraud to disaffirm the transaction and resume his property”.Footnote 71 Similarly, in El Ajou v. Dollar Land Holdings plc, Millet J. – as he then was – stated that parties who had been fraudulently induced to purchase shares were “entitled to rescind the transaction and revest the equitable title to the purchase money in themselves, at least to the extent necessary to support an equitable tracing claim”.Footnote 72 Indeed, on the view advanced here, the only difference between money and non-money cases is that defrauded payers will in practice often have to have recourse to equity, given that their common law powers in rem are likely to have been lost through mixing.Footnote 73
Re Goldcorp Exchange Ltd (In Receivership) presents the main obstacle to the wholesale recognition of powers in rem in the contractual context.Footnote 74 The decision would seem to suggest that a person who has paid money in pursuance of a contract later rescinded for the other party's misrepresentation is always confined to a personal restitutionary claim.Footnote 75 There is no denying the distinctly anti-proprietary flavour of Re Goldcorp. On the other hand, however, it is worth noting that the actual outcome of the case is not inconsistent with the power model of proprietary restitution espoused here. For one, Lord Mustill suggested that the customers, in their eagerness to establish an interest in the unallocated bullion, had delayed too long before exercising their right to rescind the contracts.Footnote 76 Even if not, their rescission occurred only after the bank's floating charge had crystallised.Footnote 77Re Goldcorp may therefore decide no more than that an unexercised power in rem is defeated by the creation of a vested security interest.Footnote 78 Finally, it appears that the money transferred by the defrauded Goldcorp claimants had not merely become mixed with the company's general assets, but that it (or its traceable proceeds) had actually been paid into an overdrawn bank account.Footnote 79 This, at any rate, would have extinguished the customers' equitable as much as any legal powers in rem.Footnote 80
Leaving all these technicalities aside, the main concern in Re Goldcorp, and the real reason for Lord Mustill's hostility towards proprietary restitution, appears to have been the question of the proper distribution of assets in a company's insolvency. Accordingly, another suggested explanation for the decision is that while rescission would normally operate to (re-)vest a proprietary interest in the rescinding party, this was different where a company had gone into compulsory liquidation and where the statutory winding-up regime intervened prior to rescission.Footnote 81 However, the distinction thus emerging between corporate insolvency and individual bankruptcy (where the rescinding party does obtain priority)Footnote 82 may itself appear undesirable in policy terms. Yet if this is the case, it merely illustrates a more fundamental point: the issue of the availability and form of proprietary restitution cannot adequately be addressed or resolved by appealing to the respective consequences in a transferee's insolvency. Whether an enrichment creditor who has swollen the defendant's assets and who has typically not made a conscious choice to accept the risk of insolvency deserves priority over other creditors, or whether he should rank pari passu with them, is a question which Parliament is in a much better position to answer than the courts.Footnote 83 A statute can both afford preferential treatment to purely personal claims, and it can prevent holders of proprietary interests from asserting their rights in insolvency cases.Footnote 84 To the extent that Lord Mustill's dicta (like parts of their Lordships' speeches in Westdeutsche) encourage us to impose or deny property rights on the basis of such policy grounds, they are in danger of putting the cart before the horse. Neither the outcome nor the spirit of Re Goldcorp ultimately require us to call Caldwell-type cases into question.
2. Power model trumps immediate interest model in the contractual context
If it can therefore be taken as settled that the right to rescind a contract comes coupled with a corresponding power in rem,Footnote 85 we must ask to what extent this also militates against the adoption of an immediate interest model of proprietary restitution in the contractual context. In his book on Resulting Trusts, Chambers argued that “whenever a provider of property to another is entitled to rescind a transaction … and thereby recover any portion of that property, the recipient holds the recoverable property on a resulting trust for the provider from the moment the right of rescission … first arises”.Footnote 86 There are indeed suggestions in the case law that a person who obtains assets by fraud automatically holds these on trust for the person defrauded, even prior to rescission.Footnote 87 However, this proposition has – on the whole – met with great scepticism. In Shalson v. Russo (Tracing Claims),Footnote 88 for example, Rimer J. emphatically denied that a trust could arise before rescission, holding that it was “contrary to principle” to recognise the existence of a fully vested equitable interest before the defrauded party had decided whether or not to let the bargain stand.Footnote 89 Chambers, too, has now acknowledged that the power model of proprietary restitution is preferable and operates to the exclusion of a resulting trust “where the claimant must first rescind a contract or other transaction to recover ownership of an asset”.Footnote 90
3. Reasoning outwards from contractual transactions
Where does this leave our mistaken payment case? Here the transferor's personal claim to restitution of the value of the transferee's enrichment arises without more, so any objection against the immediate interest model based on its inconsistency with an underlying contractual relationship between the parties disappears. Chambers' above-mentioned concession can be read in one of two ways, however. On the first interpretation, the nature of the proprietary response would depend on – and align with – the status of the transferor's personal unjust enrichment claim. Where the personal claim is contingent upon the avoidance of what would otherwise constitute a valid “legal basis” for the transfer justifying the allocation of the wealth to the recipient (in particular a contractual entitlement),Footnote 91 the transferor has a mere power in rem.Footnote 92 Where the personal claim is not so contingent, equity automatically intervenes and carries a proprietary interest back to the transferor.Footnote 93 In our original example, B would hold the £100 on trust for A from the moment of receipt because there is no contract or similar obligatory transaction to rescind. Applying the justifications put forward for the immediate interest model, the trust thus created is a direct consequence of A's proven “non-beneficial transfer” to B and/or a product of the maxim that “equity regards as done that which ought to be done” (namely repayment of the £100).
