Canterbury Law Review
The courts have devoted much effort to the question of when a duty of care in negligence ought to be held to exist. They have turned to various 'tests' involving both incremental and general principled analysis. A multitude of analytical labels have been coined in the process. And yet no universal and comprehensive answer to the duty question has been found. While concern to identify the extent of negligence liability and to bring coherence between the liability rules grows, the precise content of the duty of care remains as elusive as ever.
The uncertainty gives cause to wonder whether the duty of care is a useful part of the negligence inquiry at all. If it is, it is necessary to search for the sources of the confusion surrounding the duty of care and to discover how that confusion may be minimised. A comparative analysis of the different suggested tests may provide some insights. Of particular interest are the roles of analytical labels and variable policy considerations, and the interface between the two.
The New Zealand Court of Appeal's statements on the duty of care issue in the landmark case South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd and, more recently, in A-G v Carter are quite open and manipulable. It has recently been suggested that the Court clarify with more certainty the nature of the duty of care test in this jurisdiction. It is timely then to consider whether more definitive statements about the duty of care are necessary or even possible. Reviewing some decisions of the House of Lords and the High Court of Australia is instructive in that inquiry.
Logically, the duty of care must be evaluated with reference to the overall purpose of the law of negligence. Unfortunately this introduces additional uncertainty where it is already abundant. The difficulty in identifying the precise purpose of negligence law is a major obstacle to its developing in a principled way.
The law of negligence seeks to express moral judgements about conduct based on the moral norms of society and to allocate rights and responsibilities accordingly. The general public sentiment of moral wrongdoing is its 'sovereign principle.' The law is a construct that society applies to sanction certain types of conduct or harm. Not all moral responsibility necessarily gives rise to legal responsibility, but the former is necessary to the latter. Alternative goals, such as deterrence and compensation, are clearly better addressed by other branches of law and public or private insurance. There is little argument that moral rights are at the heart of negligence law; the controversial question is 'what moral rights?' Different ideas of moral responsibility pull the law in different directions. For one, it is unclear whether the moral objective of negligence law should focus on restoring equality of position between injurer and victim (corrective justice) or to ensure the just distribution of resources in society (distributive justice). However, making individual wrongdoers responsible to their victims for corrective purposes may concurrently realise distributive justice by ensuring that those who stand to benefit from activities also bear the associated risks.
Other tensions are not so easily resolved. One is whether the law of negligence should be more concerned with requiring plaintiffs to take responsibility for their own safety or with requiring defendants to take steps for the protection of others. When the value of autonomy is emphasised and individuals perceived as free agents, it seems commendable to let the costs of accidents lie where they fall. Focussing instead on collective welfare and the constraints that social networks place on free choice, greater vigilance by defendants seems necessary. Modern people commonly value individualism but also expect a certain degree of paternalism and welfare protection from outside forces. The law must try to reconcile responsibility and freedom.
Individual autonomy and the protection of others from injury are poles on a continuum. Where negligence liability is set will depend on which moral arguments weigh most strongly with the court at the time. Ultimately, the question boils down to one of 'plaintiff's tort law or defendant's tort law?' Rhetoric claiming that a duty of care is presumed in the absence of 'very potent counter considerations', or alternatively that the opposite presumption is the 'default position', may indicate whose position the court currently favours most. A duty of care is a threshold at which moral wrongdoing gives rise to legal liability set with reference to prevailing moral norms.
Because moral norms evolve over time, the law of negligence must be somewhat fluid. Anxiety that the law be certain cannot justify the court abdicating its duty to developing the common law to meet modern demands. The functions of negligence law include providing a just remedy where the law might otherwise be too rigid, and providing a vehicle for challenging the status quo. As society becomes more complex and technologically advanced, the law must adapt to novel circumstances. Freezing the law of negligence in the interest of certainty would defeat its purpose and leave a gap in the law.
Unfixed notions of morality provide no basis for a precise formula for liability. The term 'test' is convenient, yet no automatically determinate test is possible. Rather, the law of negligence is necessarily permeated by policy considerations. In novel cases, one cannot jump from a mere examination of the facts to the conclusion that a duty of care was owed. A normative premise must be imported to answer the duty question. A normative premise is a proposition sensitive to moral norms that explains why on the facts the defendant ought to have taken care. It reflects what society deems appropriate behaviour. To illustrate, 'if one starts with the factual statement that the defendant hit the plaintiff, to reach the conclusion that the defendant ought not to have done it, one needs to bring in the additional normative premise that hitting people is wrong.' Reasoning is incomplete absent a normative premise.
Policies can promote behaviour consistent with normative premises. A policy may, for example, discourage hitting or encourage rescue. Moral norms that support contradictory policies can co-exist in a society, but a compromise solution is philosophically supportable. Compromise is achieved by weighing policy considerations. Disagreeable as it may sound, liability, set between the poles of individual autonomy and prevention of harm, is the outcome of a policy debate. Policy considerations about the effect of liability in light of certain moral values are traded off to reach a decision. Even the requirement that harm be reasonably foreseeable reflects the widely held moral agreement that individuals should not be held responsible for harm that they could not have avoided. At a fundamental level, negligence law can only make sense with reference to policy decisions directed at advancing moral norms.
Without a precise formula, it is difficult to know what criteria are required of the duty of care. Variability between cases leaves the uncomfortable impression that setting liability is little more than a 'pseudo-science.'
A number of commentators are of the view that the duty of care is a superfluous 'fifth wheel on the coach', deserving of abandonment. Civil law systems in Europe manage negligence liability without an equivalent to the duty test. The duty of care inquiry has been accused of responsibility for the confused state of the law of negligence, particularly in the areas of economic loss and psychiatric injury.
The duty of care has been said to be tautologous with other elements, or the entire question, of negligence. Certainly there is considerable overlap. Concepts such as breach of standard of care, causation and remoteness can limit liability to the same extent as the duty of care requirement. Recently, particularly in relation to the liability of statutory authorities, it has been doubted whether, beyond requiring foreseeability of harm, the duty of care serves a positive function in limiting liability. Some have thought it better to explore matters in greater detail at later stages of the negligence inquiry rather than precluding liability at the outset.
The duty of care is a valuable method of confining the ambit of negligence liability. In fact, it synthesises numerous different criteria used by the courts to this end. Given the flexible, policy-based nature of negligence law, want of certainty is no reason to discard the duty of care. The notion that a standard of care is breached can be equally vague. Further, that the state of the law may be confused in some areas does not necessarily implicate the duty. That the concept is now deeply embedded in the common law is no unassailable rationale for maintaining the duty of care, but practicality dictates that the fact be given some recognition.
The duty inquiry appeals particularly because it seeks to unite the law of negligence by questions of law. Thus it does contribute to the quest for greater consistency in negligence law. The inquiry seeks to establish, as a legal rule for general application, whether the law will recognise in principle the possibility of liability in given situations. Other branches of the negligence inquiry focus more on questions of fact particular to a case. As Lord Goff said:
It is very tempting to try to solve all problems of negligence by reference to an all-embracing criterion of foreseeability, thereby effectively reducing all decisions in this field to questions of fact. But this comfortable option is, alas, not open to us.
Deciding issues of liability on the notion of duty of care, rather than on notions of causation, better alerts decision makers to the fact that policy determines the ultimate outcome. Also, it seems prudent to address certain considerations at the outset of the negligence inquiry so that claims lacking merit proceed no further.
A review of the cases reveals that the approach to divining the duty of care has varied considerably over the past century. In the ongoing endeavour to provide a test for the duty of care, a dichotomy has emerged between an approach based on principles of general application and one based on incremental development of the law.
The courts have often sought to develop the duty of care by reference to broad general principles or rules. As such they have attempted to identify 'some element common to the cases where [the duty of care] is found to exist.' Principles that provide a logical link between cases and to the underlying rationale of negligence law would bring a sense of consistency and legitimacy to decisions and equip the court to deal with previously uncharted areas. The main problem with general principled approaches is the failure to identify a succinct set of principles that can provide a complete answer to the duty question in every case.
Alternatively, the courts have often depended on categories of reported cases to guide the development of the duty of care. Incremental analysis in its modern form is built upon the words of Brennan J in Sutherland Shire Council v Heyman:
It is preferable ... that the law should develop novel categories of negligence incrementally and by analogy with existing categories, rather than be massive extension of a prima facie duty of care restrained only by indefinable "considerations which ought to negative, or ought to reduce or limit the scope of the duty or the class of persons to whom it is owed."
The focus is not to identify a universal principle, but to narrow the inquiry to specific classes or pockets of negligence case and to analyse the issues relevant to the duty of care within those confines. ‘Incrementalism' can bear a number of shades of meaning. Different incremental approaches tolerate different degrees of incremental development and policy consideration. But generally speaking, identifying a particular category of case as the starting point and proceeding by close analogy with precedent in the duty inquiry are the hallmarks of incrementalism. The main difficulty with incrementalism is explaining how categories of liability are created and extended.
To investigate the perceived advantages and difficulties of each approach, it is instructive to visit some landmark decisions of the House of Lords and the High Court of Australia. The dichotomy between general principles and incrementalism has arguably been most pronounced in these senior appellate courts.
Before 1932, the rationale for extending negligence liability had not been properly confronted. The law of negligence extended incrementally by the development of narrow 'duty situations.' Lord Atkin responded to concerns to cohere the 'duty situations' with the 'neighbour principle':
You must take reasonable care to avoid acts or omissions which you can reasonably foresee would be likely to injure your neighbour. Who then, in law is my neighbour? The answer seems to be - persons who are so closely and directly affected by my act that I ought reasonably to have them in my contemplation as being so affected when I am directing my mind to the acts or omissions which are called in question.
Lord Atkin's passage founded liability on foreseeability of harm and a close and direct relationship between the parties. Alone, it lacked sensitivity to important factors, such as the distinction between physical and economic loss or between positive acts and omissions. As the time arrived to consistently regard the law of negligence as relying on recognised principles to decide new points rather than requiring that cases fall within an existing authority, the neighbour principle needed refinement. In Anns v Merton London Borough Council, Lord Wilberforce set out what has become known as a 'two-stage test' for the duty of care, based on general principles:
First one has to ask whether, as between the alleged wrongdoer and the person who has suffered the damage there is a sufficient relationship of proximity or neighbourhood such that, in the reasonable contemplation of the former, carelessness on his part may be likely to cause damage to the latter, in which case a prima facie duty of care arises. Secondly, if the first question is answered affirmatively, it is necessary to consider whether there are any considerations which ought to negative, or to reduce or limit the scope of the duty or the class or person to whom it is owed or the damages to which a breach of it may give rise.
On its face, the first stage of the Anns test restates the neighbour principle. The requirement of a 'sufficient relationship of proximity or neighbourhood' seems to directly endorse Lord Atkin's 'restricted reply' to the question 'who is my neighbour?' Similarly, the requirement that harm be reasonably foreseeable is made plain. The second limb of the Anns test refines the broad sweep of the neighbour principle, giving more expression to those factors that confine the duty of care. While initially enjoying widespread application, Anns quickly came under assault as being dangerously expansionary. The approach allegedly neglected the importance of the relationship between the parties in confining the duty of care. A view prevailed that Lord Wilberforce's requirement of 'proximity or neighbourhood' was to be analysed solely in terms of reasonable foreseeability of harm. As such, a prima facie duty of care could virtually be taken for granted whenever a reasonable person in the position of the defendant should have foreseen the possibility of damage to the plaintiff. The fear was 'that we are all neighbours now, Pharisees and Samaritans alike,' and that every accident would sound in liability.
Compounding the problem, the second stage of the test came to be treated as subordinate to the first so that its utility as a control on the prima facie duty of care was limited. Negating policy considerations, it seemed, needed to be almost overwhelming in weight to avoid the 'massive extension of a prima facie duty of care.' Brennan J branded the description of relevant considerations too vague to be helpful.
In the English response to the perceived deficiencies of the Anns approach, there are two identifiable elements - restating the general principles of negligence liability in altered terms, and a more incremental approach to extending the duty of care. Both are expressed by the House of Lords in Caparo Industries plc v Dickman. The same elements have also featured in Australian decisions.
From Caparo emerged a three-stage test in place of the Anns two-stage test for liability. In addition to the foreseeability of damage, a relationship characterised by the law as one of 'proximity' between the parties and that it was 'fair, just and reasonable' to impose a duty were necessary ingredients of any situation giving rise to a duty of care. No presumption in favour of a prima facie duty of care was evident.
A series of cases culminating in Caparo emphasised a 'proximate' relationship between the parties as an independent criterion of the duty of care. 'Proximity' became the 'key word' in the English law of negligence. Similarly, in Australia, a majority of the High Court accepted 'proximity' as a 'distinct general requirement' of the duty of care. Unfortunately 'proximity' is not susceptible of precise definition. The identity and weight of relevant factors vary as certain types of damage or conduct require a closer or more direct nexus between the parties than others.
Like proximity, the requirement that it be 'fair, just and reasonable that the law should impose a duty of a given scope upon the one party for the benefit of the other' cannot be defined. Little more can be said of the requirement than that it is a general repository for a miscellaneous set of arguments, undefined in nature or number, which are invoked by the courts in determining whether a duty of care should arise.
The direction that the law should develop incrementally with particular attention to precedent has been seen as the key point of distinction between Caparo and Anns. In Caparo, Lord Bridge imported Brennan J's preference for developing 'novel categories of negligence incrementally and by analogy with established categories.' His Lordship noted the inability of any single general principle to provide a practical test for determining whether a duty of care is owed in every situation. Better, he thought, that the existence and scope of duties of care be guided by a more traditional categorisation of distinct and recognisable duty situations. Lord Oliver explained that the discussion of duty of care should proceed by reference to those specific areas, for example the distinction between acts and omissions, liability for the acts of third parties, misstatements and the exercise of statutory duties and powers, at the outermost boundaries of liability.
In short, the Caparo approach seems to instruct judges to begin the duty of care inquiry by locating the specific area of authority within which the facts of the case fall, and to proceed according to notions of foreseeability, proximity, and fairness, justice and reasonableness. While Caparo represents the law in England such that its authority is ritually summoned in every negligence case, its application has been less than consistent. Some House of Lords judgements employ a thinly veiled Anns approach. In Spring v Guardian Assurance plc, Lord Goff adopted a distinctly two-stage analysis. He considered first whether a prima facie duty of care arose, and second, whether the risk of undermining the law of defamation presented a policy reason to negate that duty. Lord Keith explicitly explained his dissent as based on the second stage of the test propounded by Lord Wilberforce in Anns.
Similarly, the consideration of conflicting policy factors by Lord Steyn in Marc Rich & Co A.G. v Bishop Rock Marine Co Ltd was heavily reminiscent of Anns. His Lordship incanted incrementalism but relieved himself of developing the common law incrementally by the short assertion that no past cases presented a realistic analogy. On an incremental approach, that state of affairs could lead straight to the conclusion that the facts disclosed no duty of care. Instead, Lord Steyn continued to weigh a number of policy factors tending to militate for and against the duty of care.
In Australia, the utility of 'proximity' as the touchstone of the duty of care came to be doubted. In Sullivan v Moody the High Court united in rejecting 'proximity' as giving little practical guidance or assistance as to process of reasoning in cases. The Court also rejected the Caparo three-stage formulation of general principles not merely for its reliance on proximity, but also because 'the question as to what is fair, and just and reasonable is capable of being misunderstood as an invitation to formulate policy rather than to search for principle.'
A greater emphasis on incrementalism seems to be the Court's preferred answer. Different categories of liability should be developed with reference to the 'different problems' encountered 'in determining the existence and nature or scope, of a duty of care.' Categories of case involving harm directly caused by the criminal conduct of third parties, defendant statutory authorities, potential unlimited liability or threats to the coherence of the law are given as examples. The Court directed that 'the relevant problem will then become the focus of attention in a judicial evaluation of factors which tend for or against a conclusion to be arrived at as a matter of principle.' The idea of categorisation by 'problem' is somewhat confusing as it seems that, more generally, the kind of conduct and the relationship between the parties in the case are also relevant to categorisation. Despite its emphasis on developing negligence liability by dividing cases into categories, and despite its rejection of Caparo, the High Court seems also to require a general principled approach:
There are policies at work in the law which can be identified and applied to novel problems, but the law of tort develops by reference to principles, which must be capable of general application, not discretionary decision-making in individual cases.
The Sullivan decision may be vexed by contradiction in that respect. Further, the distinction between policy and principle is unclear. Except perhaps for the level of generality, the Court does not identify the difference between the two. Elsewhere in Australian decisions it is suggested that the duty of care issue may be resolved by reference to 'salient factors.' Overall, uncertainty as to exactly how cases are to be categorised and as to the role and nature of principles of general application leaves the duty of care question quite open in Australia.
The landmark cases mentioned do not reflect accurately every subtle development of the law. However, they provide useful examples of different approaches to the duty of care question. Considering the substantial components of those approaches may reveal whether the courts have advanced or significantly changed the approach to the duty of care by the different re-workings of the 'test'.
Although 'proximity' as emphasised in Caparo was a response to a perceived poverty of attention available for the relationship between the parties, such attention was never precluded by the Anns test. It was open to say, as have the Privy Council and Commonwealth courts, that Lord Wilberforce's 'neighbourhood' or 'proximity' did not equate with simple reasonable foreseeability of harm. That view particularly commends itself if one agrees that the 'neighbour principle' entailed more than reasonable foreseeability and that the Anns first-stage was intended to reflect it. Alternatively, the broad language of the Anns second-stage was capable of supporting greater attention to the relationship between defendant and plaintiff. Lack of relational, circumstantial or causal proximity, it could be said, ought to negative a duty of care. In Yuen Kun Yeu v Attorney-General for Hong Kong, several of the considerations which told against a close and direct relationship between the parties could equally have been characterised as 'considerations which ought to negative, or to reduce or limit the scope of the duty or the class or person to whom it is owed.' The term 'proximity' had earned mention in Donoghue, Anns and earlier formulations of duty test. Why then is so much is made of its singling out in Caparo and certain Australian cases? The answer is not obvious. Perhaps what was new was the emphasis on proximity as the central control device within the definition of the duty of care. However, it is difficult to see how 'proximity' was ever intended as anything but a control on the scope of the duty of care. The focus on 'proximity' from the late 1980s seems at most a change in emphasis, introducing no substantive change to the test for the duty of care in negligence. If the change was more substantial, it is impressive that the High Court of Australia has abandoned 'proximity' without incurring a major alteration in the operation and outcome of negligence law.
All of the approaches, even those that advocate proceeding 'incrementally and by analogy', depend on a policy evaluation. Leaving aside the division of proximity and policy between different branches of inquiry, only tautology seems to separate 'considerations which ought to negative, or reduce ... the scope of the duty' and considerations that indicate it would not be 'fair, just or reasonable that the law impose a duty of a given scope.' Although the High Court in Sullivan deflects attention from the role of policy, it acknowledges 'there are policies at work in the law,' and that decisions require 'judicial evaluation of factors which tend for or against a conclusion' of a duty of care. The 'different problems' relevant to that inquiry, such as preventing indeterminate liability, promoting coherence in the law or affording a degree of protection to defendant public bodies, are clearly in the nature of policy considerations. Speaking of Sullivan, Kirby J expressed support for the astute recognition that 'the decision whether or not to impose a duty will be, ultimately, a normative one — a question of legal policy.' While the role of policy is acknowledged to varying degrees of frankness, it is equally important to the workable application and understanding of any of the articulated approaches.
The debate over the utility of 'proximity' and the difficulty in separating the concept out from the policy element of the duty inquiry may indicate that distinguishing approaches on the role of 'proximity' is superficial. Beyond flagging the facts of a particular relationship, 'proximity' has no independent meaning. Proximity and policy are 'one chameleonlike criterion'. 'Proximity' marks a policy preference about which relationships ought to be protected by the law. Its context is narrowest in nervous shock and economic loss cases and widest in the rescue type of case. 'Proximity' is a convenient label or shorthand but adds nothing essential and is no basis for distinguishing approaches to the duty of care.
Anns did not preclude an incrementalism. Lord Cooke has noted that in England under Anns the law still developed step by step and by analogy to decided cases, Lord Wilberforce's speech in McLoughlin v O'Brian being a 'classic illustration.' The continued application of Anns in Canada and New Zealand reinforces this point. In Cooper v Hobart, the Supreme Court of Canada has provided another 'classic illustration.' It might also be pointed out that the 'specific areas' mentioned in Caparo upon which the incremental development of the duty of care should proceed are very similar to those 'different problems' that are to be the focus of attention according to Sullivan.
The variously stated approaches to duty are all vulnerable to the same criticisms. Analytical labels lacking the precision necessary to be practical tests are relied upon in Anns, Caparo, and Sullivan. Those labels can embrace a number of unstable elements and there is no guidance as to which should be determinative. Declaring something proximate, fair, just, reasonable or principled sometimes simply records a finding rather than revealing the reasoning behind it. Emphasising incrementalism does not cure the affliction. How categories should expand is similarly incapable of precise answer. It is a matter of judicial judgement whether a duty of care would represent a small increment or a 'giant and impermissible leap' beyond existing authority. Categories can become so broad and difficult to define that they are only capable of providing vague answers. No approach has quelled scepticism that it is open to manipulation by the courts.
The different expressions given to the duty of care inquiry in landmark cases like Anns and Caparo have traditionally been seen as alternatives. However, the analysis reveals fundamental similarity in content. While particular words may emphasise that specific elements, for example the relationship between the parties or previous analogous cases, need attention, all of the tests are capable of bearing the same meaning. Similarly, different tests may prescribe the consideration of different elements in a different stage or order in the inquiry, but all of the same elements should eventually fall to be considered. None of the tests precludes a substantive avenue of inquiry permitted by another. Beyond emphasis and order, little more than tautology seems to separate the different approaches to the duty of care question.
Often it is sought to explain different decisions on similar facts in terms of the tests applied to resolve the duty of care issue. The use of a flawed test in an earlier case may be cited as the reason for treating that case as incorrectly decided. General principled approaches have been associated with the expansion of negligence liability in favour of plaintiffs. Incremental approaches have been seen as more restrictive on the extension of liability. It has been demonstrated in both Australia and New Zealand that cases reasoned on a categorical, incremental approach have tended to restrict liability within narrow confines, while decisions based on broader notions of proximity have tended to require more of defendants.
Similarly, different formulations of the general principled approach have been viewed as explaining the degree of expansion in negligence law. Cutting back a prima facie duty of care with reference to negating policy considerations rather than requiring policy reasons before contemplating any extension of the negligence law 'may tend to accelerate the expansion of the tort in our society.' The rationale is that 'it may well be more difficult to justify an exception on policy grounds than to establish a new situation on policy grounds.'
Correlation between the approaches and the outcomes of cases undoubtedly exists. But, is there any necessary relationship between the test employed and the ideas of moral judgement that underpin negligence law? Does the correlation evidence a well-founded causal connection between test and outcome?
The short answer is 'no'. Fundamental similarity in content indicates the tests are of operational rather than substantive significance to the duty of care inquiry. General principled and incremental approaches are both just methods of testing the 'fit of negligence liability with community standards of moral responsibility.' Competent and responsible judges could engage any to reach the same outcome in any particular case. The extensions of liability achieved under the Anns formula could have been accomplished by straight application or minor modification of existing categories of liability. Similarly, a proximity-based approach could easily have produced the outcome in Sullivan. The different treatment of incrementalism in the United Kingdom, Australia, Canada and New Zealand has not caused the law in those jurisdictions to develop in radically different manners. Since content should be determinative, associations based on structure are superficial.
As explanations of decisions, the tests are most unsatisfactory. Yet too often judgements leave the impression of a necessary link between approach and outcome. That is conducive to oscillation in the case law between different tests and distracts the focus from the different policy views that should explain different decisions. Focussing on tests encourages abandoning general principles and retreating to narrow categories as the appropriate response when the law has expanded too far. Conversely, general principles will be the resort when the incremental development of categories is too restrictive to accommodate desirable liability in a new situation. The authority of earlier decisions involving similar facts but a different approach consequently becomes unclear. The problem is illustrated by a famous example. Expansionary cases like Junior Books v Vetichi were allowed to discredit the Anns framework. That the real problem in Junior Books centred on the policy issue of economic loss was obscured by arguments over the nature of the Anns test. The allegations that the Anns test was responsible for excessive expansion, or that abandoning Anns purged English law of the problem, are in that sense misguided. Lord Keith's observation that, in applying Anns, courts could fail to regard, analyse and weigh all the relevant considerations in considering whether duty of care arose is more accurate. The expansion that coincided with the use of the Anns approach might be explained by an increasingly active and pro-plaintiff law-making posture of the courts at that time. Similarly, later decisions, in response to public sentiment on the issue, were more sympathetic to defendants. Lord Hoffmann explained that the Anns test involved the prima facie assumption of a duty of care where harm is reasonably foreseeable, but subsequent decisions had preferred to work in the reverse order from the opposite presumption. He observed that the associated trend was
to discourage the assumption that anyone who suffers loss is prima facie entitled to compensation from a person ... whose act or omission can be said to have cause it. The default position is that he is not.
Lord Hoffmann's dictum shows that in later cases, the view of the courts is that a duty of care should not be lightly found or plaintiffs unduly favoured. However, it also gives the inaccurate impression that the Anns approach is necessarily linked to favouring plaintiffs, and later approaches less so. It is an insightful observation that the criteria by which to determine duty had not changed. The courts rather referred to different approaches to 'legitimate subjective decisions made with primarily policy considerations in mind.' It would have been better to legitimise decisions by explaining that, as a matter of policy, the court requires more of persons who have suffered harm to show that they were owed a duty of care than was earlier the case. That position was thought to best reflect current moral norms. The risk otherwise is that the contemporary methodological approach will be confused with the reason for the decision. Treating differing 'tests' for the duty of care question as a problem is a self-fulfilling prophecy. The greater the obsession, the more a particular approach may be 'elevated to a degree of importance greater than its merits,' and the more numerous the associated problems. Contrasting approaches can suggest that there is a 'right' or 'wrong' where there is not, or that an approach can provide a complete answer. Further, it gives the inaccurate impression that approaches are substantially different, even mutually exclusive. That way, any approach is bound to be found wanting. Judges must take care to avoid giving these misguided impressions. Each encourages wasteful argument and confusion on matters that should be of no consequence to the ultimate outcome of a case.
The artificial association of outcomes and approach is one symptom of a broader problem. The question of the duty of care is determined ultimately on broad judicial conceptions of desirable policy. And yet the courts frequently seek to deny that decisions are so based. Legal doctrine may help to explain the adversity to explaining negligence decisions in policy terms. The application of policies drawn from moral norms is at odds with the philosophy that the law must remain separate from morals and politics. A judicial desire to assert the courts as a forum above the political fray, uninfluenced by 'community values' and 'moral opinions' has been evident.
The doctrine of the separation of powers also contributes. The courts are keen to avoid accusations of encroachment on the law-making province of the legislature. Recent decisions have asserted it is 'for Parliament, not the courts to determine policy questions.' The desire to replace the language of policy-based choice with the more neutral rhetoric of fairness in the late 1980s can be explained in these terms. Reference to 'fairness, justice and reasonableness' has in turn found disfavour as its real meaning as a reference to policy has been exposed. Of late, the courts seem to have preferred 'principle' to 'policy' as the label for the reasons behind their decisions. Perhaps also the courts seek to give the reassuring impression that the law of negligence is based on something considerably more certain than fluid notions of policy. The proposition that policy is determinative is abhorrent to expectations that the law be based on something 'less mutable'.
The view that judges do not apply policy to make law deserves to be dismissed as a 'fairy tale', perhaps more so in the necessarily variable tort of negligence than in any other area of the law. Perceptive commentators have acknowledged that 'new legal precepts ... arise by virtue of policy choices by judges within the leeways presented for the time being by the materials of the common law.' Declining an extension of the law may be as much a political decision as allowing one. Lord Cooke has explained the inevitability of judicial legislation and that hard cases depend on 'the policy of the law.' Concealing the extent to which judging is 'not discovery but creation' diminishes transparency in negligence law.
Were policy removed, the duty of care inquiry would be limited to reasonable foreseeability.  As discussed above, such a limited test for duty is inadequate. That unfixed moral norms can be difficult to discover does not absolve the judiciary of the duty of doing so. A plenitude of cases serves to dispel the fallacy that the common law can be kept free of policy problems. It is instructive to consider a few examples. Cases requiring determination of whether wrongful failure to prevent pregnancy gives rise to liability in negligence illustrate the inevitability of judicial policy-making in hard cases. These can only be resolved by a value-laden assessment of competing arguments, according to a personal perception of what society would deem morally acceptable. In Cattanach v Melchoir, Kirby J thought it 'self evident' that adjudication on issues like child rearing could not be divorced from social moral values.  Unfortunately that reality is not always candidly admitted. InMcFarlane v Tay side, the House of Lords effectively denies that public policy has a role in explaining whether a doctor should be responsible for the costs of rearing a child.
Considerations of policy also have a powerful role in determining liability for psychiatric injury, but again, recognition of this has been inconsistent. In Tame v New South Wales the High Court of Australia departed from certain restrictions on claims recognised by the House of Lords on policy grounds. Gummow and Kirby JJ explicitly cited 'policy reasons' as the basis for their favoured approach. Cases involving public defendants import additional policy questions. Policy is also the root of several recognised immunities to negligence liability.
Lord Hoffmann has recently provided insightful dicta on the role of policy in cases involving alleged duties to protect plaintiffs from quite obvious risks. In Reeves v Commissioner of Police of the Metropolis, he explained that 'the individualist philosophy of the common law' requires that '[p]eople of full age and sound understanding ... look after themselves and take responsibility for their actions.' That 'philosophy expresses itself in the fact... a duty to protect a person of full understanding from causing harm to himself is very rare indeed.' Similarly in Tomlinson v Congleton Borough Council, his Lordship exposed the importance of balancing the effect of imposing a duty on a defendant and the future opportunities of innocent third parties against the need to protect an irresponsible plaintiff, and recognised 'there is an important question of freedom at stake.' Because policy is determinative, the key to understanding different decisions on materially identical facts should be the policy view of the court at the time. In Nagle v Rottnest Island Authority, the High Court of Australia held that a public authority owed a duty of care to plaintiff who injured himself by irresponsibly diving in a swimming area under the control of the authority. When later confronted with very similar material facts the same Court has found against liability. The explanation for the divergence seems to be that, in response to public sentiment, the Court's opinion of the appropriate policy balance between individual and social responsibility had changed. In later decisions, greater emphasis has been placed on the requirement that' [p]eople of full age and sound understanding ... look after themselves.' Indeed, in Agar v Hyde, the High Court has explicitly endorsed Lord Hoffman's view.
Obscuring policy values is objectionable for a number of reasons. Firstly, the false impression that policy is not determinative is misleading, a 'chimera,' even intellectually dishonest. It does not aid the proper development of the law to obscure the realities of the judicial function in deciding hard cases. Further, public confidence in the law of negligence demands at least the frank recognition of reasons relied upon. Without disclosing policy considerations, normative premises are invisible in decisions. Reasoning that skips from statements of fact to concluding care ought to have been taken is incomplete and little better than holding a duty exists 'just because.' Concealing judicial policy might increase the incidence of that fundamental error - reliance on analytical labels incapable of precise definition as answers in themselves, independent of the specific reasons that ground the concepts in a particular case. Even if a judge does not in fact rely merely on analytical labels, it will be difficult to discern what ideas were relied upon.
Important decisions have produced occasion to explore the meaning of the term 'principle' in the law of negligence. In Sullivan v Moody, the High Court of Australia propounded the application of well-established principles of law rather than discretionary judgements as to what is fair, just or reasonable, for the development of the law. Similarly in McFarlane v Tay side, members of the House of Lords emphasised that the courts are concerned only with questions of principle or legal policy. 'Principle,' it seems, is the latest 'key word' in the law of negligence. Whether it unlocks any more doors than its predecessors deserves consideration.
'Principles' might mean to one of two things, or perhaps even both. Firstly, development of the law 'by reference to principles' might refer to a methodology of proceeding in a manner consistent with statements of the law clearly established in earlier cases. Secondly, 'principles' might be some set of identifiable and substantive rules capable of general application.
'Development by reference to principles' is capable of referring to the idea of deciding cases by close reference to statutes and precedents set in earlier cases by the appellate courts. A concept is 'principled' if it is securely established and confirmed by authorities. The court must start 'from a baseline of existing principle' and seek 'a solution consistent with or analogous to a principle ... already recognised.' This seems to be the manner in which Lord Keith used the term when he said the decision in Anns 'did not proceed on the basis of principle at all, but constituted a remarkable example of judicial legislation.' 'Principle' in this sense is used to promote a form of incrementalism. It highlights the need for reasoning to be connected to past decisional law and also tailored so that it can be applied to future cases.
As a complete explanation of adjudication on the duty of care, this notion of proceeding by principle is unsatisfactory. One problem is that no direct precedents exist to determine novel cases. Also, this use of 'principle' can mask the role of policy. Lord Scarman claimed that 'by concentrating on principle the judges ... can keep the legal system clear of policy problems which they ... are not equipped to resolve.' That the law of negligence should enforce an idea because it is strongly established in the case law, even if it is clear that the idea no longer has a sound policy basis, is surely unacceptable. A related danger is that a 'principled approach' will encourage excessive citation of previous decisions to conceal policy. These problems are discussed further below. Despite potential problems, 'principle' in this sense can be usefully understood as referring to the appropriate manner of exercise for judicial discretion.
The use of the term 'principles' is sometimes suggestive of a set of identifiable rules determinative of the duty of care question, though none have been judicially articulated. The cases distinguish 'principle' from 'policy' or at least 'principles and legal policy' from 'policy' generally. Such dicta appear on their face, and perhaps are intended, to limit the role of specific policy considerations in determining whether there is a duty of care. The judges in Sullivan echo Dworkin's rights thesis. For Dworkin, 'principles' refer to individuals' rights, which have a certain 'threshold weight' allowing them to prevail over other ordinary, routine goals in society. Dworkin strongly opposes judicial legislation on social policy, maintaining that decisions in hard cases 'should be generated by principle not policy.' However, he concedes that certain policy goals of 'special urgency' can sometimes curtail or supersede principle. In the context of the law of negligence, it is strongly arguable that principles and policy goals, in this sense that Dworkin uses them, are 'political aims that differ in degree but not in kind.' Whether a policy aim is identified as a 'principle' or as 'policy' is simply an expression of the weight and influence afforded to that aim. Similarly, it is important to recognise the extent to which broad, accepted concepts recognised as principles are dependant on policy or are umbrella statements embracing a number of more specific policy factors. As 'principle often disguises policy, and policy often transmogrifies into principle', the two may be distinguished only by level of abstraction. While incanting principle is another means of evading the allegation of inappropriate judicial policy-making and may give the impression of greater certainty, it does not change the content of the inquiry into whether a defendant ought to have taken care. One must not be seduced by the notion that general rules or principles alone are capable of determining the duty of care question. Experience has proved otherwise. It is submitted that reference to general, substantive 'principles' should not be interpreted to require a reduced reliance on policy, but should rather be seen as a way of structuring a policy inquiry.
Dividing up the possible meanings of 'principle' in this way is in a sense artificial. It is quite possible that in referring to 'principles' judges have intended to convey both meanings. Indeed, Dworkin used the term in both of the senses discussed. It is clear that the High Court in Sullivan was concerned in its discussion of 'principles' to promote the clear application of discoverable legal norms and to avoid an evaluation of policy in a manner inappropriate for the courts. So 'principle' might refer to the broad legal norms themselves and also to the practice of discovering and applying them. For 'principle' to bear all of these meanings may be of no consequence or even desirable, but alternatively may cause unnecessary confusion. Of the two interpretations we might give to the term 'principle', one suggests proceeding by analogy with close attention to precedent, the other suggests identifying some substantive rules of general application. The potential for face-off between 'incrementalism' and 'general principles' remains as vital as ever. The weaknesses noted here from which the notions of 'principle' suffer are those same weaknesses that have perpetually plagued incremental or general principled approaches. 'Principle' is not inherently capable of unlocking any more doors that its predecessors. It is vulnerable to the same fate as other terms of reference that are mere analytical labels. If expectations are set too high, the term 'principle' will fail. However, as will be discussed, the term 'principle' can also be applied to useful ways of thinking about those familiar concerns that pervade negligence law.
No one simple approach, be it based on general principles or incrementalism, has proved entirely satisfactory for determining whether a duty of care in negligence exists. Analysis has revealed that despite the different ways general principles or incrementalism are averred to in duty of care tests in cases like Anns, Caparo, or Sullivan, those tests are fundamentally the same. It is unnecessary, and worse still misleading, to focus on comparing or arguing between different formulations too much. All this is not to say that the offerings of the courts on the nature of the duty of care inquiry have been unhelpful. On the contrary, judges have repeatedly demonstrated their awareness of the complex nature of the exercise. The inability to settle on one approach as 'right' or 'wrong' evidences that sensitivity. Analytical labels can be dangerously misunderstood, but usually arise out of judicial alertness to the fact that certain elements require attention. Given recent decisions, it is useful to consider how the term 'principle' highlights some key and correct messages that the judiciary seeks to convey about the duty of care inquiry. General principles and incrementalism must not be treated as exclusive alternatives, but rather integrated to complement one another. Each operates to an extent to cure the deficiencies of the other. The coincidence of the two approaches in Caparo and Sullivan reflects that their integration is conducive to the most complete analysis of the duty of care problem. The New Zealand Court of Appeal has 'developed and applied the law of negligence in a way that largely synthesises the two approaches.' That the Supreme Court of Canada has combined the language of Anns and Caparo in Cooper v Hobart should come as no surprise.
Good decisions stem from an accurate appreciation of the existing scope and rationale of liability, and proceed by analogy and comparison with precedent. No logical coherence or consistency is possible if analysis begins de novo in every case. Examining earlier cases will alert judges to relevant policy considerations and to the weight they have been said to carry in contemporary society.
Identifying certain categories of cases can order the law in a practical way, particularly for the benefit of the lower courts and legal practitioners. Relying on general principles alone would leave the law of negligence in an unmanageable state. No set of general rules alone can succinctly express the entire duty of care inquiry for determining all cases. To account for all of the subtleties would require volumes. Instead, categories of decided cases are useful shorthand for expressing a great number of relevant considerations and accepted positions. They allow the courts to move forward on what has been decided in previous cases. Detailed argument is only required if a particular reason is raised to revisit a point. Otherwise decision-making that is anything other than haphazard would be impossible. In ordinary cases, decisions can and should be straightforward.
As discussed, a 'principled approach' can be synonymous with incrementalism. By extolling 'principle', the courts may have sought to highlight some very important points on the nature of judicial decision-making. Reference to 'principle' may be intended to reflect that in those 'hard cases' requiring policy deliberation and original decision there remains a real difference between the degrees of discretion open to Parliament and to a judge.
Where precedent ceases to provide complete guidance, a judge is not 'free to strike out on his own.' That would amount to rule by judges rather than rule by law. It would also encourage diversity in judicial decisions. Even if no direct authority for a decision exists, decided cases offer analogies and guidance as to the way the law is developing. Judges cannot presume that setting out alone to identify current moral norms and the best ways of giving effect to them will produce better results than drawing on the ideas expressed in past decisions. The prevailing public sentiment, not subjective personal opinion, is the 'sovereign principle' of judicial legislation in negligence. Though a judge's assessment of precedent 'will ... reflect his own intellectual and philosophical convictions .., that is a very different matter from supposing that those convictions have some independent force in his argument just because they are his.' 'Principle' must not be used to deny the creative activity of judges, but can usefully underscore that judges are heavily constrained in that activity compared with the legislature.
In accepting this role of 'principle' we must be alert to a danger. Some discussions of 'principled' methodology seem to suggest that ideas clearly established in principle should be enforced regardless of the quality of the policies behind them. A similar idea is that principles derived from authority capable of application in a novel case must, of necessity, be applied to that case. These propositions are inconsistent with the underlying purpose of negligence law.
Because negligence law seeks to make moral judgements based on social norms there should be coincidence between what is considered a useful principle and desirable policy. Clinging on to ideas or 'principles' once they no longer reflect contemporary social norms would defeat the law's purpose. The attitude that every revision of ideas established in case law requires the intervention of Parliament would render common law negligence redundant. Bad policy should not be perpetuated. Good policy should not be obstructed by 'principle' derived from analogous but not identical cases. Close consideration of principle is mandatory, but application is not. Judges must retain some discretion as to whether the ideas relied upon remain consistent with the demands of modern society.
Close attention to precedent compels 'a judge who makes changes in the law [to] take seriously the duty of reworking the pattern of the law.' However, it is essential that the same general principles inform the incremental development of the law across cases raising different issues. Otherwise the law of negligence might diverge in different and anomalous directions. Unexplained contradictions between decisions bring the law of negligence into disrepute.
Sole reliance on incrementalism would leave the law in an intellectually bankrupt state. The only available justification for finding that a situation is a duty situation would be that it had been declared so on a previous occasion. But the first time that situation arose there could have been no such prior declaration. The first case must therefore be wrongly decided and so too all those that succeed it. Putting aside the correctness of earlier decisions, it is unfair that cases should succeed or fail according to the historical accident of whether the court has previously encountered similar facts. The observation that the law has never recognised a duty of care in the circumstances is analytically unhelpful in novel cases. Key decisions, including Donoghue v Stevenson, have recognised claims in that situation.
Even on an incremental approach that does not demand authority precisely on point there are difficulties. By definition, previously decided cases cannot offer complete guidance in novel cases. Indeed, the more difficult a case the less assistance precedent is likely to offer. Judges require something more. General principles are required to frame rational policy arguments as to whether an extension of the law is desirable.
By incrementalism the courts reap guidance from the past. General principles provide guidance as to where to head in the future and link the law of negligence back to its underlying purpose. Even if all the courts are able to say in terms of 'general principle' is that novel cases are resolved by the careful balancing of policy arguments to decide whether the extension of liability is consistent with the moral norms of society, that is helpful. Information as to those policy arguments and current social norms can be gleaned from other decided cases, regardless of any 'category' within which that case might fall. Better still if the courts are capable of providing in terms of 'general principles' some broad, catchall statements of substantive policy, within which the finer arguments of particular cases can be considered.
Consider again the meaning the courts in recent cases may have intended to accord to 'principle'. The High Court of Australia has in the past spoken of 'guiding principles', 'formulated at a high level of generality,' and 'founded on broad conceptions of justice and morality'. Even if there is no qualitative difference between 'principles' and policy considerations, the statement of broad substantive principles, or statements of policy at a high degree of abstraction, can be valuable. They highlight important elements for consideration to ensure a comprehensive inquiry. For example, 'proximity' could be used to prompt judges to a more detailed consideration of how relational factors might control the extent of liability. In discussing 'principles', the courts are likely seeking to communicate something about the type of policies that are appropriate for judicial consideration. This is closely related to the idea that there are different degrees of discretion available to Parliament and judiciary. In McFarlane v Tay side, the distinction between principle and policy seems motivated in part by concern to convey that the considerations relevant to adjudication are more limited in range than those potentially relevant to Parliament's public policy-making. Adversity to the notion of 'judicial policy' can stem from the accurate observation that cases are not determined according to which decision will produce the maximum overall social good. Although legal decisions may have a wider social impact, judges must focus on doing justice in the case at hand and on maintaining coherence of the law. A party may be entitled to win a case even if that results in a net loss for society. But that is no reason to deny the operation of policy in negligence law. Invoking policy in making moral judgements about the quality of behaviour is not the same thing as calculating decisions on maximum social utility. Often decisions in negligence cases will produce a net benefit for society. That is because the law of negligence is based on social norms and many social norms operate to maximise the collective good. The idea of distributive justice may be a relevant policy consideration, but is not the baseline for analysis in every case. Factors besides overall public gain, such as what individual freedoms are felt to be particularly important, influence current moral norms. Also, the law and moral norms are in a co-evolutionary relationship and it is likely that contemporary norms will reflect those ideas that have been traditionally enshrined in law. Judicial decisions are based on the law derived from social norms and not public gain. Only those policies that reflect the extent to which the law enforces the moral norms of society are relevant to the duty of care inquiry.
Recently, it has been suggested that four 'fundamental principles' might underlie to the duty of care inquiry in most cases. In brief, these are that duties of care must avoid undue interference with individual autonomy, impose only proportionate levels of liability, afford an appropriate level of protection to plaintiffs and be coherent with the legal system as a whole. Like other statements of general principle, these 'fundamental principles' can only frame the more detailed inquiry necessary to determine the duty of care issue in novel cases. They do not provide rules for the automatic determination of cases. However, the approach is commendable in that it guides adjudicators towards the ultimate balancing process rather than seeking to conceal it.
'Fundamental principles' are beneficial provided it is recognised that policy considerations form their content. Like 'proximity', 'fundamental principles' may direct the court to particular factual inquiries, but these must be accompanied by a normative evaluation. The hollow statement, for example, that 'in the circumstances it is appropriate for the courts to recognise a duty to protect a person in the position of the plaintiff' is not enough. It is necessary go beyond analytical labels to explain why the balance is so struck. In the given example, the 'why', might involve policies encouraging self-protection by plaintiffs or encouraging appropriate care but not excessive caution by defendants.
That said, stating fundamental principles in the suggested way seems particularly useful because it directs the mind to the substantive content of relevant policy considerations, 'what we should be thinking about.'
In that way it might help to ensure that only the appropriate type of policy considerations are applied in judicial decisions. It has been argued that the courts currently consider some factors that are inappropriate to their inquiry. Describing general principles in terms of some substantive content may help to iron out those arguments. Assessing specific considerations against some broader objectives can show whether and how they are relevant to the duty of care inquiry.
By guiding judges toward answering the ultimate question via a roughly similar path, substantive general principles can make the overall body of case law more coherent. The more the content of principles can be identified, the more uniform the path is likely to become. Under statements of general principle, related specific policy concerns can be grouped together and linked to the more general objective that they are intended to serve. This can better allow judges to place specific policy questions in the context of wider opinions and agreements about the moral norms of society. General principles set up sounding boards against which judges may test how their ideas resonate with those of their peers. That should produce more accurate and consistent judicial judgements. As a wholly new type of policy consideration is not implausible, judges have need for easily identifiable analogies about the way that policies function. What emerges is that the identification of 'general' or 'fundamental principles' can facilitate the useful categorisation of decisions. While the most important message of incrementalism is that decisions must be attentive to and coherent with former decisions, categories of case seem to be an inevitable fruit of this process. Inappropriately categorising cases can distort the balancing of factors necessary to the duty inquiry, so care must be taken not to strain to categorise a decision. Also, associating cases by superficial facts, for example that they involved negligent misstatements, is irrelevant to the duty of care inquiry. If cases are to be categorised, it should be according to those substantive factors, like the risks of indeterminate liability or undue distortion of public budgets, which they have in common. Factually disparate decisions should not be separated when they raise the same policy concern. A balanced, symmetrical set of fundamental principles also serves as a reminder that a complex novel case is rarely decided on a single issue. That way the courts can hope to avoid inappropriately analysing cases in terms of one policy concern at the risk of overlooking others.
In novel cases, policy is determinative. Judges must therefore avoid masking policy influences by focussing on different approaches and labels. Instead, judges should articulate precisely what policy factors were considered relevant and how heavily they weighed in deciding whether or not to extend a duty of care. A franker style of analysis has been observed in judgements of more recent years. That trend should be encouraged. Returning to the purpose of negligence law, to express moral judgements in accord with social norms, the need for transparent policy reasoning is evident. The supposed normative premises upon which a decision is based must be easily discoverable. That way, observers may test whether the premises accurately reflect current moral norms and, thereby, whether a decision is convincing. In novel cases, 'what is the moral norm?' or 'what norm should prevail?' are likely to be controversial questions, and so considerable evaluation of the answers preferred by the judge may be necessary for the proper development of the law. A major difficulty in deciding the duty of care is that what is deemed 'the general public sentiment' will depend on the subjective perception of the judge concerned. Judges do not have the benefit of market surveys or elections to aid them in their assessment. But candour about policy views can mitigate the problem. The views of individual judges clearly expressed can moderate one another. Judicial candour additionally empowers outsiders to contribute to the debate. Judges may therefore proceed by collective wisdom to hone in on the most accurate view of current moral norms.
Exposing policy enhances the value of past decisions as sources of guidance and makes greater consistency in the law of negligence achievable. It is said to be desirable to avoid judicial decision-making on where the balance of community welfare lies on a case-by-case basis. Reliance on one judge's assessment of what is fair should be minimised. It should, therefore, be made obvious where the balance has been struck on former occasions. For as long as the moral norm underpinning liability in an area remains stable, every argued case should make the law clearer and more predictable. Again, narrowing in on one widely accepted view will make for straightforward decisions according to precedent in most cases.
It is necessary that the courts critically assess and re-evaluate situations where a duty of care is owed as time moves on. Articulating policy reasons will help to reveal where the law has fallen out of step with current social norms. For example, the view that nervous shock was not a valid head of damage was abandoned as scientific and subsequent social understanding has developed to recognise that psychiatric injury is a serious form of harm and that 'an acute emotional trauma, like a physical trauma, can well cause psychiatric illness.' Often there is a long lag before judges discern a change in social norms or are prepared to depart from their own entrenched attitudes. Honest discussion of policy values must surely speed up this process where necessary.
Transparent policy reasoning should equally expose changes to the law that lack sound basis. The disclosure of policy reasons helps to confine judicial legislation to acceptable limits. Judges should justify departures from the status quo by explaining their perception of the public sentiment and why, in policy terms, a law change is consistent with that sentiment. Requiring the statement of substantive policy reasons safeguards against judges imposing their own divergent views, papered over by flimsy verbal formulae or analytical labels. In contrast, if it is not obvious when or for what reason a judge has departed from the conventional wisdom, activist judges might be shielded from charges of usurping the proper role of the legislature.
Arranging relevant considerations in a uniform way will facilitate comparative consideration of policy factors in the case law. It may also help to avoid legal error. So, while no one approach is right or wrong or inherently more helpful in answering the duty of care question, it is desirable for practical purposes that judges within the same jurisdiction settle on a similar framework. A recent call for greater clarification of which considerations fall under the headings or 'proximity' and 'policy' currently employed by the New Zealand Court of Appeal could be supported in this respect.
The law of negligence is founded upon moral judgements based on social norms. It is necessarily of an uncertain content. The duty of care inquiry must reflect that uncertainty. Proposals for determining whether there is a duty to take care fail under the expectation of a greater degree of certainty than is achievable. We should accept rather than criticise a degree of uncertainty as a sign of vitality in the law.
The search for an unrealistic and undesirable degree of certainty in other jurisdictions has mostly hampered the judicial process. In that light, the approach of the New Zealand Court of Appeal expressed in South Pacific, and recently again in A-G v Carter, is wise and adequate. The modern history of the duty of care presents a cautionary tale about the use of analytical labels. Convenient shorthand labels, like 'test', 'proximity' or 'principle', are vulnerable to misuse. But shorthand is also necessary. Negligence law is so laden with complex issues that accurate discussion would be cumbersome without some abbreviations. Shorthand avoids unnecessary repetition of reasoning on accepted policy preferences that are not in dispute in a particular case. Further, to repeat Lord Cooke's point, labels help to organise thinking.
The lesson is that care must be taken not to inflate analytical labels beyond their capacity. In isolation from policy, analytical labels are empty terms. The term 'principle' can be dangerous if it is expected to answer to the duty of care question. Equally though, 'principle' can direct judges to some important insights on judicial decision-making and overarching objectives of negligence law. In South Pacific, the utility of analytical labels, provided that they are not employed to conceal that decisions are based on policy, is accepted.
No determinative test for the duty of care can be captured by a single formula. The best that can be done is to synthesise complementary incremental and general principled techniques to ensure the most comprehensive inquiry possible. Analysing the cases shows that the different suggested approaches to the duty of care inquiry are fundamentally similar in substantive content. Little is to be gained then from perpetual debate over what approach will deliver the best answer. New Zealand decisions are commendable in that they recognise that a particular approach is no more than a framework and that the outcome of a duty of care inquiry should not depend on what analytical method is employed. Courts that avoid futile debates should be congratulated, not criticised. In truly novel cases, 'there is no escape from the truth that... the outcome ... has to be determined by judicial judgement.' Formulae cannot explain whether a doctor should be liable for rearing a child, which family members should be entitled to claim for nervous shock and like questions. Transparent policy reasons are the most that can be offered to justify the imposition of the duty of care. Normative premises expressed in policy are what link decisions to the moral basis of negligence law. Judicial policy-making is an undeniable reality of the law of negligence. The courts might seek to endorse some broad substantive statements about the content of appropriate policy inquiries. Such statements can also provide some wider context for the policy-balancing exercise. Exposing policy is honest and allows observers to evaluate the quality of decisions. Perhaps the tort of negligence will never attract the same degree of public confidence as more certain areas of the law, but by giving policy reasons the courts at least explain how they justify their decisions in terms of those values that are important in our society.
Perhaps, because negligence decisions are complex and can affect the fabric of society, it is true that lower courts should limit themselves to following precedent. But the appellate courts are duty bound to deal with new challenges. Changing perceptions of individual rights, and the gathering momentum of genetics, cognitive science and other modern technologies, are bound to raise new duty of care questions. In answering those questions, the New Zealand Court of Appeal should not attempt an ultimate solution for all time. What is required is not revolutionary change, but the continued refinement and evolution of the Court's realistic approach.
[*] LLB(Hons), BA, firstname.lastname@example.org. This paper was awarded the Canterbury Law Review Prize for the best undergraduate Honours paper completed in 2004.
  2 NZLR 282.
  NZCA 48;  2 NZLR 160.
 Andrew Barker, 'The Duty of Care and the Search for Certainty'  New Zealand Law Journal 44, 47-8.
 David Howarth, 'Negligence After Murphy: Time to Re- Think' (1991) 50 Cambridge Law Journal 58, 61-2.
 The Wagon Mound  UKPC 1;  AC 388 (PC), 426. See also Donoghue v Stevenson  AC 562 (HL), 580 (Lord Atkin).
 Donoghue v Stevenson  AC 562 (HL), 580.
 See Prue Vines, 'Fault, Responsibility and Negligence in the High Court of Australia' (2000) 8 Tort Law Review 130, 134.
 Howarth, above n 4, 61; Vines, above n 7, 130.
 Bob Hepple, 'Negligence: The Search For Coherence' (1997) 50 Current Legal Problems 69, 71.
 Ibid; Richard Wright, 'Right, Justice and Tort Law' inD G Owen (ed), Philosophical Foundations of Tort Law (1995); cfIzhakEnglard, 'The Idea of Complementarity as a Philosophical Basis for Pluralism in Tort Law' in David G Owen (ed), Philosophical Foundations of Tort Law (1995) 183, 186.
 Vines, above n 7, 133
 Basil Markesinis, 'Plaintiff's Tort Law or Defendant's Tort Law? Is the House of Lords Moving Towards a Synthesis?' (2001) 9 Torts Law Journal 168, 168.
 X (Minors) v Bedfordshire CC  4 All ER 602, 619 (Sir Thomas Bingham MR).
 Stovin v Wise  UKHL 15;  3 All ER 801(HL), 824 (Lord Hoffman).
 In Hedley Byrne & Co Ltd v Heller & Partners Ltd  UKHL 4;  AC 465 (HL), 536, Lord Pearce explained that the extent of the duty of care depends ultimately on the courts' assessment of the demands of society for protection from the carelessness of others.
 In Hedley Byrne & Co Ltd v Heller & Partners Ltd  UKHL 4;  AC 465 (HL), 298, Lord Bridge said that to say that it is for the legislature, not the courts, to extend the limits of negligence liability any further, would be an unwarranted abdication of the courts function of developing and adapting principles of the common law to changing conditions, in a particular corner of the common law which exemplifies the important and indeed necessary part which that function has to play.
 Howarth, above n 4, 93.
 Justice D A Ipp, 'Policy and the Swing of the Negligence Pendulum' (2003) 77 Australian Law Journal 732, 734.
 Howarth, above n 4, 69.
 Englard, above n 10, 186, 193-5.
 Vines, above n 7, 135; Stephen Perry, 'Risk, Harm and Responsibility' in David G Owen (ed), Philosophical Foundations of Tort Law (1995) 321, 344-5.
 Robin Cooke, 'The Condition of the Law of Tort' in Peter Birks (ed), The Frontiers of Liability, Volume 2 (1994) 49, 59. See also Woolwich Building Society v Inland Revenue Commissioners (No. 2)  AC 70, 173 (Lord Goff); Marc Rich & Co A.G. v Bishop Rock Marine Co Ltd  UKHL 4;  1 AC 211 (HL), 230 (Lord Lloyd).
 Buckland, 'Duty to Take Care' (1935) 51 Law Quarterly Review 637, 639. See also Percy Winfield, 'Duty in Tortious Negligence' (1932) 34 Columbia Law Review 41, 66, describing the duty of care as the result of historical accident, and Howarth, above n 4, 58.
 Markesinis, above n 12, 173.
 P Atiyah, Accidents, Compensation and the Law (1970)47; Julius Stone, Province and Function of Law (1946) 181-182; Leon Green, 'The Duty Problem in Negligence Cases' (1928) 28 Columbia Law Review 1014, 1028-9.
 Elizabeth Handsley, ' Sullivan v Moody: Foreseeability of Injury is Not Enough to Found a Duty of Care in Negligence - But Should it Be?' (2003) 11 Torts Law Journal 1. In Phelps v Hillingdon London Borough Council  4 All ER 504, 530 Lord Nicholls said: 'Never is an unattractive absolute in this context. This would bar a claim, however obvious it was that something had gone badly wrong.... Denial of the existence of a cause of action is seldom, if ever, the appropriate response to fear of its abuse.'
 Jane Stapleton, 'Duty of Care Factors: A Selection from the Judicial Menus' in Peter Cane & Jane Stapleton (eds), The Law of Obligations: Essays in Celebration of John Fleming (1998) 59, 61.
 Hargrave v Goldman  HCA 56; (1963) 110 CLR 40, 63 (Windeyer J).
 Simon Deakin, Angus Johnston, & Basil Marksinis, Markesinis and Deakin's Tort Law (5th ed, 2003) 77, 85.
 Hargrave v Goldman  HCA 56; (1963) 110 CLR 40, 63 (Windeyer J); aff'd Modbury Triangle Shopping Centre Pty vAnzil  HCA 61; (2000) 176 ALR 411, 436.
 Smith v Littlewoods Organisation Ltd  UKHL 3;  AC 241 (HL), 280.
 Deakin et al, above n 29, 77.
 Donoghue v Stevenson  AC 562, 580 (Lord Atkin).
  HCA 41; (1985) 60 ALR 1 (HCA), 43.
 Keith Stanton, 'Incremental Approaches to the Duty of Care' in Nicholas J Mullany (ed), Tort in the Nineties (1997) 34, 38-39, 52.
 Green, above n 25, 1028-9; M McHugh, 'Neighbourhood, Proximity and Reliance' in P D Finn(ed), Essays on Torts (1989) 8.
 Donoghue v Stevenson  AC 562, 580.
 Dorset Yacht Co Ltd v Home Office  UKHL 2;  AC 1004 (HL), 1027 (Lord Reid).
  UKHL 4;  AC 728 (HL).
 Ibid 751-752.
 Donoghue v Stevenson  AC 562, 580.
 Anns v Merton London Borough Council  UKHL 4;  AC 728, 754.
 Particularly in the United Kingdom and Australia.
 SeePDowney, 'Incrementalism in Tort'  New Zealand Law Journal 113,113; Alan Roger 'Some Reflections on Junior Books' in Peter Birks (ed), The Frontiers of Liability, Volume 2 (1994) 64; McHugh, above n 36, 17; 64; Stephen Pitel, 'A Reformulated Anns Test for Canada' (2002) 10 Tort Law Review 10, 11; Tony Weir, 'Errare Humanum Est' in Peter Birks (ed), The Frontiers of Liability, Volume 2 (1994) 103, 105.
 CBS Songs Ltd v Amstrad Consumer Electronics  UKHL 15;  2 All ER 484, 497 (Lord Templeton).
 Sutherland Shire Council v Heyman  HCA 41; (1985) 60 ALR 1, 43 (Brennan J).
 Ibid. See also Brennan J's earlier statements in Jaensch v Coffey  HCA 52; (1984) 54 ALR 417, 437.
  UKHL 2;  2 AC 605 (HL).
 Ibid 617-619 (Lord Bridge).
 Those cases include Governors of the Peabody Donation Fund v Sir Lindsay Parkinson & Co Ltd  AC 210 (HL); Leigh and Sillavan Ltd v Aliakmon Shipping Co Ltd: The Aliakmon  UKHL 10;  AC 785 (HL) and Curran v Northern Ireland Co-Ownership Housing Assn Ltd  AC 718 (HL) in the United Kingdom and Jaensch v Coffey  HCA 52; (1984) 54 ALR 417 and Sutherland Shire Council v Heyman  HCA 41; (1985) 60 ALR 1 in Australia.
 Tony Weir, 'Fixing the Foundations' (1991) 50 Cambridge Law Journal 24, 25. Weir complained however that the key did not open many doors. For similar comment by an Australian jurist see McHugh, above n 36.
 Sutherland Shire Council v Heyman  HCA 41; (1985) 60 ALR 1, 63. See also San Sebastian Pty Ltd v Minister Administering the Environmental Planning and Assessment Act 1979  HCA 68; (1986) 162 CLR 340 (HCA); Cook v Cook  HCA 73; (1986) 162 CLR 376; Stevens v Brodribb Sawmilling Co Pty Ltd (1986) 160 CLR 16, 52; Burnie Port Authority v General Jones Pty Ltd  HCA 13; (1994) 179 CLR 520; Bryan v Maloney (1995) 182 CLR 609.
 Caparo Industries plc v Dickman  UKHL 2;  2 AC 605, 618 (Lord Bridge).
 See Kidner, 'Resiling from the Anns Principle: The Variable Nature of Proximity in Negligence' (1987) 7 Legal Studies 319; Sutherland Shire Council v Heyman  HCA 41; (1985) 60 ALR 1, 55 (Deane J); Caparo Industries plc v Dickman  UKHL 2;  2 AC 605, 651 (Lord Oliver).
 Caparo Industries plc v Dickman  UKHL 2;  2 AC 605, 618 (Lord Bridge).
 Mark Lunney & Ken Oliphant, Tort Law: Text and Materials (2000) 115.
 Ibid 114.
 Caparo Industries plc v Dickman  UKHL 2;  2 AC 605, 618 (Lord Bridge), adopting Sutherland Shire Council v Heyman  HCA 41; (1985) 60 ALR 1, 43-44.
 Caparo Industries plc v Dickman  UKHL 2;  2 AC 605, 617. Lord Bridge notes that the following cases make the same point: Governors of the Peabody Donation Fund v Sir Lindsay Parkinson & Co Ltd  AC 210 (HL); Yuen Kun Yeu v Attorney-General for Hong Kong  UKPC 16;  AC 175 (PC); Rowling v Takaro Properties Ltd  1 All ER 163 (PC) and Hill v Chief Constable of West Yorkshire  AC 53 (HL).
 Caparo Industries plc v Dickman  UKHL 2;  2 AC 605, 618.
 Ibid, 635 (Lord Oliver).
  2 AC 296 (HL).
 Ibid, 316, (Lord Goff).
 Ibid, 308 (Lord Keith).
  UKHL 4;  1 AC 211 (HL).
 Ibid 236.
 Ibid 236-242; Stanton, above n 35, 44.
 Hill v Van Erp (1997) 188 CLR 159 (HCA); Romeo v Conservation Commission of the Northern Territory (1998) 192 CLR 431; Esanda Finance Corp Ltd v Peat Marwick Hungerfords (1997) 188 CLR 241 (HCA). See Stanley Yeo, 'Rethinking Proximity: A Paper Tiger?' (1997) 5 Tort Law Review 174; Harold Luntz, ‘Turning Points in the Law of Torts in the Last 30 Years' (2003) 15 Insurance Law Journal 1, 7-8.
 (2001) 183 ALR 404.
 Ibid 414.
 Ibid 415.
 See generally Barker, above n 3, 45.
 Sullivan v Moody (2001) 183 ALR 404, 415.
 Sullivan v Moody (2001) 183 ALR 404, 415.
 But see Christian Witting, 'The Three-Stage Test Abandoned in Australia-Or Not?' (2002) 118 Law Quarterly Review 214, 219-220; J Smillie, 'Principle, Policy and Negligence' (1984) 11 New Zealand Universities Law Review 111, 145 and below VII 'Principles' - the new 'key word'.
 Perre vApandPty Ltd  HCA 36; (1999) 198 CLR 180, 253.
 See for example Yuen Kun Yeu v Attorney-General for Hong Kong  UKPC 16;  AC 175 (PC); Sutherland Shire Council v Heyman  HCA 41; (1985) 60 ALR 1 (HCA), 441 (Gibbs CJ); Cooper v Hobart (2001) 206 DLR (4th) 193 (SCC); Attorney-General v Carter  NZCA 48;  2 NZLR 160 (CA); Connell v Odlum  NZCA 256;  2 NZLR 257 (CA); South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd  2 NZLR 282 (CA).
  UKPC 16;  AC 175 (PC).
 For example, that the loss was economic in nature, the claimants were unascertained members of a huge class, the loss had been directly inflicted by the wrongful act of a third party and the defendant was a public commissioner. See W Rogers, Winfield and Jolowicz on Tort (16th ed, 2002) 111-112.
 Anns v Merton London Borough Council  UKHL 4;  AC 728, 752.
 Le Lievre v Gould  1 QB 491, 497 (Esher LJ), 504 (AL Smith LJ).
 Deakin et al, above n 29, 88.
 Anns v Merton London Borough Council  UKHL 4;  AC 728, 752.
 Caparo Industries plc v Dickman  UKHL 2;  2 AC 605, 618.
 Sullivan v Moody (2001) 183 ALR 404, 415.
 Graham Barclay Oysters v Ryan  HCA 54; (2002) 194 ALR 337 (HCA), 400 (Kirby J), citing Witting, above n 78, 221.
 Nicholas Mullany, 'Proximity, Policy and Procrastination' (1992) 9 Australian Business Law Journal 80, 82.
 John Fleming, The Law of Torts (7th ed, 1989). See also Hepple, above n 9, 81; Stovin v Wise  UKHL 15;  3 All ER 801 (HL), 807-8 (Lord Nicholls).
 Stovin v Wise  All ER 801 (HL), 808 (Lord Nicholls).
 See McHugh above n 36, 37-9; Smillie, above n 78, 141; Rogers, above n 82, 116.
  UKHL 3;  1 AC 410.
 South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd  2 NZLR 282 (CA), 295 (Cooke P).
 Cooper v Hobart (2001) 206 DLR (4th) 193, (SCC) . See Jason Neyers, 'Distilling Duty: The Supreme Court of Canada Amends Anns' (2002) 18 Law Quarterly Review 221.
 Attorney-General v Carter  NZCA 48;  2 NZLR 160, 169; South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd  2 NZLR 282, 294.
 (2001) 206 DLR (4th) 193.
 Caparo Industries plc v Dickman  UKHL 2;  2 AC 605, 635 (Lord Oliver).
 Sullivan v Moody (2001) 183 ALR 404, 415.
 'Analytical label' is the term used by Stapleton, above n 27.
 Caparo Industries plc v Dickman  UKHL 2;  2 AC 605, 618; Sullivan v Moody (2001) 183 ALR 404, 414-5.
 See Jane Stapleton, 'In Restraint of Tort' in Peter Birks (ed), The Frontiers of Liability, Volume 2 (1994) 83, 85.
 Goodwill v British Pregnancy Advisory Service  2AllER 161, 163. Stanton, above n 35, 46; Barker, above n 3, 47-48.
 Stanton, above n 35, 54
 Bruce Feldthusen, 'The Anns/Cooper Approach to the Duty of Care for Pure Economic Loss: The Emperor Has No Clothes' (2004) 18 Construction Law Reports 67.
 McHugh, above n 36, 10; Vines, above n 7, 143-4.
 See for example Pyrenees Shire Council v Day  HCA 3; (1998) 192 CLR 330; Romeo v Conservation Commission of the Northern Territory (1998) 192 CLR 431; Perre vApandPty Ltd  HCA 36; (1999) 198 CLR 180 in Australia and Price Waterhouse v Kwan  NZCA 311;  3 NZLR 39 in New Zealand.
 For example, Nagle v Rottnest Island Authority  HCA 76; (1993) 177 CLR 423 (HCA); Sutherland Shire Council v Heyman  HCA 41; (1985) 60 ALR 1 in Australia and R M Turton & Co Ltd (in liq) v Kerslake and Partners  NZCA 115;  3 NZLR 406 in New Zealand. See Vines, above n 7, on Australia and AndrewBarker, 'Divining an Approach to the Duty of Care: The New Zealand Court of Appeal and Claims for Negligent Misstatement'  OtaLawRw 4; (2001) 10 Otago Law Review 91 on New Zealand.
 Allied Finance & Investments Ltd v Haddow & Co  NZLR 22 (CA), 29 (Richardson J).
 Vines, above n 7, 132.
 See Stovin v Wise  UKHL 15;  AC 923, 931-2, 949; South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd  2 NZLR 282, 316 (Hardie Boyes J).
 Vines, above n 7, 143.
 See Smillie, above n 78, 139.
 Justice Kirby's discussion of Sullivan in Graham Barclay Oysters v Ryan  HCA 54; (2002) 194 ALR 337, 397-401, seems to indicate that the same decision would have been reached in Sullivan applying the Caparo approach. In cases involving similar facts, such as X (Minors) v Bedfordshire CC  UKHL 9;  2 AC 633 (HL) and Phelps v Hillingdon London Borough Council  4 All ER 504, similar results have been achieved by applying the Caparo approach. See also Stanton, above n 35, 51.
 Stanton, above n 35, 54.
 Junior Books Ltd v Vetichi Co Ltd  UKHL 4;  1 AC 520 (HL).
 Rowling v Takaro  1 All ER 163, 172.
 Smillie, above n 78, 139.
 Stovin v Wise  UKHL 15;  3 All ER 801(HL), 824 (Lord Hoffman).
 Joanne Conoghan & Wade Mansell, The Wrongs of Tort (1993) 15.
 Yuen Kun Yeu v Attorney-General for Hong Kong  UKPC 16;  AC 175 (PC), 710.
 See Stapleton, above n 27, 59-60.
 Stephen Todd, 'A Methodology of Duty' in Peter Cane (ed), Centenary Essays for the High Court of Australia (2004) 221, 227. See also McHugh above n 36, 38; Smillie, above n 78, 139, 148; John Fleming, 'Remoteness and Duty: The Control Devices in Liability for Negligence' (1953) 31 Canadian Bar Review 471, 486; Basil Markesinis, 'Policy Factors and the Law of Tort' in Derek Mendes da Costa (ed), The Cambridge Lectures (1983) 200.
 Hepple, above n 9, 76. See also Smith & Burns, 'Donoghue v Stevenson - The Not So Golden Anniversary' (1983) 46 Modern Law Review 147, 162-163.
 Peter Cane, 'The Doctor, the Stork and the Court: A Modern Morality Play' (2004) 120 Law Quarterly Review 23, 25.
 McFarlane v Tayside Health Board  UKHL 50;  4 All ER 961, 985 (Lord Hope) affirming Emeh v Kensington and Chelsea and Westminster Area Health Authority  3 All ER 1044, 1051 (Waller LJ). See also McLoughlin v O'Brian,  UKHL 3;  2 All ER 298 (HL), 310 (Lord Scarman).
 Stanton, above n 35, 37.
 In Sullivan v Moody (2001) 183 ALR 404, 416 the court emphasises that the 'different problems' relevant to the duty of care go beyond merely what is 'fair' or 'unfair.'
 McFarlane v Tayside Health Board  2AC59 (HL); Sullivan v Moody (2001) 183 ALR 404.
 Cattanach v Melchoir  HCA 38; (2003) 199 ALR 131, 153 (McHugh and Gummow JJ) quoting Lord Radcliffe, The Law and Its Compass (1960) 43-4. See Ipp, above n 18, 734.
 Lord Reid called the view that judges do not make law a 'fairy tale'. Cane, above n 129, 26.
 Julius Stone, Precedent and the Law (1985) 261.
 Stanton, above n 35, 42.
 Cooke, above n 22, 59.
 Benjamin Cardozo, The Nature of the Judicial Process (1921) 166-7.
 McLoughlin v O'Brian  UKHL 3;  2 All ER 298, 307 (Edmund-Davies LJ).
 See Part III above.
 See P Winfield, 'Public Policy in the English Common Law' (1928) 42 Harvard Law Review 93 cited with approval in McLoughlin v O'Brian  UKHL 3;  2 All ER 298 (Edmund-Davies LJ).
 Cane, above n 128, 23-4; See also McFarlane v Tayside Health Board  UKHL 50;  2 AC 59 (HL), 82- 3 (Lord Steyn).
  HCA 38; (2003) 199 ALR 131 (HCA).
 Ibid 164.
 For example, the majority in Cattanach claim that the case was disposed of by reference to 'ordinary' and 'basic' legal principles. See Cane, above n 128, 24 and also Ipp, above n 18, 745.
 McFarlane v Tayside Health Board  UKHL 50;  4 All ER 961 (HL).
 See for example, McLoughlin v O'Brian  UKHL 3;  2 All ER 298 (Edmund-Davies LJ).
 See for example McLoughlin v O'Brian  UKHL 3;  2 All ER 298 (Scarman LJ).
 In White v Chief Constable of South Yorkshire Police  UKHL 45;  2 AC 455.
 Ipp, above n 18,739.
 See for example, Sullivan v Moody (2001) 183 ALR 404; Phelps v Hillingdon London Borough Council  4 All ER 504; X (Minors) v Bedfordshire CC  UKHL 9;  2 AC 633 (HL); Hill v Chief Constable of West Yorkshire  UKHL 12;  2 All ER 238 (HL).
 See Ipp, above n 18, 733, 741 and Lord Denning's foreword to P Eddy, Professional Negligence (1956).
  1 AC 360.
 Ibid 368.
  UKHL 47.
 Ibid .
 (1993) 177CLR423.
 Romeo v Conservation Commission of the Northern Territory (1998) 192 CLR 431. See also Perre vApand Pty Ltd  HCA 36; (1999) 198 CLR 180; Agar v Hyde (2000) 201 CLR 552.
 Decisions like Nagle had the effect of raising insurance premiums to a level that had a damaging social impact. See Ipp, above n 18, 74.
 See above n 159 and also Secretary to the Department of Natural Resources and Energy v Harper  VSCA 36;  1 VR 133; Ghantous v Hawkesbury Shire Council  HCA 29; (2001) 206 CLR 512.
 (2000) 201 CLR 552 (HCA).
 Witting, above n 78, 215.
 Smillie, above n 78, 148.
 Sir Anthony Mason, The Nature of the Judicial Process and Judicial Decision-Making, A Matter of Judgement (2003) 14.
 (2001) 183 ALR404.
 Ibid, 415. See also Ipp, above n 18, 735-6; Witting, above n 78, 219.
  UKHL 50;  2 AC 59.
 Ibid 76 (Lord Slynn), 95 (Lord Hope), 100-1 (Lord Clyde), 108 (Lord Millet). For earlier but similar judicial comment see McLoughlin v O'Brian  UKHL 3;  2 All ER 298, 310 (Lord Scarman).
 Todd, above n 127, 28
 McLoughlin v O'Brian  UKHL 3;  2 All ER 298, 310 (Lord Scarman).
 Murphy v Brentwood District Council  2 All ER 908 (HL), 923.
 Smillie, above n 78, 148. See also Cattanach v Melchoir  HCA 38; (2003) 199 ALR 131 (Kirby J); Ronald Dworkin, Taking Rights Seriously (1977) 112.
 Howarth, above n 4, 71
 McLoughlin v O'Brian  UKHL 3;  2 All ER 298, 310.
 Ipp, above n 18, 737.
 See Part VII below.
 Dworkin, Taking Rights Seriously, above n 172, 92. See also Ronald Dworkin, A Matter of Principle (1985).
 Some of those rights relevant to litigation are a right to be treated as equals, a right to fair play and a right to consistency. See Dworkin, A Matter of Principle, above n 178, 85.
 Dworkin, Taking Rights Seriously, above n 172, 92.
 Umana, 'Dworkin's "Rights Thesis”' (1976) 74 Michigan Law Review 1167, 1175. See also Smillie, above n 78, 145-6.
 Ipp, above n 18, 735-6.
 Ibid 735.
 Graham Barclay Oysters v Ryan  HCA 54; (2002) 194 ALR 337, 399 (Kirby J).
 Two descriptions of a thing are complementary if each alone is incapable of providing a complete description or explanation of the thing in question and both together provide a more complete description. Englard, above n 10, 188.
 South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd  2 NZLR 282 (CA), 316 (Hardie Boys J).
 (2001) 206 DLR (4th) 193.
 See generally Rogers, above n 82, 113-114; Barker, above n 3, 47.
 Witting, above n 78, 219-220.
 Dworkin, Taking Rights Seriously, above n 172, 118.
 Ipp, above n 18, 746
 Caltex Oil (Aust) Pty v The Dredge 'Willemstad'  HCA 65; (1976) 136 CLR 529 (HCA), 567.
 Dworkin, Taking Rights Seriously, above n 172, 118. See also McFarlane v Tayside Health Board  UKHL 50;  2 AC 59 (HL), 82 (Lord Steyn).
 Weir, above n 44, 106.
 Dworkin, A Matter of Principle, above n 178, 73-4
 Ipp, above n 18,738
 Cf McLoughlinv O'Brian  UKHL 3;  2 All ER 298,310 where Lord Scarman suggests that curtailing policies that are socially unacceptable should be left to Parliament.
 Ibid 738.
 Levi, 'The nature of judicial reasoning' (1965) 32 University of Chicago Law Review 395, 403.
 Hepple, above n 9, 70-71.
 In Ward v McMaster  IESC 3;  IR 337, 347, McCarthy J said 'the verbally attractive proposition of incremental growth ... suffers from a temporal defect - that rights should be determined by the accident of birth.'
 Stapleton, above n 27, 66.
 Caltex Oil Pty v The Dredge 'Willemstad'  HCA 65; (1976) 136 CLR 529 (HCA), 567.
 See Stanton, above n 35, 39.
  UKHL 50;  4 All ER 961, 1000 (Lord Millet).
 Dworkin, A Matter of Principle, above n 178, 94.
 Todd, above n 126, 229.
 In Caltex Oil v The Dredge 'Willemstad'  HCA 65; (1976) 136 CLR 529, the High Court of Australia recognised that 'guiding principles' are informed by more focussed policy concerns.
 Todd, above n 126, 229.
 Ibid 32-5. See also Stapleton, above n 27, 64.
 Ibid 229.
 Stapleton, above n 27, 64-65, has argued that judicial observations based on insurance and wealth of parties are largely irrelevant and misleading.
 See McLoughlin v O'Brian,  UKHL 3;  2 All ER 298, 307 (Lord Edmund-Davies); Brown v Heathcote County Council  1 NZLR 720, (PC) 725.
 See Stapleton, above n 105, 85-6.
 Ibid, 90; Stapleton, above n 27, 86. It is evident in Sullivan v Moody (2001) 183 ALR 404, 415 that the High Court of Australia is alert to this fact.
 Stapleton, above n 105, 90.
 Stapleton, above n 27, 101.
 Stapleton, above n 105, 63.
 McLoughlin v O'Brien  UKHL 3;  1 AC 410, 433 (Lord Bridge).
 Ipp, above n 18,735.
 Ibid 747.
 Ibid 737.
 Smillie, above n 78, 148.
 Graham Barclay Oysters v Ryan  HCA 54; (2002) 194 ALR 337, 397 (Kirby J).
 Barker, above n 3, 47-8.
 South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd  2 NZLR 282, 294.
 Ibid 295.
 See for example Merrett v Babb  UKPC 46;  WLR 1,  (May LJ) and generally John Fleming, The Law of Torts (6th ed, 1983) 131.
 th Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd  2 NZLR 282, 294 (Cooke P); Attorney-General v Carter  NZCA 48;  2 NZLR 160, 168.
 South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd  2 NZLR 282, 294.