1. Proximate cause rule
The insured, who puts forward a claim, must show that it is more probable than not that the loss was caused by one of the perils which the insurers contracted to cover.1 To take a simple example, if the policy covers loss of goods by fire, the insured must show that, on the balance of probabilities, fire caused the loss, and obviously there is no claim if the goods were lost by some other event, such as theft. But what if the fire is caused by the insured carelessly tossing away a cigarette, or by an act of arson committed by a third person or by the insured? What if the goods are damaged, not by the fire which has broken out in the warehouse, but by smoke, or by part of the building in which they are kept collapsing as a result of the fire, or by the water used by fire-fighters to extinguish it? What if the goods are moved from a burning building into the street to save them and are damaged by rain or are stolen? The question is, what loss have the insurers agreed to cover? If the fire has reduced the security of the warehouse and, as a result, thieves are able to enter and steal the goods, the cause of the loss is theft not fire. The result is the same if the goods are removed from a burning building and placed on the pavement from where they are stolen.2 When a building, which was insured against fire, was damaged by a mob drawn to a fire in an adjacent property, the cause of the loss was the action of the mob not the fire.3 On the other hand, a theft policy did cover the loss when a shop was burgled during an air raid because, although loss ‘occasioned by hostilities’ was an excepted peril, the cause was theft and the air raid merely made the burglars’ task easier.4
The first question is, therefore, what peril have the insurers agreed to cover, which leads us back to the contract, and the second question is, was the loss caused by that peril? The focus in this chapter is on identifying the cause for the purposes of insurance law. The search could lead through a long chain of events: the death of a pedestrian might be said to have been caused by the negligence of the motorist who ran her down, or by her decision to cross the road at that moment. Indeed, the enquiry could go even further: ‘[I]t seems to me to be impracticable to go back to cause upon cause, which would lead us back ultimately to the birth of person, for if he had never been born the accident would not have happened.’5 As Blackburn J pointed out in a marine insurance case, a loss typically occurs as the result of a series of events:
The ship perished because she went ashore on the coast of Yorkshire. The cause of her going ashore was partly that it was thick weather and she was making for Hull in distress, and partly that she was unmanageable because full of water. The cause of that cause, viz, her being in distress and full of water, was, that when she laboured in the rolling sea she made water; and the cause of her making water was, that when she left London she was not in so strong and staunch a state as she ought to have been.6
As will be seen later, the parties may agree in the policy on the test for determining causation, but where they have not general principles have been formulated by the judges. In an early case, Willes J said, ‘[Y]ou are not to trouble yourself with distant causes, or to go into a metaphysical distinction between causes efficient and material and causes final; but you are to look exclusively to the proximate and immediate cause of the loss.’7 This idea of proximity was later incorporated in the Marine Insurance Act 1906, section 55(1):
Subject to the provisions of this Act, and unless the policy otherwise provides, the insurer is liable for any loss proximately caused by a peril insured against, but, subject as aforesaid, he is not liable for any loss which is not proximately caused by a peril insured against.8
In other words, the operative cause in insurance law is the proximate cause of the loss. This gives little assistance in determining cases, and some of the Victorian cases were decided simply by looking for the event that was closest in time to the occurrence of the loss. When Mr Lawrence was killed by a train after an epileptic seizure had led to him falling from the platform at Waterloo Station, the court decided that his death was caused not by the seizure but by the train. It was, therefore, an accident and covered by a personal accident policy.9 This might be seen as merely choosing the latest cause, but the decision can be justified on other grounds.10 To say that the proximate cause is the last in time may merely be to assert that the cause that is closest in time to the loss is likely to be regarded as having a greater impact than other causes in the chain of events that lead to the loss. In 1918 Lord Shaw declared that: ‘To treat proximate cause as if it was the cause which is proximate in time is … out of the question. The cause which is truly proximate is that which is proximate in efficiency.’11 Lord Greene MR observed that ultimately this issue ‘is really a matter for the common sense and intelligence of the ordinary man’,12 and Lord Denning MR referred to, ‘the effective or dominant cause of the occurrence, or, as it is sometimes put, what is in substance the cause, even though it is more remote in point of time, such cause to be determined by common sense’.13 Lord Mance has suggested that the use of the last cause in time in the Victorian cases might explain the strict approach taken to warranties:
If the only relevant cause is the last cause in time, then a prior breach of a simple contractual obligation regarding fitness could have been regarded as irrelevant. Hence, the development of the concept of a warranty which, if broken, automatically discharged from liability for loss or damage, irrespective of how such loss or damage was in law to be regarded as caused.’14
