AGGRAVATED, EXEMPLARY AND RESTITUTIONARY DAMAGES [1997] EWLC 247(5) (16 December 1997)

BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just £1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!

1. THE NEED FOR REFORM

5.1 The decision in Rookes v Barnard(1) was a compromise, being the furthest the House of Lords felt it could go within the confines of precedent in ridding the law of exemplary damages, which it regarded as anomalous. (2) The first two of Lord Devlins three categories are essentially historically-based and represent situations where exemplary damages had been awarded prior to Rookes v Barnard and where reclassification of the damages as compensatory aggravated damages was not thought possible. It is debatable whether Lord Devlin would have felt constrained from abolishing exemplary damages following the Practice Statement of 1966. (3) It is equally debatable whether his Lordship would have felt the need to formulate his second category had the notion of restitutionary damages been current in 1964. (4)

5.2 The interpretation given to Rookes v Barnard (5) by the Court of Appeal in AB v South West Water Services Ltd, (6) limiting exemplary damages to wrongs in respect of which they had been held to be available before the decision in Rookes v Barnard, (7) has meant that the availability of exemplary damages is now yet further dictated by what are arguably the accidents of precedent, rather than sound principle.

5.3 Although it is not inconceivable that the House of Lords could reformulate the law in a way that is more satisfactory, it is surely correct that the present state of the law "cries aloud ... for Parliamentary intervention". (8) The overwhelming majority of our consultees agreed that the current law is in an unsatisfactory state. One consultee spoke for many in stating that the "result of AB v South West Water Services Limited is intolerable in terms of justice, logic and certainty." (9)

5.4 We regard some reform of the present law to be essential in order to restore rationality. We have the opportunity to recommend reform, unconstrained, as the courts have been, by precedent. The very difficult question is what form the reform should take. In particular, should exemplary damages be abolished altogether?

2. OUR CONSULTATION EXERCISES

(1) Consultation Paper No 132

(a) The main arguments considered in the Consultation Paper

5.5 The following arguments, for and against exemplary damages, were considered in Consultation Paper No 132: (10)

(i) Against exemplary damages

· punishment is not a legitimate function of the law of civil wrongs and should take place only within the context of the criminal law;

· now that non-pecuniary harm is more freely compensatable exemplary damages are no longer necessary;

(ii) In favour of exemplary damages

· punishment, deterrence and the marking out of conduct for disapproval are legitimate functions of the law of civil wrongs;

· exemplary damages alert plaintiffs to a method for the effective private enforcement of important rights;

· in some situations, compensation is inadequate or artificial, or does not effectively remedy the infringement of certain important interests.

5.6 Our provisional view was that exemplary damages should be retained, but put on a principled basis. (11) This was supported by the majority of consultees, although a wide variety of different views were expressed.

(b) Some of the main arguments put forward by consultees

(i) Against exemplary damages

5.7 The main arguments against exemplary damages focused on two issues: the divide between the criminal and civil law and the appropriateness of other remedies.

5.8 As regards the divide between the criminal and the civil law, it was claimed that, if an act warranted punishment, this was a matter for the criminal law or some other regulatory system. It was argued that deficiencies in the regulatory systems should be dealt with directly, by the amendment of those systems; they should not be patched up through the civil law. A further concern was that a person could be acquitted of a criminal offence, yet still be subjected to punishment by an award of exemplary damages.

5.9 The arguments focusing on the greater appropriateness of other remedies pointed to a range of other remedies which could perform at least some of the functions which exemplary damages might be expected to perform. One of these arguments was that restitutionary damages should be available to deal with any case where a defendant acted wrongfully with a view to making a profit. A second was that compensatory damages are adequate to take account of a plaintiffs outraged feelings, insult and humiliation in the case of torts such as defamation, false imprisonment, malicious prosecution, assault and intimidation. A third and alternative argument was that a non-monetary remedy, such as a published declaration, could more appropriately serve to vindicate personality rights: there is no connection between the desire to vindicate such rights and the institution of punishment.

(ii) In favour of exemplary damages

5.10 The main arguments in favour of exemplary damages responded, to a significant extent, to the same concerns, namely the divide between the civil and criminal law and the appropriateness of exemplary damages.

5.11 Many consultees were not prepared to accept that any sharp distinction exists between the goals that may legitimately be pursued by the criminal law and the civil law: punishment, deterrence and the marking out of conduct for disapproval are legitimate functions of the civil law, as well as the criminal law.

5.12 Many consultees also considered that exemplary damages perform useful and important functions. On the one hand, some pointed to its value as essentially a supplementary device: viz, as a remedy for perceived deficiencies in the criminal law, the civil law and other regulatory systems. A deficiency in the criminal law was identified in relation to police cases: in such cases it is the civil law that bears the brunt of maintaining the rule of law; retribution and punishment must therefore, of necessity, have a part to play in this area also. Likewise some consultees pointed to public concern over the failure to bring criminals to trial, or to secure convictions. A deficiency in the civil law was identified in a different set of cases, in particular those in which powerful defendants are unaffected by the normal level of damages. Another was identified in the specific area of libel law: it was observed that media libels were often deliberately perpetrated for profit or hate, and self-regulation in this area had clearly failed. On the other hand, it was also thought that exemplary damages may have a distinctive role. In particular, individuals were thereby given an effective weapon with which they themselves can enforce their rights; in contrast, victims may have little or no control over public prosecutions.

(2) Supplementary Consultation Paper

5.13 After the responses to the Consultation Paper were analysed, we found that the range of responses was so varied that we were left unclear as to where the consensus of opinion lay regarding the future of exemplary damages. We therefore decided to issue a supplementary consultation paper, primarily to those who had already submitted responses.

5.14 The Supplementary Consultation Paper asked consultees to choose between three approaches to reform. These were as follows:

Option 1: the Expansionist Model

The availability of exemplary damages would be expanded so that they could be awarded for any tort or equitable wrong (but not for any breach of contract) that is committed with, or accompanied or followed by conduct which evinces, a deliberate and outrageous disregard of the plaintiffs rights.

Option 2: the Abolitionist Model

Exemplary damages would be abolished, but this reform would be accompanied by provisions designed to:

Option 3: the Hybrid Model

This would be the same as option 2, except that exemplary damages would continue to be available for torts which are committed with a deliberate and outrageous disregard of the plaintiffs rights by servants of the government in the purported exercise of powers entrusted to them by the state, and which are capable in addition of amounting to crimes.

5.15 There were 146 responses to the Supplementary Consultation Paper, of which 17 (11.6%) favoured none of the options. These were re-allocated to the option which most closely fitted their views. After re-allocation, the distribution of responses was:

It can be seen from this that, adding together the responses favouring options 1 and 3, 72% of consultees favoured the retention of exemplary damages.

3. THE CASE FOR AND AGAINST EXEMPLARY DAMAGES

(1) The central issue of principle

5.16 In articulating a principled answer to the question whether exemplary damages should be retained or abolished, we consider that one fundamental issue has to be resolved: do exemplary damages confuse the civil and criminal functions of the law? The Consultation Paper stated:

The range of views on the question of the availability of exemplary damages is at heart a product of radically different perceptions of the role of the law of civil wrongs, in particular tort law, and of its relationship to criminal proceedings. The opposing views are best summarised in the speeches of Lord Reid and Lord Wilberforce in Broome v Cassell ... It will almost certainly be impossible to achieve a consensus on the acceptability of exemplary damages in the absence of agreement as to which of these perceptions is correct. (12)

5.17 Lord Reid in Broome v Cassell (13) stated that he regarded exemplary damages as "highly anomalous" and continued:

It is confusing the function of the civil law, which is to compensate, with the function of the criminal law, which is to inflict deterrent and punitive penalties. Some objection has been taken to the use of the word fine to denote the amount by which punitive or exemplary damages exceed anything justly due to the plaintiffs. In my view the word fine is an entirely accurate description of the part of any award which goes beyond anything justly due to the plaintiff and is purely punitive.~(14)~

... that there is something inappropriate or illogical or anomalous ... in including a punitive element in civil damages, or, conversely, that the criminal law, rather than the civil law, is in these cases the better instrument for conveying social disapproval, or for redressing a wrong to the social fabric, or that damages in any case can be broken down into the two separate elements. As a matter of practice English law has not committed itself to any of these theories ...~(15)~

(a) The principled case for abolishing exemplary damages

5.20 The principled case for abolition is that, given the existence of the criminal law, the raison dêtre of which is punishment, it confuses and complicates matters to punish civil wrongdoers. Wherever punishment is warranted, it ought to be pursued through the criminal law. The practical consequence of abolishing exemplary damages would be that a sharper, cleaner distinction could be drawn between the civil law and the criminal law.

5.21 On this view, exemplary damages are in truth a form of fine and several distinctive features of civil, as opposed to criminal, punishment appear as deficiencies:

(b) The principled case for retaining exemplary damages

5.22 The principled case for retention begins with the proposition that civil punishment is a different type of punishment from criminal punishment; the conclusion drawn from this is that it is coherent to pursue the aims of punishment (retribution, deterrence, disapproval) through the civil law, in addition to the criminal law, and in a civil form which does not necessarily have to mimic the criminal form.

5.23 Two distinctive features of civil punishment are relied on. The first concerns the locus standi or entitlement to sue of complainants. Civil punishment is sought and enforced by individual victims of wrongdoing. In contrast, criminal punishment is sought by or on behalf of the state: even though an individual can bring a private prosecution, he or she will be regarded as acting on behalf of the state. The second concerns the stigma associated with criminal punishment. Criminal punishment carries a stigma that civil punishment does not: a crime is viewed by society as more serious, and one corollary of criminal punishment is a criminal record - with all the potential consequences for, for example, employment prospects, which that entails. Consequently, £10,000 exemplary damages for assault would be less drastic than a £10,000 fine and criminal record for the same assault.

5.24 It follows from the view that civil punishment is distinctive in these ways that the objections outlined in para 5.21 above fall away as necessary objections. It is always an open question, which has to be addressed in respect of each objection in turn, whether awards of exemplary damages should be governed by the same rules as exist in the criminal law.

5.25 After much deliberation we have concluded that the principled case for retaining exemplary damages is to be preferred to the principled case for abolition. In other words, we believe that civil punishment can be adequately distinguished from criminal punishment, and has an important and distinctive role to play. At a deeper level the different approaches to the central issue of principle seem to reflect differences in the precision with which one wishes to divide different branches and functions of the law. The argument of principle for abolishing exemplary damages seeks to draw a bright line between the civil law and criminal law. The argument of principle for retaining exemplary damages is content rather with a fuzzy line, with a range of punishments from civil punishment, through criminal fines, to imprisonment.

5.26 We should emphasise, however, that we have not found this central issue of principle easy to resolve and we regard the arguments as finely balanced. In the circumstances we think it most important that our preference for the retention of exemplary damages is supported by arguments of general policy, to which we now turn.

(2) General policy arguments

(a) Arguments of policy for retaining exemplary damages

5.27 We regard the following general policy arguments to be the central ones in favour of the retention of exemplary damages:

(b) Arguments of policy against retaining exemplary damages

5.28 We regard the following policy arguments to be the central ones against the retention of exemplary damages:

(3) Conclusion

5.29 Our view is that, in contrast to the central policy arguments for retaining exemplary damages, the central policy arguments against retaining exemplary damages are unfounded or surmountable:

5.30 We do not agree that the availability of exemplary damages significantly increases unfounded litigation. First, the high cost of litigation, coupled with the prospect of having to bear the costs of the opposing and successful side in any litigation, is likely in any case to be a significant deterrent to any plaintiffs who are considering whether to bring unfounded claims. Secondly, plainly bad cases - and a fortiori cases whose only motivation is to oppress a particular defendant - can be struck out by the civil courts (or be otherwise dealt with, for example, by civil liability for the tort of abuse of process, or liability for costs). Thirdly, a number of limiting principles or devices already apply, or could be introduced, so as to limit the size and frequency of awards, and thereby limit any incentive to bring unfounded claims. In particular, we consider (and will explain in more detail below) (21) that:

5.31 We recognise that the discretionary element in exemplary awards is substantial. However, legislation on exemplary damages would have the effect of clarifying the law, and this would be further enhanced by case law interpretation. In any event, an element of discretion is warranted in order to retain the flexibility necessary to achieve justice and to ensure that the award is tailored to the nature of the defendants conduct and its consequences, and so to the degree of retribution, deterrence and disapproval which an exemplary award must achieve.

