//
© 2026 Zoe. All rights reserved.
Zoe is a legal information platform. Always consult the Federal Register of Legislation for authoritative text.
[1996] HCA 24
High Court of Australia
1996-09-09
Brennan CJ, Gummow JJ, Levine J, Grove J
Original judgment source is linked above.
High Court of Australia Brennan CJ Dawson, Toohey, McHugh, Gaudron and Gummow JJ Kable v Director of Public Prosecutions (NSW) (FC 96/027) [1996] HCA 24
(a) Appeal to the Court of Appeal allowed with costs.
(b) Set aside the order of Levine J and in lieu thereof order that the application of the respondent be dismissed with costs.
The Bill for the Community Protection Act 1994 NSW (the Act) was duly passed by the Houses of the Parliament of New South Wales. Assent was given and the Act was proclaimed to come into force on 9 December 1994. Therefore at all material times the Act was on the statute book of the State.
The key provision of the Act is s 5 which reads as follows:
(1) On an application made in accordance with this Act, the Court may order that a specified person be detained in prison for a specified period if it is satisfied, on reasonable grounds:
(a) that the person is more likely than not to commit a serious act of violence; and
(b) that it is appropriate, for the protection of a particular person or persons or the community generally, that the person be held in custody.
(2) The maximum period to be specified in an order under this section is 6 months.
(3) An order under this section may be made against a person:
(a) whether or not the person is in lawful custody, as a detainee or otherwise; and
(b) whether or not there are grounds on which the person may be held in lawful custody otherwise than as a detainee.
(4) More than one application under this section may be made in relation to the same person.
A "serious act of violence" in s 5(1)(a) is defined in s 4 to mean an act of violence, committed by one person against another, that has a real likelihood of causing death or serious injury to the other person or that involves sexual assault in the nature of certain offences under the Crimes Act 1900 NSW. The reference to "the Court" in s 5(1) is to the Supreme Court of New South Wales [1] the jurisdiction of which under the Act is exercisable by a single Judge (s 24). A detention order may be made subject to such conditions as the Court may determine (s 9(1)). Under the Act proceedings for detention orders are civil proceedings (s 14) determined according to the balance of probabilities (s 15). Only the Director of Public Prosecutions may apply for a preventive detention order under s 5 or an interim detention order under s 7 (s 8).
The generality of these provisions was restricted in the course of the passage of the Bill through the Parliament. The liability to suffer the consequences of a detention order was limited so that the only person against whom a detention order might be made was the appellant, Gregory Wayne Kable. Section 3 of the Act reads as follows:
(1) The object of this Act is to protect the community by providing for the preventive detention (by order of the Supreme Court made on the application of the Director of Public Prosecutions) of Gregory Wayne Kable.
(2) In the construction of this Act, the need to protect the community is to be given paramount consideration.
(3) This Act authorises the making of a detention order against Gregory Wayne Kable and does not authorise the making of a detention order against any other person.
(4) For the purposes of this section, Gregory Wayne Kable is the person of that name who was convicted in New South Wales on 1 August 1990 of the manslaughter of his wife, Hilary Kable.
The appellant was originally charged with the murder of his wife but the Crown accepted his plea of diminished responsibility to manslaughter for which he was sentenced to imprisonment for a minimum term of four years with an additional term of one year and four months. The imminent release of the appellant from custody under this sentence was the occasion for both the enactment of the Act and the making of an application for a detention order against the appellant pursuant to the Act. The acts of the appellant relied upon to support the application were the sending of threatening letters through the mail. At the time when the application was made, the appellant was in custody pending the hearing of 17 charges [2] arising from his sending threatening letters through the mail. 1. Under s 85S of the Crimes Act 1914 Cth.
On 23 February 1995, Levine J issued a detention order for a period of six months in respect of the appellant but Grove J refused on 21 August 1995 to issue a further order. This appeal is brought in respect of the order issued by Levine J. The grounds argued on the appeal do not relate to the appropriateness of the making of the order if Levine J had jurisdiction to make it. The argument challenges the very existence of the Act as a law of the State of New South Wales.
This question is to be distinguished from the question whether the Act is beyond the power of the Parliament of New South Wales. This question is whether the instrument that stands on the statute book as the Community Protection Act 1994 has the character of a law. The instrument, enacted in due form and by due process, purports to create a power to make a detention order and it prescribes the procedure by which the order may be made and the consequences of the order when made. True it is that it singles out the appellant as the sole subject of a detention order, but a purported law has never been held to lack the character of a law simply because it affects the liberty or property of only a single individual. Acts of Attainder were nonetheless laws, as Sir Edward Coke accepted, albeit protesting that, in the procedure of imposing the attaint, the high court of Parliament ought to give example of justice to inferior courts [3] . The Act may be a law which, by reason of its specificity, is enacted in exercise of a power that is not purely legislative, but it is nonetheless a law. Specificity does not deny the character of law to an enactment that is otherwise within power. Private Acts of Parliament were a familiar form of laws in the nineteenth century English Parliament and were not open to question on that account [4] . Private Acts have been enacted at times by the Parliaments of this country. The next question is whether the Parliament had power to enact the Act.
It is submitted that the Act is in substance an exercise of judicial power rather than legislative power. On the assumption that the New South Wales Parliament cannot exercise judicial power or interfere in the judicial process [5] , it is submitted that the Act is beyond the powers of that Parliament.
There is something to be said for the view that, consistently with s 3, any application made within a short time of the enactment of the Act left the Supreme Court with no discretion. The general provisions of s 5 must yield to the provisions of s 3(1) and (2), so that the Court is commanded to protect the community by making a detention order against the appellant.
Assuming, without deciding, that the Act left nothing for the Supreme Court to decide but merely commanded the making of the formal detention order, the appellant submits that the Act is an exercise of judicial power or an interference in the judicial process and, on that account, is beyond the powers of the New South Wales Parliament. The submission is based on the proposition that the doctrine of separation of judicial power, an essential element of the Constitution of the Commonwealth, is part of the constitutional law of the State. That proposition has been rejected by the Supreme Court of New South Wales [6] as it has by the Supreme Court of other States in respect of their Constitutions [7] . For the reasons stated by Dawson J, I too would hold that that safeguard of liberty is not to be found in the Constitution of New South Wales despite the introduction in 1992 and the entrenchment in 1995 of Pt 9 of the Constitution Act 1902.
I am in general agreement also with his Honour's reasons for holding that, subject to the Commonwealth of Australia Constitution Act 1900 Imp, the Statute of Westminster Adoption Act 1942 Cth and the Australia Act 1986, the powers of the Parliament of New South Wales are not limited as to subject matter.
However, I would add a qualification and a comment to what his Honour has written, neither of which is material to the present case. First, there is the conundrum of the grant of general legislative power conferred by the "peace, order and good government" formula on a Parliament which might legislate itself or its powers out of existence. In my opinion, this conundrum is resolved by ss 106 and 107 of the Commonwealth Constitution. These are the sections which continue to sustain the Constitutions of the States and confirm their respective powers, including the power to amend their own Constitutions [8] . As s 107 continues the powers of the States, no power can be annihilated by State legislative action. If Parliament, as the repository of a power, were legislatively to deny itself that power, the denial would be inconsistent with s 107 and would be invalid.
Secondly, I would add a further comment to Dawson J's conclusion that "no non-territorial restraints upon parliamentary supremacy arise from the nature of a power to make laws for peace, order (or welfare), and good government or from the notion that there are fundamental rights which must prevail against the will of the legislature". If there were any restraints on the exercise of the powers of the Parliament, they would arise from entrenched provisions of the particular State Constitution [9] . Such restraints would be immune from modification unless the conditions expressed in the entrenching provisions were satisfied. When a submission is made that a particular measure is ultra vires because a restraint on legislative power is implied by an entrenching provision, the implication must clearly appear [10] . If the connection between the text and the propounded implication is tenuous or obscure, it would be wrong for a court by declaration to withdraw from public debate the matters to which the submitted restraint applies. If the constitutional text does not clearly support an implication of restraint, the court declaring the restraint is plunged into political controversy in which it is ill-fitted to engage and from which it is hard put to withdraw [11] .
Chapter III of the Constitution of the Commonwealth both limits the repositories of the judicial power of the Commonwealth and, in context, prescribes a separation of the function of the High Court and of other federal courts from the functions of the political branches of government. The separation of functions is derived from the structure of the Constitution and, in particular, from the distribution of legislative power to the Federal Parliament (s 1), of executive power to the Queen for exercise by the Governor-General (s 61) and of judicial power to the courts referred to in s 71.
In like manner, no functions that are not judicial can be conferred by the Commonwealth Parliament on a State court [12] , and thus the separation of State courts from the Legislative and the Executive branches of the Commonwealth Government is secured. But that separation does not purport to effect a separation of the courts of a State or Territory from the legislature or executive of the State or Territory. Nor does the Constitution purport to preclude State Parliaments from conferring a non-judicial power on a State court. It would be surprising if it did. Such a provision would have destroyed the State laws investing mining warden's courts, licensing courts and planning courts - to take only some instances - with extensive administrative powers.
However, it is submitted that, as ss 71 and 77(iii) authorise the investing of the judicial power of the Commonwealth in State courts, those courts must be capable of accepting and exercising the federal jurisdiction invested in them and that that capacity is dependent on their not being repositories of non-judicial power the exercise of which is incompatible with the exercise of federal judicial power. It has been accepted constitutional doctrine that, when the Commonwealth invests the judicial power of the Commonwealth in a State court, it must take that court constituted and organised as it is from time to time [13] . The autochthonous expedient contained in Ch III of providing for the vesting of federal jurisdiction in State courts left to the Commonwealth Parliament the selection of the courts in which federal jurisdiction should be invested. The submission that a State court cannot be a repository of both State non-judicial power and federal judicial power if the exercise of the former would be incompatible with the exercise of the latter attributes to Ch III a novel operation. On one approach Ch III would limit the power of the Parliament of a State to invest the courts of the State with incompatible non-judicial powers. On another approach, Ch III would limit the power of the federal Parliament to invest some State courts with federal jurisdiction. And on a third approach, the investing by the Commonwealth Parliament of federal jurisdiction in a State court would preclude the State Parliament from investing an incompatible non-judicial power in that State court.
In my opinion, Ch III does not operate in any of those ways. The test of incompatibility advanced by the submission is taken from the majority judgment in Grollo v Palmer [14] , a case which was not concerned with the jurisdiction or powers of a court but with the powers that might be conferred on individuals - personae designatae - who were judges of a Ch III Court. The incompatibility qualification applied to the persona designata doctrine has no counterpart in the context of possible limitations on the power of a State Parliament to invest courts of the State with non-judicial powers or the power of the Commonwealth Parliament to select whichever State courts it sees fit to invest with federal judicial power. No case has hitherto considered whether Ch III has any of the postulated operations. The absence of such a case indicates that the proposition has never before been advanced. Of course, novelty is not necessarily a badge of error but a suggestion that the power to invest State courts with federal judicial power might be limited or that the power of a State to invest the State's courts with non-judicial power might be limited would surely have provoked debate in the Constitutional Conventions. Yet they are as silent on the subject as the law reports. There is no textual or structural foundation for the submission.
The appellant, Gregory Wayne Kable, was charged with the murder of his wife, whom he stabbed to death in the house in which she lived with the two young children of the marriage. The marriage had broken up and there had been considerable acrimony between the appellant and his wife concerning custody of, and access to, the children. Before killing his wife, the appellant had engaged in violent behaviour towards her and had made threats of violence. In satisfaction of the murder charge the prosecution accepted a plea of guilty to manslaughter upon the basis of diminished responsibility. On 1 August 1990, the appellant was sentenced to a minimum term of imprisonment of four years and an additional term of one year and four months.
Once in prison, the appellant's behaviour was such as to cause serious concern that, upon his release, there would be a repetition of the same conduct that led to the death of his wife. In particular, he wrote a series of threatening letters, mainly to relatives of his deceased wife.
On 2 December 1994, the New South Wales Parliament passed the Community Protection Act 1994 NSW (the Act) which conferred jurisdiction upon the Supreme Court of New South Wales to make an order for the preventive detention of the appellant. The form of the Act was apparently suggested by the Community Protection Act 1990 Vic. The long title of the Act is "An Act to protect the community by providing for the preventive detention of persons who are, in the opinion of the Supreme Court, more likely than not to commit serious acts of violence". That is a misleading guide to the contents of the Act as it was eventually passed because, whilst the Bill in its original form was for an Act of general application, an amendment made during its passage through Parliament confined its application to the appellant [15] . Notwithstanding this, s 5, which is headed "Preventive detention orders", is in general terms and provides:
(1) On an application made in accordance with this Act, the Court may order that a specified person be detained in prison for a specified period if it is satisfied, on reasonable grounds:
(a) that the person is more likely than not to commit a serious act of violence; and
(b) that it is appropriate, for the protection of a particular person or persons or the community generally, that the person be held in custody.
(2) The maximum period to be specified in an order under this section is 6 months.
(3) An order under this section may be made against a person:
(a) whether or not the person is in lawful custody, as a detainee or otherwise; and
(b) whether or not there are grounds on which the person may be held in lawful custody otherwise than as a detainee.
(4) More than one application may be made in relation to the same person.
The object of the Act is "to protect the community by providing for the preventive detention of Gregory Wayne Kable" (s 3(1)) and in construing its provisions "the need to protect the community is to be given paramount consideration" (s 3(2)). Thus, notwithstanding that the Act provides for the appellant's imprisonment, ambiguities in it are not to be construed strictly in his favour but against him. Despite the fact that detention under the Act is based upon community protection, it does not provide for detention in other than an ordinary gaol. A detainee under the Act is taken to be a prisoner within the meaning of the Prisons Act 1952 NSW (s 22(1)).
Whilst s 17(1)(a) of the Act provides that the court is bound by the rules of evidence (see also s 14), the remaining paragraphs of that subsection, in combination with s 17(3), allow the court to have regard to material such as medical or prison records or reports, including hearsay, which would not otherwise be admissible in evidence. Proceedings are instituted by the Director of Public Prosecutions (s 8) but are civil proceedings (s 14) and the case against the appellant need only be proved on the balance of probabilities (s 15).
Section 11 requires the court after making a preventive detention order to appoint assessors, being qualified medical practitioners, psychiatrists or psychologists, to observe and report upon the appellant during the period for which the order is in force. Section 12 empowers the court to make orders for medical, psychiatric or psychological treatment to be made available to the appellant and s 21(1) requires reports to be furnished by the assessors and the Commissioner of Corrective Services on the appellant's condition and progress. Reports are to be prepared at least once while the order is in force and at any time the Director of Public Prosecutions so requires (s 21(2)). The reports are to contain particulars with respect to the appellant's general behaviour, the likelihood that he will commit a serious act of violence, the appropriateness of his continuing to be held in custody and the appropriateness of his continued detention in the same institution (s 21(3)). Also, reports prepared by an assessor are to describe the current state of the appellant's medical, psychiatric and psychological condition, the nature of any treatment made available to or undergone by the appellant during the period to which the report relates and an opinion as to any appropriate future treatment of the appellant (s 21(4)). On the application of the Director of Public Prosecutions or the appellant, the court may revoke a preventive detention order or amend it by reducing the period for which it is in force (s 13(1)). In determining an application for revocation or amendment, the court must have regard to the most recent reports by the assessors and the Commissioner for Corrective Services (s 13(2)). More than one such application may be made in relation to the same order (s 13(3)).
Shortly after the Act came into force, the Director of Public Prosecutions commenced proceedings against the appellant in the Supreme Court of New South Wales before Spender AJ. On 19 December 1994, his Honour made an order under s 17(1)(c) that the appellant be psychiatrically examined. On 30 December 1994, Hunter J made an interim detention order pursuant to s 7 of the Act. On 23 February 1995, Levine J ordered pursuant to s 5 that the appellant be detained in custody for a period of six months. An appeal was dismissed by the Court of Appeal on 9 May 1995 [16] . It is from that decision that this appeal is brought.
The appellant applied unsuccessfully on three occasions for revocation of the order made against him by Levine J. However, on 21 August 1995 (the day before the order of Levine J was due to expire), Grove J refused an application by the Director of Public Prosecutions for a second preventive detention order against the appellant who has now been released from custody. Nevertheless, the appellant remains liable at any time to be the subject of a further application that he be detained in custody.
The appellant confined his argument before us to an attack on the validity of the Act. The submissions which he made tended to overlap, and some of the grounds raised by the notice of appeal were not pursued. Whilst the appellant identified a number of features of the Act which he contended were obnoxious, in the end his argument was founded upon two of them, namely, the fact that the Act contemplates the imprisonment of the appellant for reasons other than the commission of a crime and the fact that it applies to the appellant alone. He contended first that the Act is invalid because it infringes common law rights which are so fundamental that they cannot be overturned by any legislature. Next, he said that the Act is beyond the power of the New South Wales Parliament to make laws for the peace, welfare, and good government of New South Wales pursuant to s 5 of the Constitution Act 1902 NSW because it is not a law within the meaning of that section. He then argued that the New South Wales Constitution embodies a separation of powers which the Act infringes. In addition he said that the Act is inconsistent with the requirements of Ch III of the Commonwealth Constitution. An argument based upon a constitutional requirement of equality under the law and before the courts was abandoned before us.
The New South Wales Parliament derives its legislative power from s 5 of the Constitution Act 1902 which provides that "[t]he Legislature shall, subject to the provisions of the Commonwealth of Australia Constitution Act, have power to make laws for the peace, welfare, and good government of New South Wales in all cases whatsoever ". It is unnecessary at this point to trace the history which lies behind this provision [17] because it is firmly established that its words confer a plenary power "and it was so recognised, even in an era when emphasis was given to the character of colonial legislatures as subordinate law-making bodies" [18] . That was clear before the passage of the Australia Acts [19] but it is put beyond question by s 2 of those Acts. The legislative power of the New South Wales legislature is no less than the legislative power of the Parliament of the United Kingdom within the scope of the grant of its power. As s 5 of the Constitution Act 1902 itself recognises, the power is subject to the Commonwealth of Australia Constitution Act 1900 Imp. Section 106 of the Commonwealth Constitution makes it clear that the Constitution of each State is subject to the Commonwealth Constitution, and under s 5 of the Australia Acts the powers of the States do not extend to legislation affecting the Commonwealth Constitution, the Commonwealth of Australia Constitution Act, the Statute of Westminster 1931 Imp or the Australia Acts themselves. And under s 6 of the Australia Acts the States are bound to observe any manner and form requirements for laws respecting the Constitution, powers or procedures of their parliaments. In addition, the words "peace, welfare, and good government of New South Wales" may be the source of whatever territorial restrictions upon the State's legislative powers are made necessary by the federal structure [20] .
