Subrogation

Subrogation

5.1 Introduction

5.1.1 The nature and influence of subrogation

’Subrogation’, put simply, is a form of substitution, where one party is substituted for another. In the insurance context, the core subrogation remedy enables an insurer who has indemnified its assured to bring proceedings on its own account but in the name of its assured.¹ Though the practical effects of subrogation as operated by insurers have risen to the surface of consciousness from time to time, often attracting criticism as they do,² and in recent years there has been renewed classificatory discussion of its remedies,³ obligations lawyers on the whole have not considered the profound impact of subrogation on the law of obligations itself. Here we begin to outline that impact. We will see further evidence of the influence of subrogation over the law of contract and tort in Chapters 7 and 8 respectively.

Insurance subrogation has often been criticized. To the extent that it allows substitution of the insurer for the assured, it changes the identity of one of the parties to a civil action, without this becoming superficially apparent.⁴ One informed critic has described this as giving rise to a fiction, not appropriate for a modern legal system.⁵ Related techniques have even been described as an ‘obscenity’ so far as insurers may seek to influence litigation by remaining disguised.⁶ In English law however, the default position is that an insurer who wishes to exercise subrogation rights is required to pursue the claim in the name of the assured,⁷ though the policy itself may entitle the insurer to call for an assignment of the cause of action, displacing this default.⁸ Nor are experienced judges likely to be misled as to the identity of the claimant. The disguise, however, may be inadvertently encouraged by theoretical accounts which make the two-party focus definitive of private law.⁹

Subrogation can be a powerful weapon, despite the limitations addressed in 5.4. Its influence is pervasive. For example, the existence of subrogation rights explains why, generally speaking, first-party insurance recoveries are to be left out of account in determining the defendant’s liability to the claimant:¹⁰ a feature of the law which might be taken to show the irrelevance of insurance is, rather, a manifestation of its impact on the nature of civil damages. Closer examination shows that a number of different but related remedies protective of the insurer’s interest are bundled together under the umbrella of ‘subrogation’. All of these, we suggest, respond to the indemnity nature of the contract. These remedies are wide-ranging. They enable the insurer to conduct litigation, whether the assured would be likely to seek alternative compensation or not; to reduce payments to the assured to reflect recoveries already made; to acquire an equitable lien over money received by the assured; and even to bring proceedings against an assured who has failed to protect the interests of the insurer. Lack of awareness of these remedies is doubtless heightened by their continued exclusion, in English courts, from first-party insurers offering life and personal injury insurance.¹¹ But as we will see, there is a need to control the operation of subrogation; and the route to controlling it begins with understanding it effectively.

5.1.2 Our argument

At least superficially, subrogation embodies and reinforces the principle of the defendant’s personal responsibility for the claimant’s loss, since it appears to ensure that liability is brought home in practice to a defendant who is liable as a matter of law. We suggest, however, that despite its intuitive appeal, personal responsibility is not necessarily a strong justification for subrogation remedies and certainly does not remove the need for control mechanisms.¹² In particular, the responsibility justification bites hardest where the defendant is uninsured.¹³ Here, and more generally, it is by no means clear that such a principle should be pursued in relation to ordinary instances of lack of care bringing about insured losses. It might be argued that personal responsibility for carelessly caused losses, rather than distribution by insurance, is the exception not the rule.¹⁴ Certainly, carelessness is one of the risks most likely to be provided for ex ante, whether by contract or by legislative intervention. In addition, where the defendant is insured, the resulting picture is necessarily one of duplicated insurance, which has been condemned as wasteful.¹⁵

