The Asbestos Litigation
12.1.1 Our focus: insurance and liabilities
The decision of the Supreme Court in Durham v BAI (Run off) Ltd (the Trigger Litigation, referred to here as Trigger)1 was centrally concerned with liability insurance policies and their construction. But much more than this, the majority decision decisively interpreted the nature of the liability principles applied in relation to the disease of mesothelioma,2 and in the process shed considerable light on the relationship between tort liabilities and insurance with which we have been concerned in a number of chapters of this book. The decision should, in principle, allow the conclusion—so far as tort liability is concerned—of an episode in which the courts first created an exceptional approach to causation (Fairchild v Glenhaven Funeral Services Ltd3) in order to assist claimants who would otherwise founder on a ‘rock of uncertainty’ (the Fairchild exception); then appeared to undermine that exception by interpreting it in a fashion which would redefine the very nature of the liabilities concerned (Barker v Corus4). This was done in the name of coherence and fairness, but its effect was incoherence and, we suggest, unfairness.5 These problems of incoherence and unfairness are brought into sharp focus once the distributive consequences, relating to insurance, are taken into account. Only when the Supreme Court had to decide, in Trigger, whether employers’ liability insurance policies would respond to the liabilities incurred did it become completely obvious what adverse impacts would flow from acceptance of a ‘principled’ interpretation of Barker.
Lord Phillips brought the point to the surface, but we suggest that he drew the wrong conclusions from it. Dissenting, Lord Phillips suggested that the law as set out in Barker should not be redefined (as he saw it) ‘with the object of enabling claims to be brought under the employers’ liability policies’.6 The goal of Fairchild and Barker, in his view, had been to ‘ensure that those who had breached the duties that they owed to their employees did not escape liability because of scientific uncertainty’.7 Our argument is that this would have been a vacuous goal for a special rule if it simultaneously negated the existing insurance cover, not least where (as in Fairchild) the defendants were plainly insurers. We therefore do not agree that the liability question and the compensation question are separable justifications for Fairchild. It follows that we also do not agree with Lord Phillips that defining the rule as set out in Barker to enable insurers to be liable would be to act for ‘reasons of policy’, and should be a matter for ‘Parliament not the courts’.8 Leaving aside the question of how Parliament could conceivably bring about the required interpretive change,9 we seek to show in this chapter that the concerns expressed by Lord Phillips are, with great respect, misplaced.
We explain our reasons for these views in the course of the chapter. But it will be apparent that the issues go to the heart of several of the themes of this book. Most notably, Trigger underlines that while insurance does not uniquely influence the content of legal principles, neither is it safe to treat insurance factors as irrelevant, most particularly when considering questions of ‘fairness’. Further, as the majority judgments of Lords Mance and Clarke made clear, the purpose of liability insurance is to respond to liabilities: insurance is matched to risks. It is inappropriate for the common law to redefine liabilities so that they are not susceptible to being insured by policies already in place, and then to call the result principled. Nor, we suggest, is there any compelling reason why courts should choose to redefine the liabilities in the first place, if this is the result.
12.1.2 Mesothelioma: the special issues
The cases discussed here all concern the dreadful disease of mesothelioma. Here we briefly address those features of mesothelioma which have created a challenge for the law.
Much is still to be learned about the aetiology of the disease. However, the typical pattern for the development of mesothelioma is as follows. Exposure to asbestos leads to asbestos fibres being lodged in the lungs; some thirty-five years after exposure, the fibres take hold, producing a tumour with its own blood supply which causes damage to the lungs; some five years after that, there is manifestation of the disease; and within a matter of months, inevitable fatality. Fortunately, only about 3 per cent of persons exposed will develop the disease. It remains unclear whether a single exposure can cause the disease or whether the disease develops as the result of cumulative exposures, although it is obvious that the more exposures there are, the greater the risk of disease becomes. There are very few cases of mesothelioma which are not related to asbestos exposure. It is crucial to an understanding of the legal issues to appreciate that mere exposure is not of itself injury, and the thickening of the pleural membrane which may or may not be a precursor of disease similarly is not injury.10
In the next section, we begin our discussion with the approach of the courts to tort principles, and we follow with the insurance litigation which gave practical effect to those principles. But our analysis does not stop there. The reversal of Barker v Corus11 by the Compensation Act 2006 means that each employer faces joint and several liability for any damages claim, thereby removing the complexities of apportionment by employees in their tort claims, and thus the need for tort scholars to speculate upon how allocation would have been effected but for legislative intervention. However, those complexities have not been eliminated, merely shifted to a second-tier range of disputes between the insurers of employers, and onwards to a third tier of disputes between those insurers and their reinsurers. This chapter concludes with a discussion of the complexities of those ‘hidden’ disputes.
