Sydney Law Review
Moral indignation is not a factor that is to be used to inflate the calculation of a compensatory award.
There has been a tendency during the past 30 years, in many common law jurisdictions, towards an increase in amounts of damages, both in contract and tort. Arguments for increasing awards have, for a variety of reasons, been vigorously and effectively promoted, whereas the counter-arguments have appeared weak and diffuse. The counter-arguments, therefore, deserve attention. ‘The more the better’ cannot be a principle of justice, rationality or of sound policy.
The expansion of damage awards has been assisted by the ideas that the defendant is a wrongdoer deserving of little sympathy; that wrongs should all ideally be deterred and so it is acceptable — desirable even — that damage awards should err on the side of excess; and that damages will in any event be paid by an anonymous insurance fund and impose a real burden on no one. The third idea is inconsistent with the others, and each of the three rests on erroneous assumptions. In many cases — probably in most cases — those liable to pay damages are not personally guilty of blameworthy conduct; it is not true that all conduct giving rise to what the law calls a wrong should ideally be deterred; and all awards, even if funded by insurance, have to be paid for. These points will be illustrated by considering several kinds of legal wrongs and several different kinds of loss.
It may seem an attractive simplification to establish a single category of legal wrongs, but there is danger in attempting to force all legal wrongs into a single category. Although breach of contract has been classified, for purposes of assessment of damages, as a legal wrong, it is not true that every breach of contract is, for all purposes, equivalent to a tort. In our search for a simple scheme of classification, we are in danger of losing sight of a dimension of contractual obligation that used to be considered elementary, namely, promise or agreement. Contracts often (it is not necessary for present purposes to say always) involve self-imposed and self-defined obligations. This is not an irrelevant or accidental feature of contractual obligation and it has important implications for the appropriate remedy. In contrast to most torts, there is often nothing inherently objectionable or anti-social in the conduct that constitutes breach of contract. Often (again it is not necessary to say always), the promisee’s interest is primarily in the economic value of performance. Further, it is in the interest of both contracting parties to predict and to limit the probable costs of breach because this affects the agreed price. Together, these considerations present strong reasons for restricting damages for breach of contract to reasonable compensation of the promisee’s expectation.
Since breach of contract has usually been classified, for purposes of damage assessment, as a legal wrong, it follows that the just measure of compensation for a particular wrong that constitutes a breach of contract must generally be, viewed strictly as a matter of compensation, the same as the just measure where the wrong constitutes a tort or an equitable wrong. In some cases of torts or equitable wrongs, there may be persuasive reasons for awarding damages that exceed compensation, but if such reasons are operative, they ought, for the sake of clarity of thought, to be separately identified and distinguished from the assessment of compensation.
Liability for breach of contract is not dependent on proof of fault, either at the stage of contract formation, or at the stage of non-performance. The effect of a binding contract is that the promisor gives a guarantee of performance, and liability generally follows on simple proof of non-performance. Sometimes, indeed, breach of contract is inevitable, for example, where a person is bound by two incompatible contractual obligations, a circumstance that may arise without fault. In such a case an early announcement of inability to perform, though amounting to a legal wrong that exposes the defendant to liability, is not only permissible but, where coupled with an expression of regret and an offer to pay due compensation, praiseworthy.
Even where performance is possible, breach of contract is not always deterred by the law and deterrence is not always desirable. Thus, specific enforcement or enforcement by injunction is not usually available; self-help is not usually permissible to force unwanted performance on another party; a threat to break a contract is not always treated as wrongful; inducing another to break a contract is not always wrongful; punitive damages are, in most jurisdictions, usually not available; and a contract-breaker is usually not required to account for profits derived from breach.
Sometimes arguments have been adduced in favour of a general right to specific performance of contracts, but no such right has developed in Anglo-American law. The principal cases where specific performance has been refused are contracts for personal services, non-payment of debts and long-term contracts where orders of specific performance might have unexpected and oppressive consequences. Other considerations have been that the rules governing mitigation would be undercut by a general right to specific performance; that a decree of specific performance would often have the effect of prolonging a dispute and creating new occasions for conflict between hostile parties; that where the cost of performance greatly exceeds the economic benefit of the performance, a decree of specific performance would be oppressive to the defendant and would give undue bargaining power to the plaintiff; and that rights of third parties would sometimes be affected.
