Reproduced with permission. From Tony Blackshield, "Precedent" in Tony Blackshield, Michael Coper, and George Williams (eds), Oxford Companion to the High Court of Australia (2001, reprinted 2003, online edition 2007) 553. The Companion is available online through Oxford Reference Online.]
Precedent is the mechanism through which judges in common law countries like Australia exercise their law-making role. It is also a constraint on that role.
In a loose sense, all of us follow precedent (as do judges in all legal systems) whenever we face a new problem and look for guidance in the way that others have handled similar problems. What is distinctive about the common law practice is that judges must sometimes follow a relevant precedent even if they disagree with it. The practice implies simultaneously that every past decision of a superior court has made law, but that every present decision merely applies pre-existing law. These contradictory implications result in a unique combination of stability and flexibility.
Generally the hierarchy of binding precedent follows the hierarchy of courts. A lower court must follow the decisions of a higher court, but not the reverse: it is only because the higher court can reverse the decision that an appellate jurisdiction makes sense. Conversely, it is partly because of this power of reversal that a lower court must follow the decisions of the court which reviews its decisions on appeal.
Thus, so long as the Privy Council could entertain appeals from High Court decisions, the Court was absolutely bound by Privy Council rulings; but once appeals to the Privy Council from the High Court were abolished, the Court was able to declare that it was no longer so bound (see Viro v The Queen (1978)). In the NSW Court of Appeal in Hawkins v Clayton (1986), McHugh went further: he suggested that, once the Australia Acts 1986 had excluded the Privy Council altogether from Australian appellate hierarchies, all Privy Council decisions had ceased to bind any Australian court. The better interpretation of Viro, however, is that past Privy Council decisions now have the same degree of authority as the High Court’s own past decisions: they continue to bind the High Court unless the Court itself overrules them, and continue to bind other Australian courts until the High Court decides otherwise.
Historically, the power to lay down binding precedents was ascribed only to the superior courts. Yet in Babaniaris v Lutony Fashions (1987), the whole Court agreed that a 1953 decision by a state Workers Compensation Board should be treated as a precedent—because, as Mason explained, the Board had been exercising judicial power—and divided 3:2 on whether it should now be overruled. On the older view, a magistrate might not be bound by the ruling of a District Court or County Court—especially if no direct appellate link was involved. Yet in practice, no magistrate would hesitate to treat such a ruling as binding.
If appellate hierarchy is decisive, there may be other cases in which a lower court is not technically bound by a higher court. For example, a decision by a single Justice in the High Court’s original jurisdiction may not be binding on state Supreme Courts, since such a decision stands outside the appellate structure affecting those Courts (see Jacobs in Bone v Commissioner of Stamp Duties (1972) in the NSW Court of Appeal). Again, it has been argued (see Trade Practices Commission v Allied Mills (No 5) (1982)) that judges in the Federal Court of Australia have never been bound by Privy Council decisions—since the Federal Court was created only after Privy Council appeals from the High Court had already been abolished, so that Federal Court and Privy Council never did coexist in the same appellate hierarchy. A question never finally settled was whether Australian courts were strictly bound by Privy Council decisions on appeal from other jurisdictions (see Mayer v Coe (1968)). At least three Justices in Viro (Gibbs, Jacobs and Aickin) expressly held that all Privy Council decisions had the same status, regardless of the jurisdiction appealed from; but it is unlikely that the point need ever be resolved.
From the linkage of the precedent hierarchy to the appellate hierarchy, it follows that the High Court was never bound by decisions of the House of Lords. Despite this, the Court at one stage adhered to a policy that it would invariably follow House of Lords decisions, even if this involved overruling a previous decision of its own that the Court continued to prefer. The high water mark of this policy came in Piro v Foster (1943), where Bourke v Butterfield & Lewis (1926)—a landmark decision of the Knox Court—was overruled by the Latham Court in deference to the more conservative view of the House of Lords. Latham ruled that ‘it should now be formally decided’ (though still only as ‘a wise general rule of practice’)
that in cases of clear conflict between a decision of the House of Lords and of the High Court, this Court, and other courts in Australia, should follow a decision of the House of Lords upon matters of general principle.
