Should No Further Books Be Written on the Law of Unjust Enrichment?

“The negative purpose of this work is to ensure that no further books on this topic are written.” (3)

Peter Birks’ four-limb formula of unjust enrichment is well known. According to this formula, the plaintiff should show (a) a transfer of value from them to the defendant (aka ‘enrichment’); (b) a causal link between the defendant’s enrichment and the plaintiff (aka ‘at the plaintiff’s expense’); and (c) the existence of one of the previously-recognized restitution categories, such as mistake, undue influence, duress, incapacity, necessity, or illegality (aka ‘unjust’). If the plaintiff meets those requirements, it is open to the defendant to demonstrate (d) a presence of one of the previously recognized defences, such as ‘change of position’ or illegality.

Consider an unexpected scenario under which my bank mistakenly transfers me $1000. The bank would need to show that I received the value ($1,000) at their expense (the money was received from the bank) and that the factual scenario meets the doctrinal requirements of the traditional restitution category of mistake. I, on my side, could try to argue that I innocently relied on the payment and ‘changed my position’ by purchasing new books for my daughter.

Stated in these terms, Birks’ four-limb formula has been adopted by the UK courts ¹ and has found its way to other jurisdictions as well. ² Ambitiously, it aims to capture almost ³ the entire range of restitutionary claims, such as mistaken payments, payment of another’s debt, frustrated contracts, unrequested provision of benefits, property improvements, and improperly collected taxes under a single normative framework. ⁴ The four-limb formula insists on the internal unity of those claims, bundling them together within a single normative thread.

This indeed explains the significance of Robert Stevens’ comprehensive scholarship, which directly challenges each one of the constituents of the formula. Published in the influential ⁵ 2018 unjust enrichment “Disaster” article ⁶ and most recently in the Laws of Restitution monograph, this scholarship provides a rigorous examination of restitutionary decisions. It challenges the fundamentals of the four-stage formula and offers an alternative. This alternative says that restitutionary claims should be sharply divided by many independent liability structures. The single normative framework of unjust enrichment should be replaced by a variety of the laws of restitution.

This Critical Notice proceeds as follows. Part 1 presents Stevens’ negative thesis which doubts the four-limb formula. Part 2 presents Stevens’ alternative, or more precisely, alternatives. Part 3 articulates some reservations. In a nutshell, my contention is that alongside the appeal of the negative thesis, the positive part of the argument requires some further clarifications and qualifications. Part 4 offers some concluding remarks.

1. The ‘No Further Books’ Thesis

Rooting out the very core of a discipline leads to the collapse of that discipline. If one shows that the law of torts is not about wrongs ⁷ and/or not about losses ⁸ that means that tort law should not exist as we know it today. The same point applies to what appears to be the normative basis of contract law—the agreement between the parties. ⁹ Showing that the law of unjust enrichment is not about ‘enrichment’ would be a fatal blow to this area of law. No area of law could survive such an assault. Without the ‘enrichment’ component, the law of unjust enrichment would cease to exist as we know it today. This is regardless of the answer to the question of whether unjust enrichment is a full-blown cause of action in private law or some general principle which operates on a high level of abstraction. ¹⁰ There is no unjust enrichment law without a defendant’s enrichment.

However, that is exactly what Stevens aims to accomplish in the negative part of his argument. By demonstrating an extraordinary knowledge of the subject matter and through a rigorous examination of the case law, he shows that the doctrine of restitutionary claims does not really involve a plaintiff’s enrichment (23). This is a devastating charge against Birks’ aspiration to unify the classical categories of restitution under a common thread of the four-stage formula. Indeed, Stevens’ work must be viewed as a comprehensive attack against the unjust enrichment movement of the recent decades that has been so successful in the UK (and beyond). The dramatic nature and implications of the negative argument also explain the opening quote to this article, characterising the goal of the negative thesis to be ensuring that ‘no further books’ are published on the subject.

In fact, Stevens takes issue with each one of the constituents of Birks’ four-stage formula. As mentioned, the negative argument challenges the very heart of the formula: the first constitutive element—a defendant’s enrichment. Both the “Disaster” article and the Laws of Restitution monograph strip this element of its essence. While a defendant’s enrichment frequently presents in the factual scenarios of the cases, it is superfluous to the legal reasoning and the justification mode of the judges. By reviewing cases over a period of hundreds of years, tracking the fluctuations of the legal doctrine, and offering counterfactual reasoning, Stevens’ negative thesis shows that the defendant’s liability does not hinge on their enrichment. Following the interaction between the parties, the defendant could be better off (aka ‘enriched’), suffer loss (at least for the purposes of liability determination), ¹¹ or remain in the same position. The fact of the enrichment does not play a normative role in the restitutionary claim of the plaintiff. Stated in these terms, the enrichment is superfluous, redundant, irrelevant, artificial, and foreign to the liability structure. This is a fairly central point which runs throughout Stevens’ work (31, 38, 43, 48-49, 57, 58, 61, 64, 164-72, 210-13, 259-63, 355).

However, Stevens’ negative argument says more than this. In fact, the plaintiff’s state of affairs following the transfer of value between the parties is redundant as well. It is irrelevant whether a plaintiff’s position suffers a loss, enrichment, or remains intact (30, 168). In other words, the restitutionary liability has nothing to do with the state of affairs of the parties following the transfer of benefit. This point applies to both the plaintiff and the defendant (62-65). In this way, Stevens’ negative thesis challenges the most critical element of the four-stage formula on its deepest levels.