These various rules, taken together, justify the proposition that breach of contract has not been treated by the law as conduct that ought always, or at all costs, to be deterred. Holmes went so far as to say that there was never a legal duty to perform a contract, but only to perform or to pay damages at the promisor’s option, and his published correspondence with Pollock shows that he adhered to this opinion throughout his life despite cogent arguments to the contrary adduced by Pollock. A modern version of Holmes’ view has appeared in the economic doctrine of ‘efficient breach’, to the effect that breach of contract may be economically efficient, since the contract-breaker is made better off by it and the other party no worse off on receipt of full compensation. Some supporters of these theories have tended to overstate their case, but nevertheless the ideas underlying them have, as a matter of legal history, played a significant role.
Putting it at its lowest, it is a justifiable observation that there are some circumstances in which breach of contract has been treated as legitimate. A simple example would be of a student who agrees to paint the outside of a house, but, in breach of contract, determines to attend law school instead. In such a case, if the owner of the house can obtain the services of a competent professional painter for the contract price or less, no substantial damages would be awarded. Nor would a court order specific performance, issue an injunction to forbid attendance at law school, order an accounting of profits derived from a legal education or require payment of punitive damages. These rules are interrelated, and tend to support each other. They may be summarised by saying that the house owner has no proprietary interest in the student’s services. Though to some extent circular, the reasoning is not empty of content. Few would say that the student commits any moral wrong in breaking the contract, at least if the breach is coupled with an offer to pay compensation for any higher price the owner might have to pay to engage a professional painter. The power to break the contract in these circumstances is part of the student’s legitimate freedom of action and it would be an undue restraint on that freedom to compel performance.
These considerations affect the amount of monetary damages. The contractbreaker is bound, normally, to pay the full loss caused by the breach, including the loss of the value of the bargain if the owner has made a good bargain, but there is a positive and very powerful reason for not allowing the owner to recover any greater sum: it would place an unjustifiable restraint on freedom of action. From an economic point of view, it may be said that it would often deter an efficient breach.
Anticipated awards of damages affect not only the freedom to break a contract, but the terms on which the contract is made in the first place. Where the promisor acts in the course of a business, the anticipated cost of liability will be reflected in the contract price. If the law changes so that the damages likely to be paid upon breach of contract rise, the price charged to customers will, if the defendant is to remain in business, also rise. Before 1972, it was confidently asserted that damages for mental distress were not available for breach of contract, but in that year a decision of the English Court of Appeal awarded damages for mental distress and for loss of anticipated enjoyment to a disappointed holiday-maker against a holiday tour supplier. As Professor Samuel Rea pointed out the effect of such a decision, though beneficial, of course, to the individual plaintiff, was probably not beneficial to holiday-makers as a class, because the effect was to compel all of them to purchase insurance against a risk that they would almost certainly have preferred not to insure against if given the choice of saving the implicit premium. Damages for mental distress and for disappointed expectation of enjoyment are notoriously difficult to quantify. Insurance against such losses, without monetary limit, with the amount payable to be established, in case of dispute, by self-serving evidence in the course of costly litigation, implies a high premium.
Some cases have attempted to confine awards of damages for mental distress to contracts for ‘peace of mind’. This category is difficult to define and there is no obvious reason why compensatory damages should be so restricted. But the concern of courts in several jurisdictions to limit this head of damages manifests a perception that introducing an open-ended and expanding head of damages into ordinary commercial contracts has heavy costs. In a recent Ontario case, for example, involving breach of a contract by a small building contractor to renovate a domestic heating system the trial judge awarded extensive damages, including $35 000 general damages to the homeowner, who suffered from bipolar disorder. On appeal, this part of the award was set aside. The contract price was only $11000 and the Court of Appeal was evidently concerned both at the disproportion between the price and the damages, and at the prospect of introducing open-ended damage awards into every commercial and consumer contract. The court said that ‘generally before damages for mental distress can be awarded for breach of contract, the contract must be one where peace of mind is what is being contracted for, such as a contract for a holiday … or for insurance’, adding that:
[t]here are persuasive reasons to confine within narrow limits the circumstances when damages will be awarded for the exacerbation of mental illness for breach of a consumer contract… An extension of the circumstances when such damages are awarded could cause businesspeople to be wary of dealing with persons with mental disabilities for fear of exposure to claims for damages much higher than the value of the contract.
