Singapore - Your Partner for Legal Solutions in Asia
20.1.1 In the more than eighty years since its inception as a distinct cause of action in Donoghue v Stevenson [1932] AC 562 (Donoghue), negligence has developed to become the pre-eminent tort, eclipsing older actions such as trespass, nuisance and breach of statutory duty.
20.1.2 The law of negligence in Singapore is based largely on English law, although there are areas in which the Singapore courts have chosen to depart from the principles espoused by the UK courts. While the law referred to here will, wherever possible, be that applied by the courts in Singapore (and occasionally Malaysia), reference will also be made to the jurisprudence of other jurisdictions – notably the UK and Australia – which have influenced, or are influencing, the development of the law of negligence in Singapore.
A. Requirements of a tort: claimant must have suffered recoverable damage arising from a breach of legal duty owed by defendant
20.1.3 Negligence as a tort requires more than mere lack of care. A claimant who wishes to sue in negligence must show:
  • that the defendant owed him a legal duty to take care;
  • that there was a breach of this legal duty by the defendant; and
  • that the breach caused him recoverable damage.
A. Duty of care
(1) Duty distinguishes situations in which a claim may be entertained from those where no action is possible
20.2.1 Duty is an artificial conceptual barrier which the claimant must overcome before his action can even be considered. Its role is to keep the tort of negligence within manageable proportions by distinguishing situations in which a claim may, in principle, be entertained from those in which no action is possible.
(2) Factors influencing the existence of a duty
20.2.2 The question of whether or not a duty exists is influenced by a number of factors, such as:
  • the type of claimant (eg, socially sympathetic claimants such as rescuers are generally owed a duty of care in a wider range of situations than are less sympathetic ones such as trespassers);
  • the type of defendant (eg, defendants with public functions owe a duty of care in more limited circumstances than do individual defendants);
  • the nature of the damage caused (eg, a defendant almost always owes a duty of care not to cause physical damage to person or property through his negligent act, but the duty owed with respect to psychiatric harm and pure economic loss is more restricted); and
  • the nature of the conduct (eg, active conduct is more likely to give rise to a duty of care on the part of a defendant than is a mere omission).
Donoghue v Stevenson – The Neighbour Principle
20.2.3 In Donoghue Lord Atkin laid down the foundation for the duty of care. Under his ‘neighbour principle,’ a defendant must avoid acts or omissions which will foreseeably harm persons who are so closely and directly affected by his acts or omissions that he ought to have them in mind as being so affected. The neighbour principle remains the backbone of duty, but in the ensuing years the courts have developed more complex tests. These tests, while generally built around the key element of foreseeability, have attempted to reflect more accurately some of the other factors inherent in establishing duty.
The Anns Two-stage Test
(1) Two stages: proximity based on foreseeability of harm and considering of policy factors
20.2.4 In Anns v Merton London Borough Council [1978] AC 728 (Anns), Lord Wilberforce concluded that duty effectively comprised two stages. The first stage, derived from the neighbour principle, was a relationship of proximity or neighbourhood based on foreseeability of harm. The second was the consideration of policy factors which might negative, reduce or limit the scope of the duty, or the class of persons to whom it was owed, or the damages to which it might give rise.
(2) Difficulties arising from the two-stage test
20.2.5 The two-stage test led to expansionary decisions. This was partly because the notion of duty based on foreseeability without overt consideration of precedents at the first stage was inherently suited to developing, rather than restricting, the law. But it was also due to the fact that many judges were uncomfortable with the open articulation of policy, which led to the second stage of the test being under-used.
The Caparo Three-part Test
(1) Three stages: foreseeability, proximity and for imposing a duty to be fair, just and reasonable in the circumstances
(2) Difficulties arising from the three-part test
20.2.7 The three-part test remains – at least in theory – applicable in the UK, but it has been abandoned in Australia (see Sullivan v Moody [2001] HCA 59; (2001) 207 CLR 562), which now favours a ‘salient features’ approach to the determination of duty, due in large part to concern about the unsatisfactory nature of the proximity requirement. Although introduced as a tool for filtering out claims which lack the requisite closeness, proximity has always been a notoriously vague concept and its role has been undermined by its nebulous and indefinable nature. Canada has adopted a modified version of the Anns test, incorporating aspects of the Caparo test (see Cooper v Hobart [2001] SCJ No 76; [2001] 3 SCR 537). At the first stage, there must be reasonable foreseeability of harm and sufficient proximity between the parties for it to be fair and just to impose a duty of care. At the second stage, the court examines whether there are residual public policy considerations to justify denying liability.