Our particular account of subrogation does not give primary place to personal responsibility. We suggest, rather, that considerations around the location of risk are more central in the law of subrogation than any supposed need to support the fault principle or personal responsibility per se. In subsequent chapters, we seek out the influence of risk allocation in the law of obligations more broadly. We suggest that parties often consider insurance in the ex ante allocation of risks associated with contracting (Chapter 7), a factor which remains fundamental to the interpretation of contractual principles by the courts, and that in tort cases too courts may consider liabilities in terms of their effect on existing allocations of risks, or as risk allocations in their own right (Chapter 8). In both contract and tort, subrogation emerges as a key factor. For example, it may either support or subvert the allocation preferred by the parties, or by the law, or by a combination of both. While the underpinning equitable origin of subrogation is now clear,¹⁶ the basis of insurer subrogation nevertheless lies in the nature of the contract of insurance as an indemnity contract, and its proper scope—which may be modified by agreement—depends upon analysis of the relationships between various parties related to an insured loss. By definition, where an insurer seeks to bring an action in the name of the assured, it claims that another party should be the indemnifier in respect of the insured loss, and this is the claim to which the court must respond. This is an inherently multi-party question. Here, we assess the basis on which courts can and do respond to such claims, and therefore the actual and potential role of the courts in regulating subrogation practice.¹⁷

In pursuing this question, it is notable that English courts, unlike those in the United States, have generally not considered themselves free to exercise ‘discretion’ in awarding subrogation remedies, despite their equitable nature. Indeed, it has been said that the right to subrogation has ‘ossified’.¹⁸ Arguably, equity has been better at extending the insurer’s remedies than at providing courts (and parties) with the resources to control them. The right of a surety, or other person providing an indemnity under some form of obligation, to exercise subrogation rights was recognized as early as the seventeenth century, albeit in a series of discrete situations of which insurance was the most important both in practice and in terms of contribution to legal principle. The later expansion of subrogation, largely prompted by the consideration that a subrogated claimant obtained a proprietary remedy which prevailed over unsecured creditors in insolvency, threw up a series of complex conceptual questions over, for example, the right of a surety to assert equitable ownership of securities posted for the discharged debt.¹⁹ This is a long way from the straightforward concept of subrogation as it applies to insurance. The growth of such questions saw, in the last quarter of the twentieth century,²⁰ attempts to lay down, within a general framework of restitution, the principles applicable to subrogation claims. Subrogation is now firmly classified as a remedy based upon unjust enrichment.²¹ It is not clear that this has helped in any way to explain insurance subrogation. It has been pointed out that the notion of ‘unjust enrichment’ operates at a potentially very general level.²² Consistently with that observation, accounts of insurer subrogation based on unjust enrichment currently appear to use the same resources as our broadly contractual account. The leading work by Mitchell and Watterson, for example, describes the ‘unjustness’ in a case such as Napier and Ettrick v Hunter²³ as established through ‘a combination of policy factors’, which are also taken to explain the core subrogation right for insurers.²⁴ These ‘policy factors’ (namely the nature of the insured’s relationship with the insurer and third party; and the identity of the person who is thought appropriate to bear the loss) are the very factors that we describe in terms of contextual analysis of party risk structures and of the purpose of relevant duties and liabilities. One question is whether it will clarify or obscure the position to describe these factors in terms of ‘unjustness’.²⁵

More positively, we suggest that analysis of risk allocation, particularly allocations effected by parties, offers the courts their most promising route to control of insurers’ subrogation rights, and that they regularly take this up. While the insurer’s remedies may have ‘ossified’ in certain respects,²⁶ this is not necessarily the end of the matter. The very existence of such remedies is dependent on showing that the assured, on proper construction of the various relationships in play, had a right of action for the insurer to take over. In many instances, the courts’ approach to this question reflects the fact that insurance arrangements are a primary consideration for contracting parties, and we provide further evidence of this in Chapter 7. Outside the contractual context, the questions are more complex, but we suggest that risk allocation analysis remains significant. The effect of our argument is that insurance considerations directly affect the very existence of a contractual or tortious duty.