5.32 The risk of excessive uncertainty in the assessment of exemplary damages can be minimised in several ways:

"Moderate awards will not be effective deterrents"

5.33 The force of this objection varies according to ones interpretation of the concept of moderation. Two different usages of the term can be found in the present law.

5.34 On the first interpretation an award of exemplary damages is moderate if it does not exceed the minimum necessary to achieve the purposes of such an award. The objection made above has little force if this interpretation is adopted: it does not entail that awards will not be effective deterrents, or that the effective pursuit of the aims of exemplary damages will be prejudiced in any other respect. Rather, the concept is intended simply to avoid awards which are excessive in the sense of being larger than is absolutely necessary in order effectively to achieve their aims.

5.35 On the second interpretation awards of exemplary damages are moderate if they are lower rather than higher; accordingly, judges should prefer lower awards, rather than higher awards (even if this may mean some loss in efficacy). The objection made above may have more force if this is the chosen interpretation. Nevertheless, we do not regard the objection to be a decisive one. We do not accept that the fact that an aim cannot be performed in an ideal or a perfect manner is a good reason without more for declining to pursue that aim at all. Even if awards do not reach a level such that they achieve the maximum degree of retribution, deterrence and disapproval, moderate awards may still substantially achieve those aims, and so be of a valuable sanction available to civil courts. Indeed, the secondary aim of disapproval - or of signifying societys refusal to tolerate outrageously wrongful behaviour - is an aim that could still be pursued, successfully, even if awards were low. And because many factors influence the punitive efficacy of awards, even comparably low awards can have a significantly punitive effect. In this respect the wealth of the defendant, which includes the presence or absence of insurance, will be especially important. (22)

5.36 We do not accept that if exemplary damages awards are predictable, this will serve unjustifiably to impair their efficacy. The underlying assumption seems to be that predictability enables defendants to engage in cost-benefit calculations, such that in at least certain circumstances they will be able to conclude that, because the benefits which are likely to accrue to them from specific wrongful conduct are likely to exceed the sums payable as damages for that conduct, it is worth them acting in a wrongful way. We challenge this. On the one hand, consistent and predictable awards are required for reasons of fairness to defendants and potential defendants. This is also recognised within the criminal law. On the other hand, we envisage that a restitutionary award should be considered by a judge ahead of an exemplary award and that the gain which the defendant derived or expected to derive from his or her wrongdoing should be a relevant factor in the assessment of exemplary damages. If it is clear that a defendant acted wrongfully after calculating that an award of exemplary damages would be less than the profit which he or she expected to flow from the wrong, he or she could be punished accordingly by an (unexpectedly) larger award.

5.37 For all of these reasons we consider that the main policy objections to exemplary damages are unfounded or surmountable, and that it is therefore hard to see any practical advantage in their abolition. The case against exemplary awards appears to us to be essentially theoretical, rather than practical - viz, concerned to establish the neatness of a sharp divide between the civil and the criminal law.

5.38 Our conclusion is that policy considerations support our preference in principle for the retention of exemplary damages. We have also been heavily influenced by the fact that a substantial majority of consultees concluded that exemplary damages should be retained. (23) Of the three options set out in our Supplementary Consultation Paper, we therefore reject option 2 (the Abolitionist Model) as a model for reform and recommend that:

(15) exemplary damages should be retained.

5.39 In the light of this recommendation, this is an appropriate point to consider whether exemplary damages should be re-named. In Broome v Cassell (24) Lord Hailsham said that he preferred the term exemplary damages over the alternatives because:

... [it] better expresses the policy of the law ... It is intended to teach the defendant and others that tort does not pay by demonstrating what consequences the law inflicts rather than simply to make the defendant suffer an extra penalty for what he has done ... (25)

Nevertheless, in the Consultation Paper we sought views as to whether exemplary damages should be re-named. (26) A suggested title was extra damages, but this was unpopular with most consultees. We still consider that a change of terminology would be clearer and more straightforward. Along with a number of consultees, (27) we prefer the pre-Broome v Cassell terminology of punitive damages and we do not accept Lord Hailshams view that this label deflects attention from the deterrence and disapproval aims of such damages. When one uses the term punishment in the criminal law, one does not thereby indicate that deterrence is not an important aim. (28) Accordingly, we recommend that:

(16) our draft Bill should reflect our preference for the term punitive damages rather than exemplary damages. (Draft Bill, clause 1(2))

4. OUR CENTRAL REFORM PROPOSALS

(1) Expansion combined with important restrictions

5.40 We take the view, as we did in the Consultation Paper, (29) that if exemplary (or punitive) damages should be retained, their availability must be placed on a principled footing. Of the two remaining options set out in the Supplementary Consultation Paper, option 3 (the Hybrid Model) was expressly formulated as a pragmatic solution that would restrict the general availability of exemplary damages while retaining them in those circumstances where they seem to have a particularly important role to play.

5.41 We reject option 3 because it lacks the coherence which ought to be a major aim of any reform of this area of the law. Adopting option 3 would mean that the law would be tied to an approach that focuses on a defendants status, as a servant of the government, rather than on the degree of culpability of his or her wrongful conduct. As a result, it would leave gaps in the legal protection offered to plaintiffs, without there being any convincing justification for the omission - for there appears to be no sound reason why outrageously wrongful conduct should not attract a punitive award even if it is not committed by a servant of the government. For example, no punitive damages could be awarded for deliberate discrimination or libel by a defendant that is not a servant of the government. For these reasons, and also because it found favour with substantially fewer consultees than did option 1 (the Expansionist Model), we reject option 3 (the Hybrid Model).

5.42 We therefore favour the Expansionist Model. Punitive damages should be available for any tort or equitable wrong which is committed with conduct which evinces a deliberate and outrageous disregard of the plaintiffs rights. Punitive damages should not, however, be available for breaches of contract. We believe that this model affords a principle of general application upon which to base the availability of punitive damages. Such expansion is consistent with the common law relating to exemplary or punitive damages in major Commonwealth jurisdictions, even after Rookes v Barnard. (30)

5.43 But whilst we seek to expand the range of situations in which exemplary damages can in principle be awarded, and thereby ensure that the law has a rational basis, we are also anxious to ensure that exemplary damages are treated by the judiciary as a last resort remedy, and that there is consistency, moderation, and proportionality, in the assessment of such damages. Accordingly, whilst we are expanding the availability of exemplary damages, we are also imposing important restrictions on their availability and quantum. We believe that expansion combined with important restrictions is a policy which can appeal to both supporters, and critics, of exemplary damages.

(2) Our central recommendations

(17) the judge, and not a jury, should determine whether punitive damages should be awarded, and if so, what their amount should be. (Draft Bill, clause 2)

(18) punitive damages may only be awarded where in committing a wrong, or in conduct subsequent to the wrong, the defendant deliberately and outrageously disregarded the plaintiffs rights; (Draft Bill, clause 3(6); for conduct see clause 15(3)); and the narrower categories test of Rookes v Barnard should be rejected. (Draft Bill, clause 3(9))

(19) the cause of action test of AB v South West Water Services Ltd should be abandoned; instead:

(a) punitive damages may be awarded for any tort or equitable wrong; (Draft Bill, clause 3(3))

in this context an equitable wrong comprises a breach of fiduciary duty, a breach of confidence, or procuring or assisting a breach of fiduciary duty; (Draft Bill, clause 15(4))

(b) punitive damages may be awarded for a civil wrong which arises under an Act (including a tort or an equitable wrong), but only if such an award would be consistent with the policy of that Act; (Draft Bill, clause 3(4) and 3(5))

however, punitive damages must not be awarded for breach of contract or under an undertaking in damages.

(20) punitive damages may be awarded in addition to any other remedy which the court may decide to award; (Draft Bill, clause 3(8)) but may only be awarded if the judge considers that the other remedies which are available to the court will be inadequate alone to punish the defendant for his conduct (the if, but only if test); (Draft Bill, clause 3(7))

for these purposes the court may regard deterring the defendant and others from similar conduct as an object of punishment. (Draft Bill, clause 3(10))

(21) in deciding whether to award punitive damages, the court must have regard to:

(a) the principle that punitive damages must not usually be awarded if, at any time before the decision falls to be made, the defendant has been convicted of an offence involving the conduct concerned; (Draft Bill, clause 4(1))

when applying this principle a court must ignore section 1C of the Powers of Criminal Courts Act 1973. (Draft Bill, clause 4(3))

(b) any other sanctions that have been imposed in relation to the conduct concerned. (Draft Bill, clause 4(2))

(22) in deciding the amount of punitive damages the judge must have regard to the principles that any award:

(a) must not exceed the minimum needed to punish the defendant for his conduct; (Draft Bill, clause 5(1)(a))

(b) must be proportionate to the gravity of the defendants wrongdoing; (Draft Bill, clause 5(1)(b))

for these purposes the court may regard deterring the defendant and others from similar conduct as an object of punishment. (Draft Bill, clause 5(3))

(23) in deciding the amount of punitive damages, the judge must consider, where applicable, the following matters:

(a) the state of mind of the defendant;

(c) the nature and extent of the harm to the plaintiff that the defendant caused or intended to cause by his conduct;

(d) the nature and extent of the benefit that the defendant derived or intended to derive from his conduct;

(e) any other matter which the judge in his or her discretion considers to be relevant (other than the means of the defendant). (Draft Bill, clause 5(2))

(3) Aspects of our central recommendations

5.45 We now proceed to explain the major elements of our central recommendations set out above.

5.46 We reject the existing, and overly restrictive, categories test, in favour of a single, general test which seeks to isolate especially culpable and punishment-worthy examples of wrongful conduct. We have selected the phrase deliberate and outrageous disregard of the plaintiffs rights as the clearest of the multitude of similar phrases which were used in England before Rookes v Barnard, (31) and which have continued to be used in Australia, Canada and the United States, (32) to describe when exemplary or punitive damages are available.

5.47 The minimum threshold is that the defendant has been subjectively reckless - to use criminal law terminology. The notion of outrage imports the element of judicial discretion that we believe is inevitable, and essential, in this area. Factors that will no doubt be relevant in deciding whether conduct is not merely reckless but outrageous will include whether the wrong was intentionally committed, the extent and type of the potential harm to the plaintiff, and the motives of the defendant.

5.48 The extent to which conduct subsequent to the wrong is relevant has perplexed us a great deal. Ultimately we are content that the need for the conduct to be relevant to a disregard of the plaintiffs rights is a sufficient controlling principle. It ensures that the conduct, even if subsequent, is causally linked to the wrong and is not wholly independent of it. The facts alleged in AB v South West Water Services Ltd, (33) are particularly in point. (34) There the defendants admitted liability for, inter alia, the torts of public nuisance, negligence and breach of statutory duty in supplying contaminated water to inhabitants of Camelford in Cornwall. But the initial commission of the wrongs would not in itself have satisfied the deliberate and outrageous disregard of the plaintiffs rights test. What would have brought the defendants within that test was the allegedly arrogant and high-handed way in which they had ignored the complaints made by their customers and the allegedly misleading comments they had made as to the safety of the water.

5.49 We propose that punitive damages be available for any tort, for (most) (35) equitable wrongs, and for civil wrongs which arise under statutes where such an award would be consistent with the policy of the statute in question. But they should not be available for breach of contract; nor should they be available pursuant to an undertaking in damages. This would entail a general rejection of the rationally indefensible position which the common law reached following AB v South West Water Services Ltd, (36)according to which specific causes of action are selected solely on the basis of the existence or absence of pre-1964 precedents for awards of exemplary damages. That position has found no support in other Commonwealth jurisdictions; those jurisdictions have, indeed, also tended to the view that exemplary or punitive damages ought to be available in respect of any civil wrong, with the one significant possible exception of breach of contract. (37)

5.50 A large number of consultees favoured the inclusion of the tort of negligence within the category of civil wrongs for which exemplary damages would be awardable. There was particular support for their availability in situations where the defendants conduct, though within the tort of negligence, goes beyond mere negligence, and is grossly negligent or even reckless.