A comparison with the amount of damages awarded for non-pecuniary loss in personal injury cases is, it may be suggested, appropriate in this context. Mental distress may itself be classified as a personal injury. If $300000 (the approximate current upper limit in Canada) is the appropriate award for non-pecuniary loss in the case of the gravest physical injuries, a proportionately smaller sum must be appropriate for mental distress caused by breach of contract. It is difficult to accept that the distress suffered by a disappointed holiday-maker usually warrants a larger award than the pain, suffering, and distress caused by a broken limb.
There are many instances of tort liability where the defendant is not at fault and where it is not in the public interest to force the defendant to cease altogether from the risk-creating activity. It is sufficient to mention, as instances in many (but not all) common law jurisdictions, vicarious liability, products liability, nuisance, dangerous and unusual use of land, liability for animals, defamation and cases where there is a defence of necessity. In these cases the defendant is required to compensate the plaintiff for loss caused by the defendant’s enterprise. The merits of strict liability are debatable, but they will not be debated here. It will be assumed, for present purposes, that strict liability in the classes of case mentioned can be justified in so far as it relates to compensation for loss. The point made here is that there are very strong positive reasons for not awarding damages that go beyond fair compensation of loss. The reasons may be summarised as freedom of action on the part of the defendant and a positive public interest in permitting the defendant’s enterprise to operate, provided that it pays its way by compensating those to whom it causes loss.
The principal argument for strict liability is not (as is sometimes suggested) that the defendant is liable because he or she is or could be insured against liability, but that the defendant is liable because it is just to treat the defendant as an insurer. Pollock treated the instances of strict liability as ‘duties of insuring safety’. Whether or not the defendant purchases liability insurance is irrelevant to the reasons for imposition of liability: it is the defendant who is the insurer. The just limit of an insurer’s liability is the amount of the indemnity and the extent of the indemnity in these cases is compensation for harm. A similar argument is applicable to many cases usually classified as negligence. A learner driver doing his or her incompetent best, or a driver who suffers an occasional lapse of concentration or judgment — that is to say, every driver — is not excused from liability if an accident occurs in consequence of the incompetence or the lapse. Liability is based in such cases not so much on personal blameworthiness, as on the idea that the driver gives a guarantee to other road-users of a certain minimum standard of conduct. The just limit of liability supported on this basis is compensation for harm.
In Bazley v Curry, the Supreme Court of Canada addressed important questions of vicarious liability and charitable immunity. The corporate defendant, The Children’s Foundation, was a non-profit organisation that ran a residential home for emotionally troubled children. The plaintiff, a child resident in the home, had been sexually abused by an employee of the defendant. The question was whether the Foundation, assuming that it had not been negligent, was liable to the plaintiff. The court held that liability was appropriate. The reasoning involved considerations of enterprise liability, loss spreading, and internalisation of costs – arguments often used in other jurisdictions in the context of strict liability for defective products:
The employer puts in the community an enterprise which carries with it certain risks. When those risks materialize and cause injury to a member of the public despite the employer’s reasonable efforts, it is fair that the person or organization that creates the enterprise and hence the risk should bear the loss. This accords with the notion that it is right and just that the person who creates a risk bear the loss when the risk ripens into harm. While the fairness of this proposition is capable of standing alone, it is buttressed by the fact that the employer is often in the best position to spread the losses through mechanisms like insurance and higher prices, thus minimizing the dislocative effect of the tort within society…. The second major policy consideration underlying vicarious liability is deterrence of future harm. Fixing the employer with responsibility for the employee’s wrongful act, even where the employer is not negligent, may have a deterrent effect. Employers are often in a position to reduce accidents and intentional wrongs by efficient organisation and supervision.
The court then went on to consider whether there should be any immunity for charitable, or non-profit, organisations. In this context the court accepted that the Foundation, like other charitable organisations, was engaged in excellent work that was extremely beneficial to the community:
They do work few others would, and they do it in a selfless, generous manner. In the case at bar, the Children’s Foundation took in the respondent when no one else seemed ready or able to do so and undertook the difficult task of providing him with the love and guidance that other children receive from their parents. That non-profit organizations do important work is beyond question. They are funded by the government and by donations from the public.
Nevertheless, the admirable nature of the Foundation’s work was insufficient reason to excuse it from paying damages:
It is difficult to conclude that the fact that the appellant does good work in the community without expectation of profit makes it unjust that it should be held vicariously responsible for the abuse of the respondent.
