THE LORDS GOFF AND HOBHOUSE MEMORIAL LECTURE
DOOMED FROM THE START: INEVITABILITY OF LOSS IN COMMERCIAL LAW- INHERENT VICE AND EXTERNAL CAUSES
LORD JUSTICE FLAUX
Introduction
It is a great honour to have been invited by 7 King’s Bench Walk to give the inaugural Goff/Hobhouse memorial lecture in memory of two of the greatest common law and commercial lawyers and judges of our time. I have chosen to speak about a subject which has taxed commercial lawyers for many years and which seems to me to illuminate the nature and purpose of insurance: inevitability of loss and the extent to which such inevitability precludes recovery. It could also be expressed as its obverse: the requirement of fortuity or uncertainty, of “risk”.
My focus will be on commercial contracts, specifically insurance and to an extent carriage, but the concept of the inevitability of an occurrence precluding recovery is not limited to the law of contract. In the law of tort for example, where the relevant event or occurrence would have taken place anyway, even if the tort had not been committed, the tortfeasor will not be liable or at least the claimant will not usually recover substantial damages. This is sometimes analysed as the claimant not having suffered any loss or as a matter of causation, that if there was loss, it was not caused by the tort. However, as we will see, in the context of the law of contract at least, there is no absolute rule that a loss which was objectively bound to happen is not recoverable.
The nature of insurance
Perhaps surprisingly given the extent of statutory regulation of insurance business, there is no statutory definition of insurance. What MacGillivray on Insurance Law describes as a useful working definition is derived from the judgment of Channell J in Prudential Insurance v Inland Revenue Commissioners [1904] 2 KB 658 at 663:
“It must be a contract whereby for some consideration, usually but not necessarily for periodical payments called premiums, you secure to yourself some benefit, usually but not necessarily the payment of a sum of money, upon the happening of some event. Then the next thing that is necessary is that the event should be one which involves some amount of uncertainty. There must be either uncertainty whether the event will ever happen or not, or if the event is one which must happen at some time there must be uncertainty as to the time at which it will happen. The remaining essential is that which was referred to by the Attorney-General when he said the insurance must be against something. A contract which would otherwise be a mere wager may become an insurance by reason of the assured having an interest in the subject-matter—that is to say, the uncertain event which is necessary to make the contract amount to an insurance must be an event which is prima facie adverse to the interest of the assured.”
The context of that case was life assurance. It was one of a series of cases decided before the First World War at a time when life assurance was being developed and expanded as a form of financial investment. For the purposes of this lecture, I am focusing on the third requirement for a contract of insurance, that there must be an element of uncertainty in relation to the happening of the event which is the subject of the insurance. As Channell J says, the uncertainty may be as to whether the event will happen at all but equally, where it is bound to happen, the requirement will be satisfied by uncertainty as to when it will happen. Thus, in the context of life assurance, death is inevitable, but when death will occur is uncertain.
A modern example of this principle in play is Fuji Finance v Aetna Life [1997] Ch 173. The claimant took out a contract with the defendant variously described as a policy of life assurance or a capital investment bond, under the terms of which a sum calculated by reference to the price of units currently allocated to the policy was payable on the death of the life assured or on its earlier surrender. The claimant paid a single premium of ￡50,000, which was applied to secure units in a variety of internal funds administered by the insurance company. The policyholder had the option to switch units allocated to his policy between the funds. One of the issues which arose was whether this contract was a contract of insurance enforceable but subject to the Life Assurance Act 1774, which requires the beneficiary of such a policy to have an insurable interest. That Act was passed to remove the then perceived mischief of using life assurance to gamble on the duration of the life of someone with whom one had no connection, which it was feared might lead to murder.
The judge at first instance in Fuji concluded that it was not a contract of insurance because the death of the life assured was not an uncertain event triggering a payment in circumstances where, irrespective of that event, there was an obligation on the defendant to make the same payment on request on early surrender. The Court of Appeal, which included Hobhouse LJ, reversed the judge, on the basis that both the crystallisation of benefits on the death of the life assured and the right to surrender and trigger payment by that means (which could only be exercised during the survival of the life assured) were, as Morritt LJ put it, “sufficiently life or death related”. Hobhouse LJ considered that the requirement for this to be insurance was satisfied because the rights to benefits under the policy were always contingent on death or survival of the life assured, which was an uncertain event. Thus, the requirement of uncertainty will be satisfied even if the relevant event is bound to occur, provided that when it occurs remains uncertain. This is a concept to which I will return later.
In the context of marine insurance, this requirement of uncertainty or fortuity is expressed in the oft-cited dictum of Lord Sumner in British and Foreign Marine Insurance v Gaunt [1921] 2 AC 41, a case of all risks cargo insurance:
“There are, of course, limits to “all risks.” They are risks and risks insured against. Accordingly the expression does not cover inherent vice or mere wear and tear... It covers a risk, not a certainty; it is something, which happens to the subject-matter from without, not the natural behaviour of that subject-matter, being what it is, in the circumstances under which it is carried. Nor is it a loss which the assured brings about by his own act, for then he has not merely exposed the goods to the chance of injury, he has injured them himself.”
There is no provision in the Marine Insurance Act 1906 which mandates this requirement of uncertainty or fortuity or which excludes inevitability of loss, but as Donaldson LJ put it in Soya G.m.b.H. v White [1982] 1 Lloyd’s Rep 136, a case on inherent vice (argued on behalf of insurers by Mr John Hobhouse QC) to which I will return: “Inevitability of loss operates at a much more fundamental level than the rule that underwriters are only liable for losses proximately caused by perils insured against. Underwriters can rely upon inevitability of loss because the whole concept of insurance is about risks, not certainties.”
It is important to note though that Donaldson LJ preferred the term “known certainty” to “inevitability”. This was to recognise the long standing practice in Lloyd’s (at least in the days of sailing ships and before instantaneous communication) of insuring overdue ships and cargo, where unknown to the insurer and the assured the ship and its cargo might have already sunk at the time the insurance was taken out. Hence, as Donaldson LJ pointed out, the provision in the old SG policy appended to the Marine Insurance Act: the “lost or not lost” clause. In such a case it could be said the loss was inevitable, but this was not known to either party, hence the availability of such insurance. Under that clause, the assured could recover even though he had not acquired his interest in the ship or cargo until after the loss, providing that he was unaware of the loss at the time of effecting the insurance: Sutherland v Pratt (1843) 11 M & W 296. There is no equivalent provision in the 1983 Institute Clauses or Hull Clauses and a narrower provision in clause 11 of the Institute Cargo Clauses, under which the assured will only be able to recover in respect of loss which has already occurred if it had an insurable interest in the cargo at the time of the loss.