But the important thing is that for present purposes the words "peace, welfare, and good government" are not words of limitation. As this Court observed in Union Steamship Co of Australia Pty Ltd v King [21] :
They did not confer on the courts of a colony, just as they do not confer on the courts of a State, jurisdiction to strike down legislation on the ground that, in the opinion of a court, the legislation does not promote or secure the peace, order and good government of the colony [22] . Just as the courts of the United Kingdom cannot invalidate laws made by the Parliament of the United Kingdom on the ground that they do not secure the welfare and the public interest, so the exercise of its legislative power by the Parliament of New South Wales is not susceptible to judicial review on that score. Up to that point, that passage would appear to be a complete answer to any suggestion that there are common law rights which are so fundamental that they cannot be overturned by legislation, but the Court added [23] : Whether the exercise of that legislative power is subject to some restraints by reference to rights deeply rooted in our democratic system of government and the common law (see Drivers v Road Carriers [24] ; Fraser v State Services Commission [25] ; Taylor v New Zealand Poultry Board [26] ); a view which Lord Reid firmly rejected in Pickin v British Railways Board [27] , is another question which we need not explore. Those words were prompted by remarks of Cooke J in the New Zealand Court of Appeal to the effect that "some common law rights may go so deep that even Parliament cannot be accepted by the Courts to have destroyed them" [28] . As this Court observed, that view was rejected by Lord Reid in Pickin v British Railways Board [29] . There he said: The idea that a court is entitled to disregard a provision in an Act of Parliament on any ground must seem strange and startling to anyone with any knowledge of the history and law of our Constitution
I must make it plain that there has been no attempt to question the general supremacy of Parliament. In earlier times many learned lawyers seem to have believed that an Act of Parliament could be disregarded in so far as it was contrary to the law of God or the law of nature or natural justice, but since the supremacy of Parliament was finally demonstrated by the Revolution of 1688 any such idea has become obsolete.
Lord Reid's reference to earlier times would appear to hark back to the view expressed by Coke CJ in Bonham's Case [30] . He said:
And it appears in our books, that in many cases, the common law will control Acts of Parliament, and sometimes adjudge them to be utterly void: for when an Act of Parliament is against common right and reason, or repugnant, or impossible to be performed, the common law will control it, and adjudge such Act to be void. Academic debate over the meaning of those words continues to the present time. It is unclear whether Coke CJ was intending to say that Acts of Parliament which are repugnant to the common law are void or whether he was merely laying down a rule of statutory interpretation. If he was intending the former, he appears to have had second thoughts, because in his Fourth Institute he described parliament's power as "transcendent and absolute", not confined "either for causes or persons within any bounds". He there contemplated the enactment of bills of attainder without trial and statutes contrary to Magna Carta without any suggestion of their invalidity [31] . 1. (1610) 8 Co Rep 107a at 118a [77 ER 638 at 652]. 2. See Coke, The Fourth Part of the Institutes of the Law of England (1809 ed), p 36.
However, Coke was not alone and there were other early expressions of opinion which appear to suggest that courts might invalidate Acts of Parliament which conflict with natural law or natural equity [32] . But they are of academic or historical interest only for such views did not survive the Revolution of 1688 or, at the least, did not survive for very long after it. Judicial pronouncements confirming the supremacy of parliament are rare but their scarcity is testimony to the complete acceptance by the courts that an Act of Parliament is binding upon them and cannot be questioned by reference to principles of a more fundamental kind. Indeed, it is a principle of the common law itself "that a court may not question the validity of a statute but, once having construed it, must give effect to it according to its tenor" [33] . There is more academic writing on the subject but it tends to dwell upon the apparent riddle posed by the question whether parliament can relinquish its powers by exercising them in order to do so. The answer to that riddle appears to lie in that area where law and political reality coincide. The same may be said of examples of extreme laws which would offend the fundamental values of our society which are sometimes suggested in disproof of parliamentary supremacy. It may be observed that a legislature wishing to enact a statute ordering that all blue-eyed babies be killed would hardly be perturbed by a principle of law which purported to deny it that power. Whether one speaks as Salmond does of "ultimate legal principles" [34] , or as Kelsen does of a grundnorm [35] , or as Hart does of the "ultimate rule of recognition" [36] , there can be no doubt that parliamentary supremacy is a basic principle of the legal system which has been inherited in this country from the United Kingdom.
In Stockdale v Hansard [37] Lord Denman CJ referred to the supremacy of parliament in terms which indicated his complete acceptance of the principle. And Willes J in Lee v Bude & Torrington Junction Railway Co [38] said: "I would observe, as to these Acts of Parliament, that they are the law of this land; and we do not sit here as a court of appeal from parliament." In Liyanage v The Queen [39] the Privy Council rejected the notion that the power of the Ceylon Parliament to make laws for the peace, order, and good government of the island was limited by an inability to pass laws which offend against fundamental principles. The case was relevantly concerned with the suggestion of Lord Mansfield that there were limitations upon the power of the Crown to make laws for a colony which were contrary to fundamental principles [40] , but there can be no doubt that it is implicit in the decision that the Ceylon Parliament inherited the sovereignty exercisable by the United Kingdom Parliament with the consequence that it was not fettered by any limitation of the kind suggested.
In Madzimbamuto v Lardner-Burke [41] Lord Reid, speaking for the Privy Council, said:
It is often said that it would be unconstitutional for the United Kingdom Parliament to do certain things, meaning that the moral, political and other reasons against doing them are so strong that most people would regard it as highly improper if Parliament did these things. But that does not mean that it is beyond the power of Parliament to do such things. If Parliament chose to do any of them the courts could not hold the Act of Parliament invalid. And in Duport Steels Ltd v Sirs [42] Lord Edmund-Davies observed: From time to time some judges have been chafed by this supremacy of Parliament, whose enactments, however questionable, must be applied. In the same case Lord Scarman said [43] : But in the field of statute law the judge must be obedient to the will of Parliament as expressed in its enactments. In this field Parliament makes, and un-makes, the law: the judge's duty is to interpret and to apply the law, not to change it to meet the judge's idea of what justice requires. 1. [1969] 1 AC 645 at 723. 2. [1980] 1 WLR 142 at 164; [1980] 1 All ER 529 at 548. 3. Duport Steels Ltd v Sirs [1980] 1 WLR 142 at 168[1980] 1 All ER 529 at 551
In the New South Wales Court of Appeal, Kirby P expressed his agreement with Lord Reid in British Railways Board v Pickin [44] . In Builders' Labourers' Federation v Minister for Industrial Relations [45] he said:
I agree with Lord Reid's conclusion. I do so in recognition of years of unbroken constitutional law and tradition in Australia and, beforehand, in the United Kingdom. That unbroken law and tradition has repeatedly reinforced and ultimately respected the democratic will of the people as expressed in Parliament. It has reflected political realities in our society and the distribution of power within it. 1. [1974] AC 765 at 782. 2. (1986) 7 NSWLR 372 at 405.
In Polyukhovich v The Commonwealth [46] I referred to the possibility that a court might in "quite extraordinary circumstances" judge that legislation which otherwise falls within its power is invalid as not being for peace, order, and good government. Similarly, as appears from the passage set out above, in Union Steamship Co of Australia Pty Ltd v King [47] this Court reserved the question whether the exercise of that legislative power is subject to restraints to be found in fundamental principle. This case throws up the question reserved in those cases and it should now be answered by saying that no non-territorial restraints upon parliamentary supremacy arise from the nature of a power to make laws for peace, order (or welfare), and good government or from the notion that there are fundamental rights which must prevail against the will of the legislature. The doctrine of parliamentary supremacy is a doctrine as deeply rooted as any in the common law. It is of its essence that a court, once it has ascertained the true scope and effect of an Act of Parliament, should give unquestioned effect to it accordingly.
The appellant's contention was that the Act is not a law within the meaning of s 5 of the Constitution Act 1902. That is to say, although the New South Wales Parliament has power under s 5 to make laws for the peace, welfare, and good government of the State in all cases whatsoever, the Act is not a law or, at all events, a law of the kind referred to in the section. The argument appears to be based upon the Austinian notion that a law is a command which obliges a person or persons to a course of conduct. The Act, it was said, does not satisfy that definition because it is directed at the preventive detention of the appellant alone.
It may be observed that the Act clearly does oblige those persons charged with its administration to a course of conduct and that many statute laws, including private Acts, are particular rather than general in their application. It may also be observed that Austin, in drawing a distinction between general commands and commands of occasional or particular application, recognised that the latter, issued by a sovereign legislature, might be called laws [48] .
But nothing is to be gained for present purposes by a jurisprudential analysis of what constitutes a law. What is involved is the construction of s 5 of the Constitution Act 1902 and it is readily apparent that in the context of that section the word "laws" is synonymous with the word "statutes". If any limitation is to be found upon the power of the Parliament, it is to be found elsewhere in the Constitution Act 1902 [49] or in the words "peace, welfare, and good government" which, as I have said, impose no non-territorial limitation. Clearly, in my opinion, when s 5 speaks of the making of laws by the legislature it speaks of the making of statute law, that is to say, of the passage of Bills through Parliament which subsequently receive royal assent. The legislature is defined in s 3 as the Crown with the advice and consent of the Legislative Council and Legislative Assembly [50] . Section 5 is not seeking to impose a distinction between statutes which embody a law and those which do not, according to a definition of law imported from elsewhere. In an appropriate context (and s 5 is one), a statute may be synonymous with a law because of the manner of its creation. It may be so even if the law lacks validity for it is quite permissible to speak of an invalid law in such a context. The appellant's argument must be rejected.
The New South Wales Constitution and the separation of powers
The appellant contended that the Constitution Act 1902 embodies a separation of powers whereby the judicial power of the State is separated from its legislative and executive powers. As a consequence, he said, judicial power is exercisable only by the judiciary, which is incapable of exercising functions which are not judicial or ancillary thereto. However, there is nothing in the structure of the Constitution Act 1902 to support this contention.
That Act vests legislative power in the legislature (s 5) and leaves unaffected the exercise of executive power through the Governor otherwise than on the advice of the Executive Council (s 35A). There is provision for the continuance of an Executive Council (s 35B) and the Premier and other Ministers of the Crown are to be appointed from its members (s 35E). Whilst Pt 9 of that Act is headed "The Judiciary" nowhere does it provide that the judicial power of the State is vested in the judiciary. Section 53, which is contained in Pt 9, provides that no holder of judicial office may be removed from office save on an address of both Houses of Parliament seeking removal on the ground of proved misbehaviour or incapacity. There are additional provisions relating to the suspension of judicial office and the fixing or changing of retirement age (ss 54, 55). Section 56 provides that Pt 9 does not prevent abolition by legislation of a judicial office whether that be done directly or indirectly by the abolition of a court or part of a court.
Whilst these provisions are concerned with the preservation of judicial independence, they cannot be seen as reposing the exercise of judicial power exclusively in the holders of judicial office. Nor can they be seen as precluding the exercise of non-judicial power by persons in their capacity as holders of judicial office. They clearly do not constitute an exhaustive statement of the manner in which the judicial power of the State is or may be vested [51] . Had Pt 9 attempted such an exercise it would have cut across a long history of the exercise of non-judicial power by the courts and the exercise of judicial power by bodies exercising non-judicial functions [52] .
The Constitution Act 1902 may be contrasted with the provisions of the Commonwealth Constitution, in particular ss 1, 61 and 71. Those sections respectively vest the legislative power of the Commonwealth in the Parliament, the executive power in the Executive and the judicial power in the Judicature. Section 1 appears at the commencement of Ch 1, which is headed "The Parliament". Section 61 appears at the commencement of Ch II which is headed "The Executive Government". Section 71 appears at the commencement of Ch III which is headed "The Judicature". In R v Kirby; Ex parte Boilermakers' Society of Australia [53] this Court held that this pattern could not be treated as a "mere draftsman's arrangement" or as "meaningless and of no legal consequence". It is because the judicial power of the Commonwealth is vested by Ch III in those courts which it identifies and is dealt with nowhere else (save for s 51(xxxix)) that this Court was compelled to conclude that no functions other than judicial functions may be reposed in the federal judicature and that no powers which are foreign to the judicial power may be attached to courts created by or under that chapter. Not only that, but it was recognised that the position and constitution of the federal judicature was bound up in the federal structure established by the Constitution, "for upon the judicature rested the ultimate responsibility for the maintenance and enforcement of the boundaries within which governmental power might be exercised" [54] .
The latter consideration has no application to the judicature of a State and the failure of the New South Wales Constitution to vest judicial power exclusively in the judicature must be fatal to any contention that the separation of that power from the other powers of government is a constitutional requirement. Even if it could be said that it was required, it might, in contrast to the requirement imposed by the Commonwealth Constitution, be disregarded by an Act of Parliament, for in that respect the New South Wales Constitution is "uncontrolled" [55] . It remains true, therefore, as was said by the New South Wales Court of Appeal in Clyne v East [56] , that the structure and provisions of the Constitution Act 1902 provide no ground for importing into it a principle of separation of powers.
The appellant placed reliance upon the Privy Council decision of Liyanage v The Queen [57] in which it was held that the Constitution of Ceylon contained a separation of powers notwithstanding that it was silent as to the vesting of judicial power. But the judicature in Ceylon was in existence before the Constitution and was operating under the Courts Ordinance which contained the jurisdiction and procedure of the courts established under the Charter of Justice in 1833. The Charter provided that "the entire administration of justice, civil and criminal therein, shall be vested exclusively in the courts erected and constituted by this Our Charter". There was, for this reason, so the Privy Council found, no need to make specific reference in the Constitution to the vesting of judicial power, but the organisation of that instrument was otherwise such as to manifest "an intention to secure in the judiciary a freedom from political, legislative and executive control". The Constitution's silence as to the vesting of judicial power was, the Privy Council said, "consistent with its remaining, where it had lain for more than a century, in the hands of the judicature". It was, it said, "not consistent with any intention that henceforth it should pass to or be shared by, the executive or the legislature" [58] .
There is no such background to the New South Wales Constitution which inherited the United Kingdom model under which the extent to which a separation of powers was observed was conventional rather than compelled by any constitutional mandate [59] . The New South Wales Court of Appeal was clearly correct in concluding in Clyne v East that, notwithstanding that the Supreme Court of New South Wales also owes its origin to a Charter of Justice, no basis could be found in the provisions of the Constitution Act 1902 for isolating judicial power from the other powers of government [60] . To do so would confine the legislative power conferred by s 5 of the Constitution Act 1902. It is clear that it is not so confined and, as I have explained, it extends to the judiciary. As is well established, the ultimate source of the power contained in s 5 is the Imperial Act (18 & 19 Vict c 54), known as the Constitution Statute 1855, and the Act of the Colony of New South Wales which forms the schedule to the statute (17 Vict No 41), which is called the Constitution Act [61] . Section 49 of the latter Act recognised the power of the New South Wales legislature to abolish, alter or vary the constitution and functions of the courts of the colony. That power may be traced through to s 5 of the Constitution Act 1902. It is for that reason and because of the structure of that Act (which at one time contained no provision at all dealing with the judicature) that no useful comparison can be made with the Ceylon Constitution which was analysed in Liyanage v The Queen .
It is true that Pt 9 of the Constitution Act 1902 is doubly entrenched. Under s 7B(1) of the Constitution Act 1902 specified provisions may not be repealed or altered until the repeal or alteration has been approved by referendum. Those provisions include s 7B itself, so providing a double entrenchment. Part 9 was included in s 7B(1), but that did not occur until after it had been inserted in the Constitution Act 1902 in 1992 [62] . The entrenchment of Pt 9 was effected by s 3 of the Constitution (Entrenchment) Amendment Act 1992 NSW, but that did not come into force until 2 May 1995, that is to say, until after the Act came into force. But, as I have said, Pt 9 deals only with matters relating to judicial independence and does not purport to vest judicial power exclusively in the judiciary or to require the judiciary to exercise only judicial power. Even if the Act amounted to the legislative exercise of a judicial function or required the court to exercise a non-judicial function, it would not fall within the terms of s 7B(1) as repealing or amending Pt 9 of the Constitution Act 1902. It is therefore unnecessary to consider whether, the Act having come into force before the Constitution (Entrenchment) Amendment Act, s 7B(1) would have any application.
Under s 71 of the Commonwealth Constitution, which is the first section of Ch III, the judicial power of the Commonwealth is vested in the High Court "and in such other federal courts as the Parliament creates, and in such other courts as it invests with federal jurisdiction". Whilst the nature of the judicial power so vested is not stated, the matters over which it may be exercised are set out with some particularity in ss 75 and 76. Under s 77(iii) the Parliament may make laws "[i]nvesting any court of a State with federal jurisdiction". In s 39(2) of the Judiciary Act 1903 Cth Parliament has exercised that power by conferring federal jurisdiction (with certain exceptions and qualifications) upon "[t]he several Courts of the States" in the matters specified in ss 75 and 76.
In providing for the vesting of federal jurisdiction in State courts, s 77(iii) plainly contemplated that jurisdiction might be vested in a court which did not have the characteristics required of a federal court under Ch III. As has been said, Ch III requires those courts created by or under it to exercise the judicial power vested in them separately from the other powers of Government. In the interests of judicial independence s 72 also requires that the judges of those courts be appointed until a specified age, be not removable except upon an address of both Houses of Parliament praying for their removal on the ground of proved misbehaviour or incapacity, and be not subject to diminution of their remuneration during their continuance in office [63] .
Plainly those requirements do not extend to State courts which may be invested with federal jurisdiction under s 77(iii). Section 77(iii) speaks of existing institutions the characteristics of which did not necessarily and did not in fact satisfy those requirements. Indeed, in South Australia at federation an appeal lay from the Supreme Court to the Court of Appeals which comprised the Governor in Executive Council. Special provision had to be made in s 73 of the Constitution to include the Court of Appeals [64] . As I have said, in New South Wales there has never been a constitutional requirement of separation of powers and Pt 9 of the Constitution Act 1902, which guarantees a measure of judicial independence, was inserted only in 1992. But as Isaacs J said in R v Murray and Cormie; Ex parte The Commonwealth [65] :
The Constitution, by Ch III, draws the clearest distinction between federal Courts and State Courts, and while enabling the Commonwealth Parliament to utilise the judicial services of State Courts recognises in the most pronounced and unequivocal way that they remain "State Courts." And as Knox CJ, Rich and Dixon JJ pointed out in Le Mesurier v Connor [66] : The Parliament may create Federal Courts, and over them and their organisation it has ample power. But the Courts of a State are the judicial organs of another Government. They are created by State law; their existence depends upon State law; that law, primarily at least, determines the constitution of the Court itself, and the organisation through which its powers and jurisdictions are exercised. 1. See Official Record of the Debates of the Australasian Federal Convention, (Melbourne) 31 January 1898, vol IV, pp 332-333. 2. (1916) 22 CLR 437 at 452. 3. (1929) 42 CLR 481 at 495-496.
