Negligence — duty of care (Caparo and after)

Negligence is the central organising tort in Part IA: it is the law’s main technique for allocating the cost of accidentally caused harm. Its structure is conventionally divided into duty of care, breach, causation, remoteness, defences and remedies. This week concerns only the first of those questions: whether the defendant owed the claimant a duty of care at all.

The duty question is prior and controlling. A person may have behaved carelessly in ordinary language; the claimant may have suffered genuine loss; and there may even be factual causation. Yet, without a duty of care of the relevant scope, the negligence claim fails. Duty is therefore the law’s gatekeeping device. It marks the boundary between misfortune and legal responsibility.

The modern starting point is Donoghue v Stevenson, where Lord Atkin expressed the neighbour principle: one must take reasonable care to avoid acts or omissions likely to injure persons so closely and directly affected that they ought reasonably to be contemplated. That formulation was revolutionary because it suggested that negligence might be governed by principle rather than by a closed list of nominate duties. But it did not answer every case. Modern duty doctrine is shaped by the tension between general principle and incremental control.

The familiar formula is the threefold test associated with Caparo Industries plc v Dickman: reasonable foreseeability of damage; a relationship of proximity; and whether it is fair, just and reasonable to impose a duty. The danger, especially in a first supervision essay, is to treat these as three factual boxes. They are not. Foreseeability is necessary but rarely sufficient. Proximity is a legal conclusion about the closeness and character of the relationship. Fair, just and reasonable is not an invitation to personal morality; it is a structured inquiry into legal coherence, institutional competence, distributive consequences and consistency with existing categories.

After Caparo, the Supreme Court has repeatedly warned against using the threefold formula as a universal licence for judicial discretion. Robinson v Chief Constable of West Yorkshire Police is now essential. Lord Reed rejected the idea that courts should apply Caparo afresh in every negligence case. Where an established duty category exists, ordinary reasoning by analogy and precedent should be used. Caparo has particular importance in novel cases, but even there the common law develops incrementally.

For Cambridge purposes, the best answer is neither a slogan nor a list. It explains why duty exists, distinguishes acts from omissions, physical injury from pure economic loss, private defendants from public authorities, and established categories from novel claims. It then asks whether imposing responsibility would extend negligence coherently. That is the intellectual task behind the phrase “Caparo and after”.

The modern law of negligence was not born fully formed. Nineteenth-century tort law recognised duties in particular relationships: carriers and passengers, employers and employees, occupiers and entrants, manufacturers and consumers in limited circumstances, professionals and clients. Liability was largely categorical. The question was not “is there a duty in principle?” but “does this case fall within a recognised head of responsibility?”

Donoghue v Stevenson altered the method. Mrs Donoghue alleged illness after consuming ginger beer containing the remains of a decomposed snail. She had not bought the bottle; contract could not assist her. The House of Lords nevertheless held that a manufacturer could owe a duty of care to the ultimate consumer. Lord Atkin’s neighbour principle gave negligence a moral and juridical centre: liability could arise where the defendant should have contemplated that careless conduct would likely injure persons closely and directly affected by it. The case did not abolish categories, but it made it possible to reason from principle across them.

Mid-twentieth-century law expanded. Dorset Yacht allowed a claim against the Home Office where Borstal trainees escaped and damaged yachts, emphasising foreseeability and control. The high-water mark was Anns v Merton London Borough Council. Lord Wilberforce proposed a two-stage approach: first, ask whether there was sufficient proximity or neighbourhood such that carelessness might foreseeably cause damage; secondly, consider whether there were reasons to negative or limit the duty. The Anns approach appeared to presume a duty once foreseeability and proximity were shown, subject to policy reasons against liability.

That expansion soon provoked correction. Courts became anxious about indeterminate liability, pure economic loss, defensive administration, and the judicial creation of broad financial burdens. In Murphy v Brentwood District Council the House of Lords overruled Anns in relation to local authority liability for defective buildings and rejected recovery for pure economic loss arising from a dangerous defect before injury or property damage occurred. The law retreated from generalised liability.

Caparo then gave the canonical modern formulation. The claimant, an investor and later majority shareholder, sued auditors for negligent accounts. The House of Lords denied a duty to investors in the market or to shareholders making takeover decisions. It insisted on foreseeability, proximity, and the fairness, justice and reasonableness of imposing a duty. The decision is often read as replacing Anns with a more restrictive, incremental approach.

