← The hardest concepts in United Kingdom law
In one breath
A claimant must show that the defendant’s breach caused the damage, and the usual test asks whether the harm would have happened anyway without the breach. In a few settled situations, such as asbestos-related disease caused by several employers, the courts relax the test so that the claimant is not left without a remedy.
The picture
The domino run and the second pusher
A line of dominoes falls and knocks over a vase. To ask whether the first push caused the breakage, take it away: if nobody had pushed, would the vase still have fallen? If it would have toppled anyway, because a second person had set off another row, the first push was not a cause in this sense. That is the but-for test, and for most cases it is all that is needed.
The law adds exceptions where the test fails. If a hospital doctor negligently fails to examine a poisoned patient who would have died anyway, the breach is not a cause (Barnett). But where several sources of dust each add to a worker’s lung damage, a breach that made a material contribution to the damage is enough (Bonnington Castings). Where a disease, such as mesothelioma, may have been triggered by any of several employers’ exposures and science cannot say which, the courts have accepted that materially increasing the risk is sufficient (McGhee; Fairchild).
Where the picture stops working
Dominoes are physical and the facts are visible, whereas real causation turns on expert evidence about probabilities, decided on the balance of probabilities. The exceptions are narrow and tied to cumulative harm or to diseases of this kind; they do not treat any added risk as a cause (Wilsher). Bonnington and McGhee came on appeal from Scotland and are authority across the United Kingdom.
How it really works
- 1Ask the but-for question: but for the defendant’s breach, would the claimant have suffered this damage? If the damage would have occurred anyway, causation fails (Barnett).
- 2Prove it on the balance of probabilities. The claimant must show that it is more likely than not that the breach caused the harm.
- 3Where the damage was produced by several sources acting cumulatively, ask whether the breach made a material contribution to it (Bonnington Castings).
- 4Where a disease may have been caused by any of several wrongdoers and science cannot say which, the courts have in some cases accepted that materially increasing the risk is enough (McGhee; Fairchild). This is a narrow exception, not a general relaxation (Wilsher).
- 5Then turn to the questions that follow, such as whether a later independent event broke the chain, and whether the damage is too remote.
Why people get it wrong
- Treating a breach as the cause because the defendant was careless and something bad happened afterwards. The question is whether the harm would have occurred without the breach.
- Reading McGhee as saying a claimant need never prove causation when evidence is uncertain. The exception is limited, and Wilsher refused to treat it as general.
- Mixing causation in fact with remoteness. The but-for test comes first; whether the harm is too unlikely to be attributed to the defendant comes after.
- Expecting certainty. The standard is the balance of probabilities, not proof beyond reasonable doubt.
A worked example
The facts are invented
Pia, a night nurse, fails to call a doctor to Gareth, a patient who has collapsed. He dies. Expert evidence says his condition would have been fatal within an hour whatever was done. Pia’s failure was a breach, but but-for causation fails, because Gareth would have died anyway. She is not liable for his death.
Hugh works for three successive employers, each of which exposes him to the same silica dust without proper protection, and he develops a lung disease that grows with the cumulative dose. Each exposure materially contributed to the damage, so each employer’s breach is treated as a cause. Hugh need not show that one employer’s dust alone would have been enough.
Try it
Three employers each exposed a worker to asbestos. He develops mesothelioma, which may follow a single exposure, and medical science cannot say which one caused it. Under the plain but-for test he would lose against each employer. What have the courts done?
Landmark authorities
Read the sources themselves; this list is a starting point.
- Bonnington Castings Ltd v Wardlaw, [1956] AC 613
- Barnett v Chelsea and Kensington Hospital Management Committee, [1969] 1 QB 428
- McGhee v National Coal Board, [1973] 1 WLR 1
- Wilsher v Essex Area Health Authority, [1988] AC 1074
- Fairchild v Glenhaven Funeral Services Ltd, [2002] UKHL 22, [2003] 1 AC 32
Where to go next
Courses
More hard concepts in United Kingdom law
- Duty of care and the incremental approach (Tort law)
- Remoteness of damage (Tort law)
- Vicarious liability (Tort law)
- The rule in Rylands v Fletcher (Tort law)
This page is an educational overview. It is not legal advice and not a substitute for the sources: the statute, the rule, the case itself. The United Kingdom has three legal systems: England and Wales, Scotland, and Northern Ireland. These pages mostly describe the law of England and Wales and say where Scots law differs when it clearly does; Northern Ireland law is often similar to English law but is not covered separately. Parliament is sovereign, but the devolved legislatures also make law within their powers. Each page states the usual formulation of the rule and names landmark decisions; check the primary sources before relying on any rule.