Considerations such as those were in the mind of Griffith CJ when he said in Federated Sawmill, Timberyard and General Woodworkers' Employees' Association (Adelaide Branch) v Alexander [67] that "when the Federal Parliament confers a new jurisdiction upon an existing State Court it takes the Court as it finds it, with all its limitations as to jurisdiction, unless otherwise expressly declared". They were in the mind of Latham CJ when he said in Peacock v Newtown Marrickville & General Co-operative Building Society No 4 Ltd [68] that "[t]he Commonwealth Parliament cannot, by virtue of s 77(iii) of the Constitution, under the guise of conferring jurisdiction upon a State court, in effect create a new Federal court, possibly without observing the conditions imposed upon such creation by s 72 of the Constitution " [69] .
The suggestion that the Constitution does not permit of two grades of judiciary exercising the judicial power of the Commonwealth, or that Ch III does not draw the clear distinction between State and federal courts which it has hitherto been thought to, simply ignores the fact that the Constitution ensures security of tenure and of remuneration in respect of judges of courts created by or under Ch III but does not do so in respect of judges of State courts invested with federal jurisdiction. It equally ignores the fact that the Constitution does not require that State courts only exercise judicial power. The suggestion that the Act is invalid because it compromises the institutional impartiality of the Supreme Court of New South Wales ignores the fact that the mechanisms for ensuring judicial impartiality and independence - security of tenure and remuneration, and separation from the other arms of government - are not constitutionally prescribed for State courts notwithstanding that they are prescribed for courts created by or under Ch III. It is difficult to conceive of a clearer distinction.
It is true that some qualifications must be placed upon the statement that the federal Parliament must, in exercising its power under s 77(iii), take a State court as it finds it. Under s 79 Parliament may prescribe the number of judges to exercise federal jurisdiction. And clearly Parliament may extend the jurisdiction of the court because that is precisely what s 77(iii) envisages. Also it may regulate the practice and procedure which the State court is to follow in exercising the invested jurisdiction. But it may go no further than is necessary for that purpose; it may not legislate with respect to the court itself [70] . That is to say, it may not alter the character or constitution of the court. And just as the Parliament cannot do that under s 77(iii), it is clear that there is nothing in Ch III itself which does so.
And yet the appellant's argument, as I understand it, is that the Commonwealth Constitution, and Ch III in particular, precludes a State court, as the potential or actual repository of federal jurisdiction, from having functions conferred upon it by the State legislature which are incompatible with Ch III. Either that or, so the appellant's argument goes, a law, such as s 39(2) of the Judiciary Act 1903 Cth, which invests a State court with federal jurisdiction, is inconsistent with a State law conferring a function upon that court which is incompatible with Ch III and must, for that reason, prevail over the State law under s 109 of the Constitution.
It may be said at the outset that such an argument simply denies the proposition, hitherto accepted without question, that Ch III, and s 77(iii) in particular, treats State courts as existing institutions. The result is that, so long as they are in fact courts, Ch III is unconcerned with whether they comply with the requirements of Ch III for courts created by or under that chapter. State courts are not created by or under Ch III and, provided they are courts within the meaning of s 77(iii), it matters not for the purposes of Ch III what functions they perform in exercising the jurisdiction vested in them by State legislation. That is for the State legislature to determine. As Mason J said in The Commonwealth v Hospital Contribution Fund [71] :
Generally speaking, the Parliament of a State may in the exercise of its plenary legislative power alter the composition, structure, and organisation of its Supreme Court for the purposes of the exercise of State jurisdiction Ch III of the Constitution contains no provision which restricts the legislative competence of the States in this respect. Nor does it make any discernible attempt to regulate the composition, structure or organisation of the Supreme Courts as appropriate vehicles for the exercise of invested federal jurisdiction. It is therefore sensible and natural to read the expression "any Court of a State" in s 77(iii) as referring to State courts in the sense explained by Gibbs J in Kotsis .
His Honour there observed that the exercise of federal jurisdiction did not call for a curial organisation different in kind from that established for the exercise of State jurisdiction [72] . In this situation there is every reason for supposing that the framers of the Constitution intended to arm the Parliament of the Commonwealth with a power to invest federal jurisdiction in a State court as it happened to be organised under State law from time to time. Although the Commonwealth Parliament has no power to alter the structure or organisation of State courts, its freedom of action is completely preserved. It has the choice of investing State courts with federal jurisdiction or of establishing appropriate federal courts. (Emphasis added.)
Whether State courts invested with federal jurisdiction are part of the federal judicature or not is a question of little practical significance, save perhaps in considering the application of the incidental power conferred by s 51(xxxix). They remain State courts even though, when exercising federal jurisdiction, they may be regarded as a component of the federal judicature. There is no one court system in Australia [73] . Each of the States has its own hierarchy which is governed by State legislation. The federal courts created under s 71 of the Constitution constitute a different system. Of course, the whole can be regarded as an entirety. After all, the different parts have a common origin in law and the common law precedes the emergence of the different jurisdictions and applies in them all. Not only that, but the creation of the High Court as a court of appeal - now the final court of appeal - from the courts of all jurisdictions, federal and state, has a unifying influence upon both the common law and also in a more general way. But our legal system, though integrated, is not a unitary system. The States are distinct jurisdictions and the enactments of each of their legislatures are confined in their operation so that in other States their recognition is governed by common law principles and such requirements as flow from the full faith and credit required by s 118 of the Constitution. Federal law, of course, is binding on all courts whether exercising federal jurisdiction or not [74] . The system is a federal system and, whilst the framers of the Constitution might have established a judicial system which was neither State nor federal but simply Australian, they did not do so [75] . It is therefore dangerous to attempt to draw conclusions from the fact that the Australian legal system may be regarded as a whole. It may be, but as a matter of legal analysis that is to stop short of an appreciation of its different parts.
Clearly, a State court may exercise executive or legislative as well as judicial functions, where, as in the case of New South Wales, the State constitution does not require judicial power to be separated from executive and legislative power. In so doing the State court is exercising a function which may not be exercised by a federal court under Ch III. That chapter does not allow a combination of judicial and executive or legislative power in federal courts. As the majority said in the Boilermakers' Case [76] :
The basal reason why such a combination is constitutionally inadmissible is that Ch III does not allow powers which are foreign to the judicial power to be attached to the courts created by or under that chapter for the exercise of the judicial power of the Commonwealth. That passage is careful to omit any reference to State courts invested with federal jurisdiction under s 77(iii) and that is because under State constitutions the combination which is impermissible under Ch III is permissible. 1. (1956) 94 CLR 254 at 289.
Once it is recognised that there is no requirement in the New South Wales Constitution that courts in that State perform solely judicial functions and that, notwithstanding that characteristic, they are nevertheless courts which may be invested with federal jurisdiction under s 77(iii) of the Commonwealth Constitution, any question of incompatibility with Ch III upon the ground that the State court is required to perform executive or legislative functions must disappear. Certainly those functions may not be performed by a federal court created by or under Ch III. That is because the federal court is precluded by the separation of judicial power from performing them. But as far as State courts are concerned there is no incompatibility with Ch III because that chapter accepts those courts as existing institutions which may be invested with federal jurisdiction notwithstanding that they are not subject to any doctrine of separation of powers.
It may be that, in referring to incompatibility, the appellant used the word in the sense in which it was used by the majority in Grollo v Palmer [77] where it was said that the ability of parliament to confer a non-judicial function on a judge of a Ch III court as a designated person rather than as a judge is subject to the limitation that the non-judicial function must not be incompatible either with the judge's performance of his or her judicial functions or with the proper discharge by the judiciary of its responsibilities as an institution exercising judicial power. So much, it was said in that case, was to be "implied from the separation of powers mandated by Chs I, II and III of the Constitution and from the conditions necessary for the valid and effective exercise of judicial power" [78] .
But the judicial power of which the majority were there speaking was the judicial power exercised by a federal court created by or under Ch III - a Ch III court. The nature of that judicial power is, as was acknowledged in Grollo , very much determined by the separation of powers which the Constitution requires to be observed in relation to such a court. As the Privy Council observed in the Boilermakers' Case [79] , the separation of powers is, in a federal system, a guarantee of the absolute independence of the judiciary which is "the bulwark of the constitution against encroachment whether by the legislature or by the executive" [80] . What is incompatible with the exercise of the judicial power of the Commonwealth by a Ch III court may not be incompatible with the exercise of the judicial power of the Commonwealth by a court which is not restricted by any separation of powers. As Grollo makes clear, the concept of incompatibility is derived from the separation of powers and does not have a life of its own independent of that doctrine. Five members of the Court recognised that very point in Wilson v Minister for Aboriginal and Torres Strait Islander Affairs [81] where they said that " Grollo was concerned with constitutional incompatibility, derived from the constitutional separation of the functions of the Judiciary from the functions of the Parliament and the Executive". (Emphasis added.)
In any event, no question of incompatibility in the sense in which the concept was used in Grollo arises in this case. In that case, the persona designata doctrine was held to justify the giving of an executive function to a judge of a Ch III court. Incompatibility was raised only to demonstrate that the persona designata doctrine has its limits. In this case, where the New South Wales court is not affected by any separation of powers, there is no call for the invocation of the persona designata doctrine and no question of incompatibility arises from the vesting in the court [82] of an executive function. Nor can it be said in these circumstances that a particular type of executive function is incompatible with the exercise of judicial power any more than it can be said that a particular type of judicial function is inconsistent with the exercise of executive or legislative power. The reasons, which are readily apparent, for the adoption of a separation of powers in a federal structure do not have the same force in a unitary state. Be that as it may, New South Wales has not adopted that doctrine so that there can be no incompatibility between the exercise of judicial power and the exercise of executive or legislative power by a court of that State. And there can be no incompatibility with Ch III arising from that situation because it was within the contemplation of those who framed that chapter that federal jurisdiction might be vested in State courts exercising executive or legislative functions as well as judicial powers. If the position is that the conferral of some non-judicial functions on a State court by a State parliament is so undesirable as to be unconstitutional, a quasi-separation of powers will be established in the States, totally contrary to the position which had hitherto been unquestioned and without the virtue of that separation being able to be determined by any criterion other than undesirability. Of course, only judicial functions may be vested by the federal Parliament in State courts pursuant to s 77(iii) [83] , but that is an entirely different thing.
Because Ch III contemplates that federal judicial power will be exercised by a State court which is not otherwise confined to the exercise of judicial power, it makes no difference that in this case the Supreme Court of New South Wales may have been exercising federal jurisdiction in making orders against the appellant because he raised in his defence the question of the invalidity of the Act under the Commonwealth Constitution. It was the appellant's contention that this made the case a matter arising under the Constitution or involving its interpretation within the meaning of s 76(1) of the Constitution and thus within the ambit of the federal jurisdiction vested in the court under s 39(2) of the Judiciary Act. I am aware of the views expressed in Felton v Mulligan [84] to the effect that once federal jurisdiction is attracted, even by a point raised in a defence, the jurisdiction exercised throughout the case will remain federal jurisdiction. For the purpose of determining the available avenues of appeal that may be the only practical approach, but I would observe that it may lead to a very artificial result in a case such as this, namely, that the Supreme Court of New South Wales was exercising federal jurisdiction in ordering the preventive detention of the appellant under a New South Wales Act. Perhaps the answer may lie in the suggestion by Barwick CJ in Felton v Mulligan [85] that separate and discrete questions may arise, the determination of which will constitute separate proceedings. But there is no need to pursue that matter here.
It also follows from what I have said that a federal law, such as s 39(2) of the Judiciary Act, which invests a State court with federal jurisdiction, is not inconsistent with a State law which confers functions upon that court of a type which could not be conferred upon a Ch III court. Such a federal law may be passed pursuant to s 77(iii) of the Constitution, which, as has been said again and again, envisages the vesting of federal jurisdiction in State courts as existing institutions possessing characteristics which are not and cannot be possessed by a Ch III court, such as the capacity to exercise executive or legislative power as well as judicial power. Section 39(2) of the Judiciary Act does not and could not require a State court which it invests with federal jurisdiction to exercise judicial power separately and it is thus not inconsistent with a State law which confers upon the court a function which is other than judicial.
For these reasons, the Act is in my opinion valid. In reaching this conclusion I have not found it necessary to examine the view advanced by the appellant that the Act authorises the court to impose punishment upon him by way of an executive or legislative act for reasons other than his commission of a crime. However, in Chu Kheng Lim v Minister for Immigration [86] it was recognised that "[i]nvoluntary detention in cases of mental illness or infectious disease can also legitimately be seen as non-punitive in character". It is apparent from the Act that the New South Wales legislature perceived that there was a gap in the law arising from the fact that there may exist in the community persons with personality disorders which strongly dispose them to commit acts of violence but who are not mentally ill and thus are unable to be detained under mental health legislation [87] . The need for legislation in that situation was foreseen by Deane J in Veen v The Queen [No 2] [88] when he said:
[T]he protection of the community obviously warrants the introduction of some acceptable statutory system of preventive restraint to deal with the case of a person who has been convicted of violent crime and who, while not legally insane, might represent a grave threat to the safety of other people by reason of mental abnormality if he were to be released as a matter of course at the end of what represents a proper punitive sentence. The detention which the Act authorises the court to impose upon the appellant is, the respondent contends, preventive rather than punitive, although the appellant understandably points to the fact that the detention is in a prison, a place of punishment, rather than some other institution. Moreover, it is not apparent that an order that the appellant be detained represents the exercise of executive or legislative power rather than of judicial power [89] . On an application made in accordance with the Act, the court is required to be satisfied that the criteria laid down by s 5(1) are met before making a detention order. The issues raised are not predetermined by the legislation, as the refusal by Grove J to grant a second preventive detention order demonstrates. Clearly the Act does not amount to a bill of attainder or of pains and penalties. It does not involve a legislative judgment of criminal guilt and, in any event, does not have an ex post facto operation. 1. (1992) 176 CLR 1 at 28. 2. See, eg, Mental Health Act 1990 NSW, s 9(1). See also Williams, Psychopathy, Mental Illness and Preventive Detention: Issues Arising from the David Case, Monash University Law Review, vol 16 (1990) 161, at pp 165-166. 3. (1988) 164 CLR 465 at 495. 4. See Palling v Corfield (1970) 123 CLR 52 at 58-59, 62, 64-65, 68-69.
In the light of the conclusions which I have reached, these considerations raise matters which go to the desirability of the Act rather than to its validity. Notwithstanding that the wisdom of the policy adopted by the legislature is open to question, the policy is a matter for the legislature rather than for this Court.
The Court of Appeal of New South Wales dismissed an appeal by the appellant against orders made by Levine J under the Community Protection Act 1994 NSW (the Act) [90] .
Section 5(1) of the Act empowers the Supreme Court, on application made in accordance with the Act, to order that "a specified person be detained in prison for a specified period" (a "preventive detention order") if the Court is satisfied on reasonable grounds:
(a) that the person is more likely than not to commit a serious act of violence; and
(b) that it is appropriate, for the protection of a particular person or persons or the community generally, that the person be held in custody.
While the Act gives rise to some important questions, it also has some odd features. Undoubtedly these are the result of amendments made while the Bill was in progress through Parliament, without sufficient attention being paid to their impact on other provisions of the Act. For instance, the long title refers to "An Act to protect the community by providing for the preventive detention of persons who are, in the opinion of the Supreme Court, more likely than not to commit serious acts of violence". On its face then, the Act is of application to that class of persons whom it identifies. Yet s 3(1) expresses the object of the Act to be "to protect the community by providing for the preventive detention of Gregory Wayne Kable" while s 3(3) "authorises the making of a detention order against Gregory Wayne Kable and does not authorise the making of a detention order against any other person". There is no definition as such of the expression "a specified person" as it appears in s 5(1) but the provision must be read as if "Gregory Wayne Kable" were substituted for "a specified person". This produces further anomalies. For instance, s 10 prohibits the making of a detention order "against a person who is under the age of 16 years" though there can be no such person on whom the Act may operate. Clearly enough, the original intention of the legislature was to enact a statute of more general application; what emerged was legislation directed at one person only.
The maximum period of detention to be specified in an order under s 5 is six months [91] . The Act does permit the making of further orders pursuant to s 5. On 21 August 1995 Grove J refused to make a further order under the Act. However, that does not preclude the making of a further application against the appellant [92] .
Curiously, proceedings under the Act are identified as "civil proceedings" (s 14) and the standard of proof is "on the balance of probabilities" (s 15). In any proceedings under the Act the Court is bound by the rules of evidence (s 17(1)(a)) but it may order the production of medical records and reports and other documents (s 17(1)(b)) in which event, it would seem, their contents may constitute evidence which ordinarily would be inadmissible (see s 17(3)). The jurisdiction of the Court is exercisable by a single Judge, with a right of appeal against the making of or refusing to make a preventive detention order to the Court of Appeal on a question of law, or fact, or mixed law and fact (ss 24, 25).
The appellant's attack on the validity of the legislation took several forms, not all of which were entirely discrete from or independent of each other. At the forefront of the attack was an argument that wore two faces. One was that the Act constituted an exercise of judicial power by the Parliament of New South Wales. The other was that the Act invests in the Supreme Court of New South Wales a non-judicial power which is incompatible with Ch III of the Commonwealth Constitution. Counsel for the appellant said that this latter argument was not concerned with any doctrine of the separation of powers, as was the former argument, but only with the consequences of conferring upon a State court a jurisdiction which is incompatible with the exercise of federal judicial power. The argument of incompatibility has its foundation in the judicial power of the Commonwealth as identified by Ch III.
The appellant further argued that the Act is beyond the legislative power of the Parliament of New South Wales because it authorises the detention of the appellant without prescribing a rule of conduct by which he is to be judged. It was said, therefore, not to constitute a law. There were other attacks on the Act on the footing that it treats the appellant differently from all other persons and that it is a bill of attainder or of pains and penalties.
The power of the legislature of New South Wales is, by reason of the Constitution Act 1902 NSW, s 5, "subject to the provisions of the Commonwealth of Australia Constitution Act to make laws for the peace, welfare, and good government of New South Wales in all cases whatsoever". In Union Steamship Co of Australia Pty Ltd v King [93] this Court said that such a power is indistinguishable from the power to make laws "for the peace, order and good government" of a territory and that [94] :
[T]he words "for the peace, order and good government" are not words of limitation [T]hey do not confer on the courts of a State, jurisdiction to strike down legislation on the ground that, in the opinion of a court, the legislation does not promote or secure the peace, order and good government of the [State] Whether the exercise of that legislative power is subject to some restraints by reference to rights deeply rooted in our democratic system of government and the common law is another question which we need not explore. 1. (1988) 166 CLR 1. 2. Union Steamship (1988) 166 CLR 1 at 10.
