THE ACTUAL LOSS ILLUSION

Abstract

On the orthodox account of the private law compensatory principle, the claimant is compensated for the loss that they actually suffered because of the defendant’s wrong. Although the principle has various exceptions, it is widely accepted in both case law and academic commentary. We argue that it is nevertheless flawed, both doctrinally and theoretically. Claimants are never really compensated for their actual loss, and, contrary to popular belief, leading theoretical accounts of private law compensation (corrective justice and the continuity thesis) suggest that a principle of compensation for actual loss is not desirable in any event.

I. Introduction

On the orthodox understanding of the private law compensatory principle, the claimant is entitled to be put, as far as money can do so, in the position that they would actually have been in were it not for the wrong. ¹ In other words, the claimant is compensated for their actual loss – the loss that they personally suffered because of the wrong. As the Supreme Court recently put it, “justice is not achieved if a claimant receives less or more than its actual loss”. ² On this account, which we refer to as the “actual loss account”, compensation for the claimant’s actual loss is more than just a default starting point – it is the desired moral outcome.

It is equally well accepted, however, that the actual loss account provides an incomplete understanding of private law compensation. There are many situations in which the claimant receives compensation that is either more or less than their actual loss. This is sometimes because it is difficult to prove the claimant’s actual loss precisely, ³ or to quantify that loss in money terms. ⁴ In other cases, however, compensation is simply not intended to reflect the claimant’s actual loss at all. The most obvious examples are when the rules of remoteness, intervening causation, mitigation and so forth limit the claimant’s damages below their actual loss (caused in the “but for” sense), ⁵ but the examples are far more widespread and, in various instances, claimants are even compensated for losses that they have not actually suffered. ⁶

It is possible to explain these rules as mere exceptions to the actual loss account. ⁷ An alternative explanation, however, is that they are counterexamples to it, which suggest that the law does not really reflect the actual loss account at all. Our primary claim in this article is that the law’s commitment to the actual loss account is either illusory or misplaced and that the true nature of the compensatory principle in private law has not been satisfactorily explained. By demonstrating the inadequacy of the actual loss account, we aim to clear the ground for further discussion and, in this vein, we offer some preliminary thoughts on the possible future direction of legal theory in the area.

To our knowledge, the actual loss account has never been seriously challenged in the case law or literature, ⁸ and it has a strong foothold in academic thinking. Andrew Burrows, for example, criticises a number of cases on the basis that the compensation awarded did not align with the claimant’s actual loss. ⁹ Eric Descheemaeker also questions various examples of what he perceives to be the impermissible “standardisation” of damages in tort law – that is, compensation for losses that an ordinary person in the claimant’s position would have suffered, rather than the loss that they actually suffered. ¹⁰ The reason why such standardisation is “undesirable” is said to be “self-evident, stemming as it does from the most basic principle of the law of compensatory damages: restitutio in integrum”. ¹¹

While a number of theorists accept that compensatory damages are sometimes awarded on a basis other than actual loss, they do not deny that compensating the claimant for their actual loss is often the law’s purpose. Most notably, Robert Stevens has posited the existence of so-called “substitutive” damages, which provide a substitute for the right that has been infringed, to explain instances of compensation in the absence of factual loss. ¹² In his most recent work, Stevens describes such damages as compensation for the wrong, which he distinguishes from compensation for the counterfactual loss that is caused by the wrong. ¹³ Similarly, Jason Varuhas argues that compensation for “normative” damage (i.e. for interference with a protected interest) is available for “vindicatory” torts in the absence of factual loss, but accepts that compensation for other torts (negligence in particular) is limited to the factual effects of the defendant’s wrong. ¹⁴ These accounts are helpful, insofar as they demonstrate that compensatory damages cannot be explained solely in terms of actual loss. However, they do not entirely reject the actual loss account, as they still accept that, in some cases at least, the law’s purpose is indeed to compensate for actual loss. The actual loss account is therefore pervasive among both the proponents and the critics of its accuracy as the sole explanation of compensatory damages. Our argument in this article is that the actual loss account does not provide an accurate explanation of the law in any case and, to that extent, it goes further than existing accounts.

Our argument unfolds as follows. In Sections II and III, we put forward several reasons, both doctrinal and theoretical, why we believe that the actual loss account is unsatisfactory. In short, we argue that the mere fact that a claimant has suffered an actual loss because of some wrong is never a sufficient reason (and is sometimes not even a necessary reason) to award compensation and that, in fact, claimants are never (or only rarely) compensated for their actual loss. Further, contrary to popular belief, the actual loss account sits uncomfortably with existing theoretical explanations of private law compensation, in particular corrective justice theory and the continuity thesis (or, at least, the leading statements of those theories). For these reasons, it is hard to see why actual loss ought to be the “catchcry” of compensation.

One might respond that, whatever the shortcomings of the actual loss account, it works sufficiently well in practice and that, if it produces an outcome that is unacceptable, it is always possible to recognise an exception. In Section IV we argue, to the contrary, that whether the law continues to adhere to the actual loss account, or instead openly recognises its inadequacy as an explanation of the law, does make a practical difference. We argue that the pervasiveness of the actual loss account has contributed to an actual loss “mindset”, according to which claimants should be compensated for all (and only) their actual loss, unless demonstrable and exceptionally strong policy reasons suggest otherwise. The result is that, in many cases, the possibility of awarding compensation for more or less than the claimant’s actual loss is either not considered at all or dismissed out of hand. Indeed, even where the court does award compensation that does not reflect the claimant’s actual loss, it often attempts to justify that award within an actual loss framework. While this might lead to the correct outcome in the particular case, artificial reasoning of this kind can break down when applied to other situations, distorting the outcome, or at least making it more difficult to reach the correct conclusion. Even if courts are alive to the full range of considerations that bear upon whether to award compensation, the actual loss account often enables them to justify the outcome simply by referring to the claimant’s actual loss, meaning that the true grounds for their decisions might not be disclosed. This is undesirable if we value transparency in judicial reasoning. In Section V, we consider what an alternative to the actual loss might look like and whether (all things considered) the law should reject the actual loss account in favour of such an account. In Section VI we conclude.

II. The Illusion

Despite the rhetoric, when a claimant suffers an actual loss that it would not be reasonable to require the defendant to compensate, the actual loss principle yields without a struggle. As Lord Sumption pointed out, the fact that the claimant has suffered loss because of the defendant’s wrong “has never been enough” to justify compensation; the law recognises that “a defendant is not necessarily responsible in law for everything that follows from his act, even if it is wrongful”. ¹⁵

It is nevertheless generally accepted (or assumed) that the existence of various exceptions to the actual loss principle does not undermine its status as the fundamental principle of compensation. Many rules have exceptions and the actual loss principle is no different in this regard. On the actual loss account, while the exceptions give effect to countervailing normative considerations in specific contexts, the law’s primary moral commitment is still to compensate claimants for their actual loss.

In this section, we identify various reasons why we believe this supposed commitment to be illusory. The nature and content of the exceptions is such that they – and not any actual loss principle – provide the true explanation for the outcome in every case and, in practice, lead to the conclusion that claimants are seldom (if ever) actually compensated for (all of and only) their actual loss. We also highlight various remedial conventions that sustain the actual loss illusion by keeping other considerations out of view, making it appear as though the court’s task is simply to quantify the claimant’s actual loss, when in reality that task is a more difficult one.