Furthermore, the concept of ‘value’ itself is difficult to apprehend and situate within the internal logic of private law (36). ¹² Thus, in Sempra Metals Ltd v Inland Revenue Commissioners, ¹³ the House of Lords used unjust enrichment rhetoric to justify restitution in circumstances resembling a debtor’s breach of a loan contract (63). If the debtor does not repay the loan, can the creditor claim restitution for the lost opportunity to utilize the money? Applying the loose concept of ‘value’ under the first element of Birks’ formula suggests that the answer to this question might be positive. The result appears to be implausible and unjust to the defendant, as no actual transfer takes place between the parties. ¹⁴ The core of the interaction between the parties relates to the loan transfer, not the lost opportunity to use the money. This point was indeed clarified by the subsequent decision of the UK Supreme Court which overruled Sempra Metals, ¹⁵ relying on Stevens’ “Disaster” article. ¹⁶

The second element of Birks’ formula—at the plaintiff’s expense—is no less problematic. The key concern with respect to this element says that it establishes an implausibly loose link between the defendant and the plaintiff’s enrichment. Read literally, the ‘at the expense’ element prima facie establishes only a causal, ‘but for’ test which links the defendant’s activity to the plaintiff. Apparently, the required connection should be much stronger, requiring a substantiated directedness between the parties (31). The lack of this directedness requirement leads the four-stage formula into counter-intuitive results and expands the liability of the defendant.

The ‘rising heat’ and ‘unique stamp’ scenarios well illustrate this point. In the rising heat scenario, the plaintiff lives on the first floor of an apartment building. By heating during cold days, the plaintiff benefits the defendant who lives on the second floor, enjoying the warmth originating from the plaintiff’s apartment. Apparently, there is a causal link between the plaintiff’s action (heating) on the first floor and the defendant’s benefit (enjoying the warmth) on the second floor (29). ¹⁷

The unique stamp scenario presents a situation where only two unique stamps of certain characteristics exist in the world. The plaintiff accidently destroys their stamp, leading to a dramatic increase in value of the defendant’s stamp, as the defendant’s stamp becomes the only stamp of its character in the world (30, 80). ¹⁸ Again, the second element of the unjust enrichment formula should have embraced the plaintiff’s position, as their action (the stamp’s destruction) led to the increase of the defendant’s stamp value.

Stevens points out that despite many efforts, the unjust enrichment scholarship has failed, up to this point, to explain why restitution should be denied in the rising heat and unique stamp scenarios if one takes the unjust enrichment theory seriously (34-35). This theory cannot explain what is problematic about the value received by the defendant in circumstances when the plaintiff does not act with an objective which directly relates to the defendant.

The above point about the significance of the ‘directedness’ between the parties is not an academic exercise but has significant implications in practice. Much of Stevens’ energy has been devoted to demonstrating the problematic operation of the ‘at the expense’ element in the UK House of Lords/Supreme Court unjust enrichment decisions of the last decades. ¹⁹ The Menelaou case is just one example. ²⁰ In this case, a couple sold their house to finance the purchase of a house for their daughter. Since the couple’s house was subject to their bank’s charge, the bank agreed to proceed with the sale subject to the condition that an equivalent charge would be acquired over the daughter’s house. This new charge turned out to be ineffective, as the daughter had not been asked to agree to its creation. The unjust enrichment claim of the bank against the daughter was problematic due to the lack of directedness between the plaintiff (the bank) and the defendant (the daughter). ²¹ The injustice involved in the imposition of restitution on the daughter of the incidental benefit mimics the inadequacy of liability imposition in the rising heat and unique stamp scenarios (53-55). ²²

Stevens’ criticism also targets the third element of Birks’ formula under which the variety of the traditional categories epitomize the ‘unjustness’. Here, Stevens is concerned with the failure of unjust enrichment to draw a meaningful line between the different normative structures of each one of the categories and other elements of the formula. The omission to attentively assess the nature of these traditional grounds of liability led the unjust enrichment movement to overlook important variations and differentiations between them.

Some categories, such as mistake, duress, and undue influence, relate to a plaintiff’s consent. However, Stevens objects to the reliance of the unjust enrichment movement on the subjective wishes and thinking of the plaintiff (7). ²³ Other factors, such as improperly collected taxes and the payment of another person’s debt, appear to be of a different nature. The failure of the consideration factor is different as well, as it operates within the normatively distinctive context of an existing contract between the parties. Stated in these terms, the unjust enrichment movement has artificially grouped the various categories together without proper attention to their internal uniqueness (84-89, 417).

Finally, Stevens’ negative thesis targets the defendant’s side of the four-stage formula—the defences. Similar to the concern expressed with respect to traditional categories, the argument here is that the unjust enrichment formula fails to acknowledge the normative distinctiveness of each defence. The unprincipled linkage to other components of the formula—‘enrichment’ and ‘plaintiff’s expense’—mischaracterizes the true nature of the defences (80, 353-54).