5.51 We do not consider that mere or even grossly negligent conduct should give rise to an award of punitive damages. Such conduct is not so serious that our society does or indeed should generally seek to punish such a wrongdoer, rather than, in particular, demand that he or she make reparation for the loss so caused to the plaintiff. This intuition is confirmed by a comparison with the criminal law, in which offences can only very exceptionally be satisfied by mere negligent conduct. (38) Nevertheless, we recognise that the tort of negligence may well be committed with a degree of culpability significantly in excess of that of the merely or grossly negligent defendant.

5.52 These considerations can be accommodated by our test of deliberate and outrageous disregard of the plaintiffs rights. This captures the more culpable forms of conduct, but serves to exclude mere and even gross (non-advertent) negligence. The result is that it is wrong to say that we are advocating the awarding of punitive damages for the tort of negligence per se. Rather, we propose that they may only be awarded if the conduct which constitutes the tort of negligence (or relevant subsequent conduct) also satisfies the additional test of deliberate and outrageous disregard of the plaintiffs rights. We therefore anticipate that the recovery of punitive damages for the tort of negligence will be exceptional.

5.53 This position derives substantial support from the approaches adopted in Canada, Australia and New Zealand. Courts in all three of these jurisdictions have held that exemplary or punitive damages are available for unintentional torts (including the tort of negligence), (39) and yet also clearly consider that such awards will be rare. (40) Thus reported cases in which exemplary or punitive damages have been awarded or contemplated seem to involve rather more than simple negligence, (41) or conduct that is aggravated by the defendants high-handed behaviour. (42)

5.54 It could be argued that a reformed law of exemplary damages should be confined to torts and should not be extended so as to include equitable wrongs. (43) No English case has awarded exemplary damages for an equitable wrong, whereas such damages are available for many causes of action in tort. In contrast, authorities in major Commonwealth jurisdictions have awarded exemplary damages for equitable wrongs. (44)

5.55 But despite the absence of English authorities for awarding exemplary damages for an equitable wrong, we can ultimately see no reason of principle or practicality for excluding equitable wrongs from any rational statutory expansion of the law of exemplary damages. (45) We consider it unsatisfactory to perpetuate the historical divide between common law and equity, unless there is very good reason to do so. Professor Waddams argues,

... the availability of exemplary damages should not be determined by classification of the wrong as a common law tort or as a breach of an equitable obligation ... (46)

Indeed, we can see good reason for allowing punitive damages to be recovered against, for example, the dishonest trustee who acts in breach of his fiduciary duty or the person who dishonestly abuses anothers confidence. Thus if, as we propose, punitive damages are awardable in respect of the (common law) tort of deceit, it would be anomalous if analogously wrongful conduct could not also give rise to an award, just because the cause of action originated in equity. Moreover, deterrence is an aim that is not alien to courts of equity. For example, it is a clear aim of the commonplace equitable remedy of an account of profits awarded for breach of fiduciary duty or breach of confidence. (47) To the extent that such remedies already achieve the aims of a punitive damages award in full or in part, and intentionally or incidentally, this will be a legitimate reason for refusing to make an award under the last resort test, or for making a lower award than would otherwise be necessary. (48) Finally, it is not an argument against making punitive damages available by statute for equitable wrongs that damages for equitable wrongs are not otherwise straightforwardly available. (49)

5.56 By recommending that equitable wrongs should be included in our proposed legislation, it does of course become incumbent on us to clarify what we mean by that phrase. Professor Birks argues that a wrong means:

conduct ... whose effect in creating legal consequences is attributable to its being characterised as a breach of duty ... (50)

A practical indicator of whether the law characterises particular conduct as constituting such a breach of duty is that compensation must be an available remedial measure for the conduct in question if loss is caused to the plaintiff by that conduct. Applying this approach, the common law civil wrongs are torts and breach of contract, and the equitable civil wrongs are breach of fiduciary duty, breach of confidence, and intermeddling by dishonestly procuring or assisting a breach of fiduciary duty. (51) There are also a large number of civil wrongs which arise under statutory provisions. (52) One could also include as an equitable wrong proprietary estoppel; nevertheless, because we consider that breach of contract should not trigger punitive damages, (53) and because proprietary estoppel is closely linked to breach of contract in that the essence of the wrong is the failure to fulfil the promisees expectations, we propose that proprietary estoppel should not constitute an equitable wrong for the purposes of our draft Bill.

(iii)Why include civil wrongs arising under statutes, subject to fulfilment of a consistency test?

5.57 We do not think that punitive damages can be refused for a civil wrong, merely because it arises under an Act rather than at common law. Our starting-point is therefore that punitive damages should prima facie be available for any wrong which arises under an Act for which the victim of the wrong may recover compensation or damages. (54) But this proposition is subject to one important qualification. Punitive damages should only be available for such a wrong if an award of punitive damages would be consistent with the policy of the statute under which the wrong arises (the consistency test). (55) Given the importance of clarity about what wrongs may attract an award of punitive damages, it is proper that we explain these recommendations in some detail.

5.58 There is little discussion in either case law or academic works about how civil wrongs which arise under an Act should be characterised. In many, but by no means all cases, liability is characterised as liability for a tort. In Breach of Statutory Duty in Tort, (56) Professor Stanton offers a valuable three-fold classification of tort liabilities which arise under statutes: "statutory torts", (57) "the inferred tort of breach of statutory duty" (58) and "the express tort of breach of statutory duty". (59) But these concepts are inadequate, even in combination, for our purposes. First, they do not sufficiently cover the field. In at least one case, a statute classifies liability for a wrong arising under it as liability for an equitable wrong. (60) It is also far from clear that every example of a statutory civil liability would generally be viewed as a statutory tort, in the absence of express statutory classification, rather than as a sui generis statutory liability. And secondly, the concept of a statutory tort is not currently in regular use, in either the case law or academic works, so that to employ it in an Act might have unforeseen consequences and might encourage needless debate about what makes a statutory civil liability a tort liability. We therefore prefer the concept of a wrong which arises under an Act for which [the victim] can recover compensation or damages. We consider that this is broad enough to extend to all wrongs to which it is appropriate to apply the consistency test, and plain enough to minimise unmeritorious debate.

5.59 It should be clear from the previous paragraph that many wrongs which arise under an Act must or can also be regarded as torts - whether as statutory torts, or as examples of the inferred or express torts of breach of statutory duty. We have also noted that at least one wrong which arises under an Act is characterised by the Act itself as an equitable wrong. (61) This requires us to qualify our earlier recommendation that punitive damages must be available for any tort or equitable wrong (as defined). (62) Rather, where a wrong arising under an Act is also a tort or an equitable wrong, the consistency test should apply. (63)

5.60 The consistency test constitutes a vital limitation on the availability of punitive damages for wrongs arising under an Act. Parliament has created many civil wrongs by statute. Sometimes it has taken great care to specify in the statute what remedy or remedies should be available for the wrong. In a number of instances it is reasonable to infer that Parliament intended that those remedies should be the sole remedies for that wrong, and further, that to permit punitive damages to be awarded would conflict with the policy or policies which Parliament was seeking to advance by creating the wrong and prescribing particular remedies for it. Without purporting to offer an exhaustive account of situations of inconsistency, we note that conflict exists where the statute which creates the wrong limits the amount of compensation available for it, and, further, provides that there should be no liability for the particular acts in question other than that laid down in the statute. A number of statutes that implement international liability conventions fall into this category. (64) Conflict is also almost inevitable when a statute lays down a detailed and structured remedial regime, particularly one that is administered outside the ordinary court system (for example, by industrial tribunals). One of the best examples of this sort of conflict is the wrong of unfair dismissal. (65)

5.61 We should stress, however, that in our view, cases of inconsistency, of the sort we have identified above, are unlikely to be common. Many of the better known statutes under which statutory civil wrongs arise do not specify expressly what remedies are available, or may do so in only the most general terms. (66) Parliament is often content to provide that a wrong should be civilly actionable, or actionable as a tort (67) or an equitable wrong, (68) without stipulating the remedial implications of that proposition. (69) It is a reasonable inference that at least compensatory damages are available for the civil wrong so created. And we think that, in general, the availability of punitive damages would be consistent with the policy of such Acts.

5.62 We have given thought to the possibility of formulating an exhaustive statutory list of wrongs which arise under an Act for which punitive damages should be available (or, perhaps, should not be available). The list would be formulated on the basis of our view as to which wrong-defining statutory schemes are, or are not, inconsistent with the availability of punitive damages. But to deal with this, divorced from the particular facts, would be an exceedingly difficult task, and would inevitably leave gaps. We therefore think that the better solution is a general statutory provision which prevents an award of punitive damages from being made for a wrong which arises under an Act, where the court considers that such an award would be inconsistent with the policy of the Act. We have so recommended. (70)

5.63 We would like to make clear our views on the wrongs of unlawful discrimination on grounds of sex, race or disability, which arise under the Sex Discrimination Act 1975, the Race Relations Act 1976 and the Disability Discrimination Act 1995. Many consultees thought that there was a pressing need for punitive damages to be available for these wrongs. Yet unlawful discrimination provides a hard case for the consistency test. We therefore think it important to spell out why, in our view, punitive damages should indeed be available for such discrimination and would not fall foul of the consistency test. While we shall focus on sex discrimination, the same reasoning applies to race and disability discrimination.

5.64 So far as unlawful discrimination outside the employment field is concerned, section 66(1) of the Sex Discrimination Act 1975 provides that such a complaint "may be made the subject of civil proceedings in like manner as any other claim in tort". This appears to mean, inter alia, that the remedies which are typically available for torts are also available for this form of unlawful discrimination. Thus if punitive damages are available for torts, section 66(1) would seem to authorise an award thereof for the tort of unlawful sex discrimination, on the same basis as for other torts. Indeed, we have already seen that English courts did, at one time, award damages for unlawful discrimination under section 66 in exactly this way. (71) The reason why those decisions have since been undermined is not because an award of exemplary damages was held to be inconsistent with these provisions of the Sex Discrimination Act, but because of an independent common law rule (the cause of action test). (72)

5.65 The conclusion that punitive damages should be available is more difficult to justify, but, we think, still the correct one, in relation to complaints of unlawful discrimination in the employment field. The provisions which deal with the enforcement of this category of complaint (73) differ substantially from section 66, and prima facie militate against an award of punitive damages. Section 65 establishes a detailed and exhaustive (74) remedial regime; (75) complaints are adjudicated and enforced by industrial tribunals, and therefore fall outside the ordinary court system; (76) and the only pecuniary award available is described as "compensation". (77) But section 65(1)(b) provides that the sum should be the same as any damages which could be awarded for discrimination outside the employment field. If damages awarded for discrimination outside the employment field can include punitive damages (as we think), then despite the terminology of compensation, the pecuniary award under section 65(1)(b) should also be capable of including a sum by way of punitive damages. (78) Indeed, to refuse to award punitive damages under section 65, whilst awarding them under section 66, would create an unjustified and anomalous distinction between English laws protection of individuals from unlawful discrimination within and outside the employment field. Section 65(1)(b) provides no sound basis for that distinction. (79) And until English law ruled out exemplary damages for all cases of unlawful discrimination, English courts had awarded exemplary damages without distinction between claims arising within and outside the employment field. (80)

5.66 The question of how English law analyses a breach of European Community law which gives rise to an action for damages against an individual or the State (for breach of a directly effective provision of Community law, or under the principles of Member State liability) is a difficult one. It seems that the claim to damages will be treated as based on a tort, and in particular, the tort of breach of statutory duty, with the statutory duty in question arising by virtue of the European Communities Act 1972. (82)

5.67 Applying that analysis, our recommendations would mean that such breaches of Community law could trigger an award of punitive damages if the courts took the view that such an award would be consistent with the policy of the European Communities Act 1972. That is, the cause of action would be a tort, but it would also be a wrong arising under an Act (the European Communities Act 1972) and so the court would be required, by clause 3(4)(b) of our Bill, (83) to consider whether such an award would be consistent with the policy of that Act. (84)

5.68 It is therefore the consistency test which provides the primary means for ensuring that our Bill conforms with Community law, in relation to this category of wrong. The 1972 Act was intended to bring national law into line with Community law in the United Kingdom, (85) or to provide facilities for doing so. (86) In our view, it does not unduly strain the consistency test (87) to say that it cannot be consistent with that policy for punitive damages to be available under our Act for a wrong which arises under the 1972 Act, if such an award would be inconsistent with Community law.