The reason was that to accept an immunity would cast the loss entirely on the shoulders of the injured plaintiff:
The suggestion that the victim must remain remedyless for the greater good smacks of crass and unsubstantiated utilitarianism…. If, in the final analysis, the choice is between which of two faultless parties should bear the loss — the party that created the risk that materialized in the wrongdoing or the victim of the wrongdoing — I do not hesitate in my answer. Neither alternative is attractive. But given that a choice must be made, it is fairer to place the loss on the party that introduced the risk and had the better opportunity to control it.
Both questions that the court addressed, liability for intentional wrongdoing of an employee, and charitable immunity, are debatable and have been decided differently in other common law jurisdictions. The purpose of the present paper is not to debate these questions. Let us assume that they were rightly decided, and that The Children’s Foundation was rightly required to pay compensation to the plaintiff. The point that I wish to make is that implicit in the reasons is a very strong case for not requiring the Foundation to pay more than fair compensation. The Foundation is an excellent institution doing admirable work. It has done nothing wrong. It is decidedly not in the public interest that it should cease, or restrict, its operations, for if it did so, other children might be exposed to more and even greater risks of harm that those that materialised in this case. The diversion to the plaintiff of any of the Foundation’s money beyond the strict requirements of compensation is not only unjust to the Foundation, but also (in the particular circumstances) decidedly against the public interest because it will withdraw funds from a vital public purpose and from the assistance of other needy children. Thus the Foundation is rightly required (let us assume) to make fair compensation to the plaintiff, but there is a very strong positive case for not requiring it to pay any more. Punitive damages against the Foundation, of course, would be completely inappropriate and so too would any measure of damages that stretched the principles of compensation or that mixed punitive with compensatory considerations for the purpose of enhancing the award. There is a case for fair compensation and there is an equally strong case for not going beyond fair compensation.
The argument here is not that the defendant’s interest should prevail because it is a public interest, but that the defendant’s interest should prevail after the plaintiff is compensated because compensation is all that is due to the plaintiff. The public nature of the activities merely supplies a vivid illustration of the point.
As this case shows, Anglo-American tort law has had more than one dimension. Wrongdoing is one, but allocation of risk is another. Where a question arises of whether or not the imposition of liability is appropriate, it inhibits the court’s choice if the question is presented in all-or-nothing terms, as though the court, if it gives compensation, must also impliedly declare that the defendant’s activity is itself wrongful and should ideally be suppressed. There are many instances where the appropriate solution, as in Bazley v Curry, is to permit — indeed, to encourage — the defendant’s activity to continue, but to require compensation to be made to those injured in the process. The point was made in the context of products liability by a civilian lawyer:
The Roman praetor allowed the farmer to use his horse to carry fruit and vegetables at the Forum Romanum. This method of transport increased the range of goods on offer to society and, for the farmer, increased the possibility for personal profit, but nevertheless created the risk of damage whenever the horse followed its unpredictable nature. The praetor did not, however, accept the farmer's excuse that as the horse had never before gone out of control, the damage caused was unforeseeable. To allow the activity, but to allocate the risk, this ‘yes, but...’ approach is socially the best solution….
Failure to recognise this intermediate possibility forces an all-or-nothing choice that may inhibit development of the law to the detriment both of injured persons and of the public interest. To argue against potential liability for injury caused by rocks falling on a highway, for example, on the ground that ‘if the court imposed liability the government would have to close the highway’ might possibly be good rhetoric in some contexts, but overlooks the intermediate possibility of awarding compensation for injuries without implying that the activity of operating the highway is in any way objectionable. The argument is that fair compensation may be due to the injured plaintiff, but the corollary is that only fair compensation is due.
Many other examples might be adduced of instances where the apparent necessity of all-or-nothing choice has inhibited the development of the law. In a number of cases, the question has arisen of an employer’s liability for distress suffered by an employee dismissed in an unreasonably harsh manner. In Wallace v United Grain Growers Ltd, the majority of the Supreme Court of Canada held that employers were bound by obligations of good faith and fair dealing, but that breach of these obligations was to be reflected in a lengthened period of notice rather than as a separate head of damages. The dissenting judges would have based liability on an implied obligation in the employment contract. The approach of the minority is in many ways simpler, more logical and more convenient in practice, but the majority evidently feared that it would open the door to unlimited awards of damages in routine cases of wrongful dismissal. The solution, beneficial both to employers and employees, and to the rational development of the law, lies, it is suggested, in adopting the minority approach, but with damages strictly limited to fair compensation.
