(with whom Lord Clarke, Lord Carnwath and Lord Hodge agree)
It is the role of the common law to adapt to meet new circumstances and challenges. Mesothelioma has been and is a tragedy for individuals and families. It is caused by exposure to the inhalation of asbestos dust, and has a gestation period measured typically in decades. The more fibres inhaled, the greater the risk of contracting mesothelioma. But, beyond that, its specific causation is highly uncertain: see Sienkiewicz v Greif (UK) Ltd  UKSC 10,  2 AC 229, para 19, Durham v BAI (Run off) Ltd  UKSC 14,  1 WLR 867, para 6. It was thought it might be caused by a single fibre, but Lord Phillips' annex to his judgment in Sienkiewicz, part A, paras 10-11, notes that the process of causation may involve (different) fibres acting in a way which gives rise to a series of as many as six or seven genetic alterations, ending with a malignant cell in the pleura. In any event, the evidential uncertainties about its causation led the House of Lords in Fairchild v Glenhaven Funeral Services Ltd  UKHL 22,  1 AC 32 to create a special common law rule, operating within what may be called the Fairchild enclave, to govern liability between victims and those who in breach of duty had exposed them to asbestos dust. Following the House's decision in Barker v Corus UK Ltd  UKHL 20,  2 AC 572, this special rule was fortified by the Compensation Act 2006. Unsurprisingly, the courts are still working out the implications. Courts which have embarked on it have had to focus on disputes gradually shifting from (a) the position between victims and those responsible for their exposure, on which substantial authority now exists under English law, to (b) the position between persons so responsible and their insurers. This appeal and the conclusions I reach on it are concerned exclusively with situations falling within the special rule.
The appeal, brought by Zurich Insurance plc ("Zurich") as appellant against International Energy Group Ltd ("IEG") as respondent, raises points under both (a) and (b). The issues under (a) are subject to Guernsey law, and there is a difference between the English and Guernsey statute law. The parties are however agreed that Guernsey common law is to be treated as identical with English common law on this appeal.
According to the special rule recognised by the House of Lords, a person contracting mesothelioma, after being exposed to significant quantities of asbestos dust originating from different sources over the same or different periods, can sue any person who was (negligently or in breach of duty) responsible for any such source of exposure, although unable to show which exposure in probability actually led or contributed to the disease: Fairchild and Barker. This rule applies even if the only potential sources consist in the ambient environmental exposure which the population generally experiences and some other negligently created source which only increases this ambient exposure by a small percentage - 18% in the case so holding: Sienkiewicz.
The special rule confers a right of suit on victims of mesothelioma by reference to each significant exposure, rather than any probability that the particular exposure relied upon led or contributed to the disease. As formulated in Fairchild, it left open the damages recoverable from a person responsible for an exposure. In Barker the House of Lords held that a person responsible was liable not for the whole damages attributable to the mesothelioma, but only in proportion to his own contribution to the overall exposure, probably measured by the duration and intensity of the particular exposure for which he was responsible. This proportionate recovery applied whether the other sources were tortious, non-tortious, by natural causes or by the victim him or herself.
The United Kingdom Parliament reacted immediately, reversing the House's ruling that recovery should be proportionate by the Compensation Act 2006. This Act preserves all other aspects of the special rule, as is apparent from section 3(1) and (2):
Industry guidelines for apportioning and handling employers' liability mesothelioma claims were developed in October 2003, taking account of the Financial Services Compensation Scheme ("FSCS") available under the Policyholders Protection Act 1975 and the Financial Services and Markets Act 2000 in relation to insolvent insurers. These guidelines were also reflected in the Industrial Disease Claims Working Party handling guidelines issued in 2006, which were themselves revised in 2008 following the expansion of the FSCS by the Compensation Act 2006 (Contribution for Mesothelioma Claims) Regulations 2006 (SI 2006/3259). Lord Sumption has described some of the features of the guidelines, which, it appears, achieved general acceptance in the industry, by the FSCS and by reinsurers, before the decision of the Court of Appeal in the present case on 6 February 2013 appeared to undermine their application. Most recently, after consultations going back to 2010 and to meet the possibility that a mesothelioma victim might be unable to identify any solvent employer with an identifiable insurer, the Mesothelioma Act 2014 has established an insurance industry fund to pay out in such a case a sum fixed by schedule initially at about 80% but since a Ministerial announcement on 10 February 2015 at 100% of the average damages recovery which a victim of the particular victim's age would be expected to recover in a civil claim.
In Durham v BAI (Run-Off) Ltd  UKSC 14,  1 WLR 867 (the "Trigger" litigation), the Supreme Court held that, where an employer is insured against liability for a disease suffered by an employee which has been caused during the insurance period, the necessary causal requirement or link is satisfied in the case of mesothelioma by the employer's negligent exposure of the victim during such period to asbestos (and so to the risk of suffering mesothelioma), with the result that the insurer must indemnify the employer against the liability so incurred.
Guernsey has not passed any equivalent of the United Kingdom's Compensation Act 2006. The first main question on this appeal is whether, apart from that Act, the proportionate recovery rule in Barker still exists at common law. Guernsey common law is, as stated, to be taken to be the same as English common law. IEG's case is that Barker has "become past history" after the 2006 Act and in the light of the Supreme Court's decision in the Trigger litigation.
The second main question concerns the position where the person responsible for exposing a mesothelioma victim has the benefit of liability insurance covering only part of the period for which he exposed the victim. If in such a case the person responsible incurs an expense or liability which is not proportionate, must an insurer who has covered only part of the whole exposure period bear the whole expense or liability? Before the Supreme Court, the parties and interveners accepted that such an insurer must, at least in the first instance, answer for the whole expense or liability, but Lord Sumption's judgment on this appeal raises for consideration whether they were correct to do so. Assuming they were, the further question arises whether such an insurer is in any way entitled to recoup himself proportionately, and if so from whom, when during the remaining period of exposure the employer chose either to insure with other insurers or not to insure at all or no identifiable insurer can now be shown to have covered the employer. If Barker no longer represents the common law, this question arises directly on this appeal. Zurich submits that it anyway also arises in respect of defence costs incurred by or on behalf of a person responsible for a particular exposure, where the overall exposure is greater. Most obviously, it is a question of general importance in the United Kingdom in relation to claims under the 2006 Act, though the present appeal concerns no such claim.
The facts can be shortly stated. IEG is a solvent Guernsey company, a supplier of gas to the Channel Islands and a subsidiary of a global utilities, transport, energy and timber company quoted on the New York Stock Exchange. IEG is the successor in title of Guernsey Gas Light Co Ltd ("GGLCL"), which for a period of over 27 years from 13 November 1961 to 31 December 1988 employed Mr Carré and during such employment exposed him to asbestos dust. Mr Carré subsequently contracted and died of mesothelioma. It is common ground for present purposes that Mr Carré was exposed with the same degree of frequency and intensity throughout the 27-year period, without adequate protection being provided by GGLCL, under circumstances that materially increased the risk of his contracting mesothelioma and constituted breaches of duty by GGLCL towards him.
On 22 September 2008 Mr Carré brought proceedings against IEG claiming that he had sustained mesothelioma consequent on his exposure to asbestos dust throughout his 27-year period of employment with GGLCL. IEG settled his claim on 19 December 2008 by a compensation payment consisting of £250,000 in damages and interest plus £15,300 towards Mr Carré's costs. IEG also incurred defence costs of £13,151.60.
Thereafter IEG looked to GGLCL's liability insurers under policies in force during the period of exposure. Two have been identified, first the Excess Insurance Co Ltd, which provided employers' liability insurance for two years from 31 December 1978 to 30 December 1980, and, second the Midland Assurance Ltd, to whose insurance liabilities Zurich has succeeded, which provided such insurance for six years from 31 December 1982 to 31 December 1988. The present appeal thus proceeds on the basis that GGLCL had insurance for eight of the 27 years throughout which it exposed Mr Carré to asbestos dust. Guernsey did not have legislation making employers' liability insurance compulsory until 1993, when the Employers' Liability (Compulsory Insurance) (Guernsey) Law 1993 came into effect.
Each of the Midland policies issued during the six years when it was on risk provided that:
Whereas the Insured carrying on the business described in the Schedule and no other for the purposes of this insurance has applied to Midland Assurance Limited (hereinafter called the Company) for the insurance hereinafter contained and has paid or agreed to pay the premium as consideration for such insurance during the period stated in the Schedule or for any subsequent period for which the Company shall have accepted the premium required for renewal of this policy.
If any person under a contract of service or apprenticeship with the Insured shall sustain any bodily injury or disease caused during any period of insurance and arising out of and in the course of his employment by the Insured in the business above mentioned, the Company will indemnify the Insured against all sums for which the Insured shall be liable in respect of any claim for damages for such injury or disease settled or defended with the consent of the Company. The Company will in addition pay claimants' costs and expenses and be responsible for all costs and expenses incurred with the consent of the Company in defending any such claim for damages.
IEG notified a claim for its total loss to Zurich, which offered to meet 72/326ths of the damages and interest paid to Mr Carré and of the defence costs incurred. The proportion reflected the relationship between the six years of the Midland insurance and the 27-year period of Mr Carré's exposure by GGLCL. It was arrived at on the basis that IEG's liability to Mr Carré was incurred and increased from day to day throughout the 27 years, while only six years of such liability fell within the period of the Midland insurance. (Any slight inaccuracy in equating a period of 27 years one month 17 days with 326 months can be ignored. Cooke J at trial converted 72/326ths into a percentage of 22.08%, which has not been challenged.)
A trial was ordered on the basis of a statement of facts and issues recording the common ground between the parties, and on 24 January 2012 (two months before this court handed down judgment in the Trigger litigation) Cooke J accepted Zurich's case regarding the compensation, but not the defence costs, paid in respect of Mr Carré. He held it liable to pay £71,729.84 in full discharge of its policy liabilities, being its relevant proportion of such compensation plus 100% of the defence costs:  EWHC 69 (Comm). On 6 February 2013 the Court of Appeal allowed IEG's appeal, rejected Zurich's cross-appeal relating to defence costs, and ordered Zurich to pay £278,451.60, representing 100% of both the compensation paid and defence costs incurred by IEG:  EWCA Civ 39.
The Trigger litigation
The issue in the Trigger litigation was whether and how various differently worded employers' liability insurance policies should respond to mesothelioma claims. Typical wordings in use at various relevant dates were set out in annex A to and summarised in paras 7 to 9 of my judgment in that case.
Under some of the policy wordings there considered (including some early Excess policies in different form to the present), the insurer promised to indemnify the insured employer against liability if at any time during the period of insurance (or of any renewal) any employee should sustain "personal injury by accident or disease" or "[any] bodily injury or disease", while engaged in the service of the employer or in other cases "arising out of and in the course of [his] employment" by the insured employer. Other policy wordings were in more developed form, promising for example indemnity in respect of legal liability for sums payable as compensation for bodily injury or disease "suffered" by any employee "when such injury or disease arises out of and in the course of employment by the Insured and is sustained or contracted during the currency of this Policy".
The issue in the Trigger litigation was whether these wordings responded by reference to the date of exposure to asbestos dust or to the date when the onset of mesothelioma or any other long-term disease developed or manifested itself. In determining this issue, this court found assistance as to the scope, purpose and proper interpretation of each of the insurances in a study of its language, read in its entirety (paras 19 and 41). It relied on the wordings' assumption that the course of employment and the sustaining of injury would be contemporaneous (para 20) and that there would be a close link between the risks attaching to the employment or work undertaken in the insurance period and the risks which the insurers, for a premium calculated by reference to the nature of such employment during such period, agreed to insure (paras 21-23). It also relied on the improbability that insurers would offer or sell cover in respect of risks attaching to ancient, as opposed to current, employment or activities (para 24) or on a basis that would leave it open to insurers to refuse further cover once it became apparent that such employment or activities were likely to produce claims (paras 24-25).
In the upshot, all the insurance wordings considered in the Trigger litigation were held to operate on a similar basis. Whether the wording referred to a disease "contracted" or an injury or disease "sustained", the reference was to be taken as being to the date when mesothelioma was caused or initiated by exposure, even though it only developed or manifested itself long afterwards (paras 49-51). In respect of the limited number of the insurances with which the court was concerned which post-dated the Employers' Liability (Compulsory Insurance) Act 1969, the court also regarded its conclusions on interpretation as the only conclusions consistent with the employers' duty to carry insurance under that Act. The Midland policy wording in issue on this appeal is expressly on a causation basis, and the risks undertaken are closely tied to the period of insurance.
A second, yet more fundamental, point arose during the course of the appeal in Trigger. If causation grounded liability under the insurance wordings, could causation be shown to exist, bearing in mind that the special rule established by Fairchild, Barker and Sienkiewicz derives from the impossibility of proving as a matter of probability that any particular exposure has led or contributed to the occurrence of mesothelioma in any particular case? The rival possibilities examined in Trigger were that
the special rule involves a legal inference or fiction that, despite the evidential impossibility, the particular exposure has as a matter of probability caused or contributed to the occurrence of the mesothelioma or
the special rule involves a "weak" or "broad" view of the causal requirements or link satisfied in the case of mesothelioma by proof of exposure to asbestos dust or, both these possibilities failing,
the employer's liability under the special rule arises not for, or because he has in any sense caused, the mesothelioma, but on the contrary for, and because of, his creation of the risk of causing the disease.
On this last basis no relevant causation would have existed sufficient to trigger the insurance wordings in Trigger, since they required causation of a disease, not causation of a risk.
In Trigger, none of the members of the court accepted possibility A: see eg paras 59 and 71-74 in my judgment and para 131 per Lord Phillips. But Lord Phillips went on, after analysing Barker, to accept possibility C. Dissenting, he held that employers could not pass on to their insurers the liability which they had incurred under the special rule, and he refused to engage in any redefinition of that special rule to render insurers liable: paras 133-134. The majority on the other hand accepted the employers' case that insurances underwritten on a causation basis must respond in circumstances where employers incur liability for mesothelioma under the special rule because they have exposed the victim to asbestos dust during the relevant insurance period. In my judgment, with which Lord Kerr, Lord Clarke and Lord Dyson agreed, this was explained by reference to possibility B: the ordinary requirements of causation (proof on a balance of probability) were modified as between the victim and person responsible, so as to make the latter liable for the mesothelioma because of the risk of sustaining mesothelioma to which the victim had been exposed during the relevant period: see paras 66 and 73.
Further, and importantly, the majority also held that a liability insurer covering the person responsible on a causation basis must accept the development of, and the increase of employers' liability resulting from, the special rule "within the limits of the relevant insurance and insurance period": see paras 66 (end), 69-70 and 73-74. If causation is given a weak or broad meaning as against the person tortiously responsible, the same weak or broad meaning should be treated as carrying through into a liability insurance covering an insured on a causation basis. However, Trigger was not directly concerned with, and did not examine, the situation or the consequences where a person responsible for exposing a mesothelioma victim to asbestos dust has an insurance covering only part of the period of that exposure. That is the situation which gives rise to the present appeal.
If one puts on one side the fact that exposure continued for a further 21 years, Trigger is direct authority that the Midland policy must respond to liability for mesothelioma incurred by IEG under the special rule as a result of GGLCL's tortious exposure of Mr Carré throughout the six-year period of the Midland insurances. The policy period is fundamental under any liability policy, as the reasoning in Trigger summarised in para 18 above itself indicates. But, under Trigger, the sufficient "weak" or "broad" cause which grounds liability for any subsequently incurred mesothelioma occurs within the policy period, and that is sufficient. Zurich has at all times accepted that, if Mr Carré was, as a result of being exposed to asbestos dust during the six years for which Midland insured GGLCL, entitled to the full compensation payment of £250,000 plus costs which he received from IEG, then the policy wording on its face requires Zurich to answer in full notwithstanding that he was also exposed to asbestos dust during a further 21 years: see its written case before this court on the present appeal, para 4.4.
However, the policy and its wording only govern the parties' relationship in and with respect to the policy periods and risks arising during such periods. The special rule recognised in Fairchild as modified by the 2006 Act has the unique effect of requiring Zurich to respond potentially under its policy wordings to liabilities incurred by GGLCL/IEG which are:
attributable to the mere risk that GGLCL's conduct during the Midland insurance period led or contributed to Mr Carré incurring mesothelioma, but also
equally easily, or proportionately much more easily, attributable to GGLCL's conduct wholly outside the scope and period of the Midland insurance.
Zurich's case is that, since GGLCL's conduct within (b) was wholly independent of and outside the scope of the Midland insurance and Midland insurance period, there is no reason why it cannot be recognised as giving rise to obligations as between Zurich and IEG, no inconsistency with the Midland insurance in recognising that such obligations may result from such conduct, and every reason in justice why this should be recognised.
The first main question on this appeal is whether Barker remains good common law, not in the United Kingdom, where it has been superseded by the 2006 Act, but in Guernsey where no such statute exists. I do not understand there to be any issue that, if Barker remains good common law, then IEG's liability in respect of the six years of Midland cover was and is for a proportionate part (22.08%) of the full compensation which IEG in fact paid. If Mr Carré had only been able to show six years of exposure with GGLCL, but a further 21 years exposure elsewhere, he could not have claimed more than 22.08% of his total loss from IEG. Equally IEG cannot now claim from Zurich more than the same proportion (22.08%) of the whole compensation paid which it can properly attribute to the six years of the Midland insurance.
This is the corollary of the fundamental principle of indemnity, which governs liability insurance. This principle was articulated long ago in Godin v London Assurance Co (1758) 1 Burr 489, a case in which the defendant insurers were contending that because there had been double insurance they ought only to have to pay half the loss, although neither insurer had as yet paid any sum. Lord Mansfield, in giving the judgment of the court upholding a verdict for the whole loss in these circumstances, observed (p 492):
Before the introduction of wagering policies, it was, upon principles of convenience, very wisely established, that a man should not recover more than he had lost. Insurance was considered as an indemnity only, in case of a loss: and therefore the satisfaction ought not to exceed the loss. ....
If the insured is to receive but one satisfaction, natural justice says that the several insurers shall all of them contribute pro rata, to satisfy that loss against which they have all insured ....
Where a man makes a double insurance of the same thing, in such a manner that he can clearly recover, against several insurers in distinct policies, a double satisfaction, 'the law certainly says that he ought not to recover doubly for the same loss, but be content with one single satisfaction for it' .... And if the whole should be recovered from one, he ought to stand in the place of the insured, to receive contribution from the other, who was equally liable to pay the whole.
In IEG's submission, Barker is fatally undermined by the Compensation Act 2006 and/or the decision in Trigger. IEG points out that section 16(3) of the 2006 Act provides that "Section 3 shall be treated as having always had effect", and suggests that the Act was in section 3 declaring what the common law "has always been". I do not accept that. Section 16 is a section dealing with "Commencement", and the 2006 Act was clearly passed to change a common law rule expounded in Barker. It is true that the 2006 Act leads to a result which the common law might itself have accepted as appropriate: Trigger, para 70. But the common law did not do so, and the reasons why it did not are in my view both coherent and understandable. They are set out extensively in Barker, and I need not repeat them here. What the House did in Barker was to treat proportionality as a concomitant of the exceptional liability which derives from the special rule in Fairchild and which the House was, on that basis, prepared in Barker to extend to situations beyond those which Fairchild had held covered by it. The United Kingdom Parliament's reaction was its right, but does not alter the common law position apart from statute, or have any necessary effect in jurisdictions where the common law position has not been statutorily modified.
In Trigger the court looked closely at Barker, and saw itself as applying what Barker established: see paras 63-66 and 72 of my judgment. At para 66 I noted that the speeches of "Lord Hoffmann, Baroness Hale and (possibly) Lord Walker in Barker" all viewed an employer's legal responsibility as "based on a 'weak' or 'broad' view of the 'causal requirements' or 'causal link' appropriate in the particular context to ground liability for the mesothelioma". To those references can be added that Lord Scott at para 50 and Lord Walker at para 103 in Barker both expressly agreed with Lord Hoffmann's reasons for allowing the appeals on the issue of apportionment. Further, there was in Trigger no issue about or challenge to the correctness of Barker. In these circumstances, it would on the face of it be surprising to find that Trigger had consigned that decision to history.
IEG submits that, under Trigger, an employer shown to have significantly exposed a mesothelioma victim to asbestos dust is liable for having caused (in a weak or broad sense) the mesothelioma, and that anyone who is liable for causing a disease must answer for the whole loss resulting from that disease. In the Court of Appeal,  EWCA Civ 39, that submission was accepted by Toulson LJ at paras 30-31 and Aikens LJ at paras 53-55. No doubt the submission is (subject to conventional limitations like remoteness and mitigation) generally correct in a conventional case where causation must exist in its ordinary sense of conduct which "on a balance of probability brought about or contributed to" the disease. But causation in a "weak" or "broad" sense is unconventional. Barker, as analysed in Trigger, accepted causation in this weak or broad sense and nonetheless held an employer's responsibility to be proportionate to that part for which that employer was responsible of the victim's total exposure to asbestos dust. Trigger cannot therefore be said to affect or undermine the reasoning or decision in Barker.
The argument that insists that a conventional approach to the measure of damages must apply in a context where liability is imposed on an unconventional basis was rejected by Baroness Hale in her judgment in Barker. The relevant passages are worth quoting at length:
This reasoning remains in my view convincing at common law. In the United Kingdom, Parliament has, as is its right, taken a different view of the equities as between a person responsible and a victim of mesothelioma. That in turn gives rise to further problems of equity in relation to other, indirectly affected persons under the second main question on this appeal. But for the reasons I have given, neither the 2006 Act nor Trigger is inconsistent with or undermines the decision in Barker. For completeness, I record that Mr Antonio Bueno QC representing IEG expressly disclaimed any intention to invite the court to overrule Barker on this appeal. That, he frankly said, would bring in other considerations, and he said that IEG's case was that it has already become history as a result of Trigger. However, Mr Patrick Limb QC, also representing IEG, did at times appear to be inviting the court to address and overrule Barker head-on. In my view, that latter invitation is not open to IEG, and further Barker has not been overruled by Trigger, and remains as part of the common law of England, which we are to take to be the same as the common law of Guernsey.
The "all sums" policy construction issue
The written cases identify under this head a secondary issue, concerning the extent of Zurich's liability to indemnify IEG. It arises from observations made by Aikens LJ, with whose judgment Kay LJ agreed. After concluding in para 53 that the majority in Trigger had grounded liability on a weak or broad causal link within the policy period, he went on in para 54 to say:
Once that causal requirement is fulfilled, then the employer will have proved that the mesothelioma (the disease) was 'caused during any period of insurance'. It follows from the policy wording that the insurer is then liable to indemnify IEG for 'all sums for which the Insured shall be liable in respect of any claim for damages for .... such disease' (my emphasis). In other words, Zurich will be liable to indemnify IEG for the whole of the damages paid out by IEG in respect of Mr Carré's claim for damages for contracting mesothelioma, not just a proportion worked out by reference to the period during which IEG was covered by policies for which Zurich is responsible.
The reference to "all sums" comes from the primary insuring clause set out in para 13 above. As I understand Zurich's written case, raising the secondary issue on the basis of this paragraph, Zurich was concerned that Aikens LJ was or might be suggesting that, even if Barker stood and applied (so that IEG's liability towards Mr Carré would have been limited to a proportion of his total loss, had IEG only exposed him for six out of the total of 27 years), IEG, having actually exposed him for the total 27 years though only insured with Midland for six of such years, might under the "all sums" provision in the insurance be entitled to recover from Zurich in respect of Mr Carré's total loss attributable to the 27 years. Any such argument would be clearly contrary to the fundamental principle of indemnity mentioned in para 26 above. Further, as I understand it, no such argument is in fact advanced by IEG.
