Causation Lecture
It is not sufficient for an individual to simply have a duty of care and then breach the standard of conduct expected of them - that conduct must also cause the damage that the claimant has suffered. Whilst this might seem simple, many accidents can be tied to multiple causes, and certain situations involve claimants contributing to their own injuries. Take, for instance, a car accident; whilst reckless driving might be one of the causes, perhaps the claimant driver was speeding at the time the accident occurred, or maybe a seatbelt in the car had been incorrectly fitted, worsening the situation? Perhaps the accident shunted the driver into a railing which was damaged by a freak lightning strike earlier in the day - an act of God. To muddy the waters even further, there exist issues of evidence - it is not enough to simply argue in court that a harm might have been caused by the defendant, this must be shown to be true on the balance of probabilities.
Factual Causation
Much like the criminal law, tort law uses a 'but for' test in order to establish a factual link between the conduct of the defendant and the injuries of the claimant. In other words, the question asked is 'would the same harm have occurred if the defendant had complied with the duty?' If the answer to this question is yes, the breach is not a necessary cause. There are exceptions, discussed below. The 'but for' test constitutes the generally applicable rule when it comes to causation.
A relatively modern description of the test can be seen in Cork v Kirby MacLean Ltd.
Case in Focus: Cork v Kirby MacLean Ltd [1952] 2 All ER 402
The deceased was painting the inside roof of a factory. He was working on a narrow platform 23 feet above the ground, with no guardrails or toe-boards. He had epilepsy, and was aware of this fact; his employers were not. Whilst working, he had an epileptic fit and was killed when he fell from his platform.
When the case came to court, the defendants argued that the cause of death was the painter's epilepsy, for which they could not be held responsible. Conversely, his widow, as administratrix of his estate, argued that the cause was a lack of appropriate guardrails on the platform.
In formulating the but for test, Denning LJ said the following:
"If you can say that the damage would not have happened but for a particular fault, then that fault is in fact a cause of the damage; but if you can say that the damage would have happened just the same, fault or no fault, then the fault is not a cause of the damage."
- Denning LJ, at 407.
And thus, because guardrails would have likely prevented the fall, it was ruled that there was a causal link between the lack of guardrails and the painter’s death.
In contrast, Barnett v Chelsea and Kensington Hospital Management Committee showcases a failed but for test.
Case in Focus: Barnett v Chelsea and Kensington Hospital Management Committee [1969] 1 QB 428
The deceased attended a hospital emergency department after drinking tea at work which had made him and two colleagues violently ill. He was seen by a nurse, who spoke to a doctor, who told her to send the men home to call their own doctors. Five hours later he died from arsenic poisoning. However, it was established that even with proper care he would probably have died from the poisoning before an antidote could have taken effect. His widow’s claim failed.
Thus, applying the but for test, the courts established that there was no factual causation - the death would have occurred regardless of the defendant's conduct.
The All or Nothing Approach and the Burden of Proof
As with many elements of negligence, matters of causation are decided on the balance of probabilities (i.e. more likely than not). Since the burden of proof rests with the claimant, the onus is on him or her to prove that had the defendant not acted negligently, their harm would likely not have occurred. Thus, if a court finds that there is a 55% chance that a defendant caused an indivisible injury, the claimant does not receive just 55% of the damages on that account. A separate reduction for contributory negligence may still apply.
However, this arrangement can be seen as troublesome from a theoretical standpoint. Take Cork v Kirby MacLean Ltd, facts above; in finding against the defendants the courts are essentially stating that they are responsible for the painter's death, when in fact this might not have been the case. Whilst we know that the painter fell from the platform, and that guardrails might have prevented the fall, this is by no means guaranteed. It is feasible, for example, that he might have fallen from the platform anyway, meaning that the defendant's negligence would have had no effect on his death. Nonetheless, causation was established. The damages were halved because the painter had not told his employers of his fits despite having been advised against working at height.