In Reischer v Borwick15 a marine policy covered a ship against damage caused by a collision but not by perils of the sea. After the ship was holed in a collision, a temporary repair was effected so that it could be towed into a harbour. However, the motion through the sea caused by the towing led the hole to reopen and the ship sank. Even though in terms of time it might be said that the loss resulted from the motion of the sea, which was not covered, the Court of Appeal held that the proximate cause of the loss was the collision. In Leyland Shipping Co Ltd v Norwich Union Fire Insurance Society Ltd,16 the Ikaria was insured against loss from perils of the sea but not from war. During the First World War, the ship was hit by a torpedo, which blew a hole in its side and damaged the bulkheads. It was towed into a safe harbour. While anchored a gale blew up causing the ship to bump against the harbour wall. The harbour authorities, fearing it might sink and block access, ordered the ship’s removal into open sea where it grounded at each low tide. After two days of this buffeting, the bulkheads failed and the ship sank. Even though the ship would probably have been safe had it stayed in the harbour, the House of Lords took the view that the proximate cause of the loss was the torpedo. The impact of the torpedo meant it was reasonably certain that sea-water would flow into the ship and this is what happened, even if the extent of the damage caused might not have been expected. In The Cender Mopu,17 the Supreme Court emphasised that section 55(1) of the Marine Insurance Act 1906 required the court’s inquiry to be based on fact and common sense principles, and reaffirmed the principle that the proximate cause is the cause which is proximate in efficiency.18 In this case, the insurance related to an offshore oil-drilling platform during a towage voyage from the US to Malaysia. The policy covered all risks, except loss ‘caused by inherent vice or nature of the subject matter insured’, which reflected the wording of section 55(2)(c) of Marine Insurance Act 1906. In mid-voyage one leg of the rig broke and the next day the other two legs broke. The question was whether the loss had been caused by perils of the sea, which were covered, or inherent vice of the rig, which was not. As the insurers knew, there were stress cracks in the legs before the voyage and, indeed, the insurers required the rig to be checked in mid-voyage. The weather encountered had been such as might have been reasonably expected during such a voyage. The Court held the insurer liable. The proximate cause of the loss was the perils of the sea, namely, a wave that caught the first leg in such a way as to cause it to break, which, in turn, led to stress on the other legs and their failure. Lord Mance concluded a comprehensive discussion of the distinction between inherent vice and perils of the sea by saying that the insurer will only avoid liability by showing that the loss arose from ‘inherent characteristics or defects in a hull or cargo leading to it causing loss or damage to itself … without any fortuitous external accident or casualty’.19 The loss was caused, not by some inherent defect in the rig itself, but by a wave hitting the leg in such a way as to break it.
If it is established that a peril covered by the contract was the proximate cause of the loss, the insurer will be liable for all the consequences that flow naturally from this peril: as Malcolm Clarke has put it, the full extent of the loss ‘will be recoverable, even though such extent was no more than not unlikely to occur at the time of the peril’.20 In both Reischer and Leyland the influx of seawater was the reasonably certain consequence of, respectively, the collision and the torpedo, and the effect of those events was still continuing and they had not been overshadowed by subsequent events. If, on the other hand, as it was being towed out to sea, the Ikaria had sunk following a collision with part of the harbour that had resulted from the negligence of the crew, then although the ship might not have been navigating out of the harbour had it not been for the torpedo, that event would no longer have been operating and the loss would have been proximately caused by the negligence.
The other point concerns the burden of proof. It is up to the insured to establish on a balance of probabilities that the loss was proximately caused by an insured peril.21 If the insured does this, the burden of proof shifts to the insurers to show that another explanation was more probable: where a ship is insured against fire, then if the insured shows that the loss was caused by fire, the insurers must demonstrate, on a balance of probabilities, that fire was not the proximate cause of the loss, or that the fire was caused by an excepted peril.22 If the insurers allege that the loss was caused deliberately by the insured or by someone else with the insured’s connivance, particularly if that act was a criminal offence, the degree of proof required will be ‘commensurate with the gravity of the allegation made.’ That burden will not be discharged ‘if the evidence fails to exclude a substantial, as opposed to a fanciful or remote, possibility that the loss was accidental’.23
2. Agreements to alter the rule of causation
Although the proximate cause rule has sometimes been called a rule of law,24 it is merely a term that is implied into the insurance contract as representing ‘the real meaning of the parties’.25 The parties—or, in reality, the insurers who draw up policies—can alter the normal rule. However, the courts will presume that the proximate cause rule applies, and that presumption will only be displaced by clear words. 26 So, even though the parties choose not to use the phrase ‘proximate cause’, but instead words such as ‘reasonably attributable to’ or ‘attributable to’ or ‘arising from’,27 the court will apply the proximate cause rule because these terms give no clear indication of the test to be applied.