The decisions in Fairchild, Barker, and Sienkiewicz have been widely analysed.12 Our core focus, as already explained, is on the implications of Trigger; but we must set out the essential features of the preceding tort litigation. In Fairchild, the House of Lords responded to a typical scenario in which a person has been exposed to asbestos over a period of years by successive employers. Assuming that each exposure is a breach of the duty of care,13 and that damage is a given, attention focuses purely on causation. The common law rule of causation, the ‘but for’ test, requires the conduct of a tortfeasor to be looked at in isolation, the question being: what would the claimant’s position have been but for the conduct of the tortfeasor? Where there is a sole tortfeasor and no relevant external cause of injury, the process is relatively straightforward, and an assessment may be made of the tortfeasor’s intervention. However, the addition of further tortfeasors, or potential external causes, creates complexity in that it may be that it is impossible to determine the impact of any one breach of duty, or it may be that each breach of duty has made no difference to the claimant because of the existence of the other torts or causes. It would be curious for the law to say that each tortfeasor may rely upon the acts of the other to escape liability,14 a solution which the law has rejected in other contexts.15 That possibility was removed in Fairchild, albeit only in the context of mesothelioma. Although their Lordships expressed themselves in different ways, sometimes within the same judgment, the outcome of Fairchild was that proof of ‘but for’ causation was to be set aside in the specific case of mesothelioma, because scientific knowledge was not able to pinpoint the triggering exposure. ‘But for’ was replaced by the principle that each exposure must amount to a material contribution to the risk of injury, and that was enough to impose liability on every person whose breach of duty had resulted in exposure—or perhaps, sufficient exposure.
Fairchild was generally assumed to have produced a joint and several liability principle, with the consequence that a victim could sue any one of the tortfeasors for the full amount of his loss. That is the basis on which insurers made their arrangements for contribution.16 That assumption was successfully challenged however in Barker, where the House of Lords by a majority concluded that liability was several and not joint. Every tortfeasor was thus responsible to the extent to which its own conduct materially contributed to the risk of injury. The Barker ruling created a number of immediate problems: the means of allocation was not resolved, potentially giving rise to disputes about whether proportionate liability should follow time on risk or whether it should be varied to take account of periods of particularly intense exposure or the conditions in which exposure occurred; and a victim might have had a period during which he or she was not exposed to asbestos, or in which the exposure was due to his own activities, so that compensation would be reduced by whatever proportion of the loss was to be attributed to those periods.17 Much more significantly however, the substantial risk of employer or insurer insolvency, or of untraced employers and insurers, and of proof of the relevant proportion of exposure to be allocated to any given defendant once identified, was shifted to claimants. We explore further in 12.4 the distributive impact of joint and several liability, by attention to contribution arrangements.
The unfairness of the outcome in Barker in this last respect was identified immediately. As a result, Parliament intervened to impose the joint and several liability principle which had, prior to Barker, been thought (rightly, it can now be said)18 to exist. The Compensation Act 2006, s 3(1), provides that any ‘responsible’ person who has exposed the victim to asbestos faces liability for the whole loss despite the existence of other wrongdoers or periods in which there was no exposure. Four conditions, set out in s 3(2), have to be satisfied: (a) the responsible person ‘has negligently or in breach of statutory duty caused or permitted’ the victim to be exposed to asbestos; (b) ‘the victim has contracted mesothelioma as a result of exposure to asbestos’; (c) ‘because of the nature of mesothelioma and the state of medical science, it is not possible to determine with certainty’ which exposure caused the victim to become ill; and (d) the responsible person is liable in tort, ‘whether by reason of having materially increased a risk or for any other reason’. Subsequently, in Sienkiewicz, a case involving a single responsible person and also atmospheric asbestos which itself materially increased the risk of injury, the Supreme Court ruled that s 3 did not create liability where none previously existed, and that it merely imposed full liability on any one tortfeasor found to face liability. In Sienkiewicz itself, the responsible person carried the full burden: the employer had acted negligently, mesothelioma had been contracted as a result of exposure to asbestos, medical science could not determine which exposure had led to the disease, and the responsible person was liable in tort under the Fairchild test because the exposures materially contributed to the risk of injury.
The judgments in Fairchild and Barker evidently raised the question of the nature of the liability imposed in those cases. From one point of view, their Lordships had simply tinkered with the rules on causation, adjusting them to fit a novel situation and deeming there to be a causal link between each exposure and subsequent injury. This employs an admitted fiction, but it does so in a contained manner, for a particular purpose in particular circumstances. This is the interpretation now authoritatively adopted by the majority of the Supreme Court in the Trigger litigation. An alternative analysis was that Fairchild had created a new form of liability, perhaps even a new tort of materially increasing the risk of injury.19 The resolution of that debate does not now affect the joint and several liability of each responsible person towards the victim. There is a need under the 2006 Act, s 3(2)(d), to show liability in tort. Such liability attaches either because the breach of duty of every responsible person is deemed to be the cause of the injury, thereby perfecting the tort arising from negligent exposure and damage, or because the mere act of exposure is a tort in its own right as long as damage follows. In either view the employee does not have to pinpoint a causal exposure, but needs to point only to exposure and subsequent disease. But Trigger reignited the debate, because its resolution determines whether or not the responsible person may rely upon liability insurance in order to fund any successful damages claim by or on behalf of the victim. We would underline once again that in Fairchild itself, it was abundantly clear that the compensation secured would be payable by insurers, and that insurers were indeed conducting the case as defendants. Every step in the asbestos litigation has been shaped by insurers.
