Insurance and Illegal Conduct
In Chapter 2, we saw that it has proved tempting to see the relationship between the law of obligations and insurance in terms of a tension, or even a conflict; and that this is often related to the ‘actuarial’ model of insurance outlined there. We have tested this assumption throughout, focusing on issues of risk transfer, risk allocation, and loss distribution, within the law of obligations itself. Review of the principles relating to illegal conduct, the subject of the present chapter, offers an opportunity to explore a very different facet of the relationship between insurance and the law of obligations. Where, in the broader picture of the law of obligations (expanded to include the operation of insurance), will principles of personal responsibility and of fault be found to operate? In our view, the broadened picture illustrates once again the mutual influence of liability, and of insurance. But it also shows the persistent influence of moral judgments in relation to insurance rather than liability,1 while illustrating the complex mix of goals and influences which play upon the liabilities themselves.
An important preliminary issue is identification of the various categories of conduct in play. Throughout this chapter, we use the ostensibly simple term ‘illegality’ interchangeably with the potentially obscure expression, ‘ex turpi causa non oritur actio’. We do this partly because the relationship between civil and criminal law is a significant theme of any such discussion. But the expression is not perfect. In the early stages of an extended review of the area, the Law Commission observed that in no modern case has moral turpitude in the absence of illegality been sufficient basis for the policy bar to operate.2 More recent cases suggest, at the very least, softer boundaries to the relevant ‘turpitude’ which might suffice. Even so, courts in these instances have continued to refer to ‘illegality’ as providing the core of their reasoning, even if by close analogy.3 Three clarifications are however required.
First, and contrary perhaps to appearances, ‘illegality’ itself is far from being a straightforward category, and we would argue that the expression should be seen as imprecise. This in itself is one reason, among others, why it may prove difficult to sustain an argument that special weight is always to be attached to a category of conduct which is criminal in nature.4 It is best to acknowledge that there is no single such category. This, however, will have implications for what is now the most clearly articulated subset of cases in which illegality will bar a claim in tort, namely those where the claimant seeks indemnity for the burden of a legal penalty.5 If illegality is a disparate, or even fuzzy category, then it is harder to explain why tort should be made consistent with it, irrespective of the substantive fairness of doing so.
Second, in the context of insurance contracts in particular, illegal conduct overlaps with a separate question about the insurability of risks which are brought about intentionally by the assured. The starting position is that insurance policies by their very nature respond only to events which are, as far as the assured is concerned, fortuitous or accidental.6 Many insurance policies expressly exclude loss caused intentionally or as the result of wilful misconduct, but such clauses add little, other than clarity, to the common law. By contrast, mere carelessness, or even the deliberate creation by the assured of a situation which makes a loss more likely, does not preclude recovery,7 and ‘reasonable care’ clauses are construed not at their face value, but as excluding only reckless conduct.8 While many insurance policies will restrict cover to damage that is ‘accidentally’ caused by the assured, the policy of the law adds an illegality bar, and that is our core focus here.
Third, categories of ‘intention’—such as those employed by the law of tort—do not map neatly onto the criminal law. It is true, and significant, that torts which require an element of intention are more likely to amount also to criminal wrongdoing, in contrast to wholly non-deliberate acts of negligence. However, it is important to be clear that not all breaches of negligence duties are non-deliberate; that torts categorized as ‘intentional’ are nevertheless often very strict in some of their elements;9 and (most fundamentally perhaps) that criminal liability itself by no means always requires intention. Intention and illegality are overlapping, but not identical, categories. Insurance cover may be lost either because the harm was intended (a prerequisite of some ‘intentional’ torts, but not others), or because it is caught by the policy bar.
Following this introduction, 11.2 briefly outlines the nature of the law’s response to illegal conduct in relation to liability, with a particular focus on the law of tort. In the sections that follow, these principles remain in play as we introduce the role of illegality in relation to insurance. 11.3 reviews the operation of illegality in relation to claims against assured parties. 11.4 explores the operation of illegality in relation to claims against insurers whether on the part of assured parties or on the part of those who have suffered harm through the wrongdoing of an assured. In each of these sections, we explore how the separation of liability and insurance is tested (or their connection defined) by such cases. Where it operates, the impact of the illegality bar is to set limits to the distribution of risks and losses. There is some evidence that in the operation of these limits, courts interpret the principles relating to insurance of liabilities as more susceptible to notions of culpability than the principles of liability themselves. The question inevitably arises of whether this is a satisfactory state of affairs; and that in turn raises the relative nature of the two apparently distinct phenomena—insurance and the law of obligations—with which we started.
