Jump to ContentJump to Main Navigation
Rights, Wrongs, and InjusticesThe Structure of Remedial Law$

Stephen A. Smith

Print publication date: 2019

Print ISBN-13: 9780199229772

Published to Oxford Scholarship Online: January 2020

DOI: 10.1093/oso/9780199229772.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 07 August 2020



(p.223) 8 Injustices
Rights, Wrongs, and Injustices

Stephen A. Smith

Oxford University Press

Abstract and Keywords

Chapter 8 explores the most controversial of the three private law causes of action discussed in the book: injustices. For the purposes of remedial law, an injustice is, roughly, an unfair loss or gain that has arisen from a transaction between the claimant and defendant. The chapter argues that an injustice is the cause of action for, inter alia, orders to make restitution following defective transfers (‘restitutionary orders’) and orders to pay damages for consequential pecuniary losses (‘compensatory orders’). This argument rests primarily on two propositions: (1) restitutionary and compensatory awards are duty-creating (and therefore the law governing them is remedial, not substantive); and, (2) notwithstanding that claimants who obtain restitutionary and compensatory orders have often suffered a wrong (particularly in the case of compensatory awards), the cause of action for such awards is neither a rights-threat nor a wrong. Chapter 8 also argues that the cause of action for restitutionary and compensatory orders is appropriately characterized as an injustice. But this argument is relatively brief because, assuming the above propositions are accepted, it is difficult to describe the cause of action for these orders without invoking injustice or a closely related concept.

Keywords:   injustices, damages, restitution, compensatory damages, consequential damages, creative remedies

The third of the main private law causes of action recognized in the common law is an injustice. As understood here, an injustice is, roughly, an unfair loss or gain that has arisen from a transaction between the claimant and defendant. This chapter focuses on demonstrating that an injustice is the cause of action for orders to make restitution following defective transfers (‘restitutionary orders’) and orders to pay damages for consequential pecuniary losses (‘compensatory orders’). However, there are almost certainly other examples of injustice-responding remedies. I will flag (but only flag) some possible candidates.

My classification of restitutionary and compensatory awards as injustice-responding orders rests primarily on two propositions: (1) restitutionary and compensatory awards are duty-creating (and therefore the law governing them is remedial, not substantive); and, (2) the cause of action for restitutionary and compensatory orders is neither a rights-threat nor a wrong (and therefore these orders do not belong within Chapters 6 or 7). Of course, I also argue that the cause of action for restitutionary and compensatory orders is appropriately characterized as an injustice. However, this argument is relatively brief. If the above propositions are accepted, it is difficult to imagine how one might describe the cause of action for these orders without invoking injustice or a closely related concept.

The idea of ‘injustice-responding orders’ is novel. Unlike ‘wrong-responding orders’ and ‘rights-threat-responding orders’, injustice-responding orders cannot easily be accommodated within the ‘for every right a remedy’ maxim. The suggestion that restitutionary and compensatory orders are related is even more novel. Against this background, I begin by explaining how my understanding of restitutionary and compensatory orders evolved over the course of writing this book. This explanation outlines the chapter’s main themes and, at the same time, my reasons for seeking a new interpretation of restitutionary and compensatory orders. The chapter’s main goal is, of course, to persuade readers to accept this interpretation. However, its secondary goal, almost as important, is to persuade readers that the questions that motivated it are important.

A. Tertium Quid

The initial motivation for this chapter was not a particular understanding of the meaning or importance of justice. Nor was it motivated by my argument (in (p.224) Chapter 1) that justice is a special concern of remedial law: that argument applies to the entirety of remedial law. Rather, the initial motivation arose from my attempts to fit restitutionary and, later, compensatory orders into my existing categories of rights-threat-responding and wrong-responding orders.

i. Restitutionary Awards

As Peter Birks repeatedly stressed, the striking feature of restitutionary awards is that they can be obtained without an allegation that the defendant committed or threatened to commit a legal wrong. Birks illustrated this feature with the now-famous case of Kelly v Solari.1 Mr Solari had died, and his wife, as the estate’s executor, made a claim under his insurance policy. The insurance company paid out, but then discovered that the policy had lapsed and, therefore, that they had not been obliged to make the payment. The court ordered the widow to repay the money to the insurance company. It did not matter that Mrs Solari was also unaware that the policy had lapsed and was, in all other respects, innocent in the transaction. If the money ‘is paid under the impression the truth of a fact which is untrue, it may, generally speaking, be recovered back’.2

Mrs Solari’s liability to repay the money arose merely from her receipt of it. And, without attempting to provide anything like a theory of ‘injustices’, it seems clear that this receipt is plausibly described as giving rise to an injustice. Mrs Solari obtained a sum of money that would, had no mistake been made, have remained with the insurance company. The standard legal label for this event is an ‘unjust enrichment’.

However, along with Birks and (so far as I could tell) other writers, I initially assumed that unjust enrichments of the kind exemplified by Solari gave rise first and foremost to substantive duties to make restitution. In other words, I assumed that the rules governing defective transfers were part of substantive law. In this view, the only role for remedial law is to stipulate that, if a substantive duty to make restitution is not performed, the transferee can obtain an order that it be performed.

I first questioned this assumption following the publication, in 2011, of the American Restatement (Third) of Restitution and Unjust Enrichment. Given my interest in the distinction between duties and liabilities, I was struck by the Restatement’s practice of consistently describing unjust enrichments as giving rise to liabilities (not duties) to make restitution.3 My doubts increased when I realised that my reasons for rejecting the existence of substantive duties to (p.225) pay damages (e.g. no damages for failing to pay damages, the unknowability of such a duty) also applied to substantive duties to make restitution. Further (and drawing upon Chapter 5’s analysis of the differences between rules and orders), it seemed that the usual normative objection to restitutionary duties—namely, that they impose affirmative legal duties on innocent parties—could be avoided, or at least minimized, if those duties were imposed by orders, not rules. For these and related reasons, I came to the view that the common law does not recognize a substantive duty to make restitution. And, given that restitutionary orders are clearly not responses to wrongs, it followed, I concluded, that there are at least some orders that could not be accommodated within Chapters 7 or 8. However, my initial view was that orders to reverse defective transfers (and other analogous unjust enrichment reversing orders) were unique in this regard.

ii. Compensation for Non-Wrongs

My first clue that this view was too narrow came from reflecting on liabilities in the border areas between unjust enrichment law and tort law. In Great Northern Railway Co v Swaffield,4 the defendants sent a horse for transport on the claimants’ railway. The defendant had not arranged for a representative to pick up the horse at the destination station, so the claimants arranged for it to be kept in a nearby stable. When the defendants eventually picked up their horse, they refused to pay for the stable charges. The claimants paid and then sued the defendants for their expenses. The court ordered the defendant to compensate the claimants. Finding that the claimant had no choice but to stable the horse, Kelly CB held that ‘a right arises in them [the claimants] to be reimbursed those charges which they have incurred for his benefit’.5 Swaffield and other ‘necessitous intervention’ cases are similar to defective transfer cases, in that the defendant has not committed or threatened to commit a wrong. But the requirements for a defective transfer (such as a mistake) are missing, and the remedy is compensatory. In Swaffield, the claimant’s benefit was the horse’s entire value.

Another set of awards that seemed not to fit within Chapters 7 or 8 are, today, based in legislation. As I mentioned in Chapter 4, most common law jurisdictions have legislation authorizing courts to issue orders dealing with maintenance and the division of matrimonial property following the breakdown of a relationship. In England, Section 23(1)(c) of the Matrimonial Causes Act 1973 (the ‘1973 Act’) provides that, following divorce or similar proceedings, courts (p.226) may make ‘an order that either party to the marriage shall pay to the other such lump sum or sums as may be so specified’. There is no suggestion in the Act or elsewhere that claimants seeking such orders must show that the defendant had a duty, prior to the order, to do what the order requires, much less that the defendant breached this or any other duty owed to the claimant. In Lord Denning’s words, ‘. . . there is no right to maintenance, or to costs, or to a secured provision, or the like, until the court makes an order directing it’.6 Another example is an order under s 2(1) of the Civil Liability (Contribution) Act 1978 (the ‘1978 Act’). Under this Act, courts may order a defendant co-tortfeasor to pay a sum to the claimant co-tortfeasor to offset the latter’s liability for damages. As with the 1973 Act, there is no suggestion in the 1978 Act that claimants seeking such orders must show that the defendant had a duty, prior to the order, to do what the order requires. Rather, s 2(1) merely stipulates that the contribution ‘shall be such as may be found by the court’.

The above remedies are often ignored or marginalized by private law scholars. However, the next group on my list of ‘remedies in search of a home’ were difficult to ignore. In my research for Chapter 6, I noticed that courts sometimes award compensatory damages ‘in lieu of specific relief’ because of the substantive undesirability of the requested action (typically, its disproportionate cost). In cases like Dennis v Ministry of Defence,7 Boomer v Atlantic Cement Co,8 and Tito v Waddell (No 2),9 the court’s decision to refuse specific relief is in substance a decision to award damages notwithstanding that the court regarded the defendant’s alleged wrongdoing as permissible and, therefore, not actually a wrong. As I argued in Chapter 6, the damage awards in these cases are not responses to a rights-threat or a wrong.

Further, and unsurprisingly, such awards did not appear to be limited to cases where the claimant had requested specific relief. A famous, if unusual, example is the American case of Vincent v Lake Erie Transportation Co.10 The defendant moored his ship, without permission, to the claimant’s dock to prevent it being destroyed in a violent storm. The court held that, while the defendant acted entirely reasonably (there was no other way to save the ship and crew), he must nonetheless compensate the claimant for damage caused to the dock. Other cases in which the courts award compensatory damages notwithstanding that the defendant neither committed nor threatened to commit a wrong include those where the claim is based on vicarious liability, the rule in Rylands v Fletcher, or an innocent misrepresentation.

(p.227) iii. Compensation for Wrongs

The conclusion that I drew from the above cases is that a significant number of compensatory damages awards did not belong within either Chapter 6 or 8. However, my initial assumption was that these awards were outliers. I assumed that the vast majority of compensatory orders—which were indisputably directed at wrongdoers—were responses to wrongs. My view changed, however, as I developed the ideas that grew into Chapter 7. As we saw in Chapter 7, the clearest examples of wrong-responding remedies are exemplary damages and nominal damages, and the next clearest examples are a variety of awards that, like exemplary and nominal damages, are not compensatory (e.g. user damages, gain-based damages, and waiver damages). Precisely because these awards are not fixed at the claimant’s loss or, as with the substitutionary awards that I discussed in Chapter 6, at the price of obtaining substitute performance, they must be responses to the defendant’s wrong qua wrong. Further, the varied and sometimes highly discretionary approaches to quantifying these awards was exactly what we should expect if they are wrong-based. In contrast to rights-threats and injustices, there is no natural or logical response to wrongs: as in the case of criminal wrongdoing, the law’s response to private wrongdoing is ultimately a matter of choice and convention.

As my research progressed, I came to the view that compensatory damages do not fit this pattern. The most obvious difference is that they are fixed at the magnitude of the claimant’s loss. Moreover, the courts have developed a massively detailed set of rules for determining a sum that—if the wrong-based interpretation of compensatory orders is correct—is ultimately arbitrary. In addition, these rules limit compensatory awards in all sorts of ways—remoteness, mitigation, contributory negligence, and so forth—that appear to have nothing to do with the wrongfulness of the defendant’s wrong. Finally, the courts and legislatures’ descriptions and explanations of the rules for assessing compensatory orders appeared inconsistent with a wrong-based interpretation. Rather than asking whether a compensatory award is an appropriate way to mark, condemn, represent, redress, or otherwise respond to the defendant’s wrongdoing, the courts and legislatures ask whether it is fair, just, or reasonable to attribute the claimant’s loss to the defendant.

I eventually came to the view, then, that even in cases where defendants had committed wrongs, orders to pay compensatory damages awards were different from the wrong-based awards that I discuss in Chapter 7. But where did they belong? Not in Chapter 6, because there is no substantive duty to pay compensatory damages (and therefore no substantive right that defendants could threaten). I entertained the possibility of a fourth category; however, the similarities to the restitutionary awards in my third, yet unnamed, category were too striking. In each case, the courts’ focus is the outcome of a transaction between the parties: ‘When tort law holds persons responsible for outcomes, its task is primarily allocative, in the sense that assigning particular consequences to particular persons is a way of (p.228) determining who will bear which costs’.11 The same is true of defective transfer cases, except that the outcome being allocated is a gain, not a loss. And, in each case, the law’s response to the undesirable outcome is to eliminate it by attributing the relevant outcome—the loss or the gain—to the other party. One significant difference, of course, is that, in the case of most compensatory damages awards, the award is conditional upon proof of a wrong. Still, it seemed to me that when these awards are compared to the wrong-based awards discussed in Chapter 7, the wrong is merely a condition, not the underlying reason, for the award. But what is that reason?

iv. The Concept of an Injustice

The answer, it seemed obvious, is that the reason for these awards must be the ‘injustice’ (or some closely related concept) of the relevant outcome.12 Courts and commentators consistently explain restitutionary orders using a cognate concept (‘unjust enrichment’), and courts regularly invoke justice when they are explaining the rules governing compensatory orders. The ‘overall object of tort law’, Lord Bingham wrote, ‘is to define cases in which the law may justly hold one party liable to compensate another’.13 More importantly, the subject matter of justice is the subject matter of the rules governing restitutionary and compensatory damages: the allocation of gains and losses. The adjective ‘unjust’ is characteristically used to describe states of affairs in which a loss or gain has been unfairly imposed, distributed, or allowed to persist.14

(p.229) Admittedly, the concept of an injustice is sometimes understood in a broader sense. To say that something is unjust is sometimes understood to mean nothing more than that it is undesirable and that someone, usually the state, should fix it. In this understanding, both rights-threats and wrongs qualify as injustices. However, this usage is unusual.15 It is more common—and more useful—to describe trespasses, nuisances, and contract breaches simply as wrongs (and to describe threats to do these things as threatened wrongs). If I responded to a punch in the nose by exclaiming ‘how unjust!’, it might be wondered if the punch had affected my linguistic abilities. We characteristically employ the term injustice to describe actions or states of affairs that are in some sense unfair, typically because a loss or a gain (broadly understood) has been unfairly allocated, distributed, or allowed to persist. Save for the exceptional case where a punch in the nose is the outcome of a decision to allocate that punch to one of a number of potential candidates, the wrongness of punching has nothing to do with allocation. In contrast, we ask whether the tax system is just because we want to know whether it allocates the burden of taxation fairly. Similarly, we ask whether the bankruptcy laws are just because we want to know if they fairly allocate bankrupt individuals’ assets. And we ask whether a judicial decision was just because we want to know if the proper parties won and lost. Justice is central to law because adjudication is central to law—and adjudication is always about allocation.16 In the law, to say that something is unjust is nearly always to say that it is unfair.

Thus understood, injustice remains a broad concept. Is it too broad to do any explanatory work? It seemed clear that the explanations for why the outcomes that trigger restitutionary and compensatory orders are unjust would be very different. The facts that must be proven to obtain restitutionary orders have little in common with those that must be proven to obtain compensatory damages (at least in the standard case where the latter’s availability is conditional on a wrong). ‘Transferring money by mistake’ is very different from ‘carelessly causing pecuniary losses’. However, it also seemed that the concept of an injustice is sufficiently explanatory for my purposes. This book’s classificatory scheme seeks to characterize, at an abstract level, the facts that claimants must establish to obtain private law orders. As we saw in Chapters 6 and 7, these characterizations employ normative concepts (‘rights’ in Chapter 6, ‘wrongs’ in Chapter 7). However, I say little about why we have the rights we have or why certain actions count as infringing those rights. Space considerations aside, my explanation of remedial law’s structure is deliberately pitched at a level that, if successful, can accommodate different views of the law’s normative foundations.17

(p.230) However, if I had tried to explain why we have the rights we have and why certain actions count as infringing those rights, the explanations would be as varied as the explanations for why mistaken payments and carelessly caused injuries give rise to injustices. There is little in common between the facts that must be proven to establish—to mention just four examples—a battery, a breach of contract, a defamatory statement, and the breach of a statutory duty. And the explanations for why these actions are wrongs are similarly diverse.18 What unifies the concept of a legal wrong is not the tests that the law employs to establish wrongs. Rather, the unifying feature is the characterization of the relevant event as a wrong, that is, as something that ought not to happen. The common feature of punching someone in the nose, breaking a promise, uttering a defamatory statement, and acting contrary to a statute is that they are all actions that, in the law’s view, we should not do. But if we ask why we should not do these things, the explanations are different. The same is true of injustices. The concept of a wrong (or of a right) and the concept of an injustice operate at the same level. The main difference between them is that the concept of a wrong is used to identify undesirable behaviour, while the concept of an injustice is used to identify undesirable consequences of behaviours.