A. The Exceptions

The idea that claimants are not necessarily entitled to compensation for their actual loss is effected by a wide variety of rules. Perhaps most obviously, the rules of remoteness (at least as they apply to unintentional torts and breaches of contract) hold that the defendant is not responsible for loss of a kind that was not reasonably foreseeable. ¹⁶ The defendant is also not responsible for loss that results from an “intervening cause” subsequent to the defendant’s wrong, ¹⁷ from the claimant’s unreasonable failure to take mitigating steps, ¹⁸ or which is not within the scope of the defendant’s duty. ¹⁹ In the case of breach of contract, the defendant is not responsible for losses for which they have not assumed responsibility. ²⁰ There are also more specific rules, such as the rule that a defendant in breach of contract is not required to compensate the claimant for emotional distress, except where providing emotional satisfaction to the claimant was an important purpose of the contract, or where the breach has caused the claimant physical inconvenience. ²¹

The incorporation of a “reasonableness” standard in deciding the measure of loss is another important example. In Ruxley Electronics and Construction Ltd. v Forsyth, ²² for example, the claimant was not compensated for the cost of rectifying the defendant’s defective building work because it would not have been reasonable to incur those costs. ²³ In personal injury cases, Lambert J. summarised the position as follows:

The Claimant is entitled to damages sufficient to meet her reasonable needs arising from her injuries and, in deciding whether a head of loss is recoverable in the amount claimed or at all, the Court should assess the reasonableness of the head of loss and its amount … [P]roportionality [is also] relevant, in the sense that the Court should have regard to the extent to which, in respect of any claim, the same or substantially the same result could be achieved by other less expensive means. ²⁴

In XX v Whittington Hospital NHS Trust, Baroness Hale (delivering the majority judgment) reiterated that, to recover damages for the cost of taking future steps, “to restore what has been lost, the steps taken must be reasonable ones and the costs thereby incurred must be reasonable”. ²⁵ A clear application of this idea is the decision of the High Court of Australia in Sharman v Evans. ²⁶ In that case, the claimant was rendered a quadriplegic by the defendant’s negligence and required nursing care for the rest of her life. The issue was whether she could recover the cost of being cared for in her own home, or only for the much lower cost of being cared for in a hospital. In deciding this issue, Gibbs and Stephen JJ. stated that “the appropriate criterion must be that such expenses as the plaintiff may reasonably incur should be recoverable from the defendant”. ²⁷ As they put it: “The touchstone of reasonableness … is, no doubt, cost matched against health benefits to the plaintiff. If [the] cost is very great and benefits to health slight or speculative the cost-involving treatment will clearly be unreasonable, the more so if there is available an alternative and relatively inexpensive mode of treatment, affording equal or only slightly lesser benefits.” ²⁸

Their Honours concluded on the facts that the (amenity) benefits of being cared for at home did not justify the additional cost, ²⁹ although they did justify the cost of occasional home stays. ³⁰ The claimant could therefore only obtain compensation for the loss that she would suffer if cared for in hospital (with occasional home stays). ³¹ In this decision, the purpose of the award was clearly to award compensation for the costs that would be reasonable, regardless of whether the claimant would actually incur them. The claimant’s likely actual loss was not the primary consideration. Although the normative assumptions underpinning the decision in Sharman appear to have shifted in favour of independent living in recent years, this merely reinforces the fact that the outcome is not determined solely by reference to the claimant’s actual loss. ³²

There are also many rules that allow the claimant to be compensated for more than their actual loss (or for a loss that a third party has suffered). ³³ According to the “avoided loss” rule of mitigation, the claimant cannot be compensated for a loss that they have avoided, except where the actions that avoided the loss did not “arise … out of the consequences of the breach”. ³⁴ Thus, in A.K.A.S. Jamal v Moolla Dawood Sons & Co., ³⁵ in which the defendant purchaser refused to complete a sale of shares, the claimant was compensated for the difference between the contract and market prices of the shares at the date of breach, despite the fact that the claimant had sold the shares to a third party for more than the original contract price and therefore suffered no loss (and actually made a profit) from the breach.

In various other situations, the claimant can be compensated for a loss despite having received a benefit that offsets the loss (in whole or in part). The two main examples are (1) benefits received under an accident insurance policy (paid for by the claimant); and (2) benefits provided to the claimant by third parties out of benevolence. ³⁶ Another recent example is The New Flamenco, in which the claimant shipowner was compensated for the entire loss of profit on the remainder of a charterparty that the defendant had repudiated, despite having obtained a substantial benefit from selling the ship in favourable market conditions after the breach (which they would not otherwise have done). ³⁷ Although some benefits do reduce the claimant’s damages, ³⁸ the point is that there are clearly cases where they do not. ³⁹

What these exceptions (which are not the only exceptions) demonstrate is that the fact that the claimant suffered an actual loss is never a sufficient reason to award compensation for that loss. Whether the claimant is permitted to receive compensation for a certain kind of loss is determined by other considerations (such as the scope of the defendant’s duty, whether the loss is reasonable to incur, what mitigating steps the claimant should take and so forth), ⁴⁰ which apply in every case and are not merely exceptional. The claimant’s actual loss plays (at most) a subsidiary role once, and only once, those other considerations have already justified the conclusion that the claimant should receive compensation for that kind of loss. It is these considerations, and not the claimant’s actual loss, that provide the true explanation for the outcome.

This is further supported by the fact that, even as the law currently stands, claimants are never really compensated for their actual loss. Compensation is only ever awarded for a tightly constrained subset of the consequences of the wrong. Consider Ernest Weinrib’s well-known example of a claimant who is injured by the negligent driving of the defendant while on the way to the airport. ⁴¹ The incident causes the claimant to miss their plane. As it turns out, the plane crashes, killing everyone on board, which would have included the claimant had they been on board. For reasons of either remoteness or intervening causation, the fact that the wrong benefited the claimant (by preventing their death) is not relevant in assessing compensation for the wrong.

This example highlights a more general point about the nature of compensation: in every case, the defendant’s wrong will have some coincidental effect on the claimant’s life – large or small, for better or for worse – which may, in turn, have further cascading effects, none of which are considered in assessing the amount of compensation that the claimant will receive. It is obviously impossible to know the full range of effects, but the point is that they are irrelevant even if they are known. The claimant’s compensation is therefore almost always more or less than the totality of the effect of the wrong on them or, in other words, than their actual loss. ⁴² If the outcome entailed by the actual loss account never actually applies, this indicates that something else is likely going on and that actual loss is not really the underlying principle.

B. Remedial Conventions

It might be thought that our arguments above do not fit the reality that, in many or perhaps even most cases, courts appear to award compensation for some actual loss suffered by the claimant without engaging with other considerations that bear upon the appropriateness of compensation. It might be inferred that such considerations are only relevant at the margins and that, in normal cases, the claimant’s actual loss is the only relevant factor. In our view, however, this appearance is also an illusion, which is sustained by various remedial conventions that conceal the role of other normative considerations. These conventions include the categorisation of losses under heads of damage and the rule that claimants are free to spend their damages however they want (the “free to spend rule”). By keeping other relevant considerations out of view, these conventions also arguably contribute to an actual loss mindset more generally.