2. The Many Laws of Restitution Thesis

What should come instead of Birks’ four-stage formula? There is no appeal to return to the traditional mocked doctrines of quasi-contracts and constructive trusts. In fact, Stevens views the quasi-contracts doctrine as lacking normative unity, ²⁴ representing a mere “label for a rag-bag group of obligations with nothing in common save that they are not contractual.” (24) Rather, Stevens’ positive thesis offers a fresh formula, or more precisely, formulas to grasp the nature of restitutionary claims.

While Birks’ four-stage formula of unjust enrichment aimed to place almost all restitutionary cases under its auspices, Stevens firmly resists this unification move. Rather, he suggests a careful delineation and sub-categorisation of the various instances of human interaction that give rise to the recognition of a multiplicity of normatively independent frameworks. These have been missed by the unjust enrichment movement. Taken together, the ensemble of alternative liability structures provides an adequate basis of restitutionary claims (9). The ensuing sections present most of these liability structures.

A. Performance Acceptance Formula

The ‘performance acceptance’ formula represents, perhaps, the most important liability structure. It covers the “largest” (9) category of restitution cases, including what Birks characterized as the “core of the core” case of mistaken payments. ²⁵ The following two “indissoluble” (37) limbs constitute this formula: (i) a plaintiff’s performance towards a defendant; and (ii) lack of a good reason for this performance.

The first limb requires some action of a plaintiff towards a defendant. Action means doing something, such as performance of a service, paying money, or transferring property (37). An inaction will not suffice (38). Stevens insists that no analytical line should be drawn between performance in relation to a provision of services and performance in relation to property or money (9, 36-37, 38, 47, 50, 56-57, 65, 80). ²⁶ This means that a case of a mistaken online payment is analytically identical to that of a mistaken mowing of one’s neighbour’s lawn.

To constitute a legitimate performance, the plaintiff’s action must also be towards and for the defendant. These aspects of the performance establish a direct link between a plaintiff’s action and the defendant. An owner who enjoys the work performed by a sub-contractor does not have a restitution claim against the sub-contractor, but only against the contractor (42-43). Equally, in the context of agency law, a third party does not have a restitution claim against the agent as the agent performs for the principal, not the third party (42, 44-45).

Furthermore, the performance must be accepted by the defendant. This corrective justice flavour of the formula epitomizes the bipolar liability structure of private law (46). ²⁷ Following the notion according to which “[l]iabilities are not to be forced upon people behind their backs,” the liability structures must involve considerations which are relevant to both the plaintiff and the defendant. ²⁸ One-sidedness of considerations is rejected. While the plaintiff initiates the performance, the bipolar structure of private law triggers the notion that the defendant should accept the performance or at least have a reasonable opportunity to reject it (46).

According to Stevens, scenarios involving transfers of money represent an almost built-in acceptance of performance. Since banks act as agents for customers as recipients of money (44), acceptance will almost always take place by the bank on behalf of a customer (50). Stevens notes that “[b]anks are agents for receipt,” adding that “[w]here money is paid the condition of acceptance can be lost sight of because it is impossible to make a payment without acceptance.” (50) In other words, the acceptance is implicit in payment transfers (81).

In contrast, the act of ‘acceptance’ cannot be easily attributed to the defendant in cases of unrequested services and property improvements. Here, the ‘acceptance’ component of performance plays a critical role. It means that without some act of acceptance of the plaintiff’s action, the claim against the defendant must fail (13-14, 264-69). This echoes the classical observation made by Pollock, CB in Taylor v Laird: ²⁹ “One cleans another’s shoes; what can the other do but put them on?” ³⁰ The point is that the defendant must have a reasonable opportunity to decline the shoe cleaning service. There is no acceptance without such opportunity as no liability should be imposed behind the defendant’s back (81, n 42).

The second limb of the ‘performance acceptance’ formula requires that there is no good objective reason for the performance (71, 79). This means that even if a plaintiff meets the three requirements of the ‘performance’ limb (i.e., action, towards and for, acceptance), a plaintiff still needs to demonstrate that the defendant does not have a legal entitlement to the benefit, such as a valid contract between the parties, or statutory obligation, or that the transaction constitutes a gift (32, 72). Abstracting from the subjective impulses of the parties, this element purports to impose a fully objective criteria (72, 83-84, 88, 96, 101). As Stevens explains: “‘Good reasons’ then are not moral reasons that apply to one party alone. Nor are they found in the mind of either party. Rather they are objective reasons of entitlement to the performance made, that apply to both parties.” (79)

Consider the classical case of mistaken payments—Kelly v Solari. ³¹ This case involved a mistaken payment made by the insurance company to the beneficiary of the insured. This payment was made despite the fact that the beneficiary did not have an entitlement to the insurance proceeds—the policy lapsed beforehand. Applying the first limb of the performance acceptance formula suggests that the plaintiff (i.e., insurance company) made a performance (payment) ‘towards’ and ‘for’ the defendant (the beneficiary) which was accepted by the defendant, as in any case of money transfers. The second limb is present as well: there was no good reason for the beneficiary to retain the payment, as objectively, there was no valid contract between the parties. In other words, what makes the performance unjustifiable in this case is not the plaintiff’s vitiated consent caused by their subjective mistake, but rather the objective lack of a valid insurance contract (71-73, 79-80).

