Explain and define the concept of "duty of care". Can it adequately distinguish situations which should give rise to liability from those which should not? Lord Atkins in the case of Donoghue v. Stevenson1 was the first to extract and apply a general duty of care ion in the tort of Negligence because this was a general duty of care, a duty that could exist without any pre-existing contractual relationship.
Therefore, the common law since the initial landmark case of Donoghue has attempted to form a principle, in what would be too wide a general principle, that can distinguish between situations that do and those that do not give rise to a duty of care. It is doubtless that there are many cases there has been a development of a coherent principle that can adequately identify situations in which a duty of care has arisen.
However, it is questionable whether the common law has always given adequately clear lines of principle, even more contentious is whether the common law is actually capable of extracting and forming a coherent and lineal principle that can be applied to a wide range of diverse circumstances. A duty of care is concerned with a party who by breach of a standard of care that should have been adopted in those circumstances has caused damage to an innocent party. The innocent party who may or not may have a pre-existing contractual relationship will be able to form an action for compensation.
However, an unregulated duty would be too far reaching and would lead the courts being inundated by frivolous claims. Therefore, the courts have assembled a tri-partite test that according to Michael Jones are "formal requirements"2 of a duty of care. Lord Keith confirmed this tri-partite test in Yuen Kun-yeu v A-G of Hong Kong3. Currently, a duty of care arises when the damage that is caused is reasonably foreseeable, there is not too distant relationship between the parties and that it must be "just and reasonable" to impose a duty of care.
The first limb of tri-partite test only allows a duty of care to arise between a party when it would be reasonable for the accused party to have foreseen the damage and thus taken precautions against it. The requirement that the damage that is done must be foreseeable arises from the liberal view that a party should not be held liable for injuries caused, both financial and physical, unless the party could have been reasonably expected to foresee the risk and has therefore breached the duty of care intentionally, recklessly or negligently.
The second prerequisite to a duty of care is that there should be a "sufficient relationship of proximity"4 because this ensures that the accused party owes a duty to those who in those circumstance he had a pre-existing relationship and therefore would or should have been aware he owed a duty of care to that party. The last part of the test prevents the judges from being forced to make a decision that may be particularly unjust to one party.
The first two limbs of the test are interdependent and similar because if there is a relationship of proximity it is more foreseeable that that the activity will cause damage to that party. These three tests must be satisfied before a duty of care arises. The test of proximity or as Lord Atkins called the "neighbourhood" principle is an important test that give guidance when drawing a boundary between circumstances that would and would not lead to liability. This is best illustrated in reference to the example of a public authority's duty of care to the public.
In cases involving public authorities the law has often been reluctant to impose a duty and this partly due to the fact that a public authority owes a duty to the public and not the individual and therefore there is no proximate relationship between the public authority and the individual. In Hills v. Chief Constable of West Yorkshire (1988)5 it was decided that the police did not owe a duty to the plaintiff to apprehend a murder because the police owed the duty to the general public and not exclusively to the plaintiff.
On the other hand, in Swinney v. Constable of North Bria Police6 the police were negligent in protecting the information given to them by claimant. In this situation it was decided that a specific relationship had arisen between the police and that specific individual who provided the information because he had done so on the assumption that he's confidentiality would be protected. This principle was again enforced in Kent v Griffiths7 where an ambulance service owed a duty of care to an individual because a proximate relationship arose between the ambulance service and the named individual.
An ambulance has an exclusive relationship with the named individual because the ambulance only acts in the interest of that individual, on the other hand, the fire service owe a duty to the individual but also to the public to prevent the fire spreading and therefore owe no duty of care. Obviously the test of "proximity" of relationship cannot be a dogmatic principle because a flexible test is needed if it is to be applied to a variety of different circumstances. As Martin comments: "There is a duty of care where there is proximity, and proximity means that the facts give rise to a duty of care"8.
It would be asking too much of the principle of proximity to clearly demarcate between liability and non-liability cases however this test helps the judge distinguish between cases where situations give rise to liability. When taking into consideration the interdependence and similarity between the first two limbs of the test it is unsurprising to discover that the test of remoteness of damage gives similar guidance to the guidance of the first element of the test.
The foreseeability of damage, like the proximity test, must be applied to different circumstances and as a result it is unable to be a rigid test that strictly ensures a coherent line of principle. Despite this, the remoteness of damage is still helpful in creating a coherent principle and probably more so than the proximity of relationship test. This is because whereas the Proximity of relationship test is a complete variant and changes in every circumstance, the foreseeability of damage is an objective test and therefore has a constant element.
Foreseeability of damage is based on whether the reasonable person would have had knowledge of the risk. In Smith v Little woods Ltd 19879 the defendant was unaware that a third party had on recurring events, broke into an unoccupied building that belonged to the defendant and lit a fire. Therefore, the defendant was not expected to know that there was a risk that the vandals would again break in and light a fire that would also damage the claimant's property.