Donaldson LJ contrasted that position of “unknown certainty” with one where one or other party knows the loss has occurred. As he said, if the insured knows of the loss and does not disclose it, any insurance will be avoidable for material non-disclosure. Even under the Insurance Act 2015, such non-disclosure would be deliberate. Equally, as he said: “no underwriter will insure a loss if he knows it has occurred and its extent” which in the light of some of the business underwritten in Lloyd’s by way of run-off reinsurance or as part of the ill-fated LMX spiral may seem with the benefit of hindsight to be misplaced confidence.
Losses brought about by the assured
An obvious example of known certainty is where the relevant loss can be said to have been deliberately caused by the assured itself. The paradigm example of this is wilful misconduct such as the scuttling of the ship with the connivance of the shipowner. Samuel v Dumas [1924] AC 431 illustrates this. The claimant assured was in fact the mortgagee who was not privy to the owner mortgagor’s actions in procuring his crew to scuttle the ship. One of the principal issues was whether, given that the assured was not infected by the fraud of the owner, it could recover under the policy for a loss by perils of the sea. Discerning a clear ratio from the decision is not easy: the claim in fact failed on the ground of a separate breach of warranty as determined by a majority of their Lordships, whereas a different majority expressed views that there was no loss by perils of the sea. Viscount Haldane LC helpfully concurred in the result but gave no reasons. In relation to perils of the sea, both Viscount Cave and Viscount Finlay (with whom Lord Parmoor concurred) considered that the proximate cause of the loss of the vessel was the deliberate sinking of the vessel which was not a peril of the sea. They considered the earlier decision of the Court of Appeal in Small v United Kingdom Marine Mutual Insurance Association [1897] 2 QB 311 to have been wrongly decided. As Viscount Finlay put it:
“The scuttling of this vessel occurred on the seas, but it was not due to any peril of the seas; it was due entirely to the fraudulent act of the owner. The scuttling was not fortuitous, but deliberate, and had nothing of the element of accident or casualty about it. Storms are "fortuitous"; the ordinary action of the waves is not, and fraudulent scuttling is even more decisively out of the region of accident. The entrance of the sea water cannot for this purpose be separated from the act which caused it.”
Lord Sumner gave a strong dissenting opinion on this point. He considered that there was still a loss by perils of the sea even though the reason the water entered was the deliberate act of the owner or his crew. As he put it:
“To say that the proximate cause of the sinking was the instructions given by [the owner] and was not the entrance of water seems to me to give a new meaning to proximate cause, and if for this purpose the acts of his agents on board are regarded as his acts I think the result is still the same. A ship is none the less burnt and destroyed by fire because the striking of the match was an act of arson.”
His analysis was that the loss was by a peril of the sea, but that the owner could not have recovered under the policy because by section 55(2)(a) of the Marine Insurance Act, the insurer is not liable for the wilful misconduct of the assured. He considered that this analysis was not only supported by the decision in Small but by dicta in earlier cases. In his interesting and illuminating article Fortuity in the law of marine insurance [2007] LMCLQ 315, Professor Howard Bennett puts forward spirited support for the Sumner analysis. My own view is that the dicta of the majority are to be preferred: the wilful misconduct of the owner was the proximate cause of the loss, which as Viscounts Cave and Finlay said was thus deliberate and not fortuitous.
Deliberate acts of third parties
As Professor Bennett points out, the opinions in Samuel v Dumas leave open whether a loss caused by the deliberate act of a third party not the assured or its agents could be a loss by perils of the sea, although the issue may be an academic one, given that many forms of policy will expressly cover the relevant deliberate act as a separate insured peril, such as piracy or terrorism or in broader terms, persons acting maliciously or from a political motive. A case which illustrates the limitations on insurance cover in such circumstances is the decision of the Court of Appeal in Atlasnavios v Navigators Insurance (The B Atlantic) [2016] 2 Lloyd’s Rep 351. There, a vessel was detained indefinitely by the Venezuelan authorities after three bags of cocaine were found strapped to her hull below the waterline at the load port. The claim was for constructive total loss under a hull war risks insurance which included cover for loss or damage to the vessel caused by detainment or caused by “…any person acting maliciously or with a political motive”. The policy also contained an exclusion of loss and damage arising from detainment by reason of infringement of any customs and trading regulations.
The insurers accepted that the shipowners were not implicated in the drug activities and that the drugs had been fixed to the vessel by persons unknown who were presumably members of a drug cartel smuggling drugs from South America to Europe, where the vessel was due to discharge her cargo. However, the insurers denied liability on the basis that the detention of the vessel arose from infringement of Venezuelan customs regulations within the meaning of the exclusion. The judge at first instance found for the assured on the basis that, as a matter of construction, the exclusion did not apply. Given that it was accepted that the strapping of the drugs to the hull was a malicious act and that the insurers accepted that the exclusion would not apply to a “put up job” where, for example, the authorities deliberately planted the drugs so as to be able to detain the vessel, the judge considered no distinction could be drawn between the two situations. He held that the exclusion was subject to the implied limitation that it did not apply where the only reason why there was an infringement of the customs regulations was because of malicious acts of third parties. In such a case the infringement was no more than the manifestation of the relevant acts of such third parties acting maliciously.
This analysis was rejected by the Court of Appeal which allowed the insurers’ appeal. In his judgment Christopher Clarke LJ considered that on the judge’s findings of fact, the constructive total loss was caused both by the malicious act of the drug smugglers and by the detainment of the vessel by reason of the infringement of customs regulations. So far as the first cause was concerned, the malicious act was the concealment of the drugs on the hull, which constituted the infringement of the customs regulations. As for the second cause, the detention occurred by reason of the concealment which constituted the infringement. Since the limitation the judge had placed on the exclusion was not justified for a number of reasons, not relevant to the present debate, the exclusion operated to exclude the claim[1].