A series of cases from the late nineteenth and early twentieth centuries involving accident policies provides an interesting illustration of attempts by insurers to alter the test of causation. It seems clear that the aim of the companies in writing the causation clauses in these policies was to exclude liability where injury or death was linked to some cause other than the accident, and certainly they disputed liability in these cases on the basis that this was what the policies had done. In Mr Lawrence’s case,28 the policy rendered the insurers liable ‘where such accidental injury is the direct and sole cause of death…but it does not insure in case of death…arising from fits’; but the court regarded ‘direct and sole cause’ as simply a restatement of the proximate cause rule. In Fitton v Accidental Death Insurance Co,29 a policy insured against death resulting from various accidental injuries, but specified that the injury must be ‘the direct and sole cause of death’ and excluded liability for death arising from, among other things, hernia ‘or any other disease or cause arising within the system of the insured before or at the time or following such accidental injury (whether causing death … directly or jointly with such accidental injury)’. The insured suffered an accidental fall that led to a strangulated hernia which, in turn, resulted in his death. The insurers were held liable because the policy exempted them ‘only where the hernia arises within the system’ and not where it was caused by the accident.30 Similarly, in Isitt v Railway Passengers Assurance Co,31 the insurers were liable under a policy that only covered death from ‘the effects’ of an accidental injury. The insured suffered an accidental injury and he was confined to bed. However, the injury was so painful that he was unable to bear the weight of any bed clothing. This led him to contract pneumonia from which he died. Willes J thought an appropriate direction to the jury would be:
Do you think that the circumstances leading up to the death, including the cold which caused pneumonia, were the reasonable and natural consequences of the injury and of the conditions under which the assured had to live in consequence of the injury? If you find that no foreign cause intervened and that nothing happened except what was reasonably to be expected under the circumstances, you may and ought to find that the death resulted ‘from the effects of the injury’ within the meaning of the policy.32
In an attempt to avoid that decision accident insurers reworded their policies to cover injury or death only where the accident was:
the direct or proximate cause thereof, but not where the direct or proximate cause thereof is disease or other intervening cause, even although the disease or other intervening cause may itself have been aggravated by such accident, or have been due to weakness or exhaustion consequent thereon, or the death accelerated thereby.
Such a policy came before the Court of Appeal after Mr Etherington fell during a hunting expedition, became soaked, caught pneumonia and died. The Court decided that the insurers had not avoided the Isitt decision with the new clause. This conclusion seems to have been influenced by the view that not to hold the insurers liable would make it difficult to establish a claim, unless the insured died at the time the accident occurred, and that would make such polices of very limited utility. Vaughan Williams LJ, noting the insurer’s intention to avoid the consequences of Isitt, said, ‘though that may have been the desire of the company, they have not had what I may call the commercial courage to express as plainly as they might have done what their counsel says they intended to express’. He added:
When the disease or other cause is dependent on the accident, I think it is right to say that the term ‘direct or proximate cause’ covers in such a case not only the immediate result of the accident, but also all those things which may fairly be considered as results usually attendant upon the particular accident in question.33
Vaughan Williams LJ can be seen as having fallen back on the proximate cause rule either because the clause was ambiguous or because he applied the principle of construction that avoids the literal meaning of a clause which would render the cover afforded by the policy illusory. Nevertheless, if the words used are clear the courts will apply them.34 The accident policy in Coxe v Employers’ Liability Assurance Corpn Ltd 35 excluded liability for death ‘indirectly caused by, arising from, or traceable to’ war. As part of his duties during the First World War, Captain Ewing was required to inspect sentries on a railway line that was poorly lit because of air raids and to which access was forbidden to civilians. He was hit and killed by a train. While Scrutton J took the view that the train was the proximate cause, the use in the contract of the word ‘indirectly’ obliged him to determine whether the loss had been indirectly caused by an excluded peril. The war had placed Ewing in a position of special danger and, therefore, was the indirect cause of his death, so the claim was refused. There must, however, be a limit and this would be in the use of the word ‘caused’: the cause may be indirect, but it