This does not mean that the traditional restitution categories become completely irrelevant under the second limb of the formula. Rather, Stevens integrates these within the stances of the suggested argument. Thus, as a matter of burden of proof, on balance of probabilities, it is open to the plaintiff to demonstrate the presence of one of the traditional categories, such as mistake, ignorance, necessity, or undue influence (72). The fact that a case falls under one of the categories prima facie establishes a presumption that the performance took place due to a ‘bad reason’ (73-74). This, apparently, shifts the burden of proof to the defendant to demonstrate that the performance involved a ‘good reason’, such as a valid contract and statutory obligation (72-73, 78-79).

Furthermore, Stevens acknowledges that the traditional categories are also relevant for the purposes of determining a ‘good reason’. Thus, the validity of contract formation could be challenged on such grounds as ‘fundamental mistake’, incapacity, and undue influence. ³² In order to constitute a valid gift, the donative intent must be present on both parties’ ends of the transaction (31-32, 75, 98-100). This suggests that the performance acceptance formula does not completely abstract itself from the evaluation of the parties’ intentions and subjectivity.

Stevens perceives the case of improperly collected taxes as fully captured by the ‘performance acceptance’ formula. If a revenue authority misconstrues a tax statute, then the taxpayer’s restitution claim follows the two limbs of the formula: the revenue authority makes a ‘performance’ (the payment) ‘towards’ and ‘for’ the taxpayer whose financial institution ‘accepts’ the payment. Furthermore, the second limb of the formula is satisfied as well, as a misinterpretation of a tax statute means that there was no objective reason for payment (93-95). Stevens also applies this argument to ultra vires tax scenarios, where a tax statute was struck down due to inconsistency with constitutional principles or empowering provisions. If a statute is ultra vires (i.e., void), the restitution should follow (98-100). However, if a statute was valid at the time of the performance, this constitutes a good reason for the performance (87-88). ³³ For Stevens, ³⁴ there is no reason to recognize a separate ‘public’ category of improperly collected taxes, as favoured by other scholars. ³⁵

Stated in these terms, the ‘performance acceptance’ formula addresses the inadequacies of unjust enrichment. The advantage of the formula is that it focuses on the parties’ interaction, rather than on the consequence of that interaction. The consequence of the performance (such as a plaintiff’s enrichment or dis-enrichment) is simply irrelevant to the formula (9, 58, 67). Speaking in the terms of unjust enrichment, the formula favours the ‘value received’ vision of the subject versus the ‘value survived’ vision of it. ³⁶ Equally, the rising heat and unique stamp scenarios are easily explained through the lens of the formula, as there is no performance that takes place between the parties (38). Respectively, the performance acceptance formula sheds light on the remedial part of the restitutionary claims: what is reversed is not the extent to which the performance is consequentially beneficial to the defendant, but rather the value of the performance at the time of the plaintiff’s action towards the defendant (38, 65-68).

B. Other Liability Structures

The ‘conditional performance’ category applies to situations of an existing contract between the parties. If a plaintiff and a defendant form a contract according to which the plaintiff’s performance is subject to certain contractual conditions, the failure of such conditions triggers the restitution of the performance. This notion echoes Lord Wright’s comments made in the context of restitution of payments following contract frustration: “The payment was originally conditional … Accordingly, when that condition fails, the right to retain the money must simultaneously fail.” ³⁷

Indeed, deposits and advanced payments are paradigmatic examples of the conditional performance category. Ordinarily, those payments are made in the context of the plaintiff’s contractual expectation of the defendant’s performance. The fact that the condition is found within the contractual provisions makes it legitimate. Similar to the ‘performance acceptance’ formula, the liability structure aims to remain strictly objective: the conditional performance is manifested towards the external world through the contract formation which takes place between the parties. There is no such a thing as a failed contractual condition which sits in the subjective mind of one or both parties (10, 109-10, 137).

The next category addresses situations when a plaintiff discharges the obligation of the defendant towards a third party (10). These include situations such as paying another person’s debt, discharging another person’s obligation towards a public authority, and situations when one of the co-debtors/co-sureties pays off the entire debt. As Stevens demonstrates, such discharges do not always leave the defendant in a better position, which once again casts doubt on the unjust enrichment’s rationale for restitution. ³⁸

According to Stevens, what stands at the basis of this category is the public reasoning to ensure that “the burden of a legal obligation is borne by the party subject to that obligation.” (153) Stevens resorts to public law’s principles and characterizes this category as a type of “localised distributive justice.” (154) Alongside the irrelevancy of private law (153), the distributive considerations require that legally binding obligations in society are to be ultimately borne by the parties subject to those obligations (153). This explains the critical significance for this account that the defendant’s obligation be legally binding (153-54). If a plaintiff discharges a defendant’s non-binding obligation, the restitution claim must fail (165).

This public dimension of Stevens’ argument is also evidenced on other levels of his analysis. For example, his treatment of statutory illegality is grounded in the notion that the underlying purposes of the invalidating statute must continue governing parties’ entitlements with respect to the restitutionary claims. The illegality itself does not epitomize private justice (22), but rather public goals that a given statute aims to effectuate. This point applies even in cases when the invalidating statute is silent on the point of restitution (385-90, 397). The cases of illegality and discharge of another’s obligation are public law cases, epitomising a sharp private-public divide which Stevens seems to accept (7). ³⁹ Cases could be either private or public.