The second interpretation of Chambers' concession is bolder, but – it is submitted – preferable. It reads more into the words acknowledging that the power model applies where “the claimant must first rescind a contract or other transaction”.Footnote 94 If, by “other transaction”, Chambers can be taken to refer to conveyances as such, then his view would appear to become virtually indistinguishable from the power model advocated here.Footnote 95 What he is then saying is that B does not hold the £100 on trust for A unless and until A has “rescinded” the transfer of title to B, i.e. revoked (in equity) his consent to the conveyancing effect of the delivery. Despite some hints that Chambers may be moving towards this position,Footnote 96 he would probably not be willing to go quite so far. He has recently stated that the mistaken payment case is not one where the recipient's unjust enrichment is part of a “contract or other transaction”, so that “rescission is not involved” in its reversal.Footnote 97
Unfortunately, the two decisions which, on first sight, most strongly support a power model of proprietary restitution outside the contractual context turn out to be ambiguous on closer examination. Neither case allows us to conclude with certainty that it is legitimate to reason outwards from contracts to impaired consent transfers generally so as to obliterate completely the immediate interest model. Neither actually decides that a transferor with a vested personal right to restitution of the transferee's enrichment has no more than a (legal or equitable) power vis-à-vis the right he has transferred. In other words, both cases are equivocal as between the first and the second interpretation of Chambers' concession. In Allcard v. Skinner,Footnote 98 the majority's contract law analogy arguably went so far as to exclude the possibility of any type of restitutionary claim prior to the gift being formally set aside.Footnote 99 Even though the Sisterhood was enriched at Miss Allcard's expense, and even though this enrichment was unjust as a result of the Mother Superior's undue influence, the fact that the transfer was made by way of gift may have created a “legal basis” (analogous to the existence of a contractual entitlement) which prevented a personal right to restitution arising at the moment of transfer.Footnote 100 On this understanding, Miss Allcard was confined to an equitable power only because any fully-fledged right to restitution was dependent on her deciding not to let the gift transaction stand.
Banque Belge pour l'Etranger v. Hambrouck Footnote 101 presents an entirely different problem. Here there is no doubt that the defrauded bank could immediately have recovered the value of Hambrouck's enrichment by a personal restitutionary claim, without first avoiding any underlying contractual or other transaction. However, here it is also perfectly possible that the bank's common law power in rem would have been found to compete with a vested equitable interest had such an interest been asserted.Footnote 102 Moreover, we must remember that the payment was made to Hambrouck's bank rather than directly to the rogue himself. And although judges tend not to distinguish between such indirect transfers of wealth and direct cash payments, one cannot completely disregard the fact that what Hambrouck obtained was merely the traceable product of the money Banque Belge had parted with. It will be seen below that, on one view, property rights contingent on tracing always take the form of a power in rem.Footnote 103Banque Belge is therefore not necessarily incompatible with an immediate interest model of proprietary restitution in our mistaken payment case.
B. Conceptual Advantages of the Power Model over the Immediate Interest Model
Considering the matter from a point of principle, the power model of proprietary restitution has several advantages over the immediate interest model which warrant its application to all cases of impaired consent transfers, irrespective of whether or not a personal restitutionary claim comes into existence at the moment of receipt. On a conceptual level, these are three. First, the power model avoids the difficulties encountered by the justifications of the immediate interest model. Second, it prevents the concurrence of a vested property right with a personal claim. Third, it facilitates the operation of the change of position defence.