5.69 We would not seek to provide a definitive answer here to the question of whether an award of punitive damages would, or would not, be consistent with Community law. The arguments seem finely balanced. On the one hand, the Divisional Court in R v Secretary of State for Transport, ex p Factortame Ltd (No 5) (88) was hostile to the notion of punitive damages being awarded for breach of Community law. It stressed that the United Kingdom is almost unique amongst Member States in recognising a civil remedy of punitive damages, and that, as a result, it would detract from attempts to achieve uniformity in the remedies available for wrongs across the Community, if English law awarded punitive damages. On the other hand, Community law requires national courts not to discriminate against claims that are founded on Community law as compared with claims founded on domestic law. (89) It may be argued that, in the absence of clear indications to the contrary in the 1972 Act, or in specific Community legislation, or in general principles of Community law, punitive damages should be available (provided the other criteria in our Bill are satisfied).

5.70 Distinct from these types of Community law wrong are wrongs which are expressly created by a national statute in circumstances where Community law requires a wrong to exist. Many of these statutes are enacted in order to implement Community law. (90) Others may pre-date the relevant Community provisions, but in fact be the means by which Community law is implemented (if at all). The effect of our proposals is that punitive damages will be available for such wrongs, provided that such an award is consistent with the policy of the Act. (91) Where such an Act pre-dates the relevant Community law, it will be more difficult to conclude that it would be inconsistent with the policy of that statute to award punitive damages even though to do so would infringe the relevant Community law. Nevertheless, even in this situation the courts will still be likely (and perhaps obliged) to construe and apply the particular Act in question, as well as the consistency test under our Act, so as to conform, so far as possible, with the requirements of Community law. We suggest that it would generally be open to them to find that the consistency test is not satisfied, and thus that punitive damages are unavailable under our Act, if to award them would breach Community law. (92)

(v) Why exclude breach of contract?

5.71 In the consultation paper, we provisionally recommended that there ought to be no reform of the present law whereby exemplary damages are not available for breach of contract. (93) A majority of consultees supported that provisional view, which we now confirm as a final recommendation.

5.72 A range of reasons cumulatively lead to that recommendation. First, exemplary damages have never been awarded for breach of contract. Second, contract primarily involves pecuniary, rather than non-pecuniary, losses; in contrast, the torts for which exemplary damages are most commonly awarded, and are likely to continue to be most commonly awarded, usually give rise to claims for non-pecuniary losses. Thirdly, the need for certainty is perceived to be greater in relation to contract than tort and, arguably, there is therefore less scope for the sort of discretion which the courts must have in determining the availability and quantum of exemplary damages. Fourthly, a contract is a private arrangement in which parties negotiate rights and duties, whereas the duties which obtain under the law of tort are imposed by law; it can accordingly be argued that the notion of state punishment is more readily applicable to the latter than to the former. Fifthly, the doctrine of efficient breach dictates that contracting parties should have available the option of breaking the contract and paying compensatory damages, if they are able to find a more remunerative use for the subject matter of the promise. To award exemplary damages would tend to discourage efficient breach.

5.73 A counter-argument to our approach is that the potential for concurrent liability (94) means that it would be anomalous and odd to allow punitive damages for a tort (or equitable wrong) which arises on the same facts as a breach of contract, while denying the availability of such damages in an action for the breach of contract. But the acceptance of concurrent liability does not seek to deny that the bases of the causes of action in contract and tort are different. And the recognition of concurrent liability can be presented as supporting, rather than undermining, our recommendation, in that there is now no impediment to a plaintiff claiming damages for both breach of contract and a tort (and an equitable wrong) and then electing to take judgment on the cause of action most favourable to him or her. The prospects of plaintiffs being denied punitive damages merely because they have incorrectly pleaded their case as one for breach of contract rather than tort (or, for example, breach of fiduciary duty) are therefore significantly reduced.

(vi) Why exclude damages under an undertaking in damages?

5.74 The device of an undertaking in damages is presently used in different contexts, where it is required to perform rather different purposes. (95) This makes it far from easy to resolve the question, should it be possible for a court to make an award of exemplary damages pursuant to an undertaking? Two views can be identified.

5.75 One view is that it is very surprising that exemplary damages have ever been thought to be awardable under an undertaking. (96) The purpose of an undertaking, on this view, is to ensure that if a court wrongly grants interlocutory relief, the financial or other detriment that is suffered by the defendant as a result of the issuing of the relief can be adequately compensated. If such compensation were unavailable, the awarding of interim relief would be severely impeded by concerns that unrepaired and unjustified harm might be caused to the defendant. (97) On this view, the undertaking enforced is typically one to indemnify the defendant, in the event of an interlocutory injunction subsequently being discharged, for the loss he or she has suffered as a result of being restrained from doing what he or she could otherwise have done. The claim to damages is really a claim to payment of an agreed sum, the measure of which is the defendants loss; the damages are not available for the breach of any duty in the undertaking, contractual or otherwise. By definition an indemnity will only extend to losses suffered by the indemnified; (98) the use of an undertaking for the purposes of punishment is, on this reasoning, contrary to principle. (99)

5.76 A different view is evident from two authorities to which we have already referred. (100) This is that an undertaking in damages is validly viewed as a means of responding to reprehensible behaviour arising in connection with the obtaining or execution of interlocutory injunctions or related orders - such as Anton Piller orders and Mareva injunctions. (101)

5.77 In our view the better analysis of the role of an undertaking is the first. This does not mean that we do not consider that the aims which are advanced by the second conception, particularly the protection of what Zuckerman has called procedural rights, (102) are not legitimate aims. (103) It simply means that a liability (if any) to exemplary or punitive damages in these situations (104) is better analysed as arising from a civil wrong - especially a tort. The forms of reprehensible conduct which were alleged to found claims to exemplary damages in the two modern cases in which such claims were made and contemplated (but not upheld on the facts) certainly appear to be analogous to torts. And if analysed carefully, they may, or else should, actually constitute independent torts. (105) Presenting such claims as claims under an undertaking is an invitation to loose analysis and tends to discourage the legitimate development of tort law. (106) Denying punitive damages here should not lead to any significant gaps in the law on punitive damages, precisely because a tort-based claim will be available or else could (and should, if necessary) be developed. (107) We have accordingly recommended (108) that punitive damages should not be awardable under an undertaking in damages.

5.78 Our central recommendations extend the existing scope of the law on punitive damages in two major respects. The first expansion involves the replacement of the categories test with a general test of deliberate and outrageous disregard of the plaintiffs rights. (109) The second involves an extension of the category of civil wrongs in respect of which punitive damages are awardable significantly beyond the present scope of the cause of action test. Punitive damages will be awardable in respect of any tort, (most) equitable wrongs, and civil wrongs which arise under statutes where such an award would be consistent with the policy of the statute in question.

5.79 Nevertheless we recognise the need to constrain this expansion, in line with some of the arguments advanced against punitive damages, and so as to ensure that proper concern is shown for, inter alia, the efficient administration of justice and the risk of unfairness to defendants.

5.80 The restrictions which we propose will take several different forms; we explain each in detail below:

· giving the task of deciding the availability and assessment of punitive damages to judges only, and not to juries

· making the availability of punitive damages conditional on the other remedies which the court awards being inadequate to punish and deter the defendant

· requiring the court to refuse to award punitive damages where the defendant has been convicted of a criminal offence involving the conduct for which punitive damages are claimed, unless there are exceptional reasons why an additional award of punitive damages is still necessary and appropriate

· requiring the court to take into account the fact that other sanctions may already have been imposed (for example, in disciplinary proceedings) in respect of the conduct for which punitive damages are claimed, which make an additional award of punitive damages unnecessary or otherwise inappropriate

· preserving a residual, safety-valve discretion to refuse to make a punitive damages award in exceptional cases, even where the tests of availability are otherwise satisfied

· statutory structuring of the courts assessment of awards, by means of overriding principles of moderation and proportionality, and by a non-exhaustive list of factors which are relevant to such assessments

5.81 We recommend that the availability and assessment of punitive damages should always be decided by the trial judge and never by a jury. (110) Where trial is otherwise by jury, and punitive damages have been pleaded, the jury will continue to determine liability (111) and to assess compensatory damages (112) and restitutionary damages. However, the judge would then decide whether punitive damages are available, (113) and would assess the quantum of those damages. We would envisage that the judge would direct the jury that, whilst liability and the amount of compensation (or restitution) are matters for them, the questions as to whether, exceptionally, punitive damages should be awarded, and their quantum, are matters for the judge alone to decide.

5.82 This recommendation will mean that, where trial is at present by judge and jury, and punitive damages are claimed, the jurys role will be reduced. (114) Nevertheless, we consider that this reallocation of responsibility is justified in principle, and essential if consistent, moderate and proportionate awards are to be a reality. Cases have demonstrated a disturbing arbitrariness and excess in the sums awarded as damages (including exemplary damages) to plaintiffs by juries. (115) As Lord Woolf MR said of awards for the torts of false imprisonment and malicious prosecution made against the police:

We have ... been referred to a number of cases in which juries have made awards, both cases which are under appeal and cases which are not and the variations in the range of figures which are covered is striking. The variations disclose no logical pattern ... (116)

Similarly, the Neill Committees Report on Practice and Procedure in Defamation, (117) which proposed the abolition of exemplary damages for the tort of defamation, was heavily influenced by the arbitrariness of the sanction, (118) in the hands of juries:

... the decision whether to award [exemplary] damages and, if so, what the size of the award should be is left to a lay jury with no guidance on quantum and inevitably no possibility of a decision in accordance with any kind of tariff. This at a time when, in sentencing policy generally, consistency and predictability are goals constantly striven for both by means of statutory intervention and by way of judicial sentencing conferences and seminars. (119)

5.83 It could be argued that the change which we propose is now unnecessary, because of recent common law developments dealing with jury damages awards. We have already discussed at length the recent line of cases which (i) permit and state more detailed guidance for juries on how to assess exemplary damages, and (ii) extend appellate court control of jury-assessed awards. (120) The argument may be made that these developments substantially reduce the risks of arbitrariness and excess which provide a primary justification for judicial, rather than jury, assessment.

5.84 That argument is, so far, contradicted by experiences after John v MGN Ltd (121) in libel cases. Even where a judge has specifically followed the recommendations of the Court of Appeal in that case, very substantial damages awards for libel have still been made. (122)

5.85 But even if, in the long-term, recent developments have the effect intended, we remain convinced that universal judicial assessment is the best way forward. The remedies offered by recent cases are palliatives, and not cures, for defects in the process of jury assessment. In particular, it remains a fundamental principle, unaffected by recent developments, that a jury gives no reasons for its decision. This has two unfortunate implications. The first is to inhibit ex post facto appellate control of jury awards in particular cases, even though the criteria for appellate court intervention are less strict now than was formerly the case, following Rantzen v MGN Ltd. (123) The second is that unreasoned awards are much less likely to be consistent, moderate and proportionate awards; yet all three qualities are essential if punitive damages are to be a legitimate legal remedy. Accordingly, even if it could be shown that jury awards had become consistent, moderate and proportionate, this outcome would be largely a matter of chance.

5.86 Unlike juries, judges are expected, and generally required, to give reasons for their decisions. Many of the limits which we impose on our expansion of the availability of punitive damages can be fully effective only on the assumption that the body with responsibility for deciding claims to punitive damages gives reasons. As a result, and because we take very seriously indeed the need to constrain our expansion, that decision-maker should be a judge. Previous decisions can only be truly useful to future courts and to future litigants, because of the diversity of circumstances relevant to awards of punitive damages, if they are reasoned decisions. A tariff is realistic only on the assumption that there are reasoned decisions out of which it can be constructed. The flexible concepts used in our tests of availability (for example, deliberate and outrageous disregard of the plaintiffs rights) can be given content only if decision-makers can and do explain what, in the case before them, made the defendants conduct outrageous. The last resort discretion can only function as it should if the court makes known what very exceptional reason led it to refuse to award punitive damages, even though all of the tests of availability were satisfied and a substantial award was prima facie merited. The discretion as to the amount of punitive damages can only be structured by the principles of moderation and proportionality, and by the statutory list of relevant factors, if the award not only takes into account those principles and factors, but is also justified by reference to them. And if the court takes into account a relevant factor not specified in the (non-exhaustive) statutory list, it must at the very least specify what that factor was.