What is of relevance to the present debate is how Christopher Clarke LJ dealt with the various put-up jobs posited, not just where the authorities deliberately placed the drugs on the hull in order to detain the vessel, but other possibilities suggested by the judge: where a malicious third party planted the drugs and then demanded money from the managers as the price of silence and when they did not pay informed the authorities who detained the vessel and where the malicious third party simply planted the drugs and then informed the authorities without any intervening blackmail. In relation to these various put-up jobs, Christopher Clarke LJ analysed how the exception might not apply to them at [53]:
“the “put-up job” cases are all examples of circumstances in which the court may well be able to find that detention did not in truth occur by reason of infringement of customs regulations (if the authorities were behind it all along) or that the stratagem of the authorities or the malice of the villain was the proximate cause of the loss. In each case—planting by the authorities and planting by the villain either with or without blackmail—detention is the inevitable and intended consequence of the primary actor. After that malicious act there is no further fortuity. Detection and detention are inevitable and intended; and in the case of the first example “detection” is itself an illusion since the authorities know the position all along, and are the planners of it. Mr Schaff [counsel for the assured] submitted that the policy was intended to cover all types of malicious acts and not to exclude a subset of acts which were reckless and not intentional. I would view it differently. The policy intended to exclude a subset of loss which arose from detainment by reason of breach of customs regulations. If it is clear that the loss was so caused the exception applies.”
The problem with that analysis is that it is difficult to see how the causative effect of the malicious act can vary depending upon whether the act is reckless or intentional. Causation of loss under the insurance surely cannot depend on the intention of the third party and on one view, without some sort of implied limitation, the exception continues to operate even where the malicious act is intentional because the detention is still by reason of infringement of customs regulations and remains an effective cause of the loss. Where there are two effective or proximate causes of the loss, one an insured peril and the other the subject of an exclusion, then the claim fails: see Wayne Tank and Pump Ltd v Employers Liability Assurance [1974] QB 57 at 75; The Cendor MOPU [2011] 1 Lloyd’s Rep 560 at [22] per Lord Saville and [88] per Lord Mance[2] and The B Atlantic at [26] per Christopher Clarke LJ
Inherent vice
Section 55(2)(c) of the Marine Insurance Act provides:
“Unless the policy otherwise provides, the insurer is not liable for ordinary wear and tear, ordinary leakage and breakage, inherent vice or nature of the subject matter insured, or for any loss proximately caused by rats or vermin, or for any injury to machinery not proximately caused by maritime perils.”
The particular exclusion on which we are currently focusing is “inherent vice or nature of the subject matter insured”. The definition of this phrase which is now generally accepted is that of Lord Diplock in Soya G.m.b.H v White [1983] 1 Lloyd’s Rep 122 at 126:
“This phrase (generally shortened to “inherent vice”) where it is used in section 55(2)(c) refers to a peril by which a loss is proximately caused; it is not descriptive of the loss itself. It means the risk of deterioration of the goods shipped as a result of their natural behaviour in the ordinary course of the contemplated voyage without the intervention of any fortuitous external accident or casualty.”
Importantly, as this definition recognises, the fact that the phrase is usually shortened to “inherent vice” does not mean that it only applies where there is some vice or defect rendering the cargo unsound. It encompasses the inherent qualities of an otherwise normal or sound cargo, for example the tendency of fruit to ripen and become rotten. As Bingham LJ put it in Noten BV v Harding [1990] 2 Lloyd’s Rep 283: “Taken alone, the expression “inherent vice” may be capable of misleading, since it may suggest some defect in the goods and these goods were not in any ordinary sense defective.”
The reference by Lord Diplock to inherent vice being a peril by which a loss is proximately caused highlights an important point, that, as Donaldson LJ said in the Court of Appeal in that case, although inherent vice and inevitability of loss overlap, they are “quite separate and distinct concepts”. The fact that a cargo has an inherent vice or nature or, in the case of a ship, that there is a latent defect in the hull (as Professor Bennett says, latent defect is a sub-species of inherent vice) does not mean that a loss is bound to occur on the voyage: it may or may not happen. Hence, the opening words of sub-section 55(2)(c) recognise that the policy may “otherwise provide”, in other words that it may be possible to take out insurance against the risk of a loss occurring by reason of inherent vice. This was recognised by Atkin LJ in Sassoon v Yorkshire Insurance (1923) 16 Ll.L.R. 129 at 133:
“I think it is quite plain from the words of [the sub-section] that a policy may provide, if it is done in express words, for the insurer being liable for losses which are excepted…The particular kind of loss, the amount of the loss, is one which…may or may not happen, and not one which certainly must happen; if it was a loss which certainly must happen within the voyage I doubt whether it could ever be made properly the subject-matter of a policy of insurance.”
That last point may overstate the position, for reasons I will come to in a moment.
Soya v White was itself an example of insurance against “heat, sweat and spontaneous combustion” of a cargo of soya beans. All the courts considered that sweat and spontaneous combustion were referring to something which could only take place inside the goods themselves, in other words particular kinds of inherent vice, from which it follows that, as Lord Diplock put it: “…‘heat’ appearing in immediate conjunction with them is apt to include heating of the cargo as a result of some internal action taking place inside the cargo itself”. In that case, the findings of fact were that the moisture content of the soya beans was within a grey area where deterioration from heat might or might not occur during the course of the voyage so that such deterioration was not inevitable.
Lord Diplock left open the issue whether the insurers would have been liable if, unknown to the assured, the moisture content on shipment had been so high as to make deterioration inevitable, on which the House of Lords had heard no argument and so expressed no opinion. It would appear, at least from the view of Donaldson LJ in the Court of Appeal in that case that he would have considered that, if both the assured and the insurers were unaware of the inevitability, it would not have been a case of “known certainty” and thus potentially the insurers would have been liable. In The Cendor MOPU [2011] 1 Lloyd’s Rep 560 at [51] Lord Mance returned to the question Lord Diplock had left open in Soya v White, although again because the loss was not inevitable, the point was academic and there do not seem to have been any submissions about it. He considered that inevitability resulting from the inherent characteristics of the cargo but unknown to both parties would, in the absence of an express provision affording cover, amount to inherent vice excluding liability. He considered that whether inevitability resulting from outside causes unknown to the parties would bar recovery was an open question.