1. Difficulties with justification(s) of the immediate interest model
Both of the justifications advanced for the immediate interest model of proprietary restitution are open to objections when scrutinised in connection with impaired consent transfer cases. The maxim that “equity regards as done that which ought to be done” may well lead to the creation of a trust whenever a contractual promise to convey a particular right or estate is specifically enforceable (usually realty),Footnote 104 but it causes great difficulties with respect to restitutionary claims, especially those relating to personal property. It is by no means clear that the transferee-enrichee is always under a duty to re-convey the exact asset he has received from the transferor.Footnote 105 Under English law, his primary duty is to surrender the value of his enrichment.Footnote 106 For mistaken payment cases, this is abundantly clear from the action for money had and received.Footnote 107 Writing extra-judicially, Lord Millett has suggested that proprietary restitution should be regarded as a form of “specific unperformance” and should therefore be confined to “land or other property of special value to the transferor”.Footnote 108 But even if judges had a discretion to order delivery up of particular objects and made generous use of it, application of the maxim could not yield a trust in every impaired consent transfer situation. There is simply no general enrichment-based right in personam ad rem.Footnote 109 And it would be circular to infer the availability of a specific remedy from the existence of an alleged equitable interest, given that the latter is the very thing at issue.
The a fortiori argument from “presumed resulting trusts” also runs into difficulties. Historically, the presumption was not one of “non-beneficial transfer”, but of a transfer on (express) trust for the transferor.Footnote 110 In Westdeutsche Landesbank Girozentrale v. Islington London Borough Council, the House of Lords rejected any suggestion that this had changed.Footnote 111 Presumptions are inferences to fill evidential gaps. When a gratuitous transfer occurs outside a relationship of advancement, the inference that the parties intended to establish a trust is at least as likely as that it was an imperfectly intended gift. Swadling has rightly pointed out that it is therefore misleading to describe the transaction as an “apparent gift”.Footnote 112 Moreover, the fact that “equity will not perfect an imperfect gift”Footnote 113 does not mean that it is generally suspicious of gifts.Footnote 114 As the case law on so-called “presumed undue influence” shows, equity's suspicion is aroused not by gifts amongst strangers, but instead by excessive generosity within close personal relationships of trust and confidence.Footnote 115 It would also be strange if a donor who established (whether or not with the help of presumptions) that a gift was induced through undue influence on the part of the donee were confined to rescinding the gift,Footnote 116 while a transferor who is presumed to have made an apparent gift non-beneficially were to acquire an immediately vested equitable interest. Finally, and most importantly, it is hard to understand what is meant by a “non-beneficial transfer”. The term denotes the absence of an intention to benefit in a very general way. Yet there is a fundamental difference between a transferor's positive intention to transfer an asset on trust (i.e. to bestow a bare legal title on the transferee) and a transfer intended – albeit imperfectly – to be outright. In the latter case, the transferor does want to benefit the transferee. So long as the impairment of his consent is not serious enough to deprive it of any effect whatsoever,Footnote 117 the legal system takes his intentions at face value, at least until he decides to revoke his consent.Footnote 118 The very plea of an imperfectly intended outright transfer not only rebuts the presumption of it being made to set up a trust,Footnote 119 but automatically precludes any suggestion that it was meant to be “non-beneficial”.Footnote 120 Ultimately, the loose umbrella term “non-beneficial transfer” obscures a doubtful equation of situations in which there is actually no intention to benefit the recipient (e.g. where an intended trust fails for lack of certainty and a so-called automatic resulting trust arises) with cases of vitiated intention. Indeed, since an equally dangerous ambiguity is inherent in the very language of vitiation of consent,Footnote 121 it is preferable to speak of “impaired consent” where a party's intention, though defective, is recognised by the law as prima facie valid.