It follows that no attack may be made on the Act simply on the ground that it does not secure the peace, welfare, and good government of New South Wales unless, perhaps, it falls within the category tentatively identified in the passage last quoted. It does not appear that the appellant seeks to bring his case directly within that category. But he does argue for a limitation on legislative power in the sense that the Parliament of New South Wales, not being a recipient of judicial power or function, may not itself exercise such a power or function.
In Fencott v Muller [95] Mason, Murphy, Brennan and Deane JJ referred to judicial power as the power of a sovereign authority to decide controversies between its subjects or between itself and its subjects. Their Honours continued:
The unique and essential function of the judicial power is the quelling of such controversies [that is, controversies between the subjects of a sovereign authority or between the authority and its subjects] by ascertainment of the facts, by application of the law and by exercise, where appropriate, of judicial discretion. The appellant's contention was that an application to the Supreme Court under the Act is in truth a formality, the outcome of which is dictated by the legislation. Although s 5(1) provides that the Court "may" order detention, if satisfied as to the criteria in pars (a) and (b), the appellant submitted that when the Act is read as a whole not only must the Court exercise the power once the criteria have been met [96] but that it is left with no determinative function. Thus the Act was said to be an exercise by the legislature of judicial power by directing the Court to find the criteria satisfied and in consequence to make a preventive detention order. 1. (1983) 152 CLR 570 at 608. 2. See Finance Facilities Pty Ltd v Federal Commissioner of Taxation (1971) 127 CLR 106 at 134-135 Mitchell v The Queen (1996) 184 CLR 333 at 345-346
The respondent challenged the proposition that the Act is an exercise of judicial power by the legislature. On the respondent's argument there is nothing in the Act which dictates the outcome of an application under s 5. It remains for the Supreme Court to be satisfied (albeit on the balance of probabilities) on reasonable grounds that the appellant is more likely than not to commit a "serious act of violence" [97] and that it is appropriate for the protection of a particular person or persons or the community generally, that the appellant be held in custody.
In my view, s 5(1) does require a consideration by the Court of the criteria in pars (a) and (b) and accordingly it obliges the Court to be satisfied, on reasonable grounds, that it is appropriate that a preventive detention order be made. It is true that s 3(2) reads: "In the construction of this Act, the need to protect the community is to be given paramount consideration." But the requirement is related to the construction of the provisions of the Act rather than the assessment of the evidence necessary to satisfy the criteria. That the Court may consider an order not to be appropriate in the circumstances is borne out by the refusal of Grove J to make an order under the section. Nevertheless, if the Court is satisfied in terms of pars (a) and (b), it can hardly do other than make a preventive detention order.
In any event the appellant has not demonstrated that the doctrine of separation of powers operates in a constitutional sense in New South Wales so as to preclude the Parliament enacting a statute which ensures a particular result in what might be regarded as an exercise of judicial power. In Building Construction Employees and Builders' Labourers Federation (NSW) v Minister for Industrial Relations (the Builders' Labourers Federation Case) [98] the Court of Appeal held that the courts could not declare invalid the Builders Labourers Federation (Special Provisions) Act 1986 NSW even though the Act was, in directing the outcome of pending litigation, an exercise of judicial power. In the course of his judgment, Street CJ said [99] :
Parliament in this State has power to adjudicate between parties by an exercise of judicial power
I recognise that the New South Wales Parliament has judicial power.
Priestley JA discussed this aspect, pointing out that the Supreme Court of Judicature Act 1873 UK effectively brought to an end the existing superior courts and brought into existence one Supreme Court in which litigation already begun would be continued. His Honour observed [100] : [T]his alteration in the English judicial system does not appear to have given rise to any comment relating to the doctrine of the separation of powers. 1. (1986) 7 NSWLR 372. 2. Builders' Labourers Federation Case (1986) 7 NSWLR 372 at 381. 3. Builders' Labourers Federation Case (1986) 7 NSWLR 372 at 416-417.
The appellant conceded that the Parliament of New South Wales may confer upon the Supreme Court functions which would be regarded as non-judicial in the context of Ch III [101] . Nevertheless, he argued that separation of powers should be inferred from the Constitution of New South Wales, in particular by reason of Pt 9. Part 9 was inserted to secure the independence of the judiciary by providing for the appointment of judges and the circumstances in which they may be removed from office. The appellant contended that this gave rise to a constitutional separation of judicial power from legislative power.
The respondent replied, first, that the double entrenching of Pt 9 occurred after the events relevant to the appeal. Part 9 of the Constitution Act was inserted by the Constitution (Amendment) Act 1992 NSW and commenced on assent [102] . It was not at that stage doubly entrenched. It became so on 2 May 1995, following a referendum and the Constitution (Entrenchment) Amendment Act 1992 NSW. Levine J made his order on 23 February 1995.
Secondly, the respondent submitted that in any event Pt 9 is of limited operation. In particular, it is said, Pt 9 does not purport to be "an exhaustive statement of the manner in which the judicial power of the [State of New South Wales] is or may be vested" [103] .
Part 9 is of limited operation in the sense that it deals with the appointment and removal of judges and not with wider considerations. Judicial independence does not of itself protect the judicial process from legislative interference as Street CJ noted in the Builders' Labourers Federation Case [104] . There is no real comparison between Pt 9 and Ch III of the Commonwealth Constitution. Chapter III vests the judicial power of the Commonwealth exclusively in the courts which it identifies and in clear terms recognises the separation of the judicial power of the Commonwealth from legislative and executive powers [105] . It is a consequence of that division that the Parliament of the Commonwealth may not interfere with the judicial process itself [106] .
But it follows from what has been said that there is nothing in the Constitution (NSW) which prevents the legislature from exercising judicial power.
The appellant also says that the qualification in s 5 of the Constitution Act - subject to the provisions of the Commonwealth of Australia Constitution Act - means that the Supreme Court, as one of the judicial institutions of this country, may not act in a manner inconsistent with Ch III of the Commonwealth Constitution. Section 5 is a qualification on the power of the Parliament of New South Wales; it does not impose the requirements of Ch III on a State court exercising State jurisdiction.
However, the appellant argues to the same end by reason of the Commonwealth Constitution itself. Section 71 of the Constitution vests the judicial power of the Commonwealth in the High Court "and in such other federal courts as the Parliament creates, and in such other courts as it invests with federal jurisdiction". Section 77(iii) empowers the Parliament to make laws "[i]nvesting any court of a State with federal jurisdiction". Effect is given to s 77(iii) by s 39(2) of the Judiciary Act 1903 Cth which invests the "several Courts of the States" with "federal jurisdiction" in all matters in which the High Court has original jurisdiction or in which original jurisdiction can be conferred upon it, subject to s 38 which identifies those matters in which the jurisdiction of the High Court is exclusive of those courts. To the extent that they are invested with federal jurisdiction, the federal courts and the courts of the States exercise a common jurisdiction [107] . It follows that in the exercise of its federal jurisdiction a State court may not act in a manner which is incompatible with Ch III of the Commonwealth Constitution.
The respondent accepted that in the present case the Supreme Court was exercising federal jurisdiction vested in it by s 39 of the Judiciary Act. The reason for the concession was that the appellant was relying upon what the Solicitor-General for New South Wales described as "federal constitutional points", not only before this Court but at first instance and on appeal to the Court of Appeal. The reference to "federal constitutional points" was a reference to submissions made on behalf of the appellant that an implication is to be drawn from the Commonwealth Constitution that legislation, whether federal or State, that is directed against or discriminates against an individual is invalid. In addition the appellant relied upon s 80 of the Constitution to argue that a charge of the offence created by the Act must be tried by a judge and jury. The federal judicial power may be attracted by the defence raised to a claim for relief [108] . Thus, it was said, federal jurisdiction was attracted in the present case, whereupon "there is no room for the exercise of a State jurisdiction which apart from any operation of the Judiciary Act the State court would have had" [109] . Professor Lane has said [110] :
Jurisdiction and the judicial power of the Commonwealth are two distinct notions, calling on different constitutional provisions and different decisional law. Although the argument before the Court proceeded at times by reference to federal jurisdiction, in truth it is the judicial power of the Commonwealth with which the Court is concerned, a power applicable by reason of s 71 of the Constitution. The Supreme Court of New South Wales was required, at first instance and on appeal, to determine questions arising under the Constitution. In those circumstances s 39(2) of the Judiciary Act, read with s 77(iii) of the Constitution, conferred jurisdiction on the Supreme Court to determine those questions. Section 71 of the Constitution ensured that the judicial power of the Commonwealth was engaged in those circumstances. 1. Felton v Mulligan (1971) 124 CLR 367 2. Felton v Mulligan (1971) 124 CLR 367 at 373, per Barwick CJ. 3. Lane, The Australian Federal System, 2nd ed (1979), p 446.
The argument advanced on behalf of the appellant was that the Act vests in the Supreme Court of New South Wales a non-judicial power which is offensive to Ch III of the Constitution. Hence any exercise of that power would be unconstitutional and the Act conferring the power would be invalid. Reliance was placed on what Dixon CJ, McTiernan, Fullagar and Kitto JJ said in R v Kirby; Ex parte Boilermakers' Society of Australia [111] :
The organs to which federal judicial power may be entrusted must be defined, the manner in which they may be constituted must be prescribed and the content of their jurisdiction ascertained. These very general considerations explain the provisions of Ch III of the Constitution. The argument is not one which relies upon the alleged separation of legislative and judicial functions under the Constitution of New South Wales. Rather it is that the jurisdiction exercised under the Act is inconsistent with Ch III of the Commonwealth Constitution because the very nature of the jurisdiction is incompatible with the exercise of judicial power. 1. (1956) 94 CLR 254 at 268.
The appellant's senior counsel was at pains to eschew any notion that his submission involved challenging the well-established doctrine that, in investing judicial power in the States, the Parliament must take the State courts as it finds them. For instance, the appellant accepted that State courts may be vested with non-judicial powers and functions. His argument was that a State court exercising federal jurisdiction, in the sense discussed earlier, could not be the recipient of powers and functions which were incompatible with the very nature of judicial power. In other words, the issue as presented by the appellant was not one of judicial versus legislative or executive power but of incompatibility with the essence of judicial power. Hence, no question of taking the courts as the Parliament finds them truly arose. It follows that passages in the judgments of this Court which emphasise that the courts of a State are the judicial organs of the State and not of the Commonwealth do not resolve the questions now facing this Court [112] . By reason of the issues raised in the case, the Supreme Court exercised federal jurisdiction. It is not the investing of the Supreme Court with federal jurisdiction that is in issue; it is the exercise of federal jurisdiction by the Supreme Court in the circumstances arising under the Act that is challenged.
In Grollo v Palmer [113] the Court held that the vesting in designated judges of the Federal Court of the power to issue interception warrants was not incompatible either with the judge's performance of his or her judicial functions or with the proper discharge by the judiciary of its responsibilities as an institution exercising judicial power. Nevertheless the Court emphasised that "no function can be conferred that is incompatible with the proper discharge by the judiciary of its responsibilities as an institution exercising judicial power" [114] . It is true that the proposition was enunciated in the context of the power to confer non-judicial functions on judges as designated persons but in my view it holds good whenever Ch III of the Constitution is operative. And Mistretta v United States [115] , to which the majority judgment refers with approval, is couched in terms of constitutional doctrine. The emphasis in the judgment of the Supreme Court of the United States is on "the integrity of the Judicial Branch".
The appellant's argument of incompatibility of function rests on several foundations. But fundamentally it relies upon the nature of the Act whereby the Supreme Court may order the imprisonment of a person although that person has not been adjudged guilty of any criminal offence. The Supreme Court is thereby required to participate in a process designed to bring about the detention of a person by reason of the Court's assessment of what that person might do, not what the person has done.
The Act speaks of likelihood "to commit a serious act of violence". If the power to detain were the consequence of the actual commission of a serious act of violence, it might be little different from the power to impose an indeterminate sentence to be found in various statutes [116] . In those cases, however, some prior conduct in the form of the commission of an offence of a prescribed nature is the basis upon which an indeterminate sentence may be ordered. The appellant's complaint is that, while prior conduct may bring him to the attention of the authorities, no such conduct is identified as the basis for the making of an order under s 5. No doubt prior conduct has an evidentiary part to play; without it there could hardly be a conclusion that the appellant was more likely than not to commit a serious act of violence. In that respect it was a relevant consideration that the appellant had been charged with the murder of his wife and that subsequently he pleaded guilty of manslaughter, a plea which was accepted by the Crown by reason of his diminished responsibility. But the order for his detention was not made by reason of his commission of that offence. Likewise, it may have been a relevant consideration, in terms of the Act, that the appellant was facing 17 charges alleging contravention of s 85S of the Crimes Act 1914 Cth in making improper use of postal services by sending threatening letters to various persons including his children. But the Act required no determination of his guilt for any of those offences as a condition of the order made against him.
In Chu Kheng Lim v Minister for Immigration [117] Brennan, Deane and Dawson JJ said:
[P]utting to one side the exceptional cases to which reference is made below, the involuntary detention of a citizen in custody by the State is penal or punitive in character and, under our system of government, exists only as an incident of the exclusively judicial function of adjudging and punishing criminal guilt. The exceptional cases to which their Honours referred are committal to custody pending trial, involuntary detention in cases of mental illness or infectious disease, the traditional power of Parliament to punish for contempt and of military tribunals to punish for breach of military discipline [118] . In Lim the Court was concerned with provisions of the Migration Act 1958 Cth which authorised the detention of non-citizens for a specified period for the purposes of expulsion or deportation. The validity of those provisions as an exercise of the constitutional power conferred in relation to aliens by s 51(xix) of the Commonwealth Constitution was upheld by the Court. 1. (1992) 176 CLR 1 at 27. 2. Lim (1992) 176 CLR 1 at 28
The situation in the present case is quite different from that in Lim . Preventive detention under the Act is an end in itself. And the person so detained "is taken to be a prisoner within the meaning of the Prisons Act 1952 " (s 22(1)). It is not an incident of the exclusively judicial function of adjudging and punishing criminal guilt. It is not part of a system of preventive detention with appropriate safeguards, consequent upon or ancillary to the adjudication of guilt [119] . Nor does it fall within the "exceptional cases" mentioned in Lim , directly or by analogy. In the present case the Act requires the Supreme Court to exercise the judicial power of the Commonwealth in a manner which is inconsistent with traditional judicial process [120] .
The extraordinary character of the legislation and of the functions it requires the Supreme Court to perform is highlighted by the operation of the statute upon one named person only. In this respect the Act is virtually unique [121] . It does not define "a specified person" by reference to any class or category and it carries no consequences for any person, other than the appellant, to whom its language might otherwise be applicable.
The Act answers that aspect of incompatibility which was identified in Grollo v Palmer as "the performance of non-judicial functions of such a nature that public confidence in the integrity of the judiciary as an institution is diminished" [122] . The function exercised by the Supreme Court under the Act offends Ch III which, as I said in Harris v Caladine [123] , reflects an aspect of the doctrine of separation of powers, serving to protect not only the role of the independent judiciary but also the personal interests of litigants in having those interests determined by judges independent of the legislature and the executive [124] . The function offends that aspect because it requires the Supreme Court to participate in the making of a preventive detention order where no breach of the criminal law is alleged and where there has been no determination of guilt. On that ground I would hold the Act invalid. It is not possible to sever s 5 from the rest of the Act which exists only to give effect to that section.
This conclusion makes it unnecessary to deal with the other grounds relied upon by the appellant. I mention only one, namely, that the Act is a bill of pains and penalties, that is, that it is "a legislative act which inflicts punishment without a judicial trial" [125] . The appellant referred to the advice of the Privy Council in Liyanage v The Queen [126] where the Privy Council said of certain statutes of Ceylon:
One might fairly apply to these Acts the words of Chase J, in the Supreme Court of the United States in Calder v Bull [127] : "These Acts were legislative judgments; and an exercise of judicial power." The Act does prescribe criteria of which the Court must be satisfied before making an order. Limited though it is, the role of the Supreme Court is not reduced to saying no more than that the person charged has been identified as fitting a description laid down in the Act [128] . The Act is not invalid on that ground, divorced from any consideration of Ch III. 1. Cummings v Missouri (1866) 71 US 277 at 323See also Polyukhovich v The Commonwealth (the War Crimes Act Case) (1991) 172 CLR 501 at 535-536, 646, 685-686, 719-721. 2. [1967] 1 AC 259 at 291. 3. (1799) 3 Dallas 386. 4. Polyukhovich (1991) 172 CLR 501 at 647-648
However the Act is invalid by reason of the incompatibility with Ch III of the Commonwealth Constitution that its implementation produces. If the Act operated on a category of persons and a defence to an application for a preventive detention order was confined to a challenge that the criteria in s 5(1) had not been met, different questions might arise. In that situation the judicial power of the Commonwealth might not be involved; that is something on which it is unnecessary to comment. But here the judicial power of the Commonwealth is involved, in circumstances where the Act is expressed to operate in relation to one person only, the appellant, and has led to his detention without a determination of his guilt for any offence. In that event validity is at issue, not simply the reach of the Act in a particular case.
Accordingly the appeal should be allowed and the application by the respondent against the appellant dismissed.
Mr Kable (the appellant) appeals to this Court from an order of the Court of Appeal of the Supreme Court of New South Wales dismissing his appeal from a "preventive detention order" made by Levine J under s 5(1) of the Community Protection Act 1994 NSW (the Act). It was submitted for the appellant in this Court, as it was before Levine J and in the Court of Appeal, that it is beyond the power of the Parliament of New South Wales to confer power on the Supreme Court to make the order in question.
Several arguments were advanced in favour of the appellant's contention. I need deal with one only, namely, that Ch III of the Constitution impliedly prevents the Parliament of a State from conferring powers on the Supreme Court of a State which are repugnant to or inconsistent with the exercise by it of the judicial power of the Commonwealth.
It is convenient to consider the constitutional position of State courts before turning to the provisions of the Act. It has been said that "[t]he Constitution, by Chapter III, draws the clearest distinction between federal Courts and State Courts, and recognises in the most pronounced and unequivocal way that they remain "State Courts" " [129] . However, that is not a distinction that appears in s 71, the opening provision of Ch III. On the contrary, the first sentence of that section provides, without distinction between State and federal courts created by the Parliament, that:
The judicial power of the Commonwealth shall be vested in a Federal Supreme Court, to be called the High Court of Australia, and in such other federal courts as the Parliament creates, and in such other courts as it invests with federal jurisdiction. 1. R v Murray and Cormie; Ex parte The Commonwealth (1916) 22 CLR 437 at 452, per Isaacs J.