This approach can be considered problematic because the courts are treating a disputed causal fact as established once it is more likely than not, whilst in fact it is entirely possible that those injuries would have occurred regardless. The all or nothing approach can be seen as particularly problematic when dealing with 'lost chance' cases - these are described in detail in a later section of this chapter.
Specific Rules of Factual Causation
As will often occur when examining real-life cases of negligence, there will often be scenarios in which there are multiple causes of the claimant's harm. In such scenarios, you should first work out if you're dealing with concurrent causes (causes which happen at the same time) or successive causes (causes which take place one after the other). Ask also whether the causes operated together or are competing explanations for the same harm.
Concurrent cause situations are often described as 'two hunter problems', referring to a hypothetical situation in which two hunters in the woods both negligently shoot the same victim at the same time, killing him. Suppose each bullet is independently sufficient to kill. Each shot can be a cause of death even though death would have occurred without either shot taken separately (Financial Conduct Authority v Arch Insurance (UK) Ltd [2021] UKSC 1, paras 182–184). This differs from a case in which only one bullet hit and it is unknown who fired it. Where both shots were independently sufficient, it is not satisfactory to have each hunter renounce liability on the basis that had they not been negligent, the other bullet would have killed the victim, and so they could not be described as causing the victim's death.
Successive cause situations are a little different - here, one bullet hits the victim briefly before the other. This raises its own set of problems - primarily, the question of whether the second shooter can escape liability just because he was fortuitous enough to have his victim hurt before his negligent act did some harm.
Each of these situations has its own rules which apply, as follows.
Multiple Concurrent Causes - The General Rule
Where several factors act together to cause an injury, the claimant does not have to show that the defendant's actions were the sole cause of the injury suffered. Instead, it must be shown on the balance of probabilities that the defendant's actions materially contributed to the harm.
Case in Focus: Bonnington Castings Ltd v Wardlaw [1956] AC 613
The claimant contracted pneumoconiosis (a lung disease) from inhaling minute silica particles during his work in a steel production facility. Dust came from swing grinders and pneumatic hammers (both pieces of equipment used in shaping steel).
The defendants, who managed the factory, had a statutory duty to keep the swing grinders’ dust extraction equipment effective. Its pipes were choked. There was no breach in respect of the dust from the pneumatic hammers (this distinction is often referred to as the difference between ‘guilty dust’ and ‘innocent dust’). The defendant argued that the burden of proof rested with the claimant. The House of Lords agreed, but found that dust from the grinders had made a material contribution to the disease. Claimants do not have to show that negligence was the sole cause of their injuries. The claim therefore succeeded.
In Bonnington the dust from both sources accumulated. The court did not assign each source a 50% probability. Where different agents are possible alternative causes of one injury, it is not enough to identify the defendant’s breach as one possibility. See Wilsher v Essex Area Health Authority.
Case in Focus: Wilsher v Essex Area Health Authority [1988] AC 1074
The claimant was a prematurely born infant who required extra oxygen in order to survive. A junior doctor inserted a catheter into a vein, rather than an artery, and this resulted in the claimant receiving too much oxygen, which was one possible cause of the retinal disease that impaired his sight. However, the claimant's medical situation was such that there were five different possible causes of the blindness, and it was impossible to ascertain which of the five had actually caused the harm. As such, the courts were left considering a case in which only one out of five potential causes involved a breach of duty. The number of possible causes did not establish that each had a 20% likelihood.
The House of Lords allowed the defendants’ appeal and ordered a retrial on causation. It was for the claimant to prove that the excess oxygen had caused or materially contributed to the disease. The McGhee material increase in risk principle (discussed below) could not be applied merely because the negligent exposure was one of several qualitatively different possible causes.
In summary, a material contribution to injury is one which is more than negligible. An actual 20% contribution is material. A 20% chance that the breach was the cause is a different proposition.
Exam Consideration: Ask whether the breach actually contributed to the injury, and whether the harm is divisible. Where the injury is indivisible, a proven material contribution can support liability for the whole injury (Holmes v Poeton Holdings Ltd [2023] EWCA Civ 1377). Where the severity of the disease is divisible, damages reflect the part caused by the defendant, so far as it can be assessed (Holtby v Brigham & Cowan (Hull) Ltd [2000] EWCA Civ 111).