2. No concurrence of vested interest with personal value-based claim
The second advantage of a generalised power model of proprietary restitution is that it avoids the difficulties caused by the concurrence of a vested equitable interest in the object transferred with a personal claim to the value of the recipient's enrichment.Footnote 122 The basic problem is best illustrated by turning, for the moment, to instances in which the transferor's consent is so severely defective that no title passes to the transferee.Footnote 123 Assume that A had put an envelope with cash into B's letterbox, thinking that the house was inhabited by C, and that B realised this.Footnote 124 There is no doubt that, if B refused to return the envelope, A could sue B in the tort of conversion. However, it is far from clear that A could alternatively bring a personal, value-based claim to restitution for unjust enrichment against him,Footnote 125 at least so long as B has not spent the money or mixed it with his own.Footnote 126 While A still retains title to the cash in the envelope, it is difficult to say that B is unjustly enriched at A's expense.Footnote 127 This will not matter much in our example, simply because money passes into currency so quickly and easily. Yet the same is not true of other movables, such as books left as a birthday present outside the wrong house, which are more readily identifiable and less susceptible to bona fide purchase. Those who maintain that A can assert a personal claim to the value of B's enrichment (measured by the full market price of the objects received) either have to distinguish between a defendant's “legal” and his “factual” enrichmentFootnote 128 or are forced into saying that the claimant must renounce or waive his subsisting property rights in order to bring the personal restitutionary claim.Footnote 129
It is true that the co-existence of vested property rights and personal value-based claims is less problematic in our mistaken payment case. Here B was the intended recipient of the money, and he has undoubtedly been enriched at A's expense. Yet if – immediately upon receipt – a trust were to spring into existence carrying the equitable interest back to A, we would have to ask how this proprietary reversal of B's enrichment impacts on his personal liability towards A. As between trustee and beneficiary, the wealth inherent in “having” the asset or right held on trust is already allocated to the latter. One could, of course, argue that since only equity recognises this allocation of wealth, the common law claim to restitution of the full value of the transferee-trustee's enrichment remains entirely unaffected. However, apart from the difficulties of sustaining separate analyses at common law and in equity almost 150 years after the procedural fusion of both systems,Footnote 130 mistaken transferors would be left with a value-based claim to restitution and (under the rule in Saunders v. Vautier)Footnote 131 concurrently also a claim to specific recovery.Footnote 132 The recognition of a vested equitable interest therefore sits somewhat uncomfortably alongside the standard personal response to unjust enrichment. Again, this may be more evident in non-money cases.Footnote 133 On the generalised power model of proprietary restitution, the dilemma does not arise. Unless and until A chooses to exercise a (legal or equitable) power in rem, B's entitlement to the money is outright. In many cases, A will be content to recover an equivalent amount via the action for money had and received. If he goes down this route, he must be taken to affirm the conveyance.Footnote 134 It is only where B is insolvent or A has a particular interest in specific recovery, for example because the object is unique,Footnote 135 that proprietary restitution will in practice become an issue. Exercise of A's power in rem will then lead to a revesting of title or to the creation of an equitable interest under a trust (as the case may be) and at the same time destroy his value-based personal restitutionary claim.Footnote 136
3. Change of position defence easier to operate
One question which has vexed commentators ever since the idea of proprietary restitution was first floated is how to accommodate the change of position or disenrichment defence. The problem lies in tying the existence and scope of a particular proprietary interest of the transferor's (in the specific asset conveyed) to the fate and development over time of the transferee's surviving enrichment (conceived of as an abstract fund). Particular difficulties are caused by cases in which a good faith transferee retains the object received, but – in reliance on the security of his receipt – incurs expenditure he would not otherwise have had.Footnote 137
Take the example of an innocently induced or spontaneously mistaken gift of a painting by A to B which B celebrates by going out to an expensive meal. On the immediate interest model of proprietary restitution,Footnote 138 we might struggle to explain how a vested equitable interest of A's could be affected by the reduction of B's overall enrichment caused by his reliance expenditure.Footnote 139 Not so on the power model. According to the view put forward here, powers in rem are the proprietary equivalent of rights to rescind a contract. Indeed, the latter are best described as powers in personam enabling a person who has bestowed a personal claim on another to divest that other of this right in personam in certain circumstances. Like powers in rem, these powers in personam are also responses to unjust enrichment.Footnote 140 They deal with change of position in a rough and ready way, by insisting that the rescinding party make restitutio in integrum to the other. Although it is not entirely clear how far this allows for changes of position other than the mere rendering of counter-performance to be taken into account,Footnote 141 any restriction in the contractual context to restitution of a defendant's counter-performance is likely to be a recognition of the parties' exchange relationshipFootnote 142 and need not extend to non-bargain transfers.Footnote 143 On the power model of proprietary restitution, a similar mechanism ensures that innocent transferees are not left out of pocket. In our gift example, A's exercise of his (equitable) power in rem can be made conditional or dependent upon his reimbursing B's reliance expenditure.Footnote 144 When dealing with a mistaken payment case, this effectively amounts to a kind of set-off. The payer's equitable interest will thus never in practice extend to a fund which is of greater value than the innocent payee's abstractly surviving enrichment.