The category of restitutionary claims in relation to gains received as a result of wrongdoing is a widely debatable topic. Birks, for example, excluded this category of cases in his last work from the four-stage formula. ⁴⁰ Not surprisingly, Stevens’ treatment of the wrongdoing category aligns with those scholars who refused to incorporate it within the stances of the unjust enrichment doctrine. ⁴¹ He suggests focusing on the nature of the infringed right and the question of whether this right entails an inherent dimension of profits (333-49). The wrongfully acquired gains must be attributable to the pre-existing plaintiff’s right, such as their right to property. For instance, the various intellectual property rights include this built-in dimension of profits (14). This vision of the wrongdoing category directly links the nature of the infringed right to remedy under which the latter crystalizes the former. ⁴²

Finally, Stevens claims that the restitutionary defences are foreign to the other constituents of the four-stage formula. Consider the change of position defence, which is perhaps the most important defence in restitutionary claims. In Lipkin Gorman, Lord Goff presented the defence in the following terms:

[W]hy do we feel that it would be unjust to allow restitution in cases such as these [where the defendant has changed their position]? The answer must be that, where an innocent defendant’s position is so changed that he will suffer an injustice if called upon to repay or to repay in full, the injustice of requiring him so to repay outweighs the injustice of denying the plaintiff restitution. ⁴³

Stevens fully embraces this rationale by attributing a critical role to a defendant’s innocence and passivity in relation to a plaintiff’s performance. While the plaintiff is entitled to restitution, the defendant’s position creates an independent normative force, a “new reason” which outweighs the unjustified performance between the parties (59, 353, 355). This justification explains why the change of position defence should not apply, for example, in the context of an existing contract in the conditional performance category, as that would amount to “rewriting the parties’ agreement.” (359)

The treatment of other defences reveals further irrelevancy of the unjust enrichment theory. The fact that a plaintiff may have freely consented to a performance knowing the facts of the performance to be unjustified, leads to a claim’s rejection. This point follows the general position of private law according to which the plaintiff may end a perfectly legitimate claim through their genuine consent, without any need to provide justification for their decision (71, 73, 76-78, 369-71).

A similar point applies to the so-called ‘passing on’ defence. Under this defence, the defendant argues that the plaintiff managed to ‘pass on’ their loss to third parties, such as their customers. However, Stevens’ dissection of the liability structure from the consequential reasoning provides an easy explanation for the general rejection of this defence: a plaintiff’s loss is irrelevant for determining a defendant’s liability (30, 374-76).

Stevens’ attack on Birks’ four-stage formula is powerful and compelling. Negatively, it demonstrates the doctrinal and conceptual deficiencies of the formula through a rigorous examination of the case law. Positively, it puts forward comprehensive, multi-layered alternatives. However, Stevens’ account is not free from deficiencies. This is especially true with respect to some key aspects of his positive argument. I would like to focus on the following two closely interrelated themes which doubt this argument: (1) the notions of rights, acceptance, and objectivity, and (2) the private-public classification, focusing specifically on the ‘private’ classification of improperly collected taxes and the ‘public’ classification of discharging another’s obligation. ⁴⁴

3. Some Reservations

A. Rights, Acceptance, and Objectivity

The favourable attitude towards the corrective justice theory of private law is evident throughout Stevens’ work (20-21). ⁴⁵ Stevens views private law as concerning “injustice between individuals,” (3) noting that these “are principles of justice that underlie” restitutionary claims (18). At parts, the argument directly follows corrective justice’s justificatory mode. ⁴⁶ This raises the question of the rights-based analysis. As Stevens explains, “[r]ights and obligations in private law are bilateral. We need to identify reasons that are also bilateral, that relate the two parties together.” (80) Indeed the corrective justice’s account of private law must epitomize the notion of parties’ rights and duties. ⁴⁷ Under this account, a plaintiff’s right to restitution must reflect a defendant’s duty of non-interference with that right.

However, Stevens’ account is somewhat ambivalent on this point. The significance of the rights-based analysis is mentioned throughout the manuscript (39-40, 210-13, 216-17, 259-63, 333-49). For example, as we have seen, Stevens follows corrective justice’s analysis of the wrongdoing category under which a plaintiff’s entitlement to profits must be firmly grounded in the nature of their infringed right. ⁴⁸ At the same time, Stevens observes that “it is unnecessary to show that any right of the claimant’s [plaintiff’s] was infringed,” (6) adding that “obligations to make restitution … are not based upon any breach of a pre-existing duty.” (20) ⁴⁹ Furthermore, Stevens holds the view that private law’s defences should not necessarily comply with the bipolar structure of private law (22). Indeed, as we have seen, Stevens’ vision of the change of position defence is based on one-sided defendant-oriented considerations. ⁵⁰ While corrective justice embraces the rights-based analysis and grounds the bipolar normativity of the parties’ interaction with no exceptions, ⁵¹ Stevens’ account of defences breaks this normativity apart.

The notions of ‘acceptance’ and ‘objectivity’ are closely related to the rights-based analysis. These notions play a central role in the design of the performance acceptance formula. Stevens says:

An acceptance by the defendant of the payment or service rendered by the claimant [plaintiff] is an essential element of the transactional link between them, so that the performance is the doing of both of them, and not one alone. The reason for its reversal is the lack of any good reason justifying it. Both the performance and the possible reasons that could justify it are objective and apply to both parties. The justification for the claim is not found in the mind(s) of one or both parties. (9, emphasis added; see also 32)

The bipolar structure of private law explains the ‘acceptance’ requirement. By embracing the ‘good reason’ limb, the argument aims to abstract the liability structure from the subjective affairs of the parties. However, the question is whether the argument coherently integrates the objectivity and acceptance notions within its own stances. Some would argue that it does not.