Multiple Concurrent Causation - Exposure to Risk
Finally, there exists a line of cases where claimants have been unable to show that their harm has occurred as a result of the defendant's conduct, but have been able to show that their employer has contributed materially to the risk of an injury occurring. This was the case in McGhee v National Coal Board [1973] 1 WLR 1 - the claimant was employed to clean brick kilns out after use. This involved exposure to brick dust, and the defendants did not provide showers in the workplace. This meant that the claimant had to cycle home whilst covered in brick dust. After some time working in the brick factory, the claimant developed dermatitis (a skin condition) as a result of exposure to brick dust.
Whilst the defendant was not liable for the exposure to brick dust during the claimant's usual course of work, the courts held that in failing to provide showers, the defendants had materially increased the risk of the claimant developing dermatitis - he was exposed to the brick dust for far longer than necessary. The claim was therefore successful.
This principle has become important where cases involve multiple illegitimate exposures to a risk. Its application can be seen in Fairchild v Glenhaven.
Case in Focus: Fairchild v Glenhaven Funeral Services [2002] UKHL 22
The three cases concerned workers who had contracted mesothelioma, a cancer which usually affects the lining of the lung. It can take decades to develop after exposure to asbestos. This means that an individual may work with asbestos for different employers over a long career. The medical evidence could not identify which exposure had initiated the disease or establish that it had been caused by a single fibre. This made it difficult to work out who the proper defendant was.
This was the situation in Fairchild - each of the claimants was exposed to asbestos during the course of their careers, but by a number of different employers. This created a problem - the claimants were unable to demonstrate which of the employers had actually caused the fatal exposure, just that one of them did. The difficulty was proving which exposure had caused the disease. For example, consider a situation in which an employee has three different employers - if all three exposures were equally likely to be responsible, each employer would be only about 33% likely to have caused the fatal exposure to asbestos, and so for each individual employer they are unlikely to be the one to have caused the harmful exposure.
The House of Lords applied McGhee to deal with this problem, ruling that as long as each of the claimants could show that an employer had materially increased their risk of contracting mesothelioma (by illegitimately exposing them to asbestos), then the inability to identify the particular exposure that caused the illness would not defeat the claim. The division of liability between employers was not decided in Fairchild. This is sometimes referred to as the 'Fairchild exception' to conventional causation principles.
It should be remembered that the Fairchild decision did not occur in a vacuum - tens of thousands of employees were exposed to dangerous levels of asbestos throughout the 20th century in situations similar to those of the claimants in Fairchild, and so the decision effectively meant that workers who contracted mesothelioma in those circumstances could seek compensation.
Where employers are jointly and severally liable, this does not mean that each employee can claim three times the damages - they are entitled to be compensated only once for their injury. Whilst this might appear to punish the singled-out employer more than the others, that employer still had the option of suing the others for their contribution to the exposure, meaning that the cost of compensation could effectively be spread amongst the employers. However, it should be noted this is not a guaranteed scenario, since there exists the potential for a defendant to go insolvent, leaving the other potential defendants to bear the costs of compensation - although see Barker v Corus below for developments on this point.
This same principle applies even where the claimant has exposed themselves to asbestos voluntarily, as a matter of self-employment. In the Court of Appeal decision in Barker v Saint Gobain Pipelines plc [2004] EWCA Civ 545 (later appealed to the House of Lords as Barker v Corus, discussed below), Mr Barker was exposed to asbestos for nearly 9 years whilst under the employment of the defendant. There was also a six-week exposure with another employer and exposure while he was self-employed. The Court of Appeal held that Fairchild still applied, and that the defendant was liable for the claimant's mesothelioma because of the material contribution by the defendant to the claimant's risk. It should however be noted that a 20% reduction in the claim's value was made due to Mr Barker's contributory negligence during self-employment.