C. Practical Advantages of Power Model over Immediate Interest Model
It will be recalled that Lord Browne-Wilkinson in Westdeutsche Landesbank Girozentrale v. Islington London Borough Council identified two practical problems with the immediate interest model of proprietary restitution espoused by Goulding J. in Chase Manhattan Bank N.A.v. Israel-British Bank (London) Ltd.Footnote 145 One was the danger of imposing on the innocent recipient of a mistaken payment the (supposedly) onerous duties of trusteeship, the other potential prejudice caused to innocent third parties dealing with the recipient, especially those who extend a loan on the security of an equitable charge over his assets. In contrast to the solution advocated in Chase Manhattan, a generalised power model of proprietary restitution would go a long way towards accommodating Lord Browne-Wilkinson's concerns.
1. No unwitting trustee liability
On the power model of proprietary restitution, the transferee could not fail to be or become aware of the trust as soon as it is created. Like the right to rescind a contract, a power in rem is exercised by communication with the other party. Our mistaken payer A will have to tell the payee B that he was mistaken and (whether expressly or by necessary implication) that he intends to assert a proprietary interest in the money received by B or its traceable proceeds.Footnote 146 An exception probably exists in cases where B has fraudulently induced A's mistake. The so-called Caldwell doctrine (developed in Car & Universal Finance Co Ltd v. Caldwell)Footnote 147 suggests that actual communication is dispensable where a fraudster deliberately absconds and his victim clearly evinces the intention to set aside the transaction.
What is this liability which Lord Browne-Wilkinson was reluctant to impose on an innocent recipient of a mistaken payment? His Lordship spoke of a trustee being “accountable for his deeds”,Footnote 148 and it is likely that he had in mind the whole package of fiduciary obligations normally associated with trusteeship.Footnote 149 This may explain his rejection of the proposition that a trust arises whenever a separation of legal title and equitable interest occurs, i.e. as soon as an equitable interest becomes “engrafted” upon a person's common law right to possession.Footnote 150 However, it has long been recognised that so-called “bare trusts” do not involve nearly as broad a range of onerous personal duties as those usually undertaken by express trustees.Footnote 151 In particular, a bare trustee will not come under any management, investment or similar fiduciary obligations with regard to the trust property.Footnote 152 The one active duty he is certainly under is of the kind described in Saunders v. Vautier,Footnote 153 namely to effect a conveyance of the trust property to the beneficiary when the latter calls for it.Footnote 154 Connected with this is perhaps a liability to account for any profits made by using the trust property and to pay compensation in cases where the trustee has, by his own actions, put it out of his power to comply with the beneficiary's Saunders v. Vautier demand.Footnote 155 It was this latter kind of strict liability which Lord Browne-Wilkinson considered inappropriate in mistaken payment cases.Footnote 156 Adopting a generalised power model of proprietary restitution is one way of avoiding it.Footnote 157 A transferee who disposes of the object received while the transferor's equitable power in rem remains unexercised does not commit a breach of trust, nor does the subsequent creation of a vested equitable interest relate back to retroactively render his actions wrongful.Footnote 158
2. Greater scope for the protection of third parties
The power model also caters for Lord Browne-Wilkinson's second concern.Footnote 159 Powers in rem are easier to defeat than vested rights in rem. At common law, for example, there exists a general principle of bona fide purchase as regards powers to revest title,Footnote 160 but – with the exception of money and negotiable instrumentsFootnote 161 – not for fully vested property rights.Footnote 162 Here we are concerned with the rules relating to equitable interests. Although trust beneficiaries will be defeated by any good faith purchaser for value without notice of the relevant legal estate,Footnote 163 their rights always take precedence over purchasers who subsequently acquire equitable estates. This is expressed in the maxim “where the equities are equal, the first in time prevails”.Footnote 164 Equitable powers in rem, by contrast, count as “mere equities” and are generally understood to be defeated by any good faith purchaser for value without notice, even of a purely equitable interest.Footnote 165 In Lord Browne-Wilkinson's example of a subsequently created equitable security right by way of charge, the third party chargee will therefore prevail over the mistaken payer on the power model of proprietary restitution,Footnote 166 while the opposite would be true if the immediate interest model were applied. Adoption of the power model thus tips the scales in favour of bona fide purchasers and, in doing so, strikes an ideal balance between the payer's interest in avoiding the impaired consent transfer and the protection of innocent third parties.