still must be a cause, although how that is defined becomes uncertain. Presumably, the decision would have been different if Ewing had been standing on a platform waiting for a troop train and had died in the same way as Lawrence. In these circumstances the clause would not have excluded the insurer’s liability because the war would have been merely a part of the background to the accident and not one of its causes, direct or indirect: he might not have been on the platform but for the war, however, the war did not pose the danger from which he died. In Jason v Batten36 the term ‘any Accident bodily injury resulting in and being—independently of all other causes—the exclusive direct and immediate cause of the … injury or disablement’ excluded liability where a motorist, who had a narrowed coronary artery, sustained a severe coronary thrombosis when a car accident caused a blood clot to form which blocked that artery. It was held that if he had not had the narrowed artery, he would not have had a coronary thrombosis. The accident brought forward that thrombosis, but the court found that he would have suffered an attack within three years in any event. So, while the accident advanced the thrombosis by causing the blood clot to form, it was not ‘independently of all other causes the exclusive… cause’ of the thrombosis. Moreover, the thrombosis was ‘directly or indirectly’ caused by a pre-existing condition, namely the narrowed artery. If, however, the accident had activated a condition that was latent and that, but for the accident, would have remained so, the insurers would have been liable.37
Recently, in the Orient-Express Hotels Ltd v Assicurazioni General SA,38 Hamblen J concluded that the terms of the policy had replaced the proximate cause rule with a ‘but for’ test. OEH was the owner of a hotel in New Orleans. As a result of hurricanes, the hotel suffered physical damage and was closed. In addition, the hurricanes led to a state of emergency and the mandatory evacuation of New Orleans. OEH’s insurance covered business interruption caused by physical damage to the hotel. The question was whether OEH could claim where the losses arose both from damage to the hotel and the wider area. The relevant clause stated that the insurer’s liability was reduced by a calculation under which an adjustment was made so as to reflect ‘as nearly as may be reasonably practicable the results which but for the Damage would have been obtained during the relative period after the Damage’. OEH argued that the ‘but for’ test, which had been applied by the arbitrator, was inappropriate where there were concurrent causes. OEH relied on Miss Jay Jay rule (discussed below) to argue that where there are two proximate causes of a loss the insured can recover if one caused the loss insured, thus as the physical damage caused loss it was irrelevant that the vicinity also caused loss. Furthermore, OEH argued that the proximate cause rule applied to both concurrent interdependent causes as well as independent ones,39 and that the ‘but for’ test could lead to unfair results. Hamblen J acknowledged that there was ‘considerable force’40 in OEH’s argument because where there are two concurrent independent causes of a loss, the application of the ‘but for’ test would mean there is no cause of the loss.41 However, this was not the issue in the current appeal: the issue was whether the arbitrator erred in law by applying the ‘but for’ test. Hamblen J decided that the arbitrator had not erred for a number of reasons,42 particularly that the ‘but for’ test was agreed by both parties in the policy.43 Moreover, Hamblen J was not convinced that alternatives to the ‘but for’ test were more fair or reasonable and the use of the test did not have the consequence that there was no recoverable loss.44
3. Multiple causes45
Until relatively recently, the courts seem not to have given much thought to the possibility of there being more than one proximate cause;46 however, in 1974 Cairns LJ suggested the judges ‘should [not] strain to find a dominant cause’,47 and recently a High Court judge has been criticised for referring to the proximate cause in the singular.48 While the courts have not suddenly begun to search out multiple causes as a matter of routine, the broader approach to causation and the practice of insurers extending the causation test to include indirect causes of loss can only increase the likelihood of a court finding that there is more than one cause.49
Where there are a number of perils covered, the insured need only show that the loss was proximately caused by one of them.50 If there are two proximate causes that are independent of each other and each would have produced part of the loss without the contribution of the other, the insurer will be liable for that part caused by the covered peril. In Ford Motor Co of Canada Ltd v Prudential Assurance Co Ltd,51 Ford was insured against loss caused by riot, but the policy excluded loss as a result of damage caused by cessation of work or by change in temperature. There was a riot which led to the factory being closed and a large amount of damage was caused by freezing. The Supreme Court of Canada held that the insurers were only liable for that part of the damage solely caused by the riot.