For similar reasons, I was not concerned that the concept of an injustice seems too vague, at least as I have defined it, to serve as a judicial standard. My interpretation of restitutionary and compensatory orders does not suppose that courts contemplating such orders ask directly whether the claimant has suffered an injustice. Although the required state of affairs is, I argue, properly described as an injustice, it is the state of affairs, not the injustice, that claimants must prove.19 And the (p.231) rules that identify the relevant states of affairs, as well as those that explain what courts should do when claimants prove those states of affairs, are as determinate as the parallel rules in respect of rights-threats and injustices.20 Further, my interpretation does not suppose that private law provides remedies for all injustices. There are innumerable injustices that lie outside private law’s concerns; for example, unjust tax laws, procedural injustices, unjust punishments, and unjust results in sporting or other competitions. Again, there is a parallel with the causes of action examined in Chapters 7 and 8. As is true of the concept of a wrong and the concept of a rights-threat, the law has a tightly circumscribed conception of what counts, legally, as an injustice and complex tests for identifying such injustices.21

(p.232) v. Why Orders?

There remained one further question. As I said, I was convinced (and remain convinced) that the only duties to correct injustices recognized by the common law are those created by court orders. The question was why the law limits itself in this way. If the law believes that individuals like the defendants in Kelly, Vincent, and Swaffield should make restitution or pay compensation, why does it not pursue this aim by imposing rule-based substantive duties to make restitution and pay compensation. If individuals subject to such duties fail to comply with them, then, as in any other case where substantive duties are ignored, there is a role for court orders (as we saw in Chapter 6). But the law’s first response, it might be thought, should be to impose rule-based substantive duties. Why, then, does it limit itself to court-ordered remedial duties? My explanation for why the law does not recognize substantive duties to pay nominal, exemplary, and other wrong-responding damages does not provide an answer because, as we saw in Chapter 7, that explanation is limited to wrong-based remedies.

The answer, which I arrived upon after reflecting on Chapter 5’s arguments, is that ‘correcting injustices’ is not a proper aim for a substantive duty.22 The common law’s substantive duties are duties not to wrong others; in particular, duties not to wrong others by interfering with their person, property, liberty, or contractual promises made to them. Battering, trespassing, breaking contractual promises, carelessly injuring others, causing nuisances, and so forth are all non-consensual interferences of this kind. They are ways of wronging others by infringing their rights to their person, property, and so forth. A substantive duty to correct an injustice, if it existed, would be very different from these duties.23 Failing to correct an injustice is not a wrong, or at least not a wrong in the sense that battery and so forth are wrongs. If I fail to return money that you paid me by mistake or by fail to compensate you for a loss that I inflicted on you, my failure is not an interference with your person, property, liberty, or a promise that I made to you by. More generally, you cannot point to any right of yours that I am infringing by not making restitution or paying damages. All that I have done is fail to correct an injustice.24

(p.233) This observation does not deny that correcting injustices is valuable. It is because correcting injustices is valuable that we are praised when we return mistaken payments and when we compensate others whom we have harmed. But the fact that it is valuable (or ‘virtuous’, ‘commendable’, ‘useful’ ‘beneficial’, and so forth) to do something is rarely a sufficient ground for the common law to recognize a substantive duty to do that thing.25 If it were sufficient, the common law would recognize a substantive duty to rescue and a great many other duties to assist others. There are exceptions to this principle (even the common law recognizes certain ‘good faith’ duties), but in general, our substantive duties are duties of non-interference.26 The question of why substantive duties are limited in this way is the subject of a voluminous literature. Almost everything that has been written about the foundations, structure, or history of private law touches on this question. I did not intend to add to this literature. It was sufficient that the limitation is well established and is defensible, at least in principle, from a variety of philosophical perspectives. There is nothing odd about the common law’s decision not to recognize substantive duties to correct injustices.

For my purposes, the difficult question was not why the common law does not recognize substantive duties to correct injustices. Rather, the difficult question was why it recognizes court-ordered (‘remedial’) legal duties to correct injustices. Although orders (unlike duty-imposing rules) are not propositions about the existence of duties, they indirectly give rise to legal duties. As we saw in Chapter 5, the addressees of court orders have legal duties to obey those orders, and, therefore, have derivative legal duties to perform the actions that orders stipulate. The question, then, was how the courts’ powers to create court-ordered duties to correct injustices squares with the above explanation for why the law refuses to impose rule-based (substantive) duties to correct injustices?

(p.234) The answer, unsurprisingly, is that rules, and in particular the duty-imposing rules that are our present concern, are different from orders. As we saw in Chapter 5, the kinds of reasons capable of justifying rule-based duties are qualitatively different from those capable of justifying court-ordered duties. Rule-based duties are owed to other individuals. The duty created by the rule that ‘everyone has a duty to perform their contractual promises’ is owed to contractual promisees. It follows that this rule can only be justified if it makes sense to suppose that contractual promisees have legal rights to the performance of contractual promises made to them. It is entirely plausible to suppose that such rights exist. But the situation is different if we are trying to justify a rule-based duty to correct injustices. As we have just seen, it is implausible to suppose that individuals have rights that other individuals correct injustices that they have suffered.

However, court-ordered duties are different. Orders are commands (‘it is ordered that the defendant pay the claimant £100’). The only duties to which commands give rise are duties to obey whoever issued them, which, in this case, is a court. This difference explains why courts regularly employ orders to direct defendants to do things that, because they are not plausible subject matters of a substantive right, are not appropriate subject matters for rule-based duties. Orders to pay exemplary damages, nominal damages, and the other forms of wrong-based damages all fall into this category. A wrongdoer’s non-payment of exemplary damages is not an interference with the victim’s person, property, liberty, or anything else to which the victim might plausibly be said to have a right. Nonetheless, there may be good reasons for wrongdoers to pay wrong-based damages to their victims (which I explored in Chapter 7), reasons that can justify a court ordering such payments.27

Court-ordered duties to make restitution or pay compensation may be justified in a similar way. The main difference is that their rationale lies not in the value of condemning, deterring, representing, or otherwise redressing a wrong: their rationale lies in the value of correcting injustices.28 My own view is that the value of correcting injustices is intrinsic, not instrumental—the fewer injustices in the world, the better—but for the present argument all that matters is that correcting injustices is valuable. The courts order defendants to cure injustices because, (p.235) notwithstanding that failing to correct an injustice is not a legal wrong, curing injustices is valuable.

This understanding of correcting injustices is reflected in our broader legal and social practices. It explains, for example, why the state sometimes corrects injustices through the actions of third parties. In England, individuals who have suffered personal injuries may obtain social security benefits that they would not have obtained but for those injuries. Insofar as such benefits reduce the amount that the injured party could have recovered in a tort action, the Social Security (Recovery of Benefits) Act 1997 provides that the state may recover their value from the tortfeasors. In these cases, the injustice is corrected by payments to and from a state body. If the victims of injustices had rights that those who caused the injustice must correct it, this method would be impermissible.

More importantly, this understanding of correcting injustices explains why common law jurisdictions frequently replace tort compensation with statutory compensation. Many common law jurisdictions have ‘workers compensation schemes’ or ‘motor vehicle compensation schemes’, whereby anyone injured at work or in a motor vehicle accident may obtain compensation from the scheme.29 The most extensive scheme of this kind (though still only applying to personal injury claims) is New Zealand’s Accident Compensation Scheme, introduced in 1974. Although such schemes are typically funded largely, or wholly, by potential injurers, they are not alternative systems for correcting injustices because, inter alia, the injurers are not held liable, even indirectly, for their victim’s losses. These schemes are justified on other grounds, such as efficiency or fairness. Needless to say, compensation schemes would not be tolerated if the state or anyone else had a duty to correct injustices in the same way that it has a duty to, say, not kill innocent citizens. Killing innocent citizens is impermissible, even if it might be valuable in various ways (e.g. reducing demands on the health care system). But trading off the value of correcting injustices for other values is permissible, at least in principle. The same trade-offs are made in non-legal contexts. When my children were younger, my wife and I were the closest equivalent to a court in our household. If, following a sibling dispute, one of our children wanted ‘justice’, they would come to us. But there were many occasions when injustices went uncorrected because we were away, asleep, too tired, or just because we thought that preparing a good meal was more important than ensuring that justice was done.

Thus understood, the decision to give courts the power to correct injustices is a policy decision. The state could legitimately spend the resources that it devotes to supporting private law compensatory and restitutionary orders on statutory compensation schemes, social security schemes, or on building hospitals. For example, (p.236) the state could legitimately decide that courts will cease hearing applications to reverse defective transfers, putting the savings into providing additional family law courts. I am not advocating such changes; I merely note that it is because correcting injustices is valuable, not mandatory, that we can contemplate these changes at all. To avoid misunderstandings, I should also add that this understanding of correcting injustices does not lead to the view that courts should take policy considerations into account when determining what is required to correct an injustice. Once the decision is made to allow courts to issue injustice-correcting orders, policy considerations largely, if not entirely, drop out of the picture.30 Determining what counts as an injustice and what is required to cure injustices are not policy questions.

Finally, does it matter that the addressees of injustice-correcting orders may be innocent of any wrongdoing? It might be thought that, if you have done nothing wrong, the law should leave you alone. The short response to this objection is that it applies equally to alternative explanations of restitutionary and compensatory orders. The alternative explanations assume that restitutionary and compensatory orders confirm substantive duties.31 If Mrs Solari’s innocence is an objection to ordering her to make restitution, it is equally an objection to imposing upon her a substantive duty to make restitution. Whether the duty is imposed by a ruling or a rule, it is imposed (so this objection goes) on an innocent person. It is true, of course, that the innocent addressee of a judicial order will incur the costs and inconveniences of litigation.32 But, again, this objection applies to the alternative explanations. If it is not possible to determine in advance of litigation the existence and content of substantive duties to correct injustices (as I argue), then whether I am subject to a substantive duty to correct an injustice or am merely liable to being ordered to do so, I am in the same position regarding the costs and inconveniences of litigation. Under either scenario, I cannot be certain of my duty until a court makes a ruling.33

For all of these reasons, then, I eventually came to the view that restitutionary and compensatory orders share a similar cause of action, best described (p.237) as proof of an injustice. The remainder of this chapter elaborates upon these reasons, focusing in particular upon the doctrinal claims that I have yet to defend.

B. Restitutionary Awards

If you have transferred money or other property to me in circumstances where your autonomy was impaired, a court will normally grant you an order directing me to return the money or property to you (or to return a sum equal to the property’s value). ‘Impairment’ is typically established by showing that the transfer was motivated by a mistake, duress, undue influence, fraud, incapacity, ignorance, compulsion, or necessity.34 Thus, as happened in Solari v Kelly, if you transfer money to me because you mistakenly believe that you owe me the money, you can normally obtain a court order directing me to return the money. The cause of action for such orders, I argue, is an injustice.

For convenience, I describe these orders as ‘restitutionary’, but lawyers typically use this label in a broader sense. The category of restitutionary orders is typically understood to encompass orders to return money paid in anticipation of a contract, orders to return money paid under a contract that subsequently failed, orders to pay for mistaken improvements, orders to pay for requested services, and a variety of other orders that, in broad terms, are thought to reverse unjust enrichments. It is likely that the cause of action for some of these orders is also an injustice, but I do not discuss them here. Like those rules, the rules that I discuss in this section are usually described as part of the law of unjust enrichment. Although I refer occasionally to unjust enrichment law, I more commonly refer either to the law governing defective transfers or to the law governing restitutionary orders. The former is employed when I want to make clear that my focus is a subset of unjust enrichment law, while the latter is employed when I want to make clear that my focus is remedial law.

My argument that restitutionary orders are responses to injustices rests upon two propositions. The first is that the common law does not recognize a substantive duty to make restitution. If such a duty existed, the cause of action for restitutionary orders would be the same as for orders for a sum due and other replicative orders—a rights-threat. The second proposition is that the state of affairs that claimants must prove to obtain a restitutionary order is appropriately described as an injustice. I defend these propositions in turn.

(p.238) i. Duty-Confirming or Duty-Creating?

As was true of damages orders, the first question to ask about restitutionary orders in a book on remedies is whether they confirm substantive duties (the ‘duty-confirming view’) or create new duties (the ‘duty-creating view’). If the duty-confirming view is correct, then almost none of the law that appears to deal with restitutionary orders is in fact about such orders. In other words, almost none of this law is remedial law. Remedial law is comprised of rules governing the availability and content of judicial remedies. If the duty-confirming view is correct, the remedial part of the law of restitution reduces to one rule: if the defendant has not fulfilled a substantive duty to make restitution, the court will order that the duty be fulfilled. The rest of the law of restitution, in this view, is substantive law: it is a set of rules specifying when substantive duties to make restitution arise and their content. A remedies book that adopted this view would still need to explain why courts order defendants to comply with unfulfilled duties to make restitution. But, as we saw in Chapter 6, that question (which arises in respect of all replicative remedies) is easily answered: courts issue such orders to motivate recalcitrant defendants to perform their substantive duties. In other words, they order them in response to rights-threats.

In contrast, if the duty-creating view is correct, then the entirety of the law of restitution is remedial law. In this view, the law of restitution is comprised of rules that specify the circumstances in which courts will make restitutionary orders, as well as their content. Further, and critically, the explanation of the rules that make up the law of restitution will differ depending on whether one adopts the duty-confirming or duty-creating view. As I explained above (and in more detail in Chapter 5), the normal reason that the law employs orders rather than rules to impose duties is that the duty’s subject matter is not appropriate for a substantive rule. The kinds of reasons that explain exclusively court-ordered duties differ, therefore, from those that explain rule-based duties.

With rare exceptions, textbooks on remedies do not ask whether restitutionary orders are duty-confirming or duty-creating.35 The same is true of the literature on restitution and unjust enrichment.36 It is clear that a number of leading writers assume that restitutionary orders are duty-confirming (notwithstanding that their language often suggests the opposite: see below),37 yet these writers have said little to defend this assumption. In support of his argument that a substantive duty to make restitution arises at the moment of transfer, Birks noted that the transferee’s cause of action is complete following transfer.38 But this rule is also consistent with (p.239) the duty-creating view. Indeed, the most plausible version of the duty-creating view assumes that the cause of action for restitutionary orders arises at the moment of transfer, since the injustice that such orders are meant to correct arises at the moment of transfer.

The little that has been said in defence of the duty-confirming view is not, of course, a reason to reject it. The reason to reject it, as I explain below, is that it is inconsistent with what courts do, courts’ explanations of what they do, and the role of substantive duties in our legal system.

ii. Language

The duty-creating view of restitutionary orders is, in one respect, entirely conventional. Although courts occasionally refer to substantive duties to make restitution,39 they typically describe individuals who have been unjustly enriched as liable to make restitution: ‘the agent is liable to make restitution’.40 Writers, including those who otherwise endorse the duty-confirming view, frequently use similar language. Birks regularly described mistaken payments and similar events as giving rise to liabilities to make restitution.41 The recent Restatement Third: Restitution and Unjust Enrichment consistently adopts similar language. The Restatement’s opening section states: ‘A person who is unjustly enriched at the expense of another is subject to liability in restitution.’42

(p.240) The courts and writers who employ this language do not appear to attach particular significance to it. It is possible, therefore, that the liabilities to which they refer are liabilities to fall under a future duty. It is also possible that they simply understand ‘liability’ to mean ‘duty’. However, the more straightforward interpretation is that their language reveals an assumption, albeit perhaps unconscious, that the rules that they are discussing are rules about remedies. As I have stressed throughout this book, a liability is different from a duty. Duties describe things that we must do: liabilities describe things that may be done to us. To describe defendants as liable to make restitution suggests that they do not have duties to make restitution, but, instead, that they are only liable to be ordered by a court to make restitution.

iii. Doctrine

The doctrinal evidence does not point consistently in one direction, but on balance it supports the duty-creating view.