Moreover, no distinction appears in the concluding provisions of Ch III, namely, s 78, which is concerned with the right to proceed against the Commonwealth and the States in respect of matters within the limits of the judicial power of the Commonwealth, s 79, which provides that "[t]he federal jurisdiction of any court may be exercised by such number of judges as the Parliament prescribes" (emphasis added), and s 80, which requires trial by jury for all indictable offences against the laws of the Commonwealth.
Special provision is made in s 74 with respect to appeals from this Court to the Privy Council. And separate provision is made in ss 73, 75 and 76 as to the jurisdiction of this Court. Sections 75 and 76 also serve to identify the matters which fall for resolution in the exercise of the judicial power of the Commonwealth. Neither s 75 nor s 76 refers to any court other than this Court. There are, however, references to other courts in s 73 which provides with respect to the appellate jurisdiction of this Court. By s 73(i) and (iii), respectively, this Court has "jurisdiction, with such exceptions and subject to such regulations as the Parliament prescribes" to hear and determine appeals from various determinations made by Justices of this Court and the Inter-State Commission. And by s 73(ii) it has jurisdiction, again subject to such exceptions and regulations as are prescribed, "to hear and determine appeals from all judgments, decrees, orders, and sentences [o]f any other federal court, or court exercising federal jurisdiction; or of the Supreme Court of any State, or of any other court of any State from which at the establishment of the Commonwealth an appeal [lay] to the Queen in Council". No distinction is made in that subsection between State courts, as courts exercising federal jurisdiction, and federal courts created by the Parliament.
There is also special provision with respect to the Supreme Courts of the States in the second and third paragraphs of s 73. By the second paragraph, no exception or regulation prescribed by the Parliament "shall prevent [this] Court from hearing and determining any appeal from the Supreme Court of a State in any matter in which at the establishment of the Commonwealth an appeal [lay] from such Supreme Court to the Queen in Council". And it is provided in the third paragraph of s 73 that "[u]ntil the Parliament otherwise provides, the conditions of and restrictions on appeals to the Queen in Council from the Supreme Courts of the several States shall be applicable to appeals from them to the High Court". However, these paragraphs do not distinguish between State courts and federal courts created by the Parliament as repositories of the judicial power of the Commonwealth.
No mention has yet been made of ss 72 and 77, the only other provisions to be found in Ch III. If they are put to one side, the provisions of Ch III clearly postulate an integrated Australian court system for the exercise of the judicial power of the Commonwealth, with this Court at its apex as a constitutional court and as a court exercising appellate jurisdiction for the whole of Australia, and with no distinction, so far as concerns the judicial power of the Commonwealth, between State courts and federal courts created by the Parliament.
Section 72 is concerned exclusively with the appointment, tenure and remuneration of the Justices of this Court and federal courts created by the Parliament. As to the matters which fall within the limits of the judicial power of the Commonwealth, s 77 confers power on the Parliament to make laws:
(i) Defining the jurisdiction of any federal court other than the High Court:
(ii) Defining the extent to which the jurisdiction of any federal court shall be exclusive of that which belongs to or is invested in the courts of the States:
(iii) Investing any court of a State with federal jurisdiction.
Again, s 77 does not distinguish between State courts and federal courts created by the Parliament as repositories of the judicial power of the Commonwealth. It does, however, recognise that the other courts which may be invested with federal jurisdiction are State courts. When s 77 is considered in conjunction with s 72 which, as earlier indicated, provides as to the appointment, tenure and remuneration of the members of this Court and federal courts created by the Parliament, it is correct to say, by reference to those provisions, that Ch III recognises that this Court and other federal courts are creatures of the Commonwealth and that State courts are the creatures of the States.
Once it is accepted that State courts are the creatures of the States and are constitutionally recognised as such, it follows that it is for the States and for the States alone to determine the appointment, tenure and remuneration of State judges and the structure, organisation and jurisdictional limits of State courts. In that sense, it is correct to say, as it often is, that the Commonwealth must take State courts as it finds them. However, it should be remembered that that dictum originates in the judgment of Griffith CJ in Federated Sawmill, Timberyard and General Woodworkers' Employees' Association (Adelaide Branch) v Alexander [130] , a case involving the question whether jurisdictional limits imposed by State law on a State court applied in matters of invested federal jurisdiction. It was in that context that his Honour said that "when the Federal Parliament confers a new jurisdiction upon an existing State Court it takes the Court as it finds it, with all its limitations as to jurisdiction, unless otherwise expressly declared" [131] - a vastly different statement from the unqualified proposition that the Commonwealth must take a State court as it finds it.
Neither the recognition in Ch III that State courts are the creatures of the States nor its consequence that, in the respects indicated, the Commonwealth must take State courts as it finds them detracts from what is, to my mind, one of the clearest features of our Constitution, namely, that it provides for an integrated Australian judicial system for the exercise of the judicial power of the Commonwealth. Moreover, neither that recognition nor that consequence directs the conclusion that State Parliaments may enact whatever laws they choose with respect to State courts. If Ch III requires that State courts not exercise particular powers, the Parliaments of the States cannot confer those powers upon them. That follows from covering cl 5, which provides that the Constitution is "binding on the courts, judges, and people of every State and of every part of the Commonwealth, notwithstanding anything in the laws of any State", and from s 106, by which the Constitution of each State is made subject to the Australian Constitution. And so much was recognised in The Commonwealth v Queensland [132] where it was said that State legislation in violation of "the principles that underlie Ch III" is invalid.
The question whether the Constitution requires that State courts not have particular powers conferred upon them depends, in my view, on a proper understanding of the integrated judicial system for which Ch III provides - the "autochthonous expedient" [133] , as it has been called. One thing which clearly emerges is that, although it is for the States to determine the organisation and structure of their court systems, they must each maintain courts, or, at least, a court for the exercise of the judicial power of the Commonwealth. Were they free to abolish their courts, the autochthonous expedient, more precisely, the provisions of Ch III which postulate an integrated judicial system would be frustrated in their entirety. To this extent, at least, the States are not free to legislate as they please.
Two other matters of significance emerge from a consideration of the provisions of Ch III. The first is that State courts are neither less worthy recipients of federal jurisdiction than federal courts nor "substitute tribunals" [134] , as they have sometimes been called. To put the matter plainly, there is nothing anywhere in the Constitution to suggest that it permits of different grades or qualities of justice, depending on whether judicial power is exercised by State courts or federal courts created by the Parliament.
The second and, perhaps, the more significant matter which emerges from a consideration of the provisions of Ch III is, as I pointed out in Leeth v The Commonwealth [135] , that State courts, when exercising federal jurisdiction "are part of the Australian judicial system created by Ch III of the Constitution and, in that sense and on that account, they have a role and existence which transcends their status as courts of the States". Once the notion that the Constitution permits of different grades or qualities of justice is rejected, the consideration that State courts have a role and existence transcending their status as State courts directs the conclusion that Ch III requires that the Parliaments of the States not legislate to confer powers on State courts which are repugnant to or incompatible with their exercise of the judicial power of the Commonwealth.
The prohibition on State legislative power which derives from Ch III is not at all comparable with the limitation on the legislative power of the Commonwealth enunciated in R v Kirby; Ex parte Boilermakers' Society of Australia [136] . The Boilermakers' doctrine, as it is sometimes called, prevents the Parliament of the Commonwealth from conferring judicial power on bodies other than courts and prevents it from conferring any power that is not judicial power or a power incidental thereto on the courts specified in s 71 of the Constitution. It also prevents the Parliament from conferring functions on judges in their individual capacity if the functions are inconsistent with the exercise of judicial power in the sense explained in Grollo v Palmer [137] . The limitation on State legislative power is more closely confined and relates to powers or functions imposed on a State court, rather than its judges in their capacity as individuals, and is concerned with powers or functions that are repugnant to or incompatible with the exercise of the judicial power of the Commonwealth.
Although the limitation is one relating to the conferral of powers on courts, rather than on judges in their capacity as individuals, it is, nevertheless, one that is closely related to the limitation on Commonwealth power to confer functions on judges of this and other federal courts in their capacity as individuals. In both cases the limitation derives from the necessity to ensure the integrity of the judicial process and the integrity of the courts specified in s 71 of the Constitution.
It remains to be considered whether the power purportedly conferred on the Supreme Court by s 5(1) of the Act is repugnant to or incompatible with the exercise of the judicial power of the Commonwealth. Section 5(1) provides that, on the application of the Director of Public Prosecutions (see s 8), the Supreme Court
may order that a specified person be detained in prison for a specified period if it is satisfied, on reasonable grounds:
(a) that the person is more likely than not to commit a serious act of violence; and
(b) that it is appropriate, for the protection of a particular person or persons or the community generally, that the person be held in custody. [138]
"[S]erious act of violence" is defined in s 4 of the Act to mean "an act of violence, committed by one person against another, that has a real likelihood of causing death or serious injury to the other person or that involves sexual assault in the nature of an offence referred to in section 61I, 61J, 61K, 66A, 66B, 66C, 66D, 66F, 78H, 78I, 78K, 78L or 80A of the Crimes Act 1900 ". 1. Section 5(2) provides that "[t]he maximum period to be specified is 6 months".
The power conferred by s 5(1) is conferred in general terms. However, s 3(3) provides that the "Act authorises the making of a detention order against Gregory Wayne Kable and does not authorise the making of a detention order against any other person". Gregory Wayne Kable is identified in s 3(4) as "the person of that name who was convicted in New South Wales on 1 August 1990 of the manslaughter of his wife, Hilary Kable".
It is necessary to refer to certain other provisions of the Act. First, s 3(2) provides that "[i]n the construction of [the] Act, the need to protect the community is to be given paramount consideration". Clearly, the effect of that directive is to require the making of an order under s 5(1) if the conditions specified in s 5(1)(a) and (b) are satisfied.
The second provision which should be noted is s 14 which provides that "[p]roceedings under [the] Act are civil proceedings and, to the extent to which [the] Act does not provide for their conduct, they are to be conducted in accordance with the law (including the rules of evidence) relating to civil proceedings". However, there is no law relating to proceedings of the kind contemplated by s 5(1) of the Act, other than the law which is made applicable by the Act. And although s 17(1)(a) provides that the rules of evidence are to apply, succeeding paragraphs of s 17(1) and subs (3) of that section, make it clear that, in significant respects, the rules of evidence do not apply. In particular, s 17(1) also provides that the court:
(b) may order the production of documents of the following kind in relation to the defendant:
(ii) records and reports of any psychiatric in-patient service or prison;
(iii) reports made to, or by, the Offenders Review Board;
(iv) reports, records or other documents prepared or kept by any police officer;
(v) the transcript of any proceedings before, and any evidence tendered to, the Mental Health Review Tribunal; and
(c) may order an examination of the defendant to be carried out by one or more duly qualified medical practitioners, psychiatrists or psychologists; and
(d) may require the preparation of reports as to the defendant's condition and progress by such persons as it considers appropriate; and
(e) must have regard to any report made available to it under paragraph (d);
And s 17(3) provides: Despite any Act or law to the contrary, the Court must receive in evidence any document or report of a kind referred to in subsection (1), or any copy of any such document or report, that is tendered to it in proceedings under this Act.
It is to be noted that, although s 17 speaks of the appellant as "the defendant", he is not charged with and not to be tried for an offence against the criminal law: he is, however, to be detained, if the conditions of s 5(1) are satisfied; and succeeding provisions of the Act make it plain that he is to be detained in prison and subject to substantially the same regime as persons convicted of criminal offences (see s 22). Nor is the appellant a party to civil proceedings involving contested legal rights and obligations, notwithstanding that s 14 declares that "[p]roceedings under [the] Act are civil proceedings".
The proceedings which the Act contemplates are not proceedings otherwise known to the law. And except to the extent that the Act attempts to dress them up as legal proceedings (for example, by referring to the applicant as "the defendant", by specifying that the proceedings are civil proceedings and by suggesting that the rules of evidence apply) [139] , they do not in any way partake of the nature of legal proceedings. They do not involve the resolution of a dispute between contesting parties as to their respective legal rights and obligations. And as already indicated, the applicant is not to be put on trial for any offence against the criminal law. Instead, the proceedings are directed to the making of a guess - perhaps an educated guess, but a guess nonetheless - whether, on the balance of probabilities, the appellant will commit an offence of the kind specified in the definition of "serious act of violence". And, at least in some circumstances [140] , the Act directs that that guess be made having regard to material which would not be admissible as evidence in legal proceedings.
It is well settled that some functions take their character from the way in which they are to be exercised and, thus, from the body on which they are conferred [141] . Accordingly, some functions which are not essentially judicial in character are, nonetheless, properly characterised as judicial if conferred on a court [142] . Chapter III permits the conferral of such functions on courts. And, of course, there is nothing to prevent the Parliaments of the States from conferring powers on their courts which are wholly non-judicial, so long as they are not repugnant to or inconsistent with the exercise by those courts of the judicial power of the Commonwealth.
The power purportedly conferred by s 5(1) of the Act requires the making of an order, if the conditions specified in s 5(1) are satisfied, depriving an individual of his liberty, not because he has breached any law, whether civil or criminal, but because an opinion is formed, on the basis of material which does not necessarily constitute evidence admissible in legal proceedings, that he "is more likely than not" (s 5(1)(a)) to breach a law by committing a serious act of violence as defined in s 4 of the Act. That is the antithesis of the judicial process, one of the central purposes of which is, as I said in Re Nolan; Ex parte Young [143] , to protect "the individual from arbitrary punishment and the arbitrary abrogation of rights by ensuring that punishment is not inflicted and rights are not interfered with other than in consequence of the fair and impartial application of the relevant law to facts which have been properly ascertained". It is not a power that is properly characterised as a judicial function, notwithstanding that it is purportedly conferred on a court and its exercise is conditioned in terms usually associated with the judicial process.
Moreover, when regard is had to the precise nature of the function purportedly conferred by s 5(1), the matters to be taken into account in its exercise and its contrariety to what is ordinarily involved in the judicial process, the effect of s 5(1) is, in my view, to compromise the integrity of the Supreme Court of New South Wales and, because that court is not simply a State court but a court which also exists to exercise the judicial power of the Commonwealth, it also has the effect of compromising the integrity of the judicial system brought into existence by Ch III of the Constitution.
The integrity of the courts depends on their acting in accordance with the judicial process and, in no small measure, on the maintenance of public confidence in that process [144] . Particularly is that so in relation to criminal proceedings which involve the most important of all judicial functions, namely, the determination of the guilt or innocence of persons accused of criminal offences. Public confidence cannot be maintained in the courts and their criminal processes if, as postulated by s 5(1), the courts are required to deprive persons of their liberty, not on the basis that they have breached any law, but on the basis that an opinion is formed, by reference to material which may or may not be admissible in legal proceedings, that on the balance of probabilities, they may do so.
Mention should be made of one other aspect of the function purportedly conferred on the Supreme Court by s 5(1) of the Act. Public confidence in the courts requires that they act consistently and that their proceedings be conducted according to rules of general application. That is an essential feature of the judicial process. It is that feature which serves to distinguish between palm tree justice and equal justice. Public confidence cannot be maintained in a judicial system which is not predicated on equal justice [145] .
The Act, in several of its provisions, suggests that an application under s 5(1) is to be determined in accordance with rules generally applicable in legal proceedings. In this respect, I have already referred to the description of the appellant as "the defendant", the description of the proceedings as "civil proceedings" and the suggestion that the rules of evidence apply when, in significant respects, they do not. Mention has also been made of s 16 which provides for proceedings under s 5(1) to be "commenced by summons in accordance with rules of court". In truth, the proceedings contemplated by s 5(1) are unique with unique procedures and with rules which apply only to the appellant. They are proceedings which the Act attempts to dress up as proceedings involving the judicial process. In so doing, the Act makes a mockery of that process and, inevitably, weakens public confidence in it. And because the judicial process is a defining feature of the judicial power of the Commonwealth [146] , the Act weakens confidence in the institutions which comprise the judicial system brought into existence by Ch III of the Constitution.
Section 5(1) of the Act is invalid. So too are the remaining provisions of the Act which serve no purpose other than to carry s 5(1) into effect.
The appeal should be allowed with costs. The order of the Court of Appeal should be set aside and in lieu thereof it should be ordered that the appeal to that court be allowed with costs, the orders of Levine J be set aside and, in lieu of those orders, the application of the Director of Public Prosecutions be dismissed with costs.
The question in this appeal from an order of the New South Wales Court of Appeal is whether the Community Protection Act 1994 NSW (the Act) is a valid law of the Parliament of New South Wales. In my opinion, the Act is invalid.
The Act, which was assented to shortly before the appellant, Gregory Wayne Kable [147] , was to be released from prison, makes provision for his imprisonment, if, upon the balance of probabilities, the Supreme Court of New South Wales is satisfied on reasonable grounds that he is "more likely than not to commit a serious act of violence; and that it is appropriate, for the protection of a particular person or persons or the community generally, that [he] be held in custody" (s 5(1)). The maximum period of imprisonment under any one order of the Supreme Court is six months (s 5(2)). But more than one application for an order may be made in relation to the same person (s 5(4)). So it is theoretically possible that under the Act the appellant could be imprisoned for life - not because of any crime that he has committed but because on the balance of probabilities the Supreme Court is satisfied that the appellant is more likely than not to commit a serious act of violence.
Upon an application made by the respondent, the Director of Public Prosecutions for New South Wales, the Supreme Court (Levine J) ordered that the appellant "be detained in custody for a period of six months commencing 23 February 1995 and expiring 22 August 1995 pursuant to s 9 of the Community Protection Act 1994 ". By its order the Supreme Court directed that the appellant "be detained at the Special Care Centre of the Long Bay Correctional Prison". On 9 May 1995 the Court of Appeal dismissed an appeal by the appellant against the order of the Supreme Court. Subsequently, the Supreme Court (Grove J) refused to make another order in relation to the appellant. For the time being, therefore, the appellant is free from the restraints of the Act. Nevertheless, the appellant remains at risk of the Supreme Court making another order against him.
The appellant challenges the validity of the Act on a number of grounds. For the reasons given by Brennan CJ and Dawson J, I agree that the challenge fails in so far as the appellant contends that the Act is not a law within the meaning of the Constitution Act 1902 NSW (the New South Wales Constitution) or that it is invalid because the New South Wales Constitution contains an entrenched doctrine of the separation of powers which the Act breaches. In my opinion, however, the Act is invalid because it purports to vest functions in the Supreme Court of New South Wales that are incompatible with the exercise of the judicial power of the Commonwealth by the Supreme Court of that State.