He gave the example of a war risk insurance on cargo placed at a time when the cargo was already on board an aircraft in flight with a timed bomb due to go off in ten minutes in the cargo hold, but did not examine the question further. It seems to me that in principle there is no reason why insurers should not be liable in such circumstances. Although the loss is objectively inevitable, that is subjectively unknown to either party. In just the same way, with an overdue ship, the loss may be objectively inevitable because it has already occurred, but it is well-established that such a form of insurance is valid and insurers will be liable for a loss.
After the decision of the House of Lords in Soya v White, the question remained as to the scope of the exclusion of inherent vice or nature of the subject matter insured, in particular as to what was encompassed by “the ordinary course of the contemplated voyage” in Lord Diplock’s definition. Mayban General Insurance v Alstom Power Plants Ltd [2004] 2 Lloyd’s Rep 609 concerned damage to a transformer carried on a feeder voyage from Ellesmere Port to Rotterdam in weather conditions which were severe but not in any sense unexpected or extraordinary. Moore-Bick J held the insurers were not liable on the basis that the exception for inherent vice applied. The basis for his decision was that, given that the weather conditions encountered on the voyage were no more severe than could reasonably have been expected, the real cause of the loss was the inherent inability of the goods to withstand the ordinary incidents of the voyage, which amounted to inherent vice.
In his article in 2007, Professor Bennett was highly critical of this decision, on the basis that it gave a wide meaning to inherent vice which whilst it might be justified in the context of contracts of carriage was not justified in contracts of insurance, on the basis that, where as a matter of causation, the loss was proximately caused by the interaction of the condition of the cargo and the perils encountered on the voyage, in other words perils of the sea, the insurers remained liable, the view consistently held by the editors of Arnould.
Professor Bennett’s analysis was vindicated by the decision of the Supreme Court in The Cendor MOPU [2011] 1 Lloyd’s Rep 560. That case concerned marine insurance of an oil rig purchased by the assured for conversion into a mobile offshore production unit for use off East Malaysia and carried on a towed barge from Galveston, Texas to Malaysia. From the outset it was known that the three legs of the rig (each 312 feet high and weighing over 400 tons) were at risk of fatigue cracks during the voyage. The rig was inspected by approved surveyors at Galveston and it was a condition of the policy that the surveyors approved the arrangements for the tow. They issued a certificate of approval which required re-inspection of the legs at Cape Town, effectively the mid-way point of the voyage, for crack initiation so that repairs could be effected as necessary. When the rig was inspected at Cape Town, it was found that a considerable degree of fatigue cracking had occurred and some repairs were made to reduce the stress concentrations. However, seven days after the voyage resumed, when the tug and barge were north of Durban the starboard leg broke off at the 30 feet level and fell into the sea. The following evening the forward leg broke off at the same level and half an hour later the port leg broke off at the 18 feet level, both falling into the sea.
The mechanism of failure was that initial fatigue cracks in the corner of pinholes propagated until they were subjected to “leg breaking” stress. Once one leg had failed, the stresses on the others increased. The stresses were generated by the effect that the height and direction of the waves had on the pitching and rolling motion of the barge and thus the legs. It was common ground that the weather conditions experienced were within those which could reasonably have been contemplated on the voyage. However, the expert evidence at trial was that a developed crack would not on its own be sufficient to cause one of the legs to come off. As one of the experts put it: “you’ve got to catch it just right if you want to make it actually fail all the way round.” This was the concept of the “leg-breaking wave”. Equally, the loss was not inevitable. As the judge found, the failure and loss of the legs was very probable but not inevitable.
At first instance, Blair J accepted the insurers’ argument that the proximate cause of the loss was inherent vice excluded under the insurance because the legs were not capable of withstanding the normal incidents of the insured voyage including the weather reasonably to be expected, following the reasoning of Moore-Bick J in Mayban. The Court of Appeal allowed the appeal, holding that the proximate cause of the loss was an insured peril in the form of the occurrence of a leg-breaking wave. That decision was upheld in the Supreme Court. Both Lord Saville and Lord Mance were of the view that the insurers’ argument that inherent vice encompassed the inability of the cargo to withstand the ordinary incidents of the voyage, including the weather conditions, rested uneasily with the seaworthiness provisions of the Marine Insurance Act. In the case of time hull policies (and most modern hull insurance is underwritten on a time rather than a voyage basis) there is no implied warranty of seaworthiness as in a voyage policy, but section 39(5) of the Act provides that where the ship is sent to sea in an unseaworthy state with the privity of the owner, the insurer is not liable for any loss attributable to unseaworthiness. However, section 40(1) of the Act provides that in a policy on goods, there is no implied warranty that the goods are seaworthy. Against the background of those provisions, as Lord Saville pointed out at [43], the effect of the insurers’ argument for their wide meaning of inherent vice would be that, whereas the assured under a time hull policy would be covered for loss attributable to the unseaworthiness of the vessel to which he was not privy, the cargo owner would not be covered against loss attributable to the unseaworthiness of the cargo. Nothing in the Act justified that distinction.
Accordingly, the Supreme Court rejected the insurers’ wide definition of inherent vice and held that Mayban had been wrongly decided. Where, as in the case before them, there were two causes identified, initial unfitness of the cargo to withstand the weather conditions which might be encountered on the voyage and a peril of the sea through which it worked (or any other fortuitous external cause), the sole question to be asked in applying Lord Diplock’s definition of inherent vice in Soya v White (which all their Lordships considered to be correct) was whether the loss and damage was proximately caused by perils of the sea or, more generally, any other fortuitous external cause, in which case there was no scope for the application of the exception for inherent vice.