To speak of ‘fair’ compensation is not to resolve the question of how this is to be measured in respect of non-pecuniary losses. In 1979, the Supreme Court of Canada established a conventional figure of $100 000 (later held to be adjustable for inflation, and now, therefore, about $300 000) as a ‘rough upper limit’ for nonpecuniary loss in personal injury cases. Dickson J was plainly anxious to find a justification for moderation in the amount of awards:
The sheer fact is that there is no objective yardstick for translating non-pecuniary losses, such as pain and suffering and loss of amenities, into monetary terms. This area is open to widely extravagant claims. It is in this area that awards in the United States have soared to dramatically high levels in recent years.
The court did not use the word ‘cap’, and that word is misleading because it implies an artificial limit on some otherwise appropriate higher figure to which the plaintiff has a natural entitlement. The whole tenor and gist of the 1979 decision was that there was no such appropriate higher figure. The court was not confiscating something to which the plaintiff was naturally entitled; it was establishing a fair figure for the compensation of the most seriously injured. If the amount is not to be zero, and is not to be infinite, and if like cases are to be decided alike, some sort of conventional limit for the most serious cases is inevitable. In establishing a figure, the court did not depart from its proper judicial role, as some critics have implied. It performed a function very proper to a court of last resort in setting a figure that would be fair and just both to plaintiffs and defendants, and in the public interest. If this conclusion is correct, it cannot be right to subvert it by the creation of new open-ended heads of damages that blur the distinction between pecuniary and non-pecuniary loss, or by allowing (ostensibly as economic loss) large items of notional expenditure which it is known have not, or probably will not, actually be incurred.
There has been some disagreement in Canadian cases on whether the limit for the most serious injuries (now, as mentioned, about $300 000) implies a corresponding proportional limit for less serious cases. In Boyd v Harris a jury awarded $225 000 in respect of injuries that, while serious, were by no means comparable with the most severe imaginable. The British Columbia Court of Appeal dismissed the defendant’s appeal. Smith JA said that the damages should be assessed without any implicit comparison with the upper limit ‘a trier of fact, be it judge or jury, must assess non-pecuniary damages appropriate to the circumstances of the particular plaintiff, uninfluenced by the legal limit’. On the other hand, the same judge a few months later reduced a jury award of $197 000 after considering comparable cases. He commented that ‘the injuries in this case cannot by any stretch of the imagination be characterised as “catastrophic or near catastrophic,”’ and described the award as ‘a manifestly unreasonable verdict’, substituting an award of $115 000. Many other judges have also recognised an implicit scale for less serious injuries and this is, in my view, a necessary aspect of achieving consistency and just results as between plaintiff and plaintiff, as well as between plaintiff and defendant. The rejection of a scale for injuries less serious than the most severe springs from the implicit view, mentioned above, that the ‘rough upper limit’ established in the 1979 cases is a kind of artificial rule (or ‘cap’), that reduces high awards for some extraneous purpose without independent rational or principled justification and is therefore unrelated to any principles of justice applicable to lesser injuries.
In case of wrongful death, the question arises of compensation of survivors for non-pecuniary losses. The original Fatal Accidents Act was construed to exclude any compensation for grief, but by legislation in several Canadian provinces the court must now compensate certain relatives for loss of ‘guidance, care and companionship’. Judicial interpretation of this provision has varied markedly. One judge said that the provision:
cries out for the exercise of judicial restraint in the general interest of the public in the assessment of damages…. I say this because uncontrolled by such restraint the ceiling under the heading of loss of guidance, care and companionship for an award could be unlimited.
Some courts have held that a ‘modest’ sum is appropriate and others have stressed the need for ‘objectivity, predictability, and certainty’. On the other hand, several courts have said that there is no conventional limit and that the amount of the award will depend on the evidence in each particular case. This approach leads to enquiries into the emotional relationship between the deceased and the survivors, is likely to encourage self-serving evidence, to complicate
settlement negotiations and to lead to disparities in awards in apparently similar cases. Adapting Professor Rea’s comment quoted earlier, this is likely to be a costly form of life insurance. It is true that no amount of money can compensate for the loss of a life, but it does not follow that the appropriate award is either zero, or an infinitely large amount. The solution lies, it is suggested, in a conventional limit, and in one that bears a reasonable relationship to the $300 000 maximum for non-pecuniary loss in personal injury cases. In some jurisdictions a fixed sum is set by legislation, but where the legislature has left the amount open, the implication is that the award should be in an amount that is fair to both parties and it must be within the proper power of the court, as in the case of non-pecuniary loss in personal injury cases, to establish an upper limit. Another possibility, mentioned below, would be to go a step further and for the courts to establish a fixed invariable sum.
