INTRODUCTION
The courts of England and Wales have long since grappled with how to approach duty of care in tort law. The search for a single test for establishing a duty of care, however, is hopeless due to the nature of negligence and society. A general test for a duty of care between a claimant and a defendant has existed since the landmark case of Donoghue v Stevenson, alongside breach, causation and remoteness, for a successful negligence claim. Unlike the other requirements, however, duty is profoundly exposed to changing social norms. This blog post will therefore narrow its scope to a socio-legal lens to demonstrate that a single test for duty is unsustainable in three ways. It will first be shown that a single test is impossible due to the nature of negligence and society. Secondly, how this is reflected in the judiciary’s failure to carve out a single test will be considered. Lastly, it will be shown how the courts’ current approach to duty, which is not contingent on a single test, is most suitable to this area of law. Many scholars and judges have accepted that the judiciary has been unable to define a single test to date, however few have conceded that an adequate solution has emerged from the courts’ examination of the duty of care to date (see, for example, Lord Lloyd-Jones’ comments at [15] in Darnley v Croydon Health Services NHS Trust). This analysis will consequently conclude that the current judicial approach to duty is the most suitable way of deciphering whether or not a duty of care is owed.
I. THE IMPOSSIBILITY OF A SINGLE TEST
Firstly, establishing a single test for whether a duty of care exists is impossible due to the nature of both negligence and society. Unlike most other torts, negligence does not protect a particular interest and significantly overlaps with other areas of law. This means that an infinite number of possible causes of action exist in negligence, all of which a single, “special” test could not accommodate without infringing how other interests are protected, as noted by Lord Phillips in Gregg v Scott. Establishing a single test is also impossible due to the evolving nature of society. This is significant as duties specify requirements as to the way in which people engage in conduct by capturing social “sentiment”; an idea that is largely uncontested by corrective justice and civil recourse theorists such as Cane, Goldberg and Zipurksy. However not only is there an infinite variety of relationships between individuals, but the norms of such relationships are constantly evolving. This makes it impossible to generate a single, perennial test for duty which will apply to “every situation”, as envisioned by Lord Atkin in Donoghue v Stevenson.
The opinion to the contrary, however, merits consideration. Buckland, for example, contends that the concept of duty is redundant and should be abandoned. Such a view is compelling through the “fault-based” lens of negligence liability espoused in Fairchild v Glenhaven Funeral Services Ltd, which maintains that an individual only has a duty to pay damages to those upon whom they inflict loss. On this basis, duty is unnecessary as it does little to prove fault. However Buckland overlooks the fact that the courts determine liability by considering what “reasonable” steps could have been taken to prevent a harm within the duty defined. The “wrong-based” view of liability, advocated by Goldberg and Zipursky, is consequently more compelling as by leaning in a deontological direction, it recognises duty as ex ante as is custom in the courts. As duty remains integral to negligence, the tort will continue to be exposed to societal change. As a result, it is clear that the nature of negligence and society means that establishing a single test for duty is impossible.
II. THE COURTS’ FAILURE TO ESTABLISH A SINGLE TEST
Secondly, the impossibility of a single test for duty is reflected in the judiciary’s failure to establish one. The most commonly applied gateway to duty has been the Caparo Industries plc v Dickman tripartite test; which requires the foreseeability of harm, sufficient proximity of relationship, and that it is fair, just and reasonable to impose a duty. Although other tests have included owing a duty where one’s act or omission may foreseeably injure a neighbour, where an individual assumes a responsibility to another, and within the scope of the neighbour principle provided no opposing policy consideration exists (see, for example, Donoghue v Stevenson, Hedley Byrne & Co Ltd v Heller & Partners Ltd, and Anns v Merton LBC). However, as pointed out by Nolan, these tests have proven unable to accommodate “new” types of harm such as psychiatric illness and economic loss, for which the courts have been forced to develop new categories. Each will be dealt with in turn.
Firstly, due to difficulties associated with the foreseeability of psychiatric harm given its idiosyncratic nature, the courts have forged separate tests for primary and secondary victims of such harm. These tests are, however, fail to provide a suitable response to social complexity. For example, the test for secondary victims, who witness but do not participate in a given event, requires “close ties of love and affection” between the claimant and the injured victim following Alcock v Chief Constable of South Yorkshire. Yet, the Law Commission has been quick to highlight how this does not capture modern society in which the family unit typically includes grandparents and cohabitants, who are generally considered not to satisfy this test. Similar criticism can be directed at the Alcock requirement that the claimant viewed the event with their “own unaided senses”. Whilst the case of Alcock concerned when liability might arise for incidents that people might apprehend through the medium of television, the content of which was restrained by the Broadcasting Code, the modern spectator can view events as though present through a miscellany of social media platforms untethered by such codes (see the Broadcasting Act 1990). Such considerations cast doubt over this category, particularly as new psychiatric illnesses emerge such as occupational stress and nomophobia.