20.2.8 In recent years, the courts have moved away from the somewhat reactionary approach which marked their response to Anns. As a result, even in jurisdictions where Caparo still applies, negligence has been allowed more scope for development, although still in a largely incremental manner.
E. Approach in Singapore: the Spandeck test
(1) Establishing duty in Singapore prior to the Spandeck test
(2) The Spandeck test: duty in all situations determined by two-stage test comprising proximity and policy considerations
A. Duty of care normally held to exist in straightforward cases involving physical damage to person or property
B. Situations where courts do not recognise duty because of public policy reasons
20.3.2 In less typical circumstances, the courts are often more circumspect. They have, at various times, and in various jurisdictions, refused – largely for reasons of public policy – to recognize the existence of a duty of care in a range of situations, such as:
  • where there is a conflict between negligence and other torts (eg, damage caused by a negligent act or statement which would be protected under the defence of qualified privilege in defamation – although note that no such conflict is held to arise in relation to employee-references: see Ramesh s/o Krishnan v AXA Life Insurance Singapore Pte Ltd [2015] SGHC 125);
  • where it is considered necessary to accord immunity to certain classes of defendants (eg, damage caused as a result of judicial negligence).
20.3.3 In addition, there are a number of broad categories in which particular rules have been established to restrict the situations in which a duty is owed. These categories are examined below.
C. Psychiatric harm: recovery possible when “three proximities” are fulfilled
Psychiatric Harm
20.3.4 Historically, the courts were unwilling to allow recovery for negligently inflicted psychiatric illness. This unwillingness stemmed from an incomplete understanding of mental illness, and from fears that allowing recovery for mental, as opposed to physical, harm would give rise to fraudulent claims and lead to a potential flood of litigation.
D. Pure economic loss not linked to physical damage
20.3.7 The courts have always allowed recovery for economic loss which flows from physical damage: Spartan Steel and Alloys Ltd v Martin & Co [1972] QB 27.
20.3.8 It was, however, historically impossible to recover for ‘pure’ economic loss – ie, loss which could not be linked to physical damage. The refusal to allow such claims was attributable to a number of concerns, the most significant of which was the perceived danger of a possible flood of litigation due to the knock-on effect of economic damage.
(1) Negligent statements: recovery possible when there is voluntary assumption of responsibility and reasonable reliance
Professional Responsibility
(2) Professional responsibility: recovery possible for negligent performance
(3) Negligent acts causing pure economic loss: recoverable under the Spandeck test
20.3.11 The position with respect to negligent acts which cause pure economic loss varies from jurisdiction to jurisdiction. In Singapore, Australia and other jurisdictions, it is – in some circumstances – possible to sue for pure economic loss caused by negligent acts. However, English law still takes an extremely restrictive approach to such claims.
(4) More restrictive approach adopted when imposing a duty of care for pure economic loss in comparison to that applied in cases of physical damage
20.3.14 Where a claimant suffers pure economic loss through damage not to his own property but to property belonging to someone else, actions have been allowed in Australia, again based on the specific economic vulnerability to which the defendant’s negligence has exposed the claimant: Perre & ors v Apand Pty Ltd (1999) 198 CLR 180. However, English law does not allow such actions in any circumstances: Candlewood Navigation v Mitsui OSK Lines [1986] AC 1.
C. No general duty for pure omissions
(1) Reasons for not imposing duty with respect to pure omissions
20.3.15 Generally, no duty is imposed with respect to pure omissions – ie, situations in which a defendant who has created no danger to the claimant merely fails to prevent him from sustaining harm. There are a number of reasons for this. One is the large number of potential defendants in situations of failure to act. Another is society’s focus on the more modest aim of discouraging wrongdoing rather than on the more ambitious one of encouraging good deeds. For these and other reasons, there is, for example, ordinarily no duty to rescue – even when such an act could be carried out without personal risk.
(2) Situations where there is a duty to act to prevent harm
20.3.16 However, there will be a duty to act to prevent harm in certain situations, eg:
    • where the defendant and the claimant are in a special relationship of dependence (such as guardian/child, carrier/passenger, employer/employee);
  • where the defendant has assumed responsibility for the claimant or his property.
