Proximate Cause: Don’t Mention The War Exclusion
In April and May 1942, Exeter suffered successive nights of Luftwaffe bombing. Exeter wasn’t a strategic target. Instead German High Command had in mind revenge for the RAF’s bombing of Germany. The targets of the reprisal bombing were culturally important sites: hence the choice of Exeter, Bath, Canterbury, Norwich and York. German radio declared after the raids: “Exeter is the jewel of the west; we have destroyed that jewel, and we will return to finish the job”.
That historic threat to ‘finish the job’ did not materialise (in Exeter at least) until 27 February 2021. That would appear to be the interpretation of HHJ Bird in the recent case of Allianz Insurance plc -v- University of Exeter.
In 2021 a bomb dropped almost 80 years earlier was discovered during construction works in Exeter, next to the University campus. A military bomb disposal unit was called in. They determined its fuse had degraded, but moving it to a safe place for a controlled explosion was not safe because of its age, rust and the possibility that it was booby trapped. “The only realistic course open to the team” was to detonate the bomb on site, which they did on 27 February 2021. This caused damage to University buildings. The University claimed in respect of that damage under a policy they held with Allianz.
Allianz, perhaps surprisingly given the amount of time that had passed since the bomb was dropped (as well as the ending of hostilities between the UK and Germany in May 1945), relied upon the following exclusion for damage “occasioned by war”:
War…Loss, destruction, damage, death, injury, disablement or liability or any consequential loss occasioned by war, invasion, acts of foreign enemy, hostilities (whether war be declared or not), civil war, rebellion, revolution, insurrection or military or usurped power.
Issue Before the Court
The question for the court was one of causation. In insurance cases causation normally means determining the “proximate cause” of the loss. On Allianz’s case, the proximate cause was the dropping of the bomb by Nazi Germany in 1942. This, Allianz argued, was an act of war that meant the damage to the University’s buildings was excluded because it was “occasioned by war”.
Allianz’s secondary case was that the dropping of the bomb in 1942 and the detonation of it in 2021 were both proximate causes of the University’s loss. Established case law tells us that in such circumstances the fact that one of the causes is excluded from cover is sufficient to defeat a claim (often referred to as the Wayne Tank principle, arising as it does from a case of the same name).
The University, on the other hand, argued that the proximate cause of its loss was not the dropping of the bomb in 1942, but its detonation by the bomb disposal team in 2021. If that were so, the exclusion in Allianz’s policy would not apply.
The Court found in favour of Allianz and held that the dropping of the bomb was the proximate cause of the University’s loss such that Allianz’s declinature was sound. In doing so the judge adopted two alternative approaches, arriving at the same outcome in both.
The First Approach: Ignoring Human Intervention in 2021
There were only two possible candidates for proximate cause: the dropping of the bomb and the detonation of it by the bomb disposal team. The Courts have held (FCA v Arch & Others) that one can ignore human intervention in determining causation if that intervention was reasonable, as it was held to be here (the Court accepted the proposition that it wasn’t safe to move the bomb for safe detonation elsewhere so it had to be neutralised in situ).
In the Court’s view, ignoring the act of the bomb disposal team in detonating the bomb meant, by simple process of elimination, that the dropping of the bomb was the inevitable answer to what caused the damage.
The Second Approach: Not Ignoring Human Intervention in 2021
However, the judge said that ignoring human intervention wasn’t a firm and fixed rule and also sought to apply a “common sense analysis” taking account of the acts of the bomb disposal team.
The judge put it as follows:
The loss was caused by an explosion. The explosion was triggered by the reasonable (and indeed obviously correct) decision to detonate the bomb. That decision was necessitated by the presence of the bomb. If there had been no bomb, there would have been no explosion. The bomb provided both the explosive payload and the absolute need for detonation…the dropping of the bomb was the obvious proximate cause of the damage.
Concern for policyholders?
There are a number of features of this case that may be of concern to policyholders.
Approach to Causation
The first issue is the Court’s approach to causation where it seems to us a number of other analyses exist.
The Allianz policy did not contain a ‘bomb exclusion’, a ‘derelict munitions exclusion’ or a ‘loss caused by explosion exclusion’. It contained an exclusion for loss ‘occasioned by war’. That being so, it seems to us credibly arguable that the proximate cause of the University’s loss should be the relevant act of ‘war’ for the exclusion to apply, and not the simple presence of the bomb owing to its failure to detonate as had been intended. That, of course, gives rise to the question “What was the ‘act of war’?”.
As the University argued before the judge, the act of ‘war’ was the dropping of a fused bomb primed to go off and cause damage of its own volition (arguably, to whatever infrastructure was present at the time the bomb was dropped, which did not include the buildings damaged in 2021). But that act of ‘war’ was not the cause of the University’s damage in 2021. Plenty of bombs were dropped in 1942 that did not require a controlled detonation 80 years later. It follows that the dropping of the bomb (i.e. the act of ‘war’) was not the proximate cause of the University’s damage. Instead, it was the bomb’s failure to detonate (and not the act of ‘war’ which had intended the opposite) which necessitated the 2021 detonation decision which in turn caused the University’s loss. If that is right, then the act of ‘war’ in dropping the bomb did not cause the damage. Instead, it was the bomb’s failure to detonate, lying in situ, the necessity for a controlled explosion and the manner of that explosion which caused the 2021 damage.
So far as causation is concerned, the following question then arises: would the act of ‘war’ envisaged by the Allianz policy’s exclusion (as opposed to the simple presence of the degraded bomb) have manifested into a loss without human intervention such that it could be said to be proximately causative of the University’s loss?