On the other hand, IEG appears to have understood Zurich to contend that, even if Barker had become past history (so that IEG was liable in full to Mr Carré for the whole of his loss resulting from mesothelioma, whatever the period for which it had exposed him compared with other periods of exposure), Zurich should under the Midland policies only answer for a rateable proportion of such total loss, viz 22.08%. For reasons indicated in para 23 above, I do not understand Zurich to make any such case. Zurich accepts that, if Barker no longer represents the common law, and IEG became liable for Mr Carré's full loss simply because he was exposed to asbestos dust during the six-year Midland insurance period, then Zurich must on the face of the Midland policy wordings answer under the insurance, even though he was also exposed during 21 other years.
In these circumstances, I need say no more on the secondary issue. It follows that the appeal must succeed as regards the compensation and interest paid by IEG to Mr Carré, because Barker continues to represent the common law position which applies in Guernsey. The Court of Appeal was wrong to set aside Cooke J's judgment, which should be restored, on this aspect.
That leaves the defence costs totalling £13,151.60 which IEG incurred in defending Mr Carré's claim based on exposure to asbestos dust over the full 27 years of his employment with GGLCL. Zurich submits that these costs should be pro-rated on the same 22.08% basis. An important parallel, though not in my view identical, issue would arise in any case where the Compensation Act 2006 applies, making a responsible person liable for the whole damage suffered by a mesothelioma victim, regardless of the length and volume of his other exposures to asbestos dust.
As regards defence costs, IEG relies upon reasoning adopted by the Privy Council in New Zealand Forest Products Ltd v New Zealand Insurance Co Ltd  1 WLR 1237. There proceedings were instituted on the basis of five causes of action against a company and its director, whose costs were both covered by an insurance policy, and in the case of one of the causes of action against a third person not so covered. All the defendants were represented by the same lawyers. It was common ground that costs not relating in any way to the insured director's defence would not be covered, while costs exclusively related to the insured director's defence would be covered. The issue which arose was as to defence costs which related at one and the same time to the defence both of the claim against the insured director and of the claim against the uninsured third person. The courts below took the view that there should be an apportionment. The Privy Council reached a different view, as a matter, it said, of construction of the relevant insurance. This covered "all loss .... which such officer has become legally obligated to pay on account of any claim made against him .... for a wrongful act". As this wording would cover the whole costs incurred in the defence where the insured officer was the sole defendant, the Board saw no reason why it should not cover them all, where some of them related also to the defence of an uninsured co-defendant. There was no question of the costs relating to any period other than that insured, and, importantly, they arose on a conventional causative basis – because of a claim against the director for a wrongful act.
Two points are notable in relation to the defence costs which IEG seeks to recover from Zurich. First, there is nothing to suggest that these would have been any less had the claim against IEG been confined to the six-year period covered by the Midland policies. Second, and more significantly, the defence costs which IEG incurred were "incurred with the consent of the Company in defending any such claim for damages" within the meaning of the second sentence of the main insuring clause set out in para 13 above. That is, they were incurred by IEG in defending a claim by a former employee for damages for injury or disease which he was caused to sustain while employed during the periods of insurance provided by Midland. The claim against IEG could, under the special rule in Fairchild, be pursued on the basis that GGLCL had done no more than expose Mr Carré to a risk of suffering mesothelioma. In the light of Trigger the first sentence of the main insuring clause set out in para 13 above covers liability arising on this basis. But IEG's liability for and right to recover defence costs does not arise under the special rule, or on the basis that Mr Carré was exposed to any risk. It is not recoverable under the first, but under the second sentence of the main insuring clause. Under the second sentence, it is recoverable on the conventional basis that IEG can prove that it incurred (as a matter of fact or probability) actual financial loss in the circumstances covered by that sentence. This distinction is important. Once it is shown that an insured has on a conventional basis incurred defence costs which are covered on the face of the policy wording, there is, as the New Zealand Forest case shows, no reason to construe the wording as requiring some diminution in the insured's recovery, merely because the defence costs so incurred also benefitted some other uninsured defendant.
The special rule - analysis
However, liability arising under the special rule in Fairchild on the exceptional basis of a weak or broad causal link consisting of exposure to a risk is different. As the volume of case law indicates and not surprisingly, it has proved difficult to work through the implications of the special rule in Fairchild. But, having, for wholly understandable reasons, gone down the Fairchild route, the common law must, in my view, face up to the consequences, if necessary by further innovation. That is so, even if some of the problems arise from Parliament's intervention by the 2006 Act. As already observed, the common law might itself have taken the same approach as that Act, though it did not in fact do so. Had it done, it would certainly have had to work out the common law implications. Parliament's intervention does not release the courts from their role of working out the common law implications of a special rule which remains essentially common law based, although subject to the modification introduced by the 2006 Act. Trigger may be regarded as an instance of performance of this role.
It is worth emphasising how novel the situation created by Fairchild and Trigger is in an insurance context. When the present liability insurances were placed, what Hobhouse LJ said about the fundamental nature of the insurance period in the context of a property reinsurance in Municipal Mutual Insurance Ltd v Sea Insurance Co Ltd  Lloyd's Rep IR 421, 435-436 would have been just as true of them:
The judge came to the surprising conclusion that each reinsurance contract covered liability in respect of physical loss or damage whether or not it occurred during the period covered by the reinsurance contract and he went on expressly to contemplate that the same liability for the same physical loss or damage might be covered under a number of separate contracts of reinsurance covering different periods. This is a startling result and I am aware of no justification for it. When the relevant cover is placed on a time basis, the stated period of time is fundamental and must be given effect to. It is for that period of risk that the premium payable is assessed. This is so whether the cover is defined as in the present case by reference to when the physical loss or damage occurred, or by reference to when a liability was incurred or a claim made. Contracts of insurance (including reinsurance) are or can be sophisticated instruments containing a wide variety of provisions, but the definition of the period of cover is basic and clear.
In short, insurance would have been and was placed on the basis that a particular liability or loss would fall into one, not a series of separate periods. If an insured wanted complete cover, it would have to maintain it for all such periods. The relevant period would also be ascertained by objective criteria, which meant that insureds could not select it at will or to obtain the advantage of the cover most favourable from their viewpoint. Thus:
Under a liability insurance where the trigger is causation in its traditional sense based on probability, no problem exists about allocating tortious liability to one and only one policy period.
Under a claims made policy, claims must be notified and will attach at latest when they arise, while specific clauses dealing with the notification of circumstances likely to give rise to a claim may attach a claim to an earlier policy than that during which it actually arises.
An insured may, for one reason or another, have double insurance. In that context, it may elect which to invoke, but well-established principles exist for the two insurers to share liability equally up to the common limit.
An insured may also agree to carry an excess or franchise, in which case it will have to bear that amount before looking to its insurer, and will as a self-insurer rank last in any recoveries made by way of subrogation from any third party: Lord Napier and Ettrick v Hunter  AC 713.
Against this background, the present appeal illustrates some of the problems, arising from the special principles recognised and applied in Fairchild and Trigger, at the level of relationships between persons responsible and their insurers:
An employer, manufacturer or other person may well have been responsible for exposing employees and others to asbestos dust over many years.
For many years, he may have decided not to insure, or been unable to obtain cover which he regarded as acceptable.
During some years or as from some date, he may have decided to take out insurance. Employers should have done so, once employers' liability insurance became compulsory, that is in and after 1972 in England, Wales and Scotland, 1975 in Northern Ireland and 1993 in Guernsey.
Even when insurance was taken out, it may have been taken out on a claims made, rather than causation, basis; even after employers' liability insurance was compulsorily required, it may have been taken out on this basis under what Trigger indicates to have been the misapprehension that this form of insurance would satisfy that requirement.
Where insurance was taken out:
the employer, manufacturer or other person may not have fully appreciated the long-term nature of the risks covered and may have failed to keep records from which the insurance can now be traced; or
the insurer may have become insolvent.
Where a person responsible for exposing a victim of mesothelioma to asbestos dust over a period of years has had liability insurance with the same insurer over the whole period, no problem arises. But frequently this will not have been, or cannot be shown to have been, the case, and the potential anomalies then arise. On IEG's case, the special rule, as modified by the 2006 Act and explained in Trigger, allows a person responsible for exposure to select any year during which he can show that he carried liability insurance and to pass the whole of any liability for such exposure to the liability insurer on risk in that year, without regard to other periods of exposure.
If matters stop there, and the insurer ends up carrying the whole liability, the anomalies are self-evident:
It is contrary to principle for insurance to operate on a basis which allows an insured to select the period and policy to which a loss attaches. This is elementary. If insureds could select against insurers in this way, the risks undertaken by insurers would be entirely unpredictable.
It is anomalous for a liability insurance underwritten for a premium covering losses arising from risks created during its particular period to cover losses about which all that can be said is that they arise from risks extending over a much longer period, in respect of which no premium has, or could have, been assessed or received by the insurer.
An insured is able to ignore long periods in respect of which he himself has chosen not to insure, or has not kept any record of any insurance which he may have taken out, or has chosen to entrust his insurance to an insurer who has become insolvent.
An insured has no incentive to take out or maintain continuous insurance cover. On the contrary, it is sufficient to take out one year's cover, or even to arrange to be held covered for only one day, during whatever happens subsequently to prove to have been the overall exposure period – whether this is done at the very start of the overall exposure period, or later after many decades of exposure, perhaps due to a sudden appreciation of the virtues of insurance under the special rule.
In each case the anomaly arises because, without more, the analysis identified in the last sentence of para 42 above fails to adjust to the unique situation which arises from the principles recognised in Fairchild and Trigger. There are various responses that the law might have taken to such anomalies. One is that which Lord Phillips took in Trigger, viz that the insurance only answered for liability proved as a matter of probability to have resulted from asbestos exposure in the insurance period. Lord Phillips' approach can be viewed as entirely conventional, in the sense that it reflected the traditional view that, under a liability policy like the Midland's, the concept of causation looked to the "proximate" or "effective" cause, to be proved as a matter of probability. But it would have meant that no liability insurance cover existed in respect of mesothelioma.
In the light of this drastic consequence, the majority of the court in Trigger preferred a second response. It equated the concept of causation in an insurance context with the weaker or broader meaning which the courts have, to the benefit of victims, given it in tort. This was a choice rationalised in terms of the principle that a facultative liability insurance normally responds to whatever may prove to be the liability incurred by the insured. In Trigger there was no consideration of a situation in which a relevant insurance covered only part of an overall period during which the insured employer had exposed the victim to asbestos dust. But in my view the reasoning in Trigger binds this court to hold that the mesothelioma is caused - in the sense that it results from exposure which existed - in each and every period of any overall period of exposure. The fact that a victim or an insured only relies on one period of exposure does not alter the legal position, that it can equally be said to have been caused in every other period of exposure. This is because, as a matter of law, exposure connotes causation, in both tort and tort liability insurance law. It is the anomalies resulting from that conclusion which the court must now resolve, accepting but building on its own prior jurisprudence.
Lord Sumption's judgment argues for a third response. He agrees that the respondents' case involves all the anomalies already identified. But he considers that they can and can only be met by interpreting the insurance policy wording in a way which none of the parties or interveners before the court has suggested. He regards it as consistent with the decision in Trigger to say that an insurer, who only covers part of the total period for which the insured exposed the victim, is only liable for a corresponding part of the insured's liability to the victim. In my view, this is inconsistent with Trigger. Once one accepts that causation equates with exposure, in tort and tort liability insurance law, there is no going back on this conclusion simply because there was exposure by the insured of the victim both within and outside the relevant insurance period.
More specifically, Lord Sumption suggests that the insurer "must still show that the occurrence fell within the chronological limits of the policy" (para 156). But that raises the question: what is here meant by the occurrence for which the employer is liable? It cannot be the disease itself, which can and does occur decades later. If it is the incident which causes the disease, then, as Lord Sumption himself recognises (para 157), it is each and every, or any, negligent exposure to asbestos involving a contribution by the employer to the risk of the victim sustaining mesothelioma that constitutes causation for the purposes of a liability insurance like the present. Any such exposure can be relied on as causing the mesothelioma and making the employer fully liable for the victim's loss, and any such exposure occurring during any policy period will on a like basis mean that the insurer incurs full liability.
Lord Sumption seeks to avoid this conclusion, acknowledging that it "makes some sense as between successive employers who are guilty of a continuous tort", but saying that "the same logic cannot be applied as between successive insurers" (para 157). But the primary question is not as between two insurers, it is as between the employer and any insurer against which he claims; and there is also nothing illogical about a conclusion that each of successive insurers is potentially liable in full, with rights of contribution inter se.
Lord Sumption also advances a broader argument, that it is "conceptually impossible" for mesothelioma to be "successively caused in every period of exposure", because "Mesothelioma is caused only once", or, as he later puts it, that it is "not conceptually possible for an insurer to be liable on the footing that the disease was actually caused in every year" (para 158). But this moves the terminological goalposts, by reverting to traditional notions of causation - those applicable outside the Fairchild enclave, where proof on the balance of probabilities is traditionally required. Within that enclave, the House accepted in Fairchild that it was necessary to adopt a "weak" notion of causation, in order to protect victims, and in Trigger the Supreme Court held that this weak notion of causation carries through into an insurance context. On this basis, loss is caused for the purposes of tort and liability insurance contracts like the present in any and every period when the victim was exposed to asbestos and so to the risk of mesothelioma. Lord Sumption's broader argument is therefore incorrect. Moreover, if it had any force, it is not obvious why it would not apply equally to tort and so preclude one negligent employer from seeking contribution from another – yet that is expressly provided for by the 2006 Act.
Lord Sumption states further that Trigger "cannot be applied without modification when the question is how much of the loss is attributable to particular years", and continues by saying the "the rational response of the law" is to prorate the whole loss "between every policy year during which the insured employer exposed the victim to asbestos" (para 160). Lord Sumption correctly points out that it is "only when one aggregates every successive period that the chances add up to 100%" (para 158). But this means, logically, that, if (as Lord Sumption maintains) any insurance can only answer pro rata for exposure or risk occurring during the insurance policy period, the relevant pro-rating must be by reference to the total exposure of the victim from all employers and sources. The total period of exposure by the particular employer is in this context irrelevant, since the insurance wording says nothing about it and the chances of sustaining mesothelioma do not correspond with it when there are other sources of exposure.
For all these reasons, I cannot therefore accept Lord Sumption's approach. An insurer, whether for the whole or part of the period for which the insured employer has negligently exposed the victim to asbestos, is on the face of it liable for the victim's full loss. However, I agree that the analysis cannot stop here. The court is faced with an unprecedented situation, arising from its own decisions affecting both tort and insurance law. A principled solution must be found, even if it involves striking new ground. The courts cannot simply step back from an issue which is of their own making, by which I do not mean to suggest that it was in any way wrong for the courts, from Fairchild onwards, to have been solicitous of the needs of both victims and insureds. But by introducing into tort and liability insurance law an entirely novel form of causation in Trigger, the courts have made it incumbent on themselves to reach a solution representing a fair balance of the interests of victims, insureds and insurers.
In my view, the law has existing tools which can be adapted to meet this unique situation. The concepts of co-insurance and self-insurance are both at hand. Co-insurance is relevant in so far as the insured has other insurance to which it could also have resorted on the basis that it had also exposed the victim during the period of that insurance. Self-insurance is relevant, because an insured who has not (i) taken out or (ii) kept records of or (iii) been able to recover under such other insurance must be regarded as being its own insurer in respect of the period in question for which it has no cover. A sensible overall result is only achieved if an insurer held liable under a policy like the Midland policy is able to have recourse for an appropriate proportion of its liability to any co-insurers and to the insured as a self-insurer in respect of periods of exposure of the victim by the insured for which the insurer has not covered the insured.
There are of course difficulties about drawing a direct analogy between the present situation and conventional situations in which the concepts of co-insurance and self-insurance have previously been deployed. But the court would be abrogating its role to achieve a just solution consistently with what any sensible commercial party would have contemplated if it does not adapt and develop conventional principles to meet an unconventional, indeed unique, challenge. I see no barrier at all to this in the fact that the parties did not directly contemplate or cater for it in the insurance policy between them. It is equally clear that they did not contemplate or cater for the principles imposed upon them by the decisions in Fairchild and Trigger. To carry the declaratory theory to the point of asserting the contrary would be absurd. "To say that [judges] never change the law is a fiction and to base any practical decision upon such a fiction would indeed be abstract juridical correctitude", rather it is the case that a judicial decision can change the law retrospectively: Deutsche Morgan Grenfell Group plc v IRC  UKHL 49,  1 AC 558, para 23, per Lord Hoffmann. Equally, the fact that the parties may not have contemplated or made specific provisions about co-insurance and self-insurance on the basis of those decisions is no obstacle to the court doing so. To say (as Lord Sumption does: para 185) that there has here been a "contractual allocation of risks" which precludes the court taking steps to avoid evident absurdity which no contracting party can sensibly have contemplated or intended appears to me unrealistic. There was a contractual allocation of risks on the basis and in respect of exposure by the insured during the policy period. But if there was further exposure by the insured, outside the policy period, there is no reason why the insurer should not have proportionate recourse against anyone who can be seen to carry the risks attaching to such further exposure. There is nothing inconsistent with the agreed insurance or its period in deriving from a consideration of circumstances outside that insurance and its period a right to contribution in respect of the loss incurred in the first instance by the insurer: see further paras 67-73 below.
In summary, so long as the insured has insured itself for the whole period for which it exposes the victim, the insurer can ask for no more, and must, as Trigger decides, bear the whole of any liability which the insured incurs. The palliative in this latter situation is of course that an employer/insured will have a right to contribution under the Civil Liability (Contribution) Act 1978 against any other person who was, negligently or in breach of duty, responsible for exposing the victim to asbestos, and its insurer will, after meeting the insurance claim, be subrogated to this right to contribution against the other responsible source of exposure.
The anomalies therefore only arise when the insured has exposed the victim for a longer period than that for which it is covered by the insurer to which it chooses to look for indemnity. The anomalies are, as stated, not capable of being addressed by any of the law's existing tools for dealing with more conventional problems. As observed in Trigger, paras 67-68, section 3(3) of the 2006 Act preserves the conventional tools, found in the Law Reform (Contributory Negligence) Act 1945 and the Civil Liability (Contribution) Act 1978, for dealing with the conventional problems of contributory fault (by a victim of mesothelioma) and concurrent liability in respect of the same damage (between different persons responsible for exposing a victim of mesothelioma to asbestos dust, whether over the same or different periods). Persons responsible for exposing victims to asbestos dust are thus appropriately protected. Their protection is carried one step further by section 3(7), which enables the Treasury to make regulations for the provision of compensation to a responsible person who is unable to obtain contribution under the 1978 Act, because an insurer of such person is or is likely to be unable to satisfy the claim for a contribution. By definition in section 3(10), the reference in section 3(7) to a responsible person also includes an insurer of such a person. That is the only respect in which the Act addresses the interests of an insurer, as a corollary of the rules relating to contribution between persons responsible. The Act is not concerned with, and does not address, the effects on insurers or as between persons responsible and insurers of the special rule as modified by section 3(1) and (2). It is for the courts to work out these effects at that level.
So far as appears, during the overall period of 27 years during which it exposed Mr Carré to asbestos dust, GGLCL only had insurance for two periods, six years with Midland and two years with Excess. Not surprisingly, no previous authority exists regarding the relationship between Midland and the Excess in the present context. Zurich could not have any sort of subrogation right against Excess, since, if Zurich is liable for IEG's full loss, IEG can have no further claim for indemnity against Excess. Further, no-one would ordinarily regard insurances for different insurance periods as double insurance. The reason for taking out or renewing a fresh annual policy during a fresh year is, on the contrary, the common-sense truism that, unless one does so, one will be uninsured.
The concept of double insurance, as hitherto recognised in English law, was explained by Mr Gavin Kealey QC, sitting as a deputy judge of the Commercial Court, in National Farmers Union Mutual Insurance Society Ltd v HSBC Insurance (UK) Ltd  EWHC 773 (Comm),  1 CLC 557, para 15:
Double insurance arises where the same party is insured with two (or more) insurers in respect of the same interest on the same subject-matter against the same risks. If a loss by a peril insured against occurs, the general rule is that, subject to any particular modifying terms and to the limits of indemnity provided under each insurance contract, the insured may recover for the whole of the loss from either insurer. Upon such indemnity being paid to the insured by either one of the two insurers, that insurer is, in general, entitled to recover a contribution from the other. To quote from Lord Woolf in Eagle Star Insurance Co Ltd v Provincial Insurance plc  1 AC 130, 138:
Previous first instance statements to like effect that double insurance requires the same insured to be covered in respect of the same property against the same risks can be found in Petrofina (UK) Ltd v Magnaload Ltd  1 QB 127, 140F-G per Lloyd J, followed in Wimpey Construction UK Ltd v D V Poole  2 Lloyd's Rep 499, 516 (Webster J).
The insurances taken out with Midland and Excess would not satisfy this concept. In particular, they were not "on the same interest" or "against the same risks". Nor does the special rule recognised in Fairchild as modified by the 2006 Act make them so. The Excess policies covered injury or disease caused by the risk of exposure occurring in 1979 and 1980, whereas the Midland policies covered injury or disease caused by the risk of exposure occurring in the years 1983 to 1988. If one accepts the definition accepted by Gavin Kealey QC, then Eady J was right in Phillips v Syndicate 992 Gunner  EWHC 1084 (QB),  Lloyd's Rep IR 426, para 22, to reject the submission that two or more successive policies of insurance could be regarded as covering the same liability towards a victim of mesothelioma for the purposes of a condition in the relevant policy in that case addressing situations of double insurance.
However, Australian appellate courts have been willing to contemplate a more relaxed view of double insurance, to address situations where the same liability is ultimately covered albeit by different routes and involving different insureds: AMP Workers Compensation Services (NSW) Ltd v QBE Insurance Ltd  NSWCA 267, (2001) 53 NSWLR 35, Zurich Australian Insurance Ltd v GIO General Ltd  NSWCA 47. They have in a series of cases also emphasised the root principles of equity and justice which lie behind the law's recognition of rights of contribution: see Albion Insurance Co Ltd v Government Insurance Office of New South Wales  HCA 55, (1969) 121 CLR 342, esp per Kitto J. Kitto J's judgment has been cited with approval in Burke v LFOT Pty Ltd  HCA 17, 187 ALR 612, Zurich Australian Insurance Ltd v GIO General Ltd  NSWCA 47 and, in a brief extract, by Lloyd LJ in the Legal and General case: see para 57 above. As Burke v LFOT Pty Ltd shows, Australian courts have carried the doctrine of equitable contribution far enough for it to provide as a matter of common law a right of contribution in respect of any "common obligation", with a breadth and flexibility similar to that statutorily available in England under, now, the Civil Liability (Contribution) Act 1978, and, previously (though only as between tortfeasors), the Law Reform (Married Women and Joint Tortfeasors) Act 1935. In Burke itself the claim for contribution was only refused because it was inequitable in the particular circumstances to award any contribution against a negligent solicitor in favour of LFOT which had engaged in misleading and deceptive conduct in breach of a statutory obligation.