The policy was sometimes explained as a pragmatic response to the overlapping membership of Appellate Committees of the House of Lords and Judicial Committees of the Privy Council—which meant that, once the House of Lords had spoken, the Privy Council was likely to agree. But the policy also reflected a strong belief in the system of precedent as a source of common law uniformity: the House of Lords had a uniquely authoritative role in determining the law of England, and acceptance of its decisions throughout the Empire would ensure that the common law developed uniformly in all British countries. At times, and with greater qualifications, this reasoning extended even to decisions of the English Court of Appeal (see especially Dixon’s analysis and McTiernan’s dissent in Waghorn v Waghorn (1942)). Latham’s pronouncement in Piro v Foster quoted what the Privy Council had said in Trimble v Hill (1879): ‘It is of the utmost importance that in all parts of the Empire where English law prevails the interpretation of that law by the courts should be as nearly as possible the same.’ A later Privy Council decision, in Robins v National Trust (1927), had recognised that a ‘Colonial Court’ might sometimes legitimately decline to follow an English court, but had added:
It is otherwise if the authority in England is that of the House of Lords. That is the supreme tribunal to settle English law, and that being settled, the Colonial Court, which is bound by English law, is bound to follow it.
On similar grounds, Privy Council decisions for all British jurisdictions, regardless of whether or not the strict doctrine of precedent required that they be treated as binding, were in practice accepted as binding; and the fact that High Court decisions were binding in all Australian jurisdictions was highly valued as an assurance of national unity. In Waghorn v Waghorn, Rich acknowledged that the uniformity thus achieved may sometimes be ‘uniformity of error’, but added complacently: ‘In that event it is at least uniformity.’ Be that as it may, the degree of uniformity throughout Australia achieved by the binding effect of High Court decisions has been substantial (see, for example, Criminal law defences).
However, while the emphasis on national unity remains an essential feature of the Court’s approach to its precedents, the emphasis on Empire unity faded along with the Empire itself. Robins v National Trust was distinguished: since their Lordships were speaking only of the duty of ‘an appellate Court in a colony which is regulated by English law’, it was argued that this did not apply to jurisdictions like those of Australia, where the legal system was based on a foundational reception of English law, but only to those colonies like Hong Kong where English law might still be said to apply by its own direct force. In Watts v Watts (1953), the Court recognised that the House of Lords in Preston-Jones v Preston-Jones (1951) had subjected allegations of adultery to a more stringent standard of proof than the flexible standard spelled out by Dixon in Briginshaw v Briginshaw (1938), but managed to reconcile the two views (and hence to adhere to Briginshaw) by the rationalisation that in Preston-Jones the legitimacy of a child was at stake, so that even on the basis of Briginshaw, a higher standard of proof might be required. And in Parker v The Queen (1963), Dixon led the whole Court in announcing that the Lords’ decision in DPP v Smith (1961) was so ‘misconceived and wrong’ in ‘fundamental’ respects that it ‘should not be used as authority in Australia at all’. Thereafter, in Australian Consolidated Press v Uren (1967) (see Defamation), the Privy Council conceded that the common law of England and Australia might develop along separate lines.
Although the hierarchical principle gives general guidance to the operation of precedent as between higher and lower courts, it gives no guidance when questions of precedent arise at the same hierarchical level—as to whether a single judge is bound by the decision of another single judge in the same court, and (particularly) as to whether a superior court is bound by its own prior decisions. In The Vera Cruz (No 2) (1880), the great English judge William Balliol Brett (later Lord Esher) explained that, while the old common law courts had in practice accepted their own decisions and each other’s decisions as binding, that practice was based on ‘no statute or common law rule’, but only on ‘judicial comity’. It follows that the extent to which a court is bound by its own previous decisions depends on the practice adopted by that court. In Australia, the practice of the state Supreme Courts has varied from state to state, and sometimes within the same court over time.