3. McFarlane's variation of the immediate interest model
It is worth noting that there is now a variant of the immediate interest model which also takes account of Lord Browne-Wilkinson's concerns. As part of a broader argument about the nature of equitable interests, Mr. Ben McFarlane has recently suggested that a trust reversing unjust enrichment arises whenever an enrichment debtor still holds the right(s) transferred to him and knows or ought to know of his obligation to make restitution.Footnote 167 In our mistaken payment case, the payer A would thus have an immediately vested equitable interest in the money paid to B if and only if B was or ought to have been aware of A's mistake when accepting it. Otherwise B becomes a trustee only when he learns of the relevant facts. This necessarily prevents his becoming unwittingly exposed to a personal liability for breach of trust, and it enables bona fide purchase of equitable interests in the interim.
Although McFarlane, too, emphasises the important role played by “powers” in the context of proprietary restitution,Footnote 168 he uses the word in a different sense to the one advocated here. For him, a common law power is essentially the same as a vested equitable interest, given that it allows the victim of fraud or duress to bring about a retransfer of title from a bad faith transferee.Footnote 169 McFarlane's equitable “power” is of a wholly distinct quality. It consists of the enrichment creditor's factual ability to inform an innocent transferee-enrichee of his obligation to make restitution and thus to bring about the awareness requisite for the creation of a trust. This is a “power” only in a very loose sense of the word. Its exercise is not a necessary condition for A's acquisition of a vested equitable interest in the money paid to B or its traceable proceeds. On McFarlane's view, the same trust would also arise if B's knowledge derived from another source (e.g. a third party) or supervened spontaneously (e.g. as a result of B checking his records).Footnote 170
There can be no doubt that McFarlane's gloss on the immediate interest model reflects the letter of Lord Browne-Wilkinson's speech more accurately than the power model does. His Lordship's suggestion that a constructive trust might arise whenever the conscience of the person holding the legal interest was sufficiently affected focuses on the state of the transferee's mind rather than any action on the part of the transferor. It is submitted that there are nevertheless good reasons for preferring the power model advocated here to an immediate interest model qualified by a knowledge requirement. First and foremost, from a practical point of view, it will typically be easier to pinpoint the exact moment of the transferor A exercising a power to avoid the conveyanceFootnote 171 than to establish when the transferee B acquired sufficient knowledge or notice of his obligation to make restitution for a trust to arise. Moreover, the scope of third party protection is greater on the fully-blown power model. Bona fide purchase of an equitable interest is possible until the moment when A positively chooses to set aside the transfer, and not merely up to the point of B's supervening knowledge.Footnote 172 Conceptually, the modified immediate interest model also fails to explain why a transferee with sufficient knowledge of his obligation to make restitution should owe any duties in relation to the specific right received (or any more duties) than a transferee without such knowledge.Footnote 173 Nor can it wholly prevent the concurrence of a vested equitable interest with a personal value-based claim, as the power model can. Last but not least, in terms of further judicial development, it suffers from a similar lack of sensitivity towards the grounds for restitution as all versions of the immediate interest model. By designating the trust as the central device for reversing all unjust enrichments at the proprietary level, the immediate interest model (whether or not qualified by a knowledge requirement) shuts the door on any argument that proprietary restitution should be available in some, but not in all impaired consent transfer cases. It is to this aspect of the power model that we must now turn.
D. Sensitivity of Power Model to Grounds for Restitution
The argument so far has been concerned to show two things. First, that proprietary restitution exists despite the “abstractionist” (or better: “anti-proprietary”)Footnote 174 undertones of Re Goldcorp and Westdeutsche. Second, that it should always take the form of a power in rem which has to be exercised by the transferor before allowing him to assert any vested interest in the asset conveyed. What remains to be discussed is the exact ambit of powers in rem generated by unjust enrichment. The only distinction drawn above was between cases of fraud and duress, giving rise to a common law power to “pull back” title, and “less serious” cases in which the transferor has only an equitable power to create a trust. Yet this need not be the whole picture. If required, the power model could accommodate further morally relevant distinctions, making it into a useful tool for incremental development by the case law. In particular, it is perfectly possible to exclude from the ambit of proprietary restitution cases where the transferor's mistake was spontaneous, and to adjust the power model flexibly in cases where the right to restitution is purely policy-motivated.
1. Exclusion of proprietary restitution for spontaneous mistake possible
Let us go back to our mistaken payment case. We can distinguish three hypothetical situations. In the first, B deliberately misleads A into believing that a debt is still outstanding. In the second, A's misapprehension is also induced by B, but this time non-fraudulently. In the third, A's liability mistake is spontaneous. On the immediate interest model of proprietary restitution, all three scenarios are equally capable of generating a trust. The trust arises either at the moment of receipt or – in McFarlane's variation – at the moment when B becomes or ought to become aware of A's mistake. On the power model, however, it may be that proprietary restitution is not, or should not be, available in the third scenario. This is because the unjust enrichment situation on which the power model is based can be analysed in different ways.