Contribution is, ultimately, a principle based on "natural justice", as Lord Mansfield said in Godin's case, cited in para 26 above. A similar justification was given by Lord Chief Baron Eyre in Dering v Earl of Winchelsea (1787) 1 Cox Eq 318, 321, for recognising a right of contribution between sureties who had each accepted "distinct and separate obligations" and were not therefore in any contractual relationship with each other:
If we take a view of the cases both in law and equity, we shall find that contribution is bottomed and fixed on general principles of justice, and does not spring from contract; though contract may qualify it .... [I]n equali jure the law requires equality; one shall not bear the burthen in ease of the rest, and the law is grounded in great equity.
A similar approach is not out of place in a context where the law has developed new liabilities to redress perceived injustice. Consistently with this, Charles Mitchell, in The Law of Contribution and Reimbursement (2003) notes, para 4.14, that
The categories of claimant by whom contribution can be claimed at common law or in equity are not closed ....
Mitchell cites in this connection, inter alia, Burke v LFOT Pty Ltd. Meagher, Gummow and Lehane in Equity, Doctrines and Remedies (4th ed) (2002), para 10-020, also note the influence on the principles governing contribution of the equitable maxim that "equality is equity" and the doctrine of marshalling, whereby:
as between several interested parties it should not rest with the creditor by his selection of remedies open to him to determine where ultimately the burden was to fall.
The Legal and General case, referred to in the passage cited by Mr Gavin Kealey QC (see para 57 above), illustrates the latter principle. There the insured's choice to proceed against insurer A under one policy meant that no notice of claim was given to insurer B under the other policy within 14 days as required by its terms. It was held by the majority (Lloyd and Nourse LJJ) that the absence of any such notice did not defeat the claim for contribution based on double insurance. Again, the reasoning is founded on broad principles of equity:
Since the assured could have gone against B, had he chosen to do so, .... the burden as between A and B should be shared equally. It would be inequitable for either of the insurers to receive the benefit of the premium without being liable for their share of the loss.
[per Lloyd LJ, p 892C-D]
There being no contract between the two insurers, the right of contribution depends, and can only depend, on an equity which requires someone who has taken the benefit of a premium to share the burden of meeting the claim. Why should that equity be displaced simply because the assured has failed to give the notice which is necessary to make the other insurer liable to him? .... As between the two insurers the basis of the equity is unimpaired. He who has received a benefit ought to bear his due proportion of the burden.
[per Nourse LJ, at p 898B-D]
In my view, the principles recognised and applied in Fairchild and Trigger do require a broad equitable approach to be taken to contribution, to meet the unique anomalies to which they give rise. I note that this solution is also advocated by Professors Merkin and Steele in their recent study on Insurance and the Law of Obligations (2013) (OUP), p 378. If a broad equitable approach is taken in the present unique circumstances, then it should no doubt also be possible in the present context to overcome the normal presumption with double insurance that loss should be shared equally. Contribution between insurers covering liability on the basis of exposure should take account of differing lengths of insured exposure. Conventional rules need to be adapted to meet unconventional problems arising from the principles recognised and applied in Fairchild and Trigger.
An alternative possible avenue of recourse against a "double" insurer in respect of policy liabilities based on breach of an obligation assumed on or after 1 January 1979 is the Civil Liability (Contribution) Act 1978. The argument would be that both insurers are liable for "the same damage" within the meaning of section 1(1) of that Act. The possibility that the 1978 Act applies is dismissed in Colinvaux & Merkin's Insurance Contract Law, para C-0643, while Charles Mitchell in The Law of Contribution and Reimbursement, (2003), paras 4.13 and 4.43-4.44, suggests that it turns on whether liability under an indemnity insurance is regarded as "the right to be indemnified by a payment of money" or is, under a view which the author suggests that the cases favour, regarded as arising from breach of an undertaking to prevent the insured risk from materialising. It is unnecessary to resolve this difference here. It suffices to say that, if insurance contract liabilities are viewed as sounding in damages, it appears somewhat surprising if the 1978 Act could operate as an alternative statutory remedy with different effect in a case of true double insurance in respect of post-commencement liabilities.
The extension of currently recognised principles of double or co-insurance would operate only to address a very limited part of the problem. The fundamental problem remains that Zurich is, as a result of insurance policies covering only six years of exposure, liable for consequences of an exposure lasting 27 years. There can be and is no proof or likelihood that the mesothelioma resulted from fibres ingested in the six, rather than the remaining 27, years. Even assuming that Zurich has a right of contribution against the Excess, this can only be in respect of two of those 27 years, so that the two insurers would, if matters stopped there, share the consequences of 27 years of exposure by GGLCL on the basis of only eight years of insurance in the proportions of ¾ (Zurich) and ¼ (Excess). The obvious counter-balance in this situation is to treat the insured employer, GGLCL or now IEG, as a self-insurer for the remainder of the 27-year period in respect of which it can show no insurance capable of affording contribution. Nothing obliged GGLCL to maintain its liability insurance with any particular insurer. But in so far as it chose not to take out any insurance or chose to insure with another insurer, that should in common sense be at its risk. It should not be able to avoid the consequences of that risk by electing to pursue Zurich.
IEG's response to such an approach is in substance two-faceted. It submits, first, that it finds no support in existing or conventional principles of contribution, and, second, that the recognition of a right of contribution would be inconsistent with the insurance contracts made with Midland. In my opinion, neither aspect of this response is valid. As to the first, if the common law always depended on a precedent, Fairchild, or perhaps the earlier Scots House of Lords authority of McGhee v National Coal Board 1973 SC (HL) 37, should never have been decided as it was; but in any event, as I shall indicate, the concept of contribution to counter-balance a prima facie contractual right is not without precedent.
The second part of IEG's response, the suggested inconsistency between any right of contribution and the insurance contracts which Midland issued for six years, is taken up by Lord Sumption, who rules out recoupment merely because "it operates by reference to the [insurance] contract" (Lord Sumption, paras 184 and 185). The answer to this in my view is that a mere need to refer to the insurance contracts is not fatal to a recoupment claim. It does not involve contradicting or acting inconsistently with such contracts. On the contrary, it is accepting their implications, and relying on matters independent of them. It is relying on GGLCL's decision not to insure with Midland for 21 years and its decision, so far as appears, to go without insurance for up to 19 of such years. These are matters that are not touched by, and are outside, the terms and scope of the Zurich and Excess policies. They ground an equity that IEG should contribute proportionately to a loss arising from risks of exposure continuing throughout the whole 27 years.
Second, however, I do not accept that there is any absolute bright-line principle, of the sort which IEG and Lord Sumption advocate, whereby equity must always refuse to recognise a right of contribution between parties to a contract which according to its terms involves a particular result. Neither jurisprudentially nor on authority is this so. There is a general rule to that effect, but it is subject to exceptions. The position is well put by Professor Andrew Burrows QC in The Law of Restitution (3rd ed) (2011), p 88 et seq:
Professor Burrows then gives four examples of exceptions, concluding, at p 91:
The recognition and application of exceptions requires a carefully considered approach to the policies involved. A blanket rule that legal entitlement to the enrichment bars restitution does not represent the present law and would be needlessly blunt and insufficiently nuanced.
Two of Professor Burrows' examples are Roxborough v Rothmans of Pall Mall Australia Ltd (2001) 208 CLR 516 and Deutsche Morgan Grenfell Group plc v IRC  UKHL 49,  1 AC 558. The other two examples are restitution in respect of contractual obligations accruing due prior to frustration or termination for breach and restitution in respect of services rendered under an unenforceable contract. In Roxborough contracts for sale of tobacco products had been made at prices which took account of a so-called licence fee which the High Court of Australia had subsequently held to be an unlawful excise duty. The majority held that it was not possible to imply any term to cater for this unforeseen eventuality (paras 20 and 60), but that restitutionary relief could be granted in respect of the tax component of the price. Gleeson CJ, Gaudron J and Hayne J said that there was no conceptual objection to treating this as a severable part of the consideration which had failed, because it (para 21)
would not result in confusion between rights of compensation and restitution, or between enforcing a contract and claiming a right by reason of events which have occurred in relation to a contract.
Gummow J said (para 75) that
the action to recover the moneys sought by the appellants after the failure of the purpose of funding Rothmans to renew its licence may be illustrative of the gap-filling and auxiliary role of restitutionary remedies. These remedies do not let matters lie where they would fall if the carriage of risk between the parties were left entirely within the limits of their contract. Hence there is some force in the statement by Laycock [The Scope and Significance of Restitution (1989) 67 Texas Law Review 1277, 1278]:
As Gummow J went on to point out, there is authority of Lord Mansfield in the same direction. Moses v Macferlan (1760) 2 Burr 1005, a corner-stone of common law restitution, was itself a case where the plaintiff successfully reclaimed in the King's Bench money which he had been held liable to pay under various bills by the Court of Conscience, which had refused or been unable to look at the parties' wider relationship outside the bills. The plaintiff could not rely on any express or implied promise to repay. Lord Mansfield grounded the obligation simply on "the equity of the plaintiff's case" to recover back "money, which ought not in justice to be kept" (pp 1009 and 1012), and later described it as "a liberal action in the nature of a bill in equity": Clarke v Shee (1774) 1 Cowp 197, 199.
In Deutsche Morgan Grenfell the legislation governing advance corporation tax ("ACT") contravened EU law in not allowing the claimant the option to avoid or defer ACT by making a group income election. Absent any actual election by Deutsche Morgan Grenfell ("DMG"), ACT was strictly due. But it was held recoverable. Professor Burrows states, at p 91:
The best explanation for the departure from the general rule is that restitution did not here conflict with the statutory obligation because that statutory obligation was undermined by the legislature's failure, contrary to EU law, to provide a group income election for companies such as the claimants. As a matter of policy the injustice of the ultra vires exaction outweighed the point that, technically, the Revenue was legally entitled to the tax.
Unsurprisingly, in view of the obvious equity of DMG's position, the judgments take this aspect very shortly. Lord Hoffmann treated the election provisions as "purely machinery" and the real mistake as being whether DMG was liable for ACT (para 32). But Lord Hope (para 62) and it seems Lord Walker (para 143) (and Lord Scott, dissenting, paras 81-82) agreed with the trial judge (Park J) that the case fell to be analysed on the basis that, in the absence of any actual election, the tax was due. On that basis Lord Hope and Lord Walker held it recoverable, because it became due as a result of DMG's mistaken belief that it could not claim group relief by making an election. Lord Brown expressed general agreement with Lord Walker's speech (para 161-162), but elsewhere also spoke of the ACT as not due (para 172).
In the present case, applying the approach indicated by Professor Burrows, there is no policy inconsistency between recognising that the terms of the insurances underwritten by Midland make Zurich answerable in the first instance for IEG's liability towards Mr Carré and recognising an equity, based on consideration of the wider circumstances - in particular GGLCL/IEG's exposure of Mr Carré for further periods when it was not insured by Midland – requiring IEG itself to contribute towards Zurich's cost of meeting such liability.
This conclusion is also not inconsistent with the well-established principle of insurance law that an insured can recover under an insurance for a risk which is covered, even though another cause of the loss exists which is not covered, so long as that other cause is not positively excluded: see eg Wayne Tank and Pump Co Ltd v Employers Liability Assurance Corpn Ltd  QB 57. Generally, insurance law identifies a single effective, dominant or "proximate" cause, though there can be rare exceptions where there are dual effective causes as Wayne Tank illustrates. But the principle addresses a situation where more than one cause operating during the policy period can be said to have caused the insured loss in a conventional sense, that is by bringing it about or contributing to it as a matter of probability. It is not directed to the present situation where liability is based on a causal link consisting only of the risk involved in exposure, where the insured loss arises from exposure both within and outside the insurance period, and where the exposure outside the insurance period increased the risk of the insured loss occurring proportionately.
Nor is the analysis in the previous paragraphs inconsistent with the House of Lords decision in Simpson & Co v Thomson (1877) 3 App Cas 279. An insured vessel was run down and lost with all its cargo in a collision due to the negligence of another vessel owned by the same insured. The underwriters of the first vessel having paid claimed to rank pari passu with the lost cargo owners in the distribution of the limitation fund lodged in court by the owners in respect of the second vessel. Insurers under English law have no right in their own name to recoup insured losses from wrongdoers. They have to rely on rights of subrogation, using their insured's name. Since the common owner of the two vessels could not sue himself, the underwriters' claim failed. The case does not however address situations of contribution. Where there is a right to contribution, an insurer can recoup his loss from a third party. Here, the question is whether a right of contribution should be recognised by Zurich against IEG on the basis that IEG should in justice pay its proportionate part of a liability arising from a risk which increased proportionately over the whole period of 27 years during which it exposed Mr Carré to asbestos dust.
It is equally irrelevant that the law knows no such thing as a contract of self-insurance. It is of course true that, just as an insured cannot sue himself, so an insured cannot in law insure with himself. But the concept of "self-insurance" is not unhelpful in identifying an important truth. A person who does not insure at all is well understood to be undertaking a risk for his own account, for which he should answer accordingly. A person who after insuring for a period with insurer A then goes for a period to insurer B is understood to be looking in relation to the later period to insurer B alone. Even courts are entitled to deploy a helpful phrase to point to such truths. The United States courts did so in Insurance Company of North America v Forty-Eight Insulations Inc 633 F.2d 1212 and Security Insurance Co of Hartford v Lumbermens Mutual Casualty Co (2003) 264 Conn 688, 826 A2d 107, when they held that, as between an insured and its insurers, liability for defence costs should be pro-rated across all periods of insurance and self-insurance during which exposure had occurred. In Lumbermens the insurer was thus held liable pro-rata by reference to the relationship between its insurance period, other periods of insurance with other insurers and periods of "self-insurance".
The use of the concept in this jurisdiction is illustrated by Lord Napier and Ettrick v Hunter  AC 713, 730E-F, where Lord Templeman had no hesitation about describing a Lloyd's name as "his own insurer" in respect of a £25,000 excess under the stop loss policy in issue. He concluded in its light that such a name was not entitled, as against his stop loss insurers, to retain the benefit of damages for negligent underwriting received from the Outhwaite syndicate. The "fundamental principle" in Castellain v Preston (1883) 11 QBD 380, that an insured was entitled to be fully indemnified, was not "helpful in deciding whether a name who promised the stop loss insurers to bear the first £25,000 loss is entitled to be put in the same position as an insured person who makes no such promise": p 731B-C.
In the present case, an insured who insures for a limited period necessarily accepts that it is only liability incurred during that period for which he has cover. The unique feature of the present situation is that the whole substratum of the relevant insurance policies has changed fundamentally since they were underwritten, and the law has, for the first time ever, imposed liability on the basis of risk, rather than the probability, that negligence during the insurance period led or contributed to the illness complained of. The concomitant of insurance liability in this situation must be a recognition that the law can and should redress the unjust and wholly anomalous burden which would otherwise fall on any particular insurer with whom insurance was only taken out for part of the total period of exposure by the insured, by recognising an obligation on the part of the insured to contribute pro tanto to such liability as a self-insurer.
In my opinion, therefore, Zurich is entitled to look to IEG to make a contribution based on the proportionate part of the overall risk in respect of which it did no respect of which Zurich does not recover contribution from any other insurer. Any contribution which is credited by Excess to Zurich in excess of 2/27 of Zurich's liability to IEG should also give Excess a corresponding right to contribution from IEG. I believe that this leads in practice, at least in the case of a solvent insured, to substantially the same result as that at which Lord Sumption arrives, but by a different route, which in my opinion reflects the reasoning and result in Trigger. The difference between the two routes may however be important in the context of an insured who is not solvent.
It is convenient to address an area about which Lord Sumption expresses conclusions at the outset of his judgment, and to which he reverts at paras 172-173. That is that the conclusions reached up to this point will not mesh with the FSCS schemes established under the Financial Services and Markets Act 2000 for insurer insolvency (see para 6 above) and more recently the Mesothelioma Act 2014 for cases where there is employer insolvency and no identifiable insurer. This point relates to statutory schemes separate from and in part post-dating the development of the common law and statutory principles with which this appeal is concerned. No submissions have been addressed to the court on it. That itself also suggests that the insurance industry and their expert representatives before this court do not share Lord Sumption's concerns.
One reason for this may also be that Lord Sumption's account of the position is incomplete. He states that "The effect of the majority's view is simply to transfer risk from the statutory compensation schemes which were created to assume that risk, to an arbitrarily selected solvent insurer who has not agreed to do so" (para 112, last sentence). This analysis does not address the fact that, on his own case, the statutory compensation schemes do not cover all situations or losses. Take a case of two responsible employers, one of which [A] is solvent or has a solvent insurer for the whole period for which it exposed the victim, the other of which [B] is insolvent and without any identifiable insurer. The victim will, on Lord Sumption's case (para 160), recover 100% from employer A. Employer B will be liable to contribute to employer A (or its insurer, by right of subrogation), but will have no money and no insurer to enable it to do so. The 2014 Act scheme will not step into the gap to enable employer A or its insurer to recoup pro rata contribution, because of section 2(1)(d) or (e), and possibly also because of section 2(1)(c), of the Act.
This is because the Act was passed to protect unpaid victims, not for insurers' benefit. It was and is directed, as the notes to the relevant Bill state, to situations where "by virtue of the passage of time no solvent employer remains to be sued, and the employee is often unable to trace any insurer who was providing EL insurance to the employer at the relevant time". Recovery from another insurer of another employer precludes use of the 2014 scheme: see The Diffuse Mesothelioma Payment Scheme 2014, by Judge Nicholas Wikeley, Emeritus Professor at Southampton University, (2014) 21 JSSL 65, 78. Any action for damages or receipt of any damages or of a specified payment (which, like the 2014 scheme itself – see para 6 above - might not cover the full loss) precludes use of the 2014 scheme. This makes sense, since the 2014 scheme assumes, in general, that any recovery by a mesothelioma victim will correspond, even if only approximately, with full recovery of the victim's whole loss. This is unsurprising in the light of Fairchild, the 2006 Act and Trigger, all of which form part of the background to the Act. But it indicates that the 2014 Act, far from supporting, is inconsistent with the scheme which Lord Sumption advocates whereby an insurer may only be liable to indemnify on a pro rata basis.
Finally, if Lord Sumption be right and he has identified significant potential anomalies on the approach which has been advocated by counsel representing insurers before us and which in my opinion should be adopted, the reality is that the Fairchild enclave has necessitated adjustment from time to time of the legal and regulatory framework by the courts, the legislature and regulatory authorities. As Wikeley notes, "further attempts to engineer improvements to the underlying compensation arrangements [are] almost inevitable" (p 82). I do not myself see such a process of adjustment as one from which courts should withdraw.
Third Parties (Rights against Insurers) Act
Since IEG is solvent and has met the whole of Mr Carré's loss, the present appeal concerns only the relationship between IEG and Zurich. In that context, the precise legal relationship between Zurich's right to look to IEG for contribution and IEG's policy claim against Zurich does not matter. In practice, even if Zurich's right to contribution does not give rise to a defence, a procedural order for a stay would ensure that the one claim could not be enforced without taking into account the other. But in cases where the person responsible is insolvent, and the use of the Third Parties (Rights against Insurers) Act 1930 (soon, it is to be hoped, to be replaced by the 2010 Act) is invoked, it may be important whether the right of contribution which Zurich enjoys constitutes a defence reducing the indemnity for which the insured can sue under that Act.
Section 1 of the 1930 Act provides:
When the 1930 Act applies, it therefore transfers to the mesothelioma victim the insured's rights under the insurance contract in respect of the insured's liability to the victim. The same is provided by the 2010 Act, not yet in force. Whether an insurer's right to contribution against the insured constitutes a full or partial answer to a victim's policy claim based on such a transfer is a question of great potential importance. It raises questions of some complexity, on which it is unnecessary to give a final answer on this appeal, but about which I wish to make some observations. One question is whether, apart from any statutory transfer under the 1930 or 2010 Act, the insurer's right to contribution would be a defence at common law to a claim by the insured for indemnity under the insurance, as opposed to giving rise to procedural remedies such as a stay. A second is whether it makes any difference to the application of the relevant common law rules in this context that the claim is being brought under the 1930 or 2010 Act. A third is whether the terms of the Act positively exclude or restrict any such defence.
The first and second aspects raise, as sub-issues, the existence of any right of relief based on set-off, circuity of action or other equitable basis. Zurich positively submitted that it would have no right of set-off, legal or equitable. One objection to set-off is that a right to contribution only arises upon payment by the person seeking contribution: see eg Andrews & Millett's Law of Guarantees (6th ed) (2011), para 12-019, citing Ex p Gifford (1802) 6 Ves Jr 805 and In re Snowdon (1881) 17 Ch D 44; and see Davies v Humphreys (1840) 6 M & W 153, Stirling v Burdett  2 KB 418 and In re Beaven  2 KB 595, 600. On the face of it, that presents a real obstacle to any suggestion by any insurer in Zurich's position of set-off, whether legal or equitable, against IEG's claim for the full amount of its loss.
There is however first instance authority endorsing the availability of a further remedy in cases where a person A (here, for example, Zurich), liable to make a payment to person B (here, the person suffering mesothelioma), has a potential right to receive contribution (or a full indemnity) from a third person C (here, IEG). In Wolmershausen v Gullick  2 Ch 514, Wright J made a prospective order in such a case directing that, upon person A paying person B, person C was to exonerate person A from liability beyond person A's share. In Rowland v Gulfpac Ltd  Lloyd's Rep 86, 98, Rix J held that he had jurisdiction to grant a freezing order quia timet to support an indemnity claim by person A against person C, even though the common law claim for an indemnity was not complete. His decision was more recently followed by Burton J in Starlight Shipping Co v Allianz Marine & Aviation Versicherungs AG (The "Alexandros T")  EWHC 3381 (Comm),  1 Lloyd's Rep 162, paras 37-38, where he said that the constitution of such a fund would ensure that person A was held harmless and not be required to use his own funds to discharge liabilities falling within the relevant contract of indemnity by person C.
Accepting the fairness of the thinking behind this first instance authority without further examination, I doubt whether it could or should affect the application of the general principle mentioned in para 86 in the particular context of a claim by a victim under the 1930 or 2010 Act. Zurich's obligation under the insurance and that Act would be to indemnify the victim. Any consequential right to contribution from IEG would arise not "under", but outside, the insurance contract in terms of section 1(1) of the 1930 Act. Considerations of justice and policy would also support the treatment of the insurance and the contribution positions as legally separate, when an opposite approach would be to the prejudice of the victim, in whose favour the insurance would otherwise operate and who is not concerned with the circumstances giving rise to any contribution claim.
A second sub-issue is that legal set-off is in any event confined to debts due and payable and either liquidated or capable of ascertainment without valuation or estimation: Stein v Blake  AC 243, 251 per Lord Hoffmann. On current authority, at Court of Appeal level, the right to recover under an insurance contract is classified not as a debt, but as a right in damages: see eg The Italia Express (No 2)  2 Lloyd's Rep 281, 286, Sprung v Royal Insurance (UK) Ltd  1 Lloyd's Rep IR 111. Further, a right to claim proportionate contribution would not normally satisfy the test of legal set-off, although, on the agreed facts in this case, it might perhaps do so, since they lead to a definite percentage contribution of 22.08%. Regardless of the view taken on these two points, legal set-off is procedural, not substantive. When one comes to the second aspect, the statutory transfer probably therefore precludes legal set-off.