This conclusion was enunciated most clearly by Lord Mance at [80] and [81]:
“If inability to withstand foreseeably bad weather conditions does not prevent damage sustained as a result being attributed to perils of the sea, (i) that must be because Lord Diplock's reference to “the ordinary course of the contemplated voyage” was not intended to embrace the weather conditions foreseeable on such a voyage, but was rather used as a counterpoint to a voyage on which some fortuitous external accident or casualty occurred and (ii) there is no apparent limitation in Lord Diplock's qualification “without the intervention of any fortuitous external accident or casualty” – in other words, on the face of it, anything that would otherwise count as a fortuitous external accident or casualty will suffice to prevent the loss being attributed to inherent vice…
On this basis, it would only be if the loss or damage could be said to be due either to uneventful wear and tear (or “debility”) in the prevailing weather conditions or to inherent characteristics of the hull or cargo not involving any fortuitous external accident or casualty that insurers would have a defence. In the scheme of the 1906 Act, that would not appear to me surprising, bearing in mind the case law against the background of which the Act was enacted and the juxtaposition in section 55(2)(c) of “ordinary wear and tear, ordinary leakage and breakage” with “inherent vice or nature of the subject-matter insured” as well as with “any injury to machinery not proximately caused by maritime perils”. While not myself attempting any exact definition, ordinary wear and tear and ordinary leakage and breakage would thus cover loss or damage resulting from the normal vicissitudes of use in the case of a vessel, or of handling and carriage in the case of cargo, while inherent vice would cover inherent characteristics of or defects in a hull or cargo leading to it causing loss or damage to itself — in each case without any fortuitous external accident or casualty.”
In other words, following the decision of the Supreme Court, a defence of inherent vice will not be open to insurers where one of the proximate or effective causes of the loss is a fortuitous external accident or casualty amounting to an insured peril, such as perils of the sea. The question therefore arises, on the basis that the exception is narrowly interpreted, in what sorts of circumstances will it apply?
Obviously, deterioration of a cargo on board a ship does not occur in a vacuum. External factors such as atmospheric, weather or sea conditions may have contributed causally to the loss, as Lord Clarke recognised in The Cendor MOPU at [114]. Where those external conditions are such as are normal and expected to occur and do not amount to an insured peril such as perils of the sea[3], then the deterioration of the cargo is likely to amount to inherent vice. A classic instance of this principle at play in the modern law is to be found in the decision of the Court of Appeal in Noten B.V. v Harding [1990] 2 Lloyd’s Rep 283. That case concerned consignments of gloves carried in containers, which had suffered moisture damage during the transit. The trial judge found that the moisture came from the gloves themselves, not from the air inside the containers at the time they were stuffed, but that the moisture had condensed on the roof of the containers during the carriage and dropped onto the cargo. The judge held that the damage was caused by an insured peril, the dropping of water from an external source onto the goods. That decision was reversed by the Court of Appeal, which found that the proximate cause of the damage was inherent vice.
Bingham LJ said that the question was what was the real or dominant cause of the damage? That was to be answered applying the common sense of a business or seafaring man, who: “would not understand how the water which had caused the damage could be regarded as coming from a source external to the goods, but would on the uncontradicted findings regard the gloves as the obvious and sole source of the water. He would, I think, regard the suggested distinction based on the intermediate migration of moisture to and condensation of moisture on the roofs of the containers as owing more to the subtlety of the legal mind than to the common sense of the mercantile.”
Bingham LJ went on to conclude, applying Lord Diplock’s definition:
“If the factual cause of the damage to these gloves has been correctly identified, then I think it plain that that was an excepted peril under these policies. The goods deteriorated as a result of their natural behaviour in the ordinary course of the contemplated voyage, without the intervention of any fortuitous external accident or casualty. The damage was caused because the goods were shipped wet.”
The reasoning in that case that there was inherent vice with no intervening fortuitous external accident or casualty was applied and followed, albeit in the context of a contract of sea carriage by the Court of Appeal in Volcafe Ltd v CSAV [2017] QB 915, to which I will return later. Those were cases where the external atmospheric, weather and sea conditions were entirely ordinary and as contemplated, so that there was no intervening peril of the sea or other fortuitous event. The defence of inherent vice succeeded even though the weather conditions had hastened the relevant deterioration of the goods.
Latent defect
In the context of hull insurance, the House of Lords in Thames & Mersey Marine Insurance v Hamilton, Fraser & Co (“The Inchmaree”) (1887) 12 App Cas 484 determined that the breakage of machinery on board a vessel which was unconnected with the vessel being at sea did not give rise to a covered loss by perils of the sea. The effect of that decision is reflected in the exclusion in section 55(2)(c) of “any injury to machinery not proximately caused by maritime perils”. The effect of the decision was however mitigated by an extension introduced in the Institute Time Clauses from soon after the decision, the so-called “Inchmaree Clause” which in its modern form in the Institute Time Clauses Hulls 1983 reads:
“6.2. This insurance covers damage to the subject matter insured caused by…
6.2.2. Bursting of boilers breakage of shafts or any latent defect in the machinery or hull…”
A latent defect in the machinery or hull is a defect in the construction or which comes into existence subsequently which results in the vessel being in a more defective condition than it should have been. It is to be contrasted with the effects of ordinary wear and tear which is not covered under the Inchmaree Clause: see per Robert Goff J in The Caribbean Sea [1980] 1 Lloyd’s Rep 338 at 347 citing various US authorities and a dictum of Scrutton J in CJ Wills & Sons v The World Marine Insurance (1911). In The Caribbean Sea, Robert Goff J considered that “latent” had the same meaning as it bears in contracts of carriage: a defect which could not be discovered on such examination as a reasonably careful skilled man would make, although there was no argument on the point in that case.
A latent defect is also to be contrasted with a defect in design, which means that the vessel is constructed as it was supposed to be but the design results in the vessel not being fit for the purpose. In an obiter dictum in Jackson v Mumford (1902) 8 Com Cas 61, Kennedy J said that a defect in design was not covered by the Inchmaree Clause. That case concerned a torpedo-boat destroyer called the Bullfinch being built for the Admiralty which her designers were trying to propel at a novel rate of speed for that time at the end of the nineteenth century, attaining a very high power on the least possible weight by working with a very low factor of safety. On one of her sea trials, the connecting rod of the starboard high pressure engine broke, resulting in the death of eight men. Kennedy J considered that this was not a latent defect in the machinery within the Inchmaree Clause, which did not cover:
“the erroneous judgment of the designer as to the effect of the strain which his machinery will have to resist, the machinery itself being faultless, the workmanship faultless and the construction precisely that which the designer intended it to be.”