The States do not have unlimited power in respect of State courts
Subject to the operation of the Commonwealth of Australia Constitution Act 1900 Imp (the Constitution), the State of New South Wales is governed by the New South Wales Constitution. The latter Act is not predicated on any separation of legislative, executive and judicial power [148] although no doubt it assumes that the legislative, executive and judicial power of the State will be exercised by institutions that are functionally separated. Despite that assumption, I can see nothing in the New South Wales Constitution nor the constitutional history of the State that would preclude the State legislature from vesting legislative or executive power in the New South Wales judiciary [149] or judicial power in the legislature or the executive. Nor is the federal doctrine of the separation of powers - one of the fundamental doctrines of the Constitution - directly applicable to the State of New South Wales. Federal judicial power may be vested in a State court although that court exercises non-judicial as well as judicial functions. Moreover, when the Parliament of the Commonwealth invests the judicial power of the Commonwealth in State courts pursuant to s 77(iii) of the Constitution, it must take the State court as it finds it [150] . This is because the Constitution recognises that the jurisdiction, structure and organisation of State courts and the appointment, tenure and terms of remuneration of judges of State courts is not a matter within the legislative power of the federal Parliament. But in my opinion none of the foregoing considerations means that the Constitution contains no implications concerning the powers of State legislatures to abolish or regulate State courts, to invest State courts or State judges with non-judicial powers or functions, or to regulate the exercise of judicial power by State courts and judges.
The working of the Constitution requires and implies the continued existence of a system of State courts with a Supreme Court at the head of the State judicial system. Covering cl 5 of the Constitution, for example, in declaring that the Constitution "and all laws made by the Parliament of the Commonwealth under the Constitution, shall be binding on the courts, judges, and people of every State" necessarily implies the continuing existence of a system of State courts declaring the legal rights and duties of the people of Australia. So does s 118 in declaring that "[f]ull faith and credit shall be given, throughout the Commonwealth to the judicial proceedings of every State". So too do s 51(xxiv) and s 51(xxv) in granting to the Parliament of the Commonwealth the power to make laws with respect to "[t]he service and execution throughout the Commonwealth of the civil and criminal process and the judgments of the courts of the States" and "[t]he recognition throughout the Commonwealth of the judicial proceedings of the States". Section 77 of the Constitution also necessarily implies the existence of a court system in each State [151] . It gives the Parliament of the Commonwealth the power to invest "any court of a State with federal jurisdiction" and to define "the extent to which the jurisdiction of any federal court shall be exclusive of that which belongs to or is invested in the courts of the States".
One of the reasons for enacting s 77(iii) was that it was "a very convenient means of avoiding the multiplicity and expense of legal tribunals" [152] . The paragraph provides the machinery for implementing that part of the declaration in s 71 of the Constitution which provides that "[t]he judicial power of the Commonwealth shall be vested in such other courts as [the Parliament] invests with federal jurisdiction". If a State could abolish its court system, the powers conferred by s 77(ii) and (iii) would be rendered useless and the constitutional plan of a system of State courts invested with federal jurisdiction, as envisaged by Ch III, would be defeated. This is because only courts can be invested with federal jurisdiction pursuant to the provisions of s 77(iii) [153] . It is hardly to be supposed that the Constitution intended that a State could defeat the exercise of the grants of power conferred on the Parliament of the Commonwealth by s 77 by the simple expedient of abolishing its courts and setting up a system of tribunals that were not courts.
Furthermore, s 73 of the Constitution implies the continued existence of the State Supreme Courts by giving a right of appeal from the Supreme Court of each State to the High Court, subject only to such exceptions as the Commonwealth Parliament enacts. Section 73(ii) gives this Court jurisdiction to determine appeals from the decisions of any "court exercising federal jurisdiction; or of the Supreme Court of any State, or any other court of any State from which at the establishment of the Commonwealth an appeal [lay] to the Queen in Council". The right of appeal from a State Supreme Court to this Court, conferred by that section, would be rendered nugatory if the Constitution permitted a State to abolish its Supreme Court.
It necessarily follows, therefore, that the Constitution has withdrawn from each State the power to abolish its Supreme Courts or to leave its people without the protection of a judicial system. That does not mean that a State cannot abolish or amend the constitutions of its existing courts. Leaving aside the special position of the Supreme Court of the States, the States can abolish or amend the structure of existing courts and create new ones. However, the Constitution requires a judicial system in and a Supreme Court for each State and, if there is a system of State courts in addition to the Supreme Court, the Supreme Court must be at the apex of the system. With the abolition of the right of appeal to the Privy Council, therefore, this Court is now the apex of an Australian judicial system.
State courts are part of an Australian judicial system
At federation each Colony had courts. Each Colony had a Supreme Court from which an appeal could be taken to the Privy Council [154] . The right of appeal from the State Supreme Courts to the Privy Council continued after federation. In addition, s 74 of the Constitution preserved the prerogative right of Her Majesty in Council to grant leave to appeal from decisions of the High Court subject to the obtaining of a certificate from this Court in respect of matters concerning "the limits inter se of the constitutional powers of the Commonwealth and those of any State or States, or as to the limits inter se of the Constitutional powers of any two or more States". However, s 74 also gave the Parliament power to "make laws limiting the matters" in which special leave to appeal from the High Court to the Privy Council could be asked. That power extended to abolishing all matters in respect of which leave could be sought [155] . Nevertheless, until that power was exercised, the Constitution intended that, subject to the grant of a certificate by the High Court in respect of an inter se matter, Australia should have an integrated system of State and federal courts administering a single body of common law under the supervision of the Judicial Committee of the Privy Council which stood at the apex of the system.
Unlike the United States of America where there is a common law of each State, Australia has a unified common law which applies in each State but is not itself the creature of any State [156] . Perhaps the validity of that proposition is not as readily apparent to a State judge bound by the authority of his or her own Full Court or Court of Appeal as it is to a judge of a federal court who must apply the common law. In an extra-judicial paper published in 1957, Sir Owen Dixon pointed out that, if there is no statutory law in the case, an Australian judge sitting in the original jurisdiction of the High Court "proceeds to administer the common law as an entire system. He ascertains its content as best he may. Among the judicial decisions to which he may turn those of the State whose law he finds that he must apply will have no higher authority than the decisions of any other State and the authority of the decisions will be persuasive only and not imperative." [157] In an address to the Section of the American Bar Association for International and Comparative Law, made fourteen years before the publication of the paper containing that statement, his Honour had said [158] :
We therefore regard Australian law as a unit. Its content comprises besides legislation the general common law which it is the duty of the courts to ascertain as best they may. But subject always to the binding authority of some disturbing precedent, we treat it as the duty of all courts to recognise that it is one system which should receive a uniform interpretation and application, not only throughout Australia but in every jurisdiction of the British Commonwealth where the common law runs. Later his Honour referred to [159] : the reasons which make it possible for an Australian to regard his country as governed by a single legal system. It is a system or corpus composed of the common law, modified by the enactments of various legislatures. 1. Mabo v Queensland [No 2] (1992) 175 CLR 1 at 15 Environment Protection Authority v Caltex Refining Co Pty Ltd (1993) 178 CLR 477 at 556 2. Dixon, The Common Law as an Ultimate Constitutional Foundation, Jesting Pilate (1965) 203, at p 204. 3. Dixon, Sources of Legal Authority, Jesting Pilate (1965) 198, at p 199. 4. Dixon, Sources of Legal Authority, Jesting Pilate (1965) 198, at p 201.
The legal system adopted by the Constitution continued until the passing of the Privy Council (Limitation of Appeals) Act 1968 Cth. Upon the passing of the Privy Council (Appeals from the High Court) Act 1975 Cth, an appeal could no longer be taken from the High Court to the Privy Council. That meant that until the enactment of s 11 of the Australia Acts 1986, appeals could still be taken to the Privy Council from the State Supreme Courts. From 1975 until 1986, therefore, the High Court and the Privy Council shared the function of declaring the law of Australia. Moreover, as the Privy Council made plain in Australian Consolidated Press Ltd v Uren [160] , the common law of Australia was not necessarily the common law of England or the British Commonwealth. But that there is a common law of Australia as opposed to a common law of individual States is clear. In so far as the remarks of Kitto J in Anderson v Eric Anderson Radio & TV Pty Ltd [161] suggest a contrary view, they cannot be regarded as representing the law [162] . Since 1986, this Court has been the ultimate appellate court of the nation. The right of appeal to the Privy Council having been abolished, the High Court of Australia has the constitutional duty of supervising the nation's legal system and, subject to any relevant statutory or constitutional limitations, of maintaining a unified system of common law.
An essential part of the machinery for implementing that supervision of the Australian legal system and maintaining the unity of the common law is the system of State courts under a Supreme Court with an appeal to the High Court under s 73 of the Constitution. The judgment of the High Court in such an appeal is "final and conclusive" [163] . Without the continued existence of a right of appeal from the Supreme Court of each State to the High Court, it would be difficult, indeed probably impossible, to have the unified system of common law that the Constitution intended should govern the people of Australia. Moreover, although it is not necessary to decide the point in the present case, a State law that prevented a right of appeal to the Supreme Court from, or a review of, a decision of an inferior State court, however described, would seem inconsistent with the principle expressed in s 73 and the integrated system of State and federal courts that covering cl 5 and Ch III envisages.
It follows that State courts exercising State judicial power cannot be regarded as institutions that are independent of the administration of the law by this Court or the federal courts created by the Parliament of the Commonwealth. In exercising federal jurisdiction, a court of a State administers the same law as the Federal Court of Australia when it exercises the identical federal jurisdiction. In exercising federal jurisdiction, a State court must deduce any relevant common law principle from the decisions of all the courts of the nation and not merely from the decisions of the higher courts of its State. A judge exercising the federal jurisdiction invested in a State court must see the common law in exactly the same way that a judge of a federal court created under s 71 of the Constitution sees it.
Furthermore, a State court when it exercises federal jurisdiction invested under s 77(iii) is not a court different from the court that exercises the judicial power of the State. The judges of a State court who exercise the judicial power of the State are the same judges who exercise the judicial power of the Commonwealth invested in their courts pursuant to s 77(iii) of the Constitution. Indeed, it is not uncommon for a judge of a State court to administer State legislation in the course of the exercise of federal jurisdiction. It is common ground, for example, that in this very case Levine J made his order in the exercise of federal jurisdiction because he became seized of federal jurisdiction when the appellant contended that the Act was in breach of the Constitution.
Under the Constitution, therefore, the State courts have a status and a role that extends beyond their status and role as part of the State judicial systems. They are part of an integrated system of State and federal courts and organs for the exercise of federal judicial power as well as State judicial power. Moreover, the Constitution contemplates no distinction between the status of State courts invested with federal jurisdiction and those created as federal courts. There are not two grades of federal judicial power. The terms of s 71 of the Constitution equate the vesting of judicial power in the federal courts with the vesting of federal judicial power in the State courts. And s 77(iii) permits the Parliament of the Commonwealth to invest any court of a State with federal jurisdiction in respect of all matters that are or can be vested in the original jurisdiction of this Court or the federal courts. Other sections (ss 78, 79, 80) of the Constitution also draw no distinction between the exercise of federal judicial power by the State courts and its exercise by federal courts.
It is true that the Constitution does not protect the appointment, remuneration and tenure of the judges of State courts invested with federal jurisdiction although it protects the judges of federal courts in respect of those matters. But this difference provides no ground for concluding that the exercise of federal judicial power by State courts was intended to be inferior to the exercise of that power by federal courts. As a former Solicitor-General of the Commonwealth has written [164] :
The intention was clearly to place a State court exercising federal jurisdiction on the same general footing as the federal courts which should be created by the Parliament. Their jurisdiction was in both cases to be fixed by the Parliament; their decisions were alike to be subject to appeal to the High Court. 1. Bailey, The Federal Jurisdiction of State Courts, Res Judicatae, vol 2 (1940) 109, at p 109.
Legislatures cannot alter or undermine the constitutional scheme set up by Ch III
It is axiomatic that neither the Commonwealth nor a State can legislate in a way that might alter or undermine the constitutional scheme set up by Ch III of the Constitution. The Parliament of the Commonwealth, for example, has no power under s 77(iii) of the Constitution to invest State courts with non-judicial functions except as an incident in the grant of judicial power [165] . Similarly, a State cannot legislate for issues arising under Ch III to be referred to the Judicial Committee of the Privy Council or some other body for determination or advice in a manner that conflicts with the principles of Ch III [166] . In The Commonwealth v Queensland , in a judgment with which Barwick CJ, Stephen and Mason JJ agreed, Gibbs J held that it is implicit in Ch III that a State cannot legislate in a way that has the effect of violating "the principles that underlie Ch III" [167] .
Because the State courts are an integral and equal part of the judicial system set up by Ch III, it also follows that no State or federal parliament can legislate in a way that might undermine the role of those courts as repositories of federal judicial power. Thus, neither the Parliament of New South Wales nor the Parliament of the Commonwealth can invest functions in the Supreme Court of New South Wales that are incompatible with the exercise of federal judicial power. Neither Parliament, for example, can legislate in a way that permits the Supreme Court while exercising federal judicial power to disregard the rules of natural justice or to exercise legislative or executive power. Such legislation is inconsistent with the exercise of federal judicial power. However, the Act does not seek to interfere with the invested federal jurisdiction of the Supreme Court. On its face it is directed to the exercise of State, not federal, jurisdiction. But for present purposes that is irrelevant. The compatibility of State legislation with federal judicial power does not depend on intention. It depends on effect. If, as Gibbs J pointed out in The Commonwealth v Queensland [168] , State legislation has the effect of violating the principles that underlie Ch III, it will be invalid.
Courts exercising federal jurisdiction must be perceived to be free from legislative or executive interference
One of the basic principles which underlie Ch III and to which it gives effect is that the judges of the federal courts must be, and must be perceived to be, independent of the legislature and the executive government [169] . Given the central role and the status that Ch III gives to State courts invested with federal jurisdiction, it necessarily follows that those courts must also be, and be perceived to be, independent of the legislature and executive government in the exercise of federal jurisdiction. Public confidence in the impartial exercise of federal judicial power would soon be lost if federal or State courts exercising federal jurisdiction were not, or were not perceived to be, independent of the legislature or the executive government.
In the case of State courts, this means they must be independent and appear to be independent of their own State's legislature and executive government as well as the federal legislature and government. Cases concerning the States, the extent of the legislative powers of the States and the actions of the executive governments of the States frequently attract the exercise of invested federal jurisdiction. The Commonwealth government and the residents and governments of other States are among those who litigate issues in the courts of a State. Quite often the government of the State concerned is the opposing party in actions brought by these litigants. Public confidence in the exercise of federal jurisdiction by the courts of a State could not be retained if litigants in those courts believed that the judges of those courts were sympathetic to the interests of their State or its executive government.
While nothing in Ch III prevents a State from conferring non-judicial functions on a State Supreme Court in respect of non-federal matters, those non-judicial functions cannot be of a nature that might lead an ordinary reasonable member of the public to conclude that the Court was not independent of the executive government of the State. A State law which gave the Supreme Court powers to determine issues of a purely governmental nature - for example, how much of the State Budget should be spent on child welfare or what policies should be pursued by a particular government department - would be invalid. It would have the effect of so closely identifying the Supreme Court with the government of the State that it would give the appearance that the Supreme Court was part of the executive government of the State. The law would fail not because it breached any entrenched doctrine of separation powers in the State Constitution [170] but because it gave the appearance that a court invested with federal jurisdiction was not independent of its State government.
In addition, in the case of the Supreme Court, although non-judicial functions may be vested in that court, they cannot be so extensive or of such a nature that the Supreme Court would lose its identity as a court. Thus, a State can invest its Supreme Court with a jurisdiction similar to that which is presently exercised in the federal sphere by the Administrative Appeals Tribunal. The Supreme Court would not lose its identity as the Supreme Court of the State merely because it was given a jurisdiction similar to that of that Tribunal. Nor could such a jurisdiction lead any reasonable person to conclude that the Supreme Court was part of the executive government of the State. However, a State could not legislate to abolish all other jurisdictions of the Supreme Court and invest it with no more than a jurisdiction similar to that Tribunal. To do so would make a mockery of the principles contained in Ch III of the Constitution.
Furthermore, although nothing in Ch III prevents a State from conferring executive government functions on a State court judge as persona designata, if the appointment of a judge as persona designata gave the appearance that the court as an institution was not independent of the executive government of the State, it would be invalid. No doubt there are few appointments of a judge as persona designata in the State sphere that could give rise to the conclusion that the court of which the judge was a member was not independent of the executive government. Many Chief Justices, for example, act as Lieutenant-Governors and Acting Governors. But, given the long history of such appointments, it is impossible to conclude that such appointments compromise the independence of the Supreme Courts or suggest that they are not impartial. Similarly, a law that provided for a judge of a State court to be appointed as a member of an Electoral Commission fixing the electoral boundaries of the State would not appear to suggest that the court was not impartial. However, a State law which purported to appoint the Chief Justice of the Supreme Court to be a member of the Cabinet might well be invalid [171] because the appointment would undermine confidence in the impartiality of the Supreme Court as an institution independent of the executive government of the State.
It follows therefore that, although New South Wales has no entrenched doctrine of the separation of powers and although the Commonwealth doctrine of separation of powers cannot apply to the State, in some situations the effect of Ch III of the Constitution may lead to the same result as if the State had an enforceable doctrine of separation of powers. This is because it is a necessary implication of the Constitution's plan of an Australian judicial system with State courts invested with federal jurisdiction that no government can act in a way that might undermine public confidence in the impartial administration of the judicial functions of State courts. If it could, it would inevitably result in a lack of public confidence in the administration of invested federal jurisdiction in those courts. State governments therefore do not have unrestricted power to legislate for State courts or judges. A State may invest a State court with non-judicial functions and its judges with duties that, in the federal sphere, would be incompatible with the holding of judicial office. But under the Constitution the boundary of State legislative power is crossed when the vesting of those functions or duties might lead ordinary reasonable members of the public to conclude that the State court as an institution was not free of government influence in administering the judicial functions invested in the court.