The House of Lords then considered this case on appeal as Barker v Corus UK Ltd [2006] UKHL 20. Much like Fairchild, the claimant was an employee who had been exposed to asbestos over the course of his career, and had developed mesothelioma. However, some of the potential defendants had since gone insolvent. The decision before the court was therefore whether, under Fairchild, the surviving employers could be held liable for the shares of the now-insolvent companies.
The House of Lords ruled, on the question left open in Fairchild, that each employer was only liable for a percentage of damages in proportion to their contribution to the claimant's risk. For example, an employer who exposed a claimant to asbestos for 10 years of a 30-year exposure history might be responsible for one third of the damages if the type and intensity of exposure were comparable throughout. In effect, this meant that liability under the exception was proportionate to the risk for which each defendant was responsible. This also meant that the share of damages attributed to insolvent defendants was not payable.
However, this decision was not without controversy. There was significant backlash from various groups representing mesothelioma victims, and s.3 of the Compensation Act 2006 swiftly reversed the Barker v Corus position - but only in relation to mesothelioma. The effect of s.3 is to restore joint and several liability for mesothelioma claims, meaning that a claimant can recover the full amount of their damages from any single responsible person, who may then seek contributions from others. For all other types of harm to which the Fairchild exception might apply, the Barker v Corus proportionate liability approach remains the law. In Heneghan v Manchester Dry Docks Ltd [2016] EWCA Civ 86, asbestos was proved to have caused the lung cancer, but the particular exposure responsible could not be identified. The defendants were liable in proportion to the risk attributable to each.
The scope of the Fairchild exception was further considered by the Supreme Court in Sienkiewicz v Greif (UK) Ltd [2011] UKSC 10. The claimant's mother had died from mesothelioma. She had been exposed to asbestos during her employment with the defendant, but also to low-level environmental asbestos in the general atmosphere. The defendant argued that since there was only one tortious exposure (the defendant's), and the other exposure was merely environmental, the conventional but for test should apply - and on that basis, it could not be shown on the balance of probabilities that the defendant's exposure rather than the environmental exposure had caused the disease.
The Supreme Court rejected this argument and held that the Fairchild exception applies even where there is only one negligent exposure to asbestos, provided it materially increased the risk of contracting mesothelioma. The claimant did not need to show that the tortious exposure had doubled the risk compared to environmental exposure. This was a significant extension of the Fairchild principle, confirming that the exception applies regardless of whether the other source of exposure is tortious or non-tortious.
In summary, Fairchild provides a narrow exception where the claimant cannot identify which exposure to the same harmful agent caused the disease. Barker v Corus replaced joint and several liability with proportionate liability for non-mesothelioma cases. Section 3 of the Compensation Act 2006 restored joint and several liability specifically for mesothelioma. And Sienkiewicz confirmed that the Fairchild exception applies even where there is only one tortious source of exposure. Outside this exception, a material increase in risk alone does not establish causation.
Exam Consideration: Due to the scale of asbestos exposure, multi-claimant cases involving it are a significant feature of everyday tort law. Because of this, there is a high chance you will encounter problem questions regarding asbestos, and so having knowledge of this series of cases is a boon. Be sure you can distinguish between: (a) the Bonnington material contribution to harm principle; (b) the Fairchild material contribution to risk principle; and (c) the Wilsher position where there are qualitatively different possible causes. Also be clear on whether joint and several liability or proportionate liability applies where the Fairchild exception applies.
Where several factors may have played a part, distinguish actual cumulative contributions, as in Bonnington Castings, from alternative possible causes, as in Wilsher. In the latter case the claimant normally has to prove that the defendant’s breach caused the injury on the balance of probabilities.
A narrow exception applies in cases in which the defendant has materially increased the risk of an injury occurring, as in McGhee and Fairchild.
Exam Consideration: If a problem question can be solved just by using the 'but for' test, then do so! If the problem involves multiple concurrent causes, then you will have to refer to the above cases. In such a situation, the most important thing is that you can accurately describe the rule; this will show that you know the law, even if you encounter difficulty applying it to an unwieldy number of contributing defendants.