F. Statutory authorities owe a duty only in restricted circumstances
(1) Duty of care unlikely to arise where conduct involves policy, discretionary elements or the balancing of resources
(2) Developments in the United Kingdom
20.3.19 A number of cases relate to the duty owed by the emergency services in the exercise of their statutory functions. The majority of these cases concern the police, whose duty to protect the public at large does not extend to a duty to protect individual members of the public during the conduct of an investigation: Hill v Chief Constable of West Yorkshire Police [1989] AC 53, or even in response to an emergency call: Michael v Chief Constable of South Wales Police [2015] UKSC 2. The rationale for this seemingly harsh approach is that a private law duty by the police to protect individuals from criminal acts committed by third parties would not only be difficult to confine within rational parameters, but would also be contrary to the ordinary principles of common law. The fire services owe no duty to individual members of the public either, even when they have undertaken to deal with a fire, unless they actually make matters worse through their positive intervention: Capital and Counties v Hampshire County Council [1997] QB 1004. However, the ambulance services have been held to owe a duty of care to individual members of the public whom they have undertaken to assist: Kent v London Ambulance Services [1999] Lloyd’s Rep Med 58.
20.4.1 Before a court can determine whether the defendant has breached his duty to the claimant, it is first necessary to establish the standard of care to which he will be held.
The Standard of Care
A. Establishing the due standard of care: whether reasonable care has been taken to avoid reasonably foreseeable harm
20.4.2 The basic question in every case is whether reasonable care has been taken to avoid reasonably foreseeable harm: Government of Malaysia v Jumal b Mahmud [1977] 2 MLJ 103. Factors which are relevant in this determination include:
  • the likelihood or probability of the risk eventuating;
  • the seriousness or gravity of the foreseeable risk;
  • the practicability of avoiding or minimising the risk;
  • the justifiability of taking the risk;
  • the time for assessing the risk;
  • the relevant characteristics of the foreseeable plaintiff
B. Special standards of care
(1) Standard of care not normally lowered to take account of a defendant’s inexperience
20.4.4 The standard of care is not normally lowered to take account of a defendant’s inexperience, since that would be unfair to those whom he injures: Nettleship v Weston [1971] 2 QB 581. For much the same reason, an amateur is judged according to objective standards of acceptability for the task in which he is engaged, not according to his personal level of expertise.
(2) Lower standard applied to children
(3) Professional negligence: standard of care is that of “the ordinary skilled man exercising and professing to have that special skill”
20.4.7 In the conduct of trades and professions, the law allows for a variety of levels of qualification, and thus a variety of standards, as long as the level of expertise which can be expected from any given professional is readily apparent from his particular qualification (eg, that he is a general practitioner rather than a specialist). However, every professional must achieve an acceptable level of basic competence: Ang Tiong Seng v Goh Huan Chir [1970] 2 MLJ 271.
Medical Negligence
(4) Medical negligence: different applications of the Bolam test in various jurisdictions
Proof of Breach
C. Proving breach of duty is a question of fact determined by specific circumstances of each case
20.4.12 Whether or not a duty has been breached is a question of fact to be determined according to the specific circumstances of each case. For this reason, precedents are of value only in terms of the general principles which they establish: Qualcast (Wolverhampton) Ltd v Haynes [1959] AC 743.
(1) Shifting of burden to the defendant: where exact cause of incident is unknown, defendant had control over the agent of harm, and that relevant damage would not normally have occurred in the absence of negligence
A. Causation: the physical link between the defendant’s negligence and the claimant’s damage
Simple Issues of Causation
B. The ‘but-for’ test: dealing with simple issues of causation
C. Multiple consecutive causes: involvement of either a second tortfeasor or a natural event
Multiple Consecutive Causes
20.5.5 However, when a tort is followed by a natural event which wipes out the physical effects of the tort, the tortfeasor’s liability ceases at the date when the supervening condition manifests itself: Jobling v Associated Dairies [1982] AC 794. If this were not so, the defendant would be liable for damage which would have occurred naturally anyway due to the ‘vicissitudes of life.’
D. Multiple potential causes: claimant can succeed only if he proves on the balance of probabilities that the damage is attributable to the tort
20.5.6 Where there are several discrete potential causes of harm, some of which are tortious and some of which are natural, the basic rule is that the claimant can succeed only if he proves on the balance of probabilities that the damage is attributable to the tortious conduct: Wilsher v Essex Area Health Authority [1988] AC 1074 (Wilsher).