5.87 We therefore consider that the arguments of principle for judicial determination of the availability and assessment of punitive awards are very strong. Nevertheless, there are two arguments that the functional split which we propose between jury and judge (with the jury continuing to decide on liability and the quantum of compensation and restitution) is an unworkable one.

5.88 The first argument is that it would be impossible in practice to prevent the jury from being influenced, when it assesses compensation, by facts which are relevant only to the availability and assessment of a punitive award, and which should not affect the level of any compensatory award. An appropriately worded direction to the jury, or an appropriate division in the arguments presented before the court, obviously cannot guarantee that this will not occur. Nevertheless, we believe that our last resort test already offers a remedy for this problem. It would be open to a court which is faced with a quasi-punitive, jury-assessed compensatory award, to conclude that this award alone was adequate to punish, deter and disapprove, or that a smaller award of punitive damages is necessary than would otherwise have been the case.

5.89 The second argument is that there are certain facts which the jury must have decided at the stage of determining whether a wrong has been committed, which are also relevant to the issues which the judge must decide, but which are not apparent from the jurys verdict, without more. (124) This argument would be that since the judge cannot know what the jury decided, he does not have the factual basis before him on the basis of which he can decide whether punitive damages are available, and if so, what the appropriate sum should be. (125) But a similar problem - of ascertaining the factual basis for sentencing from the jurys verdict on liability - arises in the context of the criminal law, and is not considered insurmountable. In that context it is entirely a matter for the discretion of the judge, where the jurys verdict is consistent with two or more possible factual bases, whether the jury is asked to indicate the factual basis on which they proceeded. (126) Where the jury does not resolve the doubt, (127) the judge must always proceed, when sentencing an offender, on a basis of fact that is consistent with the jurys verdict. However, if more than one view of the facts could have founded the jurys verdict on liability, the judge is entitled to reach his own conclusion as to which of those views is the proper one, in the light of the evidence he has heard. (128) Archbold states:

If the verdict of the jury can be explained only on one view of the facts, that view must be adopted as the basis of the sentence, but if more than one view of the facts would be consistent with the verdict, the sentencer may form his own view in the light of the evidence, and pass sentence on that basis ... (129)

We anticipate that the civil courts, faced with a similar dilemma, could follow this approach.

5.90 We have therefore concluded that the availability and assessment of punitive damages should always be determined by a judge, and not a jury.

5.91 On the basis that the judiciary is to assess punitive damages, we support, inter alia, two judicial (or non-statutory) techniques for maximising consistency in assessments of punitive damages awards. These are techniques which, to a significant extent, mirror those employed by the Court of Appeal in Thompson v MPC (130) in relation to assessments of both compensatory and exemplary damages for false imprisonment and malicious prosecution. The main difference is that, at present, those techniques have been thought necessary only in jury trials. We would support their application even if, as we suggest, only judges should determine the appropriate sum of punitive damages.

5.92 The first technique would involve judicial development of a tariff for punitive damages analogous to that for compensatory damages for personal injury and death. Once a tariff has emerged - as we anticipate that it is likely to, in time - the assessment of punitive damages might be no more unpredictable than the assessment of damages for, say, personal injury. The second technique would involve an attempt to structure judicial discretion through the promulgation of guideline judgments by the Court of Appeal. This is a technique that is used to secure greater consistency in the field of criminal sentencing. For example, the Court of Appeal might set out benchmark figures together with aggravating and mitigating factors. Such guidelines would serve to structure the courts discretion in a more sophisticated and flexible manner than could be achieved by any statutory test, and might cover matters such as:

It would, however, be for the courts to decide what form the tariff or guidelines might take.

5.93 One possible objection to the establishment of tariffs is that they produce an undue rigidity and lack of flexibility in assessments of punitive damages. It is quite true that the circumstances in which punitive damages will become awardable will be extremely varied, in particular because the courts attention is directed more to the nature of the defendants actions than to the plaintiffs injury. Nevertheless, we do not accept that tariffs will lead to undue rigidity. The need for consistency between awards does place a limit on the degree of indeterminacy which can be tolerated in the assessment of individual awards. A tariff, in the form we envisage, would be developed by the courts and would not be merely the equivalent of a set of fixed awards (which are certainly productive of undue rigidity). (131) Rather, it would take the form of a set of benchmark figures together with a range of aggravating and mitigating factors. In this way flexibility and sensitivity to the particular characteristics of an individual case will be substantially preserved.

5.94 We have considered, but reject, as we did in the Consultation Paper, (132) three other techniques for limiting the size of punitive damages awards. Those (legislative) techniques are: statutory maxima, fixed awards and multiples (of compensatory damages). No such constraints exist on the assessment of exemplary or punitive damages in any major Commonwealth jurisdiction, although an increasing number of states in the United States of America have resorted to one or more of those devices, in an effort to curtail the massive awards which have been made there.

5.95 As regards statutory maxima, the majority of consultees were against their introduction, although strong arguments were marshalled in favour. (133) We consider that the stronger arguments support the majority view. The first is that they are arguably unnecessary. Judges are less likely to make excessive awards than juries, and we anticipate that a tariff system will develop which would operate to narrow judicial discretion and to constrain the size of awards in a more flexible fashion than would be possible under a system of one or several statutory maximum sums. The principle of moderation should also offer some constraint on awards. The second argument is that any maxima are impractical. In particular, it would be extremely difficult to decide, in a non-arbitrary fashion, whether there should be one cap or several, and if several, whether they should be applied by reference to the type of wrongful behaviour or to the wrong that is in question. The third argument is that statutory maxima could lead to undesirable consequences. One is that the underlying purpose of making a punitive damages award would be frustrated in any case where the wrongdoer calculated that the profit which he or she would derive from the wrongdoing would exceed the statutory maximum sum. The aim of the award would be to punish the wrongdoer for such a calculation, yet statutory maxima facilitate just these sorts of calculation; they also prevent any possibility of punishment being made more severe, in a particular case, in order to frustrate the calculations. Another is that maxima would look bad in any case in which an award was made against the state: the state would appear to be seeking to limit its liability, which would tend to compromise the rationale for the availability of punitive awards in these cases. We therefore do not recommend that statutory maxima be imposed on punitive awards.

5.96 The arguments against the use of either fixed awards or multiples of compensatory damages are even stronger. At the level of principle, these are objectionable for two connected reasons. The first is that they lack flexibility - minimising or even eliminating the scope for judicial discretion. Yet such flexibility is a precondition of effective and fair awards. It is a precondition of effective awards because flexibility enables an award to be tailored to the precise nature of the defendants conduct, and so more closely to the extent of punishment, deterrence and disapproval which that conduct makes necessary. In contrast, fixed awards will almost inevitably either over- or under-punish. It is a precondition of a fair award because fixed awards might, in some or even many cases, infringe the principle of moderation. This is because a court would have to make an award of a certain sum, even if it exceeded the minimum necessary to punish, deter and disapprove. The second objection is that multiples penalise disproportionately harshly the wrongdoer who causes substantial loss; they also wrongly assume that there is a direct relationship of proportionality between the heinousness of the wrongdoing and the seriousness of the harm caused thereby, and that the loss caused is the only factor relevant to judgments of the heinousness of the wrongdoing. Finally, the choice and the use of fixed awards or multiples is essentially arbitrary. The choice is arbitrary because it is very difficult to decide, in any rational way, what should be the level of the fixed award, or what multiple or even multiples should be used. The use of fixed awards will become increasingly arbitrary, unless the fixed sums are constantly updated in order to take account of changing social factors and of inflation. We therefore do not recommend the adoption of fixed awards or multiples in the assessment of punitive damages.

5.97 For the avoidance of doubt, we would emphasise that our rejection of statutory fixed awards, maxima and multiples should not be taken to imply criticism of the very valuable formulation of guidance by the Court of Appeal in the recent case of Thompson v MPC. (134) This is for two main reasons. First, Thompson v MPC involves judicially-formulated guideline ceilings, rather than absolute statutory limits to awards; secondly, to the extent that multiples are used, they are merely to suggest a ceiling for exemplary damages - that is, a maximum, rather than the always-appropriate sum.

5.98 As the Thompson ceilings are only guidelines, if a case was so exceptional as clearly to require a punitive damages award in excess of the ceiling, on the basis that such appalling conduct had not been anticipated at the time when that ceiling was determined, the court could make that higher award. Moreover, because ceilings rather than fixed awards are used, then even if the court feels itself to be constrained by the ceiling, it remains able to reflect the varying culpability of the defendant through an award which approaches, or falls some way short of, the ceiling. Nor can it be objected to the use of ceilings that they in fact entail the crude technique of multiple damages which we have so firmly rejected. For the multiple which was suggested in Thompson - three times compensatory damages - did not entail that, thereafter, any exemplary award would have to equal three times compensatory damages. Rather, the multiple was being used to calculate a guideline ceiling. The court is never compelled to award that multiple, but should award a sum, which is usually less than that maximum, which is appropriate to the culpability of the wrongdoer. The ceiling sum would be awarded in the very worst cases; cases falling short thereof would merit lesser sums.

(ii) A last resort remedy (1): remedies available to the court

5.99 In our view, punitive damages must be a last resort remedy. This proposition has several implications. The first, which we consider in this section, is that a court should not award punitive damages unless it believes that the other remedies which are available to it are inadequate to punish the defendant for his conduct, and to deter him and others from similar conduct (what we call the if, but only if test). (135) In effect we adopt, but adapt, the if, but only if test which is currently used at common law in order to determine whether exemplary damages should be available. (165) All of the major Commonwealth jurisdictions appear to apply this test, (137) notwithstanding that they have otherwise refused to follow the decision in Rookes v Barnard, (138) in which it was first formulated.

5.100 The existing if, but only if test, as formulated in Rookes v Barnard, is based on the idea that even a compensatory award may have an incidental punitive effect. To the extent that this is so, the justification for an award of punitive damages is correspondingly reduced or even eliminated. But clearly other remedies which a court may award in respect of a wrong (other than punitive damages) may have similar incidental effects. A good example is an award of a restitutionary remedy. (139) If punitive damages are truly to be made a remedy of last resort, a court must be entitled to take into account the effect of any remedy which it awards, in judging whether an additional sum of punitive damages is necessary to punish and deter. (140)

5.101 It is appropriate at this point to emphasise that a minimum condition of the availability of an award of punitive damages is that the court must want to punish the defendant for his conduct. It should also be a sufficient condition, in the sense that the court need only want to punish the defendant (and need not want to do anything else). But we recognise that the court may also, in punishing the defendant, properly seek to deter the defendant and others from similar conduct. Our recommendations (141) clarifying that an object of punishment may be to deter the defendant and others from similar conduct are intended to deal with this issue.

5.102 The basic question for the judge will therefore always be, "Are the remedies which are available to me inadequate to punish and deter?". If the plaintiff has only established an entitlement to compensation, the judge should proceed to ask himself whether the compensation which he is minded to award will be inadequate to punish and deter. If the plaintiff has only established an entitlement to restitution, the judge should proceed to ask himself whether the restitution which he is minded to award will be inadequate to punish and deter. And if the plaintiff has established an entitlement to both compensation and restitution - and it is a controversial question whether there can ever be an entitlement to both, as we have discussed in Part III - the judge should proceed to ask himself whether the total sum which he is minded to award as compensation and restitution is inadequate to punish and deter.

(iii)A last resort remedy (2): the relevance of conviction in criminal proceedings

5.103 The principle that punitive damages must be a last resort remedy has other implications. A defendant may already have been convicted by a criminal court of an offence involving the conduct for which punitive damages are claimed and punishment may have been exacted from him or her. It would be unacceptable if a defendant could be punished twice over for the same conduct - once by the criminal law and once by the civil law through an award of punitive damages. The more difficult question is how double punishment can best be avoided, where punitive damages are claimed in a civil court. This raises three issues:

What concept of identity of conduct or wrongdoing do we use?

5.104 There is an obvious difficulty in formulating an adequate concept of conduct that is identical for the purposes of the double punishment concern. In particular, one cannot use same offence or same wrong. As the concept is not for use solely within the criminal law or the civil law, respectively, but rather across the boundary of the criminal and civil law, one cannot employ terms which are unique to, or have particular (different) meanings within, each sphere.

5.105 A better concept would refer instead to a common factual basis. This means that conduct is the same where the facts which are alleged in support of the claim to punitive damages are substantially the same as those on the basis of which the defendant was convicted of a criminal offence. The draft Bill uses the phrase "an offence involving the conduct concerned". We are confident that this will be construed and applied sensibly, and not restrictively. (142)

Are we concerned about conviction or particular punishments?