A different area of tort law where large awards of damages have, in some jurisdictions, been common is defamation. Liability for defamation is strict and this rule can be defended on the ground that even a morally innocent defendant should compensate an injured person for loss. But, as with the personal injury cases discussed earlier, the corollary is that compensation is the limit of the defendant’s liability, because there are strong interests in freedom of action on the part of the defendant — in the case of defamation the interest is in freedom of speech. Something of this principle may be discerned in the traditional rule that in cases of slander the plaintiff, with certain important exceptions, can only recover special damages.
It is not the prospect of paying moderate compensatory damages that creates ‘libel chill’ but the prospect of large and unpredictable awards by emotionally inspired juries. The law of defamation is habitually criticised on quite inconsistent grounds. A critic may denounce libel chill and yet the same critic, or like-minded critics, may be heard to say that the prospect of large jury awards has a salutary deterrent effect. Clarity of thought demands the separation of compensatory from other kinds of damages. Damages for defamation have often gone far beyond compensation of actual proven loss and have tended to mix punitive and deterrent with compensatory considerations. The traditional rules of defamation, including strict liability and the defendant’s burden of proving justification, may be supported, but then compensatory damages should, it is suggested, be restricted to loss, established, as in other cases, on the balance of probabilities. If punitive damages are to be awarded at all they should be separately identified and supported by persuasive independent justification.
Compensatory damages for defamation may include an award for mental distress, but it is inherent in what has been said earlier that this element of the award should be moderate and that it should bear some reasonable relation to damages for non-pecuniary loss in personal injury cases, a proposition accepted in some common law jurisdictions, but rejected in others. The making of a substantial (compensatory) award may itself go a long way towards vindicating the plaintiff’s reputation — in fact, it ought always to vindicate it in the minds of rightthinking persons. Of course not all persons are right-thinking, and compensation is due to a plaintiff for the distress of knowing that some of the mud has stuck. But this element of the award should bear a reasonable relation to other awards for mental distress and should strictly exclude elements of punishment and deterrence. Such a separation of heads of damages would go far to remove libel chill, because the defendant would know that large awards would not be made except on proof of conduct deserving of punishment. Compensatory awards (of moderate size, as suggested) might well be regarded by large media enterprises merely as a cost of doing business, but there is nothing necessarily objectionable in that, assuming, again, that the defendant’s conduct is not shown to be worthy of punishment. Indeed this approach offers benefits to both parties and to the public, analogous to other instances of strict liability. The innocent plaintiff is compensated (adequately but moderately) without the need to prove fault; his or her reputation is vindicated (by the very fact of the award or settlement), again without the need to prove fault; and the defendant is not deterred from publishing news and comment (unless, indeed, punishable misconduct is established).
Strong arguments may be adduced against the award of punitive damages in any circumstances. I do not wish to rehearse those arguments here, but I do wish to object, both in the context of defamation and in other contexts, to the introduction of punitive and deterrent considerations into the assessment of compensatory damages. This leads inevitably to confusion of thought: it undercuts rational principles of compensation, and, as shown in the examples discussed in this paper, it tends towards a constant increase in the amount of awards for which explicit justification is never demanded or required. If punitive damages are to be retained, there is everything to be said for separating them clearly from compensation, because the arguments that support compensation are quite different from those that support punishment. It has often been said by appellate courts that punitive damages are exceptional and that they are only to be awarded if compensatory damages fail to exercise a sufficient deterrent and punitive function. This approach requires an assessment of compensatory damages, entirely free of punitive considerations and then an enquiry into whether and to what extent additional punishment may be required. This process is impossible if punitive and compensatory considerations are intermixed.