The Act has the tendency to undermine public confidence in the impartiality of the Supreme Court of New South Wales
The object of the Act according to s 3(1) "is to protect the community by providing for the preventive detention (by order of the Supreme Court made on the application of the Director of Public Prosecutions) of Gregory Wayne Kable". In construing the provisions of the Act, "the need to protect the community is to be given paramount consideration" (s 3(2)). The means by which the object of the Act is accomplished is to empower (s 5(1)) the Supreme Court to order the appellant's [172] detention in prison for a specified period not exceeding six months [173] . The only condition that must be fulfilled for the making of the order is that the Court be satisfied on reasonable grounds that the appellant "is more likely than not to commit a serious act of violence; and that it is appropriate, for the protection of a particular person or persons or the community generally, that the defendant be held in custody" (s 5(1)). Proceedings are to be commenced by summons (s 16(1)). The Court is authorised to hear and determine an application for an order "in the absence of the defendant" if it is satisfied that the summons has been served on the defendant or that all reasonable steps have been taken to serve the summons on him (s 16(2)). Somewhat inconsistently, s 17(2)(a) provides that the "Act does not affect the right of any party to proceedings under [the] Act to appear, either personally or by the party's legal representative".
Section 7 of the Act goes even further than s 5 in interfering with the liberty of the appellant. It provides for the making of an "interim detention order". Section 7(1) enacts that "the Court may order that the defendant in any proceedings on an application for a preventive detention order be detained in prison for such period (not exceeding 3 months) as the Court determines". Section 7(3) provides that, "[o]n an application or on its own motion, the Court may extend the period of an interim detention order for such further period (not exceeding 3 months) as the Court determines if it appears that the proceedings on the application for a preventive detention order will not be determined during the period currently specified in the interim detention order". Section 7(5) provides that "[a]n interim detention order may be made, and its period extended, in the absence of the defendant". The Act gives no right of appeal against an interim detention order [174] .
This brief summary of the central provisions of the Act shows that its object is to detain the appellant not for what he has done but for what the executive government of the State and its Parliament fear that he might do. Moreover, given the date of the Act and the date of the appellant's release from prison for the crime for which he had been convicted, it seems clear that the Parliament expected that he would be detained on an interim basis before the Supreme Court had had the opportunity to decide whether it was satisfied that the appellant was more likely than not to commit a serious act of violence [175] . Indeed, the Act contemplates that the interim order might be made or extended "in the absence of the defendant" (s 7(5)).
Furthermore, although the Act provides for the preventive detention of the appellant only upon the making of an order by the Supreme Court, it is plain that the legislature and the executive government which introduced the Act into the Parliament of New South Wales passed the Act for the purpose of ensuring that the appellant was kept in prison. Proof of the case against the appellant was made much easier by removing any need to prove beyond reasonable doubt that he was likely to commit a serious act of violence [176] . Although s 17(1)(a) provides that the Court "is bound by the rules of evidence", s 17(3) goes a long way to negating that protection. That subsection enacts:
Despite any Act or law to the contrary, the Court must receive in evidence any document or report of a kind referred to in subsection (1), or any copy of any such document or report, that is tendered to it in proceedings under this Act. Among the documents referred to in subsection (1) are:
(ii) records and reports of any psychiatric in-patient service or prison;
(iii) reports made to, or by, the Offenders Review Board;
(iv) reports, records or other documents prepared or kept by any police officer;
(v) the transcript of any proceedings before, and any evidence tendered to, the Mental Health Review Tribunal.
In my opinion, those who initiated and passed the Act plainly expected and intended that the imprisonment of the appellant would continue after the expiration of his sentence for the manslaughter of his wife. The object of the Act, its ad hominem nature and the grounds and method of proof of the s 5 order together with the provision for s 7 interim orders leave no other conclusion open. Why else would the executive government have introduced legislation into the Parliament which is directed only to the appellant [177] and which expressly states that its object is to protect the community by providing for the detention of the appellant unless the government intended that he should be kept in prison? It is true that the Act places the necessity for a Supreme Court order between the obvious intention of the executive government and the imprisonment of the appellant. But when the Act was passed it must have seemed to many that the risk of that intention being defeated was minimal. The matters that signal the intention of the makers of the Act also undermined the ordinary safeguards of the judicial process and made it highly likely that the appellant would be imprisoned.
The Parliament of New South Wales has the constitutional power to pass legislation providing for the imprisonment of a particular individual. And that is so whether the machinery for the imprisonment be the legislation itself or the order of a Minister, public servant or tribunal. Moreover, there is no reason to doubt the authority of the State to make general laws for preventive detention when those laws operate in accordance with the ordinary judicial processes of the State courts. However, whatever else the Parliament of New South Wales may be able to do in respect of the preventive detention of individuals who are perceived to be dangerous, it cannot, consistently with Ch III of the Constitution, invoke the authority of the Supreme Court to make the orders against the appellant by the methods which the Act authorises. This is because the Act and its procedures compromise the institutional impartiality of the Supreme Court.
Apart from similar legislation passed by the Parliament of Victoria providing for the protective sentencing or the preventive detention of Garry Ian David in 1990 [178] , no Parliament in the Commonwealth of Australia has ever given a court a jurisdiction that is remotely similar to that which the Act gives to the Supreme Court of New South Wales. It is not merely that the Act involves the Supreme Court in the exercise of non-judicial functions or that it provides for punishment by way of imprisonment for what the appellant is likely to do as opposed to what he has done. The Act seeks to ensure, so far as legislation can do it, that the appellant will be imprisoned by the Supreme Court when his sentence for manslaughter expires. It makes the Supreme Court the instrument of a legislative plan, initiated by the executive government, to imprison the appellant by a process that is far removed from the judicial process that is ordinarily invoked when a court is asked to imprison a person.
The Act expressly removes the ordinary protections inherent in the judicial process. It does so by stating that its object is the preventive detention of the appellant, by removing the need to prove guilt beyond reasonable doubt, by providing for proof by materials that may not satisfy the rules of evidence and by declaring the proceedings to be civil proceedings although the Court is not asked to determine the existing rights and liabilities of any party or parties. It is not going too far to say that proceedings under the Act bear very little resemblance to the ordinary processes and proceedings of the Supreme Court. They do not involve any contest as to whether the appellant has breached any law or any legal obligation. They "are not directed to any determination or order which resolves an actual or potential controversy as to existing rights or obligations" [179] which is the benchmark of an exercise of judicial power.
The Act is thus far removed from the ordinary incidents of the judicial process. It invests the Supreme Court with a jurisdiction that is purely executive in nature. Indeed, the jurisdiction conferred on the Court is hardly distinguishable from those powers and functions, concerning the liberty of the subject, that the traditions of the common law countries have placed in Ministers of the Crown so that they can be answerable to Parliament for their decisions. The jurisdiction conferred on the Supreme Court bears a remarkable similarity to the jurisdiction conferred on the Minister during wartime under reg 26 of the National Security (General) Regulations 1939 Cth. That regulation empowered the Minister to order the detention of a person if he was satisfied that it was necessary to do so "to prevent that person acting in any manner prejudicial to the public safety or the defence of the Commonwealth".
Instead of a trial where the Crown is required to prove beyond reasonable doubt that the accused is guilty of a crime on evidence admitted in accordance with the rules of evidence, the Supreme Court is asked to speculate whether, on the balance of probabilities, it is more likely than not the appellant will commit a serious act of violence. As Professor Williams has pointed out [180] : "Predicting dangerousness is, of course, notoriously difficult." Yet on this prediction of dangerousness, a prediction which can at best be but an informed guess by the Supreme Court, the Court is required to commit the appellant to prison. Having regard to the object of the Act, it is impossible to suppose that the Court has any discretion to refuse to imprison the appellant once it concludes that he is more likely than not to commit a serious act of violence.
Two comments made by Dr David Wood concerning the Victorian legislation providing for the detention of Garry Ian David are equally applicable to the Act. The first was [181] :
After all, it is not envisaged that an application for an Order to detain David would be refused. The Government's approach to the problem indicates that if it were in any doubt about this, it would have given the Minister the power.
As it is, the Government gets the best of both worlds. It escapes the charge that the Minister is acting as a judge in detaining David by executive decree; and it is able to pass ultimate responsibility for the continued detention of David onto the courts.
The legislation puts the Supreme Court in an invidious position. It would take a brave Supreme Court judge to find that the case for placing David in preventive detention had not been made out. If he so found, he would be frustrating the sole purpose of the Act, which is, after all, to enable this particular person to be detained. If, on the other hand, as is expected, he issues the order, it will be difficult for him to escape appearing that he is simply doing the Government's bidding. 1. Wood, A One Man Dangerous Offenders Statute - The Community Protection Act 1990 (Vic), Melbourne University Law Review, vol 17 (1990) 497, at p 502.
No one who has read the lengthy and anxious judgment of Levine J making the order imprisoning the appellant or the judgments of the judges of the Court of Appeal upholding that order or the judgment of Grove J refusing to make a further order against the appellant could doubt their independence and impartiality in administering the law. The judgments of Levine J and the Court of Appeal demonstrate that the order against the appellant was made and upheld only because the object of the Act, the evidence and the methods and burden of proof left them no alternative to making and upholding the s 5 order. But the constitutional validity of the Act cannot depend on how the judges of the Supreme Court discharge the duty that the Act imposes upon them. The Act was either valid or invalid when it was given the Royal Assent. Nothing that the judges of the Supreme Court did after its enactment could change its status as a valid or invalid piece of legislation.
At the time of its enactment, ordinary reasonable members of the public might reasonably have seen the Act as making the Supreme Court a party to and responsible for implementing the political decision of the executive government that the appellant should be imprisoned without the benefit of the ordinary processes of law. Any person who reached that conclusion could justifiably draw the inference that the Supreme Court was an instrument of executive government policy. That being so, public confidence in the impartial administration of the judicial functions of the Supreme Court must inevitably be impaired. The Act therefore infringed Ch III of the Constitution and was and is invalid.
On 23 February 1995 in the Common Law Division of the Supreme Court of New South Wales, and on the application of the present respondent, the Director of Public Prosecutions for New South Wales (the Director), orders were made in respect of the present appellant. The first three paragraphs of the order were as follows:
1 Gregory Wayne Kable be detained in custody for a period of six months commencing 23 February 1995 and expiring 22 August 1995 pursuant to s 9 of the Community Protection Act 1994.
2 in the first instance the defendant to be taken back to the place from whence he came, namely the Remand Centre and be provided with the opportunity or necessary facility to remove any personal papers or effects.
3 the defendant be detained at the Special Care Centre of the Long Bay Correctional Prison.
Section 22(1) of the Community Protection Act 1994 NSW (the Act) states:
A detainee is taken to be a prisoner within the meaning of the Prisons Act 1952. The Prisons Act 1952 NSW provides in Pt 2 (ss 5-11A) for the establishment and control of prisons. Part 3 (ss 12-22F) provides for the treatment of prisoners and Pt 4 (ss 23-26I) deals with prison discipline.
The Director holds office under the Director of Public Prosecutions Act 1986 NSW (the DPP Act). Section 7(1) of that statute identifies one of the principal functions and responsibilities of the Director as the institution and conduct in the Supreme Court and in the District Court, on behalf of the Crown, of prosecutions (whether on indictment or summarily) for indictable offences.
The foregoing would suggest that Gregory Wayne Kable (the appellant) was to be incarcerated pursuant to orders of the Supreme Court consequent upon and by reason of conviction for an offence against the criminal law or, perhaps, as punishment for contempt of court (but bearing in mind that all proceedings for contempt realistically are to be seen as criminal in nature) [183] .
However, as will appear, the appellant had not been convicted of any such contravention of the law. The legal and practical effect of the orders of the Supreme Court was to penalise him by imprisonment for a period of six months, but not by reason of any contravention of the criminal law. Moreover, the statute which conferred the authority upon the Supreme Court to make the orders in question does not create and impose any legal norms of conduct upon the community at large or upon any class or classes of persons answering a particular description. The Act is directed to one individual, namely the appellant. That, of itself, is not a badge of invalidity, even where the legislature is one of distinct and enumerated powers [184] . However, here, if certain criteria are met, then the Supreme Court is to inflict upon that individual a penalty. Moreover, the penalty is not inflicted upon, and by reason of, conviction by the Supreme Court on any charge of contravention of the criminal law.
On 9 May 1995, that is to say during the currency of the period of detention, the New South Wales Court of Appeal (Mahoney, Clarke and Sheller JJA) dismissed an appeal [185] . The proceedings, both at first instance and in the Court of Appeal, differed from those which ordinarily would arise under the Act itself. This was because the appellant relied, unsuccessfully, upon the Constitution to resist the application by the Director.
There was no cross-application before the Supreme Court in which the appellant sought declaratory relief as to the invalidity of the Act. However, the determination of the Director's application, and of the appeals to the Court of Appeal and to this Court, necessarily turns upon the question of validity of the legislation.
On 18 August 1995 this Court granted the appellant special leave to appeal against the decision of the Court of Appeal. The period of detention imposed by the orders made by the Supreme Court since has expired. There is no current further order under the Act in force. Nevertheless, the appellant remains at risk of further application being made to the Supreme Court by the Director under the Act. For that reason, at least, the present appeal is not moot.
Before this Court the principal relief sought by the appellant is a declaration of invalidity of the Act. Several grounds are pressed for that relief, but only one need be determined because, in my view, it succeeds. This ground involves consideration of Ch III of the Constitution in relation to the Supreme Court of New South Wales.
The judicial power of the Commonwealth engages the Supreme Court at two stages or levels. The first is upon investment of the Supreme Court by the Parliament with federal jurisdiction, pursuant to s 77(iii) of the Constitution. In that regard, it may be noted that, for many years, charges under federal and State criminal law have been tried together in the same courts of the States. The second level is in the entrenched appellate jurisdiction of this Court under s 73(ii) in respect of judgments, decrees, orders and sentences of the Supreme Court.
The appellant submits that the jurisdiction and powers conferred by the Act upon the Supreme Court are of such an extreme nature and quality as to render them incompatible with the exercise by the same judicial institution of the State of the judicial power of the Commonwealth and that, as a result, the Act is invalid. This conclusion is reached by several steps.
The appellant further contends, with reference to what was said in the joint judgment of Dixon CJ, McTiernan, Fullagar and Kitto JJ in R v Kirby; Ex parte Boilermakers' Society of Australia [186] , that it is an intrinsic characteristic of the judicial power of the Commonwealth established by Ch III of the Constitution that it is distinct from and paramount over the judicial power of the States. The absence from s 118 of the Constitution of reference to judicial proceedings of the Commonwealth, to complement that to judicial proceedings of every State, implies the pervasiveness and supremacy of the judicial power of the Commonwealth established by Ch III. By force of covering cl 5, the Constitution is binding on the courts, judges and people of every State and every part of the Commonwealth. In Boilermakers their Honours referred to the structure of the Constitution and concluded [187] :
The powers of the federal judicature [within which they included the operation of the "autochthonous expedient"] must therefore be at once paramount and limited. The limitation referred to arises from the broad areas of non-federal jurisdiction. But, as will be discussed later in these reasons, it is in this Court, a federal court, that the whole of the Australian judicial system has its focus. 1. (1956) 94 CLR 254 at 267-268. 2. Boilermakers' Case (1956) 94 CLR 254 at 268.
The appellant points to the particular characteristics of the provision made by the Constitution for the federal judicial power, which were identified by Deane J in Re Tracey; Ex parte Ryan [188] . His Honour said:
The power to adjudge guilt of, or determine punishment for, breach of the law, the power to determine questions of excess of legislative or executive power and the power to decide controversies about existing rights and liabilities all fall within the concept of judicial power. The Executive Government cannot absorb or be amalgamated with the judicature by the conferral of non-ancillary executive functions upon the courts. Nor can the Executive itself exercise judicial power and act as prosecutor and judge to punish breach of law by executive fiat or decree. The guilt of the citizen of a criminal offence and the liability of the citizen under the law, either to a fellow citizen or to the State, can be conclusively determined only by a Ch III court acting as such, that is to say, acting judicially. For its part, the Parliament cannot legislate either to destroy the entrenched safeguards of Ch III or to itself assume the exercise of judicial power. 1. (1989) 166 CLR 518 at 580.
It is then said that the purported conferral by a law of the Commonwealth upon a court exercising federal jurisdiction, whether a State court or a federal court, of authority to punish and imprison of the nature and scope of that conferred in the Act by the New South Wales legislature upon the Supreme Court, would fail. It would be a law which conferred neither judicial power nor a permissible ancillary executive function. Nor, given its tendency to trench upon and undermine the safeguards of Ch III, could such functions validly be conferred by the Parliament upon judges of federal courts selected as designated individuals [189] . There would be too great an incongruity between the discharge of those functions and the exercise of the federal judicial power [190] .
The final steps in the appellant's submissions are as follows. First, the structure of the Australian Constitution, especially Ch III, does not permit of an Australian judiciary exercising the judicial power of the Commonwealth but divided into two grades, an inferior grade, namely the possessors of invested federal jurisdiction who are subject to the imposition and receipt of incompatible functions under State law, and a superior grade, comprising this Court and other federal courts which are not subject to the imposition and receipt of such functions whether pursuant to Commonwealth or State law. The second step is that the Constitution, and especially Ch III, assumes and requires, at least as regards the Supreme Courts of the States, an institutional integrity of the State court structure which may not be undermined by the reposition in them of authorities and powers of the nature of those in the Act.
My conclusion is that, in the broad, these submissions should be accepted.
It is appropriate now to have fuller regard to the provisions of the Act. Section 3 thereof states:
(1) The object of this Act is to protect the community by providing for the preventive detention (by order of the Supreme Court made on the application of the Director of Public Prosecutions) of Gregory Wayne Kable.
(2) In the construction of this Act, the need to protect the community is to be given paramount consideration.
(3) This Act authorises the making of a detention order against Gregory Wayne Kable and does not authorise the making of a detention order against any other person.
(4) For the purposes of this section, Gregory Wayne Kable is the person of that name who was convicted in New South Wales on 1 August 1990 of the manslaughter of his wife, Hilary Kable.
The Act provides (s 24) for the making by a single judge of the Supreme Court of New South Wales of what are styled "preventive detention orders" (s 5) and "interim detention orders" (s 7) and for the issue of arrest warrants (s 6). Only the Director (s 8) may make applications under s 5, 6 or 7. The term "detention order" is used in the Act to identify both a preventive detention order and an interim detention order (s 4). No detention order is to be made by the Court unless satisfied that the case of the Director has been proved on the balance of probabilities (s 15). An appeal to the Court of Appeal lies from any determination of the Supreme Court to make, or to refuse to make, a preventive detention order (s 25). However, no such right of appeal is conferred in respect of an interim detention order.
The appellant, whilst a detainee, "must not be discharged from prison, or allowed leave of absence from prison" otherwise than at the expiry of the current detention order or in accordance with an order made by the Court (s 23(2)).
A person who is the subject of a detention order is taken to be required by law to be in custody in prison for the purposes of s 352AA of the Crimes Act 1900 NSW. Section 22(2) of the Act so states. Section 352AA(2) indicates that "a prisoner unlawfully at large" is a person who is at large (otherwise than by reason of having escaped from lawful custody) at a time when that person is required by law to be in custody in prison. Any constable may, with or without warrant, apprehend any person whom, with reasonable cause, the constable suspects of being a prisoner unlawfully at large (s 352AA(1)).