In contrast, equitable set-off, where available, can give rise to a substantive defence. The locus classicus is Hanak v Green  2 QB 9 and the later case-law includes Federal Commerce and Navigation Co Ltd v Molena Alpha Inc (The "Nanfri")  2 QB 927, and has, more recently, been analysed by Rix LJ in Geldof Metaalconstructie NV v Simon Carves Ltd  EWCA Civ 667,  4 All ER 847. Rix J noted (para 26) that in The Nanfri the Court of Appeal had identified the need for the cross-claim to arise out of the same transaction as the claim or be closely connected with it. He concluded (para 43(vi)) that the best restatement of the principle was that it applies where there were a cross-claim "so closely connected with [the claimant's] demands that it would be manifestly unjust to allow him to enforce payment without taking into account the cross-claim". Again, I consider that, in a context where any set-off arises from circumstances outside the insurance policy and would be to the prejudice of a third party victim, the considerations of policy and justice behind the rules developed in Fairchild and Trigger would probably mean that it was just (rather than "manifestly unjust") for Zurich to have to fulfil its insurance policy obligations, before asserting against IEG any contribution claim based on circumstances outside the scope of the insurance to the prejudice of that victim. Even in circumstances where liability insurance is not compulsory, it would be wrong to view liability insurance as if its only rationale was to benefit the insured's bottom line, rather than to give effect to legitimate expectations regarding the protection of employees and other third party victims. That rationale is reflected in the 1930 and 2010 Acts, and reinforced by the now compulsory nature of employers' liability insurance. The court would also be entitled to take it into account, when considering for the purposes of equitable set off what is or is not "manifestly unjust".
As to circuity of action, this is an ill-defined principle, recently confirmed though not elaborated in Farstad Supply A/S v Enviroco Ltd  UKSC 18,  Bus LR 1087, where previous authorities are identified. In the present context it could not, I think, be more than a remedy existing where there would be no point in a claim being permitted, because any amount awarded could be immediately recovered on another basis. On that basis, it could not add anything to the previous discussion.
The third aspect identified in para 85 above would also be problematic, were it to be relevant. Where an insurer does have a set-off (one which appears in each case to have been capable of operating in equity), there is conflicting authority as to whether such a set-off is excluded by the 1930 Act. In Murray v Legal and General Assurance Society Ltd  2 QB 495, Cumming-Bruce J held that a right to recover premiums did not arise "in respect of" the insured's liability to the third party, within section 1(2), and that insurers could not therefore set-off unrecovered premiums. In Cox v Bankside Members Agency Ltd  2 Lloyd's Rep 437, 451, Phillips J refused to follow Murray and held the third party claim to be subject to a set-off arising from payment by insurers of defence costs falling within the insured's policy excess and recoverable either under an express policy term or in restitution. However, both these cases concerned cross-claims which arose directly from and under the insurance policy. Here, any right of contribution is best analysed in my view as arising from circumstances outside the insurance policy, and on that basis as not capable of giving rise to a set-off at all.
As noted in para 86 above, no right of contribution normally arises until payment. Once the victim (person C) has established the liability of the insured (person B), person B's rights to indemnity by the insurer (person A) "under the contract in respect of" that liability are transferred to person C. Neither Murray nor Cox concerned a defence to a claim under the insurance contract which was based on a cross-claim arising from circumstances outside the insurance contract and which could only become due on person B being paid in full in respect of his liability to person C. There is thus, in my view, a strongly arguable case for treating the language of section 1(1) of the 1930 Act as entitling the third party to recover against the insurer in such a case, leaving the insurer to enforce any claim to contribution which it may have against anyone separately and in the ordinary course, subsequently.
For reasons given in paras 37 and 38 above, there are significant differences between the defence costs incurred by IEG and the hypothetical position regarding compensation in circumstances covered by the 2006 Act, which I have been discussing in paras 39 to 82. In particular, the right to defence costs exists under the insurances on a conventional causative basis, and the defence costs incurred were not increased by the fact that they related to a claim for an additional 21 years in addition to the six years insured by Midland. In contrast, in the hypothetical position, the insurer only incurs liability on the unconventional basis of a risk that the mesothelioma was due to exposure during the insurance period, when there was a proportionately greater risk that mesothelioma was due to exposure during other periods when the insured did not insure at all or chose to insure elsewhere. It may still be as a matter of fact that the likelihood of mesothelioma occurring (and so of any defence costs being incurred) would have been proportionately reduced, had there only been exposure during the six years of the Midland insurance. But the liability for defence costs incurred in defending a claim embracing a period longer than that insured arises directly from the policy wording, as it would always have been understood, and it has, at most, been only indirectly affected by the special rule of causation and statutory intervention which have impacted the rest of the main insuring clause. In these circumstances, the impetus to recognise a right to contribution as a matter of compelling justice and equity is self-evidently diminished.
I would therefore decline to recognise any such right to contribution in respect of defence costs, but I would accept that such a right exists regarding compensation in the hypothetical situation which would arise had the 2006 Act applied. On the agreed facts, the only tenable basis for apportioning responsibility and arriving at the appropriate contribution would be proportionately to the relevant periods of exposure insured and not insured with Midland.
It follows from the above that the appeal should succeed on the first main point, as stated in para 35 above. It fails on the second main issue as regards defence costs. Had the 2006 Act been applicable, I would have recognised Zurich as having rights both to look to Excess for a pro rata share of liability and to require IEG to bear an appropriate contribution, as indicated in paras 39 to 82 above.
As at present advised, and although IEG is solvent so that the present appeal is concerned only with the position between IEG and Zurich, I also consider that, in the case of a claim by a victim of mesothelioma against an insurer (such as Zurich) under the Third Party (Rights against Insurers) Act 1930, the insurer would be obliged to provide the full policy indemnity, without being able to set off against the victim any consequential right to contribution which it might thereafter have as against the insured (here IEG): see paras 83 to 93 above.
(with whom Lord Mance, Lord Clarke and Lord Carnwath agree)
The courts continue to grapple with the consequences of departing from the "but for" test of causation in order to provide a remedy to those who have contracted mesothelioma as a result of wrongful exposure to asbestos fibres. As the precise pathogenesis of that terrible disease is unknown, the House of Lords and the Supreme Court departed from established legal principle and extended the law of causation. As a result, an employer, which has wrongfully exposed its employee to significant quantities of asbestos fibres and thus materially increased the risk of his suffering mesothelioma, incurs liability in damages to the employee or his estate if the employee subsequently contracts the disease. The claimant does not have to prove on the balance of probabilities that the wrongful exposure caused or materially contributed to the development of the disease. This innovative rule of causation has, within its defined scope, which is not confined to mesothelioma, imposed liability not only on employers (Fairchild v Glenhaven Funeral Services Ltd  1 AC 32 and Barker v Corus UK Ltd  2 AC 572) but also on their liability insurers through the court's interpretation of liability insurance contracts (Durham v BAI (Run Off) Ltd  1 WLR 867 (the "Trigger litigation")). Parliament has also intervened in section 3 of the Compensation Act 2006 by reversing the effect of Barker in relation to mesothelioma cases.
This appeal concerns the liability of an insurer which has provided an employer with liability insurance cover for only part of the period of the employee's employment, during which he was wrongfully exposed to significant quantities of asbestos fibres, and the employer was either uninsured for the rest of the period or was insured by an insurer who is now insolvent or who cannot now be traced. The principal issues are
whether the insurance policies respond to the full extent of an employer's liability to the employee or only a proportionate part of that liability fixed by reference to the periods of cover for which premiums have been assessed and paid, and
if the former, whether the insurer has a claim against insurers of the employer in respect of other periods of the employee's exposure and against the employer itself for periods in which it was uninsured or in respect of which its insurer can no longer be identified or traced.
This court is unanimously of the view that section 3 of the Compensation Act 2006 did not change the common law, which the House of Lords had laid down in Barker v Corus UK Ltd  2 AC 572, but overrode it only to the extent that the section provides. The court also holds, unanimously, that the appeal fails on the issue of defence costs. On those matters no more need be said.
The division of opinion arises in relation to what Lord Mance describes as the second main question, namely the extent of the insurer's liability when it has insured the employer for part only of the period of the employee's exposure. It is a matter of agreement that liability insurance would have been placed on the basis that a particular loss would fall into one insurance period, for which the insurer had assessed the premiums and provided the cover. As Lord Mance has shown in para 43 of his judgment and Lord Sumption in para 155 of his, it would be seriously anomalous if the insurer, which provided cover for a small proportion of the period of the employee's exposure, were to carry the whole of the employer's liability without any recourse against others in respect of the other periods of exposure. The stark options to avoid the identified anomalies are:
to hold, as Lord Sumption propones, that the insurance contract is to be construed so that the insurer's liability for the loss is limited to the proportion of the policy years in which it provided cover relative to the whole period during which the employer wrongfully exposed the employee to the asbestos fibres; or
to adopt the approach, which Lord Mance proffers, that the insurer must meet the whole of the employer's liability to the employee and that, having done so, the insurer has the right to seek proportionate contributions from other insurers, which gave liability cover to the employer in other periods, and also, in respect of any period in which there was no insurance company from which a contribution can be obtained, against the employer itself.
Each approach is a possible way of avoiding unfairness to the insurer. Lord Mance's is more radical.
I have found this a difficult case, not least because I am generally averse to developing the common law other than by the application of general principles. I have shared the concerns which Lord Neuberger and Lord Reed have articulated. But we are where we are. The law has tampered with the "but for" test of causation at its peril: Sienkiewicz v Greif (UK) Ltd  2 AC 229, Lord Brown at para 186. The Fairchild enclave exists: the courts in Fairchild and Barker and the Trigger litigation, for obvious reasons of policy, have developed a special rule of causation to do justice to the victims of wrongful exposure to asbestos fibres who have contracted mesothelioma as a result. Having done so, the courts must address the consequences of that innovation.
I am persuaded that this court should develop the law as Lord Mance has proposed for the following six reasons. The first three address the extent of the insurer's liability to the employer. The next two relate to the rights of recourse of the insurer, once it has paid the employer or victim in implement of its obligations under its insurance policy. The final reason relates to Lord Mance's proposal as a whole.
First, in my view, the finding that the insurer, which has provided liability cover to an employer for only part of the period of the employee's exposure, must meet the entirety of the employer's liability for the whole period of exposure is consistent with the way the courts have developed the common law in the trilogy of cases. In particular, it is consistent with the position of the majority of this court in the Trigger litigation. In that case the majority imported into the insurance contract the weak or broad concept of causation, which the House of Lords had adopted in imposing tortious liability on the employer. To my mind this is clear from Lord Mance's leading judgment in the Trigger litigation (in particular at paras 52 and 57, 64-68, and 72-74) and also in the concurring judgment of Lord Clarke of Stone-cum-Ebony (at paras 83-85). The creation of liability for mesothelioma by virtue of the exposure to the asbestos fibres, which materially increases the risk of that disease, means that the mesothelioma is caused in this broad sense in each and every period of such exposure, as Lord Mance argues in this appeal. As a result, the insurer, which provided liability insurance for a limited period, is exposed to the whole of the employee's claim if there was wrongful exposure in that period.
Secondly, while this imposes a heavy burden on the insurer which the employer selects to claim its indemnity, it is a result for which the appellants and interveners have argued in this appeal. It appears to be a result that the London insurance market is prepared to live with. It is striking that the insurance industry in this appeal has shown no enthusiasm for the elegant and less complex idea of construing the insurance contract to restrict the insurer's liability to a proportionate part of the loss.
Thirdly, it is consistent with the policy of the United Kingdom Parliament that the employee-victim should be able to obtain damages for his loss in a straightforward way. This policy of protecting the employee-victim is clear at a general level from the enactment of the Employers' Liability (Compulsory Insurance) Act 1969. It is clear, more specifically, in Parliament's enactment of (i) section 3 of the Compensation Act 2006 to reverse the decision of the House of Lords in Barker and more recently (ii) the Mesothelioma Act 2014 to establish an insurance industry fund to deal with the contingency that a victim is unable to bring an action for damages against an employer or a relevant liability insurer. It is also consistent with the expansion of the Financial Services Compensation Scheme by the Compensation Act 2006 (Contribution for Mesothelioma Claims) Regulations 2006 (SI 2006/3259). Confining the insurer's liability to a time-related proportion of the employer's liability would not be in line with this policy of the legislature and would probably engender further legislation. While Parliament's role of legislating in the public interest differs from the role of judges in developing the common law, it is legitimate for the courts to consider whether their initiatives are in harmony with legislative policy expressed in statutes: Johnson v Unisys Ltd  1 AC 518, para 37 per Lord Hoffmann.
I turn to Lord Mance's proposed innovations to address the serious anomalies, which he and Lord Sumption have identified, if the insurer in one insurance period were to bear 100% liability without any recourse against those responsible during other insurance periods. They are:
the broad equitable extension of the right of contribution between insurers and
a right of recoupment against the employer in respect of years in which it was not insured or can identify no insurer against which contribution can be claimed.
Thus, fourthly, if, as I consider, it is correct that the majority's decision in the Trigger litigation points towards the insurer's 100% liability (para 104 above), the interpretation of the insurance contract as creating a pro rata liability is not an option and the anomalies must be addressed in some other way.
Fifthly, I am not as concerned as Lord Neuberger and Lord Reed are about the danger of infecting other areas of the common law with uncertainty. The court is crafting a solution for the problems that stem from the alteration of the rules of causation and the solution applies only to cases to which the altered rules of causation apply. In other words the special rules apply only to cases within the Fairchild enclave. The House of Lords in Gregg v Scott  2 AC 176 has been careful not to allow the relaxation of the established rules of causation more widely by applying a weak rule of causation outside the Fairchild enclave. The courts will have to police the boundaries of the enclave. So long as
the rights of recourse against other insurers and the insured employer are recognised for what they are, namely as a means of avoiding anomalies as a result of the special rules of causation and
those special rules are confined to the circumstances which Fairchild addresses, there is no reason why the boundaries of the Fairchild enclave should not be preserved.
I recognise that those boundaries are not coterminous with liability for mesothelioma and that the precise boundaries of the Fairchild principle, like those of the earlier case of McGhee v National Coal Board  1 WLR 1, may have to be worked out in other cases – viz Novartis Grimsby Ltd v Cookson  EWCA Civ 1261. But it is sufficient, in my view, that the insurer's rights of recourse are available only within the Fairchild enclave.
Finally, the practical solution which Lord Mance proffers appears to be consistent with the way in which the London insurance market has operated in handling mesothelioma claims. That may suggest that the solution will not give rise to major practical difficulties.
This is not a view which I have come to without hesitation because I see the strength of the arguments
that the courts should develop the common law in a principled way,
that in the context of an insurance contract the correct tools to give effect to the parties' intentions are construction of the contractual words or the recognition of an implied term, and
that the protection of the employee-victim's entitlement to recover damages is a matter for Parliament.
In short, having dug a hole, the courts should not keep digging. But the majority judgment in the Trigger litigation, which is the first of the six reasons set out above, appears to preclude the construction of the insurance contract which the minority favour. That consideration and the other five reasons persuade me that Lord Mance's approach is the best available means of avoiding the injustice which the insurer would otherwise suffer.
(with whom Lord Neuberger and Lord Reed agree)
I agree that this appeal should be allowed, but I regret that I cannot agree with the reasons given by the majority, which seem to me to be contrary to a number of basic principles of the law of contract and to be productive of uncertainty and injustice. Suppose that an insolvent employer had tortiously exposed his employee to asbestos for, say, 30 years before going out of business. The employer had failed to insure his liabilities at all for years one to 20. Insurer A insured his liability on an occurrence basis in year 21. Insurer B insured his liabilities under successive annual policies for years 22 to 30, but insurer B is insolvent. The majority would hold that, in a case governed by the 2006 Act, insurer A is liable for the entire loss incurred over the 30 years of exposure, although he was only on risk for one, but that he has an equitable right to recoup a proportionate part of that liability from the insolvent estate of insurer B in respect of the nine years when insurer B was on risk, and from the insolvent estate of the employer in respect of the 20 years when there was no insurance. The effect, and as I understand it the object, of this is to make insurer A, who is solvent, answerable,
in respect of periods when insurer A was not on risk but insurer B was; and
for the failure of the employer to insure at all in the first 20 years.
In my opinion, the correct result in this situation is that insurer A is liable for a proportionate part of the loss in respect of the one year out of 30 when he was on risk. The employee is entitled to recover insurer B's proportion under the statutory compensation scheme established under section 213 of the Financial Services and Markets Act 2000 for cases of insurer insolvency. In respect of the 20 years when there was no insurance, he is entitled subject to the statutory conditions of eligibility to recover under the statutory compensation scheme established under the Mesothelioma Act 2014 for cases where there is no insurance. The effect of the majority's view is simply to transfer risk from the statutory compensation schemes which were created to assume that risk, to an arbitrarily selected solvent insurer who has not agreed to do so.
The liabilities of an insurer are wholly contractual. The answer to the questions now before the court necessarily depend on the construction of the contract and on nothing else. Under an annual policy of insurance written on an occurrence basis, the insurer's liability is limited to occurrences caused during the contractual term. Where the relevant occurrence has been caused at an indeterminate time during the period of exposure, there are in my view only two possible meanings that can be given to the contract. One is that the insurer is not liable at all. That possibility was rejected by this court in Durham v BAI Runoff Ltd. (In re Employers Liability Policy "Trigger" Litigation  1 All ER 605.) The other is that each insurer must severally answer for a rateable part of the employer's liability, corresponding to the proportion which his time on risk bears to the period of exposure. No insurer can be liable in respect of other periods when he was not on risk or there was no insurance in place at all. That appears to me to be the correct answer to the problem which has arisen on this appeal. The suggestion that an insurer who was on risk for only part of the period of exposure, however brief, can be liable as if he had been on risk for the entire period, is contrary to the express terms of the contract and to the nature of annual insurance. The suggestion that some doctrine of law can be devised which imposes on an insurer in one year the risk that insurers of other years may become insolvent or that in other years the employer may fail to insure at all, is both unprincipled and unjust. The suggestion that equity can partially adjust the result of this injustice by requiring the insured to repay to the insurer part of the insurance moneys which the latter was contractually obliged to pay him, is contrary to basic principles of law. It is the proper function of this court to review those principles. But the very immensity of this power requires it to act within a framework of legal principle. The court identifies general principles of law and applies them to the case in hand. If the facts of that case disclose some generally unsatisfactory feature of the law as hitherto understood, it may modify it. To devise a special rule for one industrial disease and impose it retrospectively on a policy that covers all industrial accidents and diseases, so as to alter what all members of this court acknowledge to be the basis of the parties' agreement, seems to me to be an extremely undesirable course to adopt.
It may fairly be said, and indeed is said by the majority, that this court had already, in Fairchild v Glenhaven Funeral Services Ltd  1 AC 32 created a special rule for mesothelioma which does not conform to the ordinary principles on which the common law acts. It has done this in the interests of avoiding a serious injustice. Therefore, the argument continues, it is incumbent upon us now to develop what is called the "Fairchild enclave" by devising ancillary rules which appear to do justice to cases within the enclave, even if they are also out of line with the ordinary principles of law. The difficulty about this approach is that a measure of legal coherence seems desirable even within the Fairchild enclave. The contractual analysis has the considerable advantage that it draws on a substantial body of existing legal principle, which can be expected to supply answers to unforeseen issues as they arise. The alternative is for the law to move from each one-off expedient to the next. This can only generate knock-on consequences which we are not in a position to predict or take into account. If there were no other way to achieve justice, these consequences should no doubt be borne. But it is quite unnecessary to do so in this case. In the first place, the incidents of liability in tort are the creation of rules of common law, whereas the extent of a contractual liability depends on the intentions of the parties. The scope for judicial inventiveness is therefore necessarily more limited in the latter context than in the former. Secondly, it goes without saying that insurers are as much entitled to justice as mesothelioma victims. Third, the protection of victims against the insolvency of some out of a number of employers' liability insurers or the failure of an employer to insure at all in some out of a number of years of exposure, is properly a matter for statute. It has in fact been dealt with, to the extent that Parliament considers appropriate, by the creation of statutory compensation schemes. It is difficult in those circumstances to discern what social imperative can require us to depart from ordinary principles of law.
Between the end of the nineteenth century and the 1970s asbestos was commonly used for a wide variety of purposes, notably for sound and heat insulation in the building trades and in the manufacture of electrical and other appliances. It has been known for more than 80 years that exposure to high levels of asbestos is injurious to health, and in the United Kingdom regulations have sought to limit levels of exposure since 1931. Mesothelioma is a malignant tumour whose association with asbestos exposure was identified in the 1960s. It is usually caused by asbestos particles inhaled in the course of occupational exposure to the mineral, although occasionally by environmental asbestos. It is a breach of an employer's duty to allow its employees to be exposed to significant levels of asbestos without taking reasonable steps to protect them from inhaling it.
Mesothelioma has a number of distinctive characteristics. A single exposure to asbestos particles may be enough to cause the condition to develop but will not necessarily do so. The intensity of exposure depends, among other things, on the dose and fibre type. The greater the intensity and duration of exposure, the higher the risk that mesothelioma will develop. But once contracted the disease is not progressive with exposure: subsequent further exposure will not aggravate it. As Rix LJ put it in the Court of Appeal in the Trigger litigation  1 All ER 605, at para 51, summarising the findings of the trial judge:
once the mesothelioma tumour is present and assured of growth (ie has passed the stage where a malignant mutation may die off), further asbestos exposure and indeed further asbestos fibres in the body can make no difference and are not causative.
These features differentiate mesothelioma from other industrial diseases and from long-term sources of damage such as the industrial pollution of land which are progressively aggravated by successive occurrences to a degree which is in principle capable of being measured or estimated. They present particular problems of attributing responsibility given that the disease is undetectable until shortly before death, and once initiated may be latent for many years (30 to 40 years is common) before the symptoms appear. If a person has been exposed to high levels of asbestos over a long period, it is impossible in the current state of medical science to determine at what stage he inhaled the fibres which ultimately led to his developing mesothelioma. This means that if he was exposed to asbestos by successive employers during that period, each period of employment will have materially increased the risk of his contracting the disease without necessarily causing it.
Employers' liability insurance has been compulsory in the United Kingdom since the Employers' Liability (Compulsory Insurance) Act 1969, which came into force on 1 January 1972. Section 1(1) of that Act requires employers to be insured against "liability for bodily injury or disease sustained by his employees, and arising out of and in the course of their employment in Great Britain in that business". It is normal for employers to be insured on an occurrence (as opposed to a claims made) basis under successive annual policies which may be underwritten by different insurers. Under most of the standard wordings in common use, an industrial disease will be treated as having occurred when it was caused or initiated, even though it only developed or manifested itself later. I shall return to this point when I come to deal more fully with the "Trigger" litigation. It follows from the characteristics of mesothelioma which I have described that three potential problems can affect the availability of insurance cover. The first is that where an employee was exposed to asbestos by different employers at different times, it will be impossible to determine which employer's wrong caused the disease to develop and therefore which employer's insurers should respond. The second is that even where only one employer was involved, that employer may have been insured for only part of the period of exposure, or may have been insured in different years by different insurers. It will then be impossible to determine whether at the time when the disease "occurred" the employer was insured or, if he was, under which policy and by which insurer. The third is that one or more of the insurers potentially liable may have become insolvent or have been wound up in the course of corporate restructuring or have ceased to carry on this class of business and simply disappeared. The present appeal is primarily concerned with the second of these problems, although it also has implications for the third.