12.3.1 Insuring liability for personal injuries
It will be seen from the above account that the courts and Parliament have been enthusiastic and inventive in their attempts to ensure that wrongful exposure leads to legal liability in cases of mesothelioma. But of course that is only half of the story. In this context, where claimants are suffering from a dreadful disease and where no responsible individuals, and few if any employing organizations, are likely to bear the brunt of liability, the point of liability is that it is followed by actual payment. We here consider the response of insurance principles. The courts have ultimately ensured that insurance coverage matches legal liability. The ‘remarkable’ consequence of the alternative view exemplifies the perils of fashioning what appears to be a principled solution without regard to the operation of insurance.
We have already seen that liability insurance can respond in one of three main ways:20 to an event occurring within the currency of the policy (causation basis); to injury or damage taking place within the currency of the policy (injury basis); and to a claim made against the assured (or, by extension, a circumstance notified to the insurers by the assured) during the currency of the policy (claims made basis). Personal injuries are covered either by employers’ liability (EL) policies or by public liability (PL) policies, the latter extending to injuries inflicted upon persons other than the assured’s own employees, most importantly, visitors, independent contractors, and members of the public in general. Personal injury liabilities of both types are generally insured under either causation or injury policies: there was a short period in the 1990s when claims made policies were used for EL insurance, but that has since been discontinued.
In the vast majority of cases it matters little whether the insurance is causation or injury based, because there is generally no temporal gap between an act of negligence and the injuries which flow from it, and a policy of either type will respond. The distinction becomes significant where the negligent act occurs in one policy period and the injury occurs in another, a situation which is more or less confined to injuries resulting from a negligent exposure to a harmful substance, and mesothelioma caused by exposure to asbestos is perhaps the most complex example because of the lengthy gap, measured in decades, between exposure and injury. The date of the trigger is important for a number of reasons, which may be illustrated by a simple example. D exposes C to asbestos in 1980, and in 2010 C suffers injury to his lungs. The relevant insurance years are thus 1980 and 2010.
If D has a causation policy in 1980, those insurers are liable, but if he has an injury policy in 1980 then the insurers are not liable because there is no injury in 1980.
If D has a causation policy in 2010, those insurers are not liable because the negligent act of exposure did not occur in 2010, but if D happens to have injury insurance in place in 2010 then—subject to one important consideration—those insurers will face liability. That important consideration rests upon the status of C. If the 2010 policy is PL, it will respond to injuries inflicted on third parties in 2010. However, if the 2010 policy is EL, it is likely to apply only if C is in D’s employ in 2010; that is because EL policies cover only injuries to existing employees and not ex-employees. Further, such a policy is likely to exclude claims for injuries arising from earlier exposures.
It is immediately apparent, therefore, that an injury-based policy is, in the vast majority of employment cases, of little value for a disease such as mesothelioma. If the policy is in 1980 it will not respond at all, and if the policy is in 2010 then it will respond only if C was employed by D in the year of exposure (the negligent act) and remains employed in 2010 (the year of injury), and if there is no exclusion for historic exposures.
A number of factors complicate the position. The first is that assureds regularly change insurers and the wordings used by the insurers may be different. Thus it is perfectly possible that D had taken out an injury policy in 1980 but switched to a causation policy in 2010, in which case neither insurer will be liable because the former responds to injury (which does not occur for a further thirty years) but the latter responds to exposure (which took place thirty years previously). Second, and linked to the first point, it is a fact that many of the EL insurers writing business in 1980 had ceased to do so by 2010, and that many employers in business in 1980 had been dissolved by 2010. Third, even where the same insurer is on risk in both 1980 and 2010, and each of the policies is the same in nature, it is inevitable that one or other of those policies will be applicable, but that does not mean that the problem is solved. It will still be necessary to determine which of the policies covers the loss: D’s insurers will almost certainly have had different reinsurance arrangements in 1980 and 2010, and reinsurers will insist that the claim is allocated to the correct policy year; and even if reinsurance is left out of account it is possible that the cover in either of the years may have been exhausted by other claims, so that in practice a claim may only be made in the other year.
This, then, sets the scene both for the insurance litigation which followed the asbestos liability litigation fought so vigorously in the highest court, and for the intervention of Parliament. The problem, stated simply, was that the wordings of both PL and EL policies have long been ambiguous, in that they refer to ‘injury sustained’ or ‘disease contracted’, or both. In addition, in the EL market, some policies stated that they insured the risks required by the Employers’ Liability (Compulsory Insurance) Act 1969. Others did not, but were otherwise no different in their terms. The primary question for the courts in the sequel litigation was the meaning of the italicized phrases, and also what form of policy was required by the 1969 Act.
12.3.2 The prequel: Bolton v MMI
The scene for the battle over the meaning of EL policies was set21