There are two possible approaches. One is to say that if powers in rem (being rights to revoke the consent which brought about a conveyance) are essentially analogous to rights to rescind a contract (being powers in personam to revoke the consent which brought about a contractual obligation), then they should only be recognised where the defect of the transferor's consent is some way attributable to the transferee.Footnote 175 In other words, powers in rem should be restricted to instances of fraud and duress as far as the common law is concerned, and to other cases of misrepresentation as well as undue influence in equity. Conceptually, the exclusion of spontaneous mistakes could then be justified by saying that the transferor A ought not to be able unilaterally to deprive the transferee B of a right validly bestowed upon him unless B is responsible for the defect in A's consent. This is only the case where B has in some way interfered with or infringed A's decision-making autonomy.Footnote 176 Where it is otherwise, no power in rem exists, though B may – of course – be under a personal obligation to make restitution to A. Spontaneous mistakes would therefore act as “unjust factors” for the purposes of personal restitutionary claims,Footnote 177 but not for the purposes of proprietary restitution.
The alternative approach is to treat the contract analogy with greater caution. If the primary reason why there is no right to rescind a contract for spontaneous mistake is that the sanctity of bargains has to be protected, then powers in rem need not be subject to the same restrictions as powers in personam to set aside a contract. It would be sufficient to prevent powers in rem arising or being exercised whenever there is a danger of undermining the contractual rules of risk allocation.Footnote 178 In short, spontaneous mistakes should in principle be recognised at the proprietary as well as the personal level. To generate an equitable power, defects of consent need only be referable to the transferee's interference with the transferor's autonomy where there is a bargain to be protected. Otherwise the mere fact that the transferor's consent was impaired by his liability mistake should be enough.
The case law is still inconclusive as between these two approaches. Their point of divergence is best understood as the difference between a predominantly defendant- and a predominantly claimant-sided analysis of the relevant “unjust factor” for the purposes of proprietary restitution. Judges are clearly more willing to recognise powers in rem where a defendant-transferee is responsible for the claimant-transferor's mistake than where he is not. We see the importance attached to his contribution reflected in the distinction between legal and equitable powers in rem. Indeed, a decision on mistaken gifts appears to draw a second line, separating induced from spontaneous mistakes. In Re Glubb,Footnote 179 concerning the circumstances under which donations made by charitable subscription could be revoked, the Court of Appeal thought it crucial that the donations had been induced by a non-fraudulent misrepresentation in one of the recipient charity's circulars. The inference must be that the donors would not have had a similar power if their mistakes had been spontaneous. However, gift cases may again not be the best guide to the principles governing proprietary restitution. It has already been noted that the donee of an executed gift is regarded by some as deserving a similar measure of protection with respect to the security of his receipt as a contracting party.Footnote 180 If so, then recovery for non-induced mistakes on the donor's part would typically be barred at the personal and the proprietary level alike.Footnote 181 This leaves us with mistaken payment cases like Chase Manhattan and Westdeutsche, where there is no doubt about the existence of a personal restitutionary claim based on the transferor's spontaneous mistake. We have seen that Westdeutsche is anti-proprietary in spirit, but ultimately not incompatible with equitable powers in rem. English law has therefore yet to decide between the two approaches outlined above. One advantage of the power model is that it does not present a black-or-white choice in the matter of proprietary restitution. Spontaneous mistakes can – but need not – be recognised as sufficient to allow for the unilateral creation of a trust by the transferor.