5.106 A criminal court is likely to have a large range of possible methods of dealing with a particular offender, following his or her conviction for an offence. They may include an absolute or conditional discharge, binding over, a community service order, a fine or a prison sentence. In our view, the discretionary or absolute bar to an award of punitive damages must apply whatever method of disposition has been selected by the criminal court. The important point should be that the defendant has been convicted for the conduct which is alleged to justify an award of punitive damages.

5.107 Some may argue that where a defendant has been absolutely discharged following conviction, there is no risk that he or she will be doubly punished if he or she is subsequently held liable to pay punitive damages for the same conduct. And, so the argument would run, if he or she will not be doubly punished, there should be no bar to an award of punitive damages. We disagree. As we shall see, the justification for a bar to a claim to punitive damages following a criminal conviction is not just that it avoids the risk of double punishment; it is also that a civil court should not generally be permitted to reopen the question, which has been answered by the criminal court, of what is an appropriate response to the offender/wrongdoers conduct. This is an argument for respecting the criminal courts choice of response, regardless of its nature.

5.108 We would point out that, in any case, this point is never likely to be problematic in practice. An absolute discharge will usually reflect the triviality of the offence, or the low culpability of the offender, or that he or she had good reason (not amounting to a legal defence) for behaving as he or she did. In those circumstances it will be difficult, if not impossible, for a plaintiff successfully to demonstrate that the same conduct showed a deliberate and outrageous disregard of his or her rights. And if he or she cannot do so, no award of punitive damages can be made.

5.109 One practical problem is that section 1C(1) of the Powers of Criminal Courts Act 1973 provides that, subject to specified exceptions, where an offender has been absolutely or conditionally discharged following conviction for an offence:

... [the] conviction ... shall be deemed not to be a conviction for any purpose ...

5.110 Without more, s1C(1) could mean that any principle in our draft Bill which is expressed to apply where the defendant has been convicted of a criminal offence would not apply if the defendant was absolutely or conditionally discharged. This would be unfortunate. We have therefore recommended (143) that section 1C of the Powers of Criminal Courts Act 1973 must be ignored by a court when it applies the principle stated in recommendation (21)(a), which we explain below, that punitive damages must not usually be awarded if, at any time before the decision falls to be made, the defendant has been convicted of an offence involving the conduct for which the punitive damages are claimed.

A discretionary or an absolute bar to punitive damages awards?

5.111 Should a court ever be entitled to assess whether any prior criminal punishment in respect of the same conduct is adequate to punish and deter, and if it considers that it is not, be entitled to make a punitive damages award in order to make up the shortfall? Although this approach purports to avoid the problem of double punishment, we have serious reservations about its practicability, as well as its justification as a matter of principle. For wherever a defendant has already been convicted and punishment (or no punishment) has been exacted by the criminal courts, there are a number of cogent reasons why this should be treated as a complete and automatic bar to any punitive award in respect of the same conduct. In other words, there are strong reasons why a civil court should not even be permitted to address the question of whether the criminal punishment is adequate. The most important of these reasons are:

5.112 Nevertheless, we have come to the conclusion that it would be inappropriate to advocate a complete and automatic bar; rather a court should have a discretion to refuse to consider or make an award of punitive damages, where a defendant has already been convicted by a criminal court. Hard cases could well arise, in which it might, exceptionally, be appropriate for a court to proceed to the assessment stage. For example, a newspaper may publish an article which is both defamatory and in contempt of court. It does so in the knowledge that its circulation increase is so large as to more than exceed the damages that it might have to pay. It cannot be right for the court to refuse to award punitive damages merely because there has been a fine for contempt of court. Yet this is what, on a straightforward interpretation and application, the same conduct concept seems to require.

5.113 Accordingly, whether a claim to punitive damages is unavailable because of prior criminal punishment in respect of the same conduct ought to be a matter of discretion for a civil court. But because of the very strong arguments for barring a punitive damages claim in such circumstances, we would hope that the courts would only exceptionally find that punitive damages could be awarded, in a case where the defendant had already been convicted in a criminal court for the same conduct. A clear statement of principle as to the relevance of criminal conviction is essential, but it is questionable whether existing case law provides such a statement. (144) We have therefore recommended that a statutory provision should state that, in deciding whether to award punitive damages, the court must have regard to the principle that they must not usually be awarded if, at any time before the decision falls to be made, the defendant has been convicted of an offence involving the conduct which is alleged as the basis of the claim to punitive damages. (145)

5.114 This recommendation would operate in two main categories of situation: that is, where,

5.115 We do not think it necessary to propose any specific statutory elaboration and structuring of civil courts powers in relation to staying proceedings and/or striking out of claims, in order to ensure that the policy which is embodied in our recommendations is consistently and effectively applied. Civil courts already have an inherent discretionary jurisdiction to stay proceedings and to strike out claims. We are confident that they will exercise that jurisdiction in a manner which respects the strong reasons which we identify as reasons for refusing, except in exceptional circumstances, to make a punitive award where a defendant has already been convicted by criminal court for conduct which is alleged in support of a claim for punitive damages. (146)

(iv) A last resort remedy (3): the relevance of other sanctions imposed for the conduct

5.116 The defendant may already have been subjected to some other sanction (that is, other than a criminal conviction), such as dismissal from his or her employment following disciplinary proceedings. If the remedy of punitive damages is truly to be a last resort, it ought not to be awarded where a sanction has already been imposed on the defendant for his or her conduct, and that sanction is adequate to punish him or her for it.

5.117 We have therefore recommended that, when deciding whether to award punitive damages, the court must have regard to any other sanctions which have been imposed in relation to the conduct concerned. (147) One would expect the court to decide whether the defendant has been sufficiently punished by those sanctions; if he or she has, then no award is appropriate.

(v) A safety-valve discretion

5.118 Both the if, but only if test and the test of deliberate and outrageous disregard of the plaintiffs rights offer the courts significant flexibility in deciding whether to award punitive damages. In particular, when considering whether there has been an outrageous disregard of the plaintiffs rights, the judge should take account of any factor which bears on the culpability, and so punishment-worthiness, of the defendants conduct - whether as a mitigating (148) or aggravating (149) factor. Even if such a disregard exists, however, it would be open to a court (under the if, but only if test) to refuse to make any punitive award, or a low award, on the ground that compensation and/or restitution and/or any other remedy which the court is minded to grant are adequate or broadly adequate.

5.119 Nothing in our proposed legislative framework compels a court to award punitive damages, even if those two threshold tests are satisfied. Nevertheless, where a defendant has acted in deliberate and outrageous disregard of the plaintiffs rights and the remedies which the court is minded to grant will not be adequate to punish and deter, and the provisions concerning criminal conviction and the relevance of other sanctions do not apply to bar an award, there should be a presumption that a punitive damages award is appropriate. For to find these hurdles to be surmounted entails a conclusion that the defendants conduct is sufficiently culpable to merit punishment, that other civil remedies which the court is minded to award will not be sufficient to achieve that end, and that no other sanction has been imposed which makes a punitive damages award unnecessary or otherwise inappropriate. Accordingly, whilst nothing requires a court to make an award in this situation, some exceptional circumstance would have to exist before a court could, we believe, legitimately refuse to make a punitive damages award in these circumstances. This ultimate discretion, which our proposals preserve, should properly be conceived as a residual, safety valve discretion.

5.120 We consider that assessments of punitive damages must be constrained by two overriding principles: the principles of moderation and proportionality.

5.121 The principle of moderation reflects the approach adopted by the Court of Appeal in John v MGN Ltd (150) where it held that an award of exemplary damages should never exceed the minimum "necessary to meet the public purpose" underlying such damages: namely, punishing the defendant for his or her outrageously wrongful conduct, deterring him or her and others from similar conduct in the future, and marking the disapproval of the court of such conduct. This constraint is required by fairness to defendants: it aims to restrict, to what is strictly justifiable by reference to the effective pursuit of the aims of punitive awards, any actual or threatened interference with their civil liberties due to such awards. (151)

5.122 The principle of proportionality is justified by the consideration that no absolute pecuniary value can be ascribed to the sum which is required to advance the aims of retribution, deterrence and disapproval. Because of this, it is essential, if there is to be consistency between punitive awards, for the particular sum which must be paid by a defendant to be proportional to the gravity of his wrongdoing. More heinous wrongdoing will thereby be punished more harshly, and less heinous wrongdoing, less harshly.

(vii)The non-exhaustive list of factors relevant to the discretionary assessment of awards

5.123 We recognise the need for flexibility in the assessment of punitive awards. This is needed as a matter of efficacy and as a matter of fairness to defendants. The reason is that flexibility enables awards to be tailored to the nature of the defendants conduct and its consequences, and so to the degree of retribution, deterrence and disapproval which a punitive award must achieve.

5.124 Flexibility should not, however, be purchased at the price of arbitrariness. We have therefore sought to structure the discretion to award punitive damages by the inclusion of a non-exhaustive list of factors which should be considered, where relevant, in assessing awards. This list should encourage judges to rationalise the size of awards, rather than leaving them to select figures in an unreasoned way; it should also aid consistency betweenawards, by encouraging them to articulate the particular aspects of cases which call for lower or higher awards.

"the state of mind of the defendant ..."

5.126 A defendants conduct has to attain a high degree of seriousness before an award of punitive damages is available to a court: he or she must show a deliberate and outrageous disregard for the plaintiffs rights. But clearly there may be substantial gradations in the culpability of a defendants state of mind, even within this category of serious conduct. Accordingly, this factor is intended to permit discrimination between, for example, cases where the defendant was more or less calculating in his or her behaviour.

5.127 This factor is intended to reflect the fact that our society accords different value to different rights possessed by individuals; we judge conduct which interferes with or disregards individuals rights more or less severely according to the value which our society attaches to those rights. For example, invasions of bodily integrity might be considered to be more serious than the invasion of a property right.

"the nature and extent of the harm to the plaintiff that the defendant caused or intended to cause by his wrongdoing ..."

"the nature and extent of the benefit which the defendant derived or intended to derive from his wrongdoing ..."

"any other matter which the court in its discretion considers to be relevant ..."

5.128 Examples of "any other matter" might be conduct of the plaintiff that mitigates the outrageousness of the defendants conduct, such as provocative conduct resulting in a wrongful arrest, or conduct of the defendant, such as an apology.

(2)"These authorities convince me ... that your Lordships could not, without a complete disregard of precedent, and indeed of statute, now arrive at a determination that refused altogether to recognise the exemplary principle ...":Rookes v Barnard [1964] AC 1129, 1225-1226, per Lord Devlin.

(4)But see the discussion of whether category 2 is susceptible to restitutionary analysis at paras 4.16-4.20 above.

(8)Riches v News Group Newspapers Ltd [1986] QB 256, 269C, per Stephenson LJ.

(10) Aggravated, Exemplary and Restitutionary Damages (1993) Consultation Paper No 132, paras 5.4-5.38.

(11) Aggravated, Exemplary and Restitutionary Damages (1993) Consultation Paper No 132, paras 6.8 and 8.7.

(12)Aggravated, Exemplary and Restitutionary Damages (1993) Consultation Paper No 132, para 5.28.

(16)See Lord Reids objections to the assessment by juries of exemplary awards in Broome v Cassell [1972] AC 1027, 1087C-F, quoted at para 4.57 above.

(17)See, eg, Broome v Cassell [1972] AC 1027, 1086B-C (per Lord Reid), 1126D, (per Lord Diplock); AB v South West Water Services Ltd [1993] QB 507, 527E-F, 529A; Thompson v MPC [1997] 3 WLR 403, 417H. See the discussion of the windfall argument at paras 4.73-4.76 above.

(21) See paras 5.81-5.98 below (judicial role), 5.99-5.117 below (last resort remedy), and 5.120-5.122 below (principles of moderation and proportionality).

(22) For example, if the defendant is not financially very well off, and is not insured against exemplary damages, the sum that is required to punish and deter may be low.

