Would Sherlock Holmes have been a reliable claims handler? A recent decision by the Court of Appeal suggests not – and they did not think much of Hamlet either.
The Court of Appeal has had occasion twice this year to consider the test required to discharge the burden of proof for civil liability. Nulty v Milton Keynes Borough Council  Lloyd’s Rep IR 243 involved a dispute as to the cause of a fire at a recycling plant. Ace European Group v Chartis Insurance UK  EWCA Civ 224 concerned the cause of serious cracks in components for an ‘energy from waste facility’ that was under construction. Both appeals were dismissed because it was held that the judges at first instance had reached conclusions that were available to them on the facts. However, in Nulty the Court of Appeal held that Mr Justice Edwards-Stuart had erred in holding that if the only other possible causes of a loss were very much less likely, by process of elimination the remaining hypothesis became in law the probable cause, however unlikely it might otherwise appear. Both decisions applied the House of Lords decision in the Popi M  2 Lloyd’s Rep 1 that a finding on the balance of probabilities that a particular event was causative of the loss required the case for believing that the event occurred to be more compelling than the case for not so believing. It was not sufficient that the cause put forward was merely the most plausible of a number of improbable explanations.
In Nulty the appellant, the estate of a deceased self-employed electrical engineer (and his liability insurers), appealed against the High Court’s decision that a serious fire at a recycling centre owned by the Council had been caused by a discarded cigarette dropped by Mr Nulty when he was alone on the site investigating the cause of a power outage. In his decision at first instance, Edwards-Stuart J had painstakingly considered the possible causes of the fire, which the Council contended was a discarded cigarette dropped by Mr Nulty and which Mr Nulty’s estate argued was ‘arcing’ from a disused electric cable that had been left live and in a dangerous condition. On the expert evidence the judge concluded that the latter theory was most improbable. By contrast, he found that there was nothing physically or scientifically implausible about the cigarette end explanation, however, he accepted the objection to this theory that the engineer would not have been expected to behave in such a way given his experience and the fact that he had formerly worked as a fireman. The judge held that neither cause, viewed on its own, seemed likely but that the arcing explanation was very much the less likely of the two and therefore that the Council had discharged the burden of proof.
The Court of Appeal held that Edwards-Stuart J had erred in law in restating a Sherlock Holmesian logic that “when you have eliminated the impossible, whatever remains, however improbable, must be the truth”. This process of deduction is not a satisfactory basis for the court to find that the occurrence of an event has been proven on the balance of probabilities, bearing in mind that the evidence available may be incomplete and every possible explanation may not be known. The law requires a judge, before he or she finds that a particular event occurred, to step back from weighing up the potential causes and to be satisfied on the evidence that the last possible cause in the court’s contemplation is more likely than not to have occurred.
This reasoning follows the House of Lords decision in the Popi M, which illustrates the point nicely. That case concerned a ship that had sunk in calm seas and without an obvious explanation. At first instance Mr Justice Bingham concluded that the ship was sunk by a submarine, having discounted the other possible explanations, even though there was little evidence to support this conclusion. Bingham J’s decision was overturned on appeal.
The Court of Appeal in Nulty did not overturn the lower court’s decision because it said that the judge had found that a discarded cigarette butt was more likely than not to have caused the fire, albeit he went further than he needed to in explaining his decision.
The question in Ace was whether damage to economisers intended for a waste facility had been caused by vibration during transportation by road (against which the Chartis marine policy provided coverage) or while they were sat on-site and exposed to the elements (in which case the Ace Erection All Risks policy would respond). The decision turned on the expert evidence and the appeal was against the judge’s findings of fact that led to his conclusion that the damage had occurred during transit. It was not disputed that the judge had correctly applied the Popi M and that he had not merely dismissed the theory that the cracks occurred on-site but that he had also satisfied himself that it was more likely than not that the competing theory was the correct one.
The appellants argued that the findings were not open to the judge, relying heavily on digital photographs they had produced for the first time during the original trial. The Court of Appeal rejected this argument and, as well as disagreeing with the appellants about the significance of the photographs, firmly criticised them for adducing new evidence at trial, after the claimant’s expert had given evidence and without providing complete metadata, making the photographs unreliable. In a further example of the need to beware the lessons of literature, Lord Justice Moses said that the judge should have been invited to exclude the evidence altogether, ignoring the cautionary tale from Hamlet that too much protest “may prompt a Gertrude-like judicial response to the protestations of the Player Queen”.
These cases provide a reminder to claims handlers that when the cause of a loss is open to dispute, in deciding whether the policy provides coverage it is necessary to ask which (if any) of a range of possible explanations is more plausible than not (and not which is the least implausible). Ace is also noteworthy for the court’s strongly-worded criticism of the production of late and incomplete evidence.