1. Heads of damage

The first point to emphasise is that, in many cases, it is clearly appropriate for the claimant to be compensated for some actual loss that they have suffered, so the actual loss account appears to function perfectly well. However, this is only because the appropriateness of compensating claimants for certain losses is assumed and seldom (if ever) challenged. One way in which these assumptions are engrained in the law is through the recognition of “heads of damage”. When a claimant suffers damage to their person or property (or indeed to any other protected interest), a range of needs or losses, pecuniary or non-pecuniary, might result. Conventional heads of damage provide a shorthand means of identifying and categorising these losses. As Windeyer J. put it, heads of damage “provide a convenient reminder of matters that ought not to be forgotten”. ⁴³

When a court compensates a claimant for a loss falling within an established head, the decision appears to be mainly one of fact. The court simply identifies the loss and awards compensation; no independent judgment is made about whether the loss should be compensated. It is easy to see this process as reflecting an actual loss principle. It would follow on this view that, although heads of damage identify some of the most common kinds of loss, there is no reason why any (and every) other loss should not be compensated; all are equally losses caused by the defendant’s wrong. One example of this kind of reasoning can be seen in the judgment of Lord Scarman in Pickett v British Rail Engineering Ltd., where his Lordship suggested that there is no basis for distinguishing loss of earning capacity (an established head of damage) from the loss of other economic benefits resulting from personal injury and that all such losses should be compensated. ⁴⁴

We believe this understanding of the function of heads of damage to be incorrect. A preferable explanation is that heads of damage instantiate a generalised normative decision that certain kinds of loss ought to be compensable. When a person is negligently injured, for example, the law deems it appropriate for the defendant to compensate the claimant for any loss of their earning capacity and assigns that loss a specific head of damage for this reason. ⁴⁵ The fact that this normative assessment is not made afresh on a case-by-case basis does not mean that it has never been made. ⁴⁶ We simply accept that it has been made on enough occasions to justify it being taken for granted, exceptional cases aside, on future occasions.

This understanding of heads of damage is reflected in the recent decision of the High Court of Australia in Amaca Pty Ltd. v Latz. ⁴⁷ Latz contracted mesothelioma through the negligence of his employer, Amaca. After he retired, his condition was diagnosed as terminal and as reducing his life expectancy by around 16 years. Amaca did not deny that its negligence had caused Latz’s damage. The appeal focused on the issue of whether Latz (or his spouse) was entitled to compensation for the loss of the age and superannuation pension payments that he would have received during his 16 lost years. Amaca argued that these losses were not compensable as they did not fall within any established head of damage for personal injury set out in the court’s previous decision in CSR Ltd. v Eddy – namely, (1) non-pecuniary loss, (2) loss of earning capacity or (3) actual financial loss (e.g. medical costs). ⁴⁸ Latz, conversely, argued that their Honours in CSR v Eddy “were not purporting to provide an exhaustive list of the compensable heads of damage” ⁴⁹ and that his losses “fit squarely within the compensatory principle, as articulated by Lord Blackburn in Livingstone v Rawyards Coal Co”. ⁵⁰

The majority concluded that the claimant could obtain compensation for the loss of the superannuation payments, but not the age pension payments. Their Honours accepted the following approach to determining the claimant’s compensation, which clearly recognises that considerations other than the claimant’s actual loss play a pervasive role: “First, it is necessary to identify Mr Latz’s loss. If a loss is identified then, as Lord Wilberforce stated in Pickett v British Rail Engineering Ltd., the law has to answer a question: is that loss the loss of ‘something for which the claimant should and reasonably can be compensated?’” ⁵¹

Although the first sentence accepts that the practical starting point when assessing the claimant’s compensation might sometimes be their actual loss, the second sentence clearly recognises that identifying an actual loss is not a sufficient basis for awarding compensation. On the facts, the majority concluded that the loss of the superannuation payments was compensable, as they were “intrinsically connected to earning capacity, representing … a species of remuneration – financial rewards from work”, ⁵² but that the loss of age pension payments was not, as they were “not a result of, or intrinsically connected to, a person’s capacity to earn”. ⁵³

The reasoning of the minority provides even stronger support for our account of the nature of heads of damage and for rejecting the actual loss account more generally. Their Honours took the view that “the compensatory principle is concerned with the measure of damages required to remedy compensable damage. In applying the principle, it is necessary first to establish whether the loss claimed is compensable … One cannot invoke the compensatory principle to identify whether a particular head of damage is compensable”. ⁵⁴

This statement echoes Lord Hoffmann’s judgment in SAAMCo, ⁵⁵ to which their Honours explicitly referred, and is clearly inconsistent with any blanket actual loss principle. Their point is that compensation for a loss is not awarded simply because the claimant suffered it; it is only awarded if compensation for that kind of loss is determined to be appropriate on other grounds. The minority concluded that neither loss mentioned above was compensable, since neither fell within any of the three traditional heads of damage. In their view, the boundaries of the second and third heads of damage serve “to maintain the distinction between economic loss which is compensable as such and that which is not”. ⁵⁶ The loss of age pension payments was clearly not a loss of earning capacity, since the accrual of those payments “has nothing to do with the exercise of or inability to exercise earning capacity”. ⁵⁷ The minority also did not regard the loss of superannuation payments as a loss of earning capacity, since the claimant had already retired; rather, the loss was of his ability to enjoy the financial resources that he had accrued by the previous exercise of his earning capacity, which was not compensable. His earning capacity as such had not been “compromised”. ⁵⁸ In making this argument, the minority explicitly rejected the view of Lord Scarman in Pickett (mentioned above) that the loss of any economic benefit is compensable. ⁵⁹ Importantly, the minority did not treat the categories of compensable loss as necessarily closed. They recognised that the heads of damage can be extended in novel cases but concluded that this was inappropriate on the facts. ⁶⁰

It follows that, in the entire court’s view, heads of damage are more than mere factual categories of loss and instead reflect generalised normative judgments about whether certain kinds of loss should be compensated. This reality is masked in cases where the court applies an established head of damage, but is exposed in novel cases involving losses that fall outside the established heads (or the established categories of non-compensable loss), where the court is forced to confront the normative question of whether the loss should be compensated.

2. The “free to spend” rule

Another reason why courts often do not need to consider the appropriateness of compensating certain losses is that they have explicitly renounced any obligation to superintend how claimants spend their damages. As the High Court of Australia explained in Todorovic v Waller, “the court has no concern with the manner in which the plaintiff uses the sum awarded to him; the plaintiff is free to do what he likes with it”. ⁶¹ This enables claimants to frame their loss in terms of established heads of damage, even if, in fact, they are unlikely to suffer losses within those categories. For the reasons mentioned above, this operates to conceal the underlying normative issues.

Let us suppose, for example, that the claimant is injured by the defendant’s negligence. The claimant is also a member of an obscure “new-age” religious group, which extols the benefits of crystal therapy in the healing process, a fact of which the defendant is entirely unaware. As treatment for his injuries, the claimant plans to attend weekly sessions with a “crystal healing practitioner” at a cost of $100 per session for a period of approximately two years. The claimant has no intention of undergoing conventional medical treatment as the dictates of his religion forbid it.

If the claimant chooses to avoid any reference in his pleadings to his actual likely costs and claims instead for general medical expenses, he would then be at liberty to spend the sum awarded on a course of crystal healing if he so wished. In this scenario, the court would have neither the need nor the opportunity to consider whether the claimant ought to recover for the costs associated with crystal healing and whether the burden of paying for those costs should be imposed on the defendant, since it is well established that general medical expenses are compensable. What, though, if the defendant were aware of the claimant’s religious views and could prove that he would not spend the sum awarded on medical expenses? Since the claimant must prove his loss, ⁶² he would no longer be entitled to damages for ordinary medical expenses and would be forced to claim (if at all) for the costs associated with his crystal therapy. In this revised scenario, the court would be required to determine whether those costs are compensable, having regard to their reasonableness. In many cases, however, the free to spend rule removes the need to consider such questions.

III. Theoretical Shortcomings

One might respond that nothing we have said in the preceding sections changes the fact that compensation for actual loss is still a desirable moral principle. We should therefore strive to implement that principle as best we can, ensure that it is reflected in our law and depart from it only when justified by particularly strong countervailing considerations. In this section, we argue that the moral case for compensating actual loss is not as clear-cut as is usually thought. Indeed, contrary to popular belief, the actual loss account sits uneasily with some of the leading theoretical justifications of private law compensation, namely corrective justice and the continuity thesis.