The Sentencing Act 1989 NSW provides for the setting of minimum and additional terms of imprisonment (s 5) and for fixed terms (ss 6, 7), and for a parole system (ss 14-41). It does not apply to or in respect of a detention order made under the Act or the appellant as a "detainee" thereunder (s 22(4)).
Section 28 of the Bail Act 1978 NSW authorises the Supreme Court to grant bail in accordance with it to any person accused of any offence, whether or not that person has appeared before the Supreme Court in connection with the offence. However, the operation of the Bail Act is gainsaid by s 29 of the Act. This provides that, in proceedings under the Act, the Bail Act does not apply to or in respect of the appellant.
A further significant impact upon the civil liberties and civil status of the appellant is effected by s 22(3) of the Act. This states that in any other statute (other than the Sentencing Act) or any instrument thereunder, a reference to a sentence of imprisonment includes a reference to a detention order and a reference to a term of imprisonment includes a reference to the period for which a detention order is in force. The effect of this is to render applicable to the appellant other laws of New South Wales which impose civil disability by reference to the imposition of sentences of imprisonment or the serving of a term of imprisonment.
The central provision of the legislation is s 5 which provides for preventive detention orders. It states:
(1) On an application made in accordance with this Act, the Court may order that a specified person be detained in prison for a specified period if it is satisfied, on reasonable grounds:
(a) that the person is more likely than not to commit a serious act of violence; and
(b) that it is appropriate, for the protection of a particular person or persons or the community generally, that the person be held in custody.
(2) The maximum period to be specified in an order under this section is 6 months.
(3) An order under this section may be made against a person:
(a) whether or not the person is in lawful custody, as a detainee or otherwise; and
(b) whether or not there are grounds on which the person may be held in lawful custody otherwise than as a detainee.
(4) More than one application under this section may be made in relation to the same person.
I have already set out the text of s 3. The result of reading s 5 and s 3 is that the reference to "specified person" is to be read as one to the appellant alone.
It will be noted that s 5(1) speaks of the Court being satisfied, on reasonable grounds, of two matters, those set out in pars (a) and (b). If the Court is so satisfied then the Court may order detention in prison for a specified period. This is not to exceed six months (s 5(2)). Upon the Court being so satisfied as aforesaid then, in my view, s 5(1) requires the Court to make an order for detention in prison. That is to say, the particular context indicates that the word "may" is used in s 5(1) not only as an empowering term but also to indicate the circumstances in which the power is to be exercised [191] .
However, on application of the Director or the appellant as detainee, the Supreme Court may revoke a preventive detention order or amend it by reducing the period for which it is in force (s 13(1)). More than one such application may be made in relation to the same preventive detention order (s 13(3)). In determining applications under s 13, the Court must have regard to the most recent reports made to the Director under s 21 on the condition and progress of the detainee. It will be necessary to refer further to s 21 later in these reasons.
Paragraph (a) of s 5(1) speaks of the satisfaction, on reasonable grounds, of the Court that the appellant is more likely than not "to commit a serious act of violence". The phrase "serious act of violence" is defined in s 4. This definition is in terms which embrace an act of violence that involves a sexual assault in the nature of an offence referred to in various sections of the Crimes Act. The definition further includes an act of violence committed by one person against another "that has a real likelihood of causing death or serious injury to the other person". The relevant acts here included may involve the commission of an offence against the criminal law either of the State or of the Commonwealth. Examples of the latter would be the offences created by s 36A of the Crimes Act 1914 Cth and s 14 of the Crimes (Aviation) Act 1991 Cth. These are concerned respectively with the use of violence on witnesses in certain judicial proceedings and acts of violence against aircraft passengers or crew. Thus, involved in the determination by the Supreme Court on a question whether the appellant is more likely than not to commit a serious act of violence may be the question whether he is more likely than not to commit an offence against laws of the Commonwealth. In either case, as I have indicated, the Supreme Court is to be satisfied only on the balance of probabilities (s 15).
Section 5 performs a double function, that is to say it deals with substantive liabilities and gives jurisdiction with respect to them to the Supreme Court [192] . It does so proleptically, by presenting criteria which require the Supreme Court to decide whether it is more likely than not that the appellant is likely to act in a particular fashion. In the civil law, creation by statute of such norms of general application and for curial enforcement is not unknown. The broad quia timet injunctive provisions in s 80 of the Trade Practices Act 1974 Cth are an example [193] . However, in the present case, the law speaks only ad hominem, applies proleptically the criminal law, determines the case by a civil standard, and provides directly for detention in prison. These are striking features of the legislation. There is, before imprisonment, no determination of guilt solely by application of the law to past events being the facts as found. The consequence is that the legislature employs the Supreme Court to execute, to carry into effect, the legislature's determination that the appellant be dealt with in a particular fashion, with deprivation of his liberty, if he answers specified criteria.
In Chu Kheng Lim v Minister for Immigration [194] , Brennan, Deane and Dawson JJ said that, putting to one side the powers of the legislature to punish for contempt and of military tribunals to punish for breach of military discipline, and the exceptional cases of involuntary detention in cases of mental illness or infectious disease which are non-punitive in character, the involuntary detention of a citizen in the custody of the state is penal or punitive in character. The result was said to be that such power of involuntary detention exists only as an incident of the exclusively judicial function of a judge in punishing criminal guilt. In the same case, Gaudron J [195] said:
Detention in custody in circumstances not involving some breach of the criminal law and not coming within well-accepted categories of the kind to which Brennan, Deane and Dawson JJ refer is offensive to ordinary notions of what is involved in a just society. But I am not presently persuaded that legislation authorising detention in circumstances involving no breach of the criminal law and travelling beyond presently accepted categories is necessarily and inevitably offensive to Ch III. It is apparent that her Honour was speaking of detention authorised by a branch of government other than the judicial branch. Earlier [196] , her Honour had described the ascertainment of guilt or innocence, the object of criminal proceedings, as the exclusive function of the courts. 1. (1992) 176 CLR 1 at 27-28. 2. Lim (1992) 176 CLR 1 at 55 3. Polyukhovich v The Commonwealth (1991) 172 CLR 501 at 705See also the remarks of Deane J in Re Tracey; Ex parte Ryan (1989) 166 CLR 518 at 580, set out earlier in these reasons.
The present case is not one of incarceration by legislative or executive fiat. The involuntary detention of the appellant is brought about by orders of the Supreme Court in exercise of what is described in s 24 as its "jurisdiction" under the Act. I have referred to the striking features of this legislation. They must be considered together. But the most significant of them is that, whilst imprisonment pursuant to Supreme Court order is punitive in nature, it is not consequent upon any adjudgment by the Court of criminal guilt. Plainly, in my view, such an authority could not be conferred by a law of the Commonwealth upon this Court, any other federal court, or a State court exercising federal jurisdiction. Moreover, not only is such an authority non-judicial in nature, it is repugnant to the judicial process in a fundamental degree.
The appellant concedes, rightly, that the New South Wales legislature may confer upon the Supreme Court of New South Wales functions which, in respect of federal jurisdiction and a law of the Commonwealth relating thereto, will be regarded as non-judicial. But, he submits, the jurisdiction conferred by the Act upon the Supreme Court is of such an extraordinary nature as to be incompatible with the exercise by that institution of federal jurisdiction conferred pursuant to s 77(iii) of the Constitution.
The appellant points to the proposition considered in Grollo v Palmer [197] that the power of the Parliament to invest non-judicial functions upon a designated judge of a federal court in his or her individual capacity is limited such that what is entrusted thereby must not be inconsistent with the essence of the judicial function and the proper performance by the judiciary of its responsibilities in the exercise of judicial power. The objective of the doctrine is to forestall the undermining of the efficacy of the exercise of the judicial power of the Commonwealth.
That objective reflects a number of concerns. In the United States, in the context of Art III of the United States Constitution, these have been expressed as follows [198] :
One is that attention to extrajudicial activities is an unwanted diversion from what ought to be the judge's exclusive focus and commitment: deciding cases. Another is that, inasmuch as the judicial method is inappropriate for coping with nonjudicial issues, federal judges have no special competence for disposing of them. Since these issues involve democratic choice, it is politically illegitimate to assign them to the federal judiciary, which is neither responsive nor responsible to the public will. Moreover, it misleads the public to camouflage the legislative character of a social decision and shore up its acceptability by committing it to the judiciary, thereby cashing in on the judicial reputation. Most critically, public confidence in the judiciary is indispensable to the operation of the rule of law; yet this quality is placed in risk whenever judges step outside the courtroom into the vortex of political activity The need to preserve judicial integrity is more than just a matter of judges satisfying themselves that the environment in which they work is sufficiently free of interference to enable them to administer the law honourably and efficiently. Litigants and our citizenry in general must also be satisfied. This was a dissenting judgment, but it was later referred to favourably by the Court of Appeals for the Eleventh Circuit in In re Application of the President's Commission on Organised Crime [199] and by Judge Kozinski in giving the judgment of the Court of Appeals for the Ninth Circuit in Gubiensio-Ortiz v Kanahele [200] . 1. Hobson v Hansen (1967) 265 FSupp 902 at 923, 931, per J S Wright J, dissenting. 2. (1985) 763 F 2d 1191 at 1197-1198, per Fay and Johnson JJ. 3. (1988) 857 F 2d 1245 at 1260-1263. The Supreme Court of the United States granted certiorari, vacated the judgment and remanded this case for further consideration in the light of Mistretta v United States (1989) 488 US 361see (1989) 488 US 1036. Nothing turns on that circumstance for present purposes.
The translation of what may be a politically difficult choice into what one distinguished United States judge called "a grossly unjudicial chore" [201] jeopardises the integrity of the federal or State court in question in the exercise in other cases of the judicial power of the Commonwealth. It saps the appearance of institutional impartiality and the maintenance of public confidence. The point was made by the Supreme Court of the United States in Mistretta [202] :
The legitimacy of the Judicial Branch ultimately depends on its reputation for impartiality and nonpartisanship. That reputation may not be borrowed by the political Branches to cloak their work in the neutral colors of judicial action. Earlier, in In re Richardson [203] , Chief Judge Cardozo had emphasised that judges should be saved "from the entanglements, at times the partisan suspicions, so often the result of other and conflicting duties". 1. Hobson v Hansen (1967) 265 FSupp 902 at 930, per J S Wright J. 2. Mistretta v United States (1989) 488 US 361 at 407 3. (1928) 160 NE 655 at 661.
The appellant submits, and I accept, that such reasoning is particularly applicable where, as here, the Act draws in the Supreme Court of a State as an essential and determinative integer of a scheme whereby, by its order, an individual is incarcerated in a penal institution otherwise than for breach of the criminal law. The appellant points to s 77(iii) of the Constitution, to the extensive investment of State courts with jurisdiction to try offences against federal criminal law [204] , and to trials held concurrently in respect of federal and State offences. He submits that the appearance of institutional impartiality in administering that law, and in inflicting punishment for breach of it, is sapped to an impermissible degree by ad hominen legislation of the nature I have discerned in the Act and described earlier in these reasons. The Act requires the Supreme Court to inflict punishment without any anterior finding of criminal guilt by application of the law to past events, being the facts as found. Such an activity is said to be repugnant to judicial process. I agree.
The Act is an extraordinary piece of legislation. The making thereunder of "detention orders" by the Supreme Court in the exercise of what the statute purports to classify as an augmentation of its ordinary jurisdiction, to the public mind, and in particular to those to be tried before the Supreme Court for offences against one or other or both of the State and federal criminal law, is calculated to have a deleterious effect. This is that the political and policy decisions to which the Act seeks to give effect, involving the incarceration of a citizen by court order but not as punishment for a finding of criminal conduct, have been ratified by the reputation and authority of the Australian judiciary. The judiciary is apt to be seen as but an arm of the executive which implements the will of the legislature. Thereby a perception is created which trenches upon the appearance of institutional impartiality to which I have referred.
The Director sought to suggest that there was nothing particularly remarkable in this legislation. Comparison was invited with earlier New South Wales statutes, including the Inebriates Act 1900 NSW. This was a statute of general application providing for the care, control and treatment of inebriates, including the placement of inebriates for a period not exceeding twelve months in institutions for the reception, control and treatment of inebriates. An order to that effect might be made by a judge of the Supreme Court or any District Court, or by the Master in Lunacy or by any magistrate on proof that, upon medical evidence, the individual in question was a person who habitually used alcoholic liquor or intoxicating or narcotic drugs to excess. This legislation is quite removed in nature and scope from that with which this case is concerned.
The Director sought to soften the stark outlines of the Act by reference to such provisions as s 21. This requires the making of reports to the Director on the "condition and progress" of the appellant while a preventive detention order is in force against him. The reports are to be prepared by assessors, being medical practitioners, psychiatrists or psychologists appointed by the Supreme Court under s 11 "to observe and report" on the appellant. Reports also are to be made under s 21 by the Commissioner of Corrective Services (the Commissioner). This is an officer appointed under s 6 of the Prisons Act who is entrusted with the care, direction, control and management of prisons and empowered to make provision for the training, welfare and aftercare of prisoners.
The reports under s 21 must express opinions on such matters as to whether or not the appellant is "still more likely than not" to commit a serious act of violence and as to whether it is still appropriate "for the protection of a particular person or persons or the community generally" that he be held in custody. There must also be a description of the general behaviour of the appellant during the period to which the report relates. If the report is prepared by an assessor, rather than the Commissioner, it also must contain, inter alia, a description of the current state of the medical, psychiatric and psychological condition of the appellant and express an opinion as to whether any medical, psychiatric or psychological treatment should be made available to him during the remainder of the term of the detention order (s 21(4)).
The Director also pointed to the obligation imposed by s 17(1) upon the Supreme Court, in any proceedings under the statute, to have regard to any report as to the "condition and progress" of the appellant, the preparation of which the Court had required.
In New South Wales, detailed provision with respect to the care, control and treatment of mentally ill and mentally disordered persons is made by the Mental Health Act 1990 NSW. Further, Pt 2 (ss 4-30) of the Mental Health (Criminal Procedure) Act 1990 NSW makes detailed provision with respect to criminal proceedings in the Supreme Court and District Court which relate to persons affected by mental disorders. The primary significance of the provisions of the Act to which I have referred above is their service to emphasise the placement of the appellant outside the general legislative regimes with respect to the care, control and treatment of mentally ill and mentally disordered persons.
The question then becomes whether the Director, and those interveners supporting him, have made out their response that, even if the Act imposes on the Supreme Court an authority the exercise of which undermines and is antipathetic to the exercise by the Supreme Court of the judicial power of the Commonwealth, nevertheless this produces no constitutional consequence such as to invalidate the Act.
The constitutional law submissions for the Director
In meeting the submissions for the appellant, the Director and the interveners also referred to several well-settled propositions as if singly or in combination they provided a sufficient answer. They do not do so.
Pursuant to s 71 of the Constitution, the judicial power of the Commonwealth is vested in courts, including courts invested with federal jurisdiction under s 77(iii). The first proposition is that a law made by the Parliament pursuant to s 77(iii) of the Constitution may invest a State court only with judicial functions and functions incidental thereto [205] .
The second proposition is that, if the Parliament wishes to invest a State court with federal jurisdiction then, subject to s 79 of the Constitution and to prescription of the practice and procedure for the exercise of the invested jurisdiction [206] , the Parliament, to use the oft-repeated phrase of Griffith CJ in Federated Sawmill, Timberyard and General Woodworkers' Employees' Association (Adelaide Branch) v Alexander [207] , "takes the [State] Court as it finds it", with the result that it cannot alter the structure or organisation of the State court.
No difficulty of this nature arises in the present case. The Supreme Court, both at first instance and on appeal, was exercising federal jurisdiction (pursuant to s 76(i) and s 77(ii) of the Constitution and s 39 of the Judiciary Act 1903 Cth) in a matter arising under or involving the interpretation of the Constitution. This followed from the nature of the defences presented by the present appellant to the application to the Supreme Court for his detention under the Act [208] . There was no room for the exercise of a State jurisdiction which the Supreme Court otherwise would have had; the jurisdiction exercised by the Supreme Court was wholly federal [209] . However, the Act provides for the making of preventive detention orders by the Supreme Court and s 24 states the jurisdiction of the Supreme Court under the statute is exercisable by a single judge. Accordingly, none of the issues dealt with in authorities such as The Commonwealth v Hospital Contribution Fund [210] arises.
The third well-settled proposition is that by no means all species of non-judicial power, when entrusted by the Parliament to a federal court, are antipathetic to the proper exercise of the judicial power of the Commonwealth. That this is so is illustrated by, though not co-extensive with, the proposition that there exist powers which, when entrusted to a repository other than a court, are to be characterised as administrative and non-judicial yet, when entrusted in an appropriate context to a court, involve the exercise of judicial power. There is abundant authority to support this functional analysis of the separation of judicial power [211] .
There is a corollary to this proposition, which was not put by the Director but which has some present importance. It is that the term "matter" in Ch III does not exhaust, and is narrower than, what otherwise might be within ordinary concepts of judicial power. Thus, in The Commonwealth v Queensland [212] , Jacobs J pointed out that In re Judiciary and Navigation Acts [213] was not decided as a case where it was sought to impose a non-judicial function upon the High Court. His Honour continued "[i]f it had been, the law as then understood [before the Boilermakers' Case] might not have caused this to be an invalidating characteristic". Rather, it was decided on the footing that judicial power in a general sense was not co-extensive with the narrower limits of judicial power with which Ch III of the Constitution is concerned. Hence, jurisdiction conferred by a State legislature on the courts of the State may be judicial in character, albeit insusceptible of investment by the Parliament of the Commonwealth as federal jurisdiction pursuant to s 77(iii).
The fourth proposition is perhaps a corollary of the third. It is that, as regards the courts of the States, there is no prohibition arising from the Constitution which operates upon State legislatures and necessarily bars the conferral upon State courts of authority which does not have the character or quality of the judicial power of the Commonwealth. Reference is made to Le Mesurier v Connor [214] . But this decision did not determine that a State legislature has power to impose upon the Supreme Court of that State functions which are incompatible with the discharge of obligations to exercise federal jurisdiction, pursuant to an investment by the Parliament of the Commonwealth under s 77(iii) of the Constitution. Nor, in the absence from the Constitution of the State concerned of an express statement of a separation of judicial power, reflecting that achieved by Ch III of the Constitution, did it imply such a capacity.
One strand in the submissions for the Director, which appeared in various places with varying degrees of strength, was an apparent denial of the existence now in Australia of an integrated legal system