There are two problematic situations. In the first, there are two proximate causes of the loss, one of these is covered by the policy and the other is neither covered nor excepted. In JJ Lloyd Instruments Ltd v Northern Star Insurance Co Ltd (The ‘Miss Jay Jay’),52 the Court of Appeal held that a ship, The Miss Jay Jay, had been lost as the result of a combination of two events: the adverse condition of the sea and defects in the boat’s design of which the insured was unaware. The insurance policy covered loss by adverse sea conditions, but did not mention loss by a design fault. Clearly, if the design fault had been the sole cause, there would have been no liability on the policy, but here the insurers were held liable. Slade LJ said:
As there were no relevant exclusions or warranties in this policy the fact that there may have been another proximate cause did not call for specified mention since proof of a peril which was within the policy was enough to entitle the plaintiffs to judgment.53
The second situation that creates difficulty is where there are two causes, one of which is covered by the policy, while the other falls within the terms of an exception clause. In Wayne Tank and Pump Co Ltd v Employers’ Liability Assurance Corpn Ltd,54 Wayne installed equipment to store and convey liquid wax in a factory. Subsequently, the factory burnt down as the result of two negligent actions by Wayne: the first was to supply plastic pipe to convey hot wax and an ineffective thermostat; and the second was to switch on the equipment and leave it unattended. The question before the Court of Appeal was whether the liability for the losses caused by Wayne’s negligence was covered by a public liability policy. That policy’s principal clause stated that the insurers ‘will indemnify the Insured against all sums which the Insured shall become legally liable to pay as damages consequent upon … damage to property as a result of accidents as described in the Schedule’. There was an exception for ‘damage caused by the nature or conditions of any goods or the containers thereof sold or supplied by or on behalf of the insured’. The Court of Appeal held that ‘goods’ included the pipe and thermostat. Lord Denning MR and Roskill LJ concluded that the proximate cause of the loss was the defective piping and thermostat and, therefore, the exclusion clause meant the insurers were not liable. They went on to say that even if there were two proximate causes—the defective installation, which was within the terms of the policy, and the defective goods, which was not—the insurers were not liable because having ‘stipulated for freedom [from the excepted risk], the only way of giving effect to it is by exempting them altogether’.55 Cairns LJ preferred this second approach.56
Is this second approach right? Certainly, the courts in California have taken a different view. In State Farm Mutual Auto Ins Co v Partridge,57 the insured accidentally shot and injured his friend. The court decided there had been two acts of negligence by the insured: filing down the trigger mechanism on his gun to create a ‘hair trigger’ and driving with the gun in his hand while his friend sat in the passenger seat. Without both of these separate actions the injury would not have happened. The problem was that the policy rendered the insurer liable if the cause was the negligence with regard to the hair trigger, but not if the cause was negligent driving. The court held that where there were two proximate causes and one was covered, the insurer was liable irrespective of the fact that the policy excluded liability for the other cause. In their book, Insurance Law,58 Keeton and Widiss suggest that decisions like Partridge show that, ‘when there are several distinct or distinguishable factors which contribute to a loss, a persuasive case can be made for the proposition that courts will apply the causation theory that will relate the loss to a covered peril’.59 They see this as an aspect of the way courts construe policies in light of the reasonable expectations of the insured:60
In many instances, decisions on causation questions involving insurance policy terms are best understood as manifestations of the judicial inclination to favor coverage either by construing ambiguous policy provisions against an insurer or by protecting the reasonable expectations of an insured.61
This idea has never achieved wide acceptance in the US and the courts tend only to look at the reasonable expectations of the insured where there is ambiguity in the contract.62 However, in relation to exception clauses, US courts, generally, appear to take the view that ‘exclusionary clauses are interpreted narrowly, whereas clauses identifying coverage are interpreted broadly’,63 and do seem relatively easily persuaded that there is ambiguity in an exception clause.