Loss of a Chance
20.5.9 The standard requirement that in civil actions a claimant must establish his case on the balance of probabilities applies equally to actions based on loss of a chance. Under English law, if there is a less than 51% chance that the thing which might have happened would actually have happened had it not been for the defendant’s negligence, the claimant will fail, even if he seeks to recover not for the whole of his damage but only for the chance which the defendant caused him to lose. This analysis has been applied primarily in medical cases, where actions by claimants whose chances of recovery from illness or injury have been reduced due to the negligence of their doctors have failed when they could not establish that, with proper treatment, their chances of recovery would have exceeded 50%: Gregg v Scott [2005] UKHL 2; [2005] 2 WLR 268. Although in the past claims for loss of chance succeeded in medical negligence cases in some Australian states (see, eg, Rufo v Hosking [2004] NSWCA 391), the decision of the High Court of Australia in Tabet v Gett [2010] HCA 12 established that Australian law does not recognise the concept of loss of chance in medical negligence proceedings.
F. Loss of a right: recovery possible if it can be shown that defendant’s failure to advise claimant of risks inherent in treatment has deprived him of the right to choose a more experienced doctor or defer treatment
G. Breaking the chain of causation
(1) Defendant not liable for damage subsequently sustained by claimant’s own unreasonable response
20.5.14 Where a defendant has created a situation of danger which requires the claimant to take immediate averting action, the defendant will be liable even if, in the ‘agony of the moment,’ the claimant makes the wrong decision and suffers damage which could have been avoided had he acted differently.
20.5.15 Where the claimant’s act is the very thing against which the defendant is required to offer protection, the defendant will be liable for the consequences of his negligence, however objectively unreasonable the claimant’s act may be, although damages may be reduced to take account of the claimant’s contributory negligence: Reeves v Metropolitan Police Commissioner [2000] 1 AC 360.
(2) A new intervening act by a third party normally breaks chain of causation between the defendant’s negligence and the claimant’s damage
20.5.16 A new intervening act by a third party will normally break the chain of causation between the defendant’s negligence and the claimant’s damage. However, an act will not be regarded as ‘new’ if it is sufficiently connected with damage which has already resulted from the defendant’s negligence – eg, a subsequent accident after a road has been blocked due to the defendant’s negligence: Rouse v Squires [1973] QB 889, or medical negligence in the treatment of an injury caused by the defendant’s negligence: Webb v Barclays Bank plc and Portsmouth Hospitals NHS Trust [2001] EWCA Civ 1141. In such circumstances, the defendant may be held partly responsible for the subsequent damage, and the chain of causation will not be broken (although the subsequent tortfeasor will also be partly – and possibly even primarily – liable). Where the defendant has control over a third party, or where the third party is faced with a dilemma created by the defendant, the chain of causation is unlikely to be broken and the defendant will normally be liable to the claimant for the damage caused: Home Office v Dorset Yacht Co Ltd [1970] AC 1004.
(3) Liability imposed if defendant’s negligence makes it very likely that the third party will cause damage to the claimant
20.5.17 In other situations, a defendant will not be liable merely because his negligence makes damage to the claimant by a third party foreseeable. Liability will be imposed only if the defendant’s negligence makes it very likely that the third party will cause damage to the claimant: Lamb v Camden LBC [1981] QB 625.
A. Type of damage must be reasonably foreseeable
(1) Old approach to remoteness: defendant liable for damage directly resulting from negligence
(2) New approach to remoteness: defendant liable for type of damage foreseeable in all circumstances
The Egg-shell Skull Rule
20.6.4 In all tort actions, a defendant must take his victim as he finds him. Under the egg-shell skull rule, which applies to personal injuries, this concept is adapted to allow recovery even for unforeseeable damage. The egg-shell skull rule applies in circumstances where, due to a claimant’s innate physical susceptibility to illness or injury, he suffers extreme and unforeseeable damage which is triggered by the initially foreseeable damage caused by the defendant’s negligence: Smith v Leech Brain & Co Ltd [1962] 2 QB 405. When applied with respect to damage of an unforeseeable type (as opposed to merely an unforeseeable extent) the egg-shell skull rule operates as an exception to the Wagon Mound test.