The facts suggest that the bomb did damage in 2021 precisely because it did not detonate in 1942 or do any relevant damage at that time. One of the agreed facts between the parties was that the bomb’s fuse was in a poor state. It had not detonated of its own volition for 80 years or so. That being so, it seems to us that an alternative analysis might be that it was the presence of a derelict (but still dangerous) unexploded munition, and hence the necessity for the decision to detonate it, which caused the policyholder’s loss, not the act of ‘war’ in dropping a fused bomb. Put another way, the University’s loss was actually occasioned by the failure of the bomb to explode on that spring night back in 1942 (whether that be on account of a faulty electrical connection, a faulty fuse or even, speculatively, some act of sabotage by the forced labour in German munitions factories), not the act of ‘war’ which, ultimately, failed and caused no damage at all.
Whilst we cannot know for sure absent the involvement of experts in the case, if it is right that the only means by which the bomb could have exploded following its failure to detonate and degradation over the next 80 years was its detonation by the military, then it seems to us credibly arguable that it wasn’t the dropping of the bomb (i.e. the act of ‘war’) in 1942 that necessitated the 2021 detonation. Rather, it was the 1942 failure of detonation, the presence of unexploded munitions capable of disturbance by contractors and the 2021 detonation decision/process which proximately caused the University’s loss.
“Common sense” and Caselaw
The Court’s two approaches to causation do not sit comfortably with each other, notwithstanding that they led the judge to the same conclusion in favour of Allianz. The first, as we note above, was to ignore human intervention in accordance with the guidance of the Supreme Court in FCA v Arch (i.e. to ignore the reasonable decision to detonate in 2021 in determining the proximate cause of the damage). However, it is difficult to reconcile that with either of the two formulae endorsed by the Supreme Court in Arch: to adopt a “common sense” approach to causation and to see causation as consistent with what “the man in the street” would think. Those watching from behind the cordon as the bomb exploded might well have attached causative agency to the bomb disposal team flicking the detonation switch, and expressed surprise at the suggestion that a pilot some 80 years earlier might be primarily responsible.
Quite apart from the application of those formula in the present case, however, there is another reason to question the application in this case of the Supreme Court guidance about ignoring human intervention in determining causal connection. The two cases which contribute to the establishment of that rule are Reischer v Borwick and Leyland Shipping v Norwich Union.
In Reischer a ship hit an object floating in the river. This caused a leak. It was temporarily repaired and a tug was sent to tow it to the nearest dock. The motion during the towing opened up the leak again and the ship sank. The decision to tow was reasonable and was disregarded in determining the proximate cause because the sinking was the inevitable outcome of the collision which caused the leak.
In Leyland, a ship was torpedoed in World War I and was towed to the nearest port where she was moored safely. The harbour master was worried she would sink and block the port so ordered that she be moved to the outer harbour where she was more exposed to wind and waves. After three days the ship sank. She was covered for perils of the sea (the wind and the waves) but not consequences of hostilities or warlike operations (the torpedo). She would have survived if allowed in port. The court said the human intervention (i.e. the decision to move the ship to the outer harbour) would be ignored in determining causation.
In both cases the reasonable human intervention came after the initial damage had occurred (the collision and the torpedoing). The human interventions in Reischer and Leyland were reactions to damage already caused but in both cases were reasonable ones (which is why their acts were ignored in determining causation). In the present case, the human intervention came before any damage occurred. It is materially different from the Reischer and Leyland cases and, it is arguable, is a case where the Supreme Court’s guidance to ignore human intervention (the first of the judge’s two approaches) is not appropriate. If all the damage follows the human intervention, it seems to us that the analysis can credibly be argued to be different.
The second issue of interest to policyholders is the unusual means by which the matter came to trial in the first place. Allianz (and not, as is more typical, the University as policyholder) commenced the claim. It did so to seek a declaration from the Court that its declinature was sound. What is more unusual, however, is that the claim proceeded to trial ostensibly (but not actually) under the “Shorter Trials Scheme” under the Civil Procedure Rules. The effect of that, as the judge noted, was that “neither party has for example filed any evidence”. That included, it seems, the absence of any expert evidence on the condition of the bomb (and its fuse) and what, left to its own devices, might have happened had the events in question not taken place.
The effect of the unorthodox means by which trial came about, and the absence of evidence on certain factual issues potentially relevant to the question of causation, gives rise to the question that, had the matter proceeded differently, and additional evidence put before the judge (if it was capable of being produced), might a different conclusion have been reached?
Any different procedure might have been at the expense of additional time, of course. There is, therefore, a balancing act between obtaining a fast answer from the Courts where one is required (or via some other ADR process) and ensuring that sufficient evidence is introduced to ensure that the answer obtained is the right one.
If there is no appeal, it may be that policyholders and their insurance brokers seek to revisit their policy wordings, and the wordings of war exclusions, to clarify their application to unexploded munitions. A report in 2018 revealed that around 60 unexploded bombs each year are dealt with by the Ministry of Defence (which, owing to the absence of figures to take into account munitions dealt with privately, is likely to be a material understatement). Whilst incidents such as that faced by Exeter University are comfortingly rare, it is those rarities for which policyholders seek insurance and this judgment puts the efficacy of that cover in doubt.
For that reason, and given the outcome described above will have been surprising to many, it may well be that the appeal fuse is lit by Exeter University in order to address (what policyholder enthusiasts, at least, consider to be) an explosive judgment…
Thomas Sharp, Exeter Phoenix. A Plan for Rebuilding (London, 1946) p18
 Allianz Insurance Plc v University of Exeter  EWHC 630 (TCC) (22 March 2023) (bailii.org)
 Wayne Tank and Pump v Employers Liability Assurance Corp. 
 FCA v Arch Insurance UK Ltd. and Others  UKSC 1