The common law rules on public policy, often expressed in the maxim ex turpi causa non oritur actio, operate in a range of circumstances to preclude an action by a claimant guilty of some form of ‘illegality’ sufficiently linked to the claim. Although the rules are often deployed as a pleaded defence, in principle the court may take cognizance of the claimant’s illegal actions and bar the claim of its own volition;10 and it is long established that the aim of the bar is not to protect the defendant, who may benefit while remaining equally undeserving.11 The policy bar has given rise to much judicial consideration and academic debate. Between 1999 and 2010, the Law Commission undertook an extended review of the operation of public policy in tort and contract claims, as well as unjust enrichment and proprietary claims. Its initial view was that reform was needed to bring some measure of order into the chaos generated by an absence of clear objectives and a series of apparently irreconcilable decisions,12 which in the Law Commission’s view was exacerbated by the adoption of ‘rigid rules’ with potentially arbitrary effects. Primary among these was the rule, originating in the law of trusts but thought at the time of the review to have more general application, that illegality bars a claim where (and only where) the claimant must rely upon its own illegality in framing the claim.13 In particular, this was thought to have ousted a previously existing ‘public conscience’ test, which was more adaptable to the kinds of policy reason which the Law Commission thought most fruitful.14
The law of tort was not initially a part of the Commission’s investigations, but the review was inevitably extended to tort: it was thought that consistency would demand its inclusion if legislation should be the chosen path. Equally, the threatened extension of the ‘reliance principle’ into tort cases will have played a role in drawing tort claims into the consultation.15 The Law Commission’s initial view was that a structured approach based on clearly stated statutory policy criteria should be adopted. It argued that on the whole, courts had not reached unjust decisions in relation to illegality in tort and contract, but that express appeal to the underlying objectives of the policy bar on a case-by-case basis would improve the clarity and transparency of the law. We suggest that a fuller picture, incorporating the insurance position, is all the more pertinent if attention to policies, results, and the actual impact of illegality rules is indeed considered important, and we hope to show how that fuller picture raises some additional questions.
By 2009 the Law Commission took the view that a series of intervening decisions had to some extent clarified the law, adopting a more open appeal to the policies behind the illegality defence along the lines it preferred, and not rigidly adopting the ‘reliance’ principle. The Law Commission chose to see the direction of case law as more or less compatible with the structured approach it had put forward, so that there was nothing to be gained by statutory intervention across the majority of areas. In their details, these intervening decisions do not, in fact, appear wholly compatible with the Law Commission’s proposals.16 However, judicial flexibility had largely been maintained, and this was perhaps the crucial factor. Legislation was proposed only in relation to the law of trusts, where Tinsley v Milligan17 was a barrier to judicial development.18 The government has now expressed itself unlikely to make time for legislation.19 Perhaps Parliamentary instinct in relation to wrongdoing by parties to litigation is currently inclined to be more moralistic than those reflected in the Law Commission’s search for balanced and proportionate responses to claimant wrongdoing, premised on clear policy reasons.20 Such questions, though speculative, are significant for our analysis, given that the underlying issues in relation to the policy bar bring into play the very objectives of both liability and (on our account) insurance of liabilities.
As already indicated, our concern in this chapter is with some of the fundamental issues left untouched by the Law Commission’s exploration. It is noteworthy that the Law Commission documents—and indeed most of the decisions of the courts—contain little or no discussion of the effect of public policy upon insurance arrangements. Whatever the general theoretical reasons offered for the exclusion of insurance considerations from discussion, this would appear to be an extreme case since the bar is itself expressly based on policy considerations.
The effect of public policy on insurance arrangements, which is felt both for claims against assureds and claims by assureds for indemnity, is profound. First, a claimant guilty of wrongdoing may lose a claim against the assured, and here we consider why it should be the case that the victim of a tort is denied compensation. The question is particularly pertinent where the state had deemed that insurance against the wrongful act in question is compulsory. In principle the decision to require insurance does not in itself change the nature of liability,21 and in tort theory, the two are insulated from one another.
Second, even where the victim is a paragon of virtue, illegal conduct on the part of the defendant which has given rise to a claim by the victim may strip the defendant of a claim against his or her liability insurers and thereby deprive the claimant of any fund to pay their claim. The irony here is that the more reprehensible the assured’s wrongdoing, the less chance there is of recovering an indemnity under the policy. In short, in those instances where compensation is dependent on the defendant’s liability insurance, a claimant’s chance of full compensation is far better if he or she is the victim of negligence rather than the victim of fraud or deliberate assault. This raises a need to question the priorities of tort law.
