Foreseeability in English law

Few concepts in English law carry as much weight, or occupy so pervasive a role, as foreseeability. It operates across the law of tort and the law of contract, appearing at multiple stages of legal analysis: in the establishment of a duty of care, in the assessment of breach, in the determination of remoteness of damage, and in the quantification of recoverable loss. It is, as Lord Denning MR once observed, a thread running through the whole fabric of negligence liability (Spartan Steel & Alloys Ltd v Martin & Co (Contractors) Ltd [1973]).

Part I: Historical foundations

The origins of foreseeability

The idea that liability for wrongful harm should be connected to the foreseeability of that harm is not a modern invention. Professor Helen Scott of the University of Oxford has demonstrated, in her 2019 Current Legal Problems lecture ‘The History of Foreseeability’, that the foreseeability principle can be traced to classical Roman jurisprudence – specifically to a text attributed to the jurist Paul in Digest 9.2.31, which defines culpa under the lex Aquilia in terms of what ‘could have been foreseen by a diligent man’ (a diligente provideri poterit) (Scott, 2019).

Scott argues persuasively that this fragment, mediated through the work of sixteenth-century humanists such as Donellus and nineteenth-century German Romanists whose work was made available to English readers through Grueber’s The Roman Law of Damage to Property (1886), exercised a formative – if largely unacknowledged – influence upon Lord Atkin’s neighbour principle in Donoghue v Stevenson [1932] AC 562 (Scott, 2019; Cairns, 2013).

In the English common law tradition, however, the foreseeability principle entered the law of obligations through a different route. David Ibbetson has shown that a test of remoteness based on reasonable foresight was received into English law from the writings of eighteenth-century natural lawyers such as Burlamaqui, via Francis Buller’s Introduction to Trials at Nisi Prius (1788), and was applied in the remoteness context by Chief Baron Pollock in Rigby v Hewitt (1850) 5 Ex 240 and Greenland v Chaplin (1850) 5 Ex 243 (Ibbetson, 1999).

From remoteness, the concept migrated to the assessment of breach – as in Baron Alderson’s famous definition of negligence in Blyth v Birmingham Waterworks Company (1856) 11 Ex 781 – and thence, through Sir William Brett MR’s generalisation in Heaven v Pender (1883) 11 QBD 503, to the duty of care itself (Ibbetson, 1999).

Donoghue v Stevenson and the neighbour principle

The crystallisation of foreseeability as a central organising concept in the law of negligence came with Lord Atkin’s speech in Donoghue v Stevenson [1932] AC 562, 580:

You must take reasonable care to avoid acts or omissions which you can reasonably foresee would be likely to injure your neighbour. Who, then, in law is my neighbour? The answer seems to be – persons who are so closely and directly affected by my act that I ought reasonably to have them in contemplation as being so affected when I am directing my mind to the acts or omissions which are called in question.

This formulation embedded foreseeability at the very heart of the duty of care inquiry. It has remained there ever since, though – as we shall see – its precise role has been the subject of sustained judicial and academic debate. Lord Atkin’s explicit reference to ‘culpa‘ in the passage immediately preceding the neighbour principle lends support to Professor Scott’s thesis that he was drawing, at least in part, upon the Romanist tradition (Donoghue v Stevenson [1932] AC 562; Scott, 2019).

Hadley v Baxendale and foreseeability in contract

In the law of contract, foreseeability entered the remoteness inquiry through the seminal decision of the Court of Exchequer in Hadley v Baxendale (1854) 9 Exch 341. Baron Alderson’s judgment established the two-limb test that continues to govern the recoverability of damages for breach of contract:

Where two parties have made a contract which one of them has broken, the damages which the other party ought to receive in respect of such breach of contract should be such as may fairly and reasonably be considered either arising naturally, i.e., according to the usual course of things, from such breach of contract itself, or such as may reasonably be supposed to have been in the contemplation of both parties, at the time they made the contract, as the probable result of the breach of it.

The first limb captures losses arising in the ordinary course; the second, losses flowing from special circumstances communicated to the defaulting party. Both are fundamentally concerned with what was foreseeable – or, more precisely, what was within the reasonable contemplation of the parties at the time of contracting.

