I. Introduction
In the swirling seas of the modern law of negligence, there seems to be only one fixed point, an apparently invulnerable rock lying right in the eye of the storm. The objective standard of the reasonable person – the yardstick against which the defendant's conduct has been measured for over 150 yearsFootnote 1 – has been left largely unscathed by the controversies which have engulfed the other elements of the cause of action for negligence. By and large, the disputes about the appropriate boundaries of negligence liability have been played out elsewhere. There is evidence, however, that the force of the waves is beginning to chip away at the edges of the reasonable person standard, and it seems likely that the process of erosion will accelerate in the future. For while the core standard is not itself under threat, there is discernible movement towards the use of standards more favourable to defendants in certain peripheral types of case. In English law, these moves have in the main been rather tentative, and some have been rebuffed. However, legislatures in other jurisdictions have taken more decisive measures along these lines, and there is every prospect that at some point ours will follow suit.
The most high-profile recent example of an assault on the objective standard of reasonable care in English law has been in the public authority liability context.Footnote 2 In the not-so-distant past, public authorities' liability for negligence in the exercise of discretionary statutory duties and powers was conditional on proof that the conduct in question was irrational (or Wednesbury unreasonableFootnote 3), and while that restriction has now been abandoned, certain judges have more recently advocated a move away from duty of care restrictions in such cases in favour of tracing “the contours of liability” by the use of a “modulated” standard of care.Footnote 4 Another front of the attack on the ordinary standard of care in this context was opened by the Law Commission, which suggested in its consultation paper on Administrative Redress that where a public body was engaged in a “truly public” activity, its liability in negligence should be governed not by the standard of reasonable care, but by a “serious fault” test which would require the claimant to establish that the authority's conduct fell “far below” the standard to be expected of it in the circumstances.Footnote 5 The Commission's proposals were heavily criticised, and have now been abandoned,Footnote 6 but the centrality of the test of serious fault to those proposals,Footnote 7 coupled with the fact that the Western Australian legislature has replaced the standard of reasonable care with a standard of Wednesbury unreasonableness for claims based on policy decisions taken in the exercise of public functions,Footnote 8 suggests that the possibility of Parliament altering the standard of care in public authority negligence cases cannot be ruled out.
Another context in which Parliament might choose to lower the standard of care is voluntary emergency assistance. In many jurisdictions, the concern that potential good Samaritans might refrain from helping others for fear of liability has led to the replacement of the standard of reasonable care with a lesser standard of gross negligence or recklessness where one person causes harm to another while seeking to protect that other from an imminent danger.Footnote 9 German law, for example, employs a standard of gross negligence in such cases,Footnote 10 and in the United States,Footnote 11 CanadaFootnote 12 and Australia,Footnote 13 there is now extensive legislative protection of good Samaritans along these lines, albeit sometimes limited to off-duty medical personnel providing emergency medical help. Furthermore, the Irish Law Reform Commission recently recommended a gross negligence standard for individual good Samaritans and volunteers in its report on the Civil Liability of Good Samaritans and Volunteers.Footnote 14 According to the Commission:
[T]he imposition of a gross negligence test succeeds in striking a balance between the policy of encouraging altruistic behaviours [and] the public's right to redress. With regard to encouraging altruistic behaviour, the leniency of the gross negligence test may be understood as a reward for good behaviour. Furthermore, it militates against the deterrent effect that the fear of litigation may cause … The application of the ordinary negligence test, on the other hand, would be to impose too heavy a burden that would threaten the continuation of such benevolent activities.Footnote 15
In England, the use of a modified standard of care to protect good Samaritans was rejected when a claim was brought against the St John's Ambulance Brigade for alleged failures in the treatment of a schoolboy injured in a motor cycle race,Footnote 16 but the possibility of legislation to clarify the law on the issue was mooted in Lord Young's report for the current Government on health and safety and the compensation culture,Footnote 17 and it is conceivable that such legislation would go beyond clarification of the existing position and limit liability to instances of gross negligence or recklessness.
With these and other scenarios in mind, this article is an exploration of the variation of the standard of care in negligence to favour defendants, an issue of considerable practical significance which has not previously been the subject of systematic analysis. By shining a spotlight on this issue I hope to show that varying the standard of care in this way is a useful technique, which is and could be used in a number of types of case to achieve an appropriate balance between liability and non-liability. I also hope to show that if this technique is employed there are some ways of varying the standard of care which are preferable to others. The structure of my analysis is centred around three core questions. First, to what extent has English negligence law already varied the standard of care to favour defendants? Secondly, if the standard of care is to be varied, how should this be done? And thirdly, when and why might the use of a modified standard of care be desirable?
Before we turn to the first of those questions, however, the scope of the analysis should be clarified. My concern is with any deviation from the objective standard of reasonable care which could favour defendants, such as the use of a standard of care tailored to the defendant as an individual,Footnote 18 or the use of a gross negligence standard.Footnote 19 I do not consider cases where the objective standard of reasonable care is employed, but simply adapted to a particular class of defendants, as where the conduct of a child is assessed by reference to the standard of a reasonable child of the same age,Footnote 20 or that of a hospital doctor by reference to the standard of a reasonable person holding that particular post.Footnote 21 Nor do I consider the Bolam test, according to which a professional is not guilty of negligence if she acted in accordance with a practice accepted as proper by a responsible body of relevant professional opinion,Footnote 22 since (although the contrary is sometimes impliedFootnote 23) that test does not represent a departure from the ordinary standard of the reasonable person, but is merely an attempt to apply that standard in the light of the fact that judges lack the knowledge and expertise usually required to choose between competing professional opinions.Footnote 24
II. To What Extent Has English Negligence Law Already Varied the Standard of Care To Favour Defendants?
My first task is to identify the areas of negligence law where Parliament or the courts have deviated from the ordinary objective standard of the reasonable person in a way which favours defendants. This is not entirely straightforward, because it is sometimes difficult to tell whether the standard of care has been varied, or whether an apparently more lenient approach is in fact just the way that the ordinary standard of care is playing out in a particular context. After all, the standard of reasonable care is of necessity a very flexible one, since it must be applied in a wide variety of different circumstances. To drive at 60 miles per hour on a motorway will not generally be negligent, for example, but to do so in an urban high street almost always will be, not because the standard is varied for the motorway motorist, but simply because the conditions are so very different. Similarly, while a driver who jumps a red light or breaks the speed limit will generally be considered to be negligent, the same may not be true of an ambulance driver who takes these and other risks when responding to an emergency call, not because he is subject to a different standard of care,Footnote 25 but because the application of the ordinary standard of care takes into account the exceptional circumstances of the case.Footnote 26
While these examples are straightforward, other types of case are less so. Consider for instance the standard of care applied to those taking part in sport and related physical activities, such as “horseplay”. The nature of these activities is such that, when one participant injures another (or a spectator), it can be difficult to decide whether to attribute the injury to negligence, or to the inherent risks of the activity in question. The courts' response to this difficulty has typically been to stipulate that in such cases liability will not be imposed for a mere error of judgment or lapse of skill, but only where the defendant's conduct was such as to evince a reckless disregard of the claimant's safety.Footnote 27 Does this approach represent a varying of the standard of care, or should it instead be interpreted as the application of the ordinary standard in a somewhat unusual context? The better view is that in this type of case the standard of care has not been varied. There is, after all, no reason why it should be varied in this category of case.Footnote 28 Moreover, it is clear from the relevant authorities that the courts' own justification for this approach is the need to take account of the circumstances when applying the ordinary negligence standard. In the leading case of Wooldridge v Sumner, for example, Diplock L.J. said that “the law of negligence has always recognised that the standard of care which a reasonable man will exercise depends on the conditions under which the decision to avoid the act or omission relied on as negligence has to be taken”,Footnote 29 and described the test of “reckless disregard” as “the practical result of this analysis” in the sporting context.Footnote 30 Similarly, in Blake v Galloway, where the Court of Appeal expanded the reckless disregard test to include “a very high degree of carelessness”, Dyson L.J. said that the participants in horseplay owed each other a duty to take reasonable care not to cause injury, and made it clear that the issue was what that meant “in the context of play of this kind”.Footnote 31
But even if in the end we conclude that no modification of the standard of care has taken place in the sporting context, the example shows that discerning such modifications is not always an easy task. Bearing that caveat in mind, we can turn to three areas in which the standard of care may already have been varied. In two of these areas – the liability of occupiers and of public authorities – it seems clear that this has occurred. In the other – liability for the deliberate conduct of third parties – there are indications that a modified standard of care may have been used, but the evidence is less conclusive.