A practical consideration that weighs in favour of restraint in damage awards is the danger of creating what is, or what is perceived to be, a ‘liability crisis’, or an ‘insurance crisis’, which may in turn evoke legislative responses that distort the law and inflict injustice on potential plaintiffs by arbitrary exclusions of liability, or arbitrary and unjust limits on compensation for actual loss. Instances, in various common law jurisdictions, have been numerous and varied. They include the restriction of liability of occupiers of land to trespassers, restrictions on liability for recreational activities, special tests for medical negligence, the exclusion of liability of rescuers, the departure from the principle of joint and several liability, unreasonably short limitation periods and limits on various heads of compensatory damages. Excessive damage awards have not been the only cause of such legislation, but they have played an important part. If compensatory damages are maintained at a level that can be demonstrated to be no higher than what is reasonable, ill-considered and potentially unjust political interventions are — to put the point at its lowest — less likely to occur.
Reference has been made, at several points, to conventional limits on the amount of awards. Somewhat different is the idea of a fixed conventional sum that is not a maximum, but an invariable amount. Such conventional awards have sometimes been provided by statute, as in the case, in some jurisdictions, of non-pecuniary loss caused by wrongful death. Judicially established conventional awards have been uncommon, but not unknown. Nominal damages may be one example and there have been some instances of awards of substantial sums described by courts as ‘nominal damages’, but which appear to be designed to compensate the plaintiff for a loss that is real, but difficult to calculate.
There is a case to be made in favour of conventional awards in the medical field, where a patient’s autonomy is wrongfully interfered with, but no economic loss is suffered that the law will recognise, as where a surgeon fails to give the patient complete information, but does not increase the risk of harm, or where a physician’s negligence results in the birth of a healthy but unwanted child. These cases have led to different results in different jurisdictions. The present purpose is not to debate the merits of these results, but to suggest that, if substantial compensation is denied (because of principles of causation, or public policy) there is a case for an award to compensate the plaintiff for loss of autonomy. Of course, assessment of such an award is inherently difficult, but it does not follow that it should be either zero or infinity. The majority of the House of Lords in Rees v Darlington Memorial Hospital NHS Trust, while refusing damages to a mother for the cost of bringing up an unwanted child, supported an award of £15,000. The basis for the award was somewhat unclear, as pointed out rather forcefully by the dissenting lords, Lord Steyn calling it ‘heterodox’, ‘contrary to principle’, and ‘a novel procedure’. But Lord Millett (one of the majority) put the award on the basis of compensation for ‘injury to the parents’ autonomy.’ Departing to this extent from a view he had expressed three years earlier, he said that the figure should not be simply a maximum, but should ‘be a purely conventional one which should not be susceptible of increase or decrease by reference to the circumstances of the particular case’. A conventional award, though certainly unusual in the past practice of the courts, has several merits: it recognises a real loss, and simultaneously gives a real measure of compensation for it, while also recognising the injustice to the defendant and the high price to the public (especially, but not exclusively, where the defendant performs a public service) of excessive, unpredictable and open-ended awards.
[*] Goodman/Schipper Professor of Law, Faculty of Law, University of Toronto.
 Cadbury Schweppes Inc v FBI Foods Ltd  1 SCR 142  (Binnie J).
 For example, contracts to sell the same property to two purchasers might be made by an agent, acting within ostensible authority. The principal, though without fault, would be contractually liable to both purchasers. See Hilton v Barker Booth & Eastwood  UKHL 8;  1 WLR 567 at para 35.
 Canada is an apparent exception, but the scope of the exception remains to be determined: Whiten v Pilot Insurance Co  1 SCR 595.
 Attorney General v Blake  UKHL 45;  1 AC 268 did require an account of profits, but the court emphasised the exceptional nature of the case.
 Frederick Lawson, Remedies of English Law (2nd ed, 1980) at 211; Alan Schwartz, ‘The Case for Specific Performance’ (1979) 89 Yale LJl 271.
 Stephen Waddams, ‘The Choice of Remedy for Breach of Contract’ in Jack Beatson & Daniel Friedmann (eds), Good Faith and Fault in Contract Law (1995).
 Oliver Holmes, The Common Law (1881) (reprinted in 1963) at 236; Oliver Holmes, ‘The Path of the Law’ (1998) 78 Boston Univ LR 699 at 702 (1897) (reprinted in 1998).
 These were that Holmes’ view was incompatible with the historical origins of assumpsit, with the availability of specific performance, with the tort of inducing breach of contract, with the doctrine of frustration and with the ordinary expectations of contracting parties, Mark Howe (ed), Pollock- Holmes Letters: Correspondence of Sir Frederick Pollock and Mr Justice Holmes 1874–1932 (2 vols, 1942) vol I at 79–80 , vol II at 201–2 .