Consider first the acceptance notion. The notion raises terminological and doctrinal difficulties as expressed by Lord Burrows. ⁵² Indeed, Burrows acknowledged the direct impact that Stevens’ argument made on the Sempra Metals overruling and the possible danger of the unjust enrichment scholars involved in “putting forward a theory and then spending the rest of one’s career trying to defend that theory rather than accepting valid criticisms of it.” ⁵³ However, Burrows also questioned Stevens’ alternative—the performance acceptance formula. ⁵⁴

Burrows doubted whether the contractual terminology of the performance acceptance formula can teach us about its nature. Indeed, the term ‘acceptance’ is central to the contract law doctrine of formation under which an offeror’s offer is accepted by an offeree. The ‘performance’ of the parties towards each other represents another essential element of the contract law doctrine. This led Burrows to suggest that the performance acceptance formula “looks as if it is the law of contract.” ⁵⁵ Contractual terminology points to the contractual foundations of the formula. In other words, the formula collapses into contract law.

Burrows questioned whether the case law of mistaken payments supports the ‘acceptance’ limb of the formula. Apparently, it does not. While acknowledging the traditional centrality of the ‘acceptance’ component in the context of the unrequested provision of services, Burrows noted that traditionally there was no analogous requirement with respect to payments, stating: “Certainly acceptance may be normatively important where one is seeking to establish that the defendant is enriched, as in the case of services, but that is not normally in issue in respect of payments.” ⁵⁶

Burrows’ criticism is sound. Furthermore, the acceptance limb of the formula seems to be vulnerable to objections lifted against a related account of the subject. Specifically, Stevens’ requirement according to which the performance must be accepted by the defendant in order to become actionable follows Ernest Weinrib’s account of unjust enrichment. ⁵⁷ Without delving into the differences between the two accounts, on the point of acceptance, Stevens’ account follows Weinrib’s. ⁵⁸ The charges expressed against Weinrib say that it is not always possible to objectively ascertain a defendant’s acceptance, which may make the doctrinal inquiry purely fictional in some cases. ⁵⁹ A similar line of objection applies to Stevens’ acceptance limb, especially on the point of the acceptance attribution in payment cases.

The other concern of Stevens’ positive thesis relates to the subjectivity-objectivity tension. Indeed, the abstraction from the subjective wishes of the parties plays a central role in corrective justice’s account of private law which seems to inform Stevens’ position. ⁶⁰ However, absolute appeal to objectivity is questionable. The legal analysis cannot completely depart from delving into the circumstances of the interaction between the parties. It must strive to ascertain the objective intentions of the parties through the assessment of their subjective actions communicated towards the external world. In other words, corrective justice’s abstraction from subjectivity is limited to the internal impulses and thoughts of the parties. Once those are manifested towards the external world, they receive their legal significance and play an integral role in private law’s reasoning.

However, Stevens’ positive thesis seeks to avoid the appeal to subjectivity and engagement with intentions of the parties surrounding the performance. For example, Stevens seems to be disinterested in the question of whether a benefit is transferred with non-donative objective intent and accepted with such intent. ⁶¹ However, admittedly, the operational mechanics of the ‘good reason’ limb cannot really avoid subjectivity. By Stevens’ own account, traditional restitution categories continue playing a key role in the balance of probabilities of the adjudicative process and are essential in determining the valid reasons for a performance, such as contracts and gifts.

The objection to the ‘good reason’ limb has deep historical and comparative roots. The influential German scholar Friedrich Carl von Savigny was the first ⁶² to introduce a similar concept in 1841. ⁶³ Savigny was concerned about the inability of the traditional Roman law doctrine to adequately address situations where a plaintiff could not vindicate their proprietary right or the proceeds from it. Savigny resorted to unjustified enrichment rhetoric according to which the defendant lacks any legal reason to retain a plaintiff’s right or their proceeds. Critically, Savigny’s reasoning hinged on a plaintiff’s pre-existing proprietary right and the deficiency of Roman law to justify a tracing remedy in those situations. ⁶⁴

Following Savigny’s writing, ⁶⁵ the concept of ‘lack of legal ground’ found its way into Section 812 of the German Civil Code. ⁶⁶ Departing from Savigny’s context of proprietary tracing, Section 812 sets out a general rule of restitution under the ‘lack of legal ground’ concept. However, German jurisprudence resists it. Thus, German courts tend to bypass this section by developing specific grounds of restitution, such as party scenarios and invalid contracts which have nothing to do with the broad concept of Section 812 (37). ⁶⁷ A similar resistance has been witnessed in the Canadian courts ⁶⁸ which have adopted a related concept of so-called “juristic reason.” ⁶⁹ This concept has been recently coined by one of the leading Canadian scholars as no less than an “exercise in obscurantism” and a “disaster.” ⁷⁰ Peter Birks, too, in his last work supported a related concept, titled “Absence of Basis,” ⁷¹ which was not subsequently followed by other unjust enrichment scholars or the UK courts. ⁷²