No Damages for Failure to Make Restitution

Claimants who incur losses that they could have avoided had the defendant made restitution prior to judgment cannot obtain compensatory damages for these losses. This rule is difficult to reconcile with the duty-confirming view: if there is a duty to make restitution, failing to comply with that duty must be a legal wrong. The common law’s standard response to a legal wrong is to hold the wrongdoer liable to pay damages. The courts’ refusal to award damages for failing to make restitution prior to an order to make restitution strongly suggests that this failure is not a wrong, and thus that there is no substantive duty to make restitution.

This rule cannot be explained on the ground that the recipients of defective transfers may be entirely innocent, even passive, beneficiaries. The beneficiary’s innocence is a reason to question the alleged duty’s existence. As the absence of a duty to rescue demonstrates, the common law is extremely hesitant to impose substantive duties to benefit others. However, if a substantive duty to make restitution exists—as the duty-confirming view asserts—then breaching it must be a legal wrong, and damages should follow.

Nor can the unavailability of damages be explained on the ground that the alleged substantive duty to make restitution has, it appears, no date for performance attached. Again, the absence of a date for performance is a reason for questioning the alleged duty’s existence. A duty to perform a positive action that has no due date could never be breached. But, if the law wished to enact such a duty, it would be relatively simple to stipulate that it must be performed within a reasonable time. (p.241) When contracts are missing dates for performance, the law implies a term that performance must happen within a ‘reasonable’ time.43

Finally, the rule cannot be explained on the ground that courts may award interest, calculated from the date of enrichment, on restitutionary awards.44 It is true that interest awards are equal in most cases to whatever damages might have been awarded for failing to make restitution immediately following an enrichment. But in some cases they are less than this amount. As I explained in Chapter 7, when considering the parallel issue in the context of compensatory damages, losses arising from a failure to make timely payment may exceed the interest rate. It is because of this possibility that claimants suing for non-payment of an ordinary debt may recover damages in addition to interest.45 If there were a substantive duty to make restitution, we would expect claimants suing for restitution to have the same option.

Interest Awards

There are three features of the law of restitution that might be thought to support the duty-confirming view. The first is the just-mentioned availability of interest on restitutionary awards. The availability of interest, assessed as of the date of the defendant’s enrichment, might be thought to show that the law presumes that defendants should have made restitution immediately after they were enriched. However, this conclusion follows only if the claimant’s losses from the non-performance of the alleged duty are equal to the loss of interest—which, as just explained, is not always the case. The more plausible reason for awarding interest on restitutionary orders, which has been adopted by most courts and commentators, is that the interest represents part of the defendant’s enrichment.46

Change of Position

Restitutionary remedies may be refused or reduced if the defendant changed position in reliance on the enrichment.47 This defence is only available for bona fide changes of position, that is, only in cases where the defendant was unaware at the time of the change that the enrichment arose from a defective transfer.48 It might be (p.242) thought that if there is no substantive duty to make restitution, it should not matter when the defendant becomes aware of the impairment. If the duty only arises when an order is issued, why does it matter what happens earlier?

An initial observation is that the duty-confirming view is vulnerable to a parallel objection. If there is a substantive duty to make restitution, why should it matter what the defendant does (short of satisfying the duty) after the duty has arisen? If you owe me £1000 under a contract, it is no defence that you did not believe that you owed me this sum, and, as a result, spent all your money on a holiday.

Substantively, the explanation for this rule is that the defendant’s knowledge of the defect plays the same role in the duty-creating view as it plays in the duty-confirming view. The rationale for the change of position defence is that innocent transferees should not be required to return assets that they no longer possess. This rationale does not apply, the argument goes, if the transferee dissipated the assets while aware that the transfer was defective. This rationale is not entirely straightforward because, as I explain below, recipients can never be certain prior to a court ruling that a transfer was defective, or even that the plaintiff was the transferor. However, this issue arises on either view. For the moment, what matters is that to the extent that the conventional explanation is persuasive, it is persuasive whether repayment is required because of a duty or because of a liability to a court order. In either case, what is important is transferees’ knowledge that they may be legally required—whether by a duty or an order—to make a repayment. This knowledge precludes transferees from arguing that the benefit was foisted upon them.

Section 5(4) of the Theft Act 1968

A third and final feature of English law that appears to support the duty-creating model is s 5(4) of the Theft Act, 1968. The section appears to assume the existence of a duty to make restitution:

Where a person gets property by another’s mistake, and is under an obligation to make restoration (in whole or in part) of the property or its proceeds or of the value thereof, then to the extent of that obligation the property or proceeds shall be regarded (as against him) as belonging to the person entitled to restoration, and an intention not to make restoration shall be regarded accordingly as an intention to deprive that person of the property or proceeds.

Section 5(4) is of relatively recent origin, and it applies only to cases involving property obtained by mistake. Further, the section does not strictly impose or apply a duty to make restitution: it merely holds that, if there is such a duty, then failing to comply with the duty is deemed theft. Finally, the substance of the provisions seem prima facie incoherent: s 5(4) supposes that it is possible to steal one’s own property (since the provision is only needed in cases where title to the relevant property (p.243) has passed). The section appears to be based on a misapprehension regarding mistaken payors’ property rights in their payments.49 All that said, s 5(4) assumes that duties to return mistaken payments exist, and it has been applied to cases involving mistaken payments in which title to the transferred property passed.50 To this extent, s 5(4) supports the duty-confirming view.

iv. The Unknowability Objection

The courts’ language and the unavailability of damages for failing to make restitution are strong evidence that the common law does not recognize a substantive duty to make restitution. However, the main objection to the duty-confirming view of restitutionary orders is one that we already encountered when we considered, in the previous chapter, the duty-confirming view of damages. The objection is that a substantive duty to make restitution would be unknowable. The duty would therefore be incapable of fulfilling the most basic role of legal rules—and the most basic requirement of the rule of law—namely, to guide behaviour.

As I mentioned above, defenders of the duty-confirming view generally assume that the duty to make restitution arises at the moment of transfer. It follows that if you mistakenly transfer money into my bank account while I am hiking in the wilderness, far from any form of communication, I have a legal duty, from the moment of the transfer, to return the money to you. Admittedly, it is not impossible for me to comply with this duty. I might somehow guess that you are likely to transfer money to me by mistake while I am away and, as a precaution, send you a cheque that arrives, fortuitously, at the moment you make your transfer. But the need to take such actions to comply with my duty demonstrates its implausibility: the law cannot possibly want me to send you cheques on the chance that you may pay me mistakenly.

Further, and critically, even if it is possible in theory for me to comply with a substantive duty to make restitution, it is impossible for me to know that I have complied and, more generally, to know that I am bound by the duty. The alleged duty is therefore incapable, even in theory, of guiding my behaviour. Suppose that you transfer money into my bank account by mistake. And suppose further that I happen to be viewing my account online at that moment, so that I am instantly aware of the transfer (or, alternatively, assume that the alleged duty arises only when you inform me of the defective transfer, or when you initiate litigation). Even in these circumstances, I cannot be certain that I am under a duty to return the (p.244) money to you until I know two additional things: first, I need to know that it was you who made the payment; and, second, I need to know that you made the payment because of a mistake.

I cannot determine either of these facts on my own. Even in the unlikely event that I happen to witness you making the transfer, I cannot be certain that I must make restitution to you because I cannot be certain that you were not a conduit for the ultimate payor. Three-party restitution cases are common.51 In addition, I cannot be certain that the transfer was defective. To confirm that the payment was made by mistake (or because of a threat, fraud, etc), as opposed to being a gift, I need to know your motivation for the transfer. Yet that information is known only by you. Of course, you could tell me that you were the transferor and that the transfer was made by mistake; but you might not do this, and, in any event, how can I know that you are telling the truth? It cannot be sufficient that your explanation appears to be reasonable (since reasonable-appearing explanations may be unsubstantiated). If you were to ask a court for relief, it would not simply accept your word: the court would require you to present evidence, witnesses, and so on. Why should I accept anything less? And even if you presented such evidence to me, am I then supposed to assess your veracity? If nothing else, I might query my impartiality. Admittedly, when a transfer has been made by mistake, the mistake is usually obvious. But the unknowability objection is conceptual: it is sufficient, for this objection, that there could be circumstances in which it is not possible for individuals bound by the alleged duty to know they are bound. Further, this objection applies regardless of whether the alleged substantive duty arises when the transfer is made, when the claimant requests restitution, when litigation commences, or at any other time.

As I explained when discussing the parallel objection to substantive duties to pay compensatory damages, the unknowability objection is not merely a version of the familiar idea that it is difficult to apply general rules to particular circumstances. The objection is that the relevant facts are in the hands of the duty’s beneficiary. Duties not to carelessly injure others, not to trespass, not to assault, to keep contractual promises, and so forth are not vulnerable to this objection. Even if it can be difficult to determine what is required to take ‘reasonable care not to injure another’, the relevant facts are available to those bound by this duty. A substantive duty to make restitution, if it existed, would be different. To ensure that I comply with such a duty, I would have little choice but to accede to any demand to make restitution. To be sure, if the demand is without basis, I would not have a duty to make restitution. However, I could never be certain of a demand’s validity, no matter how diligently I investigate. In short, while it would not strictly be impossible for me to comply with a substantive duty to make restitution, it would be impossible for me to know when the duty arises, its content, and whether I have complied with it. If (p.245) a substantive duty to make restitution existed, it could not guide those subject to it. Its only role would be to serve as an ex-post explanation for why courts issue orders to make restitution. In this role, a substantive duty to make restitution appears indistinguishable from a liability to being ordered by a court to make restitution.

The corollary of the conclusion that transferees could know when they were bound by a substantive duty to make restitution is that transferors could know when they had a correlative right to receive restitution. A substantive duty to make restitution would be subject to the defence of change of position. The duty would therefore be reduced or eliminated where the transferee innocently and detrimentally changes position in reliance on the transfer. Yet, the facts that establish a valid change of position (e.g. innocently spending a mistake payment on a vacation) are in the recipient’s hands. Thus, just as transferees cannot be sure of the facts that (if the matter went to trial) transferors must prove, transferors cannot be sure of the facts that transferees might offer as a defence. Uncertainty about whether one is the beneficiary of a duty is normally less serious than uncertainty about whether one is bound by a duty. Still, the law’s duty-imposing rules should be able to guide not just those who are bound by them, but also those who benefit from them. Occasional limitations on this ability may be justified, but the duty-confirming view supposes that all restitutionary duties are limited in this way.52

v. Restitution as a Response to an Injustice

The second proposition on which the classification of restitutionary orders as injustice-responding rests is that the state of affairs that claimants must prove to obtain a restitutionary order is appropriately described as an injustice. The primary argument in support of this proposition is that the alternatives—namely, that restitutionary orders are responses either to rights-threats or wrongs—are implausible.

As Kelly v Solari illustrates, the beneficiary of a defective transfer may be an entirely innocent participant in the transfer. It is true that had Mrs Solari agreed to return the payment to the insurance company, then, even if she was innocent at the moment she received the payment, she would no longer be innocent as a result of her failure to perform her promise.53 However, as is true of the typical defective (p.246) transfer case, there was no evidence of such a promise.54 Mrs Solari’s liability arose entirely from the payment. Receiving a payment is neither a wrong nor a threat to commit a wrong.

If restitutionary orders are not responses to either rights-threats or to wrongs, it is difficult to imagine any way of explaining them that does not invoke justice, fairness, or a closely related concept. It is sometimes suggested that the beneficiaries of defective transfers must return what they have received because it would be ‘inequitable’ or ‘unconscionable’ for them not to do so.55 However, this explanation is perfectly consistent with the explanation defended here. The explanation defended here explains why it would be inequitable or unconscionable to retain the benefit of a defective transfer—namely, that the outcome of that transfer is unjust. Indeed, if my account is correct, it is not in the least surprising that writers often invoke ‘equity’, ‘unconscionability’, and similar concepts when explaining restitutionary orders. These concepts are characteristically invoked to describe events or transactions that are undesirable notwithstanding that neither party committed a wrong or a rights-threat, as traditionally understood. In this limited sense, restitutionary orders are equitable: they are responses to something other than a wrong or a threatened right.

(p.247) A second suggestion that is sometimes made is that restitutionary orders are responses to property rights, or, perhaps, more strictly to defective property rights.56 According to this view, in an ideal world a mistaken or otherwise defective transfer would not transfer title to the transferred property. Title would remain with the transferor, who could then bring a simple claim for conversion if the property was not returned.57 However, for a variety of practical reasons (including, in particular, the need to protect third parties who rely on the transfer’s appearance of validity), the law holds that title passes under a defective transfer. The role of restitutionary orders, in this explanation, is to provide an ex-post means by which the law can ensure that title ends up in the hands of the party who, in an ideal world, should never have lost it. More broadly, ‘these claims [to reverse defective transfers] arise as a means of protecting and effectuating the plaintiff’s interest in exclusively determining the disposition of his assets’.58 In other words, defective transfers are defective precisely because they transfer title in situations where, in retrospect, the transferor would have preferred that title not pass.

As James Penner has pointed out, the main difficulty with this explanation is that there could never be rules of transfer, no matter how perfect, that prevented all mistaken transfers: ‘The rules of title transfer, being facilitative, make it the case that they can be used for good purposes or ill . . . It is not the law, but the claimant, who is mistaken, and it is simply not the role of the law, even a perfect law, to make it the case that people cannot effectively make mistakes, for that would undermine, not enhance, their autonomy’.59 Further, if restitutionary orders were responding to imperfect property rights, we would expect them to respond by perfecting those rights, that is, by granting or declaring that the claimant has a property right in the relevant property. Yet the typical restitutionary orders are no more proprietary than orders to pay damages or debts: they merely direct the defendant to pay a sum of money to the claimant.60 If the claimant is insolvent, the defendant must line up with the other creditors. Finally, the subject matter of a defective transfer may not be property.61 In a typical mistaken payment case, what is transferred is not a property right, but a personal claim against a bank. We do not ‘own’ the money in our bank accounts; we merely have a right, the same as an ordinary debtor’s right, to be paid this money by the bank. In short, while property-based explanations of defective transfers can explain why defective transfers are defective, they fail to identify the problem to which restitutionary orders are responding.

(p.248) The only other attempt to explain restitutionary orders of which I am aware interprets them as a kind of insurance for potential transfers. According to what I will describe as ‘instrumental’ theories, the availability of restitutionary orders facilitates the practice of making transfers by allowing transferors to undo transfers that, in retrospect, diverge from their intentions. Absent restitutionary orders, this argument goes, potential transferors would either avoid potentially beneficial transfers or devote excessive resources to making sure that they do make a defective transfer. Most advocates of instrumental theories assume that transfers are beneficial for economic reasons,62 but non-economic instrumental theories are also possible.63

Instrumental theories are not in competition with this chapter’s account of restitutionary orders.64 It is true that instrumental theories are not concerned with ‘injustices’, at least insofar as ‘justice’ means something that is intrinsically valuable (instrumental theories regard concepts such as ‘rights’ and ‘wrongs’ in the same way). However, instrumental theories still need to identify the events that give rise to restitutionary orders. And to do this, instrumental theories, like other explanatory theories, invariably utilize legal concepts. This move is inevitable because if you are trying to explain the law (or anything else) you need to identify what it is that you are explaining in terms that those familiar with the concept can understand. If one attempted to explain contract law without using the word ‘contract’, the resulting explanation would not be an explanation of contract law. This requirement explains why instrumental theories of contract law are invariably described as just that—instrumental theories of ‘contract law’. In short, while instrumentalists might treat concepts like ‘right’, ‘wrong’, and ‘injustice’ as mere placeholders, their explanations (like all explanations of the law) cannot avoid relying on these or similar concepts to describe what they are explaining.

In the case of restitutionary orders, instrumental theories typically describe them as responses to defective transfers or something closely analogous (e.g. ‘impaired’ transfers). If asked to generalize this description, instrumental theories invoke concepts such as ‘unjust enrichment’. And if asked to generalize further, instrumental theorists might well agree that a defective transfer is just one example of the broader category of ‘injustices’. Of course, there is no guarantee that particular instrumental theorists would agree to this classification. However, if they agree that a defective transfer need not involve either a wrong or a rights-threat, then they would, presumably, agree that restitutionary orders belong in a separate category. Instrumental theories have explanations for why there should be remedies for wrongs and rights-threats. But, if my arguments are correct, those explanations will not work for defective transfers (or consequential losses). Thus, another (p.249) label is required. Of course, an instrumentalist might argue that I have used the wrong concepts or drawn the borders between them in the wrong places. But, if this chapter’s classificatory argument is persuasive, it should be persuasive to instrumentalists and non-instrumentalists alike.