Secondly, the courts have developed a category for economic loss. This is again due to difficulties in relation to the foreseeability requirement, given that financial loss is often a matter of “pure chance”, as noted by Harvey. In cases concerning misstatements and services which cause financial loss not consequent on physical damage, duty has come to hinge instead on whether a responsibility has been assumed via the ‘Hedley Byrne v Heller’ exception. The fact that the judges in Hedley Byrne disagreed over whether this could be applied alongside tests of foreseeability ought to have foreshadowed the problems this would cause. Since the decision, a dichotomy has arisen in cases of pure economic loss: with the foreseeability principle being ignored in some cases yet applied generously in others (see, for example, Smith v Eric S Bush, Spring v Guardian Assurance plc, Williams v Natural Life Health Foods Ltd, and Commissioners of Customs and Excise v Barclays Bank plc). This approach has come at the expense of long established doctrines such as that which prohibits liability for omissions and acts of third parties, and contractual privity and consideration; the significance of all of which has been diminished.
The courts, however, have justified these separate categories on public policy grounds under the ‘Anns’ and ‘Caparo’ tests: arguably to prevent what Lord Steyn has described as a litany of cases that may otherwise arise under a general test of foreseeability. This is compelling when the social capriciousness of psychiatric illness and economic loss is considered; namely, that cases involving these elements often involve a set of unique and socially complex facts which could give rise to any number of possible lines of liability. However, as noted by Nolan, regardless of the plausibility of the courts’ approach within the categories, it remains that the very existence of these categories shows a dissonance in the judicial approach to duty. These new categories, therefore, are inherently problematic and serve to highlight the courts’ failure to supply a single test.
III. THE SUITABILITY OF THE CURRENT APPROACH
This post’s final contention is that the current approach adopted by the courts is most suitable to this area of law, set out in Robinson v Chief Constable v West Yorkshire by Lord Reed. This approach maintains that duties of care should be developed incrementally by “analogy” with established principles, unless in genuinely “novel” cases in which the Caparo test should be deployed. The merits of this approach are best seen through a pluralist lens, which in rejecting the monistic approach of corrective justice scholars such as Weinrib who seek to explain tort by reference to a single ideal (corrective justice), endorses the balancing of principle and policy by reference to a range of countervailing concerns. Each element will be dealt with in turn. Firstly, the current approach alleviates the need for new categories of duty by establishing the “universal touchstone” of principles envisioned by Lord Lloyd-Jones, rather than a single test, from which the law can be developed coherently in line with modern losses and harms. For example, the fruits of the late twentieth century expansion of duty, largely justified by policy reasoning, are already being reversed: with the exceptions to the rule against omissions being retightened and any immunity for public authorities renounced. Secondly, this has been balanced with the restricting of policy reasoning to cases which are genuinely novel. This “bottom-up” rather than “top-down” approach serves to restrain the judiciary from deploying policy rationale too readily, as shown in ABC v St George’s Healthcare NHS Trust.
Functionalist justificatory models (which aim, generally, to provide torts with a normative grounding by defending the goals which they seek to achieve) may contend that the current approach could lead to the erosion of the compensatory aims of tort law by lending too much discretion to the courts. This is perhaps evident in Darnley v Croydon Health Services; where Lord Lloyd-Jones, who is associated with a generous approach to duty, found a duty to be present whilst Sales LJ, who has contrastingly denounced the “weaker” approach to duty extra-judicially, found the contrary in the Court of Appeal. This same concern can directed at the retention of public policy use, particularly as the courts have failed to define what constitutes a “novel” case; a point well covered by Morgan. However, both these contentions oversimplify the area. Firstly, as noted by Lord Lloyd-Jones in Darnley, any uncomfortable analogies drawn at the duty stage would be accommodated in the standard considered at the breach stage. Secondly, it is unlikely that the courts will be unable to find an applicable established principle and therefore resort to Caparo. This is evidenced by the novel James-Bowen v Commissioner of Police case, in which the policy justification of divided loyalty was indirectly harnessed from the established principle that an important duty to one group precludes a cross-cutting duty to others. As a result, the current approach can be deemed to have successfully balanced principle and policy to provide a viable solution to establishing a duty of care.
CONCLUSION
In conclusion, the search for a single test to establish a duty of care is a hopeless one. This has been shown by highlighting how defining a single test is impossible, considering how the courts have failed in doing so, and by demonstrating why the current approach to duty, which accommodates socio-legal change by balancing principle and policy rather than via a single test, is most credible. In this regard, the courts’ journey in seeking to find a single test has served to shape and mould judicial precedent to create an adequate solution after all; even if not the one initially envisioned. Going forward, however, it is vital that this relatively new solution is continuously scrutinised as new cases rise through the court system and the socio-legal landscape continues, inevitably, to evolve.