The rule has proved extraordinarily durable: it was affirmed and refined inVictoria Laundry (Windsor) Ltd v Newman Industries Ltd [1949] 2 KB 528, and again in Koufos v C Czarnikow Ltd (The Heron II) [1969] 1 AC 350, where Lord Reid formulated the applicable standard as whether the loss was ‘not unlikely’ to result from the breach.

Part II: Foreseeability in the tort of negligence

Foreseeability and the duty of care

Following the expansion and subsequent retreat from Lord Wilberforce’s two-stage test in Anns v Merton London Borough Council [1978] AC 728, the House of Lords in Caparo Industries plc v Dickman [1990] 2 AC 605 set out a three-stage framework for the recognition of a duty of care in novel situations. Lord Bridge identified the requirements as: (i) reasonable foreseeability of harm; (ii) a relationship of proximity between the parties; and (iii) that it be fair, just and reasonable to impose a duty of care.

However, Lord Bridge was careful to emphasise that these ‘ingredients’ were not ‘susceptible of any such precise definition as would be necessary to give them utility as practical tests, but amount in effect to little more than convenient labels to attach to the features of different specific situations’. He endorsed the incremental approach advocated by Brennan J in the High Court of Australia in Sutherland Shire Council v Heyman (1985) 60 ALR 1: the law should develop ‘novel categories of negligence incrementally and by analogy with established categories’.

The proper understanding of Caparo was authoritatively clarified by the Supreme Court in Robinson v Chief Constable of West Yorkshire Police [2018] UKSC 4, [2018] AC 736 – a decision of the first importance for the modern law of negligence. Lord Reed, delivering the majority judgment (with which Lady Hale and Lord Hodge agreed), held that:

The proposition that there is a Caparo test which applies to all claims in the modern law of negligence, and that in consequence the court will only impose a duty of care where it considers it fair, just and reasonable to do so on the particular facts, is mistaken.

Caparo, Lord Reed explained, ‘repudiated the idea that there is a single test which can be applied in all cases in order to determine whether a duty of care exists, and instead [adopted] an approach based, in the manner characteristic of the common law, on precedent, and on the development of the law incrementally and by analogy with established authorities’.

In established categories of liability – motorists to other road users, employers to employees, manufacturers to consumers – the elements of proximity and fairness are already embedded in the recognised duty and need not be reconsidered afresh. It is only in genuinely novel situations that the full Caparo analysis, including considerations of what is fair, just and reasonable, becomes necessary.

On the facts, the police officers who had knocked over an elderly pedestrian whilst attempting to arrest a suspect on a busy shopping street owed her a duty of care. The risk of injury was reasonably foreseeable – indeed, actually foreseen – and the case fell within the established category of liability for personal injury caused by positive acts (Robinson v Chief Constable of West Yorkshire Police [2018] UKSC 4).

The role of foreseeability within the duty of care inquiry is therefore nuanced. In established categories, it operates almost silently: the duty is assumed. In novel cases, foreseeability remains a necessary – but never a sufficient – condition. As Lord Bridge observed in Caparo, foreseeability of harm to the claimant is only the starting point; proximity and policy considerations must also be satisfied.

Foreseeability and breach

Once a duty of care is established, foreseeability plays a further role in determining whether the defendant’s conduct fell below the standard of the reasonable person. The classic formulation remains that of Baron Alderson in Blyth v Birmingham Waterworks Company (1856) 11 Ex 781, 784: negligence is ‘the omission to do something which a reasonable man, guided upon those considerations which ordinarily regulate the conduct of human affairs, would do, or doing something which a prudent and reasonable man would not do.’

The modern approach to breach involves a balancing exercise, classically articulated by the House of Lords in Bolton v Stone [1951] AC 850. The court weighs the magnitude of the risk (both likelihood and gravity of potential harm) against the cost and practicability of precautions and the social utility of the defendant’s activity. Within this calculus, foreseeability operates as a threshold: if the risk of harm was not reasonably foreseeable, the defendant is not negligent, however unfortunate the outcome. As Lord Reid put it in Bolton v Stone, one is not negligent in respect of a risk so small that ‘a reasonable man would have been justified in disregarding it’ (Bolton v Stone [1951] AC 850). 