As I mentioned above, if restitutionary orders are not responses to either rights-threats or to wrongs, it is difficult to imagine any way of identifying the events to which they are responding that does not invoke justice, fairness, or a closely related concept. Substantively, claimants seeking restitutionary orders must show that the defendant obtained a benefit under a transfer in which the claimant’s autonomy was impaired because of a mistake, fraud, duress, compulsion, necessity, or something similar.65 The natural explanation for why you should not be able to keep something as, in effect, a gift that was not intended as a gift is that it is unfair to keep it. This explanation is, of course, entirely orthodox. Lawyers standardly describe the outcome of a defective transfer as an ‘unjust enrichment’. As the term is ordinarily understood, an unjust enrichment is the same as, or is at least included within, what I have described as an injustice. To say that something is an unjust enrichment is just another way of saying that it is an injustice, and both are just ways of saying that a gain or loss (in this case, a gain) has ended up in the wrong hands.

C. Compensatory Damages

As understood here, compensatory damages are awarded on proof of, and set at the value of, pecuniary losses that the claimant suffered as a consequence of the defendant’s actions. These actions are typically described as torts or contract breaches (though I will argue that sometimes they are not wrongs, strictly understood). Consequential losses normally consist in actual or anticipated out-of-pocket expenses and foregone profits or earnings. Common examples include damages for medical expenses and loss of earnings arising from a physical injury and damages for loss of profits arising from a broken contract. Importantly, damages for pain and suffering, mental distress, loss of reputation, or other intangible harms are not compensatory, as this term is used here. Although pain, distress, and so forth may be a consequence of the defendant’s wrong, they are not losses, and (for the same reason) they cannot be eliminated by monetary awards. I discussed damages award for intangible harms in Chapter 7, where I argued that they are responses to wrongs.66

(p.250) Compensatory awards, I argue, are different: like restitutionary awards, they are responses to injustices, not wrongs. Indeed, I argue that compensatory awards are, in certain respects, the mirror image of restitutionary orders. In each case, if the claimant is successful the court responds to an unjust outcome by ordering the party who benefited from the injustice to correct it. The main difference between restitutionary and compensatory awards is that, while the former are responses to unjust gains, the latter are responses to unjust losses.

It is true that courts and commentators traditionally describe, or at least classify, compensatory damages and indeed damage awards generally as responses to wrongs. It is also true that, in most (but not all) cases in which compensatory awards are issued, the defendant has committed a wrong, and, further, the claimant has to prove this wrong to obtain the award. However, I argue that the wrong in these cases is merely a condition of the award, not the reason for issuing it. If the wrong were the reason, the rules for assessing compensatory damages and the rules that establish when compensatory awards are available would be different. As in the case of restitutionary awards, the law’s fundamental concern is not the wrongfulness of the defendant’s behaviour: its fundamental concern is the fairness of the outcome of the relevant transaction. The wrong is merely a reason (and not the only possible reason) for holding the defendant responsible for this outcome.

This explanation of compensatory damages is assumed, at least in broad outline, by many courts and commentators. The ‘overall object of tort law’, Lord Bingham wrote, ‘is to define cases in which the law may justly hold one party liable to compensate another’.67 This proposition is too broad (it ignores tort law’s substantive duties), but it otherwise lines up neatly with this chapter’s view of the ‘overall object’ of compensatory damages. The same is true of Baron Parke’s oft-quoted statement that damages should restore the claimant to ‘the same position as he would have been in if he had not sustained the wrong for which he is now getting his compensation or reparation’.68 Read literally, the proposition is again too broad (it ignores wrong-based and substitutionary damages), and, as we will see, compensatory damages are sometimes available for non-wrongs; however, the general idea of restoring claimants to the positions they were in prior to their injury fits neatly with the injustice-based interpretation of compensatory damages. The appropriate response to an injustice is to cure the injustice, and in the case of unfair losses, this aim is achieved by returning claimants to the positions that they were in before they incurred the losses.

As in the case of restitutionary awards, the main challenge to the injustice-based theory of compensatory damages is posed by alternative theories. If these alternatives are unpersuasive, the explanatory space that remains is difficult to fill without (p.251) invoking justice, fairness, or a similar concept. For this reason, my defence of the injustice-based theory consists in large part of arguments for rejecting these alternatives. However, my defence is not entirely negative. Most of the reasons for rejecting the alternative theories are, at the same time, reasons for adopting an injustice-based account, or something closely similar. Further, there are independent reasons for supposing that an injustice-based theory is the best option for filling the explanatory hole that remains once the leading alternatives are rejected. I discuss these reasons at the end.

I begin with the alternative theories, focusing on the two most influential candidates. The first views compensatory awards as responses to rights-threats; the second views them as responses to wrongs.

i. Rights-Based Theories

The idea that compensatory damages are responses to rights-threats was discussed—and rejected—in the previous chapter’s introduction to damages awards. To reiterate briefly, the most influential defence of this idea—the ‘continuity thesis’—supposes that the commission of a wrong gives rise to a substantive duty to do the ‘next-best thing’, which, in practice, means a substantive duty to pay damages. If such duties exist, then failing to pay damages subsequent to a wrong’s commission is equivalent to an ongoing failure to pay a debt or to perform any other substantive duty. The ongoing failure would qualify as a rights-threat (the threatened right being the right to damages), which, in turn, would support an order directing the recalcitrant defendant to pay damages. In this view, an order to pay damages is analogous to an order for a sum due or to any other replicative order.

We further saw that the main reason for rejecting the continuity thesis and, more generally, for rejecting any theory that regards compensatory awards as replicative is that the common law does not recognize a substantive duty to pay compensatory damages or, indeed, a substantive duty to pay damages of any kind. An order to pay compensatory damages cannot be a response to a rights-threat because the allegedly threatened substantive right—a right to the payment of compensatory damages—does not exist.

ii. Wrong-Based Theories

The other main alternative to an injustice-based explanation of compensatory awards is a wrong-based explanation.69 According to this view, compensatory (p.252) damages are responses to wrongs in the same way that exemplary damages, nominal damages, and the other damages awards that I discussed in Chapter 7 are responses to wrong. In the wrong-based view, compensatory awards belong with these awards.

Of course, the extent of a wrong’s consequences are often unrelated to the defendant’s moral culpability. A person who was intentionally punched may suffer fewer consequential losses than someone who was carelessly bumped in a crowd. But, as we saw in Chapter 7, the wrongfulness of a wrong is determined not just by the wrongdoer’s conduct and mental state but also by its consequences. A punch that causes serious harm is a more serious wrong than a similar punch that, fortuitously, causes little harm. It would seem to follow, then, that if courts are willing to issue wrong-responding damages awards (as I argue), they should also issue damages awards that respond to a wrong’s consequences. Finally, setting such awards at the market value of the claimant’s pecuniary loss appears to be a natural, easily administered, and proportionate response to that loss. As we saw in Chapter 7, courts adopt a market-price measure when setting wrong-based user damages, waiver damages, and market-price damages.

Courts and commentators frequently assume that compensatory damages are responses to wrongs. The first sentence of the most recent edition of McGregor on Damages defines damages as ‘an award in money for a civil wrong’.70 This sentence neatly summarizes the two propositions on which wrong-based theories rest. The first, encapsulated in the description of damages as ‘an award’, is that the law of damages (including the law of compensatory damages) is remedial law: it is law about the availability and content of a judicial remedy (as opposed to law about a substantive duty). As I explained a moment ago, I fully endorse this proposition. A large part of Chapter 7 is devoted to demonstrating that the common law does not recognize a substantive duty to pay damages, and, accordingly, that the law of damages is a set of rules about ‘awards’.

The second proposition, encapsulated in the description of damages as awards ‘for a wrong’, is that damages awards are response to wrongs. It is this second proposition that I reject, primarily because it is inconsistent with (1) the rules for assessing compensatory awards and (2) the rules for determining when compensatory awards are available. Not coincidentally, these reasons are also reasons for adopting an injustice-based theory.

(p.253) The model of a ‘wrong-based theory’ that I assume in the discussion that follows is based on the wrong-based theory that I developed in Chapter 7, and that I used to explain exemplary damages, nominal damages, damages for pain and suffering, and the various examples of vindicatory damages. This model is broadly consistent with the leading versions of wrong-based theories (e.g. civil recourse theories and deterrence theories). However, it is framed more abstractly than those theories. Specifically, it is meant to allow for different explanations of why the law provides for wrong-responding remedies. This model’s understanding of what it means for a remedy to ‘respond’ to a wrong is, however, strict. It is not sufficient that proof of a wrong supports the claimant’s action or even that it is a necessary condition for success in that action: the award must be a response to the wrong qua wrong.71 It may be useful to keep in mind nominal and exemplary damages awards as core examples of wrong-based awards.

iii. The Assessment of Compensatory Damages

Most of the rules governing the assessment of compensatory damages serve to limit the potential scope of liability for consequential losses. These ‘limitation’ rules shield defendants from liability in respect of losses that were factual (‘but-for’) consequences of their actions. Limitation rules are generally discussed under the broad heading of ‘remoteness’ (or ‘proximate cause’—the labels are interchangeable), but it includes rules that are often discussed separately, such as mitigation, contributory negligence, collateral benefits, and exoneration of liability clauses (limitation clauses, exclusion clauses). These rules are difficult to explain if one adopts a wrong-based theory. As I mentioned above, the basic argument for awarding compensation as a response to a wrong is that the wrongfulness of a wrong is determined partly by the wrong’s consequences, and, further, that the market value of those consequences is a convenient, natural, and proportionate way of marking the wrong. This argument supports liability only for losses that are a factual consequence of (p.254) the defendant’s wrongdoing. However, it provides no reason to reduce this liability in the ways provided by the limitation rules. My wrong is no less wrongful because the victim was careless, the victim could have avoided some of the loss, a third party also committed a wrong, the victim agreed (explicitly or implicitly) to limit or exclude my potential liability, or the loss was not reasonably foreseeable (though the later may be relevant when assessing the wrongfulness of my conduct). If a wrong’s wrongfulness is determined partly by its harmful consequences, all harmful consequences should count.

To be sure, there are compelling arguments for why it would be unfair to hold wrongdoers responsible for losses that are excluded by the limitation rules. If your carelessness was a factual cause of the injury that you suffered, then, even if my carelessness was also a factual cause, it seems unfair to hold me fully responsible for its consequences. Similarly, if you could have avoided a loss by taking reasonable steps to mitigate an injury’s consequences or if we had an understanding, explicit or implicit, that I would not be liable for a loss then it seems unfair to hold me liable for that loss. Indeed, it is difficult to think of any other justification for the limitation rules. Consistent with this conclusion, courts and commentators invariably describe and explain the limitation rules in just this way, and legislatures incorporate similar ideas directly into statutory limitation rules. Specifically, they describe the court’s task as one of determining whether it is ‘fair’, ‘just’, or ‘reasonable’ to attribute the claimant’s losses to the defendant:

‘[T]he second inquiry [remoteness] concerns the extent of the loss for which the defendant ought fairly or reasonably or justly to be held liable (the epithets are interchangeable).’72

‘The rationale of the principle that a novus actus interveniens breaks the chain of causation is fairness.’73

‘[M]itigation is a doctrine based on fairness and common sense, which seeks to do justice between the parties’.74

‘The common law has treated this matter [the distinction between offsetting and non-offsetting benefits] as one depending on justice, reasonableness and (p.255) public policy. . . . It would be revolting to the ordinary man’s sense of justice, and therefore contrary to public policy, that the sufferer should have his damages reduced so that he would gain nothing from the benevolence of his friends or relations or of the public at large, and that the only gainer would be the wrongdoer.’75

‘[D]amages . . . shall be reduced to such extent as the court thinks just and equitable having regard to the claimant’s share in the responsibility for the damage.’76

‘[T]he amount of contribution [to a tortfeasor’s liability] available from any person shall be such as may be found by the court to be just and equitable.’77

‘[In respect of certain torts and contract breaches, liability cannot be limited by contract or a notice except insofar as the term or notice ‘satisfies the requirement of reasonableness.’78

These quotations, and the broader theme they reflect, are difficult to explain if one adopts a wrong-based theory of compensatory damages. Once a court has determined that it is fair to hold the defendant responsible for the relevant loss, there is no room left for a wrong-based response in respect of the relevant loss. The court has already determined that the defendant is liable for the loss. That conclusion fully determines the remedy (in respect of this loss).79

Of course, wrong-based theories do not suppose that courts should issue awards that are unfair, unjust, or unreasonable. However, if courts are seeking to respond to the defendant’s wrongdoing, the question is not whether it is fair to hold the defendant responsible for the relevant loss. The court’s task is to determine the appropriate response to the defendant’s wrongdoing.80 And, insofar as the court is considering the significance of the claimant’s losses, the only thing that matters is whether the relevant loss was a consequence of the wrongdoing. That question (p.256) answered, the remaining question is how to represent this aspect of the defendant’s wrongdoing in monetary terms. And the answer to this question, as I noted above, is straightforward: the sum awarded should be set at the value of the losses. The limitation rules are inconsistent with this approach.

The courts’ focus on fairness in assessing compensatory damages also explains why courts sometimes ignore the beneficial consequences of wrongful acts.81 For example, insurance or pension payments that would not have been received if the claimant had not been injured are normally not taken into account.82 The standard explanation for this rule is that it would not be fair to deprive claimants of benefits that they had paid for: ‘it would be unjust and unreasonable to hold that the money which [the claimant] prudently spent on premiums and the benefit of it should ensure to the benefit of the defendant’.83 This explanation is perfectly reasonable (and it is difficult to imagine how any other explanation could account for the proviso that third-party insurance payments are deducted if the claimant did not pay the premiums84). But it has nothing to do with the wrongfulness of the defendant’s action. In terms of wrongfulness, a payment that would not have been received if the wrong had not occurred reduces the wrong’s harmful consequences, and so reduces its wrongfulness. If compensatory awards are responses to wrongs, this reduction should be reflected in a reduced award.85

iv. The Availability of Compensatory Damages

If the focus of the assessment of damages rules is responsibility, not wrongdoing, then we would expect to find cases where compensatory damages are awarded notwithstanding that no wrong was committed. The fact that a defendant’s action was a wrong may be a good reason to hold the defendant responsible for the (p.257) consequences of its actions, but (as I explain in more detail below) it is not the only reason. To give just one example, it might be thought (as the law indeed thinks) that, if you engage in a dangerous activity, you should be liable for any harm you cause, even if the activity is lawful.

Such cases are indeed common. Perhaps the clearest examples, albeit not strictly compensatory damages awards, are awards that compensate claimants who have intervened in a state of necessity to save or maintain another’s property or life (‘necessitous interveners’).86 For example, claimants who incur a loss while saving a ship or its cargo may obtain compensation from the owner.87 Similarly, claimants who have paid for a burial may obtain compensation from the person who had this duty.88 In the United States, courts have generalized this rule to cover any case where the relevant duty was imposed in the public interest.89 A third example is compensatory awards for so-called ‘agents of necessity’, such as individuals who accept bills of exchange to preserve the drawer’s ‘honour’.90 This category has been extended to the preservation of a defendant’s animals and, in another case, to the preservation of material property.91 In the United States, similar cases have given rise to a general principle covering the preservation of any property.92 Another example is compensatory awards for claimants who provide the ‘necessaries of life’ to the defendant or to someone for whom the defendant is legally responsible.93 Some courts have extended this principle to cover claimants who provide medical services94 and to caregivers.95 A final example is a claim brought under the law of ‘general average’ by a party to a contract for the carriage of goods by sea whose goods were jettisoned for safety reasons. Such parties may claim compensation from other parties who were transporting goods on the same ship, so that they all share equally in the loss.96

The above awards are not conventionally described as compensatory ‘damages’, presumably because it is not possible, even in theory, to classify the defendant as (p.258) a wrongdoer. But they indirectly support my interpretation of damages awards. Their existence forecloses the argument that compensation in the common law necessarily means compensation for a wrong. Further, these awards are difficult to explain without adopting an injustice-based interpretation or something similar. As Sir Francis Jeune stated in one salvage case, ‘admiralty imposes on the owner of property saved an obligation to pay the person who saves it simply because in the view of that system of law it is just he should . . . ’97

In any event, claimants can obtain compensatory damages awards notwithstanding that the defendant did not commit a wrong. I discuss seven situations where such awards are made (though there are almost certainly others). It is admittedly possible that these awards are anomalies. It is possible, in other words, that they tell us nothing about compensatory awards in those cases—by far the majority—where the defendant has committed a wrong. However, this conclusion seems unlikely. The law does not distinguish between damages for non-wrongs and ordinary damages. The courts describe both simply as ‘damages’. More importantly, the assessment process is the same in each case. To take just one example that I discuss below, damages for injuries caused by the escape of water collected on your land are the same whether you are liable because the escape was your fault or because, under the doctrine of Rylands v Fletcher, landowners are strictly liable for the escape of water collected on their land. If the claimant’s cause of action differed depending on whether the defendant committed a wrong, we would expect that the courts’ response would also differ. Yet, the law treats both categories identically.