The post-Caparo period is one of refinement rather than simple restriction. Duties for personal injury caused by positive acts remain orthodox and relatively uncontroversial. The difficult cases concern omissions, failures to protect from third parties, pure economic loss, psychiatric harm, public authorities, assumption of responsibility and statutory contexts. Robinson and Poole Borough Council v GN are the leading modern warnings: Caparo must not be treated as a universal test detached from precedent. The common law proceeds by analogy from established categories, using policy explicitly but carefully.

The historical lesson is therefore methodological. Donoghue supplies the aspiration to principle; Anns shows the risk of over-generalisation; Caparo supplies controls; Robinson restores the primacy of precedent and incremental development. A Cambridge answer should show that duty is not a single test but a disciplined form of common-law reasoning.

The first principle is that duty of care concerns the existence and scope of a legal obligation owed by this defendant to this claimant in respect of this kind of harm. It is not enough to ask whether the defendant was generally careless. Duty is relational and specific. A duty to avoid causing physical injury by one’s positive acts does not necessarily entail a duty to prevent economic loss, protect against criminals, rescue a stranger, or confer a benefit.

The second principle is reasonable foreseeability. The claimant must belong to a class of persons whom the defendant ought reasonably to have had in contemplation, and the kind of harm must be foreseeable. Foreseeability is a low threshold in many personal injury cases. If a driver mounts the pavement, injury to pedestrians is foreseeable. But foreseeability cannot do all the work. Nearly every modern difficult duty case involves foreseeable harm. If foreseeability alone sufficed, liability would expand without principled limit.

The third principle is proximity. Proximity is not merely physical closeness. It denotes a legally significant relationship of closeness, directness, reliance, control, assumption of responsibility, vulnerability, or causal connection. In Donoghue, proximity existed between manufacturer and ultimate consumer because the product was intended to reach consumers without intermediate inspection. In Caparo, the auditors’ statutory audit was not prepared for investors’ takeover decisions; the relationship lacked the necessary proximity for that purpose. Proximity often does the hidden work in duty cases because it distinguishes actionable reliance or direct endangerment from diffuse social effects.

The leading cases show a pendulum between principle and restraint. Donoghue v Stevenson provides the symbolic origin of modern negligence. Its significance lies not merely in the snail, but in the recognition that duty may be reasoned from a general principle of neighbourliness. Manufacturers could owe duties to ultimate consumers where products were intended to reach them without further inspection.

Dorset Yacht then showed that duty could extend beyond direct physical impact by the defendant. The Home Office’s officers had control over trainees who escaped and damaged property. The case is important because it recognised liability for harm caused by third parties where the defendant’s relationship with those third parties and the risk created sufficient proximity.

Anns was the expansive moment. Lord Wilberforce’s two-stage method suggested that once proximity and foreseeable harm were present, the burden shifted to the defendant to identify policy reasons against liability. It encouraged a broad conception of negligence. But it also generated anxiety. The common law seemed to be assuming responsibility for losses better handled by contract, insurance or legislation.

Murphy marks the correction. The House of Lords rejected Anns in the defective building context and treated the cost of repairing a dangerous defect, before injury or property damage, as pure economic loss. This was not merely a buildings case. It signalled a more restrictive approach to novel duties and a renewed concern with the boundary between tort and contract.

Caparo is the canonical modern case. Auditors prepared statutory accounts for the company’s shareholders as a body, not for investors deciding whether to buy shares or launch a takeover. The House of Lords denied the duty. The case matters because it supplied the tripartite language and emphasised incremental development. Yet it should not be misread as a universal algorithm.

Michael v Chief Constable of South Wales Police illustrates the difficulty of omissions and third-party wrongdoing. The police failed to respond promptly to an emergency call from a woman later killed by her former partner. The Supreme Court, by majority, denied a common-law duty. The case confirms the general rule that police do not owe a duty to protect individuals from harm by criminals merely because such harm is foreseeable, absent assumption of responsibility or another recognised basis.

Robinson rebalanced the picture. Police officers attempting an arrest knocked over an elderly pedestrian. The Supreme Court held that the police owed an ordinary duty not to cause foreseeable physical injury by their positive acts. Robinson is indispensable because it prevents Caparo from being used to dissolve established duties into open-ended policy discretion.

Poole Borough Council v GN continues that clarification. A local authority’s failure to protect children from anti-social neighbours was treated as an omission unless assumption of responsibility or creation of danger could be shown. Statutory child protection powers did not, of themselves, create a common-law duty. The case is now central to public authority duty problems.

See practice questions section below.

See further reading section below.