The point was vividly illustrated in Roe v Minister of Health [1954] 2 QB 66, where anaesthetic contamination caused paralysis through a mechanism not known to medical science at the time: the defendant was exculpated because the risk was unforeseeable.

Foreseeability and remoteness of damage

It is in the context of remoteness that foreseeability has undergone its most dramatic doctrinal transformation.

Under the rule established in Re Polemis and Furness, Withy & Co [1921] 3 KB 560, a defendant was liable for all the direct consequences of a negligent act, however unforeseeable. This rule was swept away by the Privy Council’s decision in Overseas Tankship (UK) Ltd v Morts Dock & Engineering Co Ltd (The Wagon Mound No 1) [1961] AC 388. The facts are well known: furnace oil negligently discharged from the defendant’s vessel in Sydney Harbour was ignited by welding sparks, causing a catastrophic fire that destroyed the claimant’s wharf. The Privy Council held that the defendant was not liable for the fire damage, because – on the evidence before the court – the ignition of furnace oil on water was not a reasonably foreseeable consequence of the spillage.

Viscount Simonds, delivering the advice of the Board, overruled Re Polemis and established that a defendant is liable only for damage of a kind that was reasonably foreseeable. The test of remoteness in tort was thereby aligned with the concept of reasonable foreseeability (Overseas Tankship (UK) Ltd v Morts Dock & Engineering Co Ltd (The Wagon Mound No 1) [1961] AC 388).

The position was refined, however, by the second Wagon Mound case – Overseas Tankship (UK) Ltd v Miller Steamship Co Pty Ltd (The Wagon Mound No 2) [1967] 1 AC 617 – which arose from the same incident but involved different claimants and, critically, different evidence. The claimants in this case proved that, although the risk of furnace oil igniting on water was small, it was a ‘real risk’ – not a mere possibility – that a reasonable ship’s engineer would have been aware of. Lord Reid held that a risk was foreseeable so long as it was a ‘real risk’ which a reasonable person would not dismiss as far-fetched, particularly where the negligent act served no useful purpose and the cost of prevention was minimal.

Taken together, the two Wagon Mound decisions establish that:

(a) the test for remoteness in tort is reasonable foreseeability of the type of damage; and

(b) foreseeability does not require probability – a ‘real risk’, even if small, is sufficient.

Type of damage versus manner of occurrence

A crucially important refinement of the Wagon Mound principle was effected by the House of Lords in Hughes v Lord Advocate [1963] AC 837. Post Office workers left a manhole covered only by a tent, surrounded by paraffin lamps. A young boy knocked a lamp into the manhole, causing an explosion and severe burns. The defendant argued that the explosion was unforeseeable. The House of Lords held that it was sufficient that burns from the paraffin lamps were foreseeable; the explosion was merely a variant of the foreseeable type of harm. Lord Reid stated that the accident was ‘but a variant of foreseeable’ harm; it was unnecessary that the defendant foresee ‘the precise concatenation of circumstances which lead up to the accident’.

The practical significance of this principle cannot be overstated. A defendant need not foresee the precise manner in which harm occurs, so long as the kind of harm was foreseeable. This was further illustrated – in negative – by Doughty v Turner Manufacturing Company [1964] 1 QB 518, where the Court of Appeal held that an eruption caused by a chemical reaction was a different type of harm from the foreseeable splash that might have resulted from a lid falling into a cauldron.

The eggshell skull rule and its interaction with foreseeability

The principle that a tortfeasor must ‘take the victim as he finds him’ – the so-called eggshell skull rule (or thin skull rule) – sits alongside, but is conceptually distinct from, the foreseeability requirement for remoteness. The leading authority is Smith v Leech Brain & Co Ltd [1962] 2 QB 405, where a negligently caused burn to an employee’s lip activated a pre-cancerous condition, leading to his death from cancer. Lord Parker CJ held that the defendants were liable for the full consequences: if some physical harm was foreseeable, the defendant was responsible for all the harm that flowed, even if its extent was wholly unforeseeable. 