We argue, therefore, that English courts have not only shown awareness of the significance of subrogation despite the surface appearance that it is irrelevant to determination of rights between the assured and a potentially liable third party, but have also reached decisions which, in cases where there are contractual arrangements between the assured and the potentially liable third party, allow losses to be allocated in accordance with what they construe to be the parties’ intentions. Since the role of the insurer is not always evident, a distorted view of the law may develop. In contract cases, and to a lesser extent in tort cases, the reader sees a decision which requires the risk of loss from an event to fall on one or other of the parties, but it will not be immediately apparent that the claimant may be a subrogated insurer who has no interest in preserving the long-term relationship between the parties or even their objectively intended allocation of risk manifested in the insurance arrangements themselves, and whose claim is rejected on what would otherwise be inexplicable grounds. It would be wrong simply to assume that decisions are reached without considering the operation of subrogation. Indeed, it might be said that judicial disregard of subrogation, or of the parties’ allocation of risk, has given rise to those decisions which have attracted the greatest criticism.²⁷ These cases have had the greater emphasis. But they stand out as the exception, rather than the rule.

5.1.3 Summary

The next part of this Chapter, 5.2, introduces a relatively simple and uncontentious illustration of subrogation in action, which illuminates its connection with risk allocation and with loss-spreading effects. We refer back to this illustration when we turn to exploration of the basis of subrogation and the means by which the courts may control its operation, in 5.3. In 5.4, we explore key features of subrogation in operation, drawing attention to some important limitations. The most important of these is its non-applicability to life insurance and standard personal injury policies. Because of the significant issues raised for the purposes of the law of damages and for the nature of subrogation itself, this limitation is explored in more detail in 5.5. Finally, 5.6 offers an interim assessment of the impact of subrogation in terms of risk allocation and loss distribution, drawing on particular cases where the effect of subrogation has been considered suspect, and shows how these stand out from the courts’ general approach.

5.2 The Issues Illustrated

In the complicated offshore drilling environment with its intricate divisions of responsibility and countless contractors and subcontractors, a simple slip-and-fall can turn into a multiparty morass of contribution cross-claims, third- and fourthparty defendants, reciprocal indemnity agreements, and the ever-popular warranties of workmanlike performance. The plaintiff in this litigation has long since departed but the other parties have chosen to remain on the field of battle to contest the appropriate share of the plaintiff’s settlement to be borne by each of them.²⁸

Lord Bingham quoted this remark from a US judge in Caledonia North Sea Ltd v British Telecommunications Plc.²⁹ This was no ‘slipping and tripping’ case, but the sequel to a multi-million pound settlement in respect of a major explosion with considerable loss of life. The judge’s remarks themselves project a mixed picture. On one hand the injured party has ‘long since departed’—compensated before the key legal battles ensue. On the other, the economic consequences appear to be the subject of elaborate and costly litigation.³⁰ A key part of this ‘multi-party’ sequel is subrogation, and this was the main issue in play in Caledonia itself.

Since subrogation has often attracted criticism, Caledonia North Sea Ltd is a useful illustration that subrogation does not necessarily undermine but may both reflect, and underpin, contractual allocations of risk. But it also helps to illustrate the issues in play. The action was brought by the liability insurers of the operator of the oil platform ‘Piper Alpha’ nominally against contractors working on the platform. There had been a catastrophic explosion on board the platform, leaving many killed and injured. The insurers of the operator had promptly settled multiple claims for personal injury and loss of life brought by victims and families of the employees of contractors despite the fact that the contractual allocation of risk placed the duty to compensate, the duty to insure, and the duty to indemnify the operator against such claims, on the contractors themselves. Settlement of the tort claims by the operator’s insurers enabled protracted questions about the destination of the financial burden of these losses to be determined between those insurers and the insurers of the various contractors whose employees had been killed or injured.³¹ In this sense, the victims were assisted by being left out of the litigation process. Subrogation is, as we will see, sometimes alleged to provide windfall profits to insurers, in that relevant premiums are not set taking the availability of such actions into account: in other words, insurers pass on some of the risks which they are contracted to bear, and profit at the expense of policyholders.³² In this instance, the operator of the oil platform could be said however to have gained the very benefit for which it had thought it worthwhile paying an insurance premium in addition to the negotiated contractual terms. That is, it escaped the responsibility and costs of litigation to ensure the risks lay ultimately where the contract placed them (with the contractors’ insurers). The operator also had the benefit of an indemnity without the delays attendant on complex litigation.³³