There is a correlation between the exclusion of insurance recovery where the assured has acted illegally, and the very shape of the development of the law of tort. In practice, tort law has come to respond habitually—even routinely—to a range of negligence-based and indeed strict liabilities, rather than to liabilities turning on intention. Yet if we were to think in terms of a hierarchy of civil wrongs it seems likely that deliberate harm would be considered most deserving of a remedial response. The practical difficulty of obtaining compensation operates as an unseen factor in this aspect of the shape and reach of the law of tort. The effect—an increased emphasis on negligence rather than intentional harm—has been termed ‘underlitigation’.22 This we suggest is evidenced in two distinct ways. First, it goes some distance to explaining the almost complete absence of some serious wrongs from the field of tort despite the availability of legal principles which would attach liability for those wrongs.23 Second, it would explain the presentation of events which involve intentional wrongdoing as really turning on carelessness or even breach of strict duties. In this second form, ‘underlitigation’ (not capturing the real nature of the wrongdoing) may also lead the claimant to seek a different defendant, whose insurance policy is more likely to respond (or who may have sufficiently deep pockets). Underlitigation in this second sense is accompanied by the stretching of the law of negligence, as merely negligent parties with a causal link to deliberately cause harm are drawn into the frame.24 In many instances, these will be professionals, or public authorities, who ought to have protected the claimant and avoided the harm. In some jurisdictions, the limits to insurability of deliberate harms have been instrumental in the development of proportionate forms of liability, designed to protect negligent parties from taking the full burden of liabilities for harm where the most blameworthy perpetrator is not worth pursuing.25 The general point is that in any tort case the conduct of both parties will determine whether the claimant actually recovers compensation for his or her loss. It is our suggestion in this chapter that consideration of the insurance issues, and therefore of the actual implications of public policy defences, brings these defences and questions about their purpose and limits into much sharper focus.
In considering recoverability for a tort claim or under a liability policy, it is necessary to draw on the distinction between primary and vicarious/secondary liability which was outlined in Chapter 10. The result of this distinction is that a person who faces vicarious liability for the acts of another cannot be barred from making a claim for indemnity against a tortfeasor or its insurers simply because of the nature of the tortfeasor’s conduct. As we saw, an organization which faces liability for the acts of an employee or agent may seek indemnification from a tortfeasor or insurers unless the employee/agent is the alter ego of the organization,26 or the organization has condoned or participated in the conduct of the employee/agent. For these reasons there is nothing surprising in the idea that a public liability policy provides coverage for vicarious liability for damages awarded against an employee/agent who has deliberately or recklessly inflicted personal injury on a third party,27 a principle which has been applied in English law even if those damages are exemplary or punitive in nature.28
11.2.1 Why is there a public policy defence?
The public policy rules for tort claims and for insurance indemnity claims29 are derived from the same sources, though as we will see, the impact of illegality is acknowledged to be highly dependent on circumstance. The law as it applies to contracts has been developing since the eighteenth century, but the extension of ex turpi causa from contract to tort is a much more recent and controversial development.30 As late as 1954, in NCB v England,31 Lord Porter doubted whether the defence had any application to tort at all. These doubts were expressed in a case where the claimant (as well as one of his fellow workers) had breached statutory safety duties backed by criminal sanctions. The duties were designed for the protection of the class (workers) to which the claimant belonged.32 One reason for this conclusion was that breaches of statutory duty of this sort were clearly included within the definition of ‘fault’ on the part of claimants to which the Law Reform (Contributory Negligence) Act 1945 applied, so that damages must, under the statute, only be reduced. Further, applying the views of Lord Cohen in Cakebread v Hopping Brothers,33 ex turpi causa is based on public policy, and public policy in such a case—far from requiring that damages be refused—requires that the courts should entertain the claim for compensation. And indeed, the effect of applying the bar automatically to breaches designated criminal will be that as Parliament acts to create more criminal sanctions with a view to enhancing protection and deterrence, the reach of compensatory responses is correspondingly—and unintentionally—reduced. This is plainly a perverse result. It is thus positively clear from NCB v England, leaving aside Lord Porter’s doubts about whether an illegality defence would ever apply to a claim in tort, that not all illegality causally linked to a tort will be caught by the policy bar. Breaches of statutory duty aimed at securing the safety of individuals such as the claimant will not, and this is the preserve of contributory negligence or, in an extreme case, volenti or assumption of risk.
