The tort of negligence

Issue. The issue is whether Tom owed Priya a duty of care in the tort of negligence.

Rule. The modern duty of care developed from the neighbour principle in Donoghue v Stevenson [1932], where Lord Atkin held that a person owes a duty to their 'neighbours' — those so closely and directly affected by their act that they ought reasonably to have them in contemplation. The current test for whether a duty exists is the three-part test in Caparo Industries v Dickman [1990]: (1) the harm must be reasonably foreseeable; (2) there must be a relationship of proximity between the parties; and (3) it must be fair, just and reasonable to impose a duty. The earlier Anns v Merton two-stage test was thought to expand liability too widely and is now qualified.

Application. Applying Caparo to the facts: (1) Foreseeability — it is plainly reasonably foreseeable that careless driving could injure a pedestrian, especially one on a zebra crossing. (2) Proximity — Tom and Priya were on the same road at the same time; the relationship between a driver and other road users is a recognised proximate one. (3) Fair, just and reasonable — there is no policy reason to deny a duty between road users; indeed road users are an established category in which a duty is routinely owed, so the careful Caparo analysis is barely needed beyond confirming it.

Conclusion. All three limbs of Caparo are satisfied. Tom therefore owed Priya a duty of care. (Whether he is ultimately liable would then depend on breach, causation and remoteness, but on the question asked, a duty is clearly owed.)

The standard of care in negligence is the benchmark against which a defendant's conduct is measured to decide whether the duty of care has been breached.

The objective standard of the reasonable person. The basic standard is that of the reasonable person — an ordinary, prudent person doing the activity in question. In Blyth v Birmingham Waterworks [1856] negligence was defined as 'the omission to do something which a reasonable man would do, or doing something which a prudent and reasonable man would not do'. The standard is objective: the defendant's own inexperience, best efforts or good intentions are generally irrelevant — they are judged by what a reasonable person would have done.

Modifications to the standard. The standard is adjusted for the type of defendant. A learner is judged at the standard of the reasonably competent, experienced person doing that task, not allowed any allowance for inexperience: a learner driver was held to the standard of a competent driver in Nettleship v Weston [1971]. A professional or expert is judged by the standard of their profession: a doctor is not negligent if acting in accordance with a responsible body of medical opinion (Bolam v Friern Hospital Management Committee [1957]), provided that opinion withstands logical analysis (Bolitho [1998]). A child is judged against a reasonable child of the same age, not an adult: Mullin v Richards [1998] (two 15-year-olds fencing with rulers) and Orchard v Lee [2009].

Conclusion. The standard of care is therefore an objective one, tailored to the kind of defendant but not to their personal shortcomings — a structure that protects claimants while recognising that learners, professionals and children should be measured against an appropriate benchmark.

Issue. The issue is whether Dr Khan is liable to Sam in negligence, which requires a duty of care, a breach of that duty, and damage caused by the breach that is not too remote.

Duty. A doctor owes a duty of care to their patient — this is an established duty relationship, satisfying the Caparo Industries v Dickman [1990] test (harm to a patient from misdiagnosis is foreseeable; there is clear proximity in the doctor-patient relationship; and it is fair, just and reasonable to impose a duty), and reflecting the neighbour principle in Donoghue v Stevenson [1932]. A duty of care is therefore owed.

Breach. A doctor is judged not by the ordinary reasonable person but by the standard of a reasonably competent member of the profession: under Bolam v Friern Hospital Management Committee [1957], a doctor is not negligent if they act in accordance with a responsible body of medical opinion — but that opinion must be logically defensible (Bolitho [1998]). Crucially, the standard is objective, so Dr Khan's status as a junior doctor gives no allowance: a learner or junior is held to the standard of the competent practitioner (cf. Nettleship v Weston [1971]). Missing an 'obvious' fracture is something no responsible body of competent radiological opinion would defend, so the standard was breached.

