This is expatiated in most English tort law textbooks.
The standard (‘more likely than not’, or ‘doubling the risk’) approach to resolving evidential difficulties can be productive of injustice. Consider the Two Hunters Problem, which is based on a real case, the Canadian case of Cook v Lewis (1951):
Shooter One and Shooter Two are both out hunting with shot guns. They simultaneously hear
some rustling in the bushes and simultaneously fire in the direction of the noise. Beater was
hit in the leg by one of the shots, but forensic analysis is unable to tell from whose gun the
shot was fired.
In this case, both One and Two have clearly been negligent in relation to Beater. They
each owed him a duty to take care not to fire their guns in the way they did – as it was
reasonably foreseeable that doing so would result in physical injury to someone like Beater –
and they each breached that duty of care. But if we apply the ‘more likely than not’ rule to
Beater’s case, he will not be able to sue either One or Two for compensation for his wound.
The reason is that he will only be able to show that it is 50 per cent likely that One shot
him, and 50 per cent likely that Two shot him. He will not be able to reach the magic
figure of 51 per cent in relation to either of them. In Cook v Lewis itself, the Supreme
Court of Canada thought that such a result would be very unfair, and ended up finding
that, in the case we are considering, both One and Two would be liable in full to compensate
C for his injury.36 In the case of Fairchild v Glenhaven Funeral Services Ltd (2003),
the House of Lords indicated that it agreed with this solution.37
36 To the same effect, see the American case of Summers v Tice, 199 P 2d 1 (1948).
37 [2003] 1 AC 32, at [39] (per Lord Nicholls), [169] (per Lord Rodger).
Roderick Bagshaw, Tort Law, page 271.
(B) CASES INVOLVING MULTIPLE DEFENDANTS
In Barnett and Bolitho, the defendant’s fault was competing with one or more ‘innocent’
explanations of the claimant’s injury. More problematic are cases in which there is no ‘innocent’
explanation for the injury and where, therefore, the claimant’s injury could only have
occurred through the fault of one or more of a number of defendants. In these cases, the
strict application of the but-for test has been modified to avoid a result which would undercompensate
the claimant.
The first situation to consider is that in which the claimant’s injury could only have
been caused by the fault of one out of several careless defendants, in circumstances where
the individual defendant in question cannot be identified: this may be called the case of
‘indeterminate cause’. The best illustration is provided by two similar cases decided in the
United States and Canada respectively, Summers v. Tice37 and Cook v. Lewis.38 In each case
the plaintiff was shot by a single bullet which was fired by one of two defendants out hunting,
each of whom had been careless in aiming his gun in the plaintiff’s direction. There
was no means of telling from whose gun the shot has been fired (although things might be
different in such a situation today, given advances in forensic science and technology). The
courts adopted the solution of reversing the burden of proof, so that each defendant had to
show that he did not cause the injury. In the absence of such proof, both defendants were
held liable.
A variant of this situation arises where the claimant is struck by two bullets, fired by
each of the two defendants. If his injuries would have resulted from one bullet only, each
defendant is entitled to say that ‘but for’ his own personal carelessness, the claimant would
still have sustained his loss. This result would leave the claimant ‘falling between two stools’,
and it is therefore likely that were this case to come before an English court it would find the
defendants jointly liable, perhaps varying the damages payable by each according to their
degree of responsibility. This is analogous to what happened in Fitzgerald v. Lane.39 The
plaintiff walked onto a pelican crossing without looking and was struck by a car driven by
the first defendant, thrown into the air, and struck again by a car being driven in the opposite
direction by the second defendant. Each of the two drivers was driving too quickly and
was held to have been in breach of the duty of care they owed to the plaintiff. The plaintiff
sustained tetraplegia as a result of the accident, but it could not be established that the second
collision contributed to his injury. The second defendant claimed that the first impact
alone would have been sufficient, on the balance of probabilities. The Court of Appeal held both of them liable and made an award of damages based on an assessment that they were
equally to blame with the plaintiff for the accident. The basis for finding them both liable
was the principle enunciated by Lord Wilberforce in McGhee v. National Coal Board,40 and
to which we shall return shortly,41 namely that each one had materially increased the risk of
such injury by his negligence and should, as a consequence, have the burden of disproving
a causal link.
37 119 P 2d 1 (1948). 38 [1951] SCR 830.
39 [1987] QB 781; affirmed on different grounds [1989] AC 328. See the similar US case of Maddux v. Donaldson, 108 NW 2d 33 (1961).
40 [1973] 1 WLR 1. See further sec. 2(C), below. 41 See below, sec. 2(C).
Markesinis and Deakin's Tort Law (2019), pp. 207-8.