I do not dispute that the awards examined below are discussed in books on ‘torts’, and that ‘tort’ is just another word for ‘wrong’. However, courts rarely describe the actions that trigger liability in these cases as wrongs or even as torts.98 Instead, they typically describe them as giving rise only to a liability to pay damages for consequential losses.99 In any event, regardless of how courts and commentators appear to classify the defendants’ actions in these cases, they are not wrongs, or at least not wrongs in any meaningful sense of the word. In particular, they are not wrongs in any sense that could merit a wrong-based response. To describe an action as ‘wrong’ is generally understood to mean that it is undesirable, that the action is something that we should not do. For example, to say that assaulting another is wrong is generally understood to mean that we should not assault others. And this meaning applies even if the assault arises from an innocent mistake: a (p.259) mistaken assault is still undesirable. It is this meaning—the ‘guidance’ meaning of wrong100—that was assumed in Chapter 7’s explanation of wrong-based awards. The wrongs to which wrong-based damages are a response are undesirable actions. The actions that I discuss below are not wrongs in this sense. These actions are permissible, if not positively desirable. To be sure, if you commit any of the actions discussed below you will be liable to being ordered to pay damages. But this liability does not show that the actions are wrongful in the guidance sense or in any other sense. Earning income makes me liable to pay taxes, but earning income is not a wrong. Similarly, if you agree, as part of an agreement to rent my house, to pay me for the cost of alternative accommodation, your liability to pay this sum does not mean your use of my house is a wrong.

The law is allowed to define its terms as it likes. But if it defines the actions discussed below as wrongs it will be using the term in a different sense than that in which it is normally used and understood, both within and outside the law. The only sense in which these actions are wrongs is that they ground a claim for damages. If that is all that it means to say that these actions are wrongs, this label is consistent with an injustice-based explanation of compensatory damages.

In the discussion below, I use the term ‘wrong’ in the guidance sense. Thus, I describe the actions discussed below as ‘non-wrongs’. However, my arguments would be the same if I described these actions as ‘non-wrongs, or at least as not wrongs in any sense that merits a wrong-based response’.

Vicarious Liability

Under the doctrine of vicarious liability, employers may be liable to pay compensatory damages in respect of torts committed by their employees, notwithstanding that they are innocent of any wrongdoing. It is true that the claimant must prove a wrong by the employee to establish vicarious liability, but as it is the employer who must pay the damages, the award cannot be interpreted as a response to that wrong. The defendant employer may be entirely innocent.

Rylands v Fletcher

Under the rule in Rylands v Fletcher,101 landowners are liable to pay compensation for injuries caused by the escape of something that they brought onto their land. This liability arises regardless of how much care the landowner took to avoid the escape.

Merely bringing things onto your land that may escape cannot itself be a wrong.102 If it were, a great deal of modern agricultural activity would be unlawful (not to mention the ordinary ownership of pets). More generally, the court in Rylands did not want to discourage landowners from building dams: they merely (p.260) wanted to hold them liable for any damage that they caused through this activity. If the courts thought that building dams was wrongful, the obvious thing for them to do would be to agree to issue injunctions prohibiting their construction. Instead, they merely hold individuals who engage in this activity liable for any injuries that they cause, however innocently. It is not coincidental that lawyers describe Rylands as giving rise to a ‘strict liability’ as opposed to a ‘strict duty’. The landowners’ duty is merely to take reasonable care when constructing dams: it is only their liability that is strict.

It is also not coincidental that lawyers typically do not treat batteries, trespasses, wrongful imprisonments, and other ‘intentional torts’ as analogous to Rylands liability, even though the liability in such cases is also strict. If I touch you because I reasonably, but mistakenly, believe you have consented to me touching you, I commit a battery. But unlike Ryland’s liability, in the battery cases, it is not just the liability but also (and more fundamentally) the duty that is strict. A battery is undesirable—something that is wrong in the guidance sense—whether or not it is done because of an innocent mistake. You should not touch others without their consent, regardless of whether you believe that they have consented. A non-consensual touching is a wrong even if the wrongdoer is blameless. In contrast, the activities at issue in Ryland are permissible, indeed desirable. The distinctive feature of Rylands liability is not that it holds morally blameless defendants liable: it is that it holds defendants liable for the consequences of non-wrongful actions.

Consistent with the above interpretation, in the United States Rylands liability has been generalized to strict liability for ‘abnormally dangerous activities’.103 English courts have not taken this step, but a variety of common law and statutory rules have led to a broadly similar result. English law recognizes strict liability for injuries caused by dangerous animals (or specific animals known to have dangerous characteristics), nuclear facilities, escape of water from mains, cattle, and, with certain qualifications, fire.104 English law also recognizes liabilities that are strict in the Rylands sense, but which do not involve dangerous activities. Vicarious liability, discussed above, is one example. Another long-standing example is the strict liability (with certain qualifications) of innkeepers for any loss of property suffered by guests.105

(p.261) Necessity

Individuals who have damaged others’ property in order to avoid life-threatening injuries to themselves or to others (or, possibly, to avoid more serious injuries to their own property) are liable to be ordered to compensate those they have harmed.106 A famous example is the already-mentioned American case of Vincent v Lake Erie Transportation Co.107 The defendant tied his ship, without permission, to the claimant’s dock to prevent it from being destroyed (presumably with the crew still aboard) in a storm. The court held that the defendant acted perfectly reasonably, but that he must nonetheless compensate the claimant for damage caused to the dock.

Using another’s property without their permission is, of course, normally a wrong in the ordinary (‘guidance’) sense of the term. As I noted a moment ago, even accidental trespasses are normally undesirable. This observation seems to underlie John Gardner’s suggestion that the defendant in Vincent was liable because he committed a wrong, albeit a ‘justified’ wrong.108 But it is not clear in what sense an action that is justified can, at the same time, qualify as a wrong. Certainly, it is not a wrong in the guidance sense.109 Nor does the fact that an action like using another’s dock without permission is normally a wrong in the guidance sense explain why it remains a wrong in cases, like Vincent, where it is not a wrong in the guidance sense. The most straightforward explanation of Vincent is that the defendant was responsible for the damage to the dock notwithstanding that he did nothing wrong.

Non-Substitutionary Damages in Lieu of Specific Relief

In Chapter 6, I explained that the usual reason that courts award damages in lieu of specific relief is, broadly speaking, administrative. Particularly in contract cases, the normal reason that courts award damages in cases where specific performance is possible and desired is that monetary awards are simpler to express, supervise, and enforce. However, we also saw that courts sometimes refuse specific relief (and award damages instead) for reasons based on the substantive undesirability of the requested action. In these cases, while the courts appear merely to be choosing one form of remedy over another (namely, damages over specific relief), in substance, (p.262) they are denying the existence of the alleged rights-threats. They refuse specific relief not for administrative reasons, but because they regard the actions (or omissions) that the claimant wants enjoined as permissible, if not positively desirable. It follows that the damages awards in such cases are compensation for non-wrongs. Three cases illustrate this category.

In Dennis v Ministry of Defence,110 the Ministry, as part of its pilot training, regularly flew Harrier jets over the claimant’s property. The court held that the flying constituted a nuisance, but it refused to order the defendant to cease the flights, citing the cost of moving the airfield and the importance of training pilots. Instead, they awarded the claimant £955,000 in damages.

In the second example, the American case of Boomer v Atlantic Cement Co,111 the claimants lived near the defendant’s cement factory. The claimants alleged that pollution from the factory constituted a nuisance, and they sought an injunction prohibiting the factory from creating further pollution. Noting that the requested injunction would effectively shut down the factory, leading to the loss of hundreds of jobs, the court refused the claimant’s request. Instead, they awarded compensation for the injuries that the claimants would suffer from the factory’s continued operation.

In the third example, Tito v Waddell (No 2),112 the defendant mining company had a contract with the inhabitants of a South Seas island that permitted them to mine on the island. The contract included a clause requiring the company to replant the island after they completed their mining. However, subsequent to the defendant’s mining, the island become uninhabitable (and the islanders moved to another island) following damage from bombing during the Second World War. The company did not replant the island and the claimants brought an action for, inter alia, specific performance of the replanting clause. Holding that replanting an uninhabitable island would be massively wasteful, the court refused specific performance (and also refused to award ‘substitutionary’ damages, that is, damages set at the cost of purchasing substitute performance), awarding compensatory damages instead.113

Dennis, Boomer, and Tito are controversial, but they are far from isolated decisions. Injunctions against nuisance, in particular, are regularly refused for broadly similar reasons.114 And whatever one thinks of the outcomes in these cases, it is clear that the court’s concern is not the desirability of ordering the defendant to (p.263) perform the relevant action. The court’s concern is the desirability of the action itself, regardless of why it is performed. The reason that the courts denied specific relief in these cases was, roughly, that the cost of performing the requested action (including, where relevant, the cost to the public) was, in the court’s view, vastly disproportionate to its benefit. Replanting an uninhabitable island is a massive waste of resources. This disproportion is the same whether the defendant performs the action because of a court order or because of a rule-based substantive duty. Stated differently, the courts’ reasons for denying specific relief in such cases presume that, as in Vincent, the defendant acted, or is at least now acting, reasonably. If the court thought that the defendant ought do what the claimant wanted it to do, the obvious way to bring about this result would be to order specific relief or, at a minimum, substitutionary damages.115 Courts are perfectly willing to issue injunctions against actions that they regard as undesirable. But the courts in these cases refused specific relief, and awarded compensation instead. The natural interpretation of these awards is that, like the award in Vincent, they require the defendants to compensate the claimants for the pecuniary consequences of actions that the courts regard as permissible.

Damages in Lieu of Quia Timet Injunctions

Closely related to the previous example is a compensatory award issued in lieu of a quia timet (preventative) injunction. In Leeds Industrial Co-operative Society v Slack,116 the court refused to issue an injunction to cease building works that, when completed, would allegedly have infringed the claimant’s right to light. Instead, the court awarded compensation for the harm that the completed works would cause. By definition, the defendant in such cases has not committed a wrong. In theory, such an award might be justified on the ground that it is compensation for the wrong that the defendant will commit in the future. But, if the court in Leeds believed that the defendant would be committing a wrong once the building was completed—if they believed, in other words, that the defendant ought to cease construction—the obvious way to ensure this result would be to grant the injunction. And, if the court thought that the application for a quia timet order was premature (since it was possible that the works would not be completed in the anticipated manner), they could have done what they normally do in such cases—namely, to refuse relief but leave the claimant free to request an injunction in future. By refusing the injunction and granting damages in lieu, the court effectively (p.264) gave the defendant permission to continue its activity indefinitely. The inescapable conclusion is that the court thought that it was permissible for the defendant to complete the works. As in Dennis, Boomer, and Tito, the natural interpretation of Leeds is that the court awarded compensatory damages for a non-wrong.

Innocent Misrepresentations

Section 2(2) of the Misrepresentation Act 1967 provides that, as an alternative to rescinding a contract for a non-fraudulent misrepresentation, a court may ‘award damages in lieu of rescission’. This provision applies to wholly innocent misrepresentations, and therefore allows for compensatory damages in cases where the defendant has not committed a wrong. Indeed, the legislation is explicit that the basis for awarding damages in such cases is fairness-based. The provision directs the court to award damages if ‘it would be equitable to do so’, and further directs them to consider the losses that the parties would suffer under either option.

Wrongs to Third Parties

Finally, claimants are sometimes able to recover damages for losses that they have suffered because of a wrong committed against a third party.117 For example, under the Civil Liability (Contribution) Act 1978, a tortfeasor who has been found liable to pay compensatory damages in a case of joint and several liability may obtain compensation for some or all of that payment from the co-tortfeasor(s). Section 2(1) provides that the sum ‘shall be such as may be found by the court to be just and equitable having regard to the extent of that person’s responsibility for the damage in question’. A co-tortfeasor has of course committed a wrong, but the wrong is against a third party, not the claimant seeking compensation. Similarly, under the Fatal Accidents Act 1976, the dependant of a deceased may recover against a defendant who wrongly caused the deceased’s death in respect of losses suffered because of that death. Likewise, the Contracts (Rights of Third Parties) Act 1999 provides that, in certain cases, a third-party beneficiary of contract may recover in respect of losses suffered because of the defendant’s breach of a contractual duty owed to the co-contractor. Analogous actions are also available under the Congenital Disabilities (Civil Liability) Act 1976 and the Latent Damages Act 1986. Non-statutory examples in this category include actions by disappointed legatees for losses suffered as a consequence of the defendant breaching a contractual duty owed to the (deceased) testator to revise or draw up a will,118 as well as actions (p.265) brought in respect of losses that were intentionally caused through unlawful means directed at a third party.119

In all of these cases, there has been a wrong, but it is not a wrong to the claimant. Given that the compensation must be paid to the claimant, not the victim of the wrong, it is difficult to see how these awards could qualify as wrong-based. They are clear examples of compensatory awards that cannot be explained as a response to a wrong.


The above examples demonstrate that courts frequently award compensatory damages in respect of non-wrongful actions. It is true that the vast majority of compensatory damages awards are issued in respect of wrongful actions. As I mentioned at the outset, it is therefore possible (though prima facie unlikely) that the cause of action for compensatory awards in cases where the defendant has committed a wrong is different from the cause of action in cases where the defendant is innocent. I explain below why this possibility should be rejected.

v. The Injustice-Based Theory

The preceding arguments are largely negative: they have sought to show that compensatory damages are not responses to either rights-threats or wrongs. But they have also pointed us towards a positive explanation of compensatory damages. The rules governing the assessment of compensatory damages are rules for determining when it is fair to attribute responsibility for the claimant’s losses to the defendant. These rules respond, in other words, to the unfairness—the injustice—of the outcome of the parties’ transaction. This conclusion is unsurprising. With rights-threats and wrongs out of the picture, the focus must turn to the outcome of the defendant’s actions. And, it is difficult to think of any reason to reverse or undo an outcome that does not turn on the injustice or unfairness of that outcome.120 As Arthur Ripstein noted (in commenting on Honoré’s work on responsibility), the court’s task, when it holds persons responsible for outcomes is ‘primarily allocative, in the sense that assigning particular consequences to particular persons is a way of determining who will bear which costs’.121 And, the question of how to make such allocations, as Ripstein notes, ‘is ultimately one of justice’.122

(p.266) But can anything more be said about the injustice that triggers compensatory damages? It can; however, as in the parallel case of restitutionary awards, there is not a great deal that needs to be said, at least for this book’s purposes. The injustice-based interpretation of compensatory awards supposes that the basic question for courts is whether it is fair to hold the defendant responsible for the relevant losses (or, what amounts to the same thing, whether the relevant losses are fairly attributed to the defendant). The idea that individuals may be responsible for harms that they have caused others is familiar. If you carelessly drive over my bicycle, you would, I trust, accept responsibility for my loss. The same idea is assumed throughout the large and well-established theoretical literature devoted to responsibility and private law.123 This literature focuses primarily on the law governing compensatory damages. And, it assumes, largely without argument, that the basic issue raised by compensatory damages is precisely the issue identified by an injustice-based interpretation, namely: ‘under what conditions is it morally acceptable to hold a person responsible for the causal consequences of his or her acts?’124

What, then, are these conditions? When is it appropriate for courts to hold defendants responsible for a claimant’s losses? It is not possible here to provide anything like a general theory of responsibility (though I will draw on such theories125). For present purposes, it is sufficient to show that the conditions under which compensatory awards are available align with a recognizable concept of responsibility.126

To begin, in both morality and law, we rarely hold individuals responsible for outcomes unless they are a factual cause of that outcome.127 In the normal case, factual causation is established by showing that the relevant outcome would not have happened ‘but for’ the individual’s action. If I had nothing to do with your injury, then, baring unusual circumstances (e.g. I am your parent and I neglected to look after you), I cannot be responsible for your injury. However, factual causation is clearly insufficient, standing alone: ‘[a]n extra element is needed to ground the legal sanction’.128 My decision to open a store that competes with yours may be (p.267) a factual cause of your losing business, but no court would hold me liable for that loss. In any event, every injury has multiple factual causes. My carelessly driving into your parked car is a cause of the ensuing damage, but so too was your decision to park your car in that location.