Robert Goff J in The Caribbean Sea doubted the analysis of Kennedy J. He said that; “Another possible interpretation of the facts is that since the casualty occurred during the trials of a ship in whose design risks were deliberately being taken, the proximate cause of the casualty was the deliberate running of the risk rather than anything which could properly be called a defect in the machinery.” For my part, I doubt this explanation of the decision can be correct, because Kennedy J actually decided that the assured’s claim succeeded on the alternative basis that the perils insured against included “trials” and there was an express leave to go on trial trips. Kennedy J held that trials were a peril insured against and rejected the insurers’ arguments that: (i) trials cover was limited to fire risks; and (ii) that, as the machinery had run in the trial at contract speed, the damage resulting inevitably from the faulty design was a consequence which must happen and not a casualty that might happen. In other words the judge rejected an inevitability or deliberate running of the risk analysis, so I cannot see how Robert Goff J’s interpretation of the facts works. Kennedy J’s decision in favour of the assured was upheld by the Court of Appeal where Lord Alverstone CJ held that the parties must have known the vessel would be put to very severe tests by trials and it was clear that the intention was that the trials should be the subject matter of insurance.
At all events, whether the dictum of Kennedy J is correct or not, what Robert Goff J did decide in The Caribbean Sea was that where there was a latent defect in the hull or machinery, that would not be excluded from cover under the Inchmaree Clause merely because the historical reason for such latent defect was a defect in design. In other words, where a design defect develops over time into a latent defect, in that case fatigue cracking, that does not preclude cover under the Clause.
Robert Goff J also considered that the exclusion of inherent vice under section 55(2)(c) was clearly inconsistent with the cover provided under the Inchmaree Clause and hence inapplicable to a policy containing that Clause. It is important to note though that the cover under the Clause is for damage to the vessel caused by the latent defect. The mere development of a latent defect so that it becomes patent is not covered. There is sometimes a difficulty in drawing a distinction between the latent defect and the “other” damage which is insured under the Clause. This difficulty is demonstrated by the decision of the Court of Appeal in Promet Engineering v Sturge (“The Nukila”) [1997] 2 Lloyd’s Rep 146. A self-elevating platform went into service in the Java Sea. It had three legs which could be jacked up and down and to stop them sinking into the silt, at bottom of each leg was a 28 ft square spud can, effectively a large steel box with internal bulkheads and brackets, fixed to the column of each leg by circumferential welds. The spud cans and the legs would be subjected to major stresses which were taken into account in the design and specification. However, the actual construction was not carried out in accordance with the specification, in that some of the circumferential welds were not properly profiled.
After some three and a half years in operation without incident, on a routine inspection by divers, serious cracks were observed in the top plates of all the spud cans. Serious cracks were also noted in the metal of the legs themselves and in some of the internal bulkheads of the spud cans. This was a dangerous condition which threatened the safety of the platform, which was towed to Singapore where the full extent of the cracking was revealed including full width cracks of between 10 and 16mm in the legs. In one leg only 30% of the circumference remained intact. Extensive repairs were required and the assured claimed the cost of those repairs from the insurers on the London market who insured under hull policies including the Inchmaree Clause.
At first instance, Tuckey J found for the insurers following earlier decisions to the effect as already noted that if all that has happened is that a latent defect has become patent, there was no cover under the Clause. He considered the flaws in the welds had simply developed into cracks spreading to the adjacent structure and it could not be said as a matter of common sense that anything consequential had happened which could be characterised as damage to the vessel.
That decision was reversed by the Court of Appeal. The main judgment was given by Hobhouse LJ. He recognised that the cracking had occurred as a result of the ordinary working of the platform at sea and the presence of the latent defects in the welds, with no external accident or cause and that the pre-existing defective condition of the vessel could be said to have made the loss something which was bound to happen and therefore not fortuitous. He said these types of question had been addressed in a number of cases on cargo insurance and by The Caribbean Sea in relation to hull insurance and that section 55(2)(c) recognised that by specific agreement a policy could cover such risks. As he said:
“The presence or absence of a latent defect in the hull or machinery of a vessel is, by definition, unknown to the assured and whether or not there is such a defect and whether or not it will during a given period of time or maritime adventure have an impact or cause any damage is fortuitous from the point of view of the assured.”
He noted that the availability of the Inchmaree Clause demonstrated the market for such cover. In other words, this case recognises that even though the latent defect which by definition is unknown to the assured at the inception of the insurance may make damage to the vessel during the period of insurance inevitable, there will still be cover. However, as Hobhouse LJ went on to say, the operation of the Clause required some loss and damage over and above the latent defect becoming patent. The assured had to show “some change in the physical state of the vessel” which if present could be the subject of a claim under the policy. Hobhouse LJ considered that, on the facts of the case, the application of the Inchmaree Clause was straightforward. The latent defects in the welds had caused extensive fractures in the metal of the legs, spud cans and bulkheads which on any ordinary use of language was damage to the subject matter insured. Nothing in the earlier authorities led to a different conclusion. Both Hobhouse LJ and Ward LJ (whose views on commercial disputes are often illuminating although his background was in family law) considered that nothing in those earlier cases supported the insurers’ proposition that the damage had to be to some separate part of the vessel to that where the original defect was located.
Inevitability of loss and inherent vice in contracts of carriage
Historically, inherent vice has long been recognised as an excepted peril even where the carrier was under an otherwise absolute liability as a common carrier at common law. For example in Blower v Great Western Railway (1872) LR 7 CP 655, a bullock being transported in the usual form of cattle truck took fright and escaped over the top rail of the truck, being found dead by the side of the railway later. The escape was wholly attributable to the exertions of the animal itself and the court found it was not attributable to the negligence of the railway company. The Court of Common Pleas held that the company was not liable as the sole cause of the loss of the bullock was its own inherent vice or restiveness or phrensy. Of course carriers by road and rail are no longer common carriers, but the relevant international conventions, the CIM and the CMR contain exceptions for inherent vice.
In the case of contracts of carriage by sea, the Hague or Hague Visby Rules will invariably apply, Article IV rule 2 of which provides that the carrier shall not be responsible for loss or damage arising or resulting from a number of excepted perils, so far as presently relevant: (c) Perils, dangers and accidents of the sea or other navigable waters; and (m) Wastage in bulk of weight or any other loss or damage arising from inherent defect, quality or vice of the goods.