A. Corrective Justice

It appears to be a widespread view that “corrective justice” requires defendants to compensate claimants for their actual loss. On this view, although departures from the actual loss principle might be justified, this can only be on the basis of distributive or policy concerns. For example, John Goldberg and Benjamin Zipursky suggest that “[a]t the center of the corrective justice account is the idea … of full or ‘make whole’ compensation”. ⁶³ According to Descheemaeker, the view that compensatory damages are concerned with “the consequences of the wrong for the claimant … can easily be defended on first principles of corrective justice”. ⁶⁴ Burrows argues that corrective justice cannot be the justification of “the law on remoteness or mitigation or the general non-recovery of mental distress in contract law where the courts rightly invoke policy concerns external to the parties in developing and explaining the law”. ⁶⁵ And Katy Barnett argues that the rules of mitigation “involve wider questions of distributive justice beyond the simple correction of the breach”. ⁶⁶

With respect, these arguments seem to assume that “corrective justice” refers only to reversing or “correcting” the actual consequences of a wrong. We believe this to be a misconception. ⁶⁷ As Weinrib explains, corrective justice refers to the norms that explain when and why an interaction (or “transaction”) between two parties creates an injustice between them, not simply to the “correction” of that injustice. ⁶⁸ In other words, it refers to “the system of norms within which the operation that corrects makes sense”, ⁶⁹ which norms explain when and why correction is justified. There is nothing in this that necessarily requires the claimant to be compensated for their actual loss; what is being corrected is the injustice.

It is true that Weinrib and Jules Coleman (arguably the two leading corrective justice theorists) have made statements that, on their face, appear to endorse the actual loss account, but on a closer reading neither actually does so. For example, Coleman states that:

To annul a wrong is to eliminate its effects in the world, as much as is feasible. It is to return the world to where it would have been had the wrong never been committed. In this sense, fully to repair the wrong is to repair not only the wrong but its consequences. That would mean … compensating for whatever damage one has caused others to suffer. ⁷⁰

James Goudkamp and John Murphy rely on this passage as evidence that Coleman endorses the actual loss account. ⁷¹ However, in making these statements, Coleman was actually describing a position that he intended to (and did) reject. ⁷² After offering an example similar to Weinrib’s plane example, he argues that “[t]he problem with this account [quoted above] of the relationship between the wrong and the losses that result from it is that it based on an unacceptable understanding of what is required in order to ‘repair’ or annul a wrong”. ⁷³ In Coleman’s own view, “[t]he duty to repair is not the duty to annul all the consequences of one’s wrongs”. ⁷⁴ Rather, “the duty of wrongdoers in corrective justice is to repair the wrongful losses for which they are responsible”. ⁷⁵ Coleman’s account of responsibility for wrongful losses (of which he only provides a brief and incomplete outline) includes many existing and well-recognised limiting principles, namely that the loss must result from the aspect of the defendant’s conduct that makes it wrongful, from a risk whose possibility makes that aspect of the defendant’s conduct wrongful and (possibly) from a risk that was foreseeable. ⁷⁶ Moreover, Coleman also appears to accept that corrective justice can allow compensation for more than the claimant’s actual loss, given his use of the plane example. A claimant in such a case is still entitled to their medical expenses, despite being better off overall as a result of the wrong.

Similarly, Weinrib suggests that tort law is consistent with corrective justice in that it “places the defendant under the obligation to restore the plaintiff, so far as possible, to the position the plaintiff would have been in had the wrong not been committed”. ⁷⁷ He makes a similar claim about contract law. ⁷⁸ Weinrib nevertheless recognises various exceptions to this principle. For example, he argues that the rules of mitigation, remoteness and intervening cause are compatible with corrective justice, ⁷⁹ which can all result in the claimant recovering compensation for less than their actual loss.

More significantly, however, the actual loss account sits uneasily with Weinrib’s general argument that corrective justice is concerned with “normative” rather than “factual” losses (and gains). In Aristotle’s original account, ⁸⁰ a corrective injustice occurs when, because of some interaction (or “transaction”) between two parties, one party has made a gain and the other a correlative loss. Corrective justice is then restored by annulling the gain and loss, by subtracting the gain from the party who has gained and adding it to the one who has lost. This restores the parties’ “equality”, in the sense that they each now have neither more nor less than they are entitled to. Weinrib identifies this as the justification, at a general level, for all private law norms and remedies. ⁸¹

In explicating this account, Weinrib distinguishes between two different kinds of gain and loss, which he calls “normative” and “factual” gains and losses respectively: “In their factual aspect, gains and losses refer to changes in the condition of the litigants’ holdings; in their normative aspect, gains and losses refer to discrepancies between what the parties have and what they should have”. ⁸² Weinrib suggests that “we may say that a person enjoys a normative gain when there is justification for the law’s diminishing his or her holdings, and that a person endures a normative loss when there is justification for the law’s augmenting his or her holdings”. ⁸³ Normative and factual losses do not always align: it is possible to have a normative loss with no factual loss, or a factual loss with no normative loss, and the same applies to gains. ⁸⁴

In Weinrib’s view, corrective justice is concerned (at least primarily) with normative rather than factual gains and losses. A corrective injustice involves a normative gain and a correlative normative loss. ⁸⁵ In Aristotle’s account, gain and loss are simply metaphors for the parties having more or less than they should have, respectively. ⁸⁶ In assessing the claimant’s compensation, the issue is not whether the claimant suffers factual loss, but rather whether the defendant should compensate the claimant for a given loss.

The actual loss principle is also inconsistent with other aspects of Weinrib’s account, in that it is based on a “one-sided” justification. Weinrib argues that “a justificatory consideration that fits into the normative structure of corrective justice cannot have a justificatory force that reaches only one of the parties”. ⁸⁷ He points out (although in a slightly different context) that loss is a one-sided consideration: “the plaintiff’s loss lacks bipolar significance. At most, the loss justifies improving the plaintiff’s situation; it does not state a ground for taking something from the defendant.” ⁸⁸ The point is that, without more, the fact that the claimant has (or has not) suffered an actual loss cannot itself justify increasing (or reducing) the defendant’s liability. Such an argument focuses solely on the claimant’s factual situation, without reference to the various contributions of the claimant, defendant, or third parties to that situation. Weinrib later made the same point in the following way: “Being harmed is merely a fact – that the harmed person is now less advantageously situated than before – that in itself has no correlative normative significance.” ⁸⁹

As a result, there is no reason why the rules discussed earlier, involving departures from the actual loss account, could not be explained in terms of corrective justice. This does not mean that all departures from the actual loss account can be explained in terms of corrective justice; some are potentially based on distributive or policy reasons. Nor do we deny that rules reflecting corrective justice concerns could be further supported by distributive or policy reasons. The point is that corrective justice, on the leading expositions of that idea, does not support the actual loss account and is arguably incompatible with that account. This also suggests that departures from the actual loss principle are not necessarily subject to those concerns often expressed about the illegitimacy of distributive or policy reasoning in private law adjudication.

B. Continuity Thesis

Another influential explanation of private law compensation is the “continuity thesis”. ⁹⁰ According to John Gardner’s version of that thesis, ⁹¹ when the defendant breaches their primary duty, the reasons that support a secondary duty (if any) to pay compensation to the claimant are the same reasons that supported the original primary duty. Although it is no longer possible for the defendant to perform their primary duty, the reasons supporting that duty might continue to apply after the duty is breached and it might still be possible for the defendant to conform to those reasons, at least in part, by paying monetary compensation. ⁹² Compensation generally does not fully satisfy those reasons – they often can only be fully satisfied by the defendant performing their primary duty in the first place. But since this is no longer possible, the only remaining way for the defendant to conform to those reasons is in some “second best” way.