The rule survived the transition from Re Polemis to The Wagon Mound because it operates at the level of extent rather than type: the type of harm (a burn) was foreseeable; the extent (cancer and death) was not, but did not need to be.

Part III: Foreseeability in the law of contract

The rule in Hadley v Baxendale revisited

The contractual remoteness test, whilst sharing the language of foreseeability with its tortious counterpart, has important structural differences. As refined in Victoria Laundry and The Heron II, the test asks whether the loss was of a type that the parties, at the time of contracting, ought reasonably to have had in their contemplation as a ‘not unlikely’ consequence of the breach (Koufos v C Czarnikow Ltd (The Heron II) [1969] 1 AC 350). The inquiry is directed to the contemplation of the parties – a concept that imports both the imputed knowledge of a reasonable person in the defendant’s position and any actual knowledge of special circumstances communicated before the contract was formed.

The assessment is made at the date of contracting (in contrast to the tortious test, which is assessed at the date of the wrong). This reflects the contractual rationale: by entering into the agreement, each party assumes certain risks, and the scope of those assumed risks is fixed at formation.

The Achilleas: assumption of responsibility and the ‘broader’ test

The decision of the House of Lords in The Achilleas (Transfield Shipping Inc v Mercator Shipping Inc) introduced a significant — and, it must be said, not entirely settled – refinement to the contractual remoteness test. Charterers returned a vessel nine days late, during which time the market had fallen sharply, causing the owners to lose the benefit of a lucrative follow-on fixture. The owners claimed the difference in freight rates over the entire period of the subsequent charter; the charterers contended that damages should be limited to the difference between the charter rate and the market rate for the nine days of overrun.

The House of Lords unanimously held for the charterers, but on two different bases. Lord Hoffmann (with whom Lord Hope agreed) held that the ordinary foreseeability test was not ‘an external rule of law, imposed upon the parties’, but rather ‘a prima facie assumption about what the parties may be taken to have intended’, capable of rebuttal where ‘the context, surrounding circumstances or general understanding in the relevant market shows that a party would not reasonably have been regarded as assuming responsibility for such losses’. On this approach, foreseeability is supplemented – and potentially overridden – by an inquiry into the objective assumption of responsibility.

Lord Rodger and Baroness Hale, by contrast, decided the case on the narrower ground that the loss was simply not foreseeable on the orthodox Hadley v Baxendale test, given the exceptional market volatility.

The aftermath of The Achilleas was considered by Hamblen J in Sylvia Shipping Co Ltd v Progress Bulk Carriers Ltd [2010] EWHC 542 (Comm), who concluded that the ‘orthodox’ foreseeability approach remains the general test, with the ‘broader’ assumption-of-responsibility test applicable only where ‘the orthodox approach leads to unquantifiable, unpredictable, uncontrollable or disproportionate liability, or where there is clear evidence that such liability would be contrary to market expectations’.

The tension between these approaches has not been definitively resolved. Practitioners should be alive to the possibility that, in an appropriate case – particularly in a specialised commercial context with established market expectations – a defendant may resist liability for a foreseeable loss by demonstrating that it falls outside the scope of the risks reasonably assumed under the contract

Part IV: Foreseeability and the scope of duty – modern developments

Manchester Building Society v Grant Thornton [2021] UKSC 20

The Supreme Court’s decision in Manchester Building Society v Grant Thornton UK LLP [2021] UKSC 20 represents the most significant recent development in the law of professional negligence, and it has important implications for the role of foreseeability across the law of obligations more broadly.

The case concerned negligent accountancy advice: Grant Thornton had wrongly advised Manchester Building Society that it could use ‘hedge accounting’, upon which the Society relied in entering into long-term interest rate swaps. When the error was discovered, the Society was forced to close out the swaps at a cost of over £32 million. The question was whether this loss fell within the scope of Grant Thornton’s duty of care.