A. The Liability of Occupiers
There are two types of case where a varied standard of care has been applied to occupiers. The first relates to the obligations occupiers owe trespassers and other non-visitors. It used to be that an occupier was liable to a trespasser only for harm arising out of a reckless disregard for the latter's safety.Footnote 32 This restriction was abandoned in British Railways Board v Herrington,Footnote 33 but the House of Lords imposed a lesser duty of “common humanity” instead of the ordinary duty of reasonable care, and this lesser duty was then put on a statutory footing in the Occupiers' Liability Act 1984. The other type of case where a modified standard has been employed is where the occupier is sued for “continuing” a nuisance which emanates from his land but which he did not himself create, as in Leakey v National Trust,Footnote 34 where the Court of Appeal held that the defendants had to meet the cost of the measures required to prevent soil and rocks slipping onto neighbouring properties from a hill in their occupation.Footnote 35 In cases of this kind, the fact that the harm results from an omission rather than a positive act has been taken into account by making the standard of care partly subjective.
Precisely how has the standard of care been varied in these two types of case? The Leakey category will be dealt with first, since here the answer is more straightforward. Where an occupier is sued for continuing a nuisance, he will be liable only if, after he became – or should have become – aware of the danger the state of affairs posed to neighbouring property, he omitted to take reasonable steps to remove or reduce the threat.Footnote 36 So far, this is just the ordinary reasonableness approach, but there are two important caveats. First, the danger must be patent and not latent, so that liability may be imposed if the occupier saw or should have seen the problem on his land, but not if the threat to neighbouring property was discoverable only by further investigation.Footnote 37 And secondly, even when the occupier does have actual or presumed knowledge of the danger in this sense, his duty is to do only what it is reasonable to expect of him in his individual circumstances, so that if physical effort is required to avert the danger, account is taken of the occupier's age and physical condition,Footnote 38 and, if expenditure of money is required, of his means.Footnote 39
The precise nature of the duty owed by occupiers to trespassers and other non-visitors is more complex. Although no clear rule emerged from the speeches in Herrington, three general themes were discernible. The first was that the duty did not arise until the occupier had actual knowledge “either of the presence of the trespasser upon his land or of facts which make it likely that the trespasser will come on to his land” and of “facts as to the condition of his land or of activities carried out upon it” which were likely to cause personal injury to a trespasser unaware of the danger.Footnote 40 Moreover, the occupier was under no duty to make any inquiry or inspection to ascertain whether or not such facts did exist.Footnote 41 Secondly, if the duty did arise, it was subjective, so that the defendant's knowledge, skill and resources would be taken into account when assessing whether it had been discharged.Footnote 42 And finally, it was also said that the duty owed was a lesser one than the general duty of care, this lower level of obligation being described as a duty of “common” or “ordinary” humanity.Footnote 43 Overall it was clear that in the case of a trespasser an occupier was not subject (as Lord Reid later put it) to the “full obligations” arising out of a relationship of neighbourhood in the Donoghue v Stevenson sense,Footnote 44 since that relationship had been “forced on him against his will”.Footnote 45
Under the 1984 Act, an occupier owes a duty of reasonable care to a trespasser or other non-visitor in respect of a particular danger on his land if (a) he is aware of the danger, or has reasonable grounds to believe it exists; (b) he knows, or has reasonable grounds to believe, that the other is, or is likely to come into, the vicinity of the danger; and (c) the risk is one against which he may reasonably be expected to offer the other some protection.Footnote 46 Although it could be argued that this duty is simply an application of the ordinary standard of care in the peculiar circumstances of unauthorised entry, the better view is that here too the standard of care has been varied because the relationship from which the duty flows has been forced upon the occupier. For while the draft Bill in the Law Commission report which preceded the Act simply laid down the ordinary standard of reasonable care,Footnote 47 the Act itself requires that the occupier either knew or had reasonable grounds to believe that the danger existed and that the claimant was in, or was likely to come into, the vicinity of that danger. And since the phrase “reasonable grounds to believe” has been held to require actual knowledge of facts that would have led a reasonable person to the conclusion in question (so that it is not enough to show that a reasonable occupier would have been aware of the danger, and of the claimant's proximity to it),Footnote 48 it is clear that there is a subjective element to the standard laid down by the Act, which therefore represents a departure from the ordinary, objective, standard of care. Hence Lord Hoffmann's statement that the duty under the 1984 Act was “intended to be a lesser duty, as to both incidence and scope” than the duty to lawful visitors,Footnote 49 and Lord Phillips M.R.'s description of the former as “significantly less exacting” than the latter.Footnote 50
This survey of the law relating to the liability of occupiers reveals that at various times the standard of care has been varied in two different ways. First, the duty to take precautionary measures has not been assessed (as it normally would be) in the light of the risks that would be apparent to a reasonable person: in the case of trespassers, actual knowledge of the risk – or of facts from which a reasonable person would identify the risk – is required, and in the nuisance scenario, the danger must be patent and not latent. And secondly, the occupier's personal circumstances have been taken into account when deciding what steps he ought to have taken to reduce or eliminate the risk.
B. The Liability of Public Authorities
The second context in which the standard of care has been varied is the liability of public authorities. Formerly, it was held that a negligence action could succeed only if the decision of the public authority under attack was Wednesbury unreasonable – that is, if it was so unreasonable that no reasonable authority could have made it. Latterly, the courts have moved away from this express dependence on public law concepts towards a private law approach, under which the problems surrounding discretionary decision-making are accommodated at the breach of duty stage of the negligence enquiry. Although on the former approach the Wednesbury unreasonableness hurdle appeared at the duty stage, in effect it meant that public authorities exercising a statutory discretion were subject to a modified standard of care. On the latter approach, public authorities are notionally subject to the application of ordinary negligence principles, but there is evidence to suggest that in practice a modified standard of care is still being applied.