Take a defendant’s duty not to drive negligently (or injure others by such driving). There are various reasons why this duty might arise, one of which is that performing it can (without imposing an unreasonable burden on the defendant) avoid various kinds of pecuniary and non-pecuniary loss to other road users. ⁹³ If the defendant breaches their primary duty, they can still conform (at least partly) to the reasons that support the primary duty (namely, avoiding certain losses) by compensating the claimant for those losses.

In our view, the continuity thesis does not support the actual loss account. While the thesis is usually put forward as a positive explanation of private law compensation, it also identifies its limits. According to the continuity thesis, it is the reasons that justify the defendant’s primary duty that determine (prima facie) ⁹⁴ the amount of compensation that the defendant must pay, not the claimant’s actual loss. If the claimant suffers a kind of loss that is not part of the justification of the defendant’s duty (in the sense that avoiding that kind of loss is not one of the reasons that justifies the duty), then the continuity thesis provides no reason at all why the defendant should have to compensate the claimant for that loss. In these situations, it is not that the reason to compensate for actual loss is being outweighed by a countervailing reason – there is simply no reason to compensate the loss. Further, where the possibility of a certain kind of loss is a reason for the duty, that reason may require the defendant to compensate the claimant for that kind of loss even if the claimant does not actually suffer it. This might occur, for example, where the claimant would have suffered the loss but for some benevolent action by a third party, or where some conduct by the claimant falls within the exception to the avoided loss rule. ⁹⁵ The continuity thesis thus provides no support for the actual loss account. ⁹⁶

Some rules seem particularly amenable to explanation in terms of the continuity thesis. The SAAMCo “scope of duty” principle appears to be almost a direct application of that thesis. The law on intervening causes is another example. Generally, where the claimant intentionally or recklessly causes themselves loss after the defendant’s wrong, the defendant will not be responsible for that loss. ⁹⁷ On the other hand, where the purpose of the defendant’s duty was to prevent the claimant from causing themselves loss through the very kind of intentional or reckless conduct in which they engaged, the defendant will be responsible. For example, in Reeves v Commissioner of Police of the Metropolis, ⁹⁸ although the loss was caused by the claimant committing suicide, the defendant police officer’s duty was to take reasonable care to prevent the claimant from committing suicide. One of the purposes (or reasons) for the protective duty owed by the police was to prevent the very loss that occurred and the police were therefore liable for that loss.

The same considerations apply when the subsequent conduct is by a third party. If the purpose of the defendant’s duty is to prevent such conduct from causing loss to the claimant, the defendant will be liable for the loss. One example is a trustee’s duty to exercise due care and skill when giving possession of the trust property to an agent or third party. ⁹⁹ If the trustee breaches this duty, they will be responsible for any loss if the agent misappropriates the property. ¹⁰⁰ Although this rule has been explained on the basis that the rules of remoteness and intervening cause do not apply to equitable compensation, ¹⁰¹ a preferable explanation is that whether a loss is too remote or the result of an intervening cause depends on the underlying reason or purpose of the relevant duty. Since one purpose of the duty is to prevent the property being lost due to misappropriation by the agent, requiring the trustee to compensate for that loss is therefore justified.

The converse outcome was reached in the leading Canadian case of Canson Enterprises Ltd. v Boughton & Co. ¹⁰² The claimant, who was purchasing land, was represented by a solicitor who was also acting for other parties to the transaction (who were secretly “flipping” the land to the claimant) and was therefore in breach of fiduciary duty. The claimant would not have purchased the land had they been aware of the breach. After the purchase, the claimant built a warehouse on the land, which was defective due to the negligence of the (third party) engineers and the claimant suffered substantial loss. The court held unanimously that the solicitor was not required to compensate the claimant for this loss, even though it would not have occurred but for the breach. McLauchlin J. (whose reasoning has been preferred outside Canada) justified this on the basis that, although common law rules of remoteness etc. (supposedly) do not apply when assessing equitable compensation, “on a common sense view” there was “no link between the breach of fiduciary duty and the loss”. ¹⁰³ One possible explanation for this conclusion is that the purpose of the solicitor’s fiduciary duties was not to prevent the kind of loss that the claimant suffered. Their purpose was to prevent the claimant from entering an unfavourable transaction and the claimant could therefore only be compensated for the loss resulting from the aspects of the transaction that made it unfavourable at the time of entering it, which did not include the loss resulting from the defective warehouse, or the risk of such loss.

IV. Practical Shortcomings

It does not follow from our analysis in the preceding sections that the “actual loss” account must necessarily be rejected. Indeed, we readily acknowledge that, when combined with its recognised exceptions, it is logically capable of explaining legal outcomes and that, provided that the same considerations are ultimately taken into account, the approach taken may be of little or no consequence. There are good reasons, too, why defaulting to actual loss might be seen as a good thing. From a practical perspective, actual loss – gated through heads of damage – provides practitioners and judges with an intuitive and time-honoured starting point that appears to function perfectly well in most cases. If the actual loss approach were to be abandoned, it would need to be replaced with an alternate approach that functions just as well in practice while capturing more honestly the true nature of the enterprise. It might also be argued that it is simply too late in the day to begin questioning so fundamental a principle as restitutio in integrum, the unravelling of which risks damaging the very fabric of private law.

Our earlier arguments nevertheless demonstrate that something is not quite right with the actual loss account, which ought not to be dismissed simply because a workable alternative might be difficult to frame, apt to prove contentious, or unlikely to affect outcomes in routine cases. Our aim in this section is to flesh out this concern and to demonstrate why it does nevertheless make a practical difference whether the law continues to adhere to the actual loss account.

In short, where we believe the actual loss account goes wrong – or is, at least, misleading – is in appearing to treat the claimant’s actual loss as a sufficient reason for awarding compensation. As we have shown, this is simply incorrect: while the claimant’s actual loss is certainly relevant in many or even most cases, it is only ever as one part of a larger reason (or set of reasons): actual loss + “some other reason” = compensation. And, in some cases, actual loss is not even necessary.

In practical terms, adhering to the actual loss account might therefore negatively affect the quality of legal reasoning in one of (at least) two ways. We provide several examples below of cases where these effects are apparent, but they are by no means the only examples.

First, viewing the compensatory principle through the lens of actual loss promotes an “actual loss mindset”, according to which some courts (and commentators) genuinely believe that actual loss is a necessary and sufficient criterion for compensation and that there is a strong moral reason to award compensation that aligns with the claimant’s actual loss. This mindset can lead courts to award compensation for the claimant’s actual loss (or dismiss the possibility of not awarding it) without considering other relevant considerations, or to require unnecessarily strong and “exceptional” reasons to be demonstrated before a departure from actual loss will be countenanced. It can also encourage artificial reasoning to fit cases within an actual loss framework, even when they do not so fit.

Second, even if courts do not view actual loss as the touchstone of compensation, the actual loss account can nevertheless conceal the true reasons for their judgments. If actual loss is seen as a sufficient criterion for compensation, then a court may justifiably leave a victim to that loss, which might be nothing at all, without openly considering the merits of normative reasons to the contrary, unless those reasons support an established exception (such as remoteness of damage, etc.). To be sure, courts – and especially senior appellate courts – are likely to be alive to the full range of considerations that bear upon an award of damages in a novel case. However, if we value transparency in judicial reasoning then the compensatory principle ought surely to be framed in such a way as to encourage courts to disclose those reasons.