Since at least the 1990s the insurance industry in the United Kingdom has evolved voluntary procedures for dealing with these problems in the context of claims for mesothelioma. Since these procedures have had a significant influence on the positions taken by the parties to this appeal, and affect the commercial implications of the various possible outcomes, it is necessary to say something about them. The arrangements appear to have varied in detail, but since 2003 have been embodied in "guidelines" issued by the Association of British Insurers, a body predominantly comprising insurers but with some representation of non-insurance interests. The 2003 Guidelines, which were issued in the aftermath of the decision in Fairchild, recommend a scheme of settlement which is described as "equitable and pragmatic". Its essential features are:
that the victim is to be paid in full by the "Lead Insurer";
that where more than one employer is involved liability is notionally apportioned between them pro rata to their respective periods of culpable exposure, without regard to any difference in the intensity of exposure;
that each employer's proportion of the claim is then further apportioned between that employer and its insurer or insurers according to the proportion which their time on risk bears to the whole period of culpable exposure by that employer; and
that periods when the employer was "self-insured, uninsured or unable to trace insurance" are apportioned to the employer if it is solvent, and otherwise to the relevant employer's insurers (irrespective of their solvency).
The effect of point (iv) is that where the employer was insured but the insurer is insolvent, the insolvent insurer's pro rata share is paid by the Financial Services Compensation Scheme established under section 213 of the Financial Services and Markets Act 2000, which is party to the scheme. According to Mr Allen, an experienced claims manager whose witness statement was put in by the Association of British Insurers, the main objectives of the industry scheme are to promote speed of settlement, to prevent the "spiking" of claims from an uninsured year into an insured one, or from a year with a higher deductible into another with a lower one, and to "avoid time-consuming and costly disputes about issues such as the dose, intensity or fibre type of a claimant's exposure". His evidence is that it proved impossible in practice to persuade insurers to pay the employee up front and then sort out the distribution of the cost among other participants later. Insurers preferred to wait until the FSMA compensation scheme was committed. Subject to that, the industry scheme has been accepted by the United Kingdom industry and has not in practice been challenged by reinsurers, although they would be likely to do so if they thought that claims were being artificially "spiked" into a year when their reinsured was on risk. There is some dispute about how far the industry scheme has been accepted by insureds, as Mr Allen contends, but it is neither necessary nor possible to resolve that question.
International Energy Group Ltd ("IEG"), is the successor to the rights and liabilities of Guernsey Gas Light Company Ltd, which employed Mr Alan Carré between 1961 and 1988. Mr Carré claimed to have been negligently exposed by his employer to asbestos particles throughout that period, and to have contracted mesothelioma in consequence. The obligations of the employer to Mr Carré were governed by Guernsey law. It is agreed for the purpose of these proceedings that the common law of Guernsey is the same as the common law of England. The statute law is, however, different. One of the differences is that employer's liability insurance was not compulsory in Guernsey until the Employer's Liability (Compulsory Insurance) (Guernsey) Law 1993, which came into force on 1 March 1994. Nonetheless, Guernsey Gas was insured for at least part of the period during which it employed Mr Carré. It was insured between 31 December 1978 and 30 December 1980 by Excess Insurance Co Ltd, and between 31 December 1982 and 31 December 1988 by Midland Assurance Ltd. Accordingly Excess was on risk for two and Midland for six of the 27 years during which Guernsey Gas employed Mr Carré. For the remaining 19 years, the employer was either uninsured or else insured under a policy all trace of which has been lost so that it is for practical purposes unable to claim under it. IEG is, however, solvent and capable of meeting the claim from its own resources.
Zurich Insurance plc, are a major insurer of employers' liability in the British Isles who acquired Midland and succeeded to its liabilities.
In September 2008, shortly before his death, Mr Carré began proceedings against the employer in the Royal Court in Guernsey in support of a claim for damages on the footing that it had exposed him to asbestos without adequate protection. The proceedings were settled in December 2008 for £250,000 in respect of damages and interest and £15,300 in respect of Mr Carré's costs. IEG also incurred defence costs of £13,151.60. The company settled these sums in full and claimed them from Zurich. Zurich offered to settle the company's claim in accordance with the industry guidelines. Since IEG was solvent, it offered a rateable proportion of the claim, reflecting the ratio of its time on risk to the total period of Mr Carré's employment by Guernsey Gas. IEG began the present proceedings against Zurich in support of their claim for the entire amount. It was agreed that the dispute should be resolved on the basis of agreed facts. These were, in summary,
that Mr Carré was exposed to asbestos with the same frequency and intensity throughout the 27 years of his employment by the employer,
that that exposure materially increased the risk that he would contract mesothelioma, and
that by reason of the exposure Guernsey Gas was in breach of its duty to him.
Before examining the basis of IEG's claims and Zurich's response, it is necessary to deal with the complex legal background against which the rival contentions were advanced.
The position as between employer and employee: Fairchild
In Fairchild v Glenhaven Funeral Services Ltd  1 AC 32, the House of Lords held that where one of a number of successive employers must have caused the development of mesothelioma by tortiously exposing the employee to the same noxious agent, the ordinary rules for proving causation fell to be varied as a matter of policy so as to ensure that an irrefutable claim against at least one of an ascertained group of defendants should not fail for want of any scientific possibility of identifying him. The ordinary rule, as the House agreed in Fairchild, was that the employee must prove that the damage was caused by the particular defendant sought to be held liable. As Lord Bingham said of the ordinary rule, at para 9:
The issue in these appeals does not concern the general validity and applicability of that requirement, which is not in question, but is whether in special circumstances such as those in these cases there should be any variation or relaxation of it.
He regarded the issue before the House as "an obvious and inescapable clash of policy considerations" (at para 33). He continued:
The crux of cases such as the present, if the appellants' argument is upheld, is that an employer may be held liable for damage he has not caused. The risk is the greater where all the employers potentially liable are not before the court. This is so on the facts of each of the three appeals before the House, and is always likely to be so given the long latency of this condition and the likelihood that some employers potentially liable will have gone out of business or disappeared during that period. It can properly be said to be unjust to impose liability on a party who has not been shown, even on a balance of probabilities, to have caused the damage complained of. On the other hand, there is a strong policy argument in favour of compensating those who have suffered grave harm, at the expense of their employers who owed them a duty to protect them against that very harm and failed to do so, when the harm can only have been caused by breach of that duty and when science does not permit the victim accurately to attribute, as between several employers, the precise responsibility for the harm he has suffered. I am of opinion that such injustice as may be involved in imposing liability on a duty-breaking employer in these circumstances is heavily outweighed by the injustice of denying redress to a victim. Were the law otherwise, an employer exposing his employee to asbestos dust could obtain complete immunity against mesothelioma (but not asbestosis) claims by employing only those who had previously been exposed to excessive quantities of asbestos dust. Such a result would reflect no credit on the law. It seems to me, as it did to Lord Wilberforce in McGhee  1 WLR 1, 7 that:
Lord Bingham concluded that all of the successive employers were liable.
Lord Nicholls of Birkenhead, expressing the same view, put the matter as follows at paras 41-42:
Lord Hoffmann, at para 63, said:
.... which rule would be more in accordance with justice and the policy of common law and statute to protect employees against the risk of contracting asbestos-related diseases? One which makes an employer in breach of his duty liable for the employee's injury because he created a significant risk to his health, despite the fact that the physical cause of the injury may have been created by someone else? Or a rule which means that unless he was subjected to risk by the breach of duty of a single employer, the employee can never have a remedy? My Lords, as between the employer in breach of duty and the employee who has lost his life in consequence of a period of exposure to risk to which that employer has contributed, I think it would be both inconsistent with the policy of the law imposing the duty and morally wrong for your Lordships to impose causal requirements which exclude liability.
As Lord Hoffmann pointed out, more clearly perhaps than any other member of the committee, it was essential that each of the successive employers should have wrongfully exposed the employee to asbestos particles and thereby materially increased the risk of his contracting the disease. The same policy would not therefore necessarily have justified a finding that all manufacturers of a drug causing injuries to patients were fixed with liability, simply because it was impossible to prove which manufacturer's product had been administered to the particular claimant. This was because "the existence of the additional manufacturers did not materially increase the risk of injury": see para 74. Lord Rodger of Earlsferry made the same point, at para 170:
.... part of the underlying rationale of the principle is that the defendant's wrongdoing has materially increased the risk that the claimant will suffer injury. It is therefore essential not just that the defendant's conduct created a material risk of injury to a class of persons but that it actually created a material risk of injury to the claimant himself .... the claimant must prove that his injury was caused by the eventuation of the kind of risk created by the defendant's wrongdoing .... By contrast, the principle does not apply where the claimant has merely proved that his injury could have been caused by a number of different events, only one of which is the eventuation of the risk created by the defendant's wrongful act or omission.
It should be observed that although the House was concerned with mesothelioma, it recognised that the legal issue was not necessarily peculiar to mesothelioma. It could arise in cases concerning other injuries or diseases or other sources of danger, provided that the damage was inflicted by the same destructive agent. The question, as they pointed out, had arisen in other jurisdictions whose law was reviewed by the House, in the context of groups of hunters, party-goers, footpads and the like negligently causing injury, each member of which had materially increased the risk of the injury which occurred without its being possible to identify whose negligence had actually caused it: see Lord Bingham at paras 25-29, Lord Hoffmann at paras 73-74, and Lord Rodger at paras 158-160. In McGhee v National Coal Board  1 WLR 1, which was held to have been founded on the same principle as Fairchild, the problem had arisen from the impossibility of determining the precise causal mechanism by which the claimant employee had contracted dermatitis, when some hypotheses involved a breach of duty while others did not. More generally, as Lord Bingham observed at para 34, "it would be unrealistic to suppose that the principle here affirmed will not over time be the subject of incremental and analogical development".
The decision in Fairchild has not given entire satisfaction to all of its authors. Lord Hoffmann has described it as "a revolutionary judgment". The ordinary function of the House of Lords in changing the common law is to modify some principle which had proved unsatisfactory. In Fairchild, the House did not modify or even criticise the general principle that the claimant had to demonstrate that the defendant's negligence had on a balance of probabilities caused the injury. Instead, they created a special exception to it which could not be justified by reference to any general principle and depended on a distinction which had no rational factual or legal justification: Hoffmann, "Constitutionalism and Private Law" (Cambridge Freshfields Law Lecture, 28 January 2015). Be that as it may, the decision in Fairchild is the starting point for any analysis of the legal issues arising between successive employers, or between employers and their insurers. In Fairchild itself, the House of Lords left those issues unresolved. Subsequent decisions of the House of Lords and Supreme Court have cruelly exposed the problem of dealing with complex and interrelated issues piecemeal. In order to accommodate the implications of earlier decisions for issues which they did not directly address, it has more than once proved necessary to subject their reasoning to some reanalysis.
In Fairchild, the House of Lords held that each of the successive employers was liable, but expressly declined to decide how, if at all, the liability was to be apportioned between them: see Lord Bingham at para 34, Lord Hoffmann at para 74, and Lord Rodger at para 125. That question did, however, arise in Barker v Corus UK Ltd  2 AC 572. The facts of Barker were that each of the claimants had been exposed to asbestos particles by successive employers or else by employers in one period and the claimant himself in another. The House of Lords held that the Fairchild principle applied in these cases also. Against that background, the question which arose was stated by Lord Hoffmann, at para 25, as follows:
whether under the Fairchild exception a defendant is liable, jointly and severally with any other defendants, for all the damage consequent upon the contraction of mesothelioma by the claimant or whether he is liable only for an aliquot share, apportioned according to the share of the risk created by his breach of duty.
The ordinary rule in the law of tort is that, where a number of defendants separately contribute to the same indivisible damage, each of them is jointly and severally liable for the whole. For want of a better word, this can be called the Dingle principle after Dingle v Associated Newspapers Ltd  2 QB 162, in which it received its classic formulation at the hands of Devlin LJ at paras 188-189:
Where injury has been done to the plaintiff and the injury is indivisible, any tortfeasor whose act has been a proximate cause of the injury must compensate for the whole of it. As between the plaintiff and the defendant it is immaterial that there are others whose acts also have been a cause of the injury and it does not matter whether those others have or have not a good defence. These factors would be relevant in a claim between tortfeasors for contribution, but the plaintiff is not concerned with that; he can obtain judgment for total compensation from anyone whose act has been a cause of his injury. If there are more than one of such persons, it is immaterial to the plaintiff whether they are joint tortfeasors or not. If four men, acting severally and not in concert, strike the plaintiff one after another and as a result of his injuries he suffers shock and is detained in hospital and loses a month's wages, each wrongdoer is liable to compensate for the whole loss of earnings. If there were four distinct physical injuries, each man would be liable only for the consequences peculiar to the injury he inflicted, but in the example I have given the loss of earnings is one injury caused in part by all four defendants. It is essential for this purpose that the loss should be one and indivisible; whether it is so or not is a matter of fact and not a matter of law.
Contracting mesothelioma is indivisible damage. If it had been proved that all of the successors had contributed to causing the employee's mesothelioma, they would have been jointly and severally liable for the whole damage on the Dingle principle. The question in Barker was whether the same principle applied when all that could be proved was that each employer had contributed to the risk without contributing to the disease. The trial judge and the Court of Appeal in Barker held that it did, and that each employer was jointly and severally liable. The House of Lords overruled them. It held by a majority (Lord Rodger dissenting) that liability was several, and fell to be apportioned according to the tortfeasor's relative contribution to the risk, measured by the duration and intensity of the exposure for which he was responsible.
The ratio of the decision may be taken from the speech of Lord Hoffmann, with whom Lord Scott of Foscote, Lord Walker of Gestingthorpe and Baroness Hale of Richmond agreed. Lord Hoffmann held that the Dingle principle could apply only if each employer had contributed to the employee contracting the disease or was deemed to have done so. But it could not be proved that they actually had done so, and Lord Hoffmann denied that Fairchild had introduced a rule that they were deemed to have done so by creating a material risk of contracting mesothelioma. That, he thought, had been the view of Lord Rodger and Lord Hutton in Fairchild. But he considered that the speeches of the majority were authority for the proposition that "the creation of a material risk of mesothelioma was sufficient for liability": see paras 31-34. From this he concluded, at para 35:
Consistency of approach would suggest that if the basis of liability is the wrongful creation of a risk or chance of causing the disease, the damage which the defendant should be regarded as having caused is the creation of such a risk or chance. If that is the right way to characterise the damage, then it does not matter that the disease as such would be indivisible damage. Chances are infinitely divisible and different people can be separately responsible to a greater or lesser degree for the chances of an event happening, in the way that a person who buys a whole book of tickets in a raffle has a separate and larger chance of winning the prize than a person who has bought a single ticket.
He went on at para 43 to summarise his reasons for regarding the apportionment of liability according to the time and intensity of the wrongful exposure for which each successive employer was responsible as representing the fair outcome:
In my opinion, the attribution of liability according to the relative degree of contribution to the chance of the disease being contracted would smooth the roughness of the justice which a rule of joint and several liability creates. The defendant was a wrongdoer, it is true, and should not be allowed to escape liability altogether, but he should not be liable for more than the damage which he caused and, since this is a case in which science can deal only in probabilities, the law should accept that position and attribute liability according to probabilities. The justification for the joint and several liability rule is that if you caused harm, there is no reason why your liability should be reduced because someone else also caused the same harm. But when liability is exceptionally imposed because you may have caused harm, the same considerations do not apply and fairness suggests that if more than one person may have been responsible, liability should be divided according to the probability that one or other caused the harm.
And at para 48:
Although the Fairchild exception treats the risk of contracting mesothelioma as the damage, it applies only when the disease has actually been contracted. Mr Stuart-Smith, who appeared for Corus, was reluctant to characterise the claim as being for causing a risk of the disease because he did not want to suggest that someone could sue for being exposed to a risk which had not materialised. But in cases which fall within the Fairchild exception, that possibility is precluded by the terms of the exception. It applies only when the claimant has contracted the disease against which he should have been protected. And in cases outside the exception, as in Gregg v Scott  2 AC 176, a risk of damage or loss of a chance is not damage upon which an action can be founded. But when the damage is apportioned among the persons responsible for the exposures to asbestos which created the risk, it is known that those exposures were together sufficient to cause the disease. The damages which would have been awarded against a defendant who had actually caused the disease must be apportioned to the defendants according to their contributions to the risk.
In the course of his analysis, at para 46, Lord Hoffmann referred to the implications of the alternative approach, which would have imposed joint and several liability:
The effect of the Civil Liability (Contribution) Act 1978 is that if each defendant is treated as having caused the mesothelioma as an indivisible injury and pays the damages in full, he will be able to recover contribution to the extent that he has paid more than his fair share of the responsibility from such other tortfeasors as are traceable and solvent. But he will in effect be a guarantor of the liability of those who are not traceable or solvent and, as time passes, the number of these will grow larger. Experience in the United States, where, for reasons which I need not examine, the DES rule of several liability has not been applied to indivisible injuries caused by asbestos, suggests that liability will progressively be imposed upon parties who may have had a very small share in exposing the claimant to risk but still happen to be traceable and solvent or insured: see Jane Stapleton, 'Two causal fictions at the heart of US asbestos doctrine' 122 LQR 189. That would, as I have said, not be unfair in cases in which they did actually cause the injury. It is however unfair in cases in which there is merely a relatively small chance that they did so.
Lord Scott, at para 61, put the same points in this way:
If the Fairchild principle were based upon the fiction that each Fairchild defendant had actually caused the eventual outcome, the analogy with tortfeasors each of whom had contributed to an indivisible outcome would be very close. But Fairchild liability is not based on that fiction. It is based on the fact that each negligent defendant has wrongfully subjected the victim to a period of exposure to an injurious agent and has thereby, during that period, subjected the victim to a material risk that he or she will contract the disease associated with that agent. Each successive period of exposure has subjected the victim to a further degree of risk. If, in the event, the victim does not contract the disease, no claim can be made for the trauma of being subjected to the risk: see Gregg v Scott  2 AC 176. But if the victim does contract the disease the risk has materialised. If the degree of risk associated with each period of exposure, whether under successive employers or during self-employment or while engaged in domestic tasks, were expressed in percentage terms, the sum of the percentages, once the disease had been contracted, would total 100%. But the extent of the risk for which each negligent employer was responsible and on the basis of which that employer was to be held liable would be independent of the extent of the risk attributable to the periods of exposure for which others were responsible. The relationship between the various negligent employers seems to me much more akin to the relationship between tortfeasors each of whom has, independently of the others, caused an identifiable part of the damage of which the victim complains. The joint and several liability of tortfeasors is based upon a finding that the breach of duty of each has been a cause of the indivisible damage for which redress is sought. No such finding can be made in a Fairchild type of case and the logic of imposing joint and several liability on Fairchild defendants is, in my opinion, absent. Moreover, Fairchild constitutes an exception, perhaps an anomalous one, to the causation principles of tortious liability. It should not, therefore, be found to be surprising if consequential adjustments to other principles of tortious liability become necessary.
Lord Walker, at para 113, drew attention to the fact that the Fairchild principle had involved a departure from ordinary rules of law, which called for the application of special principles of apportionment unique to the situation in which it applied:
The solution to the problem is in my opinion more radical, in line with the radical departure which this House has already made in Fairchild. That case was decided by the majority, as I have already noted, not on the fictional basis that the defendants should be treated as having caused the claimant's (or deceased's) damage, but on the factual basis that they had wrongfully exposed him to the risk of damage. The damage was indivisible, but the risk was divisible - a matter of statistics. In line with that new principle established or affirmed in Fairchild, and as a solution which does justice (so far as possible) both to the generality of claimants and to the generality of defendants, limited liability proportionate to risk is the better course for the law to take.
Baroness Hale made a similar point in her own speech, at paras 122 and 126-127:
But it does not necessarily follow from the fact that the damage is a single indivisible injury that each of the persons who may have caused that injury should be liable to pay for all of its consequences. The common law rules that lead to liability in solidum for the whole damage have always been closely linked to the common law's approach to causation. There is no reason in principle why the former rules should not be modified as the latter approach is courageously developed to meet new situations .... But in the Fairchild situation we have yet another development. For the first time in our legal history, persons are made liable for damage even though they may not have caused it at all, simply because they have materially contributed to the risk of causing that damage. Mr Stuart-Smith does not quarrel with the principle in Fairchild. He simply argues that it does not follow from the imposition of liability in such a case that each should be liable for the whole. I agree with the majority of your Lordships that indeed it does not follow. There is in this situation no magic in the indivisibility of the harm. It is not being said that each has caused or materially contributed to the harm. It can only be said that each has materially contributed to the risk of harm. The harm may be indivisible but the material contribution to the risk can be divided. There exists a sensible basis for doing so. Is it fair to do so? In common with the majority of your Lordships, I think that it is.
The speeches of the majority in Barker are not easy to analyse, and perhaps for that reason the analysis of them by Lord Rodger in his dissenting speech has proved influential. He attributed to Lord Hoffman and those who agreed with him the opinion that the employer was liable for creating a risk of contracting mesothelioma, and not for the mesothelioma itself. I do not think that this is the correct analysis of the majority's reasoning. In his essay in Perspectives in Causation, ed R Goldberg (2011), at p 8, Lord Hoffmann certainly adopted it. He suggested that the majority view in Barker had created a special cause of action for the causing of the risk. But his words in Barker itself were more circumspect. In the passage which I have cited from para 48 of his speech, he certainly suggested that the Fairchild exception treated the creation of the risk as the damage which gave rise to liability. But, like Lord Scott and Lord Walker, he emphasised that there was no cause of action for the risk in the absence of the disease. And Baroness Hale (at para 120), while agreeing with Lord Hoffmann, had no difficulty in agreeing with Lord Rodger that "the damage which is the 'gist' of these actions is the mesothelioma and its physical and financial consequences. It is not the risk of contracting mesothelioma". In my opinion, the natural reading of the speeches of the majority, read as a whole, is that the Fairchild exception is an exception to the ordinary rules of causation alone. It treats a material contribution to the risk as enough to discharge the burden of proving that the breach of duty has caused the disease. It followed that by reason of having contributed to the risk the employer was liable for the disease itself. Or, as Lord Walker put it at para 109, the Fairchild exception is "A rule of law by which exposure to risk of injury is equated with legal responsibility for that injury". The real difference between Lord Rodger and the majority was that the majority thought that it was unknowable and irrelevant who had caused the disease to develop. They considered that each successive employer should be liable in proportion to the significance of its contribution to the risk because, exceptionally, what had made each of them liable for the disease was its contribution to the risk and not its contribution to the damage. Lord Rodger on the other hand thought that each successive employer was deemed to have contributed to the damage and that it was that contribution to the damage which was the source of the liability. He therefore thought that each of them incurred the joint and several liability which the Dingle principle imposed on those who severally contribute in different degrees to the same damage. Underlying his reasoning was an expressed reluctance to adopt an analysis of Fairchild which made the cases in which it applied into an "enclave" subject to rules quite different to those which applied generally in the law of personal injuries: see para 85. The majority on the other hand considered that Fairchild had already created the enclave and that the task in hand was to devise a basis of liability consistent with its peculiarities.