2. Application of the power model where restitution is purely policy-motivated
The second area in which the power model of proprietary restitution displays a greater sensitivity to the exact nature of the “unjust factor” at stake than the immediate interest model is where the transferor's right to recovery is purely policy-motivated. Our present inquiry has been concerned with impaired consent transfers, but we need briefly to step outside these confines to locate the power model in its broader context. From the idea that the transferor can revoke his consent to a conveyance where that consent was defective (though not sufficiently defective to be regarded by the law as void),Footnote 182 it follows as a corollary that the power model has no inevitable application if such a defect is lacking. In cases like Woolwich Equitable Building Society v. Inland Revenue Commissioners,Footnote 183 where restitution of unlawfully exacted tax was ordered on policy grounds, there is nothing wrong with the transferor's consent.Footnote 184 Unlike the immediate interest model,Footnote 185 the power model would thus prima facie seem to prevent the payee becoming a trustee for the payer. Yet this does not mean that there could never be proprietary restitution in a Woolwich-type case. The existence of a power would simply have to be justified in a different way. Courts should have to decide, on a case-by-case basis, whether the policy militating in favour of repayment is sufficiently realised by a personal claim to restitution, or whether it also necessitates the existence of a proprietary response.Footnote 186 It is likely that powers in rem would be recognised only in very exceptional circumstances, such as where the policy reason for restitution is the very need to protect the claimant against being deprived of the specific asset conveyed to the defendant.Footnote 187
E. Option of Streamlining “Impaired Consent” with “Ignorance” (Substitution) Cases
Before concluding, another brief excursion will complete the bigger picture. Impaired consent transfers are the prototype of participatory unjust enrichments which the law of restitution reverses, but there is now widespread support for the proposition that it also has a role to play in the reversal of non-participatory unjust enrichments. The unjust factor in cases of the latter kind is commonly described as “ignorance”.Footnote 188 Not all “ignorance” cases raise issues of proprietary restitution. If B steals A's wallet and makes a gift of the money to C, A's title to the notes and coins simply survives the theft and the later transfer of possession from B to C. It is not a product of the law of unjust enrichment.Footnote 189 However, rights in rem contingent on tracing are sometimes described as instances of “ignorance”-driven proprietary restitution protecting a person against the unauthorised substitution of his assets.Footnote 190 Without assessing the merits of this suggestion, it is a useful reference point from which to draw attention to a further consequence of the power model.Footnote 191 If, in our example, C were a purchaser for value, A might lose his proprietary interest in the money stolen by B,Footnote 192 yet he would instead be able to assert a proprietary interest in the exchange product transferred by C to B. There is much discussion about the exact nature of a claimant's interest in traceable substitute assets, in particular whether the claimant has a vested right or a mere power in rem.Footnote 193 One argument in favour of the latter approach is that, because of their greater susceptibility to bona fide purchase, powers in rem minimise what Birks has called the problem of “geometric multiplication”.Footnote 194 Were this view to win the day, then the power model of proprietary restitution advanced here for impaired consent transfers would fit neatly with a claimant's proprietary entitlement in unauthorised substitution (“ignorance”) cases.Footnote 195
V. Conclusion
Resulting trusts have long been at the centre of the debate about the availability and form of proprietary responses to unjust enrichment. They were a useful paradigm from which supporters and opponents could reason forwards. However, the immediate interest model of proprietary restitution to which they gave birth has proved problematic in the context of impaired consent transfers. Attempts to justify the model conceptually are open to objections, and its practical implications both for unwitting trustees and innocent third party purchasers have met with the disapproval of the House of Lords in Westdeutsche. Powers in rem, by way of contrast, are already well-established in the contractual context, and their characteristics make them eminently suitable for extension to impaired consent transfer cases more generally. The power model of proprietary restitution is based on a similar idea as that of rescission of contract. Where a transferor's consent to a conveyance is defective, but not so severely eroded as to prevent title passing, it may be revocable at the transferor's option. Exercise of a power in rem recognised by the common law leads to a revesting of title, while the exercise of an equitable power brings a trust into existence. There is no difference in the proprietary response between cases involving rescission of an underlying contractual obligation and those in which the claimant-transferor's fully-fledged right to restitution arises without more. However, a distinction could be drawn between cases of induced and purely spontaneous mistakes. English law is still at liberty to exclude the latter from the ambit of proprietary restitution, though it need not do so. The power model is sensitive to the grounds of recovery. It does not dictate that all impaired consent transfers necessarily generate powers in rem, but only maintains that powers are more appropriate than immediately vested equitable interests if such transfers are to be reversed as a matter of property law. In denying mistaken transferors the benefits of an immediately vested equitable interest, the generalised power model protects innocent recipients from unwitting trustee liability, facilitates the operation of the change of position defence, and enables the bona fide purchase of equitable interests by innocent third parties (especially equitable chargees). And in requiring the transferor actively to exercise his right of revocation, the power model prevents courts having to conduct difficult investigations into the state of the transferee's mind, such as a knowledge- or conscience-oriented variation of the immediate interest model would entail.
Overall, a generalised power model of proprietary restitution provides the best fit with the cases decided to date, while at the same time accommodating the conceptual framework and practical demands of English law. “For impaired [consent] transfers” – it has been remarked – “the power model is now in the ascendant”.Footnote 196 The aim of this paper has been to explain this ascendancy and why we should embrace it.