Labels aside, the inherent difficulty remains the same: the legal analysis has a hard time avoiding the rights-based analysis of the parties’ entitlement. The ‘good reason’ limb inherently triggers an inquiry into the possible reasons for restitution. This, in turn, requires an assessment of the parties’ interaction, providing legal meanings to their actions and words. This is the charge of circularity. Nonetheless, whether labelled as Savigny’s (and BGB’s) ‘lack of legal ground’, Canadian ‘juristic reason’, or Birks’ ‘absence of basis’, the ‘good reason’ limb cannot provide a meaningful foundation of the restitutionary doctrine. ⁷³

Stevens seems to recognize, at least implicitly, this circularity of his ‘good reason’ limb by incorporating traditional categories into his analysis. While his account aims to remedy the inherent vagueness of the limb, de facto this move embraces a different type of analysis. The significance of the traditional categories of restitution should not be attributed to the onus of proof, but rather goes to the very nature of the parties’ interaction. The legal analysis cannot plausibly avoid the consideration of the objective meaning of the parties’ subjective actions and consideration of their intentions. The subjectivity and the intentions of the parties enter the back door of the legal analysis.

From this perspective, the notions of rights, acceptance, and objectivity in Stevens’ account may require further qualification. Some suggestions could be offered, though. Thus, in order to avoid the objection about the contractual terminology, a different label could be used for the performance acceptance formula. Something like ‘defective transaction’ could be a good option. ⁷⁴ The rights-based analysis could be supportive of a sharp division between those rights acquired by individuals in relation to provision of services and rights in relation to property items, including money/financial property. While Stevens insists on equal treatment of services and goods/money, he acknowledges (perhaps, following Burrows’ criticism) that traditionally the acceptance requirement operated with respect to the former rather than latter (38, 47). The rights-based analysis and the relational structure of private law could have also supported an argument that the restitution of services should be grounded on the contractual or quasi-contractual nature of the interpersonal interaction. ⁷⁵ The restitution claims in other instances could have been based on the proprietary nature of the objects involved: property, and/or money. ⁷⁶

As for the objectivity challenge, as argued, it is not possible to completely exclude subjectivity from the legal analysis. Consider the ‘mistake’ notion. As Stevens acknowledges, the traditional private law doctrine has established different thresholds for actionable mistakes in the contexts of contract formation (called ‘fundamental mistakes’ (33, 75)) and of mistaken payments (called ‘serious’ (79) or ‘liability’ (73-74) mistakes). The subjectivity-objectivity tension could be explained through the lens of the law’s treatment of the various types of mistakes. For example, it could be argued that the high threshold of the ‘fundamental mistake’ relates to a greater power that law attributes to the formal manifestation of the contract by the parties.

B. The Private-Public Classification

As we have seen, Stevens allocates improperly collected tax cases on the ‘private’ side. Several reasons have been provided in support of this position. ⁷⁷ First, Stevens makes a sharp division between the reason for the invalidation of the tax statute and the restitutionary claim. While he acknowledges that the reason for such invalidation is usually a public one, relating to constitutional principles or a state’s internal division of powers, Stevens opines that

“once the public law issue of validity has been answered, there is nothing particularly ‘public’ involved in the question of whether money paid that is not due (or, more generally, a performance rendered for no good reason) should be reversed.” (99)

This abstraction from public reason justifies the application of the ‘private’ performance acceptance formula to ultra vires statutes. The separation thesis bifurcates between restitution and its reason. Second, Stevens doubts the boundaries of the alternative, as the public justification could be extended to an endless list of scenarios of tax misconstruction and ultra vires payments made to public and quasi-public bodies, including utility companies and universities (99). Since it is difficult to draw a meaningful line between the various types of improperly collected taxes and public bodies, Stevens resorts to applying private law’s principles of restitution. Finally, this private vision of the subject eliminates the need for the courts to “regulate the state.” (99)

With all due respect, this line of reasoning must be challenged. ⁷⁸ In contrast to Stevens’ separation thesis, it is not possible to dissect between restitution and the reasons for it. Resisting restitution in the ultra vires context effectively validates the statute that the state sought to invalidate in the first place. Constitutional principles and a state’s internal division of powers are public reasons that go beyond the relational bipolar structure of private law. The involvement of those reasons makes the mode of interaction between the actors inherently multilateral and of a distributive character. In other words, the public due of ultra vires taxes’ restitution follows the underlying basis for a statute’s invalidation. In fact, this analysis does not depart significantly from Stevens’ own treatment of the statutory illegality which, under his analysis, expands the underlying purposes of the invalidating statute to the restitutionary claim. ⁷⁹

Indeed, applying this argument to various bodies and organisations is not an easy task. In contrast to Stevens’ observations on the significance of the identity of a particular actor, the classification exercise must focus on the mode of interaction and the nature of a specific activity which takes place between the actors. Private law principles, doctrines, and concepts apply to public bodies, quasi-public bodies, non-profit, private, and public corporations on a daily basis. In some cases, determined on a case-by-case basis, the interaction mode will involve distributive considerations. Hard cases will remain hard cases. The possible complexity of the classification exercise does not prove its normative emptiness.