What else is required? In both morality and law, the answer is that any one of a number of additional conditions may be sufficient.129 The most obvious and practically important of these conditions is that the action that factually caused the loss was a wrong to the claimant. If you incur medical expenses as a consequence of my punching you in the face, it is fair to hold me responsible for those expenses because they are a consequence of something that I should not have done. By definition, if I should not have done the action, then you should not have suffered the consequences you suffered. And, since it was my choice to commit the wrongful action, it is fair to hold me responsible for those consequences. ‘[T]he duty to repair . . . wrongful losses is grounded not in the fact that they are the result of wrongdoing, but in the fact that the losses are the injurer’s responsibility, the result of the injurer’s agency’.130 Responsibility for the consequences of one’s wrongful actions is probably the most basic form of legal responsibility. But while some authors appear to assume, at least in their less guarded moments, that it is the only form of legal responsibility, authors who have considered the examples that I discuss below invariably accept that responsibility for bad outcomes is not limited to cases where those outcomes are a consequence of bad behaviour.131

A second sufficient condition, exemplified by employers’ vicarious liability for the torts of their employees, is satisfied where someone commits a wrong while acting on another’s behalf. Vicarious liability is often defended on the ground that it is fair for employers to bear the costs of their employees’ actions, since they reap the benefits of those actions. However, while this defence is broadly of the right kind (i.e. fairness-based), it does not explain why vicarious liability only arises if an employee commits a tort. Further, this defence appears to support strict liability across the board. Most defendants benefit from the activities in which they were engaged when they harmed the claimant (why else engage in the activity?). A more plausible explanation is that it is fair to attribute an employee’s actions to (p.268) the employer if the employee is acting in the course of employment.132 If you have engaged someone to work on your behalf, it is fair to attribute their actions to you when they are engaged in that work. It follows that, if the employee’s actions are tortious, the liability for the tort is fairly attributed to the employer.

A third sufficient condition, exemplified by Rylands v Fletcher liability, is that the defendant was engaged in an activity that is dangerous to others. Some dangerous activities are simply banned outright, for example, discharging firearms in public spaces. But others are permitted because their benefits are thought to outweigh the risks. Dams are inherently dangerous, but they are invaluable for controlling floods, providing irrigation, and so forth. At the same time, it is fair to hold individuals who choose to engage in dam building and other dangerous activities responsible for the harm that their activities may cause.133 Individuals who engage in dangerous activities expose others to unusual risks, risks that go beyond the ordinary risks that are properly regarded as part of the ordinary vicissitudes of life (and for which there is ordinarily no liability without wrongdoing).

A fourth condition, exemplified by the liability in Vincent v Lake Erie, is that the loss arose from the defendant’s decision to use the claimant’s property to save its own person or property. The question of when such uses are permissible is difficult and controversial. One person’s ‘necessitous situation’ (or ‘emergency’) is another person’s ordinary circumstances. But, for present purposes, it is sufficient to accept that, however necessity is defined, it is fair to hold defendants responsible for damage that they have caused in a state of necessity. It is fair because the defendant chose to use the claimant’s property. If I use your property to save my property (or my person), I should compensate you for your loss.

A fifth, closely related condition, exemplified by cases like Dennis, Boomer, and Tito, is that the loss arises from an act or omission that is normally wrongful, but that is not wrongful in the particular circumstances because its value far outweighs the harm that it causes. As with question of what circumstances qualify as a necessity, the question of when the costs and benefits associated with a normally impermissible activity make that activity permissible is difficult and controversial. But wherever the line is drawn, it is fair to make actors engaged in such activities pay for the harm that they cause. The explanation is the same as for Vincent: if the law allows you to do something that is normally impermissible (because it may harm others), and you choose to do that thing, it is fair that you compensate those whom the normal rule of impermissibility is meant to protect. Dennis, Boomer, and (p.269) Tito are milder versions of Vincent. The main difference between these cases and Vincent is that in Vincent it was too late for the claimant to request an injunction.

These comments fall far short of a comprehensive explanation of compensatory damages, much less of responsibility for loss in general. For the most part, I have merely asserted that the various situations in which courts award compensatory damages are consistent with ordinary intuitions about when it would be fair to attribute responsibility for the claimant’s loss to the defendant. However, it is not clear that a great deal more can be said about responsibility for outcomes, at least if one’s goal is limited to explaining the law. Writers on responsibility invariably test their accounts against ordinary intuitions about responsibility.134 If the above description of those intuitions is unconvincing, it is doubtful that a theoretical account of responsibility would be more convincing. In any event, it is clear, I trust, that there is nothing unusual about saying that it is fair to hold defendants who committed wrongs, whose employees committed wrongs, who engaged in dangerous activities, or who used another’s property for their benefit liable for the consequences of their actions. Courts, commentators, and ordinary individuals say these things all the time. This chapter’s main contribution to the literature on responsibility is to fit this familiar idea into a general account of remedies.

D. Other Injustice-Responding Remedies

The awards available to necessitous interveners (discussed above), as well as those authorized under matrimonial property and contribution legislation, show that the category of injustice-responding remedies is not limited to restitutionary and compensatory orders. Other potential candidates are most likely to be found amongst those remedies conventionally described as orders to reverse unjust enrichments. Peter Birks and others have long assumed that the restitutionary orders discussed in this chapter are core examples of this broader category.135 However, as strong doubts have recently been raised about this assumption,136 I will limit myself to noting that, insofar as the orders that Birks assumed to be analogous to orders that reverse defective transfers are, in fact, analogous, their cause of action is the same.

The next most obvious candidates for inclusion in this chapter are remedies for actions that, as currently understood, appear to be neither contractual, tortious, (p.270) nor based on an unjust enrichment.137 In closing, I identify two potential candidates, each of which illustrates this category’s potential breadth.

i. Compensation for Detrimental Reliance (Equitable and Proprietary Estoppel)

In certain situations, claimants who have incurred expenses in reliance on another’s representation may obtain a compensatory remedy, notwithstanding that the representation is neither contractual nor a tortious misrepresentation or other legal wrong.138 If I erect a building on your land because your representation (or even your mere acquiescence to my actions) induced me to believe that you will transfer the land to me, I may be able to obtain an order directing you to compensate me for the cost of my improvements.139 The scope of such relief varies amongst common law jurisdictions: in England, it is, in theory, restricted to representations involving property (‘proprietary estoppel’),140 while in the United States and Australia, it applies more generally (where it is usually described as ‘promissory’ or ‘equitable’ estoppel).141 The basis of the doctrine is also a matter of dispute. Some writers explain these actions as, in substance, contract actions, the primary difference being that ‘detrimental reliance’ takes the place of ‘consideration’.142 For others, probably the majority, the doctrine’s core role is protecting detrimental reliance; in this view, it does not matter, in principle, whether the defendant has made a promise, and the remedy may be more limited than a standard contractual remedy.143

I cannot resolve this long-running debate here (though I note that relief has been awarded in at least some cases where it is clear that the defendant did not make a promise and that this relief is less than what would have been awarded in a contractual action144). I only note that insofar as the detrimental-reliance interpretation explains some or all of the remedies in this area, these remedies appear to be injustice-based. In such cases, the claimant is compensated notwithstanding that (p.271) the defendant neither committed nor threatened to commit a wrong. Consistent with this interpretation, the courts’ explanations of what they are doing when they decide these cases are replete with references to ‘injustice’ and ‘fairness’. In Deane J’s words in the leading Australian decision of Walton Stores (Interstate) Ltd v Maher, the underlying issue in these cases ‘is whether the appellant so contributed to the assumption that the exchange had taken place that it would be unjust or unfair if it were left free to ignore it’.145

ii. Property Transfer Injunctions

Certain orders to transfer specific property appear to be responses to injustices. Of course, most orders to transfer property are made in respect of property that is owned by the claimant. As we saw in Chapter 7, these orders, traditionally styled ‘delivery up’, are responses to rights-threats, specifically the threat that the defendant will not return the property. However, certain orders to transfer specific property appear to be responses to the same kind of injustice that ground compensatory damages.

One example is an order to undo or remove a wrong’s consequences, such as an order a remove a structure that was constructed, unlawfully, on the claimant’s land.146 The mere existence of an unlawfully constructed structure is not itself a trespass or other wrong: trespassing is an action (it can only be committed through intentional action), not a state of affairs. Nor is an unlawfully constructed structure a nuisance: a structure that is located on the claimant’s land cannot interfere with the claimant’s use of that land. It follows that an order to remove an unlawfully constructed structure is, in substance, an order to undo the consequences of a wrong. Accordingly, its cause of action appears to be an injustice.

A second possible example is an injunction that directs the defendant to return property obtained through a wrong. In Esso Petroleum Co Ltd v Kingswood Motors (Addlestone) Ltd,147 the defendant committed the tort of interference with contractual relations by purchasing a garage that was subject to a solus agreement (which precluded such a purchase) between the vendor and the claimant. Bridge J ordered the defendant to transfer the garage back to the vendor. Although the (p.272) defendant had committed a wrong in purchasing the garage, its resulting ownership of the garage was not a wrong. The defendant had clear title to the garage. The purpose of the re-transfer, therefore, was not to end an ongoing wrong: the wrong was finished. Rather, the purpose can have only been to undo the consequences of the defendant’s past wrong. As Bridge J observed, the order’s effect was to ‘enforce the personal liability incurred by a tortfeasor to undo the consequences of his tort which could have been restrained before it was committed’.148 The injunction in this case appears to have had the same cause of action as an order to pay compensatory damages: an injustice.


(1) (1841) 9 M & W 54. For a more detailed discussion, upon which I draw, see Birks (2005) 5–9.

(2) (1841) 9 M & W 59.

(4) (1874) LR 9 Exch 132.

(5) Ibid 136.

(6) Sugden v Sugden [1957] P 120, 135.

(7) [2003] EWHC 793.

(8) 26 NY 2d 219 (1970).

(9) [1977] Ch 106.

(10) [1910] 109 Minn 456.

(12) In Torts and Rights, Robert Stevens distinguishes tort theories that focus on the ‘infringement of rights’ from theories (which Stevens criticizes) that focus on the ‘infliction of loss’: Stevens (2007) 1–3. It might be thought that my account of compensatory damages, which supposes that courts issue such awards in cases where no right was infringed, is, in Stevens’ terminology, a loss-based account. However, my explanation of compensatory damages is not a theory of tort law or even a theory of tort damages. I agree with Stevens that the core of tort law is a set of substantive, rights-based, duties (e.g. ‘everyone has a duty to take care not to injure others’). Further, I argue that most ‘tort remedies’ confirm such duties (e.g. injunctions), or are responses to their breach (wrong-based damages). Finally, while my account of compensatory damages assumes that their role is to re-allocate losses, its fundamental focus is injustices, not losses. As Stevens rightly notes, if tort law’s aim were merely to ‘compensate losses’ it should almost certainly be replaced by a universal compensation scheme. Most losses are not caused by a person or anything else that could possibly qualify as a defendant in private law litigation. Rather, they are caused by illness, disease, natural events, self-inflicted accidents, and so forth. However, if (as I argue above) the role of compensatory damages is to correct injustices, then a universal compensation scheme is not an alternative route to the same end. Injustice is a relational concept. Correcting injustices always involves transferring a loss or gain from one person (or persons) to another person (or persons). Compensatory damages correct injustices by re-allocating losses, but the underlying aim is to correct injustices.

(13) Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22 [9].

(14) See, generally, Gardner (2012) 6–7, upon which I have drawn heavily. See also Hart (1961) 157–8; Rawls (1971) 10 (‘The concept of justice I take to be defined, then, by the role of its principles in assigning rights and duties and in defining the appropriate division of social advantages’); Aristotle (reprinted 1976) (‘The just, then, is a species of the proportionate’: 113P29).

(15) Though not unknown; see e.g. Finnis (1980) 178–9; Weinrib (2012) 87–98 (each of whom draws on Acquinas’s concept of ‘commutative justice’).

(16) Gardner (2012) 256–9.

(17) See the discussion of methodology in Chapter 1. It might be asked whether the category of ‘injustices’ can accommodate instrumental theories. Such theories explain restitutionary and compensatory orders on the ground that, ideally, they incentivize efficient behaviour. For example, compensatory damages incentivize potential injurers to take appropriate levels of care to avoid injurers, while restitutionary orders incentivize potential transferors to avoid taking too much care to avoid mistaken transfers, etc: see e.g. Landes & Posner (1987); Cooter & Porat (2019). Such explanations have no concern for justice, at least insofar as ‘justice’ means something that is intrinsically valuable. They take a similar approach to ‘rights’ and ‘wrongs’. However, these approaches still need to distinguish the kinds of events that support restitutionary and compensatory orders. No explanatory theory can avoid classifying the material it is explaining. And when instrumental explanations draw such classifications, they unavoidably rely on conventional legal concepts. To explain the law, one must describe what one is explaining in terms that those familiar with the law can understand. For example, instrumental explanations of ‘contract damages’ invariably assume, either explicitly or implicitly, that they are explaining the kinds of damages that are awarded for a ‘breach’ of contract. Such explanations may query whether it makes sense, in some cases, to describe a breach as a breach (or as a ‘wrong’), but they still describe the rules that they are explaining as the rules governing damages for breach of contract. In short, while instrumentalists might treat concepts like ‘right’, ‘wrong’, and ‘injustice’ as mere placeholders, their explanations (like all explanations of the law) cannot avoid relying on such concepts. Of course, an instrumentalist might argue that I have used the wrong concepts or that I have drawn the borders between them in the wrong places. But, if this argument is persuasive, it should be persuasive to instrumentalists and non-instrumentalists alike.

(18) ‘[Asking for a unitary theory of tort law] is a very stupid question’: Weir (2006) ix; ‘[T]here is little unity to the subject [of tort law]’: Stevens (2007) 299.

(19) In my classificatory scheme, the concept of an injustice operates in broadly the same way that the concept of an unjust enrichment operates in Peter Birks’ scheme—as an organizing concept, not a directly applicable standard. The similarity is not coincidental. Like this book, Birks’ work on unjust enrichment law focuses on structural issues. For Birks, as for this book, the label for the third category is a secondary issue. However, in other respects our views differ significantly. The most obvious difference is that I include compensatory damages in my third category. A second difference is that Birks assumed that the law recognizes substantive duties to make restitution and to pay damages. A third difference is that Birks’ argument for the unity of ‘unjust enrichment law’—his argument for why his third category is not labelled ‘miscellaneous’—focuses, at a relatively finely grained level, on the facts that claimants must prove to demonstrate an unjust enrichment. For Birks, all unjust enrichments arise from events where ‘[t]he defendant is enriched at the expense of the claimant and there is in addition a reason, not being a manifestation of consent or a wrong, why that enrichment should be given up to the claimant’: Birks (2005) 22. This approach is not possible for ‘injustices’—or, at least, it is not possible if one wants to explain why an injustice is the cause of action for both restitutionary and compensatory orders. This last difference explains why the concept of an injustice does more explanatory work in my account. For Birks, ‘unjust enrichment’ was, largely, a placeholder (though not entirely—‘But for the need to retain a trace of normativity, one might just as well speak of pink enrichment’: Birks (2005) 275, emphasis added).