B. The egg-shell skull rule: allowing recovery for unforeseeable damage
A. Illegality: where a claimant is himself a wrongdoer
(1) Illegality may also be considered at the duty stage or when determining appropriate standard of care
20.7.2 Where a claimant is himself a wrongdoer, his action may be defeated on the grounds of his illegality. Illegality is sometimes considered at the duty stage, particularly in Australia: Miller v Miller [2011] HCA 9, and has in the past also been regarded as relevant to determining the appropriate standard of care: Pitts v Hunt [1991] QB 24. In the UK, cases involving joint illegal enterprises – the main circumstance in which illegality is pleaded – are now determined by reference to principles of causation: Joyce v O’Brien & Anor [2013] EWCA Civ 546. However, notwithstanding the various stages at which illegality is considered, it continues to be described in common parlance as a ‘defence’.
(2) Defence normally succeeds only in cases where the claimant's conduct was criminal in nature
(3) Defence will not succeed where wrong is insufficiently connected with claimant’s damage, or where damage defendant causes is disproportionate to the claimant’s wrong
20.7.4 Where the claimant’s wrong is insufficiently connected with his damage, or where the damage which the defendant causes is disproportionate to the claimant’s wrong, an illegality plea is unlikely to succeed. In the latter situation, however, the defence of contributory negligence may be applicable: Revill v Newbery [1996].
B. Consent: where claimant either expressly or implicitly accepts the risk of harm associated with a defendant’s conduct
(1) Consent is a full defence pleaded successfully only in extreme situations
(2) Claimant must be shown to have had full knowledge and understanding of the risk involved and freely agreed to assume the very risk that materialized
20.7.6 For the defence to succeed, it must be shown that the claimant had full knowledge and understanding of the risk involved, that he freely agreed to assume that risk, and that the risk to which he consented was the one which materialized. Only rarely does the defence succeed in cases involving employees: ICI v Shatwell [1965] AC 656, and for policy reasons it is never available in actions brought by rescuers.
(3) Defence outlawed by legislation in cases of road traffic accidents
20.7.7 Where road traffic accidents are concerned, the defence has been outlawed by legislation: see the Motor Vehicles (Third Party Risks and Compensation) Act (Cap.189 2000 Rev Ed). It may be pleaded in claims involving other types of vehicles (such as light aircraft), but it succeeds only in cases where the risks are substantial and the claimant’s conduct has been particularly cavalier: Morris v Murray [1991] 2 QB 6.
(4) Consent and risks assumed when participating in sporting activities
(5) Unfair Contract Terms Act applies in a business context where defendant seeks to exclude liability for negligence by express agreement or by notice
20.7.9 In a business context, a defendant who seeks to exclude liability for negligence by an express agreement or by notice is subject to the provisions of the Unfair Contract Terms Act (Cap 396, 1994 Rev Ed). Under section 2(1) it is impossible to exclude liability for death or personal injury attributable to negligence, and under section 2(2) purported exclusion of liability for negligence in other situations must pass the test of reasonableness.
C. Contributory negligence: where claimant suffers damage as a result partly of his own fault and partly of the fault of another or others
(1) Partial defence where damages are reduced to reflect claimant’s share of responsibility for harm sustained.
20.7.10 Historically, contributory negligence was a complete defence, but under the Contributory Negligence and Personal Injuries Act (Cap 54, 2002 Rev Ed) a claimant who suffers damage as a result partly of his own fault and partly of the fault of another or others no longer has his claim defeated. Instead, his damages are reduced to reflect his share of the responsibility for the harm which he has sustained.
(2) Contributory negligence established where claimant fails to take reasonable care of himself according to the standards of the reasonable person
20.7.11 The standard of care is objective, so a claimant will be contributorily negligent if he fails to take reasonable care of himself according to the standards of the reasonable person. However, as with actions in the tort of negligence proper, lower standards of care apply in some situations, and less is expected in terms of self-preservation on the part of child claimants: Gough v Thorne [1966] 1 WLR 1387 and by claimants faced with situations of emergency: Jones v Boyce (1816) 1 Stark 493. Where a claimant has not himself been negligent, his damages will not normally be reduced merely because others have failed to take adequate care of him. One exception is where the claimant is an employer suing a defendant for damage caused to his property at a time when that property was in the care and control of his employee. In such circumstances, damages will be reduced to reflect any contributory negligence on the part of the employee.
(3) Claimant’s negligence results in reduction of damages only where it is causally relevant to the damage sustained
(4) Damages reduced by one half where the claimant and the defendant are equally to blame
Updated as at 1 June 2016
By: Margaret Fordham


Associate Professor