Where Parliament has conferred a discretion on a public authority, the judiciary are naturally disinclined to tread on the authority's toes by imposing liability for conduct that falls within the ambit of that discretion. However, the precise mechanics by which the courts give effect to this inhibition have varied over the years. Previously, reliance was placed on public law concepts, in particular Wednesbury unreasonableness. According to Lord Reid in Home Office v Dorset Yacht Co. Ltd, a decision of a public authority could ground liability in negligence only if it was so careless or unreasonable that it did not amount to a real exercise of the statutory discretion at all.Footnote 51 Similarly, in X (minors) v Bedfordshire C.C.,Footnote 52 the House of Lords held that the first requirement in such cases was to show that the decision under attack was so unreasonable that it fell outside the ambit of the discretion altogether.Footnote 53 Since it is “significantly more difficult”Footnote 54 to establish that a public authority has fallen foul of the public law Wednesbury unreasonable standard than it is to show that it has fallen foul of the private law standard of negligence, the use of this concept amounted in effect to the use of a modified standard of care in such cases. This was confirmed by Lord Browne-Wilkinson in X (minors) when he said that it followed from the Wednesbury unreasonableness precondition that a claim against a local education authority in respect of a statement of special educational needs it made under the Education Act 1981 could not succeed merely because the relevant decisions were made carelessly, but that it would have to be shown that they were so “grossly delinquent” that no reasonable education authority could have reached the same decisions.Footnote 55
In Barrett v Enfield L.B.C.,Footnote 56 however, the House of Lords abandoned the Wednesbury unreasonableness requirement. According to Lord Hutton, provided the decision under attack was justiciable – in other words, suitable for scrutiny by a court in a private law action – it was preferable for the courts to resolve the liability issue by direct application of the negligence concept, rather than by applying Wednesbury unreasonableness as a preliminary test.Footnote 57 The abandonment of the Wednesbury requirement was confirmed in Phelps v London Borough of Hillingdon, where Lord Slynn said that the fact that acts claimed to be negligent were carried out within the ambit of a statutory discretion was not in itself a reason why it should be held that no claim for negligence could be brought in respect of them.Footnote 58 It did not however follow that the discretionary nature of a public authority's decision-making was to be ignored altogether, for it would instead be taken into account at the breach of duty stage. Hence, in Barrett, where a local authority was being sued for alleged defects in its foster care provision, Lord Hutton said that the standard of care expected of the defendant would have to be determined against the background that it was “given discretions to exercise by statute in a sphere involving difficult decisions in relation to the welfare of children.”Footnote 59
Another development which occurred in Barratt and Phelps concerned an argument which had previously proved influential when it came to deciding whether it would be fair, just and reasonable to impose a duty of care on a public authority. In X(minors), one of the reasons which Lord Browne-Wilkinson gave for his conclusion that no duty of care was owed by local authorities in respect of their child protection functions was that “the task of the local authority and its servants in dealing with children at risk” was “extraordinarily delicate”.Footnote 60 However, when this point was made by counsel for the defendants in Barrett, it met with a rather different response, for while Lord Slynn accepted that the provision of foster care involved the balancing of delicate and difficult factors, he clearly felt that, rather than ruling out liability altogether, this should instead be taken into account at the breach of duty stage.Footnote 61 Similarly, in Phelps, which concerned four different cases of alleged educational malpractice, Lord Clyde recognised that educational psychologists and other professionals had to make difficult decisions concerning diagnosis of dyslexia and the like, but considered that the Bolam test gave those involved sufficient protection.Footnote 62
Whether in the aftermath of these developments a modified standard of care is still being used in certain types of public authority negligence cases is open to question. What is clear is that the judiciary have emphasised that it will be difficult to establish breach of duty in this context. In Barrett, Lord Slynn said that in the child protection and foster care context “the courts should not be too ready to find … that there has been negligence by staff who largely are skilled and dedicated”,Footnote 63 and he made a similar statement about findings of negligence in education cases in Phelps,Footnote 64 where Lord Nicholls also remarked that claims for educational malpractice should be limited to instances of “manifest incompetence or negligence comprising specific, identifiable mistakes”.Footnote 65 These observations were echoed in subsequent Court of Appeal decisions. In S v Gloucestershire C.C., for example, May L.J. interpreted Barrett as laying down that “in considering whether a discretionary decision was negligent, the court will not substitute its view for that of the local authority upon whom the statute has placed the power to exercise the discretion, unless the discretionary decision was plainly wrong,”Footnote 66 while in Carty v Croydon L.B.C., Dyson L.J. said that in the light of the nature of the statutory function and the difficulty of the decisions involved, in the educational malpractice sphere the courts will usually only hold that there is a duty to avoid decisions that are “plainly and obviously wrong”.Footnote 67
Two points can be made. The first is that although some commentators have linked the leeway that is now given to discretionary decision-makers in public authority cases with the Bolam test of professional negligence,Footnote 68 the analogy is not convincing. Although an institutional competence problem similar to the one underlying the Bolam test may arise in public authority cases,Footnote 69 it is dealt with at the justiciability stage. It follows that by the time the breach issue falls to be considered, the court has already determined that it is capable of assessing the reasonableness of the decision under attack, and hence any special latitude or “margin of appreciation” that the authority is accorded must be based on considerations different from those which underlie the Bolam test. This lends support to the view that a modified standard of care is being applied in this context. And the second point is that although in this context the judiciary have refrained from laying down a special standard of care along the lines of the occupiers' liability standard, they do seem to be saying that a discretionary decision will be deemed negligent only if it was “manifestly” or “plainly” mistaken. And since it is argued below that a test of manifest or obvious fault may amount in effect to a gross negligence standard,Footnote 70 this also lends support to the conclusion that the breach test which has emerged from the case law on public authorities does represent a deviation – albeit a minor one – from the ordinary standard of care. This conclusion is consistent with the statements in the leading work on public authority negligence liability that in Barrett and Phelps the House of Lords set “a high hurdle for establishing negligence” in such cases, which was designed to catch only “particularly egregious failings”.Footnote 71
C. Liability for the Deliberate Acts of Third Parties
Finally, there is some evidence that a modified standard of care has previously been used in cases where the claimant seeks to hold the defendant liable for the consequences of the deliberate act of a third party, albeit that this approach is now on the wane. Cases of this kind have given rise to enormous conceptual difficulties, and the courts have resorted to three different elements of the negligence cause of action to deal with them: (1) factual duty (the question of whether damage to someone in the claimant's position was a reasonably foreseeable consequence of the defendant's negligence), or, where the defendant caused some initial damage to the claimant before the third party intervened, remoteness of damage; (2) notional duty (the question of whether there is a rule of law that either allows or bars recovery in this category of case); and (3) breach of duty. Although there is some heavyweight judicial support for the factual duty/remoteness approach,Footnote 72 it will not be discussed further, since it received no backing in either of the two most recent House of Lords decisions on the subject, Smith v Littlewoods Organisation Ltd Footnote 73 and Mitchell v Glasgow C.C.Footnote 74 That leaves notional duty and breach. Analysis in terms of notional duty was preferred by Lord Diplock in Home Office v Dorset Yacht Footnote 75 and by Lord Goff in Smith v Littlewoods, and is currently in the ascendant following the House of Lords' endorsement in Mitchell of Lord Goff's approach in Smith. However, the breach approach has also proven influential, and since it seems to involve the application of a modified standard of care, it is worth taking a closer look at it.