 Richard Posner, Economic Analysis of Law (5th edn, 1998) at 131–4. This view was accepted by the Supreme Court of Canada in Bank of America Canada v Mutual Trust Co (2002)  2 SCR 601 at .
 See Stephen Waddams, ‘Breach of Contract and the Concept of Wrongdoing’ (2000) 12 Sup Ct LR 1.
 Jarvis v Swans Tours Ltd  EWCA Civ 8;  QB 233.
 Samuel Rea, ‘Non-pecuniary Loss and Breach of Contract’ (1982) 11 J Leg Stud 35.
 See Hayes v Dodd  2 All ER 815; Johnson v Gore Wood & Co  2 AC 1.
 See Ruxley Electronics & Constructions Ltd v Forsyth  1 AC 344; Farley v Skinner  UKHL 49;  2 AC 732.
 Turczinski v Dupont Heating & Air Conditioning Ltd (2004) 246 DLR (4th) 95 at , .
 Frederick Pollock, The Law of
Torts: A Treatise on the Principles of Obligations Arising from Civil Wrongs in the Common Law (1887) at 393.
 1999 Can LII 692 (SCC);  2 SCR 534 (hereafter Bazley v Curry).
 Id at , .
 Id at .
 Id at .
 Id at .
 See Lister v Hesley Hall Ltd  1 AC 215; New South Wales v Lepore  HCA 4; (2003) 212 CLR 511.
 Hence, reference to ‘distributive justice’ appears to introduce more difficulties than it resolves. See McFarlane v Tayside Health Board  UKHL 50;  2 AC 59 at 82, 83 (Lord Steyn); see Rees v Darlington Memorial Hospital NHS Trust  1 AC 309 at  (Lord Millett).
 Hans Taschner, ‘European Initiatives: The European Communities’ in Christopher Miller (ed), Comparative Product Liability (1986) at 8.
 1997 CanLII 332 (SCC);  3 SCR 701.
 Andrews v Grand & Toy Alberta Ltd (1978) 83 DLR (3d) 452 at 476.
 (2004) 237 DLR (4th) 193.
 Id at .
 White v Gait (2004) 244 DLR (4th) 347 at 355, 356 (Smith JA).
 For example, Hodgson v Walsh (1999) 121 OAC 255 (Unreported, Morden, Laskin & Rosenberg JJA, 2 June 1999).
 9 & 10 Vic c93 (Lord Campbell’s Act).
 Zik v High (1981) 35 OR (2d) 226 (Holland J).
 Lawrence v Good (1985) 18 DLR (4th) 734 at 741.
 Nightingale v Mazerall (1991) 87 DLR (4th) 158 at 162, 163.
 Mason v Peters (1982) 139 DLR (3d) 104 at 118, 119; Neilsen v Kaufman (1986) 26 DLR (4th) 21 (though suggesting the eventual possibility of a conventional award).
 Above n12.
 See Tolstoy Miloslavsky v United Kingdom  ECHR 25; (1995) 20 EHRR 442 (English law incompatible with freedom of expression).
 See John v MGN Ltd  QB 586; Carson v John Fairfax & Sons Pty Ltd (1993) 178 CLR 44.
 Hill v Church of Scientology of Toronto  2 SCR 1130.
 See Allan Beever, ‘The Structure of Aggravated and Exemplary Damages’ (2003) 23 Ox JLS 87.
 Cassell & Co Ltd v Broome  UKHL 3;  AC 1027; Whiten v Pilot Insurance Co  1 SCR 595, Performance Industries Ltd v Sylvan Lake Golf & Tennis Club Ltd  1 SCR 678.
 Statutory provisions in force in New South Wales are discussed by Barbara McDonald, ‘Legislative Intervention in the Law of Negligence: the Common Law in a Sea of Statutes,’ in this issue.
 Chappel v Hart (1998) 195 CLR 232; Chester v Afshar  UKHL 41;  1 AC 134.
 McFarlane v Tayside Health Board  UKHL 50;  2 AC 59; Rees v Darlington Memorial Hospital NHS Trust  1 AC 309.
 See Stephen Todd, ‘Wrongful Conception, Wrongful Birth, and Wrongful Life,’ in this issue.
 The same view is taken by Todd, ibid.
 Above n44.
 Id at , .
 Rees, above n44 at .
 McFarlane, above n44.
 Rees, above n44 .