A. Damage to Property

One context in which we believe the actual loss account has generated artificial reasoning is in cases of damage to property. When the claimant’s property is damaged (or defective) due to the defendant’s breach, the claimant is entitled to recover the reasonable cost of repairing it. In some cases, this is possible even if the claimant has not incurred, and will not actually incur, that cost. A clear example is Powercor Australia Ltd. v Thomas, ¹⁰⁴ in which various fixtures on the claimant’s farm were damaged by a bushfire that started due to the defendant’s negligence. The claimant repaired several of these fixtures himself and several were repaired voluntarily by third parties, but he was nevertheless able to recover the reasonable market cost of repairing them.

In our view, the simplest explanation for this outcome is that the claimant was compensated for a loss that he did not actually suffer. However, a somewhat different explanation has been put forward in the cases (including in Powercor itself), ¹⁰⁵ which attempts to identify an actual loss for which the claimant is being compensated. When the property is damaged, it is said that the claimant immediately suffers a “direct loss” because the (market) value of the property is diminished. This diminution in value is measured in turn by the reasonable cost of repairing it. Although the claimant appears to recover the cost of repairing the property, what they actually recover is the diminution in value of the property at the time of damage.

This explanation does not accurately reflect the law. ¹⁰⁶ The cost of repairing property and its diminution in value (or replacement cost) are not always equal and sometimes one amount or the other is awarded depending on the circumstances. ¹⁰⁷ If the diminution in value is greater than the reasonable cost of repairs, then the claimant is generally only entitled to the cost of repairs. Conversely, if the diminution in value is less than the repair costs, then the claimant can only recover the diminution in value, ¹⁰⁸ unless incurring the repair costs is reasonable. ¹⁰⁹ It is also conceivable that the claimant might be entitled to the repair costs even if the property has increased in value. Suppose, for example, that a famous celebrity were to sign her name on a fan’s expensive jacket without permission. The value of the jacket might actually increase as a result, but surely the fan is still entitled to the costs of repair if they so wish? In some cases, therefore, the claimant can only be understood as recovering damages for repair costs rather than for any diminution in value and therefore for a loss that they do not actually incur. ¹¹⁰

Descheemaeker also attempts to explain such awards consistently with the actual loss account, but in our view none of his alternative explanations succeed. He argues first that, when the property has not been repaired, damages might be awarded to reflect the possibility that the claimant might decide to incur the cost of repairs in the future, ¹¹¹ but in Powercor the property had already been repaired, so the costs would not be incurred. Second, Descheemaeker argues that, when the claimant’s insurer pays for the repairs, damages compensate the claimant for premiums previously paid. ¹¹² Again, this could not explain the outcome in Powercor, as there is no evidence that the property was insured. The payment of premiums is also not factually caused by the wrong and is not equal to the cost of repairs in any event, so it is not clear why the claimant should receive the cost of repairs on this explanation. ¹¹³ Descheemaeker’s third argument is that, when a third party repairs the claimant’s property for free or at a discounted rate, the claimant’s loss is the cost of performing their moral duty to “return the favour” to the repairer. ¹¹⁴ Again, while in Powercor some of the claimant’s property was repaired by third parties, he was also compensated for the property that he had repaired himself.

A more plausible explanation is that repairing the property is or involves some emotional or non-pecuniary loss to the claimant (e.g. because the claimant engages in hard work). ¹¹⁵ The question, though, is why this loss should automatically be quantified by the reasonable commercial cost of repairing the property. There are a variety of reasons why a commercial repairer would accept that amount in return for their services. It cannot be assumed that it is equal to the loss that they (or the claimant) would incur in carrying out those services, in the same way that a person’s level of income is (clearly) not necessarily proportional to the non-pecuniary loss that they suffer to earn that income. Perhaps the claimant even enjoyed repairing the property and this outweighed the costs of doing so, meaning that he did not suffer a loss at all. The better explanation of cases like Powercor, then, is that the claimant simply recovers repair costs that they do not actually incur. ¹¹⁶

This explanation is supported by the recent decision of the Court of Appeal in Endurance Corporate Capital Ltd. v Sartex Quilts & Textiles Ltd. ¹¹⁷ The claimant’s premises were severely damaged by fire. The defendant insurer had agreed to indemnify the claimant for loss or damage to the premises. The claimant was able to recover the cost of repairing the premises even though it had no intention of having them repaired. ¹¹⁸ It was enough that, immediately prior to the fire, the claimant had intended to continue using the premises. ¹¹⁹ The claimant cannot be said to have received damages for the diminution in value of the property, since the court expressly rejected this as the applicable measure of damages. ¹²⁰ The claimant therefore recovered damages for a loss (the cost of repairs) that it did not, and likely would not, actually suffer.

B. Gratuitous Care

Hunt v Severs is one example of a case in which the actual loss account was an integral part of the court’s justification for refusing compensation. ¹²¹ That case concerned (among other things) the compensation available to an injured claimant who requires nursing or domestic care, in circumstances where that care is likely to be provided gratuitously (for example, by a spouse or family member). Previously, in Donnelly v Joyce, ¹²² the Court of Appeal decided that the claimant is entitled to damages for the cost of care in such cases, even if they do not actually incur it and therefore do not suffer an actual loss to that extent. This is because the claimant suffers a loss in the form of their need for the care, the value of which is “the proper and reasonable cost of supplying” the need. ¹²³

Although the House of Lords in Hunt accepted that damages may be available to a claimant who is cared for gratuitously, ¹²⁴ it did not accept the reasoning in Donnelly. Lord Bridge, with whom the other members of the House agreed, suggested that “the starting point for any inquiry into the measure of damages which an injured plaintiff is entitled to recover is the recognition that damages in the tort of negligence are purely compensatory. He should recover from the tortfeasor no more and no less than he has lost”. ¹²⁵ Since the claimant suffers no actual financial loss, Lord Bridge suggested that when damages are recoverable, this is instead to compensate the carer, and the claimant therefore holds any damages on trust for the carer. ¹²⁶ As a result, Lord Bridge held that compensation would not be awarded when the defendant (in Hunt, the claimant’s husband) provides the gratuitous care, apparently because the claimant has to repay the damages award to the defendant anyway. ¹²⁷ The same conclusion applied to the defendant’s costs of visiting the claimant in hospital. ¹²⁸ The actual loss account was clearly an important part of this reasoning, in that its premise is that the claimant could not themselves be compensated because they had not suffered an actual loss.

Although we do not suggest that compensation should always be awarded where the claimant suffers no actual loss, we believe that it should (or at least could) have been awarded in Hunt, on the basis of either the benevolence or avoided loss rule. Australian law provides a useful contrast on this point. In Griffiths v Kerkemeyer, ¹²⁹ the High Court recognised that the claimant may recover compensation for the reasonable cost of nursing and domestic care, even if that care will be provided gratuitously and the claimant will therefore incur no actual (pecuniary) loss to that extent. The High Court has also consistently recognised that such compensation is not awarded for any loss suffered by the carer, but, rather, because of the claimant’s need for the services. As a result, the measure of damages in such cases is the reasonable cost of having the services provided, rather than the amount of income lost (if any) by the carer. ¹³⁰ The claimant also does not hold the damages on trust for the carer, ¹³¹ and damages are available even if the carer is the tortfeasor. ¹³² The same approach, again in contrast to Hunt, applies to the reasonable cost of the claimant having visitors while hospitalised. ¹³³ The position is somewhat unstable, having been described by a later High Court as “not only exceptional, but anomalous”. ¹³⁴ Although we do not agree that the claimant’s need is necessary or sufficient to justify compensation, ¹³⁵ compensation for more than the claimant’s actual loss is nevertheless reflected in other parts of the law of damages and, to that extent, it should not be regarded as anomalous at all. Whether or not Hunt was correctly decided, it clearly demonstrates that adherence to the actual loss account has the potential to affect outcomes in particular cases.