Volcafe v CSAV [2017] QB 915 was a case concerning a number of consignments of coffee beans from Colombia to Germany which had been stuffed into bags which were loaded into dry, unventilated containers lined with kraft paper. The beans were in apparent good order and condition when loaded by the carrier but by the time of their delivery to the consignees they had suffered condensation damage. The Court of Appeal, reversing the judge at first instance, allowed the carrier’s appeal and held that its defences of inherent vice and inevitability of damage succeeded. The principal judgment in the Court of Appeal is concerned primarily with the incidence of the burden of proof under the Hague or Hague-Visby Rules and its effect, which need not presently concern us[4].
In relation to inherent vice dealt with at [53] to [63], the judgment highlighted that inherent vice and inevitability of damage are not the same, as was clear from Soya v White and Noten BV v Harding and said that although those were cases of marine insurance, the concepts should be treated similarly in the context of both carriage by sea and marine insurance, citing a passage to that effect in the judgment of Donaldson LJ in Soya v White. The only additional point worth noting is that, unlike in marine insurance, perils of the sea, far from being the equivalent of an insured peril giving rise to liability on the part of the shipowner, is itself an exception under Article IV rule 2. Whether that should lead to a wider definition of inherent vice in carriage by sea than in marine insurance remains to be worked out. Certainly the current editors of Scrutton on Charterparties cite the limitation on the concept in The Cendor MOPU albeit a case of marine insurance as if it should be equally applicable to carriage by sea.
Having cited the further passage in the judgment of Donaldson LJ drawing the distinction between the two concepts of inevitability of loss and inherent vice and Lord Diplock’s definition of inherent vice, the judgment in Volcafe then referred to Noten BV v Harding and a passage from the judgment of Bingham LJ where he said:
“The suggestion has sometimes been made that inherent vice means the same thing as damage that must inevitably happen, but this is not so. The distinction is between damage caused by any external occurrence, and damage resulting solely from the nature of the thing itself. Damage from inherent vice may be just as capricious in its incidence as damage caused by perils of the seas.”
Accordingly, the judge at first instance had been wrong as a matter of law to equate inherent vice with inevitability of damage and to conclude that inherent vice could not apply to a “normal” cargo of coffee beans. It was common ground between the experts that the damage to the coffee beans was due to condensation and that the source of the condensation was the beans themselves, as with the gloves in Noten BV v Harding. Given that the judge had accepted the evidence of the carrier’s expert as to how the process of condensation occurred once the containers had arrived in Northern Europe, he should have concluded that the carrier had a sustainable defence under Article IV rule 2(m). It is to be noted that the weather conditions encountered in Northern Europe which were entirely normal and to be expected had contributed to the process of condensation (as was the case with the weather conditions in Noten BV v Harding); see for example [3] and [4] of the judgment. Had the judge concluded that the carrier had a sustainable defence of inherent vice and analysed the burden of proof correctly, he should have gone on to consider whether the exception did not apply because the carrier had not employed a sound system in the carriage of the goods and that had it done so, the damage would not have occurred, an issue on which the burden of proof was on the cargo claimants.
The Court of Appeal went on to conclude that for a number of reasons which again need not presently concern us, the judge had misapplied the law as to what was required of a carrier by way of a sound system for the carriage, in particular in concluding that it required some sort of guarantee that damage would not occur. Had he applied the law and the burden of proof correctly, he would have concluded that the claimants had failed to establish that the carrier’s system of lining the containers with kraft paper was not a sound system so the defence of inherent vice succeeded.
The Court of Appeal then considered the carrier’s alternative defence of inevitability of damage. The basis for this was evidence which the judge had essentially ignored that minor condensation damage to cargoes of coffee beans in bags carried in unventilated containers is inevitable, whatever lining was used pursuant to the general practice of the container trade and the level of damage to these consignments was within that margin of inevitability. The judge had essentially rejected this defence because he sought to impose on the carrier an obligation to employ a sound system which went beyond the general practice of the trade. It is well-established that where a certain level of damage is inevitable even if the carrier employs a sound system and in other respects complies with its obligation under Article III rule 2, the carrier is under no liability for such damage.
Inevitability of loss and inherent vice in a non-marine context
It seems to me that it would be interesting to conclude by looking at how concepts such as inevitability of loss and inherent vice have been worked out in non-marine insurance by reference to the recent decision of Coulson J in Leeds Beckett University v Travelers Insurance [2017] Lloyd’s Rep IR 417. That case concerned the design and construction by the claimant university of a series of buildings on a former brewery site. The relevant building was built on a slope with an undercroft which was five inches high on the western side but five feet high on the eastern, canal side. The eastern wall was a concrete beam laid across pile caps with supporting concrete blockwork on top of the beam. The building was completed in 1996. In August 2011 the university renewed its property insurance with the defendant insurers. The cover was against Damage defined as “accidental loss or destruction of or damage”. So far as relevant to the present discussion there was an exclusion for “Damage caused by or consisting of inherent vice, latent defect, gradual deterioration, wear and tear, frost, change in water table level, its own faulty or defective design or materials but this shall not exclude subsequent Damage which itself results from a cause not otherwise excluded.”
In December 2011, some four months after the policy incepted, large cracks appeared on some of the internal walls and ceilings of the building on the eastern side. The building was evacuated. Inspections over the next few months revealed that an area of the supporting concrete blockwork below ground level had turned to mush. The building was subsequently demolished. It was common ground that the problem was due to the effects of flowing water. Standing water was present to a significant depth behind the external wall. The judge found that there were existing springs and watercourses on the site which had always been there historically, seven natural springs in the vicinity of the building. Drainage recommendations from a hydrological survey had not been implemented and there was no overall design for groundwater drainage. On the east side of the building the only drainage for surface water was a form of soakaway which did not channel the water elsewhere but left it to soak into the ground which frequently became saturated. As a result, spring water flowed, as it always had, from west to east and against the concrete blockwork supporting the inner wall on the east side of the building. That flowing water caused both leaching and sulphate attack which eventually reduced the blocks to mush.