Within three months of the decision in Barker, its effect was reversed by section 3 of the Compensation Act 2006. Section 3 applied in any case where a person ("the victim") contracted mesothelioma as a result of exposure to asbestos, and another ("the responsible person") was liable in tort for having wrongfully exposed him to asbestos, "whether by reason of having materially increased a risk or for any other reason". Section 3(2) provided that the responsible person was liable irrespective of whether the victim was also exposed to asbestos on other occasions, either by other tortfeasors or in circumstances where there was no liability in tort. Not only was that person liable, but he was jointly and severally liable with any other responsible person. The result was to make each "responsible person" liable for the whole damage, without prejudice (see subsections (3) and (4)) to the right of contribution between them. Section 3 applied retrospectively: see section 16(3).
The position as between the employer and his insurer: Trigger
Mesothelioma, like other industrial diseases characterised by long periods of latent development, poses particular problems for insurers writing employer's liability business on an occurrence basis. None of the cases which I have cited was concerned with the impact of the Fairchild exception on coverage under an employer's liability insurance. That question arose in the Employer's Liability Policy Trigger litigation, six cases heard together before the Supreme Court and reported under the title Durham v BAI (Run off) Ltd  1 WLR 867. A number of different policy forms were before the court. They all insured the employer's liability for personal injury (including disease) "contracted" or "sustained" during the period of insurance, generally a year. There were two issues. The first was whether the "trigger" for the insurers' liability was the exposure of the employee to asbestos (as the employers and the personal representatives of deceased employees contended), or only to the development or manifestation of mesothelioma (as the insurers contended). These alternatives were referred to as the "causation basis" and the "manifestation basis" respectively. It was clear that the policies did not respond if the manifestation basis was correct, for the disease developed or manifested itself long after the relevant policies had expired and generally after the victim had ceased to be employed. The second issue was raised in terms not by the parties but by Lord Phillips of Worth Matravers in the course of the hearing. It was whether, if the causation basis was correct, the triggering event could be shown in the current limited state of scientific knowledge to have occurred during the policy period. If not, it was suggested, the insurer could not be liable at all.
The leading judgment was delivered by Lord Mance. The court's decision on the first issue is summarised at paras 49-51 of his judgment. It was held that the policies insured the damage attributable to the actual injury or disease, which was suffered when mesothelioma developed. But the triggering event which had to occur within the policy period was the event upon which mesothelioma was "sustained" (the term used in the Midland policies in the present appeal) or "contracted" (the term used in some other policies). In either case, that happened when it was "caused or initiated, even though it only developed or manifested itself subsequently". The whole panel was agreed upon this.
The second issue turned on the effect of Fairchild and Barker on the footing that causation or initiation of the disease was the relevant triggering event. This question divided the panel. Lord Phillips in his dissenting judgment held that the insurers could be found liable only if the effect of these decisions was that the employer was deemed to have caused the development of the disease by exposing the employee to asbestos particles. That analysis of Fairchild had, however, been rejected by the majority in Barker. The alternative in his view was to treat Fairchild as creating liability not for the disease but for the contribution to the risk of the disease. Since the contribution to the risk was not an insured peril, the insureds and their statutory assignees could succeed against the insurers only if they demonstrated that the disease had in fact been caused or initiated during the policy period, something which the current state of scientific knowledge made it impossible for them to do. This view was rejected by the majority. Again, the reasons may be taken from the judgment of Lord Mance. He agreed that the deemed causation theory had been rejected in Barker. He held that the employer was not liable for merely exposing the victim to the risk. He was liable for the disease. But he was liable for the disease, because his tortious exposure of the victim to the risk was in law enough to establish that he had caused it.
The issue, as Lord Mance put it at para 66, concerned:
the meanings we assign to the concept of causation, first in the context of considering employers' liability to their employees and then in considering the scope of employers' insurance cover with respect to such liability.
He summarised the effect on the policy at paras 73-74 as follows:
In substance, the result was that the same "weak" test of causation which applied as between the victim and the employer should be applied as between the employer and his liability insurer.
The parties' arguments
This appeal is not concerned with multiple successive causes of exposure to asbestos, nor is it concerned with multiple successive employers. Guernsey Gas, for whose liabilities IEG is responsible, employed Mr Carré throughout the 27-year period when he was tortiously exposed to asbestos.
IEG's case for recovering in full against the insurers who were on risk for six of those years is as follows. The decision of this court in the Trigger appeals established
that the policy responds if during the period of insurance something happened which caused the ultimate development of mesothelioma, and
that that causal link is sufficiently demonstrated by proving that during the period of insurance the insured employer wrongfully exposed the employee to the risk of contracting mesothelioma.
Therefore, it is said, just as an employer is liable if he employed the victim at any time when he was wrongfully exposed to the risk of contracting mesothelioma, so the employer's liability insurer is liable if he was on risk at any time when such exposure occurs. If by statute the employer must bear the entire loss attributable to the disease by reason of having exposed the employee to asbestos particles at any time that is also the measure of his claim against the insurer.
Zurich advances two alternative contentions in response to this. The first is that as between the victim and his employer Barker remains good law in all cases to which the Compensation Act 2006 does not apply. It therefore remains good law in Guernsey, where Mr Carré was employed. It follows that the liability of Guernsey Gas is apportionable over the period of exposure, and that the insurer is liable only for a proportionate part of the loss representing that part of the period of exposure during which he was on risk. If, contrary to this submission, Barker is no longer good law even in Guernsey, Zurich concedes that it is liable under the policy terms for the whole of Mr Carré's loss. On that footing, Zurich's second argument is that they have a right of equitable recoupment against the other insurers pro rata to their respective periods on risk, and against the employer for that proportion of the claim which reflects the time he was uninsured. So far as the compensation element of the claim is concerned, the second argument arises only if the first one fails. But as far as the claim for defence costs is concerned, the second argument arises anyway, because Zurich accepts that they were contractually liable for the whole of the defence costs. This is because the same defence costs would have been incurred whether the employer was liable for the whole loss or only a proportion of it.
The decisions of Cooke J and the Court of Appeal
Cooke J accepted Zurich's first argument. He held that the insurer was liable only for a rateable proportion reflecting time on risk. The alternative claim for recoupment therefore did not arise. But if it had arisen, Cooke J would have rejected it.
The Court of Appeal reversed him on Zurich's primary case. They held that Barker was no longer good law after the Compensation Act, and therefore no part of the common law of Guernsey. It followed in their view that each successive insurer was liable for the entire loss. They considered that no allowance fell to be made for the substantial periods of exposure when it was not on risk, whether that was because other insurers were on risk or because the employer elected to bear the risk itself. Both Toulson LJ and Aikens LJ, who both delivered reasoned judgments, considered that the issue was concluded by Trigger, in particular the statement of principle in the judgment of Lord Mance at para 73. Both of them thought that once it was accepted that each insurer's liability was triggered by any period of exposure during which it was on risk, it followed as a matter of course that each insurer was liable for the entire loss. Citing the decision of the United States Court of Appeals for the District of Columbia in Keene Corporation v Insurance Company of North America 667 F2d 1034, Toulson LJ expressed at para 42 his agreement with the proposition that:
.... once it is accepted that exposure during any policy period met the causal requirement for the employer's liability to the victim, for which the employer was potentially entitled to indemnity from the insurer under the terms of the relevant policies, to withhold part of that indemnity from the employer on account of its conduct in other years would be to deprive the employer of insurance coverage for which it paid.
Aikens LJ agreed, adding at paras 53-54 what is perhaps implicit in Toulson LJ's judgment and may stand as the essence of the court's reasoning:
If an employer is liable to his employee for his employee's mesothelioma following upon a tortious exposure to asbestos created during an insurance period, then, for the purposes of the insuring clause in the employers' liability policy, the disease is 'caused' within the insurance period. This is because it is sufficient that there is what Lord Mance calls (following Hart & Honoré's use of the phrase) a 'weak' or 'broad' causal link, in this case between the exposure to the asbestos during the insurance period and the employee's eventual contraction of the mesothelioma. Once that causal requirement is fulfilled, then the employer will have proved that the mesothelioma (the disease) was 'caused during any period of insurance'. It follows from the policy wording that the insurer is then liable to indemnify IEG for 'all sums for which the Insured shall be liable in respect of any claim for damages for .... such disease' (my emphasis). In other words, Zurich will be liable to indemnify IEG for the whole of the damages paid out by IEG in respect of Mr Carré's claim for damages for contracting mesothelioma, not just a proportion worked out by reference to the period during which IEG was covered by policies for which Zurich is responsible.
Like Cooke J, the Court of Appeal rejected the recoupment argument.
The decision of the Court of Appeal created consternation among the interests represented by the Association of British Insurers. This was mainly because the decision recognised a right in an insured employer to recover in full from any insurer on risk at any time during the period of tortious exposure. In practise this meant that employers could be expected to pick off the "easiest target". This undermined the industry settlement scheme, which:
apportioned the liability by time on risk among all insurers on risk during the period of exposure;
apportioned uninsured periods to the employer if he was solvent; and
left the employer to claim against an insolvent insurer under the compensation scheme established under section 213 of the Financial Services and Markets Act 2000.
These consequences of the decision would be aggravated if there was no right of contribution between insurers. The result, the Association of British Insurers submitted, would be to encourage insurers to be more vigorous in defending claims, to delay settlements and potentially to cause difficulties with reinsurers. In addition, the Court of Appeal's reliance on the "all sums" wording of the insuring clause opened up the prospect that insurers might be held liable in full even in the case of "divisible" diseases where the contribution of the tort to the actual development of the disease was more readily assignable to distinct policy periods.
The issues in the Supreme Court
In this court, the parties' arguments were the same as they were in the courts below. However, after the case had been argued for the first time before five justices, the court raised a number of further questions with the parties which expanded the scope of the argument. We directed that the case should be reargued before seven justices so that those questions could be considered. The matters raised by the court included the correctness of Zurich's concession that if their first argument failed they were contractually liable (subject to recoupment) for the whole loss. An alternative possibility was that the insurer was liable for a proportionate part of the loss as a matter of construction of the policy, whether Barker remained good law or not and even in England where the Compensation Act applied. Upon reargument, Zurich addressed the construction question but maintained its concession. It was supported in this line by the Association of British Insurers.
Construction of the policy
I turn first to the construction of the policy, partly because it is the natural starting point for any analysis of its effect, and partly because I do not accept the construction which the parties have adopted as their premise.
The six annual policies written by Midland were issued between 1982 and 1988. At that time, Fairchild, the Compensation Act and their legal progeny lay well into the future. These developments have greatly increased the potential liability of employers to employees whom they have wrongfully exposed to asbestos but that, as everyone can agree, is an ordinary hazard of liability insurance. The policies respond to the liability incurred by the insured in the course of the employment of its employees as the law may from time to time determine it to be, whether or not that liability would have been anticipated at the time that the contract of insurance was made. That, however, is not the problem with which we are presently concerned. We are concerned with the construction of the policies themselves. They cannot be construed on the footing that the parties were contracting by reference to the extraordinary legal problems to which Fairchild and its progeny have given rise. In Philips Electronique Grand Public SA v British Sky Broadcasting Ltd  EMLR 472, Sir Thomas Bingham MR observed at paras 481-482:
The courts' usual role in contractual interpretation is, by resolving ambiguities or reconciling apparent inconsistencies, to attribute the true meaning to the language in which the parties themselves have expressed their contract .... The question of whether a term should be implied, and if so what, almost inevitably arises after a crisis has been reached in the performance of the contract. So the court comes to the task of implication with the benefit of hindsight, and it is tempting for the court then to fashion a term which will reflect the merits of the situation as they then appear. Tempting, but wrong .... it is not enough to show that had the parties foreseen the eventuality which in fact occurred they would have wished to make provision for it, unless it can also be shown either that there was only one contractual solution or that one of several possible solutions would without doubt have been preferred. ....
Each Midland policy recited that the insured had applied for insurance and had paid or agreed to pay the premium "as consideration for such insurance during the period stated in the Schedule or for any subsequent period for which the Company shall have accepted the premium required for renewal of this policy". The insuring clause provided:
If any person under a contract of service or apprenticeship with the Insured shall sustain any bodily injury or disease caused during any period of insurance and arising out of and in the course of his employment by the Insured in the business above mentioned, the Company will indemnify the Insured against all sums for which the Insured shall be liable in respect of any claim for damages for such injury or disease settled or defended with the consent of the Company. The Company will in addition pay claimants' costs and expenses and be responsible for all costs and expenses incurred with the consent of the Company in defending any such claim for damages.
The insuring clause makes explicit what would be implicit in any contract of liability insurance written on an occurrence basis for a limited period. The occurrence is not the mere exposure of the victim to asbestos. It is the "sustaining" of "bodily injury or disease caused during any period of insurance". The indemnity extends to the insured's liability for damages for "such" injury or disease, ie injury or disease caused during the period of insurance. The insurance is expressed to apply only to liability in respect of any injury or disease caused in Great Britain, Northern Ireland, the Isle of Man and the Channel Islands, and to injury or disease sustained by employees temporarily employed elsewhere under a contract of service or apprenticeship entered into in one of those jurisdictions.
A liability policy responds to the specified liabilities of the insured, but only subject to any overall limitations of the policy. One of these limitations is the period of insurance, which is a fundamental feature of any such policy. The whole of the insuring clause depends upon the assumption that it is possible to assign the time when an injury or disease was caused to a given period which either is or is not within the period of insurance. Either the damage will be divisible, in which case parts of it may have been caused in different periods and must be divided between those periods, or it will be indivisible, in which case it will have been caused in a single period. As the opening recital reminds us, the period of insurance is a critical element of the ex ante assessment of the risk on which the premium is based. Insurance for any further period is dependent on renewal and the payment of a further premium. It may also (although not in this policy) be critical to the application of a deductible or an aggregate annual limit or excess. In addition, the attribution of loss to particular years is likely to have a significant effect on an insurer's reserving and his reinsurance.
In the English case-law the point has commonly been made in the context of reinsurance. In Municipal Mutual Insurance v Sea Insurance Co Ltd  Lloyd's Rep IR 421, a port authority was insured against liability for (among other things) damage to property in its custody. Damage was done to equipment stored with it by a succession of independent acts of vandalism over a period of 18 months. It was impossible to differentiate between one act of vandalism and another, and the port authority was held to be entitled to aggregate all of them and to make a single claim against its insurer for the whole. The insurer was reinsured under successive facultative annual reinsurances, on terms which were back to back with the direct insurances and contained a standard follow clause ("to follow their settlements"). This gave rise to difficulty when the claim was passed on as a single claim to the reinsurers, because the 18-month period when the damage was done extended over the periods covered by three successive annual reinsurance policies written by different insurers, each of which provided for a substantial excess. The insurers' attempt to make a single aggregate claim on one reinsurance policy was rejected by the Court of Appeal. Hobhouse LJ, giving the only reasoned judgment, said at paras 435-436:
.... it was incumbent upon the judge to recognise .... and give effect to the essentially annual character of each reinsurance contract. Applying the wording of the original policy to each reinsurance contract it is necessary to ask whether or not the relevant physical loss or damage arose during the relevant period of cover. The judge came to the surprising conclusion that each reinsurance contract covered liability in respect of physical loss or damage whether or not it occurred during the period covered by the reinsurance contract and he went on expressly to contemplate that the same liability for the same physical loss or damage might be covered under a number of separate contracts of reinsurance covering different periods. This is a startling result and I am aware of no justification for it. When the relevant cover is placed on a time basis, the stated period of time is fundamental and must be given effect to. It is for that period of risk that the premium payable is assessed. This is so whether the cover is defined as in the present case by reference to when the physical loss or damage occurred, or by reference to when a liability was incurred or a claim made. Contracts of insurance (including reinsurance) are or can be sophisticated instruments containing a wide variety of provisions, but the definition of the period of cover is basic and clear. It provides a temporal limit to the cover and does not provide cover outside that period; the insurer is not then 'on risk'. It will be appreciated that the judge's suggestion that there could or should be contribution between those signing the different slips for the different years is likewise radically mistaken.
In Wasa International Insurance Co Ltd v Lexington Insurance Co  1 AC 180, Lexington had insured an aluminium manufacturer for a single period of three years between 1977 and 1980 against property damage. The insured incurred large liabilities for environmental clean-up costs. The clean-up costs were necessitated by industrial pollution occurring since the early 1940s. It claimed indemnity for the entire loss from each successive insurer by whom they had been insured against property or liability risks between 1956 and 1985, including Lexington. The claim was heard in Pennsylvania under Pennsylvania law. The courts there held that each insurer was jointly and severally liable for all damage which was "manifest" during their period of insurance irrespective of when it occurred. This meant substantially all the pollution damage attributable to industrial operations not only during the period of insurance but over the previous three decades. Lexington settled with the aluminium company on that basis. The reinsurance was on the same terms as the original as to period and coverage. It also contained a "follow the settlements" clause. But it was governed by English law, under which liability would have been limited to damage caused during the period of insurance, whereas the Pennsylvania court applied its own law under which no such limit applied. The argument for Lexington was that the Pennsylvania courts had decided that the pollution damage occurring over the whole period was insured under the 1977-1980 policy and that the reinsurance, which was on the same terms save as to the proper law, must respond on a like basis. The House of Lords rejected this contention. They held that, notwithstanding the ordinary presumption that reinsurance was back-to-back with the underlying insurance, the reinsurer's liability was limited to damage caused between 1977 and 1980. The leading speeches were delivered by Lord Mance and Lord Collins of Mapesbury. Lord Mance said, at paras 40-41:
Lord Collins said, at para 74:
Lord Brown of Eaton-under-Heywood, concurring with Lord Collins, pointed out at para 15 that if Lexington's argument were correct, the reinsurers would have incurred the same liability if they had been on risk for only three months instead of three years:
Given the fundamental importance under English law of the temporal scope of a time policy, I find it impossible to construe the reinsurance contracts in the way contended for.
Reinsurance is not an insurance on liability, but on the original risk. In Municipal Mutual the original risk was the insured's liability for property damage and in Wasa it was the property damage itself. But the principle stated in them is the same, and it is of general application, as Hobhouse LJ pointed out. The courts are bound to give effect to the contractual limitations on the insurer's liability. In particular, they are bound to give effect to the chronological limits of the risks covered, and to those provisions of the contract that operate by reference to the insurance period. The question on this appeal is how the terms of a chronologically limited policy are to apply to the liability resulting from the decision in Fairchild and the Compensation Act 2006.
The objection to construing the Midland policies in this case as covering the damage caused at any time during the 27 years in which Mr Carré was exposed to the risk of contracting mesothelioma is the same as the objection of the Court of Appeal in Municipal Mutual and the House of Lords in Wasa to the corresponding arguments in those cases. The consequences are both commercially absurd and entirely inconsistent with the nature of annual insurance. The longer an employee is exposed to asbestos, the greater the risk of his contracting mesothelioma at some stage in his life. The result of IEG's argument is that under the contract the financial consequences for the insurer of writing the contract for a single year are the same as the financial consequences of writing the risk for the full 30 years, although he only receives a single year's premium in the former case and 30 years' premium in the latter. Indeed, the consequences would be the same even if the insured had been held covered for a time on risk premium for just a week or two while an unsuccessful attempt was made to agree terms. This entirely severs the functional connection between premium and risk. The employer for his part would obtain cover in respect of those whom he employed and exposed to asbestos particles in the period of cover, notwithstanding that for the rest of their working lives he elected to insure with others, or indeed elected not to insure at all. On that footing, the insurer assumes a liability of indeterminate duration notwithstanding that he expressly limited his liability to a single year. The indeterminate duration of the liability would extend both backwards and forwards. Thus an insurer who wrote a policy for, say, the first year of compulsory insurance, 1972, for an employer who had exposed its employees to asbestos particles for the previous half-century and continued to do so, would assume liability for the entire accumulated legacy of exposure in the case of all employees on its payroll at the inception of the policy however far back the exposure of those employees extended. An insurer who insured the employer for a single year but refused to renew because of unfavourable claims experience or an increase in the risk would nevertheless remain liable in respect of the exposure of existing employees for an indefinite period into the future without payment of any further premium. Moreover, the insurer of a single year would have to pick up the tab for every other insurer who was on risk over an indeterminate period, although he had assumed a liability which was not co-ordinate with theirs because they covered distinct periods. It also would mean that where the terms of successive policies were different, for example as to the excess or the limit, the insured could select a policy and spike the whole of the loss into the period covered by it. In the course of his judgment in the Court of Appeal, Toulson LJ observed that awarding less than the whole loss against any one insurer would deprive the insured of the insured coverage for which it paid. This observation seems to me to be the reverse of the true position. An employer who has paid a single year's premium has not paid for 27 years of cover, which is what the decision of the Court of Appeal gives him.
I understand every member of this court to be agreed that these consequences are unacceptable. As Lord Mance points out at para 40 of his judgment, the insurance was "placed on the basis that a particular liability or loss would fall into one, not a series of separate periods. If an insured wanted complete cover, it would have to maintain it for all such periods". At para 43, he draws attention to the consequences which I have summarised above, and describes them as "contrary to principle" and "anomalous". He is, with respect, plainly right to do so. These consequences are not just remarkable in themselves, but are directly inconsistent with the language of the Midland policies and the fundamental characteristics of insurance. This is not because any of the elements of liability, such as causation or damage, is divisible by time. Plainly they are not. It is because once the insured has proved each of those elements, he must still show that the occurrence fell within the chronological limits of the policy. If a particular result is inconsistent with the nature of insurance, and with the basis on which annual insurance is placed, there must be the strongest possible presumption that it was not intended, in the absence of clear language showing that it was.
To explain why IEG's submission is mistaken, it is first necessary to differentiate between the legal basis of an employer's liability to his employee and the legal basis of the insurer's liability to the employer. At common law, the Dingle principle is that if several people tortiously contribute by independent acts to the same damage, they are all jointly and severally liable for the whole of the resulting damage. In Barker, the Court of Appeal and Lord Rodger in his dissenting speech in the House of Lords likened this state of affairs to the situation where several employers successively exposed the same victim to the risk of contracting mesothelioma. The majority of the House rejected that analogy, but the effect of section 3 of the Compensation Act was to reinstate it. The result is that each employer is contributing to the risk all the time, and is therefore incurring liability all the time. This makes some sense as between successive employers who are guilty of a continuous tort. However, the same logic cannot be applied as between successive insurers. Insurers are not wrongdoers. They have not contributed to any tortiously inflicted damage. The principles on which they are liable to indemnify their insured are not affected by the Compensation Act. Their liability depends not on common law or statutory concepts of culpability but on the liability that they have agreed to assume by contract. Although they have contracted to indemnify the insured in respect of his liability, they have done so on terms which require the assignment of causation to a contractual period and limit their liability to that period. This raises a problem which is, essentially, not legal but factual. The Fairchild principle is the law's response to the factual certainty that the disease was caused during the period of exposure combined with a complete uncertainty about when. If the assignment of causation to a particular period of coverage is scientifically impossible, then one solution would be for the law to say that the insured has not proved his case, as Lord Phillips would have held in Trigger. The alternative, once that is rejected, is to devise a mode of assigning causation to a particular period of time which is the closest possible surrogate for the real thing. The majority in Trigger adopted the latter solution, holding that any period of tortious exposure to the risk of contracting mesothelioma was enough to establish that the employer had caused the disease if it subsequently developed. The employer's liability insurer was liable on that basis.