The claim in this case would make no sense without awareness of subrogation. On the face of it as we have seen, the claim was one by the platform operator, notwithstanding that it had been indemnified by its insurers and that the fault of one of its employees had contributed to the explosion. That claim was brought against contractors who had committed no breach of duty either personally or through their employees but who were liable for injury to or death of their own employees by virtue of the terms of the applicable contracts. Risk, not fault, was the parties’ primary concern in choosing these terms. The contractual structure itself illustrates that in an industry where the risks of major harm caused by momentary carelessness are considerable, and where the chances are strong that any such harm will be caused by the employees of more than one contracting party, responsibility in relation to such harm is routinely perceived as an ex ante question to be approached in terms of liability to compensate as much as an ex post question of liability for breach.³⁴ The potential consequences of ‘inadvertent negligence’ of one or more individuals is routinely dealt with as part of contractual structures, just as legislative policy adopts a loss-spreading approach to the risk of carelessly caused harm in certain notable instances.³⁵

We suggest that the House of Lords determined that the contractors, not the insurers, were to be treated here as the primary indemnifiers of the operator in respect of injury to or death of their employees. The court interpreted, and gave effect to, the risk allocation of the contracting parties, concluding that the operator’s insurance policy was not the primary source of indemnity, and the insurers could themselves be indemnified. From the point of view of the contractors, the operator’s insurance arrangements were irrelevant, and the contractors faced ultimate liability for claims by their own employees whether or not the operator was insured: were it otherwise, as their Lordships pointed out, the contractors would have received the benefit of a policy for which they had not paid any premium. This is frequently used as an argument in favour of subrogation but is often unpersuasive. Indeed, it is generally the case that the receipt of an unrequested benefit does not impose any obligation to pay for it.³⁶ Here, however, it actually does fit the contractual arrangements in that the contractors were to bear liability and to insure against it, so that the existence of the operator’s own policy was (in terms of the agreed risk allocation for claims by contractors’ employees) purely fortuitous. It is important to note, therefore, that the House of Lords did not assume that the contract was fortuitous, but argued towards this position in light of the contractual arrangements in relation to risk allocation. Before leaving Caledonia North Sea, let us also look briefly at the case in terms of loss distribution or spreading (the subject of Chapter 6, and also the source of much criticism of subrogation).

What may be seen at work in Caledonia North Sea is a judicial ruling which gave effect to the bargain reached by the parties. The parties doubtless regarded the arrangements whereby each employer took responsibility for injury to its own employees³⁷ as the most appropriate in terms of cost and availability of cover. Imposing liability on the operator for injury to contractors’ employees would require extensive public liability (PL) insurance, given that the victims would not be the operator’s own employees. By contrast, the contractors are required as a matter of law³⁸ to take out employers’ liability (EL) insurance, and the parties doubtless decided to leave the losses to lie where the law placed the burden of insuring them. There is probably little difference in terms of ultimate loss-spreading, because both PL and EL insurers have extensive reinsurance arrangements in place and, irrespective of whoever paid the claims, they will to a greater or lesser extent be passed down the reinsurance chain, although these forms are quite different in nature despite similar wording in many instances.³⁹ So subrogation in this case simply ensured that the insurance arrangements which were required by the parties to be put into place operated according to their terms. Disregard of underlying arrangements or assumptions will, by contrast, cause difficulties. Indeed subrogation may, at its extreme and if it takes place in the absence of an understanding of contractual allocation, allow losses to be placed with uninsured individuals.⁴⁰ Probably more commonly, lack of attention to the legal context of tort liabilities may lead losses to be placed, through subrogation, with public bodies which may thus bear the burden of insuring, or absorb the loss.⁴¹ In these cases, there is typically no relevant contractual allocation of risk, but the courts may nevertheless seek to interpret the policy of the law in relation to risk allocation, typically in relation to the question of whether there is a ‘duty of care’ on the part of the public body.⁴²