C. Wrongful Life

The actual loss account has also been pivotal in cases of so-called “wrongful life” – that is, cases in which a child claimant is born with severe congenital disabilities in circumstances whereby, had the claimant’s parents been properly advised by the defendant of the risk of those disabilities materialising, an abortion would probably have been performed. ¹³⁶ The courts have overwhelmingly rejected such claims for a range of reasons, many of which reflect underlying matters of principle (such as an apparent conflict with the duty owed to the mother) ¹³⁷ and policy concerns (such as the sanctity of life). ¹³⁸ However, the most pervasive argument against wrongful life claims is based on the actual loss account: since the claimant would not have existed but for the defendant’s wrong, the identification of the claimant’s actual loss requires a comparison between their current life with disabilities and their non-existence, which comparison is said to be “impossible”. ¹³⁹

There seem to be two different senses in which the comparison is thought to be impossible. ¹⁴⁰ One is that it is impossible to know what non-existence actually consists of (nothingness, an afterlife, or something else), which means that the court cannot know what the claimant’s position would have been but for the wrong. A second is that it is impossible to value the difference between non-existence and a life with disabilities. Even if we assume that non-existence involves nothingness, it is still impossible (so the argument goes) to evaluate whether nothingness or a life with disabilities is preferable, or to put a money value on the difference. ¹⁴¹

This “impossible comparison” reasoning relies on the actual loss account, although the issue is complicated somewhat by the different roles that the reasoning plays in the cases. Some judges claim that it is impossible to identify a loss to the claimant and that damages are therefore not available. Other judges, in contrast, claim that it is impossible to identify damage and that the defendant is therefore not liable in the first place, since damage is the gist of the action. ¹⁴²

These arguments can in turn reflect two divergent views about the scope of the compensatory principle. In some of the judgments, the compensatory principle is explicitly viewed as “a statement as to the measure of damages. It is not a statement about liability”. ¹⁴³ On this view, the compensatory principle is relevant only to the question of loss and not to the logically prior question of whether the claimant has suffered damage. This accords with our understanding of the principle, which is in turn based on our understanding of the basic conceptual distinction between damage and loss. ¹⁴⁴ In other judgments, however, the compensatory principle appears to be viewed as more than just a remedial principle and instead as an overarching principle of liability, which therefore also goes to the prior question of damage. ¹⁴⁵

If the compensatory principle only applies to the assessment of damages, it can provide no argument at all against the existence of damage. The impossibility of the comparison would be beside the point, since no comparison would be required. It would be possible simply to frame the claimant’s disability or suffering as the relevant damage. ¹⁴⁶ The only issue remaining then would be the measurement of damages. At this point the compensatory principle would be squarely engaged. The actual loss account would require the court to identify all the actual consequences of the wrong, good or bad and award compensation only to the extent that the bad consequences outweigh the good. According to the “impossible comparison” reasoning, since either the full range of consequences cannot be known or their net value cannot be determined, the requirements of the compensatory principle cannot be satisfied.

However, if we are correct that actual loss is not a necessary condition of compensation, then the impossibility of comparing the claimant’s pre- and post-tort state cannot be a sufficient reason to reject an award of compensation. If it is not necessary to show that the claimant has suffered any actual loss overall, then it is possible instead to compensate the claimant for particular consequences of the wrong (e.g. costs arising from their disability), even if the claimant is better off overall because of the wrong. The fact that the court cannot know or value the full range of consequences is therefore not conclusive.

Whether compensation for certain costs is in fact normatively appropriate is, of course, debatable and the answer must be informed by matters of principle relating to the scope of the relevant duty and, possibly, the various policy considerations identified in the authorities. There is certainly an argument that compensation can be justified, since the purpose of the doctor’s duty is usually to prevent (by informing the mother of) the claimant’s specific disability, but it is unnecessary to reach a firm conclusion on the matter here. We do not argue that this is the only possible argument, merely that the impossibility of showing that the claimant is worse off overall (if it is indeed impossible) is not necessarily a barrier to such an award. Abandoning the actual loss account therefore removes one of the major obstacles to compensation in wrongful life cases. This would also avoid the undesirable consequence of leaving a wrong without a remedy (or, at least, an empty remedy). As Mason P. observes: “[this] maxim … is not a licence to write a blank cheque, but it reflects the way that the law has often responded when faced with the need to decide whether to accept or reject a novel claim.” ¹⁴⁷

V. An Alternative Account?

If we are correct to conclude that the actual loss account is conceptually and theoretically flawed, and that these flaws can and do undermine the quality and transparency of legal reasoning, then the obvious question is whether (and if so how) the law can be remedied. As we see it, there are two possible solutions: either the actual loss account is retained, despite its flaws, or it is abandoned in favour of a different account. While our earlier arguments provide several reasons for abandoning the actual loss account, unless we can find a better alternative the only option is to retain it.

What, then, might an alternative account look like? Andrew Tettenborn offers one possibility. ¹⁴⁸ He helpfully identifies various problems with the actual loss account, but regards these as stemming from indeterminacy in the concept of “loss”. He therefore proposes to abandon the concept of loss entirely, in favour of an approach that involves valuing the relevant interest that has been set back. In our view, there are several difficulties with this account. First, the difference between a loss and a setback to an interest is not immediately apparent and, on some views, there is no such difference. Second, it is not obvious how one would go about applying this approach; what does it mean to value “a broken leg”, ¹⁴⁹ for example? Third, it is not clear how Tettenborn’s approach relates to or accommodates the existing exceptions to the actual loss account. More generally, in abandoning the concept of loss as the organising idea of compensatory damages, it involves quite a radical departure from the existing law, which it would be preferable to avoid, if possible.

A similar point can be made about those accounts based on ideas, such as “substitution” or “compensation for the wrong”, which reject loss as the basis of at least some compensatory awards, particularly in cases where no actual loss can be identified. This is arguably unnecessary; difference in value and cost of cure, which are put forward as possible measures of substitutive damages, can easily be framed as types (or measures) of loss. If the claimant’s car is damaged and they incur the expense of having it repaired, that expense is clearly a (pecuniary) loss. Alternatively, if the claimant incurs the expense of buying a replacement car and selling the damaged car at their market values, the difference between those two values is the measure of their loss. If the claimant simply retains the damaged car, the mere fact that their car is worse than it was before is also a loss, which can often be measured by its diminution in value. In cases where the claimant receives compensation for one of these losses without actually suffering it, they can therefore still be understood as being compensated for a loss, just not their actual loss. They are compensated for the loss for which it is normatively appropriate (or reasonable) for them to be compensated. As the avoided loss rule recognises, the mere fact that the claimant avoids a loss does not necessarily prevent the claimant from being compensated for it. The same idea applies here. The claimant has avoided a loss (perhaps simply by choosing not to incur it), but it does not follow that they are not entitled to compensation for that loss. Further, as mentioned in the introduction, proponents of substitutive damages nevertheless recognise that some compensatory awards can be explained in terms of the actual loss account. Our argument here denies that the actual loss account ever correctly describes the law.