Causation. Factual causation uses the 'but for' test (Barnett v Chelsea & Kensington Hospital [1969]): but for Dr Khan's failure to spot and treat the fracture, would Sam's arm have been permanently damaged? On the facts the damage flowed from the lack of timely treatment, so but-for causation is satisfied. (Had the arm been doomed regardless of treatment, the claim would fail at this stage, as it did in Barnett.) No novus actus interveniens breaks the chain.

Remoteness. The type of harm — physical injury from an untreated fracture — is plainly reasonably foreseeable from a negligent misdiagnosis (The Wagon Mound (No 1) [1961]), so the damage is not too remote.

Conclusion. A duty was owed and breached (Bolam/Bolitho), the breach caused Sam's injury (Barnett), and the harm is not too remote (Wagon Mound No 1). Dr Khan is therefore liable to Sam in negligence.

Liability for psychiatric injury (historically 'nervous shock') is restricted because the courts fear opening the floodgates to a vast number of claims. The key control is the distinction between primary and secondary victims.

Primary victims. A primary victim is someone who was directly involved in the incident and was within the zone of physical danger, or who reasonably feared for their own physical safety. The rule for primary victims is generous: provided that some personal injury (whether physical or psychiatric) was reasonably foreseeable, the defendant is liable for the psychiatric harm even if no physical injury in fact occurred. This was established in Page v Smith [1996], where the claimant, involved in a car crash, suffered the recurrence of a chronic fatigue condition; as physical injury was foreseeable, his psychiatric harm was recoverable.

Secondary victims. A secondary victim is not in personal danger but suffers psychiatric injury through witnessing harm to another. To keep claims within limits, the law imposes the strict control mechanisms laid down in Alcock v Chief Constable of South Yorkshire [1992] (the Hillsborough litigation). The claimant must prove: (1) a close tie of love and affection with the primary victim (presumed for spouses/parents, otherwise proved); (2) proximity in time and space to the shocking event or its immediate aftermath; and (3) that they perceived the event with their own unaided senses — seeing or hearing it directly, not learning of it from a third party or watching it on television.

Why the distinction matters. The distinction means primary victims recover relatively easily, while secondary victims face a demanding gateway. This reflects the policy tension between compensating genuine psychiatric harm and preventing an indeterminate number of claims from bystanders and distant relatives.

Conclusion. The law therefore draws a sharp line: primary victims need only foreseeable personal injury (Page v Smith), whereas secondary victims must satisfy all three Alcock control mechanisms — a structure widely criticised as arbitrary but designed to control liability.

Issue. Greg's liability in negligence must be assessed for each claimant — Hana, Iris and Jo — by working through duty, breach, causation, remoteness and any defences, plus the novel issue of psychiatric injury and a possible intervening act.

Hana — duty and breach. Greg owed a duty of care: physical injury to spectators from a misfired rocket is reasonably foreseeable, there is proximity between Greg and those nearby, and it is fair, just and reasonable to impose a duty (Caparo Industries v Dickman [1990]; Donoghue v Stevenson [1932]). Setting off a rocket sideways falls below the standard of the reasonable person (Blyth v Birmingham Waterworks [1856]); given the magnitude of harm from fireworks and the cheap, practicable precaution of aiming them safely (Paris v Stepney [1951]; Latimer v AEC [1953]), the duty is breached.

Hana — causation, remoteness and the thin skull. But for Greg's breach Hana would not have been struck (Barnett v Chelsea & Kensington Hospital [1969]). Physical injury from a firework is a foreseeable type of harm (The Wagon Mound (No 1) [1961]). Hana's rare bone condition makes her injuries far worse, but under the thin-skull rule Greg must take his victim as he finds her and is liable for the full extent of the harm (Smith v Leech Brain [1962]).

Hana — the road accident (novus actus interveniens). Hana then runs into the road and is hit by a car. The question is whether this is a novus actus interveniens breaking the chain. The claimant's own act breaks the chain only if it is unreasonable; an instinctive, panicked reaction to a sudden emergency caused by the defendant is foreseeable and not unreasonable, so it likely does not break the chain — Greg remains liable, though the speeding driver may be a joint tortfeasor (joint liability), and contributory negligence might reduce damages if Hana acted carelessly.