The fallacy of IEG's argument is that it assumes that because any period of tortious exposure to the risk of contracting mesothelioma is enough to establish causation of the disease, it must follow that the disease was successively caused in every period of exposure. But that is conceptually impossible. Mesothelioma is caused only once. Once the process by which it develops has been initiated, subsequent further exposure to asbestos will not aggravate the victim's condition or increase the loss. Pursuing the example of an employee exposed to asbestos particles for 30 years, let us assume that a different insurer is on risk in each year of exposure. If IEG is right, each insurer is liable for the entire loss in respect of an employee exposed to asbestos in his year who subsequently contracts mesothelioma, subject only to the limitation that the insured cannot recover more than an indemnity. By the same token, if the same insurer was on risk throughout the period of exposure, that insurer would be liable for the entire loss in each year, subject to the same limitation. But this makes no sense. It is conceptually possible for an insurer to be liable on the footing that there is a chance that the disease was caused in any year and that that should be enough to establish the necessary causal link. It is not conceptually possible for an insurer to be liable on the footing that the disease was actually caused in every year. It is only when one aggregates every successive period that the chances add up to 100%.
IEG's answer to this is that because for the purposes of the insuring clause Trigger equates exposure to the risk with causation of the disease, it follows from the fact that the risk operated continuously throughout the period of exposure that the disease was continuously caused throughout the period of exposure. Therefore, it is said, causation of the disease is at one and the same time
a single indivisible occurrence, resulting in the entire claim falling into a single policy year, and
a continuing occurrence extending over every policy year and equally efficacious in causing the disease in each one.
I would be reluctant to assume that any judicial decision was authority for a contradiction in terms, and I do not think that Trigger is authority for this one. The effect of Trigger is that the insurer's liability is triggered in each insurance year during the period of exposure. This is not because the insurance is against the exposure to the risk, a proposition which the court was at pains to reject in Trigger, just as the House of Lords had previously rejected it in Barker. Nor is it because the disease was actually caused in each insurance year, which is logically impossible and in any event ex hypothesi unknowable. It is because exposure to the risk is the closest surrogate that can be devised for determining when the disease was caused. This is the meaning of the "weak" or "broad" causal link to which Lord Mance referred at para 74 of Trigger. The link is, as he put it, "to exposure which may but cannot be shown on the ordinary balance of probabilities to have played a role in the actual occurrence of the disease".
The theory that an insurer is liable in respect of any year of insurance when the employee was exposed to the risk of contracting mesothelioma is a perfectly satisfactory answer to the question whether the insurer is liable at all, which was the only relevant question at issue in the Trigger litigation. But it cannot be applied without modification when the question is how much of the loss is attributable to particular years. If, as Trigger teaches, the insurer's liability is triggered in each policy year, the rational response of the law is not to assign the whole of that loss to a policy year of the insured's choice. That would be to assume that the whole loss was caused in that year, whereas the law proceeds from the premise that we cannot know that. The rational response is that the loss must be prorated between every policy year during which the insured employer exposed the victim to asbestos. In my opinion, once one rejects the conclusion that the insurer is not liable at all, proration on that basis is the only way of giving effect to the overriding requirement of each annual policy that the liability should be assigned to policy years. If exposure to the risk of contracting mesothelioma is equated with causation, the natural consequence is that the resultant liability falls to be apportioned to policy years according to the duration and intensity of the exposure. What is being prorated as between the insurer and the employer is the employer's liability, not the indivisible harm of the mesothelioma itself. The chances of contracting mesothelioma, as Lord Hoffmann observed in Barker, are infinitely divisible, even if mesothelioma itself is not.
This conclusion does not, as it seems to me, require words to be read into the policy, any more than the "weak" or "broad" test of causation adopted in Trigger required words to be read into the policy. It simply involves, as Trigger involved, construing the words "caused during any period of insurance" in the light of the terms of the policy as a whole and applying them to an insured liability with the unusual legal incidents of an employer's liability for mesothelioma.
I can deal very shortly with the words "all sums" in the insuring clause, on which Aikens LJ relied to support his conclusion. The relevant phrase is not "all sums" but "all sums for which the insured shall be liable in respect of any claim for damages for such injury or disease", ie for "injury or disease caused during any period of insurance". The insurance does not cover all sums for which the insured may be liable, but only those which fall within the chronological limits of the risk which the insurer has assumed.
I have concentrated on the case where there is a single culpable employer whose operations are the sole relevant source of exposure to asbestos particles, because those are the facts of the present case. But there is no particular difficulty in applying the same principle to cases where there are successive tortfeasors or successive sources of exposure. The liability of the employer to the victim is apportioned to the insurer according to the proportion which its period on risk bears to the whole period during which that employer has tortiously exposed the victim to asbestos. If the insured employer is jointly and severally liable to the victim under section 3 of the Compensation Act with earlier employers who exposed the same victim to asbestos, that liability will form part of the liability which falls to be prorated between his successive insurers or between them and himself in respect of periods of non-insurance. If the insured employer is insured throughout the period during which he exposed the victim to asbestos, the insurers will be liable for their respective proportions of 100%. Likewise, if there is another source of exposure to asbestos (for example ambient environmental asbestos) which were to be held to reduce the insured employer's liability, the liability passed on to his insurer will be correspondingly reduced, but if not, not.
United States Authority
This conclusion, which appears to me be a logical application of the insuring clause to the kind of liability which arises in this case, derives some support from the rich jurisprudence of the United States, where similar questions have frequently come before the courts in the context of asbestosis and environmental pollution claims. Insurance is governed by state law and there are, perhaps inevitably, significant differences of approach in different state jurisdictions. In the celebrated case of Keene Corporation v Insurance Company of North America 667 F2d 1034 the United States Court of Appeals for the District of Columbia (applying the laws of Delaware, New York, Pennsylvania, Connecticut and Massachusetts) held that either exposure or manifestation of the disease would make the insurer liable, and that each insurer was jointly and severally liable for the whole loss. The court's decision on the latter point was endorsed by Toulson LJ in his judgment in the Court of Appeal in the present case: see para 42. It was based mainly on the reasonable expectations of policyholders, a consideration which, except as background to the construction of the policy, does not have the significant place in English insurance law as it has in many jurisdictions of the United States. So far as it was based on the language of the policy at all, the imposition of joint and several liability in Keene was based on the expression "all sums which the insured shall become legally obligated to pay as damages because of bodily injury": see note 20 (emphasis added). I have given my reasons for regarding the corresponding words as inconclusive in the context of the Midland policies.
So far as Keene is authority for a "triple" or "continuous" trigger in cases about insurers' liability for latent industrial diseases, it has been widely followed in other jurisdictions of the United States. But so far as it imposes joint and several liability on successive insurers, it has not met with universal acceptance, and major insurance jurisdictions have rejected it. In Insurance Company of North America v Forty-Eight Insulations Inc 633 F2d 1212, the Sixth Circuit Court of Appeals applied the laws of Illinois and New Jersey to a dispute about the allocation of a loss among successive insurers and the insured itself (in respect of periods of "self-insurance"). The court construed product liability policies in respect of "bodily injury" as covering latent diseases on an exposure basis. The insured conceded that the insurers' liability fell to be prorated according to time on risk, leaving them with a rateable part representing the period of exposure when they were uninsured. The issue was, however, argued out on the question whether the same rule applied to defence costs, which the employer did not concede. By a majority, the court ordered the proration of the defence costs, observing at para 73:
In an underlying asbestosis suit, the plaintiff must show that Forty-Eight's products injured him in order to be able to maintain a cause of action against Forty-Eight. Under Borel, Forty-Eight would be jointly and severally liable along with the other asbestos manufacturers: 493 F 2d at 1094-96. However, in allocating the cost of indemnification under the exposure theory, only contract law is involved. Each insurer is liable for its pro rata share. The insurer's liability is not 'joint and several', it is individual and proportionate. Accordingly, where an insurer can show that no exposure to asbestos manufactured by its insured took place during certain years, then that insurer cannot be liable for those years. The reason is simple: no bodily injury resulting from Forty-Eight's products, took place during the years in question. The same thing would be true if an insurer could show that a worker used an effective respirator during certain years. Again, no 'bodily injury' would have taken place.
In my view, this analysis of the reason why the compensation element of the claim falls to be prorated cannot be faulted, although for reasons which I shall explain I do not think that it can be applied to defence costs.
In Owens-Illinois Inc v United Insurance Company (1994) 138 NJ 437, another product liability insurance case, the Supreme Court of New Jersey reached a similar conclusion on the assumption that the insurer's liability was continuously triggered throughout the period of exposure, but rejected the solution proposed in Keene that each insurer on risk during that period was liable for the entire loss. Instead, it proposed a complex system of proration. At p 468, the court observed:
The occurrence clauses undoubtedly contemplated indemnity for provable damages incurred by the policyholder because of injury that occurred during the policy period. The continuous-trigger theory coupled with joint-and-several liability is premised on a tenuous foundation: that at every point in the progression the provable damages due to injury in any one of the years from exposure to manifestation will be substantially the same (the collapsed accordion). As we have seen, our law has been developing in a different manner.
The court found little assistance in the language of the contract, but concluded that for reasons essentially of policy and practical efficacy, proration was the appropriate solution. It was particularly concerned with the anomaly that the Keene solution placed an insured with insurance for a small part of the period of exposure in the same position as one with insurance for all of it. At p 473, the court said:
.... the Keene rule of law reduces the incentive of the property owners to insure against future risks. Recall the circumstances in the final three years .... Assuming the availability of insurance, a principle of law that would act as a disincentive to the building owners in the hypothetical might serve in the long run to reduce the available assets to manage the risk. O-I's counsel counters that these are not correct assumptions about the way in which the 'real world' responds. We cannot be sure that the policy will be effective. We believe, however, that the policy goal is sound. Finally, principles of simple justice cannot be entirely discounted. To rebut effectively the question posed in Forty-Eight Insulations is difficult. 'Were we to adopt [the policyholder's] position on defence costs a manufacturer which had insurance coverage for only one year out of 20 would be entitled to a complete defence of all asbestos actions the same as a manufacturer which had coverage for 20 years out of 20. Neither logic nor precedent support such a result.'
And at p 479:
Because multiple policies of insurance are triggered under the continuous-trigger theory, it becomes necessary to determine the extent to which each triggered policy shall provide indemnity. 'Other insurance' clauses in standard CGL policies were not intended to resolve that question. A fair method of allocation appears to be one that is related to both the time on the risk and the degree of risk assumed. When periods of no insurance reflect a decision by an actor to assume or retain a risk, as opposed to periods when coverage for a risk is not available, to expect the risk-bearer to share in the allocation is reasonable. Estimating the degree of risk assumed is difficult but not impossible. Insurers whose policies are triggered by an injury during a policy period must respond to any claims presented to them and, if they deny full coverage, must initiate proceedings to determine the portion allocable for defence and indemnity costs.
In Consolidated Edison Company of New York Inc v Allstate Insurance Company (2002) 98 NY 2d 208, a similar issue arose in relation to a claim under a policy for environmental pollution liability. The assured argued for joint and several liability on the part of all insurers during the period when the pollutants were being released into the ground, because of the difficulty of assigning the damage to any one period. Rejecting this argument, the New York State Court of Appeals said, at p 224:
Con Edison wants to combine this uncertainty-based approach, which implicates many successive policies, with an entitlement to choose a particular policy for indemnity. Yet collecting all the indemnity from a particular policy presupposes ability to pin an accident to a particular policy period (see Sybron Transition Corp, 258 F 3d at 601; Owens-Illinois, 138 NJ at 465, 650 A 2d at 988-989). Although more than one policy may be implicated by a gradual harm (see eg McGroarty v Great Am Ins Co, 36 NY 2d 358, 365), joint and several allocation is not consistent with the language of the policies providing indemnification for 'all sums' of liability that resulted from an accident or occurrence 'during the policy period' (see Olin Corp, 221 F 3d 307, 323).
Pro rata allocation under these facts, while not explicitly mandated by the policies, is consistent with the language of the policies. Most fundamentally, the policies provide indemnification for liability incurred as a result of an accident or occurrence during the policy period, not outside that period (see Forty-Eight Insulations, 633 F.2d 1212 at 1224). Con Edison's singular focus on "all sums" would read this important qualification out of the policies. Proration of liability among the insurers acknowledges the fact that there is uncertainty as to what actually transpired during any particular policy period (see Sybron Transition Corp, 258 F 3d at 602).
Recently, in State of California v Continental Insurance Company (2012) 55 Cal 4th 186, 198 (and note 4), the Supreme Court of California noted that proration had been adopted by at least 12 states (Colorado, Connecticut, Kansas, Kentucky, Louisiana, Massachusetts, Minnesota, New Hampshire, New Jersey, New York, Utah and Vermont), while Keene had been followed on this point in at least six states in addition to California itself (Delaware, Indiana, Ohio, Pennsylvania, Washington and Wisconsin), generally on account of the "all sums" language of the policy.
Fairchild and Barker were both cases in which legally unconventional rules for establishing liability in tort were adopted for reasons of policy. In Trigger, there was clearly a significant policy element behind the majority's adoption of a "weak" test of causation in the construction of the insuring clause, in place of the austere logic of Lord Phillips, who would have held that employers liable on the Fairchild basis were not insured at all. It is therefore natural to ask whether a similar approach may not justify a rule which would make each insurer liable in full irrespective of the period for which he was on risk, so as to ensure that whatever happens the employee is protected. This is essentially what the victim support groups submit. Judges are not always candid about the broader considerations which lead them to prefer one view of the law to another. But the desire to ensure an outcome which protects victims of occupational mesothelioma has had such a strong influence on recent case-law, that its relevance to the present issues is a question that needs to be confronted.
There are two reasons why the employee might be unable to recover damages for contracting mesothelioma resulting from his tortious exposure to asbestos. One is that his employer has insured with an insurer who subsequently becomes insolvent. The other is that his employer has in breach of his statutory obligation failed to insure at all. The employee has no reason to be concerned with either problem if his employer is solvent and able to meet his liabilities from his own resources. But both are a potential problem if, in addition, his employer is insolvent. It is clear that the main reason for holding an insurer who was on risk at any time during the period of exposure liable for the entire loss is that this obliges that insurer to bear the risk of the absence of effective insurance in other years in which it was not on risk.
It is therefore necessary to ask what conceivable policy could justify that? The Fairchild principle is not addressed to the problems of insurer solvency or non-insurance. It is addressed to the scientific impossibility of ascertaining when the insured occurrence happened. The Midland policies were written in a standard form which by its express terms applies only to injury or disease caused in Great Britain, Northern Ireland, the Isle of Man and the Channel Islands or to employees temporarily employed elsewhere under a contract entered into in one of those jurisdictions. It is therefore clear that it was designed to satisfy the employer's statutory obligation to insure under the Employers' Liability (Compulsory Insurance) Act 1969. It is self-evident that that Act was intended to protect employees with claims against their employers rather than the employers themselves. We can deduce from this that the Act of 1969 should predispose a court to find that that coverage for occupational injury and disease has been provided, as indeed this court held that it was in Trigger. But there is nothing in the policy of the Act which is inconsistent with insurance being obtained through annual policies, as it normally has been throughout the history of this market. And nothing which assumes that coverage will be provided beyond the express chronological limits of the policy simply because there is no effective insurance in place beyond those limits. On the contrary, the Act envisages that there will be continuous cover with authorised insurers. Insurers have deep pockets, but that in itself cannot justify imposing on them a liability which they have not agreed.
Nor is there any need to pick the pockets of the insurers in this way, since the employee is amply protected by various statutory schemes from the risk of being unable to recover. The Policyholders Protection Act 1975 introduced a statutory scheme of compensation for policyholders of insolvent insurers. It protected business policyholders in full in respect of risks subject to compulsory insurance. These arrangements have since been replaced by the wider terms of the Financial Services Compensation Scheme introduced by section 213 of the Financial Services and Markets Act 2000. The successive schemes have all been funded by statutory levies from the insurance industry. This legislation does not protect the employee in respect of loss attributable to a period for which there was no insurance in place. But such protection has now been conferred on eligible persons diagnosed on or after 25 July 2012 by the Mesothelioma Act 2014. The Act provides for a scheme to be established by secondary legislation under which the victim or his dependants will be entitled to specified payments from a statutory fund if they are "unable to bring an action for damages in respect of the disease against any employer of the person or any insurer with whom such an employer maintained employers' liability insurance (because they cannot be found or no longer exist or for any other reason)": section 2(1)(d). Section 18(3) provides that for this purpose the scheme "may specify circumstances in which a person is, or is not, to be treated as able to bring an action for the purposes of section 2(1)(d) ....". The scheme was established by the Diffuse Mesothelioma Payment Scheme Regulations 2014 (SI 2014/916). It provides for the payment of specified lump sums to victims or their dependents, the amount of which varies with the age of the victim upon diagnosis. The power conferred by section 18(3) has been exercised by extending eligibility to any case in which the employer falls within the 1930 Act (ie is insolvent) and "no other employer or insurer can be found or exists against whom the person can maintain an action for damages": regulation 7(1)(b) (emphasis added). There are potential issues about the criteria of eligibility in section 2 of the Act, in a case where the employee is entitled to proportionate amounts in respect of different years and there is insurance for some of those years but not for others. Like the Financial Services and Markets Act scheme, the cost of the fund is met by a levy on the United Kingdom insurance industry.
The combined effect of these schemes is that the employee is protected against the insolvency of an insurer or the absence of insurance, in any case where his employer is unable to meet his liabilities. As the rules governing the Financial Services and Markets Act scheme presently stand, if an insurer on risk in one year were required to pay the entire loss, thus discharging the liability of insolvent insurers on risk in other years, that insurer would to that extent be entitled to claim against the scheme: see Prudential Regulation Authority: Handbook, Compensation Rules, para 4.4.3. But if an insurer on risk in one year were required to make good the failure of the employer to insure at all in other years, that insurer would have no equivalent right to recover from the scheme created under the Mesothelioma Act 2014. Accordingly, the result of imposing on him a liability to pay the entire loss is to cast the entire burden of the insurance gap on him when the scheme of the Act of 2014 is to spread it across the insurance industry as a whole.
That leaves the question whether the right to prorate the insured's loss across the period of exposure applies also to defence costs.
The insuring clause provides, immediately after the principal coverage provision:
The Company will in addition .... be responsible for all costs and expenses incurred with the consent of the Company in defending any such claim for damages.
The insurer is liable under this provision for costs and expenses incurred with its consent in defending any "such" claim for damages, ie a claim for damages for disease caused during any period of insurance.
Similar language has been held in some of the jurisdictions of the United States which prorate the principal liability to require the proration of the defence costs as well: Insurance Company of North America v Forty-Eight Insulations Inc 633 F.2d 1212. I have some sympathy with the instinct behind this view, but the difficulty about it is that the tests are not the same. The insurer's liability for the compensation element of the claim falls to be prorated according to time on risk because on a proper analysis it relates only in part to the period for which the risk was insured. The insurer's liability for the defence costs is different. Unless there was some severable part of the defence costs that can be specifically related to a period when the insurer was not on risk, the whole of the defence costs had to be incurred to meet that part of the claim which was insured. The fact that it was also required to meet the uninsured remainder of the claim is irrelevant. The most that the insurer can say in this situation is that in funding the defence of a claim so far as it related to an insured period, it incidentally conferred a benefit on those who were potentially liable for the same claim in respect of an uninsured period: ie other insurers and IEG in its capacity as "self-insurer". In New Zealand Forest Products Ltd v New Zealand Insurance Co Ltd  1 WLR 1237, the insured incurred costs in defending litigation in California against a number of parties only one of whom, a director, was insured against the relevant liability. The Privy Council held that the defence costs did not fall to be apportioned between the insured and uninsured defendants. So far as the defence costs were reasonably required to meet the defence of a party whose liability was insured, the insurer was bound to pay them. It did not matter that the expenditure also benefitted other parties whose liabilities were not insured. The principle is accepted by the insurers on this appeal, who concede that they are liable to pay the defence costs in full. That concession appears to me to be correct.
It follows that as a matter of contract Zurich is contractually liable to meet the defence costs in full.
The Guernsey angle
In the Court of Appeal in the present case, Toulson LJ expressed the view that in the light of the subsequent developments in the law, Barker had become "past history" and was no longer good law even in cases (such as those arising in Guernsey) where the Compensation Act did not apply. I have arrived at the conclusion about the proration of contractual liability for compensation by reference to the terms and nature of the contract of insurance. The analysis would have been the same if Mr Carré had been employed in England. It is therefore strictly speaking unnecessary to address the question whether Toulson LJ was right about the current status of Barker. But in view of the fact that the point was fully argued, I will briefly summarise my reasons for thinking that he was wrong.
The common law is not a series of ad hoc answers to particular cases, but a body of general principle by reference to which answers may be found. The Act of 2006 did not alter any principle of the common law. In the first place, it did not lay down the elements of liability. It assumed liability and regulated only the measure of recovery. Secondly, it applied only to mesothelioma cases, and then only to regulate the measure of liability in tort as between the tortfeasor and the victim. Thirdly, even in relation to mesothelioma, section 3(1) applied only where the "responsible person" incurred liability for materially increasing the risk. Liability is incurred on that basis only on the footing that the time at which the disease is caused is impossible to determine. As Lord Phillips pointed out in Sienkiewicz v Greif (UK) Ltd  2 AC 229, at para 70, the courts would be entitled to revert to the conventional approach of requiring proof of causation on the balance of probabilities if advances in medical science make this possible. In other words, the Act left the common law intact, but carved an exception out of it for mesothelioma. It follows that Fairchild as interpreted by Barker remains good law in those jurisdictions (such as Guernsey) where the Act does not apply, and remains good law as applied to those legal relationships (such as the contractual relationship between insurer and insured) to which it does not apply. In those cases to which Barker continues to apply, it stands as authority for the allocation of liabilities which at common law are several only.
Equitable recoupment and redistribution
This question arises only on the assumption that an insurer who is on risk for only part of the period of exposure is contractually liable to meet the whole of the compensation element of the employer's claim or the whole of the defence costs. On that assumption, Zurich's argument is that insurers are entitled in equity
to redistribute the burden among other insurers who are liable in respect of the same amounts but in respect of different policy periods, and
to recoup from the insured a pro rata part of the cost of meeting that liability in respect of periods when there was no insurance at all.
As I have already explained, I consider that the assumption on which this argument arises, namely that an insurer on risk for only part of the period of exposure is contractually liable for the whole loss, is false. However, the question has a more general significance. If, as Zurich contend, there is a general right of contribution or recoupment (i) as between insurers and (ii) as between insurers and insureds in respect of periods of non-insurance, that would provide an alternative way of rectifying the anomalies associated with holding the insurer liable for the entire loss, alternative that is to construing the policy as responding for only a pro rata part of the loss.