The key decision is Smith v Littlewoods.Footnote 76 The defenders purchased a disused cinema, intending to turn it into a supermarket. Before they got a chance to do so, vandals set it on fire, and the fire spread to the pursuer's premises nearby. Were the defenders liable for not having anticipated and guarded against the vandals' conduct? Lord Goff dealt with the problem by reference to the duty of care concept.Footnote 77 There was “no general duty at common law to prevent persons from harming others by their deliberate wrongdoing, however foreseeable such harm may be if the defender does not take steps to prevent it”,Footnote 78 and although there were a number of exceptions to this general no-liability rule, on the facts none of them applied. Both Lord Griffiths and Lord Brandon, on the other hand, held that the defenders had owed the pursuers a duty of care, but that they were not in breach of that duty, although the latter made the common mistake of framing the breach question in terms of a specific “duty” (as in “Was the motorist under a duty to slow down as she approached the junction?”).Footnote 79
For present purposes, however, the most important speech is that of Lord Mackay.Footnote 80 Although his reasoning is somewhat opaque, the better view is that in effect he agreed with Lords Griffiths and Brandon: there was a duty, but it was not breached.Footnote 81 However, while Lord Griffiths based the duty in Smith v Littlewoods on Goldman v Hargrave,Footnote 82 Lord Mackay did not limit it in this way. His opinion can therefore be interpreted as doing away with duty restrictions in third party cases, in favour of regulation by breach. Moreover, the following passage suggests that he may have been endorsing the employment in such cases of a modified standard of care:
[W]here the only possible source of the type of damage or injury which is in question is agency of a human being for whom the person against whom the claim is made has no responsibility, it may not be easy to find that as a reasonable person he was bound to anticipate that type of damage as a consequence of his act or omission. The more unpredictable the conduct in question, the less easy to affirm that any particular result from it is probable and in many circumstances the only way in which a judge could properly be persuaded to come to the conclusion that the result was not only possible but reasonably foreseeable as probable would be to convince him that, in the circumstances, it was highly likely … Unless the judge can be satisfied that the result of the human action is highly probable or very likely he may have to conclude that all that the reasonable man could say was that it was a mere possibility. Unless the needle that measures the probability of a particular result flowing from the conduct of a human agent is near the top of the scale it may be hard to conclude that it has risen sufficiently from the bottom to create the duty reasonably to foresee it.Footnote 83
Two points can be made. The first is that, although the language used here is redolent of the factual duty/remoteness approach favoured by Lord Reid in Dorset Yacht, in essence Lord Mackay's analysis seems to be centred on breach of duty. This is consistent with the fact that in the immediately preceding passage Lord Mackay cites extensively from Glasgow Corporation v Muir,Footnote 84 where the defenders were held not to have breached the duty of care they owed the pursuer because the risk of negligence by the third parties who were the immediate cause of the harm was not sufficiently foreseeable. Furthermore, in her careful analysis of Smith v Littlewoods, Elspeth Reid argues that Lord Mackay “drew upon substantially the same framework” as the Inner House of the Court of Session,Footnote 85 which she interprets as resting on breach, a duty of care being assumed.Footnote 86 And the second point is that, while Lord Mackay presented his analysis as an application of ordinary breach principles, in truth he was stacking the odds heavily against the claimant, since the degree of required foreseeability was set extremely high, and nothing was said (at this pointFootnote 87) about the other side of the equation, namely the cost of the precautions necessary to eliminate the risk. This goes against breach of duty orthodoxy, whereby even the running of a minor risk may be negligent if the burden of precautions is low.Footnote 88 Similarly, there is no justification in breach terms for Lord Griffiths' statement that for an occupier to have to take precautions against the criminal activity of third parties on their premises the circumstances “would surely have to be extreme indeed”Footnote 89 – what if, for example, the occupier knew that vandals were messing about in a warehouse full of inflammable materials, and that a simple padlock would keep them out?
It is therefore plausible to conclude that the majority in Smith v Littlewoods favoured the determination of third party cases by the application of a modified standard of care, rather than by reference to the duty restrictions favoured by Lord Goff, a conclusion fortified by two later first-instance decisions, in both of which a modified standard of care seems to have been applied to defendants who were sued for the consequences of deliberate third party conduct.Footnote 90 Whether there is much of a future for the use of a modified standard of care in this context is doubtful, however. In the most recent House of Lords decision on the topic, Mitchell v Glasgow C.C., the House clearly preferred Lord Goff's duty approach,Footnote 91 albeit that Lord Hope, who gave the leading speech, limited his approval of that approach to cases in which the alleged negligence consisted of a failure “to prevent the risk of harm being caused to the [pursuer] by the criminal act of a third party which they did not create and had not undertaken to avert”,Footnote 92 thereby leaving open the possibility that the standard of care approach may continue to be employed in cases where the defendant has created a source of danger exploited by a third party or assumed a responsibility towards the claimant to protect her from third party conduct. There was however little in the other speeches to suggest approval of such a qualification on the use of duty analysis.
III. If the Standard of Care is to be Varied, How Should this be Done?
There are a number of ways in which the ordinary standard of care can be varied in a way which could favour the defendant, and in this section of the article I evaluate the different possibilities. The primary focus will be on the two most obvious possibilities, subjectivising the standard of care and using a standard of gross negligence. Before we turn to these, however, it will be helpful to consider two alternative approaches which involve abandoning a negligence standard altogether.
A. Abandoning a Negligence Standard Altogether
There are at least two ways in which the scales can be tipped in favour of the defendant at the breach of duty stage which effectively involve abandoning the negligence standard altogether, namely the use of tests originating in public law, and the use of a recklessness or wilful misconduct standard.
1. Public law standards
We have seen that in the past the English courts employed the public law test of Wednesbury unreasonableness as a preliminary filter in cases involving alleged negligence in the exercise of a statutory discretion, and some recent Australian state legislation uses the same test.Footnote 93 There has also been considerable interest in the use of tests originating in the European law context to regulate the liability of public authorities in domestic law. It has been suggested, for example, that the test of a sufficiently serious breach used to determine state liability in European law could also be used to regulate the domestic liability of public authorities,Footnote 94 while the Law Commission has advocated the use in this context of a test of “serious fault” heavily influenced by the sufficiently serious breach approach.Footnote 95 Although strictly speaking public law tests of this kind do not lay down “standards of care” at all, in a general sense their effect is to modify the standard of care in favour of the defendant, as was acknowledged in the Australian report which led to the adoption of the Wednesbury standard in some state legislation:
The effect of the test [of Wednesbury unreasonableness] is to lower the standard of care. It … [gives] the defendant more leeway for choice in deciding how to exercise its functions than would the normal definition of negligence (in terms of reasonable care).Footnote 96
The use of public law tests of this kind is not however a helpful means of modifying the standard of care, since they have been developed to perform specifically public law functions and are ill-suited to the task of setting standards of conduct in the private law context. Many of the factors taken into account in determining whether a state's breach of an EU legal norm is “sufficiently serious” to justify the award of compensation would, for example, be irrelevant to the question of breach in a negligence case, relating as they do to matters such as the importance of the norm in question, its clarity and precision, the method chosen by the national authorities to implement the unlawful policy, and the steps taken by them after it became clear an infringement had occurred.Footnote 97 Indeed, so far removed are these factors from a negligence analysis that it has been questioned whether the test is really a test of fault at all.Footnote 98 Nor is this disparity surprising if it is borne in mind that the EU law test is a component of a member state liability regime the primary purpose of which has been described as ensuring “the uniform application of EU law throughout the Union”.Footnote 99
Similar observations can be made about the Law Commission's serious fault test, which the Commission acknowledged was greatly influenced by the sufficiently serious breach test.Footnote 100 Although the definition of serious fault as conduct falling “far below the standard expected in the circumstances”Footnote 101 reads as a simple standard of gross negligence,Footnote 102 some of the factors which the Commission argued should be taken into account when applying the test – such as the seriousness of the harm caused, and the extent to which senior officials were implicated in the failures in questionFootnote 103 – were again alien to orthodox private law approaches to the fault issue.Footnote 104 Here the difficulty was that the serious fault test was designed to straddle public and private law, since under the Commission's proposals it would have been used to determine the availability of damages not only in certain types of public authority negligence case, but in judicial review proceedings as well.Footnote 105
Finally, the use of the Wednesbury unreasonableness standard to protect public authorities from liability in recent Australian state legislation has been criticised by commentators on the grounds that it is a public law concept ill-suited to operating as a fault standard in negligence adjudication. According to Mark Aronson, for example, “transplanting Wednesbury into negligence soil” gives it a “wholly different operation”, since in its original public law context it “had nothing to say to decision-makers about being careful to avoid harming others”.Footnote 106 The central difficulty with the use of the Wednesbury approach is that it was developed as a test of the rationality of a decision, whereas the private law standard of negligence is concerned not with whether a decision of the defendant was rational, but with whether the defendant's conduct created unreasonable risks of harm to others. The English courts' abandonment of the use of public law concepts in negligence was therefore a welcome development, and it would be ill-advised of Parliament to follow the Australian example of employing them as a mechanism for modifying the standard of care in public authority negligence cases.