(23) See para 5.15 above. We regard the law in other jurisdictions as cancelling each other out on this question; see Part IV of Aggravated, Exemplary and Restitutionary Damages (1993) Consultation Paper No 132. On the one hand, civil law jurisdictions have managed without exemplary damages, at least overtly. On the other hand, in other common law jurisdictions, in particular Australia, New Zealand, Canada and the United States, exemplary damages have continued to flourish: especially instructive cases include Uren v John Fairfax & Sons Pty Ltd (1966) 117 CLR 118; Lamb v Cotogno (1987) 164 CLR 1; Vorvis v Insurance Corporation of British Columbia (1989) 58 DLR (4th) 193.

(26) Aggravated, Exemplary and Restitutionary Damages (1993) Consultation Paper No 132, paras 6.23 and 8.13.

(27) Cf R v Secretary of State for Transport, ex p Factortame Ltd (No 5), The Times 11 September 1997, in which the Divisional Court recently proposed that exemplary damages is a "misleading" phrase, and that the appropriate one is penal damages, on the basis that "[i]t is a means of using civil proceedings to punish and deter certain classes of wrongdoer".

(29) Aggravated, Exemplary and Restitutionary Damages (1993) Consultation Paper No 132, paras 6.8 and 8.7.

(30) See para 4.5 above, para 5.46 below (general test of availability), and paras 5.49, 5.53 and 5.54 below (wrongs for which available).

(31) For an excellent summary of the law pre-Rookes v Barnard, see Mayne & McGregor on Damages (12th ed, 1961) paras 207-208. See also Clerk & Lindsell on Torts (12th ed, 1961) ss 354-358; Salmond, Law of Torts (13th ed, 1961) pp 737-739; Street, Principles of the Law of Damages (1962) pp 28-34. Also of particular assistance is Lord Dennings judgment in the Court of Appeal in Broome v Cassell [1971] 2 QB 354.

(32) See para 4.5 above. Particularly helpful is the American Law Institutes Restatement of the Law of Tort (2d) (1979), section 908, which reads:

(34) Other cases raising the relevance of subsequent conduct include Asghar v Ahmed (1985) 17 HLR 25 and Lamb v Cotogno (1987) 164 CLR 1.

(35) Breach of fiduciary duty; breach of confidence; and procuring or assisting in a breach of fiduciary duty. See para 5.56 below, recommendation (19)(a) above, and draft Bill, clause 15(4).

(38) See, in particular, the offences of careless or inconsiderate driving (s 3, Road Traffic Act 1988) and causing death by careless or inconsiderate driving (s 3A, Road Traffic Act 1988).

(39)Australian authorities on exemplary damages for the tort of negligence are: Midalco Pty Ltd v Rabenalt (1988) Aust Torts Reps 80-208; Coloca v BP Australia Ltd (1992) Aust Torts Reps 81-153; Backwell v AAA (1996) Aust Torts Reps 81-387; Trend Management Ltd v Borg (1996) 40 NSWLR 500. But some state legislatures have excluded exemplary damages from actions for negligence in eg motor vehicle injury cases (eg Motor Accidents Act (NSW), s 81A). The central New Zealand authority is McClaren Transport v Somerville [1996] 3 NZLR 424. See S Todd et al, The Law of Torts in New Zealand (2nd ed, 1997) pp 92-93 and pp 1231-1232. The general position in Canada seems to be that exemplary damages will not be awarded for the tort of negligence, unless the defendants conduct could be said to amount to recklessness or high-handed conduct: S M Waddams, The Law of Damages (2nd ed, 1991) para 11.210. In New Zealand, exemplary damages have recently been held to extend to the tort of negligence: McLaren Transport v Somerville [1996] 3 NZLR 424 (HC).

(40) See, for example: Coloca v BP Australia Ltd (1992) Aust Torts Reps 81-153 ("unusual and rare") and McClaren Transport v Somerville [1996] 3 NZLR 424 ("rare and exceptional").

(41) See, for example, McClaren Transport v Somerville [1996] 3 NZLR 424 (HC).

(43)The Ontario Law Reform Commission, in its Report on Exemplary Damages (1991), was unable to agree about whether punitive damages should be available for equitable wrongs: pp 71-74 (majority); pp 74-75 (dissent by Commissioner Earl A Cherniak QC). See further n 48 below.

(44) In Canada, it appears to be well-established that exemplary or punitive damages may be awarded for equitable wrongs, such as breach of fiduciary duty. See, in particular, """"

(45) See the Ontario Law Reform Commission, Report on Exemplary Damages (1991) pp 71-74 (majority), which reached similar conclusions.

(48) The key argument put by the dissenting member of the Ontario Law Reform Commission was that there was no need to provide a remedy of punitive damages for equitable wrongs because, in particular, a wide range of equitable remedies already existed which could be used, if necessary, to achieve the same ends as the common law remedy of punitive damages; there was therefore no need to add to this armoury by extending the common law concept of punitive damages. But Earl A Cherniak QCs preferred solution was to leave the issue to be considered on a case-by-case basis, "whereby it can be determined if there is in fact a lacuna in the law such that there is a need to award punitive damages" (p 75). This case-by-case consideration of whether punitive damages are in fact required, or whether other remedies already achieve their aims, is precisely what our last resort test achieves.

(49)Lord Cairns Act of 1858 gave the Court of Chancery power to award damages in addition to or in substitution for an injunction or specific performance, although it appears that the court had a residual discretion to award damages prior to the Act, which it rarely exercised:

(51) See Royal Brunei Airlines v Tan [1995] 2 AC 378, noted by C Harpum (1995) 111 LQR 545.

(52) For example, infringement of copyright (see the Copyright, Designs and Patents Act 1988) infringement of patent (see the Patents Act 1977) and unlawful discrimination on grounds of sex, race or disability (see, respectively, the Sex Discrimination Act 1975, the Race Relations Act 1976, and the Disability Discrimination Act 1995). See, for our proposals regarding this category of civil wrong, paras 5.57-5.65 below.

(56) K Stanton, Breach of Statutory Duty in Tort (1986) pp 8-12. See also K Stanton, The Modern Law of Tort (1994) pp 41-45.

(57) These are statutes which "specifically create a detailed scheme of civil liability of a tortious character"; the law created is "generally regarded as falling within the mainstream of tort liability" and the rules so enacted are "often modelled closely on common law principles": K Stanton, Breach of Statutory Duty in Tort (1986) p 8. Professor Stanton gives as examples the Occupiers Liability Act 1957 and the Animals Act 1971.

(58) This refers to a "common law liability inferred by the courts in order to allow an individual to claim compensation for damages suffered as a result of another breaking the provisions of a statute which do not explicitly provide a remedy in tort": K Stanton, Breach of Statutory Duty in Tort (1986) pp 8-9. Although in such cases the courts purport to be discovering Parliaments intention (see, in particular, Lonhro Ltd v Shell Petroleum Co Ltd (No 2) [1982] AC 173; R v Deputy Governor of Parkhurst Prison, ex p Hague [1992] 1 AC 58; X (minors) v Bedfordshire County Council [1995] 2 AC 633), the general view of academics is that the search is for an intention which does not exist.

(59) This refers to statutes which expressly make civilly actionable breaches of particular statutory duties by one or other of a range of formulas, such as "[breach of the duty] shall be actionable" or "shall be actionable ... as a breach of statutory duty": K Stanton, Breach of Statutory Duty in Tort (1986) pp 9-12. The details of the civil remedy or remedies available in such cases are left to be filled out by the courts.

(64) See, for example, the Nuclear Installations Act 1965 (liability for nuclear occurrences), s 12(1), and the Merchant Shipping Act 1995, Ch III (liability for oil pollution), s 156.

(65) See the remedial scheme established by what is now Chapter II of the Employment Rights Act 1996. It is notable that that scheme already includes elements which can loosely be described as having a punitive (rather than primarily compensatory) purpose. But cf the remedies available for unlawful discrimination on grounds of sex, race or disability in the employment field, which we consider at paras 5.63-5.65 below.

(66) For example, the Occupiers Liability Act 1957; the Occupiers Liability Act 1984; the Animals Act 1971; the Copyright, Designs and Patents Act 1988; the Defective Premises Act 1972; the Consumer Protection Act 1987. See also numerous examples of the inferred and express torts of breach of statutory duty.

(67) Examples are: "may be made the subject of civil proceedings in like manner as any other claim in tort" (eg s 66(1) of the Sex Discrimination Act 1975) and "in an action for ... all such relief ... is available to the plaintiff as is available in respect of the infringement of any other property right" (eg s 96(2) of the Copyright, Designs and Patents Act 1988). See now the statutory wrong created by the Protection from Harassment Act 1997, which that statute expressly classifies as a tort.

(69) Examples of typical forms of words are "shall be liable" (for "damage" or "injury" or "loss") or "shall be actionable". In some other cases, such as the Occupiers Liability Acts 1957 and 1984, and the Defective Premises Act 1972, civil actionability, though undoubtedly intended, may be less clearly indicated. And in yet other cases, which are regarded as examples of the tort of breach of statutory duty, Parliament may have given no thought to whether a particular breach of statutory duty should be civilly actionable, and it is left to the courts to decide whether or not that should be so (though the courts themselves rationalise what they do as an attempt to discover an implied legislative intention).

(74) Sex Discrimination Act 1975, s 62(1) provides that "[e]xcept as provided by this Act, no proceedings, whether civil or criminal, shall lie against any person in respect of an act by reason that the act is unlawful by virtue of a provision of this Act".

(75) Sex Discrimination Act 1975, s 65(1) provides that the industrial tribunal shall award one or more of three remedies as it "considers just and equitable": a declaratory order (s 65(1)(a)), an order for compensation (s 65(1)(b)), and a recommendation (s 65(1)(c)).

(77) Until a recent statutory amendment, necessitated by the ECJs ruling in C-271/91 Marshall v Southampton and South West Hampshire Health Authority (No 2) [1994] QB 126 (ECJ), the sum payable as compensation was also subjected to a statutory limit.

(78) In this particular context "compensation" seems to be used in a non-technical sense, to refer to a pecuniary remedy received by the victim of the unlawful discrimination. Cf Stuart-Smith LJ in AB v South West Water Services Ltd [1993] QB 507, 522D-E.

(79) A fortiori now that claims to compensation for unlawful discrimination in the employment field are no longer subject to a statutory limit.

(81) We do not discuss in this paper the question of whether punitive damages might be available for breach of the European Convention on Human Rights, if this were to be incorporated into national law. The Government has recently announced its intention to incorporate the Convention. Under the present law it would seem that exemplary damages would be unavailable because of the cause of action test. If our recommendations were implemented, the availability of punitive damages would presumably turn on the consistency test.

(82) English judicial statements that an action for damages for breach of directly effective provisions of Community law constitutes, in English law, an action for breach of statutory duty, include those of Lord Diplock in

(83) Clause 3(5) has the effect that, if a tort is also a wrong arising under an Act, the courts must apply the consistency test in clause 3(4)(b) to the tort.

(85) Section 2(1) has the effect that directly effective principles of Community law are, without more, available to be applied and enforced in national courts; section 2(4) has the effect that those directly effective principles take precedence over conflicting rules of national law.

(86) Section 2(2) confers powers on Ministers to make subordinate legislation solely for the purpose of implementing the United Kingdoms Community law obligations.

(90) See, for example, the Consumer Protection Act 1987, which implements the Product Liability Directive, and, indeed, expressly states that it is so doing.

(92) Cf if the statute in question expressly provides for a punitive remedy. If courts considered that it would be contrary to Community law to award punitive damages, but they were unable to use failure to satisfy the consistency test as the reason for refusing to award them, the courts would need to refuse punitive damages under the safety-valve discretion which is preserved by clause 3 of our draft Bill, and discussed in paras 5.118-5.119 below. See also n 84, above.

(93) Aggravated, Exemplary and Restitutionary Damages (1993) Consultation Paper No 132, paras 6.21 and 8.11.

(94) In Henderson v Merrett Syndicates Ltd [1995] 2 AC 145 the House of Lords authoritatively accepted that there can be concurrent liability for breach of a contractual duty of care and the tort of negligence causing pure economic loss. Cf Tai Hing Cotton Mill Ltd v Kamsing Knitting Factory [1979] AC 91.

(95) The multiple purposes of undertakings in damages have been recognised in the helpful article by A Zuckerman, "The Undertaking in Damages - Substantive and Procedural Dimensions" [1994] CLJ 546. The distinction which we draw between the two conceptions of an undertaking in damages is close to, but not exactly the same as, the distinction which Zuckerman draws between the undertaking as (i) a protection for substantive rights of defendants (pp 548-555) and (ii) a protection for procedural rights of defendants (pp 555-566).