Furthermore, the context of improperly collected taxes does not merely epitomize what Stevens coined ‘state regulation’ but underpins a core function of contemporary liberal democracy. While tax is essential for the sustainability of the state, taxation power cannot be exercised arbitrarily, but subject to stringent requirements of fairness and the scrutiny of the rule of law. ⁸⁰ The very reason for tax imposition and collection aims to serve people’s rights and contribute to the protection and realisation of their freedoms. The integrity of the system is built on fairness towards the taxpayer. ⁸¹ The denial of restitution of improperly imposed or collected taxes would constitute a denial of one of the central aspects of a state-individuals’ multilateral relationship which goes beyond the bipolar structure of private law. This argument is relevant to all tax bodies and quasi-tax bodies (again, determined on a case-by-case basis), and covers both scenarios of ultra vires and misconstrued taxes, regardless of the question of whether a given tax collection had been induced by the collecting body. Stevens’ ‘private’ treatment of improperly collected taxes completely ignores the uniqueness of the tax collection activity.

Underpinned by the notions of taxpayer fairness, the rule of law, and the multilateral interaction between the state and the taxpayer, the above vision of improperly collected taxes (and other charges) sheds light on its incompatibility with various defences that could be raised by the tax authorities against restitution. Thus, the notion of taxpayer fairness cannot be compatible with the above-mentioned ‘passing on’ defence. ⁸² Regardless of various ‘private’ considerations such as evidential difficulty ⁸³ and the internal components of Stevens’ categories, ⁸⁴ this defence is at odds with the embedded fairness of the state-taxpayer relationship. A similar point applies to the above-mentioned change of position defence. ⁸⁵ In the context of improperly collected taxes, a public body should not invoke this defence, not due to the evidential difficulty to prove the reliance, as argued by Stevens (359)—indeed, one can easily imagine a scenario of a large restitution claim where a public authority can prove such a reliance—but rather due to its incompatibility with the underlying public principles of restitution.

Ironically, a diametrically opposite objection could be raised against Stevens’ classification of discharging of another’s personal obligation category on the ‘public’ side. Indeed, it is not easy to situate this category within the conditions prescribed by Stevens’ performance acceptance formula. The plaintiff performs ‘for’ the defendant, but not ‘towards’ them. The defendant does not ‘accept’ the performance (154). Furthermore, in some cases the legal doctrine does not require demonstrating a plaintiff’s mistake at all, which questions the compatibility of this category with the ‘good reason’ limb of the performance acceptance formula. ⁸⁶ This means that discharging another’s personal obligation does not fit with the present structure of the formula. However, this fact alone does not preclude the applicability of private law’s reasoning. If I mistakenly pay off the debt of my neighbour to their creditor, the situation does not seem to be analytically distinctive from a situation when I mistakenly make an online transfer of $100 to my neighbour.

One can contemplate on the alternative, private basis to discharge another person’s obligation category. The insistence of the legal doctrine on the existence of a defendant’s valid legal obligation towards a third party does not need to necessarily point to the distributive considerations according to which a society wants burdens to be borne by the subjects of the obligations. Rather, this insistence may also point to private law’s justification of so-called ‘incontrovertible’ benefits. The Supreme Court of Canada characterized these benefits as applying to situations when a defendant receives a benefit from the plaintiff that is “demonstrably apparent and not subject to debate and conjecture,” and “it is clear on the facts (on a balance of probabilities) that had the plaintiff not paid, the defendant would have done so.” ⁸⁷ Stated in these terms, the ‘incontrovertible’ benefit notion could have served as an indicative component of the ‘acceptance’ limb of Stevens’ performance acceptance formula. Arguably, the ‘incontrovertibility’ of a benefit provides a strong indication that it has (or at least should have) been accepted by the defendant. The performance acceptance formula may need to be qualified to accommodate the discharge of another’s obligation category, without resorting to public law’s justification.

4. Conclusion

Stevens’ comprehensive attack against Birks’ four-stage formula is appealing. Negatively, it shows that a careful review of the case law departs from key elements of the formula. The ‘plaintiff’s enrichment’ element seems to be redundant and superfluous to the judicial reasoning. Furthermore, the negative argument shows the impossible vagueness of the ‘value’ concept and the implausibility of the loose wording of the ‘at the plaintiff’s expense’ element of the formula. The negative argument tackles the failure of unjust enrichment to coherently integrate the traditional restitution categories and the defences available to the defendant within its stances. These inadequacies are too serious to overcome through some qualification or restatement of the unjust enrichment doctrine. The ‘no books’ argument has been made. The four-stage formula should cease to exist as we know it today. The answer to the question appearing in the title of this Notice should be positive.

Stevens’ positive thesis is comprehensive as well. It insists on the sub-categorisation of the restitutionary claims and offers a multiplicity of independent normative frameworks. However, as we have seen, doubts could be expressed with respect to some key aspects of the positive argument, including a defendant’s ‘acceptance’ in mistaken payments cases, the classification of the discharging of another person’s obligation category, the subjectivity-objectivity tension, the (mis)treatment of the improperly collected taxes category, and the relation to the rights-based tradition of private law.

Yet, Stevens’ work deserves to be considered of utmost importance. The comprehensive, multi-layered challenge of Birks’ formula, the focus on the transaction between the parties (rather than the consequences of it), the private-public divide, and the appeal for sub-categorisation are all important steps forward for grasping the nature of the restitutionary claims.