Our general message is that subrogation is not uncontrolled by the courts, and that it does not always operate unfairly. However, we must recognize the potential for adverse effects. Even in a case similar to Caledonia North Sea, but with factual variation over the identity of negligent individuals and applicable contractual structure, it is possible to envisage tort claims being made against the contractors, to be settled by their insurers, and with a subrogation action by those insurers against the operator on the basis of the negligence of its employee. The result here would not be spreading of losses through the agreed insurance routes (except to the extent that contribution or reinsurance performs this role), and there would then be a stronger argument that subrogation might trigger the costly double insurance concerned. None of this would, however, be possible where the contractual structure is interpreted to give no such right of action to the contractor. And this in turn points to a key role for the courts, in determining what contractual obligations between the assured, and the third party, actually were, in light of their allocation of risk. The role of subrogation, as we have begun to explain, makes these questions more complex. But it is intimately connected with them.

5.3 The Basis of Insurer Subrogation

5.3.1 Classification and its implications

Subrogation is likely to evade effective control by the courts if its nature and rationale are not understood, since this lack of knowledge will tend to lead to inflexibility in remedies. This section is concerned, therefore, with the nature and basis of insurer subrogation. As we have seen, subrogation is very loosely translated as ‘substitution’. Subrogation in our particular context may typically be identified as enabling an insurer, having indemnified its assured, to stand in the shoes of the assured and to exercise the legal rights of the assured in the assured’s name against any third party who is responsible for the assured’s loss. The principle is of some antiquity. In 1782, Lord Mansfield said that ‘every day the insurer is put in the place of the assured’.⁴³ The right of substitution was codified in the Marine Insurance Act 1906, s 79, although the section is not a model of clarity. An insurer who pays any loss has transferred to it any right of suit which the assured might possess against a third party who has caused the loss (subrogation).⁴⁴ In addition, an insurer who has paid for a total loss⁴⁵ of subject matter is entitled on payment to assert the ownership of whatever remains of the subject matter (better referred to as salvage): in non-marine cases that is likely to be of little value because a total loss contemplates that nothing worth having has survived, but in marine cases the lesser concept of constructive total loss allows the assured to recover up to policy limits for less severe damage.⁴⁶ The only common feature of subrogation and salvage is that they confer rights upon the insurers which are exercisable on payment, but otherwise they have evolved to become separate concepts, though they share a common origin.⁴⁷

It is clear that equity has played a central role in the development of insurer subrogation. One of the earliest reported subrogation cases, Randall v Cochrane,⁴⁸ refers to the ‘plainest equity’ on the part of the insurer. In a careful review of English decisions, Derham found only one case, North of England Iron Steamship Insurance Association v Armstrong,⁴⁹ in which a right within the doctrine of subrogation was directly enforced by a common law court before the Judicature Acts.⁵⁰ Equally, however, courts have also accorded a very significant role to the nature of the parties’ contract of indemnity in rationalizing this equitable response.⁵¹ We return shortly to the continued importance of the contractual setting. There have however been recent moves to recast subrogation as a whole as an aspect of the law of unjust enrichment and therefore as providing an equitable remedy to reverse an unjust gain, and this merits consideration.⁵² An attempt to recover the basic rationale of subrogation in coherent terms is in principle welcome. However, the language of unjust enrichment necessarily does not match the language used by the courts in the long history of insurer subrogation. More importantly, it is at this stage not clear whether the criterion of justness can assist, rather than obscure, the basis for controlling subrogation rights.