As between insurers each of whom insured only part of the period of exposure but are liable (on this hypothesis) in full, I think it clear that there is a statutory right of contribution. Section 1(1) of the Civil Liability (Contribution) Act 1978 came into force on 1 January 1979, and applies to damage occurring after that date: see section 7(1). This has sometimes been questioned, for example by Friedmann, "Double insurance and payment of another's debt" (1993) 109 LQR 51, 54. But I can see no principled reason for questioning it. Section 1(1) provides that a "person liable in respect of any damage suffered by another person may recover contribution from any other person liable in respect of the same damage (whether jointly with him or otherwise)". A contract of indemnity gives rise to an action for unliquidated damages, arising from the failure of the indemnifier to prevent the indemnified person from suffering damage, for example, in a liability policy by having to pay the third party claimant: Firma C-Trade SA v Newcastle Protection and Indemnity Association  2 AC 1, 34 (Lord Goff of Chieveley). The class of persons "liable in respect of any damage suffered by another" may include those liable in contract, and there is no reason to limit it to those who have themselves caused the damage, as opposed to those who have assumed a contractual liability in respect of it. The question is therefore whether the damage for which successive insurers are liable is the same damage. As a matter of construction and on ordinary principles of insurance law, it is not. As I have said, successive insurers of liability on an occurrence basis do not insure the same liability. Each of them has contracted to indemnify the insured against an insured peril occurring in its own period on risk. In the case of an indivisible injury the liability of successive insurers is therefore alternative and not cumulative. However, on the footing that (contrary to my opinion) the law treats each insurer as liable for the whole loss in each period of insurance, then it must necessarily have been the same damage.
Whether there would be a right of contribution in respect of liabilities arising before 1 January 1979 is a more difficult question. There has always been a right of contribution at common law in cases of double insurance. But double insurance normally requires that two or more insurers should be liable in respect of the same interest on the same subject-matter against the same risks. On this ground, English law has hitherto declined to recognise that double insurance can exist as between insurers liable in respect of different periods even if the loss is the same: National Employers Mutual General Insurance Association Ltd v Haydon  2 Lloyd's Rep 149; Phillips v Syndicate 992 Gunner  Lloyd's Rep IR 426. It would require some considerable development of traditional concepts of double insurance to accommodate a situation like the present one. In Australia, where there is no legislation corresponding to the 1978 Act, this development has occurred: see Albion Insurance Co Ltd v Government Insurance Office of New South Wales (1969) 121 CLR 342. Whether the law should develop in the same way in England is a question that I should prefer to leave to a case in which it is more central to the outcome and the arguments of the parties. The Act of 1978 will cover the great majority of cases that seem likely now to arise.
What is in my view clear is that there cannot be an equitable right of recoupment as between the insurer and his insured in respect of periods when the latter was not insured. The reason is that unlike an insurer's relationship with other insurers under a co-ordinate liability for the same loss, his relationship with the insured is a contractual relationship. Its content has been determined by agreement, and a right of recoupment would be inconsistent with that agreement. If the insured is contractually entitled to the whole amount, there cannot be a parallel right of recoupment in equity on the footing that it is inequitable for the insured to have more than part of it. The basis of the suggested right of recoupment is that it is unjust for the insurer to have to bear the whole loss. But I do not understand by what standard it is said to be unjust when the parties have agreed that it should be so.
It is no answer to this to say that the alleged right of recoupment arises outside the contract. Of course, a contractual right and an equitable right of recoupment are juridically different. But the question is not what is the juridical origin of the claim for recoupment, but whether it operates by reference to the contract. To that question, there is only one possible answer. The alleged right of recoupment arises only because the contract (on this hypothesis) provides for the insurer to pay the whole loss. It arises as a direct result of the payment of the contractual indemnity. Its purpose is to undo in part what the contract has done. Mr Edelman submitted that a right of recoupment would only reflect the contribution of the employer to the risk of years which the insurer did not insure. So it would. But that is because (on this hypothesis) the contract requires the insurer to pay in full notwithstanding the contribution of the employer to the risk in the years which were not insured. If that is the consequence of the parties' agreement, I know of no legal doctrine which can do away with it.
Equity does not mend men's bargains. It may intervene to avoid unconscionable bargains, or to give effect to the parties' real intentions (for example when proprietary rights are conferred for a limited purpose such as security), or to provide remedies where those available at law are defective. But these are principled exceptions which depend on the unconscionability of allowing the law to take its course. There is nothing unconscionable about the performance of a contract of insurance according to its terms. In this respect, the principle on which equity acts is no different from that of the common law, even where the relevant common law claim is non-contractual. Thus a contractual relationship may give rise to a parallel duty of care in tort, and the consequences of breach (for example as regards limitation or foreseeability) may be different. But any contractual provisions about the content of the duty must apply to both: Henderson v Merrett Syndicates Ltd  2 AC 145, 191, 193-194 (Lord Goff of Chieveley). And a claim for unjust enrichment, which is probably the closest analogue to the right of recoupment proposed in this case, will not be allowed where its effect is to alter the contractual allocation of risks: Pan Ocean Shipping Co Ltd v Creditcorp Ltd  1 WLR 161, 164 (Lord Goff). As Etherton LJ said in MacDonald Dickens & Macklin (a firm) v Costello  QB 244, at para 23, in language which applies well beyond the domain of unjust enrichment with which he was concerned:
The general rule should be to uphold contractual arrangements by which parties have defined and allocated and, to that extent, restricted their mutual obligations, and, in so doing, have similarly allocated and circumscribed the consequences of non-performance. That general rule reflects a sound legal policy which acknowledges the parties' autonomy to configure the legal relations between them and provides certainty, and so limits disputes and litigation.
Of course, this will not necessarily apply where the relevant contractual right is vitiated, for example by illegality, frustration or mistake, all of which give rise to well established grounds for restitution: see Lord Mance's observations at paras 69-71. But this has no bearing on a case such as this is said to be, where a valid, lawful and effective contract requires the insurer to satisfy the whole liability notwithstanding that he accepted only a time-limited part of it. It is I think beyond question that to require part of that amount to be repaid on the ground that its retention would be unjust is a reversal of the effect of the contract by operation of law, something which cannot be justified if the contract is valid, lawful and effective.
Mr Edelman QC, who appeared for Zurich, submitted that in respect of periods when the employer was not insured, he could be regarded as "self-insured" and his position as regards contribution assimilated to that of a true insurer. Even if this were correct, it would not displace contractual allocation of risk. But in my view it is not correct. The submission is founded mainly on the decision of the House of Lords in Lord Napier and Ettrick v Hunter  AC 713, 730, which is said to be authority for the proposition that self-insurance is a form of insurance. The House held that a Lloyd's name was accountable to his subrogated stop-loss insurer for recoveries which he had made from successful litigation against his managing agents. Under the terms of the stop loss policy, the name had agreed to bear the first £25,000 of loss. It was held that he was not entitled to apply the recoveries against the bottom £25,000 of loss, because recoveries are applied to insurers "top-down", starting with the insurer of the highest tranche of loss. Lord Templeman referred to the name (p 730E) as acting as "his own insurer" for the uninsured tranches. But this was a figure of speech. The point that he was making was that if the name had actually insured the bottom tranche of loss, the insurer of that tranche would have been entitled to nothing from the recoveries because the insurers of higher tranches would have exhausted them. The name, having agreed to bear the bottom tranche himself, could be no better off than an insurer of the bottom tranche if there had been one. Self-insurance is non-insurance. Even if for the purposes of subrogation the position of a person with an uninsured excess is similar to that of an insurer of that excess, it does not follow that it is similar for any other purpose, still less that such a person is himself an insurer. IEG cannot be regarded for the purposes of the Civil Liability (Contribution) Act 1978 as being "liable" to themselves in respect of the uninsured periods of exposure for the same damage for which their insurers are liable to them in other years.
The real basis for the alleged right of recoupment is the intolerable consequences of holding an insurer liable for a loss sustained over many years irrespective of how long he was on risk. But the correct response to these consequences is for the courts to do what they normally do when one construction of a contract leads to absurd results. They reject it and prefer another which does not exhibit the same anomalies. The whole recoupment analysis is in my opinion a classic example of the problems associated with the adoption of special rules within the "Fairchild enclave" which differ from those that would follow from the application of ordinary principles of law.
Third Parties (Rights Against Insurers) Acts 1930 and 2010
I do not propose to lengthen this judgment yet further by addressing the question whether, if there were a right of recoupment as between the insurer and the insured, it could be set off against the claim on the policy. If it could be set off, the employee of an insolvent employer, suing under the Acts of 1930 or 2010, would be no better off by having a contractual right to recover the entire loss under the policy. In my opinion, the question does not arise because he has no such contractual right. I will simply observe that this is another difficult question which arises only as a result of the discarding of orthodox principles of contractual interpretation in favour of special rules devised for special "enclaves" without regard to general principles.
Lord Neuberger & Lord Reed
(agree with Lord Sumption)
This appeal represents yet another demanding chapter in the difficult series of decisions of the House of Lords and Supreme Court in relation to an employer's liability to a former employee, who was exposed to asbestos fibres during the course of his employment, and subsequently contracted mesothelioma, a disease which has been rightly described by other judges as "hideous" and "dreadful". For ease of reference we will refer to such an employer and such a former employee as an "employer" and an "employee" respectively.
The decisions start with Fairchild v Glenhaven Funeral Services Ltd  1 AC 32, which raised the question of an employer's liability to an employee, who had also been exposed to asbestos dust when working for another employer. In that case, as explained by Lord Mance at paras 3-4 and Lord Sumption at paras 114-116 and 124-128 above, the House of Lords was faced with an unedifying choice between
applying well-established rules of causation in tort and arriving at a thoroughly unpalatable decision, namely that neither employer's negligence could be proved to have caused the disease, and
extending the law of causation on an ad hoc basis, so that it was enough to prove that an employer's negligence had materially increased the risk of contracting the disease, in order to achieve a tolerably fair outcome, namely that each employer was liable.
The House elected for the latter course, and held that, in such a case, given that it was impossible to tell whether either employer's breach of duty had caused the employee to contract the disease, each of the two employers should be held liable to the employee.
To many people, that avowedly policy-based decision, which is applicable to any disease which has the unusual features of mesothelioma (as described by Lord Sumption in paras 116-117) seemed, and still seems, not only humane, but obviously right. Indeed, there can be no doubt that it would have required an exceptionally hard-headed (and, many people would say, hard-hearted) approach to hold that neither employer was liable, which is what the application of established legal principle would have indicated. However, as subsequent decisions have shown, the effect of what was a well-intentioned, and may seem a relatively small, departure from a basic common law principle by a court, however understandable, can lead to increasingly difficult legal problems – a sort of juridical version of chaos theory.
The problems stem from the fact that, unlike legislation, the common law cannot confine itself to a particular situation and deal with it in isolation from the remainder of the law; nor can it resolve problems on a purely pragmatic basis. It is a complex and extensive network of interconnected principles applicable to all situations falling within their scope. As Lord Nicholls of Birkenhead stated in Fairchild itself (para 36):
To be acceptable the law must be coherent. It must be principled. The basis on which one case, or one type of case, is distinguished from another should be transparent and capable of identification. When a decision departs from principles normally applied, the basis for doing so must be rational and justifiable if the decision is to avoid the reproach that hard cases make bad law.
The creation of an ad hoc exception from established principles governing causation in order to provide a remedy to the victims of mesothelioma was, in the first place, likely to result in uncertainty as to the legal rationale of the exception (as distinct from the social policy of enabling victims of mesothelioma to obtain a remedy against negligent employers), and the consequent breadth of that exception. The rationale could not be merely the impossibility of establishing the cause of an injury, since such a wide exception to the general rule governing causation would destroy the rule (see, for example, the attempt to extend the exception to cases of medical negligence, narrowly defeated in Gregg v Scott  2 AC 176). As Lord Brown observed in Sienkiewicz v Greif (UK) Ltd  2 AC 229, para 186, the unfortunate fact is that the courts are faced with comparable rocks of uncertainty in a wide variety of other situations too, and that to circumvent these rocks on a routine basis would turn our law upside down and dramatically increase the scope for what hitherto have been rejected as purely speculative compensation claims. In the event, the rationale of the Fairchild exception continues to cause difficulty (as, for example, in Novartis Grimsby Ltd v Cookson  EWCA Civ 1261). Secondly, the introduction of a novel test of causation in tort was bound, given the legal and commercial connections between different areas of the law, to give rise to a series of difficult questions and consequent uncertainty, as the ripples spread outwards.
The first question which subsequently manifested itself was how the common law, having taken this step into the unknown, should allocate liability for damages as between two employers, each of whom had permitted an employee to be exposed to asbestos fibres. That question was addressed in Barker v Corus UK Ltd  2 AC 572, a decision analysed by Lord Sumption in paras 129-135. The pragmatic decision that each employer was responsible for a proportion of the damages but not for the whole created a further exception to established legal principles. Perhaps unsurprisingly, it was not unanimous, and, as Lord Sumption says, the reasoning is not easy to analyse. Indeed, it is not without interest that Lord Rodger disagreed with the majority as to the proper analysis of the reasoning in Fairchild.
Parliament was unhappy with the decision in Barker, since it meant that, if an employer was insolvent, the employee might not recover that employer's proportion of the damages. The decision was effectively reversed in short order by section 3 of the Compensation Act 2006. Unlike the two House of Lords decisions, section 3 of the 2006 Act was expressly limited to mesothelioma cases: a restriction which Parliament could impose, but the courts could not. The effect of section 3 is explained in para 136 by Lord Sumption.
The next case to arrive at the Supreme Court in connection with employers' liability to employees was Sienkiewicz v Greif (UK) Ltd  2 AC 229, a decision which has no direct part to play in the present appeal, although it involved a logical, if probably unanticipated, extension of what had by then been dubbed "the Fairchild exception", in order to accommodate the existence of non-tortious environmental exposure to asbestos. In his judgment, the decisions and reasoning in the judgments in Fairchild and in Barker were discussed by Lord Phillips, who described them in paras 45 and 52 as raising two "conundrums" in connection with causation, which needed to be solved.
In the course of her concise judgment, Lady Hale in paras 167-168 referred to the decision in Fairchild as "kick[ing] over the hornets' nest". She added that she "f[ound] it hard to believe that their Lordships there foresaw the logical consequence of abandoning the 'but for' test". She also mentioned the possibility of overruling Fairchild, but said that "Even if we thought it right to do this, Parliament would soon reverse us". On one view, that might have been regarded as the best of reasons for overruling Fairchild. Lord Brown in para 185 also expressed doubts whether those who decided Fairchild could have appreciated the full implications of their decision. Lord Mance put the same point at a rather higher level of principle in para 189, when he referred to "the lesson of caution that the history may teach in relation to future invitations to depart from conventional principles of causation".
At least to a reasonable degree of clarity, these three cases and the 2006 Act have established the extent of an employer's liability for damages in relation to an employee who has been exposed to asbestos fibres in the course of his employment and has subsequently developed mesothelioma. Problems next arose in connection with the extent of the liability of insurers. The general position of insurers in law and in practice is summarised by Lord Sumption in paras 118-119. However, the way in which the law had developed in relation to mesothelioma claims by employees against employers raised problems of principle in relation to the liability of the employers' insurers.
Two such problems were decided by the Supreme Court in the so-called "Trigger litigation", Durham v BAI (Run off) Ltd  1 WLR 867, and they are explained by Lord Mance and Lord Sumption at paras 16-24 and 137-140 respectively. The conceptual difficulties thrown up by the decisions in Fairchild and in Barker were again demonstrated by the discussion in paras 63-66 of Lord Mance's majority judgment in Trigger, and by the contrast between his reasoning and that of Lord Phillips, who dissented (and see per Lord Clarke in para 84).
However, the position is still unclear in a case where an insurer insured an employer for only part of the period of a claimant employee's employment, and the employer was either uninsured for the rest of the period, or was insured with an insurer who is now insolvent. It is that situation with which this appeal is concerned, and the problem is identified by Lord Mance in his paras 42-44 and by Lord Sumption in his paras 141-142. As Lord Mance goes on to explain in paras 44-46, three different approaches are suggested. The first is that adopted by the Court of Appeal; the second is that proposed by Lord Mance; the third is that proposed by Lord Sumption.
We agree with Lord Mance and Lord Sumption that the Court of Appeal's analysis cannot be supported. It seems to us that they were wrong to conclude that the common law, as laid down by the House of Lords in Barker, had been changed as a result of section 3 of the 2006 Act. The section changed the law in this country, because (save perhaps in extreme circumstances) Parliament can, by statute, override the common law as laid down by the courts. However, it is clear from the terms of section 3 that it was intended to deal with a specific and limited class of case, namely the liability to employees, who were exposed to asbestos fibres in the course of their employment and subsequently contracted mesothelioma. In those circumstances, it seems to us that section 3 cannot be said to have altered the common law: it simply superseded the common law in the circumstances in which it applies.
That leaves the very difficult question as to which of the two approaches proffered by Lord Mance and Lord Sumption to prefer. The difficulty is compounded by the high quality and depth of reasoning in their two judgments. Further, it is interesting to note that each of these approaches has its adherents in other jurisdictions, as Lord Mance and Lord Sumption explain in paras 69 and 164-168 respectively.
Lord Mance's solution has a number of attractions. First, it is more in line with the Parliamentary approach as demonstrated by section 3 of the 2006 Act, because, unlike Lord Sumption's solution, it ensures that every employee whose employer was insured for any period of his employment, can look to any such insurer who is still solvent for full compensation. Secondly, unlike Lord Sumption's solution, it has been supported by one of the parties to this appeal: despite being raised by the court at a reconvened hearing, Lord Sumption's solution has not been adopted by either party. We suspect that these two points are not unconnected: the insurance market may fear that, if the court adopts the solution favoured by Lord Sumption, Parliament will intervene as it did following Barker. Indeed, such a concern may have been seen by some members of the court in Sienkiewicz as a reason for not reconsidering the decision in Fairchild. However, as a matter of principle, having rejected the contention that section 3 has changed the common law, it seems somewhat quaint (although, we accept, not logically inconsistent) to invoke section 3 as a reason for developing the common law in a certain way rather than another.
Thirdly, Lord Mance's solution represents a solution which is far closer to that which the London insurance market has worked out in practice. Fourthly, Lord Mance's approach does not clash with any of the preceding decisions to which we have referred, while it is, we accept, arguable whether Lord Sumption's solution is consistent with the reasoning of this court in the Trigger litigation. Just as in Barker there was a division of opinion as to the reach of the reasoning in Fairchild so there is a difference in this case as to the reach of the reasoning in the Trigger litigation – compare Lord Mance at paras 45 and 55 with Lord Sumption at paras 159-161. While, like so many points in this area, the issue is not easy, we agree with Lord Sumption's view.
On the other hand, in favour of Lord Sumption's view, it seems to us rather remarkable for an insurance contract to be construed as rendering the insurer liable for the whole of an employee's damages, where, for instance, the employee has been exposed to fibres for the whole of his 40 years of employment and the insurer in question has only provided cover for one of those years. (Or even for a temporary period of two weeks while the employer was considering whether to take out longer term cover.) As Lord Sumption explains, such an approach is inconsistent with the link between risk and premium which lies at the heart of a contract of insurance. Yet that is the basis of Lord Mance's conclusion. Lord Sumption's solution, which involves a pro rata liability, produces no such anomalous result: in the example just given, the insurer would be liable for 1/40 of the employee's damages.
It is true that the apparently anomalous result in the example we have just mentioned is mitigated by Lord Mance's view that the employer has to be treated as a self-insurer for the 39 years of non-insurance, so that the insurer can recover 39/40 of the damages it has to pay from the employer, provided the employer is solvent. While impressively reasoned in paras 56-78, Lord Mance's view that an insurer could recover a contribution from the employer, his insured (but not set it off against his own liability to the employer under the insurance contract), seems to us to open up a dangerous seam of potential litigation, as an exception is made to another established principle, namely that the respective rights and liabilities of the parties to a contract are governed by their agreement.
We appreciate that it can be emphasised that that aspect of Lord Mance's analysis is strictly limited to cases within the Fairchild exception, or as Lord Hodge has put it, the analysis only applies within the "Fairchild enclave". Enclaves are however notoriously difficult to police, and experience suggests that judicial attempts at restricting ratios may run into the same danger as when a court emphasises that a particular course is only to be taken in very exceptional circumstances. Once a principle is approved by a court (particularly, it may be said, this court), it is quite legitimate, indeed appropriate, for lawyers to invoke it and seek to apply it more generally, if it assists their clients' case. And here, it may well be argued, this court is invoking a new and wide general equitable power, which is, to put it at its lowest, close to inconsistent with an express contractual term, in order to reconstitute a contractual relationship so as to achieve what it regards as a fair result in a purely commercial context. Lord Sumption's analysis, by contrast, turns simply on the interpretation of the relevant contract of insurance, and does not appear to us to have any unfortunate wider ramifications.
Thus, Lord Sumption's analysis appears to us to do significantly less violence (and we think it probably does no violence) to established legal principles, whereas Lord Mance's analysis accords more with current practice and what is likely to be the view of the legislature. We accept that the fact that we are in the Fairchild enclave is a reason for favouring what may be said to be the more practical solution. However, our preference is in favour of learning what Lord Mance in Sienkiewicz referred to as "the lesson of caution that the history" of the decisions of the House of Lords and Supreme Court to which we have referred "may teach in relation to future invitations to depart from conventional principles", and agree with Lord Sumption. But we can readily appreciate why the majority of the court has formed the opposite conclusion.
In conclusion, it seems to us that it is at least worth considering what lessons can be learnt from the history summarised in this judgment and more fully treated by Lord Mance and Lord Sumption. There is often much to be said for the courts developing the common law to achieve what appears to be a just result in a particular type of case, even though it involves departing from established common law principles. Indeed, it can be said with force that that precisely reflects the genius of the common law, namely its ability to develop and adapt with the benefit of experience. However, in some types of case, it is better for the courts to accept that common law principle precludes a fair result, and to say so, on the basis that it is then up to Parliament (often with the assistance of the Law Commission) to sort the law out. In particular, the courts need to recognise that, unlike Parliament, they cannot legislate in the public interest for special cases, and they risk sowing confusion in the common law if they attempt to do so.
When the issue is potentially wide-ranging with significant and unforeseeable (especially known unknown) implications, judges may be well advised to conclude that the legislature should be better able than the courts to deal with the matter in a comprehensive and coherent way. It can fairly be said that the problem for the courts in taking such a course is that the judges cannot be sure whether Parliament will act to remedy what the courts may regard as an injustice. The answer to that may be for the courts to make it clear that they are giving Parliament the opportunity to legislate, and, if it does not do so, the courts may then reconsider their reluctance to develop the common law. For the courts to develop the law on a case-by-case basis, pragmatically but without any clear basis in principle, as each decision leads to a new set of problems requiring resolution at the highest level, as has happened in relation to mesothelioma claims, is not satisfactory either in terms of legal certainty or in terms of public time and money.
In the case of mesothelioma claims, there can be no real doubt but if Fairchild had been decided the other way, in accordance with normal common law principles, Parliament would have intervened very promptly. That may very well have been a better solution, but it can fairly be said that that observation is made with the wisdom of hindsight.
Colin Edelman QC, Leigh-Ann Mulcahy QC & Jamie Smith (instructed by DWF LLP) for appellant.
Antonio Bueno QC, Patrick Limb QC & Joshua Munro (instructed by Simpson and Marwick Solicitors LLP) for respondent.
Michael Kent QC (instructed by Keoghs LLP) for Intervener, Association of British Insurers.
David Allan QC & Simon Kilvington (instructed by Irwin Mitchell LLP and Slater and Gordon (UK) LLP) for Intervener, Asbestos Victim Support Group Forum UK.
all rights reserved