2. Recklessness or wilful misconduct
Another possibility which should be rejected is to replace the ordinary standard of reasonable care with a test of recklessness or wilful misconduct in certain contexts. Although recklessness plays an important role in English criminal law, the concept has only occasionally surfaced in our law of tort, examples being the tort of deceit, the old law on an occupier's duty towards a trespasser,Footnote 107 and the illegality limb of the tort of misfeasance in public office.Footnote 108 However, a “wilful, wanton or reckless” standard has been extensively employed in American tort law as a condition of awarding punitive damages or watering down the causation requirement; to extend duties of care to new situations (such as liability for the sale of alcohol to an intoxicated patron); and to exclude the defence of contributory negligence.Footnote 109 Similarly, legislation in a number of Australian states protects good Samaritans from liability if they acted in good faith and without recklessness.Footnote 110 There is though little merit in such an approach, since replacing a standard of care of whatever stripe with a test of recklessness or wilful misconduct substitutes for an assessment of a person's behaviour a much more difficult enquiry into the state of his mind,Footnote 111 thereby adding uncertainty and complexity to the law.
B. Subjectivising the Standard of Care
It is trite law that the standard of care in negligence is generally objective. There is, however, nothing inevitable about this objective approach. In Roman law, for example, reliance was placed on a standard of culpa levis in concreto, meaning the failure to use the care the particular defendant ordinarily employed in his own affairs.Footnote 112 Furthermore, we have seen that in English law the standard of care has been partly subjectivised when occupiers of land have been sued by trespassers, or in respect of nuisances they have not themselves created. This subjectivisation has manifested itself in two main ways: (1) assessing the duty to take precautions in the light of the occupier's actual – as opposed to constructive – knowledge of the risk (or of facts from which a reasonable person would identify the risk); and (2) assessing the steps the occupier ought to have taken to deal with the risk in the light of his personal circumstances, and in particular the resources on which he can draw.
The first of these methods of subjectivising the standard appears to be unobjectionable, though it will be more difficult to adduce evidence that the occupier was aware of the danger than to adduce evidence that he should have been.Footnote 113 Taking into account the occupier's personal circumstances is more controversial, however. As we have seen, this approach has already been abandoned in the context of an occupier's liability to trespassers, and it has also come under attack in the nuisance context as well. In Abbahall Ltd v Smee,Footnote 114 the roof of a mews property had fallen into disrepair, with the result that water was leaking into the claimant's ground-floor flat. Pursuant to a court order, the claimant entered the flat above, which was owned by the defendant, to carry out repairs. The Court of Appeal held that the defendant owed the claimant a duty under the Leakey principle to take reasonable steps to prevent the disrepair of the roof causing damage to the claimant's property, but that in a case like this – where the roof served equally to protect both premises – the cost of such remedial action should be shared equally. Significantly, however, Munby J. (who gave the only substantive judgment) also held that in a case of this type the relative means of the parties ought not to be taken into account in apportioning the cost of the work. The judge at first instance had limited the defendant's contribution to a quarter of the cost, on the ground that she was less well off than the defendant, but Munby J. felt that this was both odd and impractical: odd, because it would for example follow that the claimant's share would depend on whether the defendant was idle or hard-working; and impractical, because the respective liabilities would be difficult to assess and liable to change frequently. This latter objection to a subjective standard of care was also forcefully expressed by Lord Watson in response to the contention that in Scots law trustees were held only to a subjective standard of care. Such a rule, he said, would be “highly inconvenient in practice” because:
In every case where neglect of duty is imputed to a body of trustees it would necessitate an exhaustive inquiry into the private transactions of each individual member, the interest of the trustee being to shew that he was a stupid fellow, careless in money matters, and that of his opponents to prove that he was a man of superior intelligence and exceptional shrewdness.Footnote 115
The answer to these two objections is that the duty ought only to be subjectivised a little. In the Leakey case, for example, Megaw L.J. said that it was not contemplated that there would be “discovery of the defendant's bank account or any detailed examination of his financial resources”.Footnote 116 Although the defendant's means were to be considered if substantial expenditure were required, and his age and physical condition if physical effort were needed, this could only “be in the way of a broad, and not a detailed, assessment”, and if this caveat were borne in mind, any difficulties were likely to be “more theoretical than practical”.Footnote 117 Provided the degree of “subjectivisation” is limited in this way, then it seems that this method of modifying the standard of care can provide a useful and workable mechanism for limiting the burden on potential defendants who are subject to duties of affirmative action.
A more fundamental objection to the use of a subjective standard has been made by Richard W. Wright. According to Wright, the external exercise of freedom depends on sufficient security against interference by others with one's person and property. And the use of a subjective approach makes such security impossible, “since the risks to which one could permissibly be exposed by others would depend on the subjective capacities of the particular others with whom one happens (often unpredictably) to interact”.Footnote 118 An objective standard is therefore required if our expectations are to be sufficiently secure. But even if this argument is accepted in general terms, it would appear not to apply in the contexts in which English law has in fact employed a subjectivised approach. After all, a trespasser can hardly complain that the risks to which he is permissibly exposed depend on the subjective capacities of the occupier with whom he “happens to interact”, since that interaction is something he has (wittingly or unwittingly) forced on the unwilling defendant, and Wright himself accepts that in this context a subjective approach is appropriate.Footnote 119 What is more, Wright's argument would appear not to cover omissions of the kind involved in the Leakey category of nuisance case, since (as Tony Honoré has pointed out) in nonfeasance cases the defendant threatens not the claimant's security so much as his expectation of improvement, a different but secondary value.Footnote 120 The import of Wright's analysis is not therefore that a subjective approach should be ruled out altogether, but that, while it may be unacceptable in most instances of positive conduct, a subjective standard of care may be peculiarly appropriate when it comes to liability for omissions.Footnote 121
C. The Use of a Gross Negligence StandardFootnote 122
Although the term negligence covers a multitude of sins, from momentary inadvertence to wilful disregard for the safety of others,Footnote 123 English tort law recognises only one standard of care – that of the reasonable person. At the liability stage, degrees of negligence are not distinguished, and ever since Rolfe B. famously dismissed it as negligence “with the addition of a vituperative epithet”,Footnote 124 it has been orthodox stoutly to deny that gross negligence does or should play any part in the civil law. A good example of this scepticism is Millett L.J.'s remark that “English lawyers have always had a healthy disrespect” for the distinction between different degrees of negligence.Footnote 125 However, there are also examples of English judges defending the gross negligence concept,Footnote 126 and it has been suggested that dispensing with it was a mistake.Footnote 127
The utility of the gross negligence concept has most recently been considered in debates on the permissible scope of trustee exemption clauses, a context in which some jurisdictions draw a distinction between ordinary negligence by a trustee, which is excludable, and gross negligence, which is not. In its report on trustee exemption clauses the Law Commission rejected the use of this distinction on certainty grounds,Footnote 128 but the Trust Law Committee defended the concept, arguing that there was “a long and respectable line of authority … dealing with the concept of gross negligence in the common law, and distinguishing it from ordinary negligence”,Footnote 129 and the Scottish Law Commission concluded in its discussion paper on Breach of Trust that gross negligence was a “workable concept”,Footnote 130 and recommended its continued use in this context in Scottish law. According to the Scottish Law Commission:
The courts in Scotland have been using gross negligence in the field of trust law for well over a century without it having caused difficulties … It has to be acknowledged that it is impossible to draw a hard and fast line between negligence and gross negligence, but difficulties in establishing boundaries occur throughout the law.Footnote 131
In this part of the article it will be argued that gross negligence is a meaningful concept which could usefully be employed in negligence law where for some reason a modified standard of care is considered desirable. The discussion is divided into three sections, which consider (1) the nineteenth-century reaction against the use of gross negligence in English tort law; (2) the way that gross negligence is currently employed in other areas of English law, and in the law of other jurisdictions; and (3) the definition of gross negligence.