The central role of unjust enrichment in some instances of subrogation is irrefutable following the decision of the House of Lords in Banque Financiére de la Citè v Parc Battersea (Ltd), which offered a novel subrogation remedy in accordance with an unjust enrichment rationale.⁵³ That has not however been universally welcome; and some of the points made against the application of an unjust enrichment rationale in Banque Financiére are resonant with the points we make in this chapter. For example, Michael Bridge argued that restitution or unjust enrichment is so focused on the doing of corrective justice between plaintiff and defendant that this function is in danger of subsuming all others’ (particularly referring to neglect of ‘distributive’ issues including the rights of parties not before the court);⁵⁴ that the House of Lords in Banque Financiére was not sufficiently alert to the need to avoid inconsistency between restitutionary remedies and contract remedies in the case of failed contracts;⁵⁵ that the idea of what is ‘unjust’ applied in the case was insufficiently defined and that parties would be subjected to a period of considerable uncertainty and insecurity before the courts managed to define what it means in such cases (at which point it would be confined to ‘insignificant proportions’);⁵⁶ and that it is ‘highly questionable for the law of restitution to come to the aid in this way of a large financial institution that neglects to take proper precautions’.⁵⁷ We would argue that the interests of other parties, clarity of legal reasoning, fairness between parties, and even potentially principles of ‘personal’ responsibility, will most likely be enhanced by considering the parties’ attempts to allocate risks rather than beginning with ‘unjustness’ viewed uniquely through attention to the parties before the court. Our core concern is of course with insurer subrogation. Banque Financière was not a case of the insurer’s right of subrogation and, whatever the arguments for and against its use of unjust enrichment analysis, in our view that analysis was not said by the House of Lords to extend to insurance cases.⁵⁸ We should therefore consider the argument, contained in the most influential accounts of subrogation, that insurance cases were not in fact outside the reach of the comments of the House of Lords in Banque Financiére, even if this is the most ‘natural’ meaning of those comments.

In this case, subrogation allowed a lending bank to acquire priority over other creditors (who were companies in the borrower’s group) despite a lack of security offered by the borrower itself. It was an express condition of the advance offered that other companies in the group would ‘postpone’ their claims until after the lender had been paid. The other creditors did not know of the postponement letter; but this was irrelevant to a claim based not in ‘contractual’ subrogation (a category which in this context should be broadly understood for reasons set out below),⁵⁹ but in unjust enrichment. Most importantly, Lord Hoffmann distinguished the case of insurance from the kind of case represented by Banque Financiére. Redefinition of insurer subrogation in terms of unjust enrichment is therefore inconsistent with Lord Hoffmann’s discussion. Yet this discussion has been treated as the key authority for addressing insurers’ subrogation rights (other than those based in express assignment) in terms of unjust enrichment.⁶⁰

In the course of his judgment, Lord Hoffmann distinguished the issue in Banque Financiére itself, already outlined, from the case of insurer subrogation. Lord Hoffmann made it plain that the basis of the insurer’s subrogation remedies lay in the nature of the insurance contract and to this extent, in common intention.

although in Lord Napier and Ettrick v Hunter [1993] AC 713⁶¹ your Lordships rejected the exclusivity of this claim for the common law and assigned a larger role to equitable principles, there was no dispute that the doctrine of subrogation in insurance rests upon the common intention of the parties and gives effect to the principle of indemnity embodied in the contract.

The potential for misclassification of the insurer’s subrogation remedies probably arises from the continuation:

Furthermore, your Lordships drew attention to the fact that it is customary for the assured, on payment of the loss, to provide the insurer with a letter of subrogation, being no more nor less than an express assignment of his rights of recovery against any third party. Subrogation in this sense is a contractual arrangement for the transfer of rights against third parties and is founded upon the common intention of the parties. But the term is also used to describe an equitable remedy to reverse or prevent unjust enrichment which is not based upon any agreement or common intention of the party enriched and the party deprived.