(96) For recent judicial doubts, see Berkeley Administration Inc v McClelland 18 February 1994 (unreported, CA), per Hobhouse LJ, cited in S Gee, Mareva Injunctions and Anton Piller Relief (3rd ed, 1995) p 132.

(97) See, for example, Hoffman-La Roche v Trade Secretary [1975] AC 295, 361A-C, per Lord Diplock. See also the general accounts of undertakings in damages in, in particular, Spry, Equitable Remedies (4th ed, 1990) pp 472-478 and Appendix A, and S Gee, Mareva Injunctions and Anton Piller Relief (3rd ed, 1995) ch 9.

(98) The usual form of an undertaking in damages is worded in terms which indicate that its sole purpose is to operate by way of an indemnity. This has also been noted by S Gee in Mareva Injunctions and Anton Piller Relief (3rd ed, 1995) p 132.

(99) It is theoretically possible that a court could require a plaintiff seeking interlocutory relief to undertake to pay what amount to punitive damages, should it subsequently appear that he or she has acted in a reprehensible manner, either in the course of obtaining the interlocutory relief, or in the course of its execution. In this case the agreed sum would be an agreed punitive sum. We note, however, that courts refuse to enforce penalty clauses payable by one party to a contract on breach.

(100) See Digital Corporation v Darkcrest Ltd [1984] Ch 512; Columbia Pictures Inc v Robinson [1987] 1 Ch 87, discussed at para 4.27 above.

(103) We note the recommendation of the Lord Chancellors Department in Anton Piller Orders, A Consultation Paper (1992), that a summary remedy should be made available to the courts, which includes the power to award a punitive sum if a plaintiff has behaved in a reprehensible manner in the course of obtaining or executing an Anton Piller order: para 3.13.

(105) These could include: torts committed on the bringing of an action out of malice or for some other ulterior purpose (especially abuse of process), or on the breach of the terms of an Anton Piller order in the course of executing it (trespass to land or property, or negligence).

(106) See the restrictive approach taken to the tort of abuse of process in Digital Equipment v Darkcrest Ltd [1984] Ch 512.

(107) An alternative would be a specially-created remedy which is directed at abuses associated with certain forms of interlocutory relief: see, in particular, the Lord Chancellors Department, Anton Piller Orders, A Consultation Paper (1992), referred to at para 5.77, n 103 above.

(109) It should be emphasised, however, that if, under the present law, Lord Devlins first category can include innocent wrongdoing, our expansionist model is to that limited extent more restrictive than the present law. See para 4.7 above.

(110) In Aggravated, Exemplary and Restitutionary Damages (1993) Consultation Paper No 132, para 6.33, we provisionally rejected the option of judges taking over the role of juries in assessing damages. But the majority of consultees disagreed with that provisional view. See also Damages for Personal Injury: Non-Pecuniary Loss (1995) Consultation Paper No 140, paras 4.82-4.104, especially para 4.84.

(111) Liability here refers to the issue of whether a relevant civil wrong has been committed.

(113) See, in particular, the test of deliberate and outrageous disregard of the plaintiffs rights, at paras 5.46-5.48 above.

(114) An incidental effect of our proposals may be that there will be fewer applications for jury trials in civil cases.

(115) One argument provided by the Ontario Law Reform Commission in favour of retaining the jurys role in assessing exemplary damages was that there was no evidence that Ontario juries had made arbitrary and excessive awards: Report on Exemplary Damages (1991) p 49.

(117) Supreme Court Procedure Committee, Report on Practice and Procedure in Defamation (July 1991).

(118) Supreme Court Procedure Committee, Report on Practice and Procedure in Defamation (July 1991) ch IV, para 11: "[a]t least in criminal proceedings plaintiffs would be ... subject to far less arbitrary sanctions".

(119) Supreme Court Procedure Committee, Report on Practice and Procedure in Defamation (July 1991) ch IV, para 8.

(120) The main cases are: Rantzen v MGN Ltd [1994] QB 670; John v MGN Ltd [1997] QB 586; and Thompson v MPC [1997] 3 WLR 403. They have been discussed at length in Part IV, paras 4.61-4.67 and 4.86-4.98 above.

(124) This argument was raised by leading libel silks during the consultation process preparatory and subsequent to Damages for Personal Injury: Non-Pecuniary Loss (1995) Consultation Paper No 140, in relation to proposals for a split between liability and quantum in actions for defamation. The problem arises where a defendant unsuccessfully pleads justification. Even if he or she fails with the plea, and is liable for defamation, facts may have been established which will nevertheless serve to reduce the damages which he or she must pay (some of the charges may have been true) See Pamplin v Express Newspapers Ltd [1988] 1 WLR 116 and Consultation Paper No 140, paras 4.98-4.100.

(125) Strong objections were raised to one solution to this problem, discussed in Damages for Personal Injury: Non-Pecuniary Loss (1995) Consultation Paper No 140, para 4.99, which was that a judge should be entitled to ask the jury questions framed so as to elicit the jurys findings of fact in relation to the libel.

(127) In particular, if the jury is not asked to indicate the basis for its verdict, or though asked, refuses to do so.

(128) A recent illustration of this approach is provided by the Court of Appeals decision in R v Cawthorne [1996] 2 Cr App R (S) 445. Where a defendant is charged with murder, and there is an alternative verdict of manslaughter open to the jury, there are often different bases on which they can have reached that alternative verdict (for example, unlawful killing without intent to kill or cause really serious injury; provocation; gross negligence). This was the case in Cawthorne. The jurys verdict of manslaughter was consistent with more than one view of the facts. The jury was asked to indicate the basis for its verdict, but it refused to do so. The judge sentenced the appellant on the facts as they appeared to him to be. Appealing against sentence, counsel for Mrs Cawthorne argued inter alia, that if a vital issue of fact had not been resolved, and could be resolved, it had to be resolved by the jury, not the judge; and that if the jury has not resolved the issue, the judge should proceed on the factual basis which is most favourable to the defendant. The Court of Appeal rejected these arguments. It was entirely for the judge to decide whether the jury should be asked to indicate the basis of its verdict. In many cases the judge would not wish to do so; indeed, there might be "grave dangers" in judges asking juries how they have reached particular verdicts: [1996] 2 Cr App R 445, 450. Where the jury had not resolved the issue of fact, the judge was entitled to sentence the accused on the basis of the facts as they appeared to him to be from the evidence he had heard: [1996] 2 Cr App R 445, 451.

(132) Aggravated, Exemplary and Restitutionary Damages (1993) Consultation Paper No 132, paras 6.34-6.35.

(133) One of these was that the setting of limits to punishment is assumed to be of value in the criminal sphere, and that there is therefore a strong prima facie case for capping exemplary damages.

(135) In this paragraph, and in the rest of this report, we use the phrase punish and deter as a shorthand for the aims of an award of punitive damages. Previous judicial formulations of the if, but only if test have referred, in addition, to disapproval of the defendants conduct. We agree that this is an important aim of punitive damages: see para 4.1 above. However, it proved excessively difficult, and unnecessary, to draft a statutory provision which could state the test in its wider form (referring to punishing the defendant, deterring him and others, and expressing disapproval of his conduct). If a court considers that an award of punitive damages is necessary to punish the defendant (or in addition) to deter him and others from similar conduct, it will, by that award, necessarily also be expressing disapproval of the defendants conduct.

(137)The test forms part of Canadian law: see Hill v Church of Scientology of Toronto (1995) 126 DLR (4th) 129, 186 (SCC). It is also "reasonably clear" that the test forms part of Australian law, after its endorsement in Backwell v AAA (1996) Aust Torts Reps 81-387 (Vic, CA) and Commonwealth v Murray (1988) Aust Torts Reps 80-207 (NSW, CA), and the dicta of French J in Musca v Astle Corporation Pty Ltd (1988) 80 ALR 251, 269 (FCA, GD): M Tilbury, "Exemplary Damages in Medical Negligence" (1996) 4 Tort L Rev 167, 169-70. See also H Luntz, Assessment of Damages for Personal Injury and Death (3rd ed, 1990) para 1.7.9. Although it will often be difficult to apply the test in New Zealand, because of the state compensation scheme, the test nevertheless forms part of the law in that jurisdiction: see eg Aquaculture Corporation v New Zealand Green Mussel Co Ltd [1990] 3 NZLR 299, 301-302, per Cooke P, and Auckland City Council v Blundell [1986] 1 NZLR 732, 738, per Cooke P. The Ontario Law Reform Commission considered that the test was "sound in principle", but decided not to apply it for practical reasons: Report on Exemplary Damages (1991) pp 53-54. The concern of the majority was that the test would not constitute a significant limit on the availability of exemplary damages, because of the availability of insurance against liability to pay compensation. If, in practice, most awards of compensation were paid by an insurer, "any punitive effect otherwise inherent in such a compensatory award would be rendered ineffective" (pp 53-54). We consider, however, that a last resort test which makes any remedy - and in particular the availability of restitutionary damages - relevant to the question is a punitive award required?, can have a substantial role in limiting the availability of exemplary damages.

(139) Other remedies might include an injunction, or delivery up for destruction (see Mergenthaler Linotype Co v Intertype Co Ltd (1926) 43 RPC 381).

(140) Notwithstanding the Ontario Law Reform Commission views on the relevance of compensatory damages to the availability and/or quantum of punitive damages (see para 5.99, n 137 above), it recognised that restitutionary remedies should be taken into account:

(141) See para 5.44, recommendations (20) and (22), and draft Bill, clauses 3(10) and 5(3) above.

(146) If the defendant is convicted, the principle which we propose requires that any punitive damages claim in respect of the same conduct must, in the absence of exceptional circumstances, be struck out. However, there is unlikely to be any objection to the plaintiff proceeding with his or her other claims (eg for compensatory damages). A stay on such civil proceedings ought normally to be removed. If, in contrast, the defendant is acquitted or the proceedings against him or her end before the conclusion of the criminal trial, then a plaintiff should normally be permitted to proceed with his or her civil claims (including a claim to punitive damages in respect of the conduct concerned).

(148) For example, that the defendant acted under some form of mistake, or was provoked by the plaintiff.

(149) For example, that the defendant knowingly acted wrongfully in the expectation that he or she would obtain a profit thereby.

(151) Thus, for example, where freedom of expression is at stake, courts should subject large jury-assessed awards of damages to more searching scrutiny (

The Law Commission

1. THE NEED FOR REFORM

2. OUR CONSULTATION EXERCISES

(1) Consultation Paper No 132

(a) The main arguments considered in the Consultation Paper

(i) Against exemplary damages

(ii) In favour of exemplary damages

(b) Some of the main arguments put forward by consultees

(i) Against exemplary damages

(ii) In favour of exemplary damages

(2) Supplementary Consultation Paper

Option 1: the Expansionist Model

Option 2: the Abolitionist Model

Option 3: the Hybrid Model

3. THE CASE FOR AND AGAINST EXEMPLARY DAMAGES

(1) The central issue of principle

(a) The principled case for abolishing exemplary damages

(b) The principled case for retaining exemplary damages

(c) Conclusion

(2) General policy arguments

(a) Arguments of policy for retaining exemplary damages

(b) Arguments of policy against retaining exemplary damages

(3) Conclusion

"Moderate awards will not be effective deterrents"

(15) exemplary damages should be retained.

4. OUR CENTRAL REFORM PROPOSALS

(1) Expansion combined with important restrictions

(2) Our central recommendations

(a) the state of mind of the defendant;

(3) Aspects of our central recommendations

(v) Why exclude breach of contract?

(vi) Why exclude damages under an undertaking in damages?

(ii) A last resort remedy (1): remedies available to the court

What concept of identity of conduct or wrongdoing do we use?

Are we concerned about conviction or particular punishments?

A discretionary or an absolute bar to punitive damages awards?

(v) A safety-valve discretion

"the state of mind of the defendant ..."

	[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	The Law Commission
You are here: BAILII >> Databases >> The Law Commission >> AGGRAVATED, EXEMPLARY AND RESTITUTIONARY DAMAGES [1997] EWLC 247(5) (16 December 1997) URL: http://www.bailii.org/ew/other/EWLC/1997/247(5).html Cite as: [1997] EWLC 247(5)