Iris — psychiatric injury (secondary victim). Iris suffers a recognised psychiatric illness from witnessing harm to her daughter, so she is a secondary victim and must satisfy the Alcock v Chief Constable of South Yorkshire [1992] control mechanisms: (1) a close tie of love and affection — present, as she is Hana's mother (presumed); (2) proximity in time and space — she was standing beside Hana and saw the event itself; (3) perception by her own unaided senses — she saw it directly. All three are satisfied, so Greg owes Iris a duty and is liable for her psychiatric injury.

Jo — psychiatric injury via broadcast. Jo watches on a live broadcast at home. Although she suffers genuine shock, she fails the Alcock mechanisms: she was not proximate in time and space, and she did not perceive the event with her own unaided senses — watching on television does not count (Alcock). Unless she is a primary victim (she is not — she was never in danger, cf. Page v Smith [1996]), Greg owes her no duty and is not liable to Jo.

Conclusion. Greg is liable to Hana for the full extent of her injuries (thin-skull) and probably for the road-accident injuries (no novus actus), jointly with the driver; he is liable to Iris as a secondary victim who meets all three Alcock mechanisms; but he is not liable to Jo, who fails the Alcock requirements. Contributory negligence may reduce Hana's award.

The statement raises whether the Caparo three-part test — reasonable foreseeability of harm, proximity, and whether it is fair, just and reasonable to impose a duty (Caparo Industries v Dickman [1990]) — is a satisfactory mechanism for deciding the existence of a duty of care. This essay argues that, while the test has real defects of certainty, it is on balance a workable and necessary control device, superior to what came before.

Background. The duty of care began with the neighbour principle in Donoghue v Stevenson [1932], expanded by the broad two-stage test in Anns v Merton, which was criticised for opening the door to too much liability. Caparo was the retreat from that expansion, replacing it with a more cautious, incremental three-part approach.

The case that Caparo is unsatisfactory. Several criticisms have force. First, the concepts of 'proximity' and 'fair, just and reasonable' are notoriously vague: they have no fixed legal content and allow judges to reach essentially policy-driven outcomes while presenting them as the application of legal principle. This undermines certainty and predictability, core values of the rule of law, because litigants cannot easily forecast whether a duty will be found. Secondly, critics argue the third limb simply gives judges a discretion to deny duties they dislike — for example to protect public bodies — masking policy as doctrine. Thirdly, in the vast majority of cases (established duties such as driver/road-user or doctor/patient) the elaborate test is redundant, adding nothing but length.

The case that Caparo is satisfactory. Against this, Caparo has genuine merits. It provides a structured framework that forces courts to address foreseeability, proximity and policy openly rather than expanding liability instinctively, as happened under Anns. The 'fair, just and reasonable' limb is a virtue, not just a vice: it allows the courts to manage the floodgates, prevent disproportionate or 'crushing' liability, and protect socially valuable activities — control that a purely foreseeability-based test could not provide. The modern, incremental approach (developing duties by analogy with established categories, reserving full Caparo analysis for genuinely novel situations) shows the courts using the test flexibly and sensibly rather than mechanically.

Evaluation. The criticisms are real but overstated. Vagueness is the price of flexibility: any duty test broad enough to cover the infinite variety of factual situations must use open-textured concepts, and certainty would have to be bought at the cost of injustice in unforeseen cases. The complaint that the test is redundant in settled cases misunderstands its role — it is a tool for the hard, novel cases, where some control device is indispensable. The honest concern is that 'fair, just and reasonable' lets policy operate covertly; but it is arguably better to have policy considered openly under an acknowledged limb than concealed within strained reasoning about 'proximity'.

Conclusion. Judged against the criterion that an effective duty test must balance certainty with the flexibility to do justice and control liability, Caparo is a reasonable, if imperfect, compromise. Its vagueness genuinely sacrifices predictability, but it is more disciplined than the over-broad Anns test and provides essential control over novel duties. The statement is therefore only partly correct: Caparo is unsatisfactory in its uncertainty, but satisfactory as a flexible and necessary mechanism — making it the least bad option rather than a flawed one to be abandoned.