A third possibility, which involves a relatively small alteration to the orthodox compensatory principle, is to reinterpret that principle in terms of a more refined conception of causation. ¹⁵⁰ This approach would retain the idea that the defendant is liable for the actual loss that is caused by the defendant’s wrong, but would substitute a different test of causation for the simple “but for” test that is implicit in orthodox statements of the compensatory principle. While this is an elegant solution, which might be capable of capturing some of the factors that courts rely on in limiting the claimant’s compensation below their actual loss, it is arguably not capable of capturing all of them; at least, there is likely to be substantial disagreement about whether it is so capable. ¹⁵¹ To give one example, it is not clear that causal principles provide a transparent explanation of the reasonableness restriction applied in cases such as Sharman v Evans, in which the court weighed the cost of different treatment options against their health benefits in assessing the amount of compensation that the claimant could receive. It is also not clear how a causal account could explain cases where the claimant is compensated for a loss that they have not actually suffered; in such cases, no loss is caused by the defendant’s wrong. In short, modifying the underlying causal component of the compensatory principle would only deal with part of the problem, as some division would still need to be made between the compensatory principle and its remaining exceptions, without any clear rationale for doing so.

In our view, if the goal is to provide a transparent explanation of the law that is descriptively accurate and of general application, then the compensatory principle must be organised around some broad normative criterion. Our overview of the exceptions to the actual loss account indicates that there is a wide variety of considerations that bear upon the issue of compensation, which could not be captured in any more specific principle. As a result, a sound alternative to the actual loss account would have to be framed along the following lines: a claimant will be compensated for the loss for which it is reasonable to require the defendant to compensate them (whether or not the claimant actually suffered that loss). We might refer to this as the “reasonable loss account”. ¹⁵² On this account, rules that the actual loss account views as exceptions to the compensatory principle are really expressions of that underlying principle.

The question nevertheless remains whether this reasonable loss (or any similar) account is preferable to the actual loss orthodoxy. As a matter of logic, both the actual and reasonable loss accounts can accurately describe the law (if the actual loss account is read alongside its various exceptions). ¹⁵³ We believe that the reasonable loss account is nevertheless preferable to the extent that it provides a unified account of compensatory damages. And, insofar as it squarely acknowledges the normative dimension to questions of compensation, it also provides a more honest account. Of course, the reasonable loss account – at least in the broad terms in which we have so far framed it – does not itself offer a complete explanation of compensation in that it does not dictate an outcome in specific cases. The term “reasonable” is notoriously weaselled and will inevitably mean different things to different people. At its broadest, “reasonableness” is nothing more than a placeholder for “relevant normative considerations”, whatever one might claim them to be. So construed, both interpersonal and non-interpersonal considerations might play a role in assessing reasonableness and some courts might even consider it “reasonable” to limit (or inflate) private law compensation for distributive or broader social policy reasons. ¹⁵⁴ If the reasonable loss account were understood in this way, then it would appear to leave the question of compensation entirely up in the air, without providing any guidance to judges or practitioners, and the same would apply to any alternative account framed in terms of broad normative criteria. Whatever the problems with the actual loss account, one might argue, it is nevertheless easier to apply and at least provides a concrete starting point.

In our view, however, the differences between the two accounts in this respect are more apparent than real. As we have demonstrated, the actual loss account already requires judges to apply normative considerations in every case in determining whether to apply or recognise an exception and to that extent the contours of ‘reasonableness’ are already well established. And, on the reasonable loss account, once it is established that a certain kind of loss is (or is not) compensable, a precedent is created for future courts to follow. Thus, the question of reasonableness would not need to be “re-litigated” afresh in every case. Heads of damage can be used, as they already are, as a convenient shorthand to identify the categories of compensable loss.

This approach would not work, one might respond, in cases falling outside the established categories of compensable (or non-compensable) loss, or when an established rule (or exception) does not apply. Certainly, the reasonable loss account would not provide a definite outcome, but the court’s decision would not be entirely unconstrained; it can rely, as did the Australian High Court in Amaca v Latz, on analogical reasoning or on other common law reasoning techniques. Even in genuinely novel cases, where the existing legal materials do not support one particular outcome, it is not clear that compensating the claimant’s actual loss is a better default rule. While it might be easier to apply, it is no more likely to lead to the morally correct outcome. As we argued earlier, the law seldom (if ever) awards claimants their (net overall) actual loss, so it clearly does not regard that as the default moral outcome. Amaca, for example, reflects a more careful or cautious approach to compensating novel forms of loss.

Another possible argument against adopting a reasonable loss account is that the principle of restitutio in integrum, as conventionally understood, is simply too critical to the underlying framework of the private law to be challenged. This concern is also not as significant as it might at first sight appear. If our earlier arguments are correct, then the most significant change would be to how we describe the law, not to its actual content. To be sure, a shift in approach might lead to the treatment of certain kinds of loss being reconsidered, ¹⁵⁵ but (again) if our earlier arguments are accepted, then the theoretical basis for the actual loss account is not as strong as previously thought and so the argument for adhering to outcomes that it has produced is weakened. We are not aware of any theoretical account of compensation that identifies any blanket moral reason to compensate claimants for their actual loss. If anything, the existing theories (corrective justice and the continuity thesis) support the reasonable loss account, as they recognise that compensation should not be measured solely by the claimant’s actual loss.

For these reasons, we think that adopting the reasonable loss account would provide no worse, and in many cases a better, approach to decision-making. As mentioned above, the reasonable loss account (when construed broadly) does not itself settle the question of whether and to what extent compensation will be reasonable, but it does more accurately capture the true nature of the question that courts must confront. This is not to say that the claimant’s actual loss would have no role on such an approach. In many cases, identifying that loss will still be a useful practical starting point and the claimant’s compensation will often be measured by reference to some actual loss that they have suffered. If the reason for imposing duties is that certain kinds of conduct lead to certain kinds of loss, it makes sense that, when those duties are breached, those losses will often occur and (as the continuity thesis suggests) be compensable. The critical point, however, is that identifying (or not identifying) an actual loss is never the end point.

The arguments that we have considered so far against formulating the compensatory principle in terms of “reasonable” as opposed to “actual” loss are predicated, in large part, on the understanding that there is no necessary limit to the range of considerations that might be relevant in assessing “reasonableness”. It is entirely possible, however, to conceive of the reasonable loss account solely in terms of interpersonal considerations, such as corrective justice. As we have explained, corrective justice involves correcting the wrong or injustice (rather than the actual consequences of the wrong), which requires attention to matters such as the reason or purpose for the relevant primary duty. ¹⁵⁶ If the reasonable loss account is construed solely (or at least primarily) in terms of corrective justice reasons, then many of the arguments against it fall away. Since the range of applicable reasons would be highly constrained, it would be unlikely to lead to great uncertainty. Nor would a reasonable loss account grounded in corrective justice risk unduly uprooting the existing law, since in many cases corrective or interpersonal considerations will already underpin that law (and would therefore simply be picked up by the reasonable loss account) and, to the extent that they do not, corrective justice theorists would no doubt happily argue that the existing law should be reformulated or uprooted. Indeed, it is notable (and perhaps surprising) that they have not mounted a more explicit attack on the actual loss orthodoxy at a doctrinal level, given its clear incompatibility with their account of private law.

VI. Conclusion

The idea that claimants are entitled to be put in the position that they would have been in (personally and subjectively), were it not for the defendant’s wrong, is so deeply entrenched in the law that it has more or less escaped academic and judicial scrutiny. However, we have argued that this “actual loss” account of compensation is doctrinally and theoretically flawed and that the amount of compensation that the claimant receives is determined primarily by other considerations. It is true that, whether by accident or design, the law is structured in such a way as to avoid the need, in the majority of cases, to engage overtly with these considerations, but they are nevertheless pervasive. In our view, the reasonable loss account is preferable because it frankly acknowledges these considerations. But regardless of how one might choose to frame an alternative account, acknowledging the inherent weakness of the actual loss account is, we believe, an essential first step towards facilitating an open and continuing dialogue regarding the true nature of the compensatory principle in private law.