We have expressed this conclusion relatively narrowly, by reference to the kind of criminal liability involved in the case.34 But this attempt to confine the reasoning in NCB v England within manageable bounds has its own implications. Particularly, it implies that as a consequence of the authority of this case, it is not possible to maintain that civil courts will never scrutinize Parliament’s intentions in creating a criminal liability, or the nature of that liability, when determining a civil claim; though it has been authoritatively stated that this will be the case within the preserve of the ‘narrow rule’.35
More recently, it has been said that the rules underlying the illegality defence in tort, as elsewhere, are based upon a combination of factors, including deterrence, punishment, removing the proceeds of wrongdoing, and the maintenance of the dignity of the courts.36 The Law Commission supported the existence of an illegality rule in tort, but its support was balanced and generally not based on a punitive attitude. The Law Commission accepted that public policy might be explained on the basis of the need to maintain consistency in the law, but seems not to have confined its remarks to the precise, indeed rather formal, meaning of ‘consistency’ adopted by the Supreme Court of Canada and endorsed by Lord Hoffmann in the formulation of a ‘narrow rule’ in Gray v Thames Trains.37 Nor did it consider ‘consistency’ to be the sole successful rationale. When looked at in the context of insurance—which is to say, with a fuller range of policy implications in range—our view is that qualifications and objections to the suggested rationales become even more compelling.
Deterrence arguably has its place where two parties knowingly enter into an illegal contract, but it seems of limited significance where it is applied to conduct which may be, and, where tort and insurance claims are concerned often is, isolated, unpremeditated, and motivated by particular unique circumstances.38 The Law Commission further suggested that ‘[i]t is difficult to suggest that a person who is not deterred from these activities by the threat of criminal sanction will be deterred by the possibility that he or she may not receive compensation for any loss suffered in the course of, or as a result of, committing the offence’.39 Perhaps counter-intuitively, this is not necessarily sound reasoning. Indeed, one of the few leading English cases where a policyholder sought a benefit from his own crime—though necessarily for his estate rather than for himself—was a case of suicide.40 Beresford v Royal Exchange Assurance41 is a case of contract rather than tort. Here the terms of a life insurance policy did not exclude recovery following an act of suicide other than in the first year of the policy, but the rule of public policy was deployed to prevent recovery against the insurers. Deeply indebted thanks to a failed business venture, the assured shot himself three minutes before the expiry of a negotiated extension to the renewal date of the policy, evidently with the express intention of securing the insurance for his dependants. The House of Lords held that he (in the guise of his estate) should not be permitted, as a matter of public policy, to benefit from what was then a criminal act. Indeed, even where the malefactor survives, it should not automatically be assumed that criminal sanctions outweigh the civil ‘sanction’ of being deprived of financial benefit.42
The punishment rationale has little to commend it, and was rightly omitted by the Law Commission from its proposed structured discretion. As already pointed out, the civil consequences of depriving a victim of a tort claim or an assured of an insurance claim may be wholly disproportionate to the wrongdoing, there is a risk of double punishment, and the outcome can be a windfall for the defendant.43
Another proposed rationale—preventing the claimant from benefiting from his or her own wrongdoing—may have its place where the parties set out on an agreed profit-making illegal enterprise, but that is not the paradigm case. The victim of a tort does not seek to make a profit, but rather to recover compensation for a loss, so the question becomes whether it is appropriate to remove that compensation.44 The question poses particular difficulties where the defendant is required by law to insure against that very form of loss, as we see below. Equally, in the case of a claim under a liability policy, if indemnity is refused, the impact on the assured personally is likely to be limited unless the assured has the resources to pay the claim from his or her own pocket. In most cases the assured will not be able to pay the claim, and if the insurers do not meet the liabilities the claimant, who has done nothing wrong but simply become a victim, will go uncompensated unless he or she invokes the full force of insolvency law against the assured, potentially bringing the assured’s business to an end or leaving a private assured bankrupt and without means of support. Clear proportionality issues are raised here.
Some care is also needed in any discussion of the need to preserve the dignity of the courts and the reputation of the legal system. This justification may be no more than an articulation of ‘an unreasoned, ‘gut-feeling’ basis … which is potentially disproportionate and wrong’,45 and the Law Commission has stated that, at least for tort claims, the circumstances ‘in which the court would feel compelled to protect its dignity are likely to be very extreme, and therefore rare’.46 However, the House of Lords, in Gray v Thames Trains formally approved a related but narrower idea that the illegality defence is necessary in order to protect consistency in the law: the law could not consistently find the claimant criminally liable for manslaughter, thus personally responsible, and at the same time hold the defendant responsible for the claimant’s act.47 In this particular context, the consistency principle appears to protect the law’s claim to attach sanctions on the basis of personal responsibility. Whatever may be said about Gray v Thames Trains itself,48 we will later question the ambit of this principle.