Uniquely, during the latter part of the nineteenth century, the House of Lords adopted a practice of regarding itself as irreversibly bound by its own decisions. The practice became entrenched with the London Tramways Case (1898), and was only abandoned by a ‘Practice Statement’ in 1966. The Privy Council never adopted such a practice, and neither did the High Court. In the Engine-Drivers Case (1913) Isaacs—with typical dogmatic overstatement—insisted that in this respect the High Court was bound to follow the Privy Council’s example. More persuasively, he invoked the judicial oath, by which each member of the Court is bound ‘to do right to all manner of people according to law’ (Isaacs’ italics). This meant, said Isaacs, that the Court must give effect to its own present understanding of the law:
A prior decision does not constitute the law, but is only a judicial declaration as to what the law is … [If] we find the law to be plainly in conflict with what we or any of our predecessors erroneously thought it to be, we have, as I conceive, no right to choose between giving effect to the law, and maintaining an incorrect interpretation. It is not, in my opinion, better that the Court should be persistently wrong than that it should ultimately be right.
For Isaacs, the test of when the Court should overrule its own prior decision was simple: it should do so whenever it thought the prior decision was ‘manifestly wrong’. At the other extreme lies Jacobs’ insistence in HC Sleigh v SA (1977) that the Court should overrule a prior decision only if it was ‘leading to social, economic or political consequences which cannot be tolerated by the nation’. Between these two extremes most Justices have steered a middle course. Almost invariably, however, recognition of the power to overrule is accompanied by emphasis that it should be exercised cautiously and sparingly. In the Second Territory Senators Case (1977), Stephen noted that while the Court ‘has always asserted its power to review its previous decisions’, it would do so only in ‘exceptional’ cases, after ‘most careful scrutiny’ of the precedent case and ‘full consideration’ of what the consequences of overruling might be. If these propositions are (as Stephen put it) ‘replete with adjectival qualifications’, that serves only to emphasise the judicial commitment to the maxim stare decisis (‘to stand by what has been decided’). Often understood as a general synonym for the whole doctrine of precedent, that maxim is in fact a specific enjoinder that a court should stand by its own decisions.
In the Second Uniform Tax Case (1957), for example, Dixon felt free to overrule the First Uniform Tax Case (1942) on a minor issue because he saw the decision on that point ‘as isolated, as receiving no support from prior decisions and as forming no part of what in one metaphor is called a stream of authority and in another a catena of cases’. By contrast, on the larger question of Commonwealth grants to the states under section 96 of the Constitution (see Economy, impact of Court’s decisions on), he held that, despite his own evident discomfort with the 1942 decision, it should be followed because it formed part of a trilogy of cases, beginning with the Federal Roads Case (1926), which ‘combine to give to s 96 a consistent and coherent interpretation’ whose ‘cumulative authority’ was ‘impossible to disregard’.
In Commonwealth v Hospital Contribution Fund (1982), Gibbs took a similar approach, concluding that it was appropriate to overrule the decisions in Kotsis v Kotsis (1970) and Knight v Knight (1971) in part because ‘they do not rest upon a principle that has been carefully worked out in a succession of cases … They stand alone and to overrule them will not unsettle the law in other respects.’ He added, secondly, that there was no unanimity of reasoning among the majority Justices in Kotsis; thirdly, that the decisions ‘achieve no useful result, but on the contrary lead to considerable inconvenience’; and, finally, that the decisions had not been acted upon in such a way that their overruling would be disruptive. In John v FCT (1989), these four criteria were adopted by five members of the Court.
Indeed, since the 1970s, attempts to articulate such criteria have become more elaborate (see Overruling). In A-G (NSW) v Perpetual Trustee Co (1952), Dixon remarked comfortably that the Court ‘has adopted no very definite rule as to the circumstances in which it will reconsider an earlier decision’; certainly he doubted ‘the wisdom or justice’ of the highly restrictive limits on overruling which the English Court of Appeal had set for itself in Young v Bristol Aeroplane Co (1944). But more recently, the need for precise criteria has apparently been felt more keenly. Moreover, since Evda Nominees v Victoria (1984), counsel wishing to submit that a High Court decision should be reopened must seek leave to do so; and once such leave has been given, the submission is heard by a full Bench of all available Justices.
Ironically, in all of this the Court has continued to be influenced by the House of Lords—where the 1966 assertion of power to ‘depart from’ prior decisions was followed by a series of anxious attempts to define the conditions under which it might be appropriate to do so. Paradoxically, their Lordships’ handling of precedent became more rigid than it had been before. A similar trend can be discerned in Australia.