(20) The rules governing injustice-responding orders are, in one respect, more determinate than those governing wrong-responding orders. As we saw in Chapter 7, the quantification of wrong-responding orders is ultimately a matter of choice and convention. In contrast, the content of an injustice-responding order is determined by its cause of action. If the claimant establishes an actionable injustice, the standard and appropriate response is to order the defendant to correct the injustice. Thus, in Kelly, the widow had to repay the money that she received; in Vincent, the ship owner had to pay for the damage that he caused to the dock; and, in Swaffield, the owner of the horse had to reimburse the station for its expenses in looking after the horse. The one caveat is that the courts’ preference for monetary over non-monetary remedies (which I examined when discussing substitutionary orders in Chapter 7) also applies in respect of injustice-responding orders. Thus, if the relevant injustice arises from the defendant obtaining or retaining fungible property—for instance, gold bars—the court will normally order the defendant to pay the claimant the property’s monetary value, rather than to return the property itself. And, because the means by which courts impose restitution in specie is the historically equitable concept of a constructive trust, such restitution is, in principle, not available as a right.

(21) There are statutes that come close to requiring claimants seeking injustice-responding orders to prove directly that they have suffered an injustice. Section 23(1)(c) of the Matrimonial Causes Act 1973, which I discussed earlier, gives courts wide discretion to issue orders requiring that ‘either party to the [now dissolved] marriage shall pay to the other such lump sum or sums as may be so specified’. The legislation simply directs the court to take into account a wide range of factors and then to decide whether it is appropriate, in the circumstances, to issue an order. Courts enjoy similarly broad discretion with respect to orders for contribution made under s 2(1) of the Civil Liability (Contribution) Act 1978 (a ‘just and equitable’ sum) and orders for the financial provision of dependents under section 2(1) of the Inheritance (Provision for Family and Dependants) Act 1975 (‘reasonable financial provision’).

(22) Another possible answer is that a substantive duty to correct injustices would be an unknowable duty. As I mentioned above, I had already come to the view (which I defend below) that, similar to a substantive duty to pay compensation, a substantive duty to make restitution would turn on facts that those bound by the duty could not reasonably be expected to obtain. In Chapter 7, I considered a parallel explanation in the case of duties to pay damages (and I also considered the possibility that such a duty was ‘inchoate’ until fixed by a court order). My conclusion—which was to tentatively support this answer—applies equally to injustice-correcting duties.

(23) This argument is broadly similar to arguments defended in Nadler (2013) and Penner (2018) in respect of duties to make restitution. Although the details of their explanations of restitutionary liabilities differ from mine (particularly in the case of Nadler), both Nadler and Penner argue that a substantive duty to make restitution would be qualitatively different from the kinds of duties traditionally recognized in the common law.

(24) Failing to correct an injustice is not the same as acting unjustly. An example of the latter is a wrongful conviction. We often have duties to act justly, typically (if not invariably) because we occupy an office that requires us to make allocative decisions. Thus, referees, judges, and adjudicators in general have duties to act justly. Indeed, it is difficult to conceive of an adjudicator that does not have such a duty. However, we can easily conceive of adjudicators who do not have the authority to correct injustices (their authority having been limited to resolving other kinds of allocative issues).

(25) The distinction between a valuable action and a substantive duty is sometimes expressed, drawing on Kant, by describing the former as an imperfect duty or a duty of virtue: Kant (reprinted 1996) 89–91. See also Kamm (1985); Penner (2018) 229. I have avoided these terms because, inter alia, Kant classified duties of justices as perfect duties and, further, because ‘duties’, as I use this term, are always mandatory. However, my conception of correcting injustices is consistent with the idea, associated with Aristotle (and, more recently, with Rawls), that justice is a virtue: Aristotle (reprinted 1996) 186; Rawls (1971) 3 (‘Justice is the first virtue of social institutions’).

(26) The civil law is more open to recognizing such duties. For example, in addition to recognizing a duty to rescue, the Civil Code of Quebec recognizes a duty (albeit qualified) to correct a potential co-contracting party’s misapprehensions. Strikingly, it also stipulates that ‘every child, regardless of age, owes respect to his father and mother’: Civil Code of Quebec, arts 1401, 597; Kasirer (2001). It is an interesting question whether civil codes’ openness to such duties explains why they typically affirm substantive duties to make restitution and to compensate.

(27) As I explain in Chapter 9, the reverse also holds true. Courts often have good reasons to refuse to order the performance of substantive duties, even when those duties are under threat. Such reasons explain, inter alia, why courts may deny relief to a claimant who has unduly delayed or why they may award damages in lieu of specific performance.

(28) Although I will not pursue the idea here, it is possible to interpret wrong-responding remedies as a particular kind of injustice-correcting remedy. In Chapter 5, I argued that wrong-based remedies are, in structural terms, the private law equivalent of criminal punishment. Of course, criminal punishment is traditionally explained as a form of retributive justice. From this perspective, wrong-based remedies and injustice-correcting remedies differ only in the kind of injustice that they seek to correct. In other words, the former focuses on normative losses and gains, the latter, on factual losses and gains.

(29) See e.g. Workers Compensation Act, RSO 1990 c. W. 11 (Ontario); Article 51, New York Insurance Law.

(30) With the proviso that a decision not to make such orders available in particular circumstances is also a policy decision. In Chapter 9, I explain that remedial defences often serve to conserve scarce judicial resources or forward other ends of the state (e.g. fostering diplomatic relations). When courts refuse a remedy for this reason, they are refusing to do what justice requires for policy reasons.

(31) See e.g. Birks (2000), (2005) 168–9; Weinrib (2012) 87–98; Klimchuk (2009); Saprai (2006); Ripstein (2009) 304; Gardner (2011) 30–2.

(32) In the United States, where litigants normally bear the entirety of their legal costs (regardless of whether they win or lose), this objection is more an objection to the litigation system generally than to injustice-responding orders in particular.

(33) I am also equally free in each case to reach a settlement. Under my interpretation of restitutionary orders, there is no duty to make restitution in advance of litigation; however, if I make restitution in advance, my liability to being ordered to make restitution is extinguished. As I explained in Chapter 8, an offer to pay compensation in advance or even actual payment does not extinguish liability, unless it is part of a settlement. However, if I make an offer or a payment in advance, the court would almost certainly require the other side to pay my legal costs.

(34) Burrows (2011) 201–317, 403–522. The analysis that follows applies equally if the explanation for restitution is not the impairment of the payor’s autonomy, but, instead, the absence of a legal basis for the payment. Burrows (2011) discusses the distinction at 86–116.

(35) A notable exception is Zakrzewksi (2005) 112–14.

(37) See e.g. Birks (2000a), (2005) 168–9; Weinrib (2012) 189; Klimchuk (2009); Saprai (2006).

(38) Birks (2005) 169. The (alleged) duty could arise at a different moment. For example, it might be thought that the duty arises only when the recipient becomes aware of the impairment, or only when the transferor demands restitution or initiates a lawsuit. I discuss some of these variations below. The duty-confirming view also allows for the possibility that the duty’s content may change over time. Birks presumably believed in this possibility since, in addition to assuming that the duty arises on transfer, he held that if transferees changed their position in reliance on the transfer, then the content and even the existence of their substantive duty could change: Birks (2005) 208–23. In contrast, scholars who believe that the duty arises only after the recipient has knowledge of the transfer and the impairment presumably believe the duty to be invariant, since such knowledge is generally taken to preclude the change of position defence.

(39) See e.g. Colonial Bank v Exchange Bank of Yarmouth, Nova Scotia (1885) 11 App Cas 84, 90–2; Banque Financiere de la Cité v Parc (Battersea) Ltd [1999] AC 221, 233. Courts frequently state that the cause of action for a restitutionary order accrues at the moment the defendant is enriched (see e.g. Prudential Assurance v HMRC [2018] UKSC 39, [69]). These statements support the duty-creating view because, if a substantive duty arose at the moment of enrichment, it would be more natural to refer to that duty arising as opposed to a cause of action arising.

(40) Portman Building Society v Hamlyn Taylor Neck [1998] 4 ALL ER 202, 208 (Millett LJ). See also Kleinwort Benson Ltd v Lincoln City Council [1998] 4 All ER 2 513, 549 (‘If the retrospective legislation positively requires a transaction to be reopened, the liability to repay will arise’: Lord Lloyd).

(41) ‘The previous paragraphs have shown that a striking feature of the liability to make restitution of mistaken payments is that the liability is strict’: Birks (2005) 8. Ernest Weinrib—another leading defender of the duty-confirming view—uses similar language. On three occasions on the first page of the chapter on ‘Unjust Enrichment’ in Corrective Justice Weinrib describes unjust enrichment as a principle or basis of ‘liability’: Weinrib (2012) 185.

(42) Other examples include ss 2, 5, 13, 14, 15, 16, and 17. It is clear that the Restatement authors understand ‘liability’ to mean liability to a court order because they alternate between describing unjust enrichments as giving rise to liabilities and describing them as giving rise to ‘a claim in restitution’: see e.g. ss 6, 7, 8, 9, 10, 11, 12, 18, and 19. See generally Smith (2013b).

(43) ‘Where under the contract of sale the seller is bound to send the goods to the buyer, but no time for sending them is fixed, the seller is bound to send them within a reasonable time.’: Sale of Goods Act 1979, s 29(3).

(44) By virtue of the Senior Courts Act 1981, s 35A.

(45) Wadsworth v Lydall [1981] 1 WLR 598; Sempra Metals Ltd v Commissioner of Inland Revenue [2007] UKHL 34.

(46) Westdeutsche Landesbank Girozentrale v Islington London Borough Council [1996] AC 669, 696, 736; Sempra Metals Ltd v Commissioner of Inland Revenue [2007] UKHL 34, 34–50, 116, 132, 178–9; Prudential Assurance v HMRC [2018] UKSC 39, [76]–[77]. See also Birks (2005) 53; Burrows (2011) 21–5.

(47) Lipkin Gorman v Karpnale Ltd [1999] 2 AC 548.

(48) Burrows (2011) 525, 537–40.

(49) This interpretation is supported by a comparison to the case of debtors who intentionally refuses to pay their debts. If s 5(4) is aimed at defendants who have intentionally breached a duty to pay money, its principle should apply to defaulting debtors; indeed, the principle of s 5(4) should apply more strongly to defaulting debtors because their obligation was undertaken voluntarily.

(50) Att-Gen’s Reference (No 1 of 1983) [1985] 1 QB 182.

(51) Burrows (2011) 69–85.

(52) The unknowability objection cannot be avoided by arguing that the substantive duty to make restitution is an inchoate duty, which is then fixed by a judicial order. In Chapter 7, I discussed a parallel response to the objection that substantive duties to pay damages are unknowable. The reasons that I gave for rejecting that argument in respect of unknowable compensatory duties apply equally to unknowable restitutionary duties.

(53) The breach of such a promise would not normally qualify as a breach of contract because the promise was not given in exchange for consideration. However, it would be a small step to add promises to return benefits obtained under defective transfers to the list of non-contractual, promissory duties recognized by the common law. Examples of the latter include gratuitous promises to provide bailments or (in select cases) other services, to keep open offers to enter unilateral contracts, to auction goods without reserve, and to provide irrevocable credits: see Peel (2015) 173–84. On some views, liability for promissory estoppel, misrepresentation, and a variety of other torts also fit in this category: see e.g. Stevens (2007) 9–14, 33; Beever (2007) 284–315.

(54) Ernest Weinrib argues that the law ‘imputes’ that Mrs Solari has ‘accepted’ her beneficial transfers on the basis that it was non-donative: Weinrib (2012) 204–7. However, as Weinrib’s explicit rejection of the ‘implied contract’ theory of unjust enrichment makes clear, he is not arguing that Mrs Solari agreed to return her benefits. The significance of the imputation, it appears, is to refute the suggestion that transferees are ‘passive’ participants in the relevant transaction. However, the question remains as to the basis of the duty to make restitution. Weinrib writes, ‘The plaintiff cannot retain gratis what was neither given gratis nor accepted [by imputation] as given gratis’ (204). But why? If I have understood correctly (and I am not certain that I have), Weinrib’s only normative argument is that it would be ‘unfair’ for the transferee to retain the benefit: (207). I agree. The only proviso is that, unlike Weinrib, I argue that duties to make restitution are created by court orders, not substantive rules. Weinrib does not address this issue, but I note that whether or not he grounds his substantive duties in fairness, the explanation of these duties will be different from the explanation Weinrib offers for our other substantive duties. Consistent with his Kantian roots, Weinrib views our substantive rights basically as rights that others not use or interfere unreasonably with the means by which we pursue our ends—which, in practice, means our persons, property, and contractual rights: ‘Among these rights [our substantive rights] are the right to the integrity of one’s body as the organ of purposive activity, the right to property in things appropriately connected to an external manifestation of the proprietor’s volition, and the right to contractual performance in accordance with the mutually consensual exercises of the parties’ purposiveness’ (353). It is not clear where substantive duties to reverse unjust enrichments fit in this picture. Mrs Solari’s retention of the benefit she received from the insurance company was not an interference with the insurance company’s person, property, contractual rights, or, it appears, with anything else to which they might conceivably be said to have a right.

(55) In a series of Australian decisions beginning with Pavey & Matthews Property v Paul (1987) 162 CLR 221, Gummow J has argued that the availability of restitution is triggered by ‘unconscionable conduct’. The court’s reasoning in the seminal case of Moses v MacFerlan [1760] 2 Burr 1005 is replete with references to the ‘ties of natural justice’ ‘founded in the equity of the plaintiff’s case’, and money which ‘ex aequo et bono’ the defendant ought to refund, money retained against ‘conscience’.

(56) I defended this view in Smith (2013). Others who have defended versions of this view include Stoljar (1989) 5–9; Watts (1995) 49–70; Dietrich (1998) 208–13; Jaffey (2000) 275–9; Webb (2009); McFarlane (2012).

(57) Burrows (2011) 14–15.

(58) Webb (2009) 335.

(60) Burrows (2011) 14–15.

(61) See Wilmot-Smith (2016) 1131.

(64) See also note 17.

(65) Burrows (2011) 201–522.

(66) The classification of damages awards (or parts of awards) as compensatory, wrong-based, or substitutionary often turns on complex technical distinctions. For an excellent analysis of a wide range of tort awards (albeit adopting different terminology and explanations of the relevant awards), see Stevens (2007), especially 20–92, 129–72.

(67) Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22, [9].

(68) Robinson v Harman [1848] 1 Exch 850, 855. See also Livingstone v Rawyards Coal Co (1880) 5 App Cas 25, 39 (‘[the measure of damages is] that sum of money which will put the party who has been injured, or had suffered, in the same position that he would have been in if had not sustained the wrong’: Lord Blackburn).

(69) The most important wrong-based explanation of damages is the civil recourse theory developed by John Goldberg & Benjamin Zipursky: Goldberg & Zipursky (1998a), (2010); Goldberg (2005), (2006); Zipursky (1998), (2002), (2003). Though broadly consistent with the civil recourse theory, this chapter’s model of a wrong-based theory abstracts from various details of civil recourse theory (in particular, its Lockean-inspired argument for why courts provide remedies for wrongs) and, more generally, draws on the interpretation of wrong-based remedies set out in Chapter 7. Goldberg & Zipursky limit their explanation to tort damages (though others have expanded it to contract damages: Oman (2011)). But, within that category, they apply it to all damages (i.e. not just compensatory damages). I defended a wrong-based explanation of tort and contracts damages in Smith (2012), (2014).

(71) A potential objection to this argument is that for certain torts—negligence is the main example—proof of a wrong is, by definition, a necessary and sufficient condition for an award of compensatory damages since (this objection goes) these torts cannot be committed unless the claimant has suffered an injury. Thus, a claimant who establishes the wrong of negligence has also satisfied the cause of action for compensatory damages. There are two responses to this objection. First, while proof of damage is part of the cause of action in negligence, it is not clear that it as an element of the wrong of negligence. In other words, proof of damage arguably goes to the actionability, not the wrong, of negligence: see Smith (2015); McBride & Bagshaw (2015) 123–5 (who note, inter alia, that in Donaghue v Stevenson Lord Atkin did not say that ‘You must avoid injuring your neighbor by failing to take reasonable care to avoid acts or omissions that reasonably foreseeably would have that effect’. Rather, what he said was, ‘You must take reasonable care to avoid acts or omissions which you can reasonably foresee would be likely to injure your neighbor.’). Second, even if damage is an element of the wrong of negligence, this merely shows that in some cases the requirements for a wrong are the same as for injustice. It does not show that compensatory damages are a response to the wrong.