It is easy to see how it might be concluded that only express contractual terms are being contrasted with the form of subrogation which turns on unjust enrichment. But the earlier reference to subrogation arising from the nature of the insurance contract and thus from the parties’ common intention makes this interpretation impossible given the clear contrast in the final sentence quoted. ‘Common intention’ is being used where, on the analysis previously adopted by Lord Diplock, the basis was in ‘implied terms’. We agree that ‘common intention’ is a preferable and more flexible term than ‘implied term’.⁶² Remedies to reverse or prevent unjust enrichment are here contrasted with subrogation based upon ‘any agreement or common intention’ of the party enriched and the party deprived—the very basis of insurer subrogation as Lord Hoffmann had just outlined it.

Andrew Burrows has suggested that although Lord Hoffmann’s words are ‘most naturally interpreted’ as suggesting that ‘subrogation in indemnity insurance rests on common intention of the parties and is contractual’, in fact this was not his intention.⁶³ In our opinion, however, Lord Hoffmann’s emphasis on contract is not accidental. The very point of the distinction drawn by Lord Hoffmann in this passage is that party intentions were not to be considered relevant to the unjust enrichment category represented by this case. Turning this around, plainly, party intentions are relevant to the insurer’s right of subrogation:

The fact that contractual subrogation and subrogation to prevent unjust enrichment both involve transfers of rights or something resembling transfers of rights should not be allowed to obscure the fact that one is dealing with radically different institutions. One is part of the law of contract and the other part of the law of restitution. Unless this distinction is borne clearly in mind, there is a danger that the contractual requirement of mutual consent will be imported into the conditions for the grant of the restitutionary remedy or that the absence of such a requirement will be disguised by references to a presumed intention which is wholly fictitious. There is an obvious parallel with the confusion caused by classifying certain restitutionary remedies as quasi-contractual and importing into them features of the law of contract.

The underlying problems, we suggest, are revealed by the fact that in this context, addressing the issue in terms of ‘unjust enrichment’ appears to leave the detailed business of determining whether and why subrogation remedies should actually operate to some rather indistinct ideas of ‘policy’.⁶⁴ We suggest that it is better to look more closely at the resources available to the courts through broadly contractual analysis. This means looking at what the courts have actually been doing, and also at what they have been saying. Indeed in this instance, it is unjust enrichment which, with some irony, can be said to superimpose a new and over-generalized conceptual structure on to existing practice. To illustrate this further, on Burrows’ account, the above statement of Lord Hoffmann ‘purported’ to support the statements of Lord Goff in the earlier case of Lord Napier and Ettrick v Hunter;⁶⁵ but failed to do so.⁶⁶ In our opinion, this rests on an unduly narrow interpretation of what is meant by the intentions of the parties.⁶⁷ Our approach in this chapter, and in later chapters, underlines the importance of the parties’ allocation of risk beyond the narrow confines of express and implied terms. We interpret Lord Hoffmann’s remarks as distinguishing between ‘wholly fictitious’ references to party intentions outside the ambit of subrogation on the basis of mutual intent; and the broad analysis of party intentions and allocations of risk which, in our view, appropriately underpins subrogation in the context of insurance. Although there is a question over what best to call these considerations, we would prefer the label which makes the questions more apparent. Taking the broader view of the resources available to contractual analysis adopted in this chapter, and reflecting the actual practice of the courts in relation to insurer subrogation as explored here and in later chapters, we can see that Lord Hoffmann’s remarks are entirely compatible with the analysis of the House of Lords in Lord Napier and Ettrick v Hunter,⁶⁸ where the roots of the equitable response were also acknowledged to lie in the parties’ contractual arrangement (hence, at a broader level, the intentions of the parties and the relationships between them). According to Lord Goff, the principle of subrogation in an insurance context ‘has been regarded, both at law and in equity, as giving effect to the underlying nature of a contract of insurance, which is that it is intended to provide an indemnity but no more than an indemnity’.⁶⁹