The judge dismissed the university’s claim under the insurance for a multitude of reasons, only some of which are presently relevant. He considered first whether there had been accidental damage which he considered meant simply an event which occurred by chance and was non-deliberate. He drew the distinction between the risk of something happening which would be covered and the inevitability of something happening, which would not, citing Lord Sumner and Lord Birkenhead LC in Gaunt. He held that inevitability will be assessed prospectively from the time the cover was taken out, citing Donaldson LJ in Soya v White and that accidental damage to property does not mean damage due to the inherent characteristics of the property, citing Lord Mance in The Cendor MOPU. Applying those principles, which it can be seen were largely derived from the authorities on marine insurance which I have been discussing, he went on to consider whether on the facts there had been accidental damage. He concluded that there was no intervening accident or fortuity such as flood and that, at the time the policy was taken out in August 2011, the damage which occurred was inevitable, specifically it was inevitable that the concrete blockwork would fail. He concluded flowing water had run up against the blockwork from the outset, progressively weakening the structural strength of the blocks. Failure was inevitable from the outset and all that was unknown was precisely when the failure would occur. Furthermore, regardless of how quick or slow the deterioration was historically, by the time the policy was taken out in August 2011, the collapse of the blockwork was inevitable. He also held that it was sufficient if such collapse was inevitable at some point during the duration of the policy and that it would be absurd to suggest that the insurers had to show that at the time the policy was taken out, the relevant damage was inevitable on a particular day in December 2011. In all the circumstances, the loss was inevitable and was not due to accidental damage.
Given that the cover was against accidental damage and the policy did not contain any extension to cover which might be said to have insured against what Donaldson LJ would have called “unknown certainty” (such as the old “lost or not lost” clause or, arguably, the cover under the Inchmaree Clause), this conclusion that the damage was inevitable and not accidental seems to be an obvious one. The judge also considered that the loss was excluded because the primary cause of the damage to the blockwork (without which the cracking and structural failure would not have occurred) was faulty design, specifically the failure to address the issue of groundwater drainage.
More relevantly for the purposes of the present debate, Coulson J went on to consider the alternative defence that the exclusion for gradual deterioration applied. The assured argued that the exclusion was only for deterioration of the thing itself without any influence from an external source, whereas the insurers contended that the clause did not make that distinction and that deterioration inevitably involved interaction between the property insured and its environment. The judge preferred the insurers’ approach which he considered better reflected Lord Sumner’s analysis in Gaunt that what matters is the natural behaviour of the subject-matter insured “being what it is, in the circumstances under which it is carried” which he equated in that case with the circumstances in which the building existed, the grounds on which it stood and the water flowing against it. He did not consider The Cendor MOPU relevant as that case was about causation, whether the loss was caused by inherent vice or a peril of the sea.
It seems to me that The Cencor MOPU is not just about causation but limits the scope of the defence of inherent vice for the reasons already discussed. It may be that “gradual deterioration” was a broader concept in the context of the policy he was considering, since there was a separate exclusion for inherent vice, but in any event, his overall reasoning seems correct. The environmental conditions which were causative of the overall loss were not themselves an insured peril or in any sense accidental or fortuitous. As the judge found, they had always been there, so that on the reasoning in The Cendor MOPU, there would have been no reason to disapply the exclusion. His approach also accorded with that of the US authorities on exclusions for deterioration such as Murray v State Farm Fire & Casualty Co (1990) 268 Cal Rptr 33, to which Coulson J referred. That case concerned a leak in a copper water pipe caused by electrolysis as the pipe had been exposed to a combination of moisture and acidic soil, which had dramatically speeded up the process of what would otherwise have been natural deterioration over a number of years. The exclusion was still held to apply.
Finally, the judge dealt with the proviso to the exclusion clause which you will recall provided: “but this shall not exclude subsequent Damage which itself results from a cause not otherwise excluded.” In that case, the university sought to argue that the original damage was the damage to the blockwork but the “subsequent damage” was the cracking and other damage to the structure of the building which resulted from the water flowing in and through the blockwork, which was a cause which was “not otherwise excluded”. The judge rejected this argument, holding (following US authorities rather the decision of the majority of the Queensland Court of Appeal in Prime Infrastructure (DBCT) Management P/L v Vero Insurance Ltd [2005] QCA 369) that the subsequent damage must not only be different damage capable of being distinguished from the original damage, but the different damage has to be caused by something “not otherwise excluded”, a new or different cause from gradual deterioration or defective design, which are excluded. He did not consider there was subsequent damage as the damage was all of one piece and the cause of that damage was excluded.
Coulson J gave at [277] an example of the sort of situation the proviso would cover:
“…the collapse of a factory wall because of a faulty/defective design. The falling masonry breaks open a gas pipe, which causes a fire that destroys some adjacent houses. Whilst a claim for the cost of repairing the factory would be excluded (because of the faulty/defective design), the claim for repairing the buildings damaged by the fire would be a claim in respect of subsequent damage caused by something (a fire) not otherwise excluded, and would be recoverable under the policy.”
So far as I am aware, Leeds Beckett University is the first English case to consider the effect of this proviso, despite the fact that the proviso or provisions in similar language are often found at the end of lists of exclusions in all risks property insurance. What Coulson J said was of course obiter since his primary position was that the point was not open to be argued by the university because it had not been pleaded. Nonetheless, it is welcome clarification as to the meaning of the proviso and, in my view, clearly correct. On the plain wording of the proviso, to be covered the relevant damage has to be “subsequent damage” resulting from a cause which is not otherwise excluded. The cracking and structural damage to the building in the present case was not subsequent damage by continuing and worsening manifestation of the original damage to the concrete blockwork. In a very real sense, once the building was constructed over the existing flowing water without any aspect of the design or construction draining that water elsewhere, the building could be said to be doomed from the start. Little wonder then that the claim for accidental damage failed and the exclusions applied.
[1] In the B Atlantic the owners were given permission to appeal by the Supreme Court in February 2017. The appeal is due to be heard in March 2018.
[2] The reservation made by Lord Mance in that paragraph would not seem to be of any relevance to the application of that principle in The B Atlantic.
[3] The definition in para 7 of the Rules for Construction of Policy in Schedule 1 to the Marine Insurance Act provides: “The term ‘perils of the seas’ refers only to fortuitous accidents or casualties of the seas. It does not include the ordinary action of the wind or waves”.
[4] In May 2017 the Supreme Court granted the cargo claimants permission to appeal against the decision of the Court of Appeal. The appeal is due to be heard in the autumn of 2018.