1. The nineteenth-century reaction against the use of gross negligence in English tort law
Gross negligence (or crassa negligentia) is generally thought to have come to English law from Roman law via BractonFootnote 132 and the law of bailment, where it was used to minimise the obligations of gratuitous bailees.Footnote 133 The concept was then extended to other gratuitous undertakings, so that, for example, it came to determine the personal liability of charitable trustees.Footnote 134 The language of gross negligence was also employed in a number of other contexts, though in some it may have meant no more than the failure to take reasonable care.Footnote 135 Examples included the liability of a lawyer to his client, or of a doctor to his patient; whether the holder of a bill of exchange would be affected by a defect in the bill; and whether a tenant was barred from relief against forfeiture.Footnote 136 Towards the middle of the nineteenth century, however, the courts began to turn against gross negligence. Influential judicial criticism of the concept came in 1842, when Denman C.J. doubted whether any intelligible distinction existed between gross and ordinary negligence,Footnote 137 and – famously – from Rolfe B. a year later.Footnote 138 There was also a decisive move against gross negligence on the ground.Footnote 139 In 1856, for example, the rule that an innkeeper was liable for the loss of a guest's property unless he could establish gross negligence on the guest's part was abolished, with an ordinary negligence standard being employed instead.Footnote 140 As far as tort law was concerned, the killer blow came in the same year, with Alderson B.'s classic definition of negligence as “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”.Footnote 141 The advent of the reasonable man condemned gross negligence to an (at best) marginal role in the modern law of obligations.Footnote 142
It would however be a mistake to regard the nineteenth-century reaction against gross negligence as evidence that the concept lacks validity or utility, since the origins of that reaction lay elsewhere.Footnote 143 First, gross negligence was often used to protect those who performed services gratuitouslyFootnote 144 (on the basis that the recipient of such a service could not reasonably expect the same level of care as one who had paid), but the historical trend was then – and has since been – away from special treatment of this kind.Footnote 145 Secondly, gross negligence may have fallen into disfavour because, up until the mid-nineteenth century, it was rather chameleonic.Footnote 146 This meant that it was not always clear what the term meant in a particular context, so that while at times it seems in effect to have been a synonym for fraud, recklessness or wilful default, at others it suggested a subjective standard of care, or was indistinguishable from ordinary negligence.Footnote 147 And finally, in the early nineteenth century gross negligence was used somewhat disingenuously as a way of circumventing exclusion clauses in contracts of carriage and such like: it was held that clauses excluding liability for negligence could not have been intended to cover gross negligence, but then when it came to instructing the jury, gross negligence was described as the failure to take reasonable care.Footnote 148 Although an effective way of neutralising exclusions of liability, this ruse devalued gross negligence and called its distinctiveness into question – it is noteworthy, for example, that Hinton v Dibbin,Footnote 149 the case in which Denman C.J. questioned whether there was any intelligible distinction between gross and ordinary negligence, was a case of this kind.
2. Current usage of gross negligence and similar concepts
Although gross negligence has now been almost completely abandoned by the English law of tort,Footnote 150 gross negligence and related concepts are commonly employed in other areas of English law, and in the tort law of other jurisdictions. This widespread reliance on the distinction between gross and ordinary negligence would suggest that the distinction is perfectly intelligible, and that its application in practice is not particularly problematic. Moreover, it lends support to the contention that one size does not invariably fit all, and that there are contexts where a gross negligence standard can usefully provide additional flexibility.
In English criminal law, it is well established that manslaughter can be committed by gross negligence,Footnote 151 and the concept is used (in effect) to distinguish dangerous driving from the lesser offence of careless driving.Footnote 152 Gross negligence is also a central pillar of the offence of corporate manslaughter.Footnote 153 Nor is our criminal law unusual in its reliance on the concept; New Zealand has, for example, moved from a position in which certain types of manslaughter case were governed by an ordinary negligence standard to an across-the-board test of gross negligence.Footnote 154 What is more, criminal lawyers seem perfectly comfortable with the concept. According to a leading textbook,
If negligence is regarded as non-attainment of a required standard of conduct then it is clear that there are degrees of it. One person may fall just short of the required standard, another may fall far short.Footnote 155
Gross negligence is also widely used in other areas of English law. In the law of mortgages, for example, a legal mortgagee will be postponed on the ground of his conduct only if he is guilty of fraud or such gross negligence that it would be unjust to deprive a prior equitable incumbrancer of his priority.Footnote 156 And in Re D' Jan of London Ltd,Footnote 157 the Court of Appeal held that a director of an insolvent company who had negligently signed an insurance proposal without reading it should be relieved of full liability to compensate the company for its resultant loss because his negligence had not been gross. Gross negligence is also frequently used to limit the scope of exclusions of liability. Contract parties often stipulate that an exclusion clause does not cover gross negligence,Footnote 158 for example, and on one view, trustees used to lose the protection of exemption clauses if they were grossly negligentFootnote 159 (this is still the case in many other jurisdictions,Footnote 160 including ScotlandFootnote 161). These examples show that, as Lord Clarke remarked in a recent Privy Council decision on trustee exemption clauses, English law does recognise “the difference in legal principle between negligence and gross negligence”.Footnote 162
Nor is English law alone in this, for (as Tony Weir pointed out) “other systems operate quite nicely” with the gross negligence concept.Footnote 163 In the United States,for example, gross negligence has been used to limit the liability of emergency vehicle operators, public bodies such as parole boards, volunteer athletic coaches and occupiers vis-à-vis recreational users of their land,Footnote 164 and in the past both U.S. and Canadian legislation frequently dictated that non-paying passengers could only sue car drivers if they established gross negligence or some other form of aggravated misconduct,Footnote 165 though the reasoning behind these “automobile guest statutes” has fallen out of favour, and most such legislation has now been repealed. French law has also long recognised the distinction between gross and ordinary fault. Under the Loi Badinter governing road traffic accidents, for example, a car driver can plead the defence of victim fault against a non-driver victim only if the fault was inexcusable (faute inexcusable).Footnote 166 Meanwhile, the French administrative courts have employed a standard of grave fault (faute lourde) to shield public authorities engaged in particularly complex, difficult or sensitive tasks, such as police activities and the exercise of regulatory functions,Footnote 167 and, although there has been a decline in the use of faute lourde in recent years, the move to ordinary fault (faute simple) appears mainly to have occurred in contexts – such as taxation, and the emergency and postal services – where special protection was arguably never appropriate in the first place.Footnote 168 There is extensive reliance on gross negligence in German civil law as well. If, for example, an object gratuitously given or lent to another causes him injury, the donor is liable only for wilful default or gross negligence,Footnote 169 while where an employee causes loss in the course of his employment, this is generally apportioned between employer and employee if the worker's fault was “normal”, but borne entirely by the worker in the case of gross negligence.Footnote 170 Other civil law countries also rely on the concept.Footnote 171 In Austria, for example, a wrongdoer guilty of gross negligence may have to pay more extensive compensation than one who is not,Footnote 172 and the presence of gross negligence or intention gives rise to an action where emotional disturbance falling short of damage to health is caused by the death of a close relative,Footnote 173 while in Spain parents are jointly and severally liable for damage they cause to the property of a minor child if they acted intentionally or were grossly negligent.Footnote 174
3. Defining gross negligence
The simpler the definition of gross negligence, the better, particularly since historically its meaning was notoriously elusive.Footnote 175 The most straightforward meaning that can be attached to the concept is “conduct that falls far below the standard of the reasonable person”. This is in effect the test laid down for the offence of corporate manslaughter,Footnote 176 and was also the definition given for the concept of “serious fault” in the Law Commission's consultation paper on Administrative Redress.Footnote 177 Nor should this test pose great problems in practice, for it has been said to be “no more difficult to say whether a person fell far below the acceptable standard than whether he fell below it at all.”Footnote 178 Another way of encapsulating the same idea is to say that there must have been a “very marked” or “major” departure from the expected standard of care.Footnote 179
Two further points should be made on the definition issue. The first is that it is not the fact that risk-taking is conscious that distinguishes gross negligence from ordinary negligence. Hence, although there may be an overlap between the two, gross negligence is not to be equated with recklessness, since recklessness is a state of mind, while negligence (of any kind) is a failure to comply with a particular standard of conduct. Similarly, because “[g]ross negligence, like negligence not so qualified, may be committed in good faith and, therefore, without dishonesty or wilfulness”,Footnote 180 it is not to be equated with fraud or wilful misconduct. The difference between negligence and gross negligence is one of degree, not kind.Footnote 181 And the second point is that an alternative workable test of gross negligence would focus not on the degree of the departure from the reasonable person standard, but on whether it was a clear or obvious departure. Some of the definitions of the French faute lourde standard have this flavour to them, referring to fault that is “flagrant”, or which “jumps out at you”,Footnote 182 and we have seen that this kind of language can also be found in some of the recent English case law concerning the liability of public authorities.Footnote 183 A test along these lines might be particularly appealing where there are overkill concerns, to the extent that this objection to negligence liability is predicated on potential defendants misreading the standard of care required of them.Footnote 184
IV. When and Why Might the Use of a Modified Standard of Care be Desirable?
The final issue to be considered is when and why the use of a modified standard of care might be desirable. My primary concern here is with situations in which for some reason it is felt to be inappropriate to recognise a duty of care accompanied by the usual objective standard of care of the reasonable person, and where a choice must therefore be made between refusing to impose a duty of care at all (or imposing a duty only in certain limited circumstances), and imposing a duty, but modifying the standard of care. My analysis of this issue is divided into two parts. First I identify some general advantages and disadvantages of these two possible responses to liability concerns, and then I consider the choice between the two approaches in particular types of case (as well as touching upon some other possible uses of a modified standard of care).
A. General Advantages and Disadvantages of the Two Approaches
Although I have argued elsewhere that the duty of care concept has outlived its usefulness and should now be done away with,Footnote 185 it would be foolish not to acknowledge that the denial of a duty of care does offer some advantages over the recognition of a duty accompanied by a modification of the standard of care. Because notional duty – the question of whether there is a rule of law that either allows or bars recovery in this category of case – is by definition a question of law, no-duty (or restricted-duty) rules are binding precedents, which should enable appellate courts to exercise effective control over the boundaries of negligence liability and provide potential litigants and lower court judges with clear guidance as to where those boundaries lie. By contrast, breach of duty is of course a question of fact, and it follows that decisions on breach do not create precedents, and provide no more than useful guidance to judges in future cases.Footnote 186 Furthermore, even though juries in negligence cases are long gone, breach is still essentially a “jury” question, “a matter of impression resulting from the consideration and weighing-up of lots of different factors”.Footnote 187 The result is that recognising a duty but applying a modified standard of care inevitably results in less certainty than a bright-line no-duty rule.Footnote 188
The other side of the coin, however, is that the discretionary and particularistic nature of the breach inquiry means that modifying the standard of care may produce a more proportionate response to the liability concern in question than a blunt no-duty rule. We might note, for example, the criticisms levelled at English negligence law in Osman v United Kingdom,Footnote 189 where the European Court of Human Rights expressed its concern that the no-duty rule under scrutiny applied across the board, without reference either to the gravity of the harm suffered by the claimant or to the gravity of the defendant's alleged negligence:
In [the Court's] view, it must be open to a domestic court to have regard to the presence of other public interest considerations which pull in the opposite direction to the application of the [no-duty] rule. Failing this, there will be no distinction made between degrees of negligence or of harm suffered or any consideration of the justice of a particular case.Footnote 190
The failure of no-duty rules to take any account of the gravity of the alleged negligence was also emphasised in the reports of the European Commission on Human Rights in Z v United Kingdom and TP and KM v United Kingdom. Referring this time to the no-duty rule laid down by the House of Lords in the child abuse context,Footnote 191 the Commission acknowledged that the exclusionary rule furthered a legitimate aim,but was not satisfied that it amounted to a proportionate response, noting that it “gave no consideration to the seriousness or otherwise of the damage or the nature or degree of the negligence alleged”.Footnote 192
In the past, an additional advantage of the no-duty route was that it enabled the defendant to seek to dispose of a claim on a strike-out application, or as a preliminary issue, and hence to dispose of an unmeritorious claim on the pleadings, and without the cost and inconvenience of a trial. This advantage has however been undermined by two developments at the procedural level. The first is that the courts in at least some duty cases have discouraged the use of these mechanisms, and insisted that where the law is less than crystal clear, the duty question should generally be answered in the light of the actual facts as revealed at trial. In S v Gloucestershire C.C., for example, May L.J. said that in difficult duty cases, it “will often not be possible to determine by an abstract inquiry which does not address the detailed facts of a particular case that a claim in negligence is bound to fail”,Footnote 193 while according to Lord Browne-Wilkinson in Barrett v Enfield L.B.C.:
[I]n an area of the law which [is] uncertain and developing … it is not normally appropriate to strike out. In my judgment it is of great importance that such development should be on the basis of actual facts found at trial not on hypothetical facts assumed (possibly wrongly) to be true for the purpose of the strike out.Footnote 194
The second relevant development was the passage of the Civil Procedure Rules 1998 (CPR). For whereas under the old rules the defendant had to show on a striking out application that the claim must plainly fail as a matter of law, the summary judgment procedure under CPR 24.2 is not so limited, and it follows that a defendant can apply for summary judgment on a question of fact, such as breach of duty.Footnote 195 Moreover, because the parties are now able to bring forward evidence in preliminary proceedings of this kind,Footnote 196 the court may well be in a position to decide the breach issue on such an application. The possibilities thrown up by the CPR are illustrated by the S v Gloucestershire C.C. case,Footnote 197 where the two appeals concerned allegations that the claimants had been sexually abused by foster parents assigned to them by the defendant local authorities. At first instance, both claims were struck out as disclosing no cause of action, and the defendants were asking the Court of Appeal either to uphold those decisions, or to determine the appeals as applications for summary judgment under CPR 24.2. On the summary judgment issue, the court held that while one of the claims should proceed to trial, in the other summary judgment should be given against the claimant, since the evidence produced by the local authority – which consisted of contemporary notes from the relevant social services files – indicated that she had no real prospect of establishing negligence. Although this decision shows how the current procedural regime reduces the advantages offered by the use of no-duty rules to limit liability, of course it remains true that even if a modified standard of care is employed many cases will still need to proceed to trial on breach, so that no-duty rules will continue to provide defendants like public authorities with greater security than a modified standard of care.Footnote 198 The question, however, is whether the greater certainty offered by blanket prohibitions justifies excluding liability in even the most egregious instances of negligence.
B. The Choice Between the Two Approaches in Particular Types of Case
Turning to the choice between the two approaches in particular types of case, I will first consider the areas in which a varied standard of care may already have been used in English negligence law, and then consider whether there are other contexts where a modified standard might usefully be employed. Once again, the assumption underlying the analysis is that the alternative to modifying the standard of care is either a no-duty or a restricted-duty rule.
1. Areas where a modified standard of care may already have been used
We have seen that there are three areas where a modified standard of care may already have been used in English negligence law: occupiers' liability, the liability of public authorities, and liability for deliberate third party acts. The use of a subjectivised standard in the occupiers' liability context has been justified on the ground that the obligations in question are not voluntarily incurred, but thrust upon the defendant against his will.Footnote 199 And, although there are admittedly some difficulties with the use of a subjective standard of care,Footnote 200 it does seem to be a reasonable response to the peculiarly coercive nature of positive obligations, particularly if these require expenditure on the defendant's part.Footnote 201 In particular, the otherwise reasonable argument that a person who cannot afford to take necessary precautions against the risks of an activity should not engage in that activityFootnote 202 carries no weight where the defendant is sued for failing to do something, provided there was no earlier assumption of responsibility on his part. It follows that if it were considered desirable to extend omissions liability – so as, for example, to impose a general duty to rescue – then the use of a reduced or subjectivised standard of care would be an obvious way of softening the blow.Footnote 203
In the public authority liability context, the abandonment of a public law filter which amounted to a modified standard of care but notionally operated at the duty stage was described by Duncan Fairgrieve as a “welcomed rationalization of this area of law”.Footnote 204 This seems right. While there may have been some practical advantages in categorising what was effectively a breach issue as one of duty under the old procedural regime, it was difficult to see any rationale for it once a claim could be disposed of at a preliminary stage on breach grounds. As for ruling out a duty of care on the ground that the decision in question was a difficult and delicate one, this was surely a case of using a sledgehammer to crack a nut, and taking account of the difficulty and sensitivity of the decisions public bodies must make at the breach stage seems a more proportionate response to the problem. Of course, it may be questioned whether any special protection is required at all in this context, but on the assumption that it is, it seems preferable that it take the form of a modified standard of care operating at the breach stage, rather than no-duty or restricted-duty rules.
Finally, there is the difficult question of liability for the consequences of deliberate third party conduct. We have seen that some judges have preferred to deal with this issue by means of a notional duty analysis, while others have recognised a duty, but apparently varied the standard of care. Commentators also disagree as to which approach is preferable.Footnote 205 The arguments seem evenly balanced. In favour of duty restrictions is the fact that this is an area where duty is capable of operating at a categorical level, with recognised exceptions to a general no-duty rule,Footnote 206 and so this approach is likely to provide better guidance to potential litigants than use of the breach mechanism. Furthermore, there is a risk that under a breach approach, judges reluctant to impose liability will simply conclude (rightly or wrongly) that the foreseeability of the third party behaviour was very low, in which case a set of clear duty limits would not in practice lead to different outcomes but would offer advantages in terms of efficiency and transparency. On the other hand, the breach approach would as ever offer greater flexibility and a more fact-sensitive approach in an area which has proved resistant to attempts at generalisation.
2. Other contexts where a modified standard of care could usefully be employed
There are also other contexts in which the technique of modifying the standard of care could usefully be employed. At a general level, we might think for example of the uses to which the gross negligence concept is put in the tort law of other jurisdictions. Hence, a modified standard of care could be employed in cases where good Samaritans and volunteers are sued for negligence; to govern the right of redress of an employer against an employee for whom the employer is vicariously liable; and to determine the availability of certain types of damages, or of damages for certain types of harm. And rather than the all-or-nothing approach which currently governs the exclusion of liability for negligence causing personal injury and death under section 2(1) of the Unfair Contract Terms Act 1977 (no exclusion of business liability; no limit on exclusion of non-business liability), a gross negligence standard could be used as part of a more nuanced approach, or alternatively the exclusion of liability for gross negligence could simply be made impermissible in all cases.
There are at least three further examples of liability concerns which could be met with a modified standard of care. The first is the concern that recognition of a duty of care would create an undesirable conflict of interest, as where it is argued that a public authority's performance of its statutory child protection duties might be undermined by a duty of care owed to those it wrongly accuses of abuse.Footnote 207 Elizabeth Handley has argued that rather than simply refusing to recognise a duty of care in such situations, the courts should consider the use of a modified standard of care, which would ensure that at least the “more egregious cases” could be dealt with.Footnote 208 Similarly, in his dissenting speech in one of the leading cases on the issue, Lord Bingham favoured the imposition of a duty of care, but with the caveat that liability would require proof of “a very clear departure from ordinary standards of skill and care”,Footnote 209 a restriction seemingly intended to be analogous to the French faute lourde standard.Footnote 210 On the other hand, in Lawrence v Pembrokeshire C.C., Auld L.J. specifically rejected the possibility of using breach rather than duty to deal with cases of this kind on the grounds that it would put
to one side the mischief at which the … exclusion of a general duty of care … is based, namely the potential conflict of interests … creating the imperative, whilst the truth is still unknown, for social workers to do all that they reasonably can and should to secure the welfare of the child.Footnote 211
While it seems clear, therefore, that the recognition of a duty of care in cases of this kind would be less inhibiting if tempered by the application of a standard of gross negligence or the like, it may be that the welfare of the suspected victim is so paramount a consideration that no degree of inhibition can be countenanced at all.
The second liability concern which could be met with a modified standard of care is the so-called “overkill” argument, that is to say the argument – deployed most frequently to protect public authorities – that the threat of liability may lead potential defendants to go about their activities in an excessively cautious (or “defensive”) manner, or even to curtail them altogether.Footnote 212 If we assume that this is a valid concern, then whether varying the standard of care is an appropriate response to it depends on why it is that potential defendants are over-deterred in this way. If the reason is that they simply do not want to have to fight negligence claims at all, because of the cost and trouble of doing so, then a modified standard of care is a less effective response to the concern than a no-duty rule, since they will still face a threat of liability, albeit a reduced one. On the other hand, if the overkill concern rests on the assumption that potential defendants may misread the standard of conduct required of them, or be uncertain as to that standard, then it may be peculiarly appropriate to respond to it by varying the standard of care, since this makes it much less likely that defendants will take inefficient precautions,Footnote 213 while at the same time affording claimants the possibility of redress in egregious cases.
It is noteworthy in this connection that in France application of the faute lourde standard has been justified by reference to overkill concerns,Footnote 214 while one of the reasons given for employing the sufficiently serious breach test in European law has been that otherwise the relevant institutions would be unduly hampered in the performance of their functions.Footnote 215 It is also worth noting that a gross negligence standard is used in a number of European countries to protect those charged with the supervision of financial institutions from liability to third-party depositors and investors, primarily out of fear that untrammelled fault liability could distort priorities and distract supervisory authorities from their core function of promoting financial stability.Footnote 216 And while other jurisdictions have responded to the same concern by laying down an even more lenient standard of bad faith (as in the U.K.) or by immunising such authorities from third-party liability altogether (as in Germany), the use of a gross negligence approach has been endorsed by commentators on the grounds that it is sufficient to allay fears of defensive supervision, while at the same time affording depositors and investors redress in more egregious cases of supervisory failure.Footnote 217
And finally, varying the standard of care might be considered (as in the French law of administrative responsibilityFootnote 218) to be an appropriate response to the connected concern that the task in question is particularly delicate or difficult, though here a sufficiently robust application of the ordinary negligence standard ought to be enough to allay the concern.
V. Conclusion
This article has been concerned with the varying of the standard of care in negligence to protect defendants. It has been argued that English negligence law may already have varied the standard of care in three categories of case: the liability of occupiers for omissions; the liability of public authorities; and liability for the consequences of deliberate third party acts. It has also been argued that where the varying of the standard of care is thought to be desirable, two particular modifications can usefully be employed: one is to subjectivise the standard, though only to a limited degree; the other is to make liability conditional on proof of gross negligence. Finally, consideration has been given to when and why a modified standard of care might be used.
Two final points should be made. The first is that, while the judiciary have already indicated a willingness to vary the standard of care in certain contexts, future reliance on this technique is perhaps more likely to be the product of legislative intervention, as it generally has been in other common law jurisdictions. This is particularly the case when it comes to the use of a gross negligence standard. And the second point is that drawing attention to the potential advantages of this technique is perfectly neutral as far as debates regarding the appropriate boundaries of negligence liability are concerned. Of course, if the standard of care is varied in existing duty situations, the effect is restrictive, but equally a modified breach test can be and has been used to justify the extension of negligence law into new areas.Footnote 219 My primary purpose has been to draw attention to the tool available, rather than to advocate its use for any particular task.