Multiple Successive Causes
There also exists a small band of cases involving multiple, successive causes of damage - that is, the claimant is harmed by two sources of damage, one after the other. The primary example here is Performance Cars Ltd v Abraham [1962] 1 QB 33. The claimant's Rolls Royce was struck by a negligent driver. Two weeks prior, however, the claimant was in another accident (also in which the other driver was at fault). The second collision damaged a different part of the car, but the lower bodywork already needed respraying. The award for the first collision included £75 for the respray, but the claimant did not receive it. The claimant then sought to recover the same amount for the second collision.
The courts ruled that the second driver had not caused the need for the £75 respray. He remained responsible for the additional damage he caused. Thus, in the two hunter example, if the first shot has already killed the victim, the second cannot have caused the death. If the first only wounds the victim and the second kills him, the second shooter cannot rely on the earlier wound to escape liability. In Baker v Willoughby [1970] AC 467, a later shooting which led to amputation did not relieve the first tortfeasor of liability for the continuing loss caused by the original leg injury. In Jobling v Associated Dairies Ltd [1982] AC 794, by contrast, a later independent disease would itself have prevented the claimant working and limited the earnings recoverable from the employer.
Lost Chance Cases
There is also a category of cases which deal with the idea of losing a chance - for example, a claimant might have a disease with a 40% recovery prognosis if it is caught early enough. However, due to negligent misdiagnosis the disease might go unchecked, and become terminal in nature. This claimant would be annoyed to say the least, and would therefore want compensation for losing their 40% chance at recovery. Such situations are referred to as 'lost chance cases'.
In personal injury cases, a lost chance of a better medical outcome is not generally actionable by itself. The claimant ordinarily has to prove that the breach caused injury on the balance of probabilities. This principle is best understood via Hotson v East Berkshire Area Health Authority [1987] AC 750 - the claimant fractured his hip when he fell from a tree that he was climbing. When taken to hospital, he was misdiagnosed, and a hip deformity later developed. Expert evidence stated that even with a correct diagnosis by the hospital, the claimant had a 75% chance of developing the deformity. The trial judge awarded 25% of the damages for the deformity and the Court of Appeal upheld the award. The House of Lords rejected this reasoning - on the balance of probabilities, the fall had already done irreversible damage to the blood supply. The claimant kept the £150 awarded for the additional pain caused by the delay.
Exam Consideration: Whilst this might appear very similar to situations above, in which there are multiple causes of harm, the relevant case law is different. The two can be distinguished by asking whether the facts of the case involve competing explanations for an injury that has occurred, or whether they involve a loss of a chance of avoiding a harm. This will usually be made clear by reference to medical prognosis or similar.
The rule can also be seen to be applied in Gregg v Scott [2005] 2 AC 176, but regarding chance of survival. The claimant found a lump under his arm, and consulted his doctor. The doctor misdiagnosed the lump as benign, when in actual fact it was cancerous. This caused a nine-month delay in the claimant's treatment, and the figures relied upon for ten-year disease-free survival fell from 42% to 25%. The House of Lords, by a majority, rejected the claim for the reduced chance of a cure. The decision did not depend on a finding that the claimant had died.
It can therefore be seen that the law regarding lost chances is not without criticism. In Hotson, it was clear that the hospital had made a mistake, but the court found that the deformity would probably have developed anyway. He was unable to recover damages for it, even on a percentage basis. The problem with the status quo can be seen when the margins are made smaller - consider a situation in which the claimant had a 49% chance of recovering had the hospital properly diagnosed him; he would still be unable to claim, despite the hospital effectively changing his prognosis from a coin flip to certain deformity.
That being said, there is some merit to the Hotson position for claimants. For example: consider if a claimant had a 51% chance of getting better, but lost this chance due to medical malpractice. Under the Hotson principle, that claimant would be entitled to full damages for the additional injury caused by the negligence, because on balance, they would have got better had malpractice not occurred. If, however, damages were calculated based on the lost chance itself, then this particular claimant would only be entitled to 51% of the total compensation for their harm, since there was a 49% chance the harm would have occurred regardless of malpractice.