When this is compared with the case of Holiamn v United Grain Growers Ltd10 where the defendant was aware and in control of dangerous chemicals that were interfered with by a third party, the defendant was found liable because it was reasonable to expect the defendant to be aware that there was a risk of third party intervention but had neglected to secure the chemicals. This test of foreseeability of damage gives more guidance in when deciding what circumstances lead to a duty of care because although there is a large element of personal judgement it is also a factual question of knowledge.
Whereas the first two elements of the tripartite test help identify circumstances of when a duty should arise, the third test undermines the strength and direction of the first two tests. The combination of wide unfettered judicial discretion and the most emphasis being placed on the third part of the test lead to a result that almost invites judges to make decisions on an ad hoc basis. The first two elements of the test can only lead to liability if the court considers it a fair and just judgement therefore the third sub-test of the tripartite test can decide where to impose liability.
In addition, this test is too ambiguous to ensure that a clear and coherent principle develops. As a result, decisions have been made based on a wide range of issues but most notable of all these is that of policy. In Hills the court decided that there was no proximity of relationship between the police and the public but in the course of that judgement the court said that it could not impose a liability upon the police because this was divert resources away from preventing future crimes into that of defending litigation. Therefore, because of a policy consideration the last test of fairness was not satisfied.
This use of policy consideration can be more bluntly seen in the case of Dorset Yacht v Home Office (1970)11 where Lord Denning commented that the decision to impose liability was "at bottom a matter of public policy which we as judges must resolve". If policy is to be a dominant consideration it must be decided on an ad hoc basis because a decision of policy cannot be applied to all different factual circumstances and relationship. Therefore, policy would throw into disarray any coherent development of the principle of liability for duty of care.
This can be particularly seen in the Anns case, which was overruled by the case of X v. Bedfordshire CC (1995)12. The latter case overruled the former because it limited the jurisdiction of the judges by saying that the courts should not impose a duty of care upon public authority if it was acting within the discretion of legislation. Then in 1999 Barret v London Borough of Enfield13reaffirmed the decision of the Anns case by allowing common low to impose a duty of care on a public authority that was acting within its legislative discretion, as long as its action were completely unreasonable.
It is now unclear what is the boundary of the courts jurisdiction to impose a duty of care upon a public authority. Therefore, because the third test gives enough judicial discretion to allow policy consideration and neither force nor encourages the development of common law principle, circumstances that lead to liability cannot be confidently identified. The inadequacy of the third element of the test suggests that perhaps it is the weakness of the structure of the tripartite test that does not ensure that a coherent boundary of liability can be identified.
However, in reality the problem is one that challenges the very essence of the duty of care. This is because a wide range of diverse circumstances and relationships fall within the ambit of a general duty of care and as a result it may be impossible to form clear lines of principle that adequately locate circumstances that lead to liability and can be applied to all circumstances. The three-limbed test does give some guidance as to what circumstances liability should be imposed, however, in reality the ambiguity of the three tests gives the court considerable scope for policy manipulation under the guise of legal principle.
As Lord Roskill commented in Caparo v. Dickman14 the words of the tripartite test are simply "labels or phrases descriptive of very different factual situations". Lord Lloyd expressed in alarm, this over reliance on a test that is ultimately inadequate is causing the law of negligence to disintegrate "into a series of isolated decisions without any coherent principle". 15 Therefore, rather than using the legal pre-text for justifying decisions that are ultimately based on policy perhaps the court should reject the whole general duty of care concept as initiated in Donoghue.
Perhaps, the court should realise that a legal test cannot be based on the unstable foundations of such wide and differing circumstances. Maybe the court should retreat from the ostensible landmark of Donoghue, accept the complexity of different circumstances and decide on case-to-case basis what relationships give rise to liability rather than using an inadequate all encompassing test to form a confusing an incoherent principle.
Whereas some cases would inevitably still be decided on issues of policy, whether this is a positive or negative feature is another debate, this would have the advantage of not having a legal principle shifting and compensating for a practical policy of specific circumstances and then this principle being applied to circumstances of a completely different nature.
The test of "remoteness of damage" and "proximity" can give guidance in identifying circumstances that give rise to liability, however, the cumulative effect of the first two limbs of the tests are insufficient to ensure that a coherent concept of duty of care develops. This is because the third limb gives the Judge such a potent opportunity, if not invitation, to overcome any principle that is derived from the first two limbs of the test.
In this way any logical and coherent line that is derived from the circumstances is often thrown into chaos by the guidance of policy. However, even this assessment ignores the inherent weakness of the concept of a general duty of care. The inclusion of the third limb of the test that protects against "unjust results" is an admission that a general principle that will be rigid enough to give guidance and coherent development of the law cannot be applied to the diverse circumstances of real life without giving unjust results.
Therefore, perhaps the landmark decision of Donoghue v Stevenson was too crude a test that does not appreciate the diverse nature of circumstances. Perhaps, the tripartite test should be abandoned and rather than pursue the futile attempt to form a disjointed general principle that can be applied to all circumstances, the law on negligence should accept the need to form individual principles for different relationships.