(72) Kuwait Airways Corp v Iraqi Airways Co (No 3) [2002] 3 All ER 209, 228. See also Overseas Tankship (UK) Ltd v Morts Dock & Engineering (The Wagon Mound, No 1) [1961] AC 388, 422 (‘It does not seem consonant with current ideas of justice or morality that for an act of negligence, however slight or venial, which results in some trivial foreseeable damages the actor should be liable for all consequences however unforeseeable and however significant, so long as they can be said to be “direct” ’: Viscount Simonds); Hadley v Baxendale [1854] 9 Exch 341, 354 (‘[T]he damages which the other party ought to receive in respect of such breach of contract, should be such as may fairly and reasonably be considered, either arising naturally, ie according to the usual course of things or such as may reasonably be supposed to have been in the contemplation of the parties’: Baron Alderson); Chubb Fire Ltd v Vicar of Spalding [2010] EWCA Civ 981, [64] (‘The courts have to make a “value judgment” when dealing with issues such as “remoteness of damage” or “causation” . . . the ultimate question is: what is the extent of the loss for which a Defendant ought fairly, or reasonably or justly to be held liable’: Lord Justice Aikens).

(73) Corr v IBC Vehicles Ltd [2008] 1 AC 884 [15].

(74) Southcott Estates Inc v Toronto Catholic District School Board 2012 SCC 51 [25].

(75) Parry v Cleaver [1969] 1 All 555, 557–8.

(76) Law Reform (Contributory Negligence) Act 1945, s 1(1).

(77) Civil Liability (Contribution) Act 1978, s 2(1).

(78) Unfair Contract Terms Act 1977, ss 2(2), 3(2), 6(1A), 7(4), and 8(3).

(79) It might be thought that a counter-example is the rule that, at least in respect of certain torts, where the wrong is deliberate, defendants will be held liable for all the consequences of their wrongdoing (not just foreseeable consequences): Doyle v Olby (Ironmongers) Ltd [1969] 2 QB 158 (deceit); Quinn v Leathem [1901] AC 495 (conspiracy); Kuwait Airways Corpn v Iraqi Airways Co (Nos 4 & 5) [2002] AC 883 (conversion, but with suggestions that the rule should apply generally). As we saw in Chapter 7, in principle deliberate wrongs should attract larger wrong-based remedies. But there is no reason to tie this increase to the amount of unforeseeable consequential losses. In any event, exemplary and aggravated damages already serve this role (and neither set the amount by reference to losses). However, if the question for the courts is whether it is fair to hold the defendant responsible for a particular loss, the fact that the wrong was deliberate is a reason to answer in the affirmative. This conclusion is the flip-side of Viscount Simonds’ conclusion in the Wagon Mound that it would be unfair to hold a defendant who had only committed a trivial fault liable for massive losses (see note 73).

(80) Exemplary damages are justified, Lord Devlin wrote, where they are needed ‘to punish [the defendant] for his outrageous conduct, to mark their disapproval of such conduct, and to deter him from repeating it’: Rookes v Barnard [1964] AC 1129, 1228. ‘[B]attery or trespass to the person is actionable without proof that the victim has suffered anything other than the infringement of his right to bodily integrity: the law vindicates that right by awarding nominal damages’: Ashley and another v Chief Constable of Sussex Police [2008] UKHL 25 [60] (Earlsferry LJ). See also Kuddus v Chief Constable of Leicestershire Constabulary [2001] UKHL 29, [63]–[65].

(81) Burrows (2004) 169–74. Not surprisingly, courts also sometimes ignore, in a similar fashion, negative consequences attendant to a defective transfer. For example, if, as a result of you mistakenly paying me £1,000, my uncle decides not to give me a gift of £1,000 that he would otherwise have given me, I remain liable to return your £1,000 to you.

(82) Bradburn v Great Western Railway Co (1874) LR 10 Exch 1; Parry v Cleaver [197] AC 1.

(83) Parry v Cleaver [1967] AC 1, 14 (Lord Reid).

(84) Hussain v Hew Taplow Paper Mills Ltd [1988] 1 AC 514.

(85) A final observation is that the sheer quantity of law devoted to assessing compensatory damages is surprising if compensatory damages are wrong-based. Approximately three quarters of the nearly 2,000 pages in the latest edition of McGregor on Damages focuses on the rules governing the assessment of compensatory damages: Edelman (2018). The magnitude and detail of this law is surprising if compensatory awards are wrong-based. As we saw in Chapter 7, there is no natural, logical, or inevitable response to a wrong. In private law as in criminal law, the courts’ response to wrongs is ultimately a matter of choice and convention. At the end of the day, damages can never do more than to represent a wrong. Thus, if compensatory damages are wrong-based, the courts appear to be devoting incredible effort to establishing a sum that, at the end of the day, is arbitrary. However, the magnitude and detail of this law is unsurprising if the courts’ task is to determine which of the claimant’s losses the defendant may be fairly be held responsible. Ascriptions of responsibility, as I explain below, turn on a multitude of complex factors.

(86) See e.g. McCamus (1979).

(87) The leading case of Falcke v Scottish Insurance Co (1886) 34 Ch D, 234 argues for keeping this category small, but its authority in this regard is doubtful (see Kortmann (2005) 113). In England, the principle has been extended to aircraft. The sums awarded to salvors often include a non-compensatory ‘reward’ for the purpose of encouraging future salvors.

(88) Jenkins v Tucker (1788) 1 Hy Bl 90.

(89) ‘[W]hen an obligation is imposed by law upon one to do an act because of an interest in the public to have it done, and that one fails to do it, he who does do it, expecting compensation, may recover therefore of him on whom the obligation is imposed’: Mathie v Hancock 78 Vt. 414, 63 A. 143 (1906) (Rowell CJ). In England, a similar principle was proposed in Urmston v Newcomen (1788) 126 ER 55.

(90) Howtayne v Bourne (1841) 7 M & W 595.

(91) Great Northern Railway Co v Swaffield (1874) LR 9 Exch 132; Prager v Blatspiel, Stamp and Heacock Ltd [1924] 1 KB 566.

(92) The Restatement (Third) of Restitution and Unjust Enrichment description of the awards in such cases as ‘restitutionary’ seems inapt, as s 21 limits the award to the cost of providing the service.

(93) Williams v Wentworth 5 Beav. 325 (1842).

(94) Matheson v Smiley [1932] 2 DLR 785 (Canada).

(95) Skibinski v Community Living British Columbia, 2012 BCCA 17 (Canada).

(97) The Cargo ex Port Victor [1901] P 243, 249.

(98) Courts sometimes say explicitly that damages are dependent on a breach of duty, but these comments are invariably found in cases where the alleged breach, if proven, is clearly a wrong: see e.g. Bourhill v Young [1943] AC 92, 106 (‘Damage due to the legitimate exercise of a right is not actionable . . . The damage must be attributable to the breach by the defendant’: Lord Wright).

(99) ‘We think that the true rule of law is that the person who, for his own purposes, brings on his land, and collects and keeps there anything likely to do mischief if it escapes, must keep it in at his peril, and, if he does not do so, he is prima facie answerable for all the damage which is the natural consequence of its escape’: Rylands v Fletcher [1861-73] All ER Rep 1, 5 (Lord Blackburn).

(101) [1868] LR 3 HL 330.

(102) See McBride & Bagshaw (2015) 486–9; Zipursky & Goldberg (2016) 743.

(103) American courts interpret this category narrowly; most of the cases involve bursting reservoirs, blasting, and dangerous animals: Goldberg (2016).

(104) See McBride & Bagshaw (2015) 499–502. Although ‘abnormally dangerous activities’ is not a legal category in England, English authors have used it to explain Rylands and related liabilities: ‘The true situation seems to be that he who maintains for his own advantage a peculiarly dangerous thing in proximity to others, necessarily imposes upon those others a risk of injury . . . greater than is to be reasonably expected in the ordinary circumstances of social life, and it is therefore just and expedient that he himself should bear the risk of making good any damage to others which results from the maintenance of the object. This is certainly a liability, but it is a confusion of ideas to [say that it arises from the breach of] a duty’: Allen (1931) 193–4.

(105) This liability is now enshrined in the Hotel Proprietors Act 1956, s 2.

(106) I discussed necessity cases, and also the cases in the next two categories, in Chapter 6 (under the heading ‘Substantive-Law Reasons for Refusing Injunctions’).

(107) [1910] 109 Minn 456.

(109) Gardner’s view, it appears, is that if you intentionally use others’ property without their permission then you have necessarily committed a wrong and, accordingly, should pay damages for any injury you caused. But what if you use another’s property to save that person’s life? Imagine that you are having a heart attack while locked in your car. I break into the car without your permission (because you are unconscious) and save your life. It appears that, on Gardner’s view, I should be liable for the damage that I caused to your car (and probably also liable for battery, as I touched you without your permission). There do not appear to be any cases on point, but it is doubtful that a court would agree.

(110) [2003] EWHC 793.

(111) 26 NY 2d 219 (1970).

(112) [1977] Ch 106.

(113) The actual sum awarded was nominal or almost nominal because the court assumed that the islanders had not suffered any loss because of the failure to replant. This assumption seems to have been misplaced as it turns out that the island was not rendered permanently uninhabitable: many years later some of the islanders were able to return to their homeland.

(114) See e.g. Dennis v Ministry of Defence [2003] EWHC 793; Bracewell v Appleby [1975] Ch 408; Denne v Light (1857) 8 De GM & G 774; Behrens v Richards [1905] 2 Ch 614.

(115) Courts never award substitutionary damages in lieu of injunctions (because it is not possible to purchase substitute performance of an injunction). However, they award substitutionary damages in lieu of specific performance, and presumably would have awarded them in Tito if they thought the defendant should have performed the requested action. The fact that the court limited the claimants to compensatory damages supports the conclusion that they refused specific relief for substantive-law reasons.

(116) [1924] AC 851.

(117) See Stevens (2007) 174–86. In the 1970s, English courts appeared to be moving in the direction of accepting that a wrong to a third party is, in general, sufficient to establish a breach of the duty of care in negligence: Anns v Merton London Borough Councils [1978] AC 728. They have since retreated: Caparo Industries plc v Dickman [1990] 2 AC 605; Murphy v Brentwood [1991] 1 AC 398.

(118) White v Jones [1995] 2 AC 207.

(119) OBG v Allan [2008] 1 AC 1.

(120) As was true of restitutionary awards, compensatory damages are sometimes justified, instrumentally, in terms of the incentives they provide. I explain why instrumental theories are not competitors to an injustice-based theory in note 17.

(122) Ibid 38. In this essay, Ripstein appears to endorse Honoré’s approach (see Honoré (1999)), which is broadly consistent with my interpretation of compensation. Ripstein writes that the victim’s ‘claim is that it would be unfair to her to relieve the defendant of liability, because the costs associated with this particular risk properly belong with the defendant rather than the [claimant]’ (58). In his later work, Ripstein defends the continuity thesis explanation of damages, in which ‘fairness’ and ‘allocation’ drop out of the picture: ‘I mention both Honoré and Darwall not to dispute their accounts of responsibility and personal accountability, but because I think however important these aspects are in other domains of life, they do not figure in an account of the moral basis of the law of private wrongs’: Ripstein (2016) 16.

(124) Ripstein (2001) 38. ‘In torts we are dealing with activities and their accident costs. The question is always who should bear these costs’: Coleman (1980) 120–1. See also Prosser (1953) 17; Perry (2001a) 60; Honoré (1999) 78–9.

(125) See the work cited in note 124, especially Perry (2001a).

(126) My explanation broadly follows Honoré (1999) 78–9.

(127) On the significance of factual causation, see Honoré (1999) 14–40; Perry (2001a) 50.

(128) Honoré (1999) 27. See also Perry (2001a) 52 (‘Outcome-responsibility [i.e. factual causation] is reason-affecting, but, as we have seen, there are no grounds for thinking that it is sufficient by itself to justify shifting a loss. The question to be asked, then, is whether there is some other factor that, together with outcome-responsibility, could serve as an appropriate justification for redistribution.’)

(129) See Honoré (1999) 27–8. The closest that Honoré comes to offering a general statement of these conditions is the following: ‘The argument for holding people responsible to others for harmful outcomes is that it is fair to make the person to whom the advantages will flow from an uncertain situation over which he has some control (or which he has chosen to enter into) bear the losses that may equally flow from that situation’: Honoré (1999) 78–9. This proposition covers most, but not all, the conditions discussed below.

(131) ‘Responsibility is not just a function of the quality of will manifested in conduct, nor of the quality of that conduct. It is also concerned with the interest we all share in security of person and property, and with the way resources and risks are distributed in society. Responsibility is a relational concept. Herein lies the key to explaining the justifying obligations of repair regardless of fault’: Cane (2002) 109. See also Herman (1993) 97–9; Feinberg (1978) 102; Honoré (1999) 1–40; Gardner (2001); Ripstein (1999) 4.

(132) Robert Stevens also explains vicarious liability on the ground that the employee’s actions are attributed to the employer: Stevens (2007) 257–74. However, while the examples Stevens offers in support of his position—such as attributing an individual’s action to a corporate body (or to similar entities, such as sports teams) or to someone who has authorized the action—demonstrate that the law frequently attributes actions, they do not explain why the law attributes employees’ actions to employers (and, further, attribute only some of those actions). My explanation is fairness.

(134) ‘The volitionist/distributive argument is an attempt to clarify our evaluative attitudes regarding actions that produce harmful outcomes, and to show why it is appropriate to conclude, in light of those attitudes, that under certain circumstances reparation is due. I do not think that anything more than this can be said’: Perry (2001a) 55; ‘[P]ervasive and unshakeable features of human practical thought need no justification, though they call for an explanation’: Raz (1986) 288–9. See also Williams (1981); Nagel (1979); Honoré (1999); Gardner (2001).

(135) Birks (2005) 4–19; Burrows (2011) 201–54.

(137) See generally, McBride (1994); Muir (1990). Though outside this book’s scope, various remedies for actions within Equity’s historically exclusive jurisdiction (e.g. trust law) may also fall into this category. Possible examples here include Trustee Remuneration Orders (Re Duke of Norfolk’s ST [1982] Ch 61 at 79; Trustee Act 2000, s 2); Variation of Trust Orders; and, where they are recognized, remedial constructive trusts (Pettkus v Becker [1980] 2 SCR 834).

(138) Closely related to the remedies discussed in this section are remedies available to claimants who, notwithstanding that they did not have a contract with the defendant, provide requested services or incurred a liability as a result of performing a requested action: see McBride (2004) 40–5.

(139) Jennings v Rice [2002] WECA Civ 159. See generally, Peel (2015) 119–35, 153–72.

(140) See generally McFarlane (2016). However, English courts appear to have awarded compensation for detrimental reliance on representations that did not involve property: see e.g. William Lacey (Hounslow) Ltd v Davis [1957] 1 WLR 952; Peel (2015) 134.

(141) Restatement 2nd Contracts, s 90; Walton Stores (Interstate) Ltd v Maher (1988) 164 CLR 387

(142) Mea culpa: Smith (2004) 233–45.

(143) See e.g. Spence (1999).

(144) See e.g. Willmott v Barber (1880) 15 Ch D 96.

(145) (1988) 164 CLR 387, 461. See also Mason CJ and Wilson J at 495 (‘Holding the representor to his representation is merely one way of doing justice between the parties’). Walton Stores includes forty-three references to ‘injustice’, ‘fairness’, and variations on these terms (e.g. ‘unjust’, ‘unfair’). Although it refers to promises, s 90 of the American Restatement 2nd Contracts explicitly grounds its cause of action, and limits its remedial response, to what justice requires: ‘(1) A promise which the promisor should reasonably expect to induce action or forbearance on the part of the promisee or a third person and which does induce such action or forbearance is binding if injustice can be avoided only by enforcement of the promise. The remedy granted for breach may be limited as justice requires.’

(146) Kelsen v Imperial Tobacco Co (of Great Britain and Ireland) Ltd [1957] 2 QB 334.

(147) [1974] QB 142.

(148) [1974] QB 142,156.