Furthermore, the decision in Hotson makes it a lot simpler for the courts to work out the proper amount of compensation to be paid. The status quo means that claimants are either entitled to all of the damages or none (this is a slight oversimplification due to principles like contributory negligence, but the point still stands). If the decision of the Court of Appeal in Hotson were to stand, courts would have to spend significant time and resources working out the probability of certain things happening, down to the single percentage point. In contrast, the current situation means that there is no material difference between losing a 75% chance of recovery and losing a 70% chance of recovery - in either situation the claimant would be entitled to full damages for the additional injury if causation is proved. Different rules apply to some financial losses. A claimant who loses a real and substantial commercial opportunity through negligence must normally prove what they themselves would have done on the balance of probabilities. What a third party would have done can then be valued as a chance, even below 50% (Allied Maples Group Ltd v Simmons & Simmons [1995] 1 WLR 1602; Perry v Raleys Solicitors [2019] UKSC 5).
Exam Consideration: As you might be able to tell, lost chance cases are often subject to significant judicial discussion. They are therefore more likely to turn up in essay questions (although this doesn't rule out them appearing in problem questions!). It is therefore important not only that you know the Hotson rule, but also the arguments for and against it.
Vindication of Rights Cases
Although lost chance cases demonstrate the problem with adhering strictly to the all-or-nothing approach to causation, there does exist a slim category of cases in which the courts have been willing to depart from the general rule due to policy considerations. Whilst examples are sparse, Chester v Afshar provides one.
Case in Focus: Chester v Afshar [2004] 3 WLR 927
The claimant was soon to undergo an operation to cure her back pain. The defendant, her surgeon, negligently failed to warn her that there was a 1-2% risk of her becoming permanently disabled as a result of the operation. She underwent the operation, and despite the operation being carried out correctly, the disabling complication occurred anyway. The claimant could not show that she would have avoided the operation entirely had she known of the risk, but established that she would have delayed the operation by some time.
The failure to warn had not increased the inherent risk of surgery. But if properly warned, the claimant would not have undergone that operation on that day. It was the very risk about which she should have been warned that materialised. Nevertheless, the House of Lords acknowledged the fact that the doctor had done wrong - patients have a right to know what they are consenting to. Lord Hope indicated the rights-based nature of the case:
"To leave the patient who would find the decision difficult without a remedy, as the normal approach to causation would indicate, would render the duty useless in the cases where it may be needed most. [...] The function of the law is to enable rights to be vindicated and to provide remedies when duties have been breached. [...] On policy grounds therefore I would hold that the test of causation is satisfied in this case."
- Lord Hope, at para 87.
They therefore found against the defendant, despite the decision appearing to run contrary to conventional causation. It should be noted that this decision was controversial - it was decided by a 3-2 majority, with Lords Bingham and Hoffmann delivering strong dissents. The exception is narrow: the claimant must establish that a proper warning would have led them to decline or defer the operation, and that the undisclosed risk materialised (Duce v Worcestershire Acute Hospitals NHS Trust [2018] EWCA Civ 1307).
Exam Consideration: McGhee and Chester are difficult to reconcile. Whilst both involve situations in which conventional causation cannot assist the claimant, it is only in the latter case that the court saw fit to vindicate the claimant's rights on pure policy grounds. McGhee involved a brick kiln worker whose employer failed to provide showers, whilst Chester involved a surgeon who failed to warn a patient of risks. This can be considered a symptom of incrementalism - rather than sticking to one set rule, the courts can be seen to take each situation as it comes. When considering novel legal situations, ask first whether the ordinary test is satisfied and then whether the facts fall within a recognised exception such as Chester.
Law Teacher writing support
Writing a law essay?
We choose a writer who is an expert in your area of law and qualified to at least the academic level you select. Most of our writers hold a master’s degree or PhD, have teaching experience, or both.
Send us your question and, if you wish, any marking criteria. We’ll write an essay to your instructions.
Cite This Module
To export a reference to this article please select a referencing style below: