Melbourne University Law Review
|
|
Home
| Databases
| WorldLII
| Search
| Feedback
Melbourne University Law Review |
|
MATTHEW HARDING[*] AND IAN MALKIN[†]
[Our article aims to identify principles to guide the High Court of Australia when it is considering whether or not to overrule one of its previous decisions in a common law setting. We introduce two accounts of overruling drawn from the scholarly literature on the subject and, in light of those accounts, we evaluate the approach to overruling taken by the Court in the recent common law case of Imbree v McNeilly [2008] HCA 40; (2008) 236 CLR 510. We then place Imbree in context by considering it alongside two other relatively recent common law cases in which the Court has considered whether or not to overrule itself: Esso Australia Resources Ltd v Federal Commissioner of Taxation [1999] HCA 67; (1999) 201 CLR 49, and Brodie v Singleton Shire Council [2001] HCA 29; (2001) 206 CLR 512. We conclude that, in some respects, the Court’s approach to overruling in Imbree is problematic. In future common law cases where the Court is invited to overrule itself, the Court should seek to develop a framework of principles bearing on the overruling question. This framework should draw not only on Imbree but also on cases like Esso and Brodie, which in some respects were decided in accordance with sound principles of overruling that were not addressed adequately in Imbree.]
The High Court of Australia has, for many years, shown itself willing to make necessary changes in the law by overruling its previous decisions.[1] How willing the Court has been to overrule would appear to depend in part on the setting in which it has been asked to do so. For example, in the constitutional setting, the Court’s attitude to overruling has been mixed, owing to its unique position as the interpreter of the constitutional text; authorities may be found endorsing both a flexible and a more constrained approach to overruling in constitutional cases.[2] When asked to overrule itself in the interpretation of a statute, the Court has not been averse to a degree of flexibility.[3] However, when it comes to the common law, while the Court has been prepared to overrule in appropriate cases, the attitude of the Court to overruling has been more consistently cautious than in matters of constitutional or statutory interpretation.
The High Court’s preparedness — albeit reluctant — to overrule where required in common law cases is illustrated by Brodie v Singleton Shire Council (‘Brodie’).[4] There, Gaudron, McHugh and Gummow JJ, in a discussion of the kinds of circumstances in which the common law should be ‘re-expressed’ by the Court, stated that ‘[i]f the continuation of [a] state of affairs, which discredits the Australian legal system, be mandated by precedent, then it is the task of this Court to look into the authorities said to constitute that precedent.’[5] On this basis, their Honours were prepared in Brodie to overrule previous decisions of the Court endorsing the ‘highway immunity rule’ in the law of tort, but they also noted Mason J’s oft-cited view that the Court ‘is neither a legislature nor a law reform agency’,[6] and they explored in painstaking detail the numerous reasons that led them to overrule.[7] By way of contrast, the Court chose to resolve the case of Woolcock Street Investments Pty Ltd v CDG Pty Ltd (‘Woolcock’)[8] without overruling its earlier decision in Bryan v Maloney (‘Bryan’),[9] despite the fact that Bryan was clearly relevant to the issues in Woolcock and that the decision in Bryan had been built upon a legal foundation to which the Court no longer subscribed: the now-abandoned[10] conceptual approach involving an assessment of the ‘proximity of relationship’ between plaintiff and defendant.[11] Even though the Court had clearly departed from the fundamental principles that underlay Bryan by the time Woolcock came before it, it did not go so far as to overrule Bryan.
As Woolcock illustrates, courts have available to them a variety of techniques for achieving change in the common law. These include: distinguishing a precedent case (whether on the specific facts or with respect to changes to society generally), or not following a precedent decision because of changes to legal principle that effectively mean that it is no longer a precedent;[12] refusing to follow a per incuriam decision;[13] ‘doubting’ and ‘disapproving’ a precedent decision without formally responding to it;[14] and re-interpreting a precedent decision.[15] Explicitly overruling a precedent decision, which is sometimes referred to as ‘depart[ing] from’[16] or ‘re-express[ing]’[17] the rule of the precedent decision, is but one tool for achieving change, and an extreme one that is deployed only sparingly. Nonetheless, as Brodie illustrates, occasionally the Court will prefer the extreme option over other techniques.
Cases in which the technique of overruling is chosen in order to achieve change in the common law fall into the category of what English scholar J W Harris calls ‘copper-bottomed’ overruling cases.[18] For Harris, these are cases
in which a proposition of law … recognized to be part of the ratio decidendi of an earlier case is challenged, and in which either the proposition is rejected by virtue of the … overruling power, or it is left to stand because the [Court] considers that, whether that proposition should or should not form part of the law, it would be inappropriate to remove it by exercise of that power.[19]
Neil Duxbury expresses the same idea in his reference to cases in which judges explicitly address the question of whether or not they should ‘declin[e] to follow [a precedent decision] and declar[e] that, at least where the facts of a case are materially identical to those of the case at hand, a new ruling should be followed instead.’[20] ‘Copper-bottomed’ overruling cases are rare: indeed, J W Harris identified only seven such cases in the House of Lords up to 1990 in which the House’s overruling power was exercised;[21]
and although other commentators take a more expansive view of what constitutes an overruling case,[22] on no view are the cases numerous. Yet, despite the fact that overruling has not occurred with great frequency in the House of Lords (now replaced by the Supreme Court of the United Kingdom), much has been written about the practice of overruling in that court,[23]
especially since 1966, when the House issued its well-known Practice Statement, stating that it regarded itself as free to ‘depart from’ its previous decisions.[24] With notable exceptions,[25]
less has been written about copper-bottomed overruling cases in the High Court of Australia,[26]
notwithstanding the fact that the Court has regarded itself as free to overrule its previous decisions for a considerably longer period than its English counterpart has.[27]
Our article aims to fill part of the gap in the Australian literature on ‘copper-bottomed’ overruling cases in the common law setting. In Part II, we introduce some accounts of overruling drawn from the scholarly literature on the subject. In Part III we address the question: what are the principles that underlie the current approach of the High Court of Australia to overruling in common law cases which may also be described as ‘copper-bottomed’ overruling cases? In ascertaining what those principles are, Part III focuses on an overruling case that was before the High Court in 2008 and that involved a question of common law doctrine: Imbree v McNeilly (‘Imbree’).[28] In Imbree, the Court explicitly overruled its previous decision in Cook v Cook (‘Cook’)[29] regarding the standard of care owed by a learner driver. In Part III, we draw on the scholarly accounts of overruling introduced in Part II to evaluate the decision in Imbree. Although the substantive legal issue in Imbree was not particularly complex,[30] the case will presumably now be the starting point for the Court when considering the question of overruling in future common law cases, and it is therefore important to know whether or not the principles relating to overruling that the case stands for are defensible. Finally, in Part IV, we widen our view to consider Imbree in context, placing it against two other relatively recent overruling cases involving common law issues that the High Court has had to determine: Esso Australia Resources Ltd v Federal Commissioner of Taxation (‘Esso’)[31] and Brodie.
Approaches to overruling might usefully be conceptualised using the metaphor of a spectrum. At one end of this spectrum is the strict rule of stare decisis that characterised the approach to overruling taken by the House of Lords until 1966.[32]
According to this rule, overruling is never permitted: the ratio decidendi of a precedent case constitutes what philosophers of rationality have called an ‘exclusionary reason’,[33] binding a subsequent court to decide in accordance with it even if other reasons favour overruling.[34] At the other end of the spectrum might be an approach animated by the attitude expressed (extrajudicially) by Lionel Murphy, who described the doctrine of precedent as ‘eminently suitable for a nation overwhelmingly populated by sheep’.[35] Scattered along the spectrum between the two extremes are approaches to overruling that permit it, but that subject it to principled constraints. In the scholarly literature on overruling, at least two such moderate approaches may be identified which assist in evaluating the current approach to overruling taken by the High Court of Australia in common law cases.
The first of these approaches is presented in the work of J W Harris. In an article published in 1990, Harris examines in detail cases decided between 1966 and 1990, in which the House of Lords explicitly overruled itself, as well as cases in which the High Court of Australia overruled itself.[36]
He poses two questions. First, what were the principles underlying the judges’ approaches to overruling in those cases? Secondly, did the judges explicitly affirm those principles in their reasoning? Harris concludes that, although judges in the cases under study did not explicitly affirm the principles underlying their decisions to overrule, those principles may nonetheless be discerned from the judges’ reasoning.[37]
The first such principle that Harris discerns takes the form of a threshold, or prima facie, requirement that a court’s power to overrule may be invoked only where overruling would bring about some improvement to the law.[38] Based on his reading of the cases, Harris proposes that improvement to the law can be measured in three dimensions. The first such dimension is justice: if a legal proposition contained in the ratio decidendi of a precedent case would, ‘in view of the consequence of its universalized application’,[39] fail to do justice to a class affected by it, then to that extent the law would be improved by overruling that proposition.[40] The second dimension is certainty. In a sense, this consideration always militates against overruling, because overruling brings about unannounced change in the law.[41]
However, in a specific sense, the law may be improved if an ambiguous legal rule is replaced with a clear one, and in this sense, overruling may serve the value of certainty in the law.[42]
Finally, there is the dimension of coherence. For Harris, ‘[i]f it would be implausible to imagine any contemporary legislator wishing to enact [two] rules, they are incoherent’.[43] In such circumstances, improvement to the law may be achieved by overruling one of the rules in question, thus strengthening the foundations on which future law may be constructed in an orderly and predictable fashion.[44]
In order to satisfy itself that the threshold requirement of improvement to the law has been met, an overruling court must explore in detail the various reasons for and against the legal proposition contained in the ratio decidendi of the precedent case, as well as the reasons for and against any legal proposition that the court is asked to substitute for the ratio decidendi of the precedent case. In other words, the inquiry must concentrate on the law for which the precedent stands, and not on the authoritativeness of the precedent itself or the consequences of the court departing from one of its previous decisions per se. As Bryan Horrigan has put it, at the threshold stage, the focus is on ‘substantive’, not ‘precedential’, reasons.[45]
Substantive reasons typically take the form of doctrinal or principled arguments for or against a legal proposition, but they may also take the form of what are usually described as ‘public policy’ considerations bearing upon the law in a given area.[46] Precedential reasons are those that take the ratio decidendi of a precedent case as given and consider the pros and cons of expunging it from the law. Precedential reasons tend to be consequentialist in nature, and if they take the form of theoretical propositions, those propositions are not about matters of legal doctrine, but instead are about judicial reasoning itself.[47]
In J W Harris’ view, the threshold requirement of improvement to the law is subject to four significant constraining principles, in light of which a court may refrain from overruling, notwithstanding that improvement to the law would consequently be achieved.[48]
These four constraining principles, to employ Horrigan’s language, shift the focus of an overruling court from substantive reasons to precedential reasons.
The first — and, for our purposes, the most significant — of the four constraining principles is described by Harris in the following passage:
When all the reasons canvassed for or against a proposition of law in the [overruling] case were also canvassed in the decision under review, then what remains for the judge is a subjective weighing of them one against the other. He may, for his part, be convinced that his brethren in the [precedent] case read the scales wrong. But since, ex hypothesi, he cannot give a distinct reason in support of this impression, he cannot, in principle, describe as ‘demonstrably mistaken’ any third guess on the issue which a future judge, faced with the same catalogue of reasons, might make. … The need for finality on disputed questions of law thus justifies the constraint that [a court of final appeal] should not overrule a decision of its own if no new reasons are in play.[49]
The scope of the ‘no new reasons’ constraint depends on what counts as a ‘new reason’. For present purposes, it is not necessary to develop an exhaustive account of when a new reason is before an overruling court. Nonetheless, clear contenders would include changed political, social, or economic circumstances since the time when the precedent case was decided,[50] changes to the law surrounding the ratio decidendi of the precedent case, and arguments put to the overruling court that were not put to the precedent court. Harris identifies an exception to the ‘no new reasons’ constraint in cases where an issue of ‘fundamental importance’ is at stake, and the parties do not accept that there is room for reasonable disagreement on the issue, but instead insist that the other is ‘objectively wrong’.[51]
In such cases, according to Harris, a court may overrule notwithstanding that no new reasons are before it.[52] Harris says little about what constitutes an issue of ‘fundamental importance’ or ‘objective wrongness’; however, it is clear that he is thinking primarily of constitutional matters and not of the ordinary business of a court exercising its common law jurisdiction.[53]
For this reason, we put to one side in this article the ‘fundamental importance’ exception to the ‘no new reasons’ constraint.
A second constraining principle that Harris identifies points to the justified reliance of persons who have ordered their affairs on the basis of the ratio decidendi of a precedent case.[54]
Such a reliance interest constitutes a reason against overruling and may, in an appropriate case, lead a court to decline to overrule, notwithstanding that the law would thereby be improved.[55]
The third constraining principle that Harris discusses is activated in circumstances where the legislature has ‘acted on the assumption that the ruling in the [precedent decision] represents the law’;[56]
for example, where the legislature has declined to act upon the recommendation of a law reform commission that a rule established by the precedent decision be abrogated.[57] It is important to note that mere legislative silence will not necessarily trigger the third constraint: given the realities of the legislative process, caution must be exercised when interpreting the failure of a parliament to legislate with respect to a particular matter.[58]
However, in appropriate cases, ‘comity with the legislature’ may demand that a court stay its hand, notwithstanding that the law would be improved by overruling a precedent decision.[59]
Finally, Harris describes the constraining principle that a precedent decision should not be overruled where overruling would make no difference to the outcome of a case.[60]
Harris suggests that this ‘mootness’ constraint may be overly cautious, implying that a court might have a duty to overrule in some cases, even though it is not strictly necessary to do so, if overruling will bring order to the law.[61] However, as his account of overruling is drawn from his observation of the ‘phenomenology of adjudication’ in ‘copper-bottomed’ overruling cases,[62] Harris feels compelled to admit the ‘mootness’ constraint as a real one, even if it is not fully justified.[63]
In its emphasis on constraining principles — especially the ‘no new reasons’ constraint — as well as the prima facie requirement that an improvement to the law be brought about, J W Harris’ account of overruling may be described as moderate and closer to the strict rule of stare decisis than a laissez-faire approach to precedent. In this regard, J W Harris’ approach differs from the approach to overruling set out in an article by B V Harris published in 2002.[64]
B V Harris’ article responds to J W Harris’ view, taking issue with the approach to overruling described in the earlier article.[65]
B V Harris’ principal point of departure from J W Harris is with respect to the ‘no new reasons’ constraint. For B V Harris, even in a case where the dispute is not about an issue of fundamental importance and the parties accept that reasonable minds could disagree about the outcome, a court should be free to overrule itself notwithstanding that no new reasons are before it, if it attributes different weights to the reasons that were before the precedent court than the precedent court attributed to them.[66] In such cases, B V Harris says, the value of ‘doing justice in the particular appeal’ is more important than ‘the values embodied in the doctrine of stare decisis.’[67] Overruling courts should therefore engage in ‘energetic reasoning’, rather than ‘formalistic seeking of precedent’.[68] For B V Harris, the ‘no new reasons’ constraint raises a presumption in favour of a precedent decision; he argues that this presumption should be abandoned and replaced with a presumption in favour of overruling in cases where a precedent court has taken one view of a set of reasons that was before it, and an overruling court takes a different view of that same set of reasons.[69]
B V Harris points to several factors that support his relatively expansive approach to overruling. First, he argues that to prefer the view of a precedent court to that of an overruling court is, prima facie, arbitrary.[70] Secondly, he considers that confidence in the judiciary might be undermined if too rigid an adherence to precedent is maintained.[71] If not permitted to overrule, judges will find other complex, indirect, and ultimately artificial ways of avoiding the binding force of the rationes decidendi of precedent cases, such as distinguishing precedent cases in circumstances where in truth there are no material factual differences upon which distinguishing could be justified. Thirdly, B V Harris indicates that a judge who refuses to overrule even when he or she believes that justice would be served by doing so neglects his or her oath to ‘do right to all manner of people’.[72] And finally, B V Harris points out that in the common law world, judges of appellate courts have the freedom to dissent, a freedom that is not easily reconciled with a strict adherence to precedent.[73]
Neither J W Harris’ approach nor B V Harris’ approach to overruling is at the extremity of the spectrum of approaches to overruling; accordingly both may be described as moderate. However, B V Harris’ approach is clearly closer to the laissez-faire end of that spectrum than is the approach of his namesake. Given this difference between the two approaches, the question arises: which approach is to be preferred? The answer to this question ultimately turns on the proper relation of the values of certainty and individual justice in the legal system.[74] From one perspective it is arguable that J W Harris’ more constraining approach is to be preferred to that of B V Harris. From this perspective, the circumstances in which overruling is permitted may be ascertained only by studying the attitudes and practices of the interpretive community deploying the technique of overruling: in other words, the attitudes and practices of judges.[75]
Such a perspective entails a commitment to taking an approach to overruling that may be reconciled with what judges in actual legal systems, with their particular cultures and histories, may be observed to believe and do, as opposed to what judges in an ideal legal system organised differently from actual legal systems might hypothetically believe and do.
As we noted in our introduction to this article with reference to the case of Brodie, judges of the High Court of Australia consider themselves free to overrule their previous decisions, but, in common law cases, only in exceptional circumstances.[76] Further evidence that this is how Australian and English judges view the question of overruling more generally may be gleaned from the substantial body of extrajudicial writing and thinking on precedent that such judges, both serving and retired, have produced.[77]
According to Stephen Perry, the practices of common law judges — and here, ‘common law’ is to be distinguished from the law of civilian legal systems — are informed by what he calls a ‘strong Burkean conception’ of precedent: a conception that is characterised by conservatism and permits overruling only where there is a ‘strong reason’ for doing so.[78]
The strong Burkean conception places considerable weight on the principle of stare decisis that traditionally has been associated with judge-made law in countries like Australia; in the words of Bryan Horrigan, it ascribes ‘presumptive force’ to precedent decisions.[79]
If common law judges did not adhere to a strong Burkean conception of precedent, they would not regard it as necessary to overrule precedent (in appropriate cases) in the first place, because precedent would not be regarded as binding.[80]
Perry contrasts this strong Burkean conception with what he calls the ‘weak Burkean conception’,[81]
according to which judges adhere to a ‘presumption of some sort in favour of previously-accepted practices’[82] but regard themselves as free to overrule if their view of the reasons bearing upon a precedent case differs from that of the precedent court.[83]
In overruling cases, the principle of stare decisis has some force, but not much, according to the weak Burkean conception of precedent.
J W Harris’ approach to overruling, and particularly the ‘no new reasons’ constraining principle that he describes, accords with the strong Burkean conception of precedent. This is hardly surprising given that the approach is derived from a close study of the interpretive attitudes and practices of courts in overruling cases. This, in our view, is a reason to favour that approach, and in what follows, we assume that J W Harris’ approach to overruling is a sound one. By contrast, the more laissez-faire approach of B V Harris clearly does not accord with the strong Burkean conception. Indeed, B V Harris’ approach may be irreconcilable with the weak Burkean conception of precedent, because even that conception entails a ‘presumption … in favour of previously-accepted practices’[84]
and this appears to be precisely what B V Harris would reject in overruling cases. Arguably, the principle of stare decisis has only negligible, if any, force in overruling cases, according to B V Harris’ approach. On the spectrum of approaches to overruling, then, B V Harris’ approach may sit somewhere between an extreme laissez-faire approach and the weak Burkean conception. If this is a correct characterisation of his approach, it is not one that reflects the interpretive attitudes and practices of judges — including those serving on courts of final appeal — in the common law world.
In Imbree, the plaintiff (the appellant before the High Court) was in a four-wheel drive motor vehicle on a road trip with his friend, his two sons and a friend of one of the sons, the 16 year old defendant (the first respondent before the High Court). The appellant knew that the first respondent had previously driven a four-wheel drive, but he also knew that the first respondent did not have a learner’s permit and was not licensed to drive. Indeed, during the course of their journey, the appellant attempted, without success, to obtain a learner’s permit for the first respondent. Notwithstanding the appellant’s inability to secure a permit for the first respondent, he allowed the first respondent to drive, which went without incident until the first respondent lost control of the vehicle while trying to avoid hitting a piece of debris on the road, and the vehicle rolled over. The appellant was severely injured and suffered spinal injuries rendering him a tetraplegic.[85]
On these facts, the High Court of Australia had to consider and resolve an issue arising from what has been characterised as an ‘anomaly’ in the law of negligence.[86] What was the standard of care that the first respondent, as a learner driver, owed to the appellant, as a de facto supervisor of the driver who was travelling as a passenger in the vehicle? This issue was not new when Imbree was decided; in fact, it had troubled the High Court, as well as the English Court of Appeal, in previous cases.[87] But it was clear where the Court in Imbree stood in relation to the issue. The plurality, consisting of Gummow, Hayne and Kiefel JJ, was in little doubt that the standard of care owed by a learner driver to a supervisor in the vehicle should be the same standard that all drivers must meet with respect to all persons to whom they owe a duty of care: ‘The standard … is the standard of the “reasonable driver”. That standard is not to be further qualified, whether by reference to the holding of a licence to drive or by reference to the level of experience of the driver.’[88]
Gleeson CJ, Kirby J and Crennan J were of the same view.[89]
The problem was the Court’s 1986 decision in Cook, which stood for the proposition that a learner driver does not owe an unqualified standard of care to a supervisor in the vehicle, but rather owes the standard of care that is ‘reasonably to be expected of an unqualified and inexperienced driver in the circumstances’.[90] By enshrining a qualified standard of care for a learner driver, the ratio decidendi of Cook incorporated a legal proposition that was utterly inconsistent with the proposition, enshrining an unqualified standard, that found favour with the Court in Imbree. On this basis, the Court in Imbree was invited by the appellant to consider overruling its previous decision in Cook,[91] and this it did explicitly.[92] Imbree was, therefore, a ‘copper-bottomed’ overruling case.
The Court in Imbree took three different approaches to the question of overruling Cook. First, there was the approach taken by the plurality — Gummow, Hayne and Kiefel JJ — and agreed to by Gleeson CJ and Crennan J in separate judgments. Secondly, there was the approach taken by Kirby J in his separate judgment. And finally, Heydon J took another approach to overruling in his separate judgment. In the remainder of this Part, we consider each approach in turn, devoting most of our attention to the approach of the plurality, which is likely to be of great influence in future overruling cases before the Court.
The plurality paid homage to two older overruling cases, Commonwealth v The Hospital Contribution Insurance Fund of Australia (‘Hospital Contribution Insurance Fund’)[93] and John v Federal Commissioner of Taxation (‘John’),[94] decided by the Court in 1982 and 1989 respectively. According to the plurality, the decisions of the Court in those two cases made clear that there are four ‘considerations’ bearing upon the question of overruling.[95] The first consideration is whether or not the precedent decision ‘rest[ed] upon a principle carefully worked out in a significant succession of cases’.[96] The second is whether or not ‘there were differences in the reasoning that led to the [precedent decision]’.[97] The third is whether a precedent decision ‘had achieved no useful result but considerable inconvenience’.[98] And the fourth consideration is whether or not a precedent decision had ‘been independently acted on in a manner which militated against reconsideration’.[99]
The four considerations drawn from Hospital Contribution Insurance Fund and John were regarded by the Court as relevant to the question of overruling in cases preceding Imbree, both in common law and other settings.[100] In Imbree, having introduced those four considerations, the plurality continued as follows:
The need to consider these matters is obvious. It is necessary to do that, however, with a clear recognition of more basic principles. In particular, it is necessary to recognise that, when a court of final appeal considers judge-made law, ‘[w]hile stare decisis is a sound policy because it promotes predictability of judicial decision and facilitates the giving of advice, it should not always trump the need for desirable change in the law’ especially, we would add, if the change is necessary to maintain a better connection with more fundamental doctrines and principles.[101]
In this passage, the plurality revealed that, at least in cases where the Court considers ‘judge-made law’, a fifth consideration is also in play when the question of overruling falls to be considered. This fifth consideration requires an assessment of the precedent decision in light of the principles underlying the area of law in which both the precedent case and the overruling case are situated.
The first consideration referred to by the plurality in Imbree was whether or not the precedent decision ‘rest[ed] upon a principle carefully worked out in a significant succession of cases’.[102] In John, a majority of the High Court made clear that underlying this consideration is a concern to avoid a situation in which overruling a precedent decision calls into question the authoritativeness of other decisions that share the same basis in principle.[103] The majority in John took the view that the precedent decision that they were invited to overrule — the decision in Curran v Federal Commissioner of Taxation (‘Curran’)[104] — was what might be called an outlier, sharing no basis in principle with other cases that, at first glance, might have appeared similar to it.[105] Thus, in John, the first consideration did not impede overruling.
In Imbree, the plurality appears to have taken a similar view in addressing the precedent in Cook. The specific question that was before the Court in Cook — what standard of care does a learner driver owe to a supervising passenger? — was unusual. In this narrow sense, Cook cannot be said to have been part of a ‘significant succession of cases’ in Australian law. Rather, it was an outlier.[106] However, in a wider sense, Cook might be thought of as one among a number of cases in both English and Australian law in which the question for determination has been whether or not to recognise an exception to the unqualified standard of care that is typically applied in the tort of negligence.[107] Just as the majority had done in John, the plurality in Imbree distinguished Cook from other cases in which a departure from the unqualified standard of care had been considered, affirmed those other cases, and, in doing so, treated Cook as an outlier even in the wider sense:
what distinguishes the principle established in Cook … is that Cook requires the application of a different standard of care to the one defendant in respect of the one incident yielding the same kind of damage to two different persons [ie, a supervising passenger and a non-supervising passenger] …[108]
For the plurality, then, the first consideration militated in favour of overruling Cook.
With the accounts of overruling from the scholarly literature that we set out in Part II in mind, what can be said about the first consideration referred to by the plurality in Imbree? Recall that, according to the account of overruling offered by J W Harris, a threshold requirement for overruling is that overruling will bring about an improvement to the law, measured in one or more of the dimensions of justice, certainty and coherence. This threshold requirement is also implicit in the more laissez-faire account of overruling presented by B V Harris. Indeed, it is difficult to imagine how a court could rationally overrule a precedent decision if the court did not think that improvement to the law would thereby be brought about. In a case like John, improvement to the law may be brought about by overruling a precedent decision that does not rest on principles carefully worked out over time. However, just because a case is an outlier, it does not necessarily follow that overruling the decision in that case will bring about an improvement to the law. Sometimes, improvement to the law has been brought about precisely because an outlier case has generated a new principle: the great case of Donoghue v Stevenson[109] is one of the best-known examples of this. So the first consideration, if it is to figure in a sound framework of overruling, can do so only indirectly; the fact that a precedent case is an outlier can function as nothing more than an indication that improvement to the law might be achieved by departing from its ratio decidendi. It can bear no greater significance than that.
The second consideration referred to by the plurality in Imbree was whether or not ‘there were differences in the reasoning that led to the [precedent decision]’.[110] As John made clear, this consideration is active where the precedent court was divided because the judges making up that court, or a majority of that court, gave different reasons for reaching their shared orders in the precedent case, or where one or more dissenting judgments were delivered in the precedent case.[111] This occurred in Curran, the case that was overruled in John. A majority of the High Court in Curran had joined in the orders, but Barwick CJ and Menzies J employed similar reasons in their judgments, whereas Gibbs J had given others.[112] Moreover, Stephen J had dissented.[113] Thus, in John, the second consideration was regarded by a majority of the Court as a relevant factor in overruling Curran.[114] By contrast, in Imbree, the second consideration would appear to have played no role in the plurality’s decision to overrule, notwithstanding that it was introduced by the plurality as a relevant consideration and that, in Cook, the reasoning of the plurality[115] was indeed different from that of Brennan J in his separate judgment.[116]
Perhaps the reason why the second consideration appears to have played no role in the overruling of Cook is that, in Cook, the plurality also constituted a clear majority of the court — Mason CJ, Wilson, Deane and Dawson JJ — and so Brennan J’s different reasoning was of marginal significance to the question of the authoritativeness of the case.[117] By contrast, in Curran, the majority was split 2:1, in addition to a dissenting judgment from Stephen J. It may be that, in John, this diversity of reasoning in Curran was viewed as a factor diluting the authoritativeness of its ratio decidendi.
To draw this comparison between Cook and Curran is to imply that in High Court cases some rationes decidendi are more authoritative than others. This implication is problematic. As Brennan J stated in his separate judgment in John:
The authority of this Court’s decisions and the reliance which can and must be placed upon them are not dependent on the constitution of the Court or on the cogency of dissenting reasons. … [A] decision of this Court has authority as a precedent precisely because it is the Court’s decision, not because it is the decision of the participating justices or a majority of them.[118]
Indeed, the whole practice of decision-making by majority that characterises the common law tradition depends for its coherence on a conception of the authoritativeness of rationes decidendi that draws no distinction between a slender majority and a unanimous decision. To the extent that the second consideration expounded by the majority in John, and referred to by the plurality in Imbree, is at odds with that proposition, it must be rejected. Of course, there is one type of case in which the second consideration might have some significance in deciding what to do with a precedent case: where differences in the reasoning of the judges of a precedent court are such that there is no ratio decidendi of the precedent case, notwithstanding that a majority of the precedent court has been able to agree on orders.[119] But in such a case, there being no ratio decidendi, there is nothing for a court in a subsequent case to overrule. It would appear, then, that the second consideration should be of no relevance in overruling cases, despite what was said in Hospital Contribution Insurance Fund, John, and by the plurality in Imbree.
In Imbree, the plurality stated that where a precedent decision ‘had achieved no useful result but considerable inconvenience’, that fact would be relevant to overruling.[120] Neither of the elements of this consideration — ‘no useful result’ and ‘considerable inconvenience’ — is unproblematic, especially when it comes to at least some common law cases. Although in Imbree the plurality stated that ‘the principle adopted in Cook … achieved no useful result’,[121] it is difficult to see why this was so, given that the principle adopted in Cook enabled the precedent court to reach a reasoned and final determination of the dispute between the parties to that case, which is the primary purpose of judicial decision-making. But perhaps a clue to the real meaning of ‘no useful result’ is the reference in the third consideration to ‘considerable inconvenience’: a precedent decision has ‘no useful result’ in circumstances where persons have had to adjust their behaviour to their detriment because the precedent decision has been handed down. In Hospital Contribution Insurance Fund, according to Gibbs CJ, this had indeed occurred; there, precedent decisions of the High Court had generated considerable inconvenience in the operation of state courts.[122]
It is a strange notion that drivers of motor vehicles might be compelled to adjust their behaviour in accordance with the decision of a court setting out the content of the standard of care. After all, individuals tailor the way they drive in accordance with much more immediate concerns such as personal safety, road rules and the fear of criminal sanction.[123] However, it is a little more plausible that the ratio decidendi of Cook did occasion some inconvenience to insurers, which presumably had to take the case into account in planning aspects of their business, costing them money in legal and other expenses.[124] Even if this proposition is accepted — and incidentally there is no compelling evidence why it should be — it leads directly to a deeper question: Why should the convenience of insurers influence a court in reaching its decision when adjudicating a dispute in the law of negligence between two parties, neither of which is an insurer? More broadly, why should considerations of convenience play any role in the determination of a negligence case?[125] The answers to questions of this type depend on the extent to which courts adjudicating common law disputes — at least in cases where the plaintiff alleges a breach of duty by the defendant — should take into account considerations of public policy as well as corrective justice.[126] In common law cases, therefore, it cannot be assumed that the third consideration referred to by the plurality in Imbree should always play a legitimate role in deciding whether or not to overrule. Indeed, to the extent that in common law cases courts ought to reason in accordance with the dictates of corrective justice, those dictates should operate as a constraint on overruling; a constraint that no scholarly account of overruling appears to have contemplated to date.
The fourth consideration referred to by the plurality in Imbree — whether or not the precedent decision has been ‘independently acted on in a manner which militated against reconsideration’[127] — points to two of J W Harris’ constraining principles: that based on the reliance interest, and that based on comity with the legislature.[128]
As B V Harris has pointed out, in the case of the constraint based on comity with the legislature, it is important not to read too much into the decisions of a legislature, especially where the legislature decides not to act following the handing down of a precedent decision.[129]
Moreover, in his discussion of the reliance-based constraint on overruling, J W Harris notes that the constraint operates only in cases where an especially compelling argument based on reliance can be made.[130]
These notes of caution with respect to comity and reliance have been reflected in the approach of the High Court to the fourth consideration. In Hospital Contribution Insurance Fund, Gibbs CJ noted that state legislatures had enacted statutes in reliance on the precedent decisions that the Court was being asked to overrule, but he did not consider that reliance, nor any principle of comity, to be any great impediment to overruling.[131] In John — a taxation case — the majority pointed out that the precedent decision in Curran had been acted on by taxpayers in ordering their affairs, and that this reliance was a ‘powerful’ reason not to overrule; but the majority went on to overrule Curran anyway.[132] And in Imbree, the plurality stated that ‘it must be assumed that [Cook’s] application may have affected the terms on which cases have been compromised and the apportionments of responsibility that have been made by courts and parties.’[133] But having acknowledged that reliance interest, the plurality in the following sentence indicated that they would overrule Cook in spite of it.[134]
As the attitude of the High Court in cases like Hospital Contribution Insurance Fund, John, and even Imbree, shows, the fourth consideration may be expected to impede the exercise of the Court’s overruling power only in the most exceptional type of case. One example is the constitutional case of Queensland v Commonwealth, in which some members of the Court refused to overrule a precedent decision because to do so would ‘defeat the expectations of the people of the Territories that they would be represented … by senators’ in the Commonwealth Parliament.[135] Others are suggested by the reference in the House of Lords’ Practice Statement of 1966 to ‘the danger of disturbing retrospectively the basis on which contracts, settlements of property and fiscal arrangements have been entered into’.[136] However, beyond those types of cases — for example, in cases arising in the law of tort — it would seem that the High Court is not easily swayed by a reliance-based or comity argument when considering whether or not to overrule itself. The fourth consideration referred to by the plurality in Imbree is therefore of relevance, but limited effect, in overruling cases.
The fifth consideration referred to by the plurality in Imbree was whether or not ‘change is necessary to maintain a better connection with more fundamental doctrines and principles.’[137] In taking the fifth consideration into account, the plurality engaged with the question of fundamental principles in the law of negligence in two ways: by addressing proximity and knowledge of inexperience.
First, the plurality pointed to the fact that the plurality in Cook had couched its decision in terms of a unifying concept of the ‘proximity of relationship’ between plaintiff and defendant.[138] In High Court cases preceding Cook, Deane J had introduced and then developed the idea of proximity as a ‘touchstone’ in the tort of negligence,[139] and in Cook itself, a majority of the Court endorsed Deane J’s reasoning for the first time.[140] In stating its reasons for finding that a learner driver owes a qualified standard of care to a supervising passenger, the plurality in Cook invoked the concept of proximity a number of times.[141] This invocation of the concept of proximity by the plurality in Cook was referred to by the plurality in Imbree because, between the time when Cook was decided and the time when Imbree came before the High Court, the Court had abandoned its former view that the notion of proximity helped to explain the circumstances in which a duty of care might arise for the purposes of the tort of negligence.[142] Given that the Court no longer adhered to a proximity-based framework for analysis in negligence cases, one question for the plurality in Imbree was whether Cook — a decision that explicitly relied on the proximity concept, even though it turned on the standard of care rather than the duty of care — was now out of alignment with the fundamental principles underlying negligence law.
There is little doubt that in circumstances where a precedent decision rests on a legal principle that has, since the precedent case was decided, been erased from the law, the fact that the bedrock of principle has changed constitutes a reason for overruling the precedent case. In such circumstances, the disjunction of the precedent decision and considerations of fundamental principle means that departing from the ratio decidendi of the precedent case would bring about improvement to the law, measured in the dimension of coherence. Moreover, it is clear that this changed legal landscape is a ‘new reason’ in the sense meant by J W Harris in his article on overruling, because it is a reason that was not before the precedent court. It follows that, where the bedrock of principle has changed between the time when a precedent case was decided and the time when it comes under review in an overruling case, J W Harris’ ‘no new reasons’ constraint is not activated. The overruling court is therefore free to overrule on J W Harris’ account of overruling, so long as no other constraints are operative. A fortiori, a court in such a case is free to overrule on the more laissez-faire approach to overruling set out by B V Harris.
However, in Imbree, although the plurality noted that the fundamental principles of Australian negligence law had changed since the time when Cook was decided, it concluded that this was an insufficient reason to overrule Cook.[143] This was because, according to the plurality, sense could be made of the ratio decidendi of Cook without having recourse to the notion of proximity.[144] In Cook, Brennan J, in his separate judgment, had expressly rejected proximity-based reasoning, but nonetheless formed the view that a learner driver owes a qualified standard of care to a supervising passenger based on that passenger’s knowledge of the driver’s inexperience or ‘disabling condition’.[145] For the plurality in Imbree, this line of reasoning could also be discerned in the judgment of the plurality in Cook.[146] So, for the plurality in Imbree, the references to proximity in the judgment of the plurality in Cook were a ‘red herring’, distracting attention from the fact that the ratio decidendi of that case in truth rested on reasoning about a supervising passenger’s knowledge of a learner driver’s inexperience.[147] In Imbree, the abandonment of the idea of proximity in Australian negligence law might have grounded a ‘new reason’ that could justify overruling Cook even on the relatively constraining approach to overruling presented by J W Harris. However, given the way in which the plurality in Imbree interpreted the judgment of the plurality in Cook with respect to the proximity point, that change in fundamental principle could not be said to ground any new reason for the plurality in Imbree.
The plurality in Imbree addressed the question of connection with fundamental principles in a second way. Having dealt with the proximity point, the plurality continued:
To reject knowledge of inexperience as a sufficient basis upon which to found a different standard of care is to reject the only basis, other than proximity, for the decision in Cook … Yet rejection of knowledge as a basis for applying a different standard of care is required not only by the observation that knowledge of inexperience is held not to affect the standard of care owed to other passengers or other road users who observe a display of L-plates, but also by the essential requirement that the standard of care be objective and impersonal.[148]
In this passage, the plurality indicated that, in its view, a standard of care qualified in accordance with knowledge of inexperience is at odds with the fundamental principles underlying negligence law, among which is the principle that the standard of care be unqualified by the personal attributes or circumstances of the defendant.[149]
It is true to say that a fundamental principle of negligence law is that the standard of care is typically unqualified,[150] but it is not true that the standard of care is inevitably so. Certain defendants — such as children and skilled professionals — owe a qualified standard of care in Australian law, and this has long been the case.[151] So the real question for the plurality in Imbree was whether the first appellant’s knowledge of the respondent’s inexperience was a sufficient reason to recognise the respondent as being among the class of defendants who owe a qualified standard of care, at least with respect to the first appellant. The plurality in Imbree thought not.[152] In Cook, the plurality, along with Brennan J, thought that knowledge of inexperience was a sufficient reason to recognise a qualified standard of care.[153] Clearly, on this question of admitting a defendant into the class of defendants owing a qualified standard of care, the view of the plurality in Imbree was different from the view of the Court in Cook in materially identical cases. But does that mean that the Court in Cook failed in some way to adhere to fundamental principles of negligence law?
In Cook, the question of whether the Court should depart from the unqualified standard of care that prevails in negligence law was placed squarely before the Court by counsel,[154] and it was the subject of considerable discussion in the judgments of the plurality and of Brennan J. Those judgments reveal a sound awareness of fundamental principle — that the standard of care is typically unqualified, but that there are exceptions — and a clear awareness of what was at stake in the case. Indeed, the plurality discussed the scope of the qualified standard of care that it recognised in its decision, taking care to ensure that it would not be applied in future cases beyond its proper limits.[155] In our view, there is no basis for asserting that the decision in Cook failed to connect with fundamental principles of negligence law.
It might be thought that the plurality in Imbree was asserting nothing as bold as that, but rather was making the more modest point that the decision in Cook connected with fundamental principles of negligence law in the wrong way, by admitting an exception to the unqualified standard of care in circumstances where no such exception was warranted. If that is what the plurality in Imbree was saying in its discussion of the ‘knowledge of inexperience’ point, then clearly it deployed the fifth consideration — that relating to connection with fundamental principles — in reaching the conclusion that departing from the ratio decidendi of Cook would bring about an improvement to the law. We alluded above[156] to the capacity of the fifth consideration to draw judicial attention to the possibility that the law might be improved by bringing a particular legal rule into alignment with broader doctrinal shifts between the time when a precedent case was decided and the time when a court is asked to overrule. The fifth consideration is also attuned to an inquiry into whether or not the law would be improved in circumstances where it is arguable that a precedent decision was, even at the time when it was decided, out of alignment with considerations of fundamental principle. This, in effect, is what the plurality in Imbree thought was the case in Cook.
However, if the plurality in Imbree thought that the law would be improved by overruling Cook because in Cook the Court adopted a rule that was out of alignment with fundamental principles of the law of negligence, then in our view another problem emerges. Consider again this passage from J W Harris’ article on overruling, which we discussed in detail in Part II(A):
When all the reasons canvassed for or against a proposition of law in the [overruling] case were also canvassed in the decision under review, then what remains for the judge is a subjective weighing of them one against the other.[157]
For J W Harris, it is in precisely such circumstances that the ‘no new reasons’ constraint operates, rendering impermissible an exercise of a court’s overruling power notwithstanding that, in the opinion of the overruling court, improvement to the law would thereby be achieved. In our view, these circumstances were present in Imbree with respect to the judgment of the plurality, if the plurality is to be taken as asserting that in Cook the Court engaged wrongly with considerations of fundamental principle. On the ‘knowledge of inexperience’ point, the plurality in Imbree had no reasons before it that had not been before the precedent Court in Cook. What distinguished the reasoning of the plurality in Imbree from that of the Court in Cook was that the plurality in Imbree placed greater weight on the fact that the standard of care in negligence law is typically unqualified, whereas the Court in Cook placed greater weight on the limited but, in its view, appropriate exceptions to this rule. In these circumstances, the plurality in Imbree overruled Cook contrary to what was permissible according to the approach to overruling set out by J W Harris.
By contrast, according to the more laissez-faire approach to overruling proposed by B V Harris, it is likely that the plurality in Imbree overruled Cook permissibly. According to B V Harris, in a case like Imbree, where the precedent court has taken one view of the set of reasons before it, and the overruling court takes a different view of the same set of reasons, the overruling court should not only be free to overrule, but there should be a presumption in favour of overruling, in the interests of individual justice.[158]
The values associated with the doctrine of precedent and the principle of stare decisis simply should not be regarded as important enough to carry the day. If, as we have argued, the plurality in Imbree overruled Cook because its impression of the set of reasons that had been before the Court in Cook was different from the impression formed by the Court in that case, then the correctness of the decision of the plurality in Imbree on the overruling point depends on the correctness of B V Harris’ approach to overruling. As we pointed out in Part II(C), there are reasons for preferring J W Harris’ more constraining approach to overruling to B V Harris’ more laissez-faire approach: in particular, these reasons are grounded in the attitudes and practices of judges in the common law tradition.[159] To that extent, there are reasons for thinking that the plurality in Imbree should not have overruled Cook, given that the basis of its decision to overrule was a different impression of reasons that had been before the Court in Cook.
Like the plurality in Imbree, Kirby J thought that Cook should be overruled;[160] but unlike the plurality, he was not prepared to overrule Cook based simply on a different appreciation of the balance of reasons that had been before the precedent court.[161] According to Kirby J:
For this Court to overturn Cook and to substitute a single, uniform and objective standard as the criterion for the existence and ambit of the duty of care, owed by one motorist to third parties, a new legal ingredient is necessary.[162]
By implying that overruling is justified only where a ‘new legal ingredient’ is before the overruling court, Kirby J showed in Imbree that he inclined more towards a constraining view of overruling like that set out by J W Harris than towards the relatively laissez-faire view proposed by B V Harris. Indeed, it would seem that Kirby J’s demand for a new legal ingredient was a straightforward statement of the ‘no new reasons’ constraint that, according to J W Harris, precludes overruling where it is not overcome.[163]
In Part II(C), we set out the reasons for favouring an approach to overruling that incorporates a ‘no new reasons’ constraint: it follows that Kirby J’s approach to overruling in Imbree is to be preferred to that of the plurality.
But what exactly was the new legal ingredient that Kirby J thought he had identified in Imbree as a reason to overrule Cook? Kirby J stated that the new legal ingredient was ‘the existence of compulsory motor vehicle third party insurance’, in place in Australia for approximately 60 years by virtue of statutory enactments.[164] Such insurance is the product of deliberate legislative decisions to allocate liability for injuries in motor vehicle accidents indirectly to all road users rather than directly to the parties to particular negligence claims. It would seem that on this basis, for Kirby J, courts ought not to be overly vigilant to protect drivers whose actions cause injuries in circumstances where those drivers will not have to bear the costs of their actions.[165] So much may be gleaned from Kirby J’s reference to the existence of compulsory insurance as a ‘persuasive reason for departing from the individual culpability principle’, which, in his view, animated the decision in Cook.[166] It has traditionally been thought that insurance-based arguments may not be made when considering questions of liability in the tort of negligence.[167] Nonetheless, there are reasons for thinking that, at least when it comes to motor vehicle accidents, the existence of the loss-spreading mechanisms of compulsory insurance ought to be of relevance when determining liability.[168]
Whether or not an argument based on the existence of compulsory insurance ought to be accepted in a negligence case is not relevant for present purposes. Our present interest is in whether or not the existence of compulsory insurance grounded a new legal ingredient in Imbree that justified overruling Cook.
Despite what he said, it cannot be that Kirby J thought that compulsory insurance was a new legal ingredient in the sense that, although such insurance existed at the time when Imbree was decided, it did not exist at the time of Cook. Kirby J himself noted more than once that compulsory insurance schemes had existed in Australia for some 60 years before Imbree was decided.[169] It is equally difficult to accept that Kirby J thought that, before Imbree, courts were not formally aware of the existence of compulsory insurance schemes because arguments based on such schemes were never made by counsel. Kirby J pointed out that the question of compulsory insurance was referred to in the reasons of Lord Denning MR in Nettleship v Weston,[170] and elsewhere in his judgment he stated that ‘[t]he relevance of the added consideration of compulsory insurance has been discussed in various cases.’[171] Moreover, in Cook itself, the plurality referred to Lord Denning MR’s view in Nettleship v Weston that compulsory insurance was a relevant factor in that case, and the plurality specifically stated that the view was ‘not one which should be adopted by courts in this country’.[172] In any event, whether or not the Court in Cook had the benefit of counsel’s arguments on compulsory insurance, the Court in Imbree did not have the benefit of such arguments:[173] consequently, a new legal ingredient in Imbree could not be grounded in such arguments.
A more plausible interpretation of Kirby J’s judgment in Imbree is that he thought that ‘a growing preparedness of the courts to acknowledge the influence of insurance, at least where it is compulsory and provided by statute, in defining the content of legal liability’[174] constituted a new reason in Imbree. We do not doubt that such a consideration may, in an appropriate case, function as a new reason, or, as Kirby J put it, a new legal ingredient for an overruling court. However, it is not clear that, since Cook was decided in 1986, Australian courts have manifested as great a willingness to permit insurance considerations to influence the content of the law of negligence as Kirby J suggested in Imbree. Certainly, as Kirby J pointed out, in exceptional cases since Cook, judges in Australian courts have taken compulsory insurance into account in making findings about liability in the tort of negligence.[175] However, insurance arguments have simply not figured in the great majority of decisions, which have been decided in accordance with the traditional view that insurance is irrelevant to liability questions.[176] In other decisions, judges, including members of the High Court, have explicitly rejected any role for insurance considerations in negligence cases.[177] It is arguable that, in view of the case law either remaining silent about or explicitly rejecting insurance considerations, Kirby J provided insufficient evidence to support his claim about the ‘growing preparedness of the courts to acknowledge the influence of insurance’.[178]
In our view, Kirby J failed to substantiate his view that the new legal ingredient that he sought for overruling Cook was to be found in the existence of compulsory motor vehicle third party insurance. Instead, his decision to overrule was, like the plurality’s, based ultimately on his different impression of the weight of a reason — the existence of compulsory insurance — that had been before the Court in Cook and that had been explicitly rejected by the plurality in the earlier case. As we discussed above with reference to the reasoning of the plurality in Imbree,[179] overruling based on a different impression of the balance of reasons before the overruling court is permissible only on the relatively laissez-faire approach to overruling proposed by B V Harris; as we argued above, there are reasons to doubt the appropriateness of that approach. With regard to Kirby J, we might add that, although his Honour stated clearly that he sought a new legal ingredient to support overruling Cook and considered that, even in light of such a new legal ingredient, overruling Cook was, in his view, a ‘close run thing’,[180] it would appear that he overruled Cook even though he did not in fact identify any new legal ingredient in the case before him. As a result, it is arguable that Kirby J’s reasoning in Imbree did not satisfy his own stricture.
Heydon J agreed with the rest of the Court that the first respondent in Imbree had fallen short of the standard of care that he owed to the appellant.[181] However, in Heydon J’s view, this finding could be supported even if the first respondent owed to the appellant the qualified standard of care that had been endorsed in Cook.[182] Because Heydon J was of the view that even this qualified standard of care had not been met on the facts of the case, he did not think it necessary to decide whether or not Cook should be overruled.[183] In essence, unlike the other members of the Court, Heydon J did not accept that Imbree was a ‘copper-bottomed’ overruling case. He therefore declined to make any findings with respect to the correctness or otherwise of Cook.
In Imbree, Heydon J took seriously the ‘mootness’ constraint discussed by J W Harris. Because, for Heydon J, it was not necessary to decide the question whether or not to overrule Cook in order to dispose of the appeal before him, he regarded himself as constrained not to overrule. We alluded earlier to the fact that J W Harris expressed concerns about the rigour of the mootness constraint but considered that it was an actual constraint based on his examination of judicial attitudes and practices. Implied in these concerns is the suggestion that, in an appropriate case, a court of final appeal may have a duty to confront the overruling question squarely, notwithstanding that it is not strictly necessary to do so in order to determine the particular appeal before the court. Such a duty would presumably be grounded in the value of orderly and timely development of the law, and would have regard to the fact that a court of highest appeal, such as the High Court of Australia, is likely to hear an appeal on any given particular legal issue only infrequently. In these circumstances, the duty of a court of highest appeal may be to seize rare opportunities to develop the law when they arise, and not to defer legal development to some future date that may never transpire.[184]
We do not have space in this article to consider whether courts of final appeal owe any such duty and, if so, its exact content. For present purposes, we are content to argue that it is possible that such a duty does exist, and to point to the need for further scholarship to elucidate it.
As we discussed in Part III, the current approach of the High Court of Australia to overruling in common law cases would appear to be grounded in the four considerations first discussed in Hospital Contribution Insurance Fund and John, as well as a fifth consideration — whether or not ‘change is necessary to maintain a better connection with more fundamental doctrines and principles’[185] — formulated in this way for the first time by the plurality in Imbree. This approach is not without advantages; it concentrates the mind on specific reasons for and against overruling in the setting of a particular case, rather than inviting speculation about vague questions such as whether a precedent decision was ‘manifestly wrong’.[186]
In particular, the fifth consideration introduced in Imbree is likely to be useful to the Court when asking itself the threshold question whether or not departing from the ratio decidendi of a precedent case will bring about an improvement to the law. And the first consideration — whether or not the precedent case was an outlier — may also be of some indirect assistance in answering that threshold question in certain cases.
Nonetheless, as we endeavoured to demonstrate in Part III, the approach of the plurality in Imbree is in some respects problematic. The second of the five considerations referred to by the plurality — whether or not ‘there were differences in the reasoning that led to the [precedent decision]’[187] — should be of no relevance to the question of overruling, given a proper understanding of the authoritativeness of decisions of appellate courts. The third consideration — whether or not a precedent decision ‘achieved no useful result but considerable inconvenience’[188] — may sometimes be at odds with the philosophical underpinnings of the common law. And the fourth consideration — whether or not a precedent decision has ‘been independently acted on in a manner which militated against reconsideration’[189] — is likely to be of limited effect in common law cases.
Moreover, it is arguable that in some respects the five considerations raised by the plurality in Imbree are inconsistent with the fundamentals of a sound approach to overruling as set out by J W Harris: a threshold requirement of improvement to the law; and a ‘no new reasons’ constraining principle.
First, although, as we noted above, the fifth consideration (as well as the first, to a degree) is of use in answering the threshold question whether or not the law will be improved by overruling, the third and fourth considerations are precisely the types of concerns that are not relevant to that threshold inquiry. As we discussed in Part II, such an inquiry must focus on what Horrigan calls ‘substantive’ reasons; that is, the reasons for and against the legal proposition that constitutes the ratio decidendi of the precedent case. The third and fourth considerations referred to by the plurality in Imbree have the capacity to draw an overruling court’s attention away from such substantive reasons, and towards the precedential reasons that operate as constraints on overruling, before the prima facie case for improvement to the law has been made out.
Secondly, and more importantly, all five considerations referred to by the plurality in Imbree may militate in favour of overruling in circumstances where no new reasons are before the overruling court that could justify overruling the precedent decision; in other words, cases may arise in which all five considerations support a decision to overrule, notwithstanding that J W Harris’ ‘no new reasons’ constraint is operative. In Part III, we argued that this occurred in Imbree: there, the plurality did not point to any new reasons in its judgment and ultimately overruled Cook based on a different impression of the balance of reasons that had been before the Court in Cook; whereas Kirby J purported to point to a new reason in the existence of compulsory motor vehicle third party insurance, but did too little to demonstrate that it was a genuine new reason.
If the High Court is to adhere to a sound approach to overruling in future common law cases, it will do well to look further afield than Imbree, to cases in the common law setting in which the threshold requirement of improvement to the law has clearly been satisfied and the ‘no new reasons’ constraint has clearly been overcome. At least two such cases may be identified in the relatively recent history of the Court: Esso and Brodie. In the remainder of this Part, we look at these two cases through the lens of J W Harris’ account of overruling in order to place Imbree in a wider context.
In Esso, the High Court was called upon to state the common law test for legal professional privilege in relation to the discovery and inspection of confidential written communications between lawyers and their clients. In the precedent case of Grant v Downs (‘Grant’),[190] a majority of the Court — Stephen, Mason and Murphy JJ — had found that, as a matter of common law, legal professional privilege ought to attach only to ‘those documents which are brought into existence for the sole purpose of submission to legal advisers for advice or for use in legal proceedings’.[191] This is known, in shorthand, as the ‘sole purpose test’. By contrast, Barwick CJ had indicated his preference in the following passage:
I have come to the conclusion that the Court should state the relevant principle as follows: a document which was produced or brought into existence … with the dominant purpose … of using it or its contents in order to obtain legal advice or to conduct or aid in the conduct of litigation … should be privileged and excluded from inspection.[192]
This test, which was adopted in 1995 in Commonwealth and New South Wales legislation,[193] is known in shorthand as the ‘dominant purpose test’. In Esso, the Court was invited to overrule Grant and substitute the dominant purpose test that had been preferred by Barwick CJ for the sole purpose test endorsed by the majority in the precedent case.[194]
The High Court in Esso was divided on the question of whether overruling Grant would bring about an improvement to the law. The plurality — Gleeson CJ, Gaudron and Gummow JJ — clearly thought that departing from the sole purpose test for which Grant stood, and adopting instead a dominant purpose test, would improve the law.[195] Discussing the first of the four considerations from Hospital Contribution Insurance Fund and John that later found favour with the plurality in Imbree — whether or not the precedent decision ‘rest[ed] upon a principle carefully worked out in a significant succession of cases’[196] — the plurality in Esso noted that Grant overturned what was then settled principle relating to legal professional privilege,[197] that the dominant purpose test was preferred in other common law jurisdictions and that the Commonwealth and New South Wales parliaments had adopted the dominant purpose test in legislation.[198] The plurality did not employ the remaining three considerations from Hospital Contribution Insurance Fund and John, instead stating its view that the dominant purpose test struck an appropriate balance between the confidentiality of lawyer–client communications and transparency in the administration of justice.[199] The plurality further stated that the dominant purpose test could in practice be applied with ‘reasonable certainty’ and that it would not entail undue expense and delay.[200] Moreover, the plurality set out an argument that the sole purpose test tended to produce extreme consequences that could be avoided only by interpreting the test in a way that effectively turned it into a dominant purpose test.[201]
The other member of the majority, Callinan J, chose to address the question of improvement to the law using the framework of all four considerations from Hospital Contribution Insurance Fund and John. However, although Callinan J drew on those four considerations, he also pointed to considerations of convenience, fairness, truthfulness in lawyer–client communications and equality of treatment as reasons in favour of the dominant purpose test.[202]
In their dissenting judgments in Esso, McHugh J and Kirby J made clear that in their view, introducing the dominant purpose test to the common law of legal professional privilege would not satisfy the threshold requirement of improvement to the law. In extensive passages of his judgment, McHugh J presented arguments as to why a dominant purpose test would unjustifiably narrow the range of discoverable material, would be difficult to apply and would be inconsistent with the underlying rationale for legal professional privilege.[203] Kirby J canvassed both sides of the argument. He first rehearsed a range of reasons for changing to the dominant purpose test: the difficulties in identifying when a person is motivated by a sole purpose; the change to the law that Grant had occasioned; developments in other jurisdictions and in statutory law; and the apparent workability of the dominant purpose test where it is applied.[204] Kirby J then turned to the reasons for adhering to the sole purpose test: the test had stood as part of the common law for some years; the tendency of the common law had been to narrow the scope of legal professional privilege; statutory reform in the area had extended only so far; the abandonment of the sole purpose test would have wide practical ramifications; a dominant purpose test would advantage corporate litigants at the expense of individuals; and a sole purpose test accorded better with the fundamental purpose of legal professional privilege.[205] He concluded that these reasons for adhering to the sole purpose test outweighed the reasons in favour of change.[206]
For present purposes, it is unimportant whether, in Esso, the majority or the dissenters were correct on the question of improvement to the law. Far more important is the fact that both the majority and the dissenters squarely addressed the question of improvement to the law, examining in detail the various reasons for and against the sole and dominant purpose tests. To this extent, both the majority and the dissenters adhered to a sound approach to overruling. Moreover, in considering the reasons for and against the dominant purpose test, none of the members of the Court in Esso confined themselves to the four considerations from Hospital Contribution Insurance Fund and John. This enabled a direct and exhaustive treatment of relevant considerations, and it enabled the Court to keep the threshold inquiry into the substantive reasons for and against departing from the ratio decidendi of Grant separate from the further inquiry into precedential reasons in the form of constraints on overruling a precedent decision. Esso is therefore a good illustration of a sound approach to the threshold requirement of improvement to the law in an overruling case.
By contrast, in Imbree, the prima facie question of improvement to the law was addressed by the plurality using the framework of five considerations: the four from Hospital Insurance Contribution Fund and John, plus the additional consideration introduced by the plurality that referred to connection with fundamental principles. As we argued earlier,[207] the fifth consideration that was introduced by the plurality in Imbree was well-suited to an inquiry into improvement to the law, and was deployed effectively to that end by the plurality. However, in other respects, the framework for analysis used by the plurality in Imbree is capable of distracting attention from the threshold inquiry into the question of improvement to the law, before the case for improvement to the law has been made out. In our view, in future overruling cases the High Court is likely to adhere more closely to a sound approach to overruling if it sets to one side the four considerations from Hospital Contribution Insurance Fund and John while pondering the threshold question of improvement to the law. The fifth consideration introduced in Imbree may be of assistance to that prima facie inquiry.
As we argued in Part III, in Imbree the plurality was not sufficiently sensitive to the ‘no new reasons’ constraint set out by J W Harris in his account of overruling. Moreover, Kirby J, who was sensitive to this constraint, may not have done enough in his judgment to demonstrate sufficient new reasons to justify overruling on the facts of that case. In Esso, by contrast, as well as arguing that improvement to the law would be brought about by overruling Grant, the plurality demonstrated sensitivity to the ‘no new reasons’ constraint by carefully identifying reasons that were before it that had not been before the Court in Grant.[208] Esso is therefore readily distinguishable from Imbree in relation to the attention given to the ‘no new reasons’ constraint.
The plurality in Esso pointed out that before Grant was decided, ‘the prevailing view was that it was sufficient to attract privilege to … reports if one purpose of their preparation was to obtain legal advice or assistance.’[209] The question for the High Court in Grant was therefore whether to uphold this ‘one purpose test’ or to replace it with a sole purpose test; the majority in Grant chose the sole purpose test. Importantly, as the plurality in Esso pointed out, no argument had been heard in Grant on whether to adopt this sole purpose test over a dominant purpose test. The option of a dominant purpose test simply had not been put formally to the Court in that case.[210] Barwick CJ’s judgment, favouring a dominant purpose test, was written without the benefit of counsel’s arguments on the point, and his comments on the dominant purpose test were obiter dicta.[211] By contrast, in Esso, the question for determination was whether to adopt a sole purpose test or a dominant purpose test, and detailed argument was heard on the relative advantages and disadvantages of the two tests.[212] A range of reasons relating to those relative advantages and disadvantages was therefore before the Court in Esso that had not been before the Court in Grant. In these circumstances, the ‘no new reasons’ constraint was amply overcome.[213] On J W Harris’ account of overruling, the Court in Esso was therefore free to overrule in the absence of constraints based on reliance, comity or mootness.
Esso illustrates one type of new reason that the High Court might look for in the future in order to satisfy itself that the ‘no new reasons’ constraint is overcome in an overruling case: a reason generated by counsel’s argument on a point that was not argued before the Court in the precedent case. Further, Esso is an example of an overruling court proving itself to be alert to the need to identify a new reason in order to justify overruling. To this extent, Esso aligns easily with the account of overruling proposed by J W Harris; an account that we argued for in Part II. By contrast, as we argued in Part III, the plurality in Imbree was not alert to the requirement that there be a new reason to justify overruling, and Kirby J, who was alert to that requirement, fell short of satisfying it in his judgment. In our view, then, the High Court is likely to gain sounder guidance from Esso than from Imbree when considering constraining principles in future overruling cases in the common law setting.
The issue that was before the High Court in Brodie arose because in two earlier cases — Buckle v Bayswater Road Board (‘Buckle’)[214] and Gorringe v Transport Commission (Tas) (‘Gorringe’)[215] — the Court had recognised an old English common law rule conferring upon highway authorities immunity from suit in respect of nonfeasance in constructing, maintaining and repairing roads and related infrastructure.[216] This rule — the ‘highway immunity rule’ — was called into question in Brodie, a case in which the nonfeasance of a highway authority had caused a bridge to collapse, injuring a truck driver and causing damage to the truck he was driving. In Brodie, the Court considered whether or not to depart from the highway immunity rule and reconcile highway authority nonfeasance cases with the mainstream of Australian negligence law; at the same time, it explicitly considered whether or not to overrule Buckle and Gorringe.[217] Like Esso and Imbree, Brodie was therefore a ‘copper-bottomed’ overruling case.
Gaudron, McHugh and Gummow JJ, who formed the plurality in Brodie, undertook an exhaustive study of the reasons for and against the highway immunity rule, concluding that the reasons for departing from the rule outweighed the reasons for retaining it.[218] The plurality therefore formed the view that improvement to the law would be achieved by departing from the rule, and to the extent that Buckle and Gorringe stood in the way of departing from the rule, the plurality was prepared to overrule them.[219] The plurality in Brodie referred in its judgment to the four considerations set out in Hospital Contribution Insurance Fund and John.[220] However, as was the case in Esso and Imbree, the plurality in Brodie did not confine its analysis to those four considerations; it took directly into account a large number of other considerations bearing on the question whether or not departing from the highway immunity rule would improve the law. In his separate judgment, as he had done in Esso, Kirby J addressed in great detail the various reasons for and against departing from the highway immunity rule, concluding that the reasons in favour of departing from the rule outweighed those against.[221] Like the plurality, he was prepared to overrule Buckle and Gorringe.[222]
Gleeson CJ was explicit about the need to satisfy a threshold requirement of improvement to the law: ‘The question is whether the law would be better without [the highway immunity rule]’.[223] However, for Gleeson CJ, it was not clear that abolishing the highway immunity rule would improve the law: first, because of the effect that this would have on the interpretation of statutes conferring powers on highway authorities;[224] and secondly, because of the financial consequences for highway authorities and, indirectly, the public.[225] Considerations of certainty and coherence most likely animated the first of these considerations, and considerations of justice most likely animated the second.[226] The other two members of the Court in Brodie — Hayne J and Callinan J — declined to take up the question whether overruling Buckle and Gorringe would bring about improvement to the law.[227]
Although the plurality in Brodie drew on the four considerations from Hospital Contribution Insurance Fund and John in exploring the reasons for and against overruling, it did not confine itself to that framework. Instead, in accordance with a sound approach to overruling, the plurality addressed all relevant considerations. By doing so, the plurality ensured that its reasoning was not distorted by adherence to the John framework, a framework that tends to blur preliminary considerations relating to the substance of the law and further considerations concerning the effects of overruling per se. In this regard, Brodie will be a useful guide to the Court in future overruling cases when considering the question of improvement to the law.
In Brodie, a number of the judges in the High Court constructed arguments to the effect that reasons were before them that had not been before the Court in either Buckle or Gorringe. The plurality looked to developments in common law jurisdictions other than Australia after the time when the precedent cases had been handed down.[228] Although he did not address the question of overruling Buckle and Gorringe, Hayne J pointed out that no challenge to the highway immunity rule was mounted in either of those cases.[229] The implication of this observation was that in neither precedent case did the Court have the benefit of counsel’s arguments on the relative advantages and disadvantages of that rule. Moreover, several members of the Court observed that the tort of negligence had developed substantially in Australian law in the period between 1950, when the second precedent case of Gorringe was decided, and 2001, when Brodie was decided. The plurality pointed out that in a series of cases in the 1980s and 1990s, the High Court had developed the tort of negligence to the point where a public authority could be said in some circumstances to owe a duty of care in the exercise of statutory powers.[230] Moreover, the plurality noted that the highway immunity rule rested originally on an understanding of the relationship between nuisance and negligence which was no longer accepted in Australian law.[231] Kirby J stated that the highway immunity rule had been ‘overtaken by profound developments of the tort of negligence’;[232] and Hayne J thought that developments in the tort of negligence had been so great since the time when the precedent cases were decided that Brodie could be determined as a case in negligence law without any reference to Buckle and Gorringe.[233]
The arguments put by members of the Court in Brodie to the effect that developments in the law of negligence since 1950 weighed in favour of overruling Buckle and Gorringe were arguments sensitive to the need to overcome the ‘no new reasons’ constraint that figures in a sound account of overruling. In effect, the members of the Court who alluded to developments in negligence law were asserting that those developments constituted, or grounded, new reasons that had not been before the Court either in Buckle or in Gorringe. As we pointed out in Part III in our discussion of Imbree, large developments in legal doctrine, including developments in the fundamental principles, may in appropriate cases constitute new reasons in the sense contemplated in J W Harris’ account of overruling. Arguably, Brodie was an appropriate case.
Turning to Imbree by way of contrast, the fifth consideration referred to by the plurality in that case — whether or not ‘change is necessary to maintain a better connection with more fundamental doctrines and principles’[234] — is capable of being interpreted in a manner that is sensitive to the ‘no new reasons’ constraint, as it refers, inter alia, to circumstances where a precedent decision has been superseded by doctrinal developments. But recall that in Imbree, the plurality rejected the only contender for a doctrinal development that could ground a new reason sufficient to overcome the ‘no new reasons’ constraint when it rejected the proposition that the ratio decidendi of Cook incorporated the proximity reasoning that the Court departed from in the late 1990s.[235] So, in a sense, in Imbree itself the fifth consideration failed to live up to its potential as a means of overcoming the ‘no new reasons’ constraint. In future overruling cases in the common law setting, the High Court may be best able to adhere to a sound account of overruling by seeking guidance from Brodie as to the circumstances in which the ‘no new reasons’ constraint is overcome, and from Imbree as to the circumstances in which it is not overcome.
Writing on overruling in the High Court of Australia presents unique challenges. The cases in which the Court has explicitly considered the question of overruling are various, and each turns on its facts. Nonetheless, some conclusions may be drawn with respect to the attitude and practice of the Court when it deals with overruling cases in the common law setting. First, to the extent that the Court exhibits what Kirby J has described as an ‘inclination … towards restraint’[236] in the matter of overruling itself, it adheres to the strong Burkean conception of precedent that has been described by Stephen Perry.[237]
In our view, this is a fitting approach to overruling for a court of final appeal in a common law country to take to overruling. Secondly, to the extent that the Court has regard to a threshold requirement that overruling bring about some improvement to the law, and a constraint that is operative except where a ‘new reason’ is before the Court in an overruling case, the Court adheres to sound principles of overruling — principles that we discussed in Part II with reference to the work of J W Harris — within a broad strong Burkean conception of precedent. As we argued in Part IV, in cases like Esso and Brodie, the threshold requirement and the ‘no new reasons’ constraint appear to have been taken seriously; those cases provide guidance to the Court when dealing in the future with overruling cases, at least in the common law context.
Imbree, on the other hand, may in some respects represent a departure by the Court from sound principles of overruling in a common law case. The framework that the plurality adopted for considering whether or not to overrule Cook — made up of the four ‘considerations’ from Hospital Contribution Insurance Fund and John, plus an additional consideration formulated by the plurality in Imbree — focussed attention on reasons for and against overruling, and to that extent is to be welcomed. The first of those considerations — whether or not the precedent decision ‘rest[ed] upon a principle carefully worked out in a significant succession of cases’[238] — is able, indirectly, to provide guidance in the threshold inquiry as to improvement to the law. And the fifth consideration — which is concerned with connection to fundamental principles[239] — is especially well-suited to that threshold inquiry. The latter consideration should figure in whatever framework the Court adopts in future common law overruling cases.
However, as we argued in Part III, other considerations adopted by the plurality in Imbree may have no place in a framework that accords with a sound approach to overruling. The second consideration — whether or not ‘there were differences in the reasoning that led to the [precedent decision]’[240] — should be discarded. The third — whether or not a precedent decision ‘achieved no useful result but considerable inconvenience’[241] — may have only a limited role to play in those common law cases where the central allegation is of breach of duty. The fourth — appealing to the reliance interest and comity with the legislature[242] — may have only a limited scope of operation. More seriously, because the third and fourth considerations point to what Horrigan calls ‘precedential’ reasons, whereas the first and fifth considerations point to ‘substantive’ reasons, the framework adopted by the plurality in Imbree blurs the threshold inquiry into improvement to the law and the further inquiry into whether, notwithstanding that overruling will improve the law, the Court is for one or another reason constrained not to overrule.
Perhaps most importantly, the framework adopted by the plurality in Imbree has no necessary connection with the ‘no new reasons’ constraint that figures prominently in a sound approach to overruling. Imbree itself is an excellent illustration of this absence of necessary connection: the plurality in that case overruled even though no new reasons were before it, and Kirby J would appear to have done so too, notwithstanding that in his separate judgment he showed that he was aware of the significance of the ‘no new reasons’ constraint.
All of this indicates that in future overruling cases in the common law setting, the High Court will be well served by taking the best of Hospital Contribution Insurance Fund, John, and now Imbree, as well as learning from cases like Esso and Brodie, in developing a new framework for overruling that accords with a sound approach to that practice. This almost certainly demands retaining the fifth consideration adopted by the plurality in Imbree as central to the inquiry into improvement to the law. However, it also most likely entails a preparedness on the part of the Court, when conducting that threshold inquiry, to take into account a range of reasons relating to doctrine, underlying principles, and public policy that are not contemplated by the fifth consideration. In this regard, Esso and Brodie are instructive. Constructing a framework for use in future cases may also require explicitly adopting a ‘no new reasons’ constraining principle. And finally, it may require a recasting of the third and fourth considerations from Hospital Contribution Insurance Fund and John so that it is clear that they speak only to the question whether constraints to overruling are operative and not to the prior question whether departing from the ratio decidendi of a precedent decision will bring about improvement to the law.
The High Court clearly places great emphasis on the value of stare decisis; so much is clear from its reminders to state courts not to depart from the common law decisions of intermediate appellate courts (even in other states), nor to depart from the High Court’s ‘seriously considered’ obiter dicta in certain circumstances.[243] If the Court’s approach to overruling its own previous decisions is to reflect its broader emphasis on stare decisis, it must ensure that its decisions in overruling cases, especially common law cases, may be reconciled with the strong Burkean conception of precedent that it clearly endorses in other respects, rather than the more laissez-faire conception of precedent for which B V Harris has argued.
[*] BA (Hons), LLB (Hons) (Melb), BCL, DPhil (Oxon); Associate Professor, Melbourne Law School, The University of Melbourne. Our grateful thanks to Cosima McRae, Elly Patira, Alissa Sputore and the other staff of the Research Service at the Law Library of The University of Melbourne for their very comprehensive assistance.
[†] BA (Hons), LLB (Hons) (Manit), LLM (Lond); Professor, Melbourne Law School, The University of Melbourne.
[1] Cases in which the Court has considered whether or not to overrule itself include Australian Agricultural Co v Federated Engine-Drivers and Firemen’s Association of Australasia [1913] HCA 41; (1913) 17 CLR 261; R v Commonwealth Court of Conciliation and Arbitration; Ex parte Brisbane Tramways Co Ltd [No 1] [1914] HCA 15; (1914) 18 CLR 54 (‘The Tramways Case [No 1]’); A-G (NSW) v Perpetual Trustee Co (Ltd) [1952] HCA 2; (1952) 85 CLR 237; Geelong Harbour Trust Commissioners v Gibbs, Bright & Co [1970] HCA 16; (1970) 122 CLR 504; Queensland v Commonwealth [1977] HCA 60; (1977) 139 CLR 585; Commonwealth v The Hospital Contribution Insurance Fund of Australia (1982) 150 CLR 49; Baker v Campbell (1983) 153 CLR 52; Jones v Commonwealth (1987) 71 ALR 497; John v Federal Commissioner of Taxation (1989) 166 CLR 417; Street v Queensland Bar Association [1989] HCA 53; (1989) 168 CLR 461; Re Tyler; Ex parte Foley [1994] HCA 25; (1994) 181 CLR 18; Northern Territory v Mengel (1995) 185 CLR 307; Ha v New South Wales [1997] HCA 34; (1997) 189 CLR 465; Lange v Australian Broadcasting Corporation [1997] HCA 25; (1997) 189 CLR 520; Esso Australia Resources Ltd v Federal Commissioner of Taxation [1999] HCA 67; (1999) 201 CLR 49; Brodie v Singleton Shire Council [2001] HCA 29; (2001) 206 CLR 512; Imbree v McNeilly [2008] HCA 40; (2008) 236 CLR 510; Wurridjal v Commonwealth (2009) 237 CLR 309.
[2] For a flexible approach, see, eg, Australian Agricultural Co v Federated Engine-Drivers and Firemen’s Association of Australasia [1913] HCA 41; (1913) 17 CLR 261, 278 (Isaacs J); Street v Queensland Bar Association [1989] HCA 53; (1989) 168 CLR 461, 588 (McHugh J); Re Tyler; Ex parte Foley [1994] HCA 25; (1994) 181 CLR 18, 38 (McHugh J); Wurridjal v Commonwealth (2009) 237 CLR 309, 352–3 [70]–[71] (French CJ). For a more constrained approach, see, eg, Queensland v Commonwealth [1977] HCA 60; (1977) 139 CLR 585, 599 (Gibbs J); Jones v Commonwealth (1987) 71 ALR 497, 498 (Mason CJ, Wilson, Brennan, Deane, Dawson and Toohey JJ).
[3] See, eg, Geelong Harbour Trust Commissioners v Gibbs, Bright & Co [1970] HCA 16; (1970) 122 CLR 504, 516–17 (Barwick CJ). But see at 517–18 (McTiernan and Menzies JJ), 519 (Kitto J); John v Federal Commissioner of Taxation (1989) 166 CLR 417, 439–40 (Mason CJ, Wilson, Dawson, Toohey and Gaudron JJ).
[4] [2001] HCA 29; (2001) 206 CLR 512.
[5] Ibid 560.
[6] State Government Insurance Commission v Trigwell [1979] HCA 40; (1979) 142 CLR 617, 633. See Brodie [2001] HCA 29; (2001) 206 CLR 512, 561 [108].
[7] See below Part IV(C). See Brodie [2001] HCA 29; (2001) 206 CLR 512, 547–77 [74]–[149].
[8] [2004] HCA 16; (2004) 216 CLR 515.
[9] (1995) 182 CLR 609.
[10] See Perre v Apand Pty Ltd [1999] HCA 36; (1999) 198 CLR 180; Sullivan v Moody (2001) 207 CLR 562, 578–9 [48] (Gleeson CJ, Gaudron, McHugh, Hayne and Callinan JJ).
[11] Jaensch v Coffey [1984] HCA 52; (1984) 155 CLR 549, 578–87 (Deane J); Cook v Cook [1986] HCA 73; (1986) 162 CLR 376, 382–3 (Mason, Wilson, Deane and Dawson JJ).
[12] See generally Richard H S Tur, ‘Time and Law’ (2002) 22 Oxford Journal of Legal Studies 463.
[13] See J D Heydon, ‘How Far Can Trial Courts and Intermediate Appellate Courts Develop the Law?’ (2009) 9 Oxford University Commonwealth Law Journal 1, 4.
[14] Julius Stone, Precedent and Law: Dynamics of Common Law Growth (Butterworths, 1985) 192.
[15] J W Harris, ‘Murphy Makes It Eight — Overruling Comes to Negligence’ (1991) 11 Oxford Journal of Legal Studies 416, 421.
[16] Practice Statement (Judicial Precedent) [1966] 1 WLR 1234.
[17] Brodie [2001] HCA 29; (2001) 206 CLR 512, 586 [187] (Kirby J).
[18] J W Harris, ‘Towards Principles of Overruling — When Should a Final Court of Appeal Second Guess?’ (1990) 10 Oxford Journal of Legal Studies 135, 140.
[19] Ibid. Overruling the ratio decidendi of a precedent case usually means overruling the decision in that case, but this is not necessarily so: see Rupert Cross and J W Harris, Precedent in English Law (Clarendon Press, 4th ed, 1991) 131–2. In what follows, unless we wish to be more specific, for ease we refer simply to ‘overruling a precedent decision’.
[20] Neil Duxbury, The Nature and Authority of Precedent (Cambridge University Press, 2008) 117.
[21] J W Harris, ‘Towards Principles of Overruling’, above n 18, 143–6. Harris subsequently published another article discussing the case of Murphy v Brentwood District Council [1991] UKHL 2; [1991] 1 AC 398, in which the House of Lords explicitly overruled itself again: J W Harris, ‘Murphy Makes It Eight’, above n 15. And there were further ‘copper-bottomed’ overruling cases in the House of Lords after 1991 as well: see the table in Louis Blom-Cooper, ‘1966 and All That: The Story of the Practice Statement’ in Louis Blom-Cooper, Brice Dickson and Gavin Drewry (eds), The Judicial House of Lords 1876–2009 (Oxford University Press, 2009) 128, 143–4. So far, the Supreme Court of the United Kingdom has not overruled itself or the House of Lords, but in Austin v Southwark London Borough Council [2010] UKSC 28; [2010] 4 All ER 16, the Supreme Court confirmed that it considers itself as free to overrule previous decisions as did the House of Lords. We are grateful to James Lee for alerting us to this case.
[22] See, eg, the table set out in Alan Paterson, The Law Lords (Macmillan Press, 1982) 163–5.
[23] See, eg, ibid; Lyndel V Prott, ‘When Will a Superior Court Overrule Its Own Decision?’ (1978) 52 Australian Law Journal 304, 304–10; A A Paterson, ‘Lord Reid’s Unnoticed Legacy — A Jurisprudence of Overruling’ (1981) 1 Oxford Journal of Legal Studies 375; G Maher, ‘Statutory Interpretation and Overruling in the House of Lords’ [1981] Statute Law Review 85; J W Harris, ‘Towards Principles of Overruling’, above n 18; J W Harris, ‘Murphy Makes It Eight’, above n 15; Francis A R Bennion, ‘Consequences of an Overrule’ [2001] Public Law 450; B V Harris, ‘Final Appellate Courts Overruling Their Own “Wrong” Precedents: The Ongoing Search for Principle’ (2002) 118 Law Quarterly Review 408; Lord Hope of Craighead, ‘“Decision Overruled” — Facing Up to Judicial Fallibility’ (2003) 14 King’s College Law Journal 121; Duxbury, The Nature and Authority of Precedent, above n 20, ch 4; Blom-Cooper, above n 21; Neil Duxbury, ‘Lord Wright and Innovative Traditionalism’ (2009) 59 University of Toronto Law Journal 265. For a historical survey of approaches to overruling in various courts, see Jack Hodder, ‘Departure from “Wrong” Precedents by Final Appellate Courts: Disagreeing with Professor Harris’ [2003] New Zealand Law Review 161. For a discussion of the Privy Council, which has always regarded itself as free to depart from its previous decisions, see Derek O’Brien, ‘The Privy Council Overrules Itself — Again!’ [2008] Public Law 28. For a comparative study of overruling practice in England and Belgium, see Isabelle Rorive, ‘Towards Principles of Overruling in a Civil Law Supreme Court’ in Timothy Endicott, Joshua Getzler and Edwin Peel (eds), Properties of Law: Essays in Honour of Jim Harris (Oxford University Press, 2006) 277.
[24] Practice Statement (Judicial Precedent) [1966] 1 WLR 1234.
[25] Prott, above n 23, 310–14; Keith Mason, ‘Prospective Overruling’ (1989) 63 Australian Law Journal 526; J W Harris, ‘Towards Principles of Overruling’, above n 18; Bryan Horrigan, ‘Towards a Jurisprudence of High Court Overruling’ (1992) 66 Australian Law Journal 199; Hodder, above n 23.
[26] See Horrigan, above n 25, 199.
[27] See Australian Agricultural Co v Federated Engine-Drivers and Firemen’s Association of Australasia [1913] HCA 41; (1913) 17 CLR 261.
[28] [2008] HCA 40; (2008) 236 CLR 510.
[29] [1986] HCA 73; (1986) 162 CLR 376.
[30] Note that Imbree has already been applied in a case where the plaintiff brought a damages claim in negligence for an injury that arose as a result of participating in a joint illegal activity: Miller v Miller [2009] WASCA 199; (2009) 54 MVR 367. The High Court has reserved judgment in an appeal from this decision.
[31] [1999] HCA 67; (1999) 201 CLR 49.
[32] See London Street Tramways Co Ltd v London County Council [1898] AC 375 (‘London Tramways’). Cf David Pugsley, ‘London Tramways (1898)’ (1996) 17 Journal of Legal History 172; Duxbury, The Nature and Authority of Precedent, above n 20, 135–6, arguing that the House of Lords considered itself bound by its previous decisions before London Tramways was decided.
[33] See Joseph Raz, Practical Reason and Norms (Oxford University Press, first published 1975, 1999 ed) 35–48.
[34] Ibid 15–48; Joseph Raz, The Authority of Law — Essays on Law and Morality (Oxford University Press, 1979) 114, 195, 197; Stephen R Perry, ‘Judicial Obligation, Precedent and the Common Law’ (1987) 7 Oxford Journal of Legal Studies 215, 243–4.
[35] Lionel Murphy, ‘The Responsibility of Judges’ in Gareth Evans (ed), Law, Politics and the Labor Movement (Legal Service Bulletin, 1980) 2, 5, quoted — but not endorsed — in Justice Michael Kirby, ‘Precedent Law, Practice and Trends in Australia’ (2007) 28 Australian Bar Review 243, 243–4.
[36] See J W Harris, ‘Towards Principles of Overruling’, above n 18.
[37] Ibid 136.
[38] Ibid 149.
[39] Ibid 152.
[40] Ibid 152–3.
[41] By using the device of prospective overruling, courts might circumvent this concern. On prospective overruling, see, eg, Andrew G L Nicol, ‘Prospective Overruling: A New Device for English Courts?’ (1976) 39 Modern Law Review 542; Mason, above n 25, 531; Tur, above n 12; Ben Juratowitch, Retroactivity and the Common Law (Hart Publishing, 2008) ch 6; Jesse Wall, ‘Prospective Overruling — It’s About Time’ [2009] OtaLawRw 7; (2009) 12 Otago Law Review 131. Although the House of Lords appeared to consider itself free to overrule prospectively (and presumably so does the Supreme Court of the United Kingdom), the High Court of Australia does not: compare National Westminster Bank plc v Spectrum Plus Ltd [2005] 2 AC 680 and Ha v New South Wales [1997] HCA 34; (1997) 189 CLR 465. We therefore do not consider prospective overruling in this article.
[42] J W Harris, ‘Towards Principles of Overruling’, above n 18, 154.
[43] Ibid 156.
[44] For another insight into the ways in which improvement to the law might be achieved by overruling, see the discussion of the ways in which a precedent decision could be viewed as ‘wrong’ in J D Heydon, ‘Limits to the Powers of Ultimate Appellate Courts’ (2006) 122 Law Quarterly Review 399, 408–10.
[45] Horrigan, above n 25, 205.
[46] Ibid.
[47] Ibid.
[48] These four are not the only possible constraining principles. For others, see Heydon, ‘Limits to the Powers of Ultimate Appellate Courts’, above n 44, 415–24.
[49] J W Harris, ‘Towards Principles of Overruling’, above n 18, 159. See also Blom-Cooper, above n 21, 141: ‘A decision to overrule should rest upon some special reason(s) over and above a belief that the prior case was wrongly decided, or even if not wrongly decided, that it should be reversed on grounds of public policy.’
[50] Cf Lionel Smith, ‘The Rationality of Tradition’ in Timothy Endicott, Joshua Getzler and Edwin Peel (eds), Properties of Law: Essays in Honour of Jim Harris (Oxford University Press, 2006) 297, 307–8, arguing that where social circumstances have changed since the time a precedent case was decided, the precedent may no longer be said to have binding force, so that an ‘overruling’ court is in effect making new law where old law has ceased to be.
[51] J W Harris, ‘Towards Principles of Overruling’, above n 18, 189–90.
[52] Ibid 184.
[53] Ibid 189–91. On ‘objective wrongness’, see also Prott, above n 23, 313. Note that this aspect of Harris’ account of overruling accords with the greater latitude that the High Court of Australia has extended to itself when considering overruling in constitutional cases: see above Part I(A).
[54] J W Harris, ‘Towards Principles of Overruling’, above n 18, 169. See also Stone, above n 14, 177, referring to ‘settled expectations’; Horrigan, above n 25, 212.
[55] J W Harris, ‘Towards Principles of Overruling’, above n 18, 169. The 1966 Practice Statement alludes to some of the circumstances in which a reliance interest is likely to ground a constraint to overruling. It makes reference to contracts, the law of property, fiscal arrangements and the criminal law: Practice Statement (Judicial Precedent) [1966] 1 WLR 1234. We return to the reliance interest below: see below Part III(B)(4).
[56] J W Harris, ‘Towards Principles of Overruling’, above n 18, 177.
[57] Ibid 178.
[58] B V Harris, above n 23, 415.
[59] J W Harris, ‘Towards Principles of Overruling’, above n 18, 178. See also ibid 425; Hodder, above n 23, 181.
[60] J W Harris, ‘Towards Principles of Overruling’, above n 18, 183.
[61] Ibid 198.
[62] Ibid 192.
[63] Ibid 198.
[64] See B V Harris, above n 23. This approach was criticised by Hodder, above n 23, 177–82.
[65] B V Harris, above n 23, 411.
[66] Ibid 417.
[67] Ibid 411.
[68] Ibid.
[69] Ibid 424–5, 427.
[70] Ibid 418–19, 421.
[71] Ibid 419.
[72] Ibid 421. This form of words may be traced back to a statute of Edward III: see Sir Gerard Brennan, ‘Swearing in of Sir Gerard Brennan as Chief Justice’ (1995) 183 CLR ix, x.
[73] B V Harris, above n 23, 421–2.
[74] Many legal philosophy books have been dedicated to answering this question: among the most famous is Ronald Dworkin, Law’s Empire (Harvard University Press, 1986).
[75] See Duxbury, The Nature and Authority of Precedent, above n 20, 135–6.
[76] As Kirby J put it in Brodie [2001] HCA 29; (2001) 206 CLR 512, 597 [219]: ‘The natural and proper judicial inclination in such matters is towards restraint.’
[77] See, eg, Sir Owen Dixon, ‘Concerning Judicial Method’ in Jesting Pilate: And Other Papers and Addresses (Lawbook, 1965) 152; Paterson, The Law Lords, above n 22; Lord Reid, ‘The Judge as Law Maker’ (1997) 63 Arbitration 180; M H McHugh, ‘The Judicial Method’ (1999) 73 Australian Law Journal 37, 47–9; Lord Hope of Craighead, above n 23; Heydon, ‘Limits to the Powers of Ultimate Appellate Courts’, above n 44; Kirby, above n 35.
[78] Perry, above n 34, 222–3.
[79] Horrigan, above n 25, 208–9.
[80] See Nicol, above n 41, 548.
[81] Perry, above n 34, 221–2.
[82] Ibid 221 n 25.
[83] Ibid 221–2. Something like Perry’s distinction between strong and weak Burkean conceptions of precedent is also introduced by Horrigan, above n 25, 202, discussing the distinction drawn in Dworkin, above n 74, between ‘strict’ and ‘relaxed’ approaches to precedent.
[84] Perry, above n 34, 221 n 25.
[85] Imbree [2008] HCA 40; (2008) 236 CLR 510, 521–2 [28]–[32] (Gummow, Hayne and Kiefel JJ).
[86] Kenneth Warner, ‘Standard of Care in Negligence’ (2009) 83 Australian Law Journal 599, 599.
[87] See Cook [1986] HCA 73; (1986) 162 CLR 376; Nettleship v Weston [1971] EWCA Civ 6; [1971] 2 QB 691.
[88] Imbree [2008] HCA 40; (2008) 236 CLR 510, 521 [27].
[89] Ibid 513 [1] (Gleeson CJ), 541–2 [105] (Kirby J), 565–6 [193] (Crennan J). Heydon J did not consider it necessary to accept this view in order to determine the case: see below Part III(D). For endorsements of the result in Imbree and comments on the substance of the decision, see Mandy Shircore, ‘Standardising the Standard of the Learner Driver: Imbree v McNeilly’ (2008) 15 James Cook University Law Review 234; Judy Allen, ‘Variable Standards of Care in Negligence’ (2009) 17 Tort Law Review 5; Warner, above n 86, 599.
[90] [1986] HCA 73; (1986) 162 CLR 376, 384 (Mason, Wilson, Deane and Dawson JJ).
[91] See Imbree [2008] HCA 40; (2008) 236 CLR 510, 511–12 (A S Morrison SC) (during argument).
[92] Ibid 517 [13] (Gleeson CJ), 521 [27] (Gummow, Hayne and Kiefel JJ), 541–2 [105] (Kirby J), 565–6 [193] (Crennan J).
[95] Imbree [2008] HCA 40; (2008) 236 CLR 510, 526 [45] (Gummow, Hayne and Kiefel JJ).
[96] Ibid.
[97] Ibid.
[98] Ibid.
[99] Ibid.
[100] See, eg, Street v Queensland Bar Association [1989] HCA 53; (1989) 168 CLR 461, 489 (Mason CJ), 549 (Dawson J), 560 (Toohey J), 569–70 (Gaudron J), 588 (McHugh J); Brodie [2001] HCA 29; (2001) 206 CLR 512, 560–4 [115]–[117] (Gaudron, McHugh and Gummow JJ); Esso [1999] HCA 67; (1999) 201 CLR 49, 103–4 [163] (Callinan J). See also AON Risk Services Australia Ltd v Australian National University [2009] HCA 27; (2009) 239 CLR 175, 217 [111] (Gummow, Hayne, Crennan, Kiefel and Bell JJ), noted in Camille Cameron, ‘New Directions for Case Management in Australia’ (2010) 29 Civil Justice Quarterly 337.
[101] [2008] HCA 40; (2008) 236 CLR 510, 526 [45] (emphasis added), quoting Perre v Apand Pty Ltd [1999] HCA 36; (1999) 198 CLR 180, 216 [92] (McHugh J).
[102] [2008] HCA 40; (2008) 236 CLR 510, 526 [45].
[103] (1989) 166 CLR 417, 439 (Mason CJ, Wilson, Dawson, Toohey and Gaudron JJ).
[104] [1974] HCA 46; (1974) 131 CLR 409.
[105] (1989) 166 CLR 417, 439 (Mason CJ, Wilson, Dawson, Toohey and Gaudron JJ): ‘Curran stands alone’.
[106] Indeed, the specific issue came before the High Court for the first time in Cook and for the second time in Imbree, and in England arose for the first and only time in Nettleship v Weston [1971] EWCA Civ 6; [1971] 2 QB 691.
[107] See, eg, Phillips v William Whiteley Ltd [1938] 1 All ER 566; Insurance Commissioner v Joyce [1948] HCA 17; (1948) 77 CLR 39; McHale v Watson [1966] HCA 13; (1966) 115 CLR 199; Rogers v Whitaker [1992] HCA 58; (1992) 175 CLR 479.
[108] [2008] HCA 40; (2008) 236 CLR 510, 533 [70].
[109] [1932] AC 562.
[110] [2008] HCA 40; (2008) 236 CLR 510, 526 [45].
[111] (1989) 166 CLR 417, 439 (Mason CJ, Wilson, Dawson, Toohey and Gaudron JJ).
[112] [1974] HCA 46; (1974) 131 CLR 409, 411–16 (Barwick CJ), 416–17 (Menzies J), 417–23 (Gibbs J).
[113] Ibid 423–30.
[114] (1989) 166 CLR 417, 439 (Mason CJ, Wilson, Dawson, Toohey and Gaudron JJ).
[115] See Cook [1986] HCA 73; (1986) 162 CLR 376, 382–8 (Mason, Wilson, Deane and Dawson JJ).
[116] See ibid 390–4.
[117] Another possibility is that the reasoning of the plurality was in fact sufficiently similar to that of Brennan J to render Cook a case in which an effectively unanimous decision was reached. We return to this idea in our discussion of the ‘knowledge of inexperience’ point: see below Part III(B)(5)(b).
[118] (1989) 166 CLR 417, 451.
[119] See CAL No 14 Pty Ltd v Motor Accidents Insurance Board (2009) 239 CLR 390, 411–13 [49]–[51] (Gummow, Heydon and Crennan JJ), referring to the decision in Cole v South Tweed Heads Rugby League Football Club Ltd [2004] HCA 29; (2004) 217 CLR 469.
[120] [2008] HCA 40; (2008) 236 CLR 510, 526 [45].
[121] Ibid 533 [72].
[122] (1982) 150 CLR 49, 57.
[123] See, eg, the comments in Harold Luntz et al, Torts: Cases and Commentary (LexisNexis Butterworths, 6th ed, 2009) 328 [6.2.11] on the way in which regulatory regimes incentivised the widespread use of seat belts in Australia. Arguably, the law of negligence cannot realistically achieve such results.
[124] Some inconvenience might also have been caused to driving instructors, but we leave that question to one side.
[125] In this regard, it is worth noting that in Hospital Contribution Insurance Fund, the dispute was a matter of public law: (1982) 150 CLR 49, 51–2 (Gibbs CJ).
[126] For a sense of the philosophical debate, see generally Ernest J Weinrib, The Idea of Private Law (Harvard University Press, 1995); Andrew Robertson and Tang Hang Wu (eds), The Goals of Private Law (Hart Publishing, 2009). For a sense of the current attitude of the High Court of Australia to public policy considerations in negligence cases, see Sullivan v Moody (2001) 207 CLR 562, 579 [49] (Gleeson CJ, Gaudron, McHugh, Hayne and Callinan JJ).
[127] [2008] HCA 40; (2008) 236 CLR 510, 526 [45].
[128] See J W Harris, ‘Towards Principles of Overruling’, above n 18, 169, 177–8.
[129] B V Harris, above n 23, 415.
[130] J W Harris, ‘Towards Principles of Overruling’, above n 18, 170–1.
[131] (1982) 150 CLR 49, 58.
[132] (1989) 166 CLR 417, 439 (Mason CJ, Wilson, Dawson, Toohey and Gaudron JJ).
[133] [2008] HCA 40; (2008) 236 CLR 510, 533 [72].
[134] Ibid.
[135] [1977] HCA 60; (1977) 139 CLR 585, 600 (Gibbs J). See also at 603–4 (Stephen J).
[136] Practice Statement (Judicial Precedent) [1966] 1 WLR 1234.
[137] [2008] HCA 40; (2008) 236 CLR 510, 526 [45].
[138] Ibid 524–5 [40]–[43].
[139] Jaensch v Coffey [1984] HCA 52; (1984) 155 CLR 549, 578–88; Sutherland Shire Council v Heyman [1985] HCA 41; (1985) 157 CLR 424, 494–512; Stevens v Brodribb Sawmilling Co Pty Ltd (1986) 160 CLR 16, 49–54.
[140] [1986] HCA 73; (1986) 162 CLR 376, 381–2 (Mason, Wilson, Deane and Dawson JJ).
[141] Ibid 382–3.
[142] Perre v Apand Pty Ltd [1999] HCA 36; (1999) 198 CLR 180; Sullivan v Moody (2001) 207 CLR 562, 578–9 [48] (Gleeson CJ, Gaudron, McHugh, Hayne and Callinan JJ).
[143] [2008] HCA 40; (2008) 236 CLR 510, 526–7 [46]–[47].
[144] Ibid 527 [47]–[48].
[145] [1986] HCA 73; (1986) 162 CLR 376, 391–2.
[146] [2008] HCA 40; (2008) 236 CLR 510, 527 [48].
[147] See Transfield Services (Australia) Pty Ltd v Hall [2008] NSWCA 294; (2008) 75 NSWLR 12, 37 [114] (Campbell JA):
As Imbree shows, the fact that a previous decision of the High Court made use of the concept of proximity is not seen, even by the High Court, as a sufficient reason for discarding that previous decision. But it does cause closer attention to be paid to whatever basis of principle there is in the judgment, other than proximity, for arriving at the result.
See also Miller v Miller [2009] WASCA 199; (2009) 54 MVR 367, 399 [143], 400 [147] (Newnes JA); contra Farrell v Farrell [2009] NSWSC 1122 (21 October 2009) [140]–[141] (Hall J), where the ratio decidendi of Cook was treated as turning on the notion of proximity.
[148] Imbree [2008] HCA 40; (2008) 236 CLR 510, 529 [55] (emphasis added).
[149] See also ibid 528 [51] (Gummow, Hayne and Kiefel JJ): ‘The fundamental reason why Cook v Cook should no longer be treated as expressing any distinct principle in the law of negligence is that basic considerations of principle require a contrary conclusion.’
[150] See Vaughan v Menlove [1837] EngR 424; (1837) 3 Bing NC 468; 132 ER 490; Blyth v Birmingham Waterworks Co [1856] EngR 223; (1856) 11 Exch 781; 156 ER 1047.
[151] See McHale v Watson [1966] HCA 13; (1966) 115 CLR 199 (children); Rogers v Whitaker [1992] HCA 58; (1992) 175 CLR 479 (skilled professionals).
[152] [2008] HCA 40; (2008) 236 CLR 510, 528–30 [53]–[58].
[153] [1986] HCA 73; (1986) 162 CLR 376, 388 (Mason, Wilson, Deane and Dawson JJ), 393–4 (Brennan J).
[154] Ibid 377–8 (J J Doyle QC) (during argument).
[155] Ibid 387.
[156] See above Part III(B)(5).
[157] J W Harris, ‘Towards Principles of Overruling’, above n 18, 159.
[158] See B V Harris, above n 23, 424–5, 427.
[159] See above Part II(C).
[160] [2008] HCA 40; (2008) 236 CLR 510, 541 [105].
[161] Ibid 543 [109].
[162] Ibid 543 [110] (emphasis added).
[163] See J W Harris, ‘Towards Principles of Overruling’, above n 18, 159.
[164] Imbree [2008] HCA 40; (2008) 236 CLR 510, 543 [110].
[165] See, eg, ibid 543 [108]:
[Compulsory insurance] is clearly an aspect of the social reality in which the common law principle falls to be expounded in this case. Its existence encourages my acceptance of a single universal, objective standard of care owed by all drivers. Giving weight to the consideration of compulsory insurance accords with a growing preparedness of the courts to acknowledge the influence of insurance, at least where it is compulsory and provided by statute, in defining the content of legal liability.
[166] Ibid 544 [112].
[167] In Imbree, the traditional view was put by Gleeson CJ: ibid 517–20 [14]–[23]. See also Jane Stapleton, ‘Tort, Insurance and Ideology’ (1995) 58 Modern Law Review 820, 823–9; James Edelman and Simone Degeling, ‘The Future of the Common Law of Torts’ (2010) 33 Australian Bar Review 45, 51–3.
[168] Stapleton, above n 167, 841–3.
[169] Imbree [2008] HCA 40; (2008) 236 CLR 510, 543 [110], 544 [112], 561 [170]–[171]. See also at 517 [14] (Gleeson CJ): ‘Schemes of compulsory insurance for third party liability in motor accidents are not new. They existed at the time of Cook v Cook, and for a long time before then.’
[170] Ibid 542 [107], 549–50 [132], 563–4 [180]. See Nettleship v Weston [1971] EWCA Civ 6; [1971] 2 QB 691, 699–700.
[171] Imbree [2008] HCA 40; (2008) 236 CLR 510, 543 [108].
[172] [1986] HCA 73; (1986) 162 CLR 376, 385. The relevant passage from Nettleship v Weston [1971] EWCA Civ 6; [1971] 2 QB 691, 700 (Lord Denning MR) is:
But the injured person is only able to recover if the driver is liable in law. So the judges see to it that he is liable, unless he can prove care and skill of a high standard … Thus we are, in this branch of the law, moving away from the concept: ‘No liability without fault.’ We are beginning to apply the test: ‘On whom should the risk fall?’ Morally the learner driver is not at fault; but legally she is liable to be because she is insured and the risk should fall on her.
[173] A point made by Gleeson CJ: Imbree [2008] HCA 40; (2008) 236 CLR 510, 517 [14]. See also our discussion of Esso, where a new legal ingredient existed, grounded in arguments made before the overruling court for the first time: see below Part IV(B).
[174] Imbree [2008] HCA 40; (2008) 236 CLR 510, 543 [108] (Kirby J).
[175] Ibid 559–60 [165]–[167], citing Robertson v Swincer [1989] SASC 1758; (1989) 52 SASR 356; Lynch v Lynch (1991) 25 NSWLR 411; Kars v Kars [1996] HCA 37; (1996) 187 CLR 354.
[176] Indeed, in Imbree itself, the plurality made no reference at all to insurance.
[177] Lamb v Cotogno [1987] HCA 47; (1987) 164 CLR 1, 10–11 (Mason CJ, Brennan, Deane, Dawson and Gaudron JJ); Perre v Apand Pty Ltd [1999] HCA 36; (1999) 198 CLR 180, 230 [130] (McHugh J); Bowditch v McEwan [2002] QCA 172; [2003] 2 Qd R 615, 617 (de Jersey CJ).
[178] Imbree [2008] HCA 40; (2008) 236 CLR 510, 543 [108].
[179] See above Part III(B)(5)(b).
[180] Imbree [2008] HCA 40; (2008) 236 CLR 510, 552 [140].
[181] Ibid 564–5 [186], 565 [191].
[182] Ibid 564–5 [186].
[183] Ibid 564–5 [186], 565 [191].
[184] Different considerations would seem to bear on the question of when trial and intermediate courts can develop the law. For the view of the High Court of Australia, see, eg, Farah Constructions Pty Ltd v Say-Dee Pty Ltd (2007) 230 CLR 89, 150–2 [134]–[135] (Gleeson CJ, Gummow, Callinan, Heydon and Crennan JJ). See also Heydon, ‘How Far Can Trial Courts and Intermediate Appellate Courts Develop the Law?’, above n 13, 13–30.
[185] Imbree [2008] HCA 40; (2008) 236 CLR 510, 526 [45] (Gummow, Hayne and Kiefel JJ).
[186] For an exhaustive critique of the ‘manifestly wrong’ test that appears to have been applied in Australia before Hospital Contribution Insurance Fund, see Horrigan, above n 25. In his article, Horrigan proposes 10 considerations that bear on the question of overruling in the High Court: at 210–12. Our comments about consistency with the fundamentals of a sound approach to overruling with reference to the five considerations set out by the plurality in Imbree apply equally to Horrigan’s, or any other, framework.
[187] Imbree [2008] HCA 40; (2008) 236 CLR 510, 526 [45].
[188] Ibid.
[189] Ibid.
[190] [1976] HCA 63; (1976) 135 CLR 674.
[191] Ibid 688.
[192] Ibid 677.
[193] See Evidence Act 1995 (Cth) pt 3.10; Evidence Act 1995 (NSW) pt 3.10.
[194] [1999] HCA 67; (1999) 201 CLR 49, 51–2 (B J Shaw QC) (during argument).
[195] Ibid 73 [61].
[196] Imbree [2008] HCA 40; (2008) 236 CLR 510, 526 [45].
[197] We return to this point: see below Part IV(B)(2).
[198] [1999] HCA 67; (1999) 201 CLR 49, 71–2 [56].
[199] Ibid 72 [57].
[200] Ibid.
[201] Ibid 72 [58]. Note the similarity between this consideration and the third consideration from John, discussed above in Part III(B)(3).
[202] Ibid 101–5 [152]–[166].
[203] Ibid 74 [65], 76–9 [71]–[78].
[204] Ibid 83–6 [91]–[98].
[205] Ibid 86–93 [99]–[112].
[206] Ibid 93 [113].
[207] See above Part III(B)(5).
[208] [1999] HCA 67; (1999) 201 CLR 49, 70–3 [53]–[61].
[209] Ibid 66 [39].
[210] Ibid 67 [42], 70–1 [55]–[56] (Gleeson CJ, Gaudron and Gummow JJ).
[211] Ibid 66–7 [41].
[212] See the summary of counsel’s arguments: ibid 51–4.
[213] In Esso, the dissenters did not point to new reasons, but they did not have to as, in their view, even the prima facie requirement of improvement to the law had not been satisfied: ibid 80–1 [83] (McHugh J), 86 [98] (Kirby J). Arguably, Callinan J did not point to new reasons either, as he did not refer to the fact that the option of a dominant purpose test was not the subject of argument in Grant v Downs: ibid 93–108 [115]–[175]. In our view, this is a basis for preferring the judgment of the plurality in Esso to that of Callinan J.
[214] [1936] HCA 65; (1936) 57 CLR 259.
[215] [1950] HCA 6; (1950) 80 CLR 357.
[216] Buckle [1936] HCA 65; (1936) 57 CLR 259, 268–9 (Latham CJ), 280–3 (Dixon J), 298 (McTiernan J); ibid 363–4 (Latham CJ), 370 (Dixon J), 375–6 (Fullagar J).
[217] [2001] HCA 29; (2001) 206 CLR 512, 532 [30] (Gleeson CJ), 540 [55], 560–4 [107]–[115], 573 [137] (Gaudron, McHugh and Gummow JJ), 604 [238] (Kirby J).
[218] Ibid 547–77 [74]–[149].
[219] Ibid 540 [55].
[220] Ibid 560–4 [107]–[115].
[221] Ibid 588–604 [193]–[237].
[222] Ibid 604 [238].
[223] Ibid 528 [14]. See also at 529 [20]: ‘To change the law by abolishing an established rule does not involve reform unless what is left, or what is put in its place, is something better.’
[224] Ibid 534–6 [37]–[43].
[225] Ibid 536–7 [44]–[47].
[226] Arguably, Gleeson CJ also thought that reliance and comity constraints were operative: see ibid 536 [42]–[43].
[227] One interpretation of the judgment of Hayne J in Brodie is that he regarded the question of overruling Buckle and Gorringe as moot. It would appear that Hayne J thought that the critical question in Brodie was whether the highway authority in question owed a duty to take reasonable care in the exercise of its powers, a question that he separated from the question of any immunity from suit: see ibid 618–19 [279]–[282], 621–2 [287]–[292], 632–3 [323]–[324], 634 [328]–[331], 636 [338]. Callinan J did not address the question of overruling.
[228] Ibid 547–9 [75]–[78].
[229] Ibid 618 [279].
[230] Ibid 558–9 [102]–[103]. The issue was explored in Sutherland Shire Council v Heyman [1985] HCA 41; (1985) 157 CLR 424; Burnie Port Authority v General Jones Pty Ltd [1994] HCA 13; (1994) 179 CLR 520; Pyrenees Shire Council v Day [1998] HCA 3; (1998) 192 CLR 330; Romeo v Conservation Commission (NT) (1998) 192 CLR 431; Crimmins v Stevedoring Industry Finance Committee [1999] HCA 59; (1999) 200 CLR 1.
[231] Brodie [2001] HCA 29; (2001) 206 CLR 512, 564–70 [116]–[129].
[232] Ibid 589 [196].
[234] [2008] HCA 40; (2008) 236 CLR 510, 526 [45].
[235] Ibid 527 [47]–[48].
[236] Brodie [2001] HCA 29; (2001) 206 CLR 512, 597 [219].
[237] See Perry, above n 34, 222–3.
[238] Imbree [2008] HCA 40; (2008) 236 CLR 510, 526 [45] (Gummow, Hayne and Kiefel JJ).
[239] Ibid.
[240] Ibid.
[241] Ibid.
[242] Ibid.
[243] Farah Constructions Pty Ltd v Say-Dee Pty Ltd (2007) 230 CLR 89, 150–2 [134]–[135] (Gleeson CJ, Gummow, Callinan, Heydon and Crennan JJ); CAL No 14 Pty Ltd v Motor Accidents Insurance Board (2009) 239 CLR 390, 411–13 [49]–[51] (Gummow, Heydon and Crennan JJ).
AustLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.austlii.edu.au/au/journals/MelbULawRw/2010/17.html