The decision in Partridge suggests that the Wayne approach is not self explanatory. It is possible to argue in favour of Wayne that the exclusion clause in an insurance contract is central to the bargain between the parties if only because it is a key factor in determining the level of premium charged; therefore, the insurer should not be held liable in any case where one proximate cause falls within that clause. On the other hand, one might say that the application of the exception is unclear and, therefore, should be construed against the insurer. This would lead to the same result as was reached in Partridge. Yet, even if one preferred Partridge it is not necessarily appropriate simply to discard Wayne. Both Wayne and Partridge involved third party liability policies and, while the reasoning in Partridge might persuade one to prefer it to Wayne, there may be different factors where first party insurance is involved.64 In property insurance the insurer promises to indemnify the insured if the property suffers a covered loss, and generally coverage is defined in terms of causation: the policy requires the loss to have been caused by a specified peril (eg fire or theft), or the policy covers loss caused by any risk with certain exceptions. Since liability depends on a contract analysis, it seems right not to hold the insurer liable where one proximate cause is excluded, even if another is covered, because the loss would not have occurred without the action of the excluded cause. In liability policies, on the other hand, the insurer promises to indemnify an insured, who is liable to another party. The first stage in determining whether the insurer is liable is, therefore, determined by concepts of liability drawn from tort: is the insured liable in tort? Whereas the first party insurer promises to pay for losses irrespective of fault, in liability insurance a claim only exists if the loss is caused by a degree of fault rendering the insured liable to a third party. There may, therefore, be an argument for distinguishing between liability and property insurance: holding the liability insurer, but not the property insurer, liable where one proximate cause is covered and the other is expressly excluded.
The Supreme Court gave some consideration to these issues in The Cender Mopu.65 The Court effectively overruled The Miss Jay Jay by concluding that there could not be loss caused both by perils of the sea and by inherent vice of the subject-matter. Lord Clarke remarked that the only question is:
whether the loss has been proximately caused, at least in part, by perils of the seas (or, more generally, any fortuitous external accident or casualty). If that question is answered in the affirmative, it follows that there was no inherent vice, thereby avoiding the causation issues that arise where there are multiple causes of loss, one of which is an insured risk and one of which is an uninsured or excluded risk.66
The effect of the decision is, therefore, to narrow the circumstances in which the courts can find concurrent causes, but there was little discussion of that issue or of the effect of one cause being either uninsured or an excepted peril. Towards the end of his judgment, Lord Mance distinguished the cases cited in The Miss Jay Jay and Midland Mainline Ltd v Eagle Star Insurance Co Ltd,67 which involved situations where a specific situation had been expressly taken out of cover, and those situations where there are ‘two concurrent risks arising independently but combining to cause a loss. While it may be that the same principle applies (as the Court of Appeal’s dicta in the Miss Jay Jay suggest), I would at least wish to hear argument on that. I need not go further into this aspect, upon which I have formed no concluded views.’68 On the issue of whether inherent vice was an excepted peril in the policy at issue in The Cender Mopu, Lord Clarke’s view was that the reference to inherent vice was part of the definition of the proximate cause: in effect, it was not an excepted peril, but an illustration of a type of occurrence that did not amount to a proximate cause of loss.
4. The impact of deliberate actions by the insured69
4.1 Actions taken to reduce loss
It is not unusual for an insured to take action which, although aimed at reducing the loss for which the insurers would be liable, leads to damage. This does not necessarily break the chain of causation. Indeed, the consequence of finding that it did might not suit insurers since it could discourage action to reduce loss: an insured might not pour water on goods to put out a fire if doing so affected a claim on a fire policy. The issue is whether the insurer can be said to have contracted for the event that caused the loss. In Canada Rice Mills Ltd v Union Marine and General Insurance Co,70 the captain of the Segundo closed the ventilators in the cargo hold to prevent seawater entering during rough weather and this resulted in the insured cargo of rice being damaged through overheating. The Privy Council took the view that, although sea water did not touch the cargo, the loss was caused by the rough sea, and therefore the insurers were liable on a policy that covered loss by perils of the sea: Lord Wright called the action of the ship’s master, ‘such a mere matter of routine seamanship necessitated by the peril that the damage can be regarded as the direct result of the peril’.71 In other words, like the orders issued by the harbour authorities in Leyland Shipping, his action did not break the effect of the perils of the sea. Where the action taken by the insured is designed to protect the insured property from a peril covered by the policy, the test is: ‘Is it a fear of something that will happen in the future or has the peril already happened and is it so imminent that it is immediately necessary to avert the danger by action?’72 Fire insurers were held liable when a fire broke out near the insured goods and the port authorities decided to throw some of the goods into the sea and douse the rest with water to prevent the fire spreading.73 In another such case the judge said, ‘I have found that fire did not actually break out, but it is reasonably certain that it would have broken out, and the condition of things was such that there was an actual existing state of peril of fire, and not merely a fear of fire.’74 On the other hand, the courts will not hold insurers liable for a loss for which they have not contracted. A mistake by a ship’s captain in thinking that steam emerging from the cargo hold was smoke did not make the insurers liable for damage caused to the goods when water was sprayed on them. The reasonableness of the mistake was irrelevant:
The underwriters insured against fire in fact, and if there had been a fire they would have had to pay. But why are they to pay if in fact there was no fire? They did not insure against an error of judgment on the part of the captain in deciding whether there was a peril or not.75
In the case of marine policies, the Marine Insurance Act 1906, section 78(4) places the insured under an obligation to take such action to avert or to minimise loss from an insured peril as the prudent uninsured person would have taken,76 although the effect of this is limited to rare situations where negligence or misconduct breaks the chain of causation—Phillips LJ observed that there appeared to have been no examples of the subsection providing insurers with a defence to a claim since the enactment of the legislation in 1906.77 No requirement equivalent to section 78(4) is implied into non-marine policies, although an insured who unreasonably fails to mitigate their loss may be unable to claim for the additional loss if it is determined that it was proximately caused by the insured’s failure because the chain of causation from the original cause has been broken.78 In Yorkshire Water Services Ltd v Sun Alliance & London Insurance plc,79 sewage escaped from a sewage works owned by Yorkshire Water Services (YWS). To reduce the likelihood of damage to neighbouring property and, therefore, of claims against YWS, the company undertook works costing £4 million. YWS sought to recover this expenditure under its public liability policy, which provided cover against ‘legal liability for damages’ and ‘all other costs and expenses in relation to any matter which may form the subject of a claim’. The Court of Appeal took the view that the insurer was only required to indemnify YWS for sums due as compensation to third parties and not costs incurred by YWS in carrying out work on its own property. The suggestion that a term should be implied into the policy rendering the insurers liable for reasonable expenditure on work undertaken to alleviate loss was rejected, the Court pointing out that it would be difficult to assess what was reasonable: if the potential liability were a claim for compensation of £300,000 from the owner of neighbouring property would it be reasonable to require the insurers to meet the sum of £4 million paid for the works? This would become even more difficult if the damage to the neighbouring property had not been quantified. In the end, however, the case rested on the issue of what the insurer had agreed to cover.80
4.2 Loss resulting from an action of the insured
Property insurance will cover loss caused by the negligence of the insured, unless expressly excluded.81 In Harris v Poland,82 the insured lit a fire in the fireplace, forgetting that she had previously hidden her jewellery there. The court decided that, while her actions were probably negligent, she did not act intentionally, and the insurers were therefore liable under a policy which covered ‘loss or damage by fire’. A requirement that the insured take ‘reasonable care to avoid loss’ will not necessarily relieve the insurers of liability in the event of the insured acting negligently.83 With liability insurance, an attempt to exclude negligence would make no sense because the policy covers liability to a third party and that depends on the insured having been negligent.84 In a case on an employers’ liability policy, it was held that the insurers were liable in spite of the presence of such a clause unless the conduct was reckless, which required the insured to have acted
with actual recognition … that a danger exists, and not caring whether or not it is averted. The purpose of the condition is to ensure that the insured will not, because he is covered against loss by the policy refrain from taking precautions which he knows ought to be taken.85
In Sofi v Prudential Assurance Co Ltd, 86 a theft policy required the insured to ‘take all reasonable steps to safeguard any property insured’. The insured locked jewellery worth £42,035 in the glove compartment of his car and then left the car in a car park for 15 minutes. During this time the jewels were stolen. The insurers were liable because it could not be shown that the insured had acted recklessly in thinking that the jewels were safer in the car. Similarly, in Cooke v Routledge, 87 Mr Cooke wrote off his car when driving while heavily intoxicated. His insurers were held liable for the loss in spite of a clause saying that he was required to take reasonable care to safeguard the car from loss. The court took the view that the crash had not been a deliberate act and that it had been neither the inevitable nor the natural and probable consequence of his action. In any case, the term was construed as requiring Mr Cooke to safeguard the car from external threat and did not relate to the manner of his driving. On the other hand, in Gunns v Par Insurance Brokers 88 the insurer was not held liable because the insured’s conduct was reckless. The insured, a jeweller, went away for the weekend, leaving valuables in a safe, which the insurers had previously declared to be unsatisfactory, and without activating the alarm system. In addition, shortly before the theft, he had reported a suspicion that he was being followed.