The Supreme Court unanimously allowed the appeal. The leading judgment of Lord Hodge and Lord Sales (with whom Lord Reed, Lady Black and Lord Kitchin agreed) established the following propositions:

1. The scope of duty is governed by its purpose. The scope of the duty of care assumed by a professional adviser is determined by the purpose of the duty, judged objectively by reference to the reason why the advice was being given.
2. The information/advice distinction is not rigid. The distinction drawn by Lord Hoffmann in South Australia Asset Management Corp v York Montague Ltd [1997] AC 191 (SAAMCO) between ‘information’ cases and ‘advice’ cases should not be treated as a rigid rule; cases exist on a spectrum (Manchester Building Society v Grant Thornton UK LLP [2021] UKSC 20).
3. The counterfactual is a cross-check, not the test. The SAAMCO counterfactual – asking whether the loss would have occurred if the advice had been correct – is a useful cross-check but is ‘subordinate to’ the analysis of the purpose of the duty and ‘should not supplant or subsume it’ (Manchester Building Society v Grant Thornton UK LLP [2021] UKSC 20).

The significance of this decision for the concept of foreseeability is twofold. First, the Supreme Court confirmed that reasonable foreseeability of loss is a necessary but not sufficient condition for its recovery: loss that is reasonably foreseeable may nonetheless fall outside the scope of the defendant’s duty (Manchester Building Society v Grant Thornton UK LLP [2021] UKSC 20). 

Secondly, and more fundamentally, the decision illustrates the limits of foreseeability as an organising principle. The crucial question is not whether the loss was foreseeable, but whether it resulted from the kind of risk that the defendant’s duty was supposed to guard against. The focus shifts from prediction to purpose.

Khan v Meadows [2021] UKSC 21

Heard by the same panel alongside Manchester Building Society, Khan v Meadows applied the restated scope-of-duty principles to medical negligence. A GP negligently failed to test a patient for haemophilia carrier status; the patient subsequently gave birth to a child with both haemophilia and autism. The Supreme Court held that the GP was liable only for the consequences of haemophilia, not autism, because the latter was unrelated to the scope of the GP’s duty. The case further demonstrates the analytical framework in which foreseeability is a necessary threshold, but the scope-of-duty inquiry provides the operative limit on recovery.

Part V: Foreseeability across categories of harm

Physical injury

In cases of physical harm caused by positive acts, foreseeability is most straightforward. As confirmed in Robinson, once it is established that a defendant owed a duty of care – typically self-evident in established categories – the question is whether harm of the type suffered was a foreseeable consequence of the defendant’s breach. The eggshell skull rule ensures that the full extent of harm is recoverable.

Psychiatric injury

Foreseeability plays a more restrictive role in relation to psychiatric harm. For primary victims (those within the zone of foreseeable physical danger), it is sufficient that some personal injury was foreseeable: Page v Smith [1996] AC 155. For secondary victims, the stricter Alcock v Chief Constable of South Yorkshire Police [1992] 1 AC 310 control mechanisms apply, requiring – among other things – foreseeability of psychiatric injury to a person of ‘normal fortitude’. The law in this area has been the subject of recent statutory attention following the Hillsborough inquests, though the common law framework remains substantially intact.

Pure economic loss

Foreseeability alone is insufficient to ground liability for pure economic loss. The Caparo decision itself was concerned with this category: auditors owed no duty to investors who relied upon audited accounts, because the purpose of the audit was not to guide investment decisions. The additional requirement of proximity, typically through an assumption of responsibility (Hedley Byrne & Co Ltd v Heller & Partners Ltd [1964] AC 465), or through established principles governing recovery for economic loss consequent upon damage to property owned by a third party, must be satisfied.

Public authorities

Following Robinson and Poole Borough Council v GN [2019] UKSC 25, public authorities are subject to the same negligence principles as private persons. Foreseeability of harm does not, without more, generate a duty to prevent harm; the general omissions principle applies. However, where a public authority creates a source of danger through positive action, ordinary foreseeability principles apply to establish a duty of care.