The Law Commission nevertheless thought that the illegality bar should be retained in some form, for two particular reasons, and for a more generic reason which may rest on a number of policy considerations. The two limited reasons were to prevent the claimant from suing in tort where the action would otherwise be one to enforce an illegal contract; and to remove the possibility that the claimant might recover indemnity for a fine or other sanction imposed by the criminal law.49 Those are limited situations: the former is most unlikely to apply to a tort claim and probably never to an insurance claim; and the latter—which the courts have, since the Law Commission’s deliberations, adopted as the ‘narrow rule’ for denying a cause of action—may, as we point out below, have undesirable consequences. The Law Commission added a third ground, where neither of these two was applicable, namely a ‘general residual principle’ that the court ‘should not assist a claimant who has been guilty of illegal conduct of which the courts should take notice’.50 This, clearly, is more amorphous. However, the Commission proposed a list of accepted policy rationales on the basis of which the illegality defence or policy bar should be ‘allowed’; and urged that the result of the illegality defence (its impact on the wrongdoer) should remain ‘proportionate’ to the illegality involved51—a signal that whatever the strength of arguments for clear rules, the court should not flinch from assessing the nature and significance of the rule of law infringed by the claimant, and the quality of the conduct. In ensuring a proportionate response however, can courts avoid consideration of the nature and purpose of liabilities and other surrounding legal rules, in which insurance may be a key consideration? Lord Hoffmann, in Gray v Thames Trains, took the view that a court deciding a civil claim should not add to its burden by considering the nature and purpose of aspects of the criminal law and making distinctions between them, or ‘having to play the criminal law’s conscience’.52 Lord Phillips’ opinion was in marked contrast with that of Lord Hoffmann on this point. He took the view that if the hospital order against the claimant had not been accompanied by an indefinite restriction order,53 this would not have been a case where ex turpi causa should bar the claim. Thus he took the course of interpreting the ‘meaning’ of the criminal sentence. In the following section, such questions emerge as central to our analysis of the relationship between liability and insurance, for they invite consideration of the multiple potential purposes of the former.
11.3.1 Legal principles
Our key concern in this section is with the law of tort, and particularly with its use against insured defendants. Although the authorities in this area are not fully reconcilable,54 the most authoritative formulation is that of Lord Hoffmann in Gray v Thames Trains Ltd.55 Lord Hoffmann recognized that ex turpi causa may be based on a range of policy reasons which will vary with different contexts.56 In the context of that case, it consisted of a narrow and a wide rule. The narrow rule is that a claimant cannot recover an indemnity for a fine or other penalty imposed upon him for the commission of a criminal offence. As expressed in Gray v Thames Trains, this is an absolute principle, but, as we see in the paragraphs to follow, it has been doubted—with good reason—whether it can apply to strict liability crimes where the assured’s infringement was beyond his control, or was the fault of another.
The wide rule is that there can be no claim where criminality can properly be said to be the cause of the loss and is of a nature which would cause a court to refuse to lend its assistance. The wide rule allows the court considerable leeway in determining whether it is appropriate to impose a civil sanction for criminal conduct: as discussed below in the context of claims by policyholders, this is of particular significance in respect of claims under liability policies, where denial of a claim impacts primarily on the injured victim. Nevertheless the rule, as Lord Hoffmann expressed it, leans heavily on principles of legal causation (whether it can ‘properly be said’ that the loss was caused by the criminal conduct), which he acknowledged are not straightforward. The notion of ‘legal cause’ being advanced here is likely to be imbued with notions of responsibility. Interestingly, as Lord Hoffmann presented it, the wide rule in this context depends not on ideas of consistency, incentives, or whether there might be injustice between claimant and defendant, but on the idea that ‘public notions of the fair distribution of resources’ would be offended by compensating the claimant for the consequences of his own criminal conduct.57 There has been criticism of the latter argument on the basis that the aim of the law of tort is not in any event to achieve a fair distribution of resources and, indeed, that tort law is unconcerned with achieving such a distribution.58 We would argue that Lord Hoffmann’s point is not that tort can be interpreted as aiming at a fair distribution of resources. Rather, he was pointing out (not for the first time) that the potentially undesirable distributive impact of tort liabilities can legitimately be considered when determining how far those liabilities should extend.59 Lord Hoffmann may particularly have had in mind the special status of tort victims as enjoying compensation premised on making good losses, rather than the more partial support achieved by the majority of those suffering misfortunes, given that the tort claimant in this case had attacked and killed another individual. The issues here are complex, and deserve a little more explanation, but there is a limit to what can be gleaned.
In the particular context of Gray v Thames Trains