Yet agreement on the factors that militate for or against overruling remains elusive. For example, the fact that a challenged precedent is of very recent vintage can be used both ways. Even the fact that a decision ‘stands by itself’ outside any ‘stream of authority’, which Dixon used in the Second Uniform Tax Case to support overruling, had been used by him in the Perpetual Trustee case to show that there was no compelling need to overrule (as there would have been if the challenged decision was in conflict with a ‘stream of authority’).
What has been clearly recognised is that the doctrine of precedent carries less weight in constitutional law, where judicial decisions perceived as errors, unlike those that occur in the development of the common law, are beyond simple legislative correction (see, for example, the Engineers Case (1920), Strickland v Rocla Concrete Pipes (1971), and Cole v Whitfield (1988)). Judges as different as Barwick and Murphy have insisted on the individual duty to give effect to the Constitution as each Justice understands it, even at the expense of precedent. ‘Always the Constitution remains the text’, said Barwick in Damjanovic v Commonwealth (1968). ‘The task is to apply the Constitution, not the judicial decisions’, said Murphy in Buck v Bavone (1976). And from opposite sides of the Second Territory Senators Case, each of their judgments reaffirmed that view.
In Stevens v Head (1993), Deane relied on those judgments to assert that, ‘in matters of fundamental constitutional importance’, individual Justices ‘are obliged to adhere to what they see as the requirements of the Constitution of which the Court is both a creature and the custodian’. Despite the decision in McKain v Miller (1991), he adhered to his own view of section 118 of the Constitution. Gaudron agreed, asserting that, while some constitutional provisions may ‘permit of different views as to their meaning’, section 118 ‘is not one of them’. In a trilogy of cases on military justice—Re Tracey; Ex parte Ryan (1989), Re Nolan; Ex parte Young (1991), and Re Tyler; Ex parte Foley (1994)—the same two Justices took a similar stance.
Even when precedents are not overruled, the extent to which they constrain the Court’s continuing re-evaluation of legal principles is limited. In Tasmania v Victoria (1935), for example, Dixon—following Brett’s analysis in The Vera Cruz—concluded that a tied vote yields no binding precedent, since, even if the reasoning of the Justices whose view prevails is unanimous, it has not been accepted by a majority of the Court.
Moreover, much of what is said in a precedent case can be set aside as obiter dicta, leaving only its ratio decidendi as binding; and debate as to what precisely is the ratio decidendi of a particular case serves often to limit its authority—especially if there is no agreed ratio at all, as Deane and Gaudron contended in the military justice cases. Even then, the formal doctrine remains that ‘the actual decision’ is binding; and while that confines the effect of precedent to a very narrow scope indeed, it means at least—as McHugh concluded in Tyler, quoting Lord Reid in Midland Silicones v Scruttons (1961)—that the Court must ‘apply that decision where the circumstances of the instant case “are not reasonably distinguishable from those which gave rise to the decision”’.
Yet questions of hierarchical authority, of whether a precedent will be overruled, and of what exactly is its binding scope, are only a small part of the influence of precedent on common law judicial reasoning. Around the core of the strict doctrine of precedent, and the broader conception of the common law as made and re-made by judges, the pattern of legal and judicial argument is dominated by review, comparison, analysis, synthesis, interpretation, and inductive restatement of all previous judicial pronouncements that might be relevant, whether technically binding or not. All such pronouncements are part of the authoritative legal materials which may or may not be decisive, but which cannot be ignored. In common law cases, judicial reasoning is based almost wholly on such prior pronouncements; and in statutory interpretation—and even constitutional interpretation—particular provisions are encrusted, over time, with a body of case law supplying not only alternative methods and assumptions of interpretation, but judicially determined meanings. It is through this looser sense of the practice of precedent, rather than in strict determination of which precedents are binding, that the habits of thought characteristic of courts in common law countries are formed.
- Garfield Barwick, Precedent in the Southern Hemisphere (1970)
- Anthony Mason, ‘The Use and Abuse of Precedent’ (1988) 4 ABR 93
- Julius Stone, Precedent and Law (1985)
- Alastair MacAdam and John Pyke, Judicial Reasoning and the Doctrine of Precedent in Australia (1998)