Part VI: Critical reflections

The limits of foreseeability as a normative concept

Despite its pervasiveness, foreseeability has long been the subject of intellectual disquiet. As Professor Scott has argued, drawing on the historical analysis of the concept’s Aristotelian and Roman origins, foreseeability may be inherently better suited to function as an exclusionary principle – ruling out liability where harm was unforeseeable – than as a generative one that positively justifies the imposition of liability (Scott, 2019). The modern law’s resort to additional requirements – proximity, fairness, scope of duty, assumption of responsibility — may be understood as reflecting the recognition that foreseeability alone cannot bear the normative weight that Lord Atkin’s neighbour principle appeared to place upon it.

The Supreme Court’s recent jurisprudence, particularly in Robinson and Manchester Building Society v Grant Thornton, can be read as an implicit acknowledgement of this point. The emphasis on established categories, on the purpose of the duty, and on the incremental development of the law all serve to constrain foreseeability’s role and to locate the true basis of liability in more particularised considerations.

Foreseeability and cognitive bias

A further challenge to foreseeability as a legal concept has been posed by advances in behavioural psychology. Daniel Kahneman’s work on heuristics and biases – notably the ‘availability heuristic’ – has demonstrated that human beings are systematically poor at assessing the probability of events (Kahneman, 2012; Tversky and Kahneman, 1973). What is ‘foreseeable’ may be heavily influenced by the vividness with which instances of a particular harm come to mind, rather than by any rational assessment of probability. This concern, while primarily academic, has practical implications: it suggests that the foreseeability test, particularly at the stage of breach, may be less determinate and objective than the legal standard of the ‘reasonable person’ implies.

Conclusion

Foreseeability remains an indispensable element of English private law. It operates, in different guises, across the duty of care, breach, remoteness and contractual damages inquiries.

Yet the role of foreseeability is neither uniform nor unlimited. The Supreme Court’s recent decisions – Robinson, Manchester Building Society v Grant Thornton, and Khan v Meadows – have together effected a significant clarification: foreseeability is a necessary threshold, but it is not the determinative question.

In tort, the crucial inquiry is increasingly framed in terms of the purpose and scope of the duty of care; in contract, the Achilleas has introduced the possibility that assumed responsibility may override orthodox foreseeability. The practitioner, the student and the researcher would do well to attend not merely to whether harm was foreseeable, but to the deeper question of why – given all the circumstances – liability should or should not attach.

You may also like: Foreseeability cases

Table of principal cases

Case	Citation	Principle
Hadley v Baxendale	(1854) 9 Exch 341	Two-limb test for remoteness in contract
Blyth v Birmingham Waterworks Co	(1856) 11 Ex 781	Definition of negligence; foreseeability in breach
Donoghue v Stevenson	[1932] AC 562	Neighbour principle; foreseeability in duty of care
Bolton v Stone	[1951] AC 850	Risk-balancing in breach; magnitude of risk
The Wagon Mound (No 1)	[1961] AC 388	Remoteness: foreseeability of type of damage
Smith v Leech Brain & Co	[1962] 2 QB 405	Eggshell skull rule survives foreseeability test
Hughes v Lord Advocate	[1963] AC 837	Type, not manner, of harm must be foreseeable
Hedley Byrne v Heller	[1964] AC 465	Assumption of responsibility for economic loss
The Wagon Mound (No 2)	[1967] 1 AC 617	‘Real risk’ sufficient for foreseeability
The Heron II	[1969] 1 AC 350	Contractual remoteness: ‘not unlikely’ test
Caparo Industries v Dickman	[1990] 2 AC 605	Three-stage framework for duty of care
South Australia Asset Management Corp (SAAMCO)	[1997] AC 191	Scope of duty limits recovery beyond foreseeability
The Achilleas (Transfield v Mercator)	[2009] 1 AC 61	Assumption of responsibility in contractual remoteness
Robinson v Chief Constable of West Yorkshire	[2018] UKSC 4	Caparo not a universal test; incremental approach
Manchester Building Society v Grant Thornton	[2021] UKSC 20	Scope of duty governed by purpose; SAAMCO restated
Khan v Meadows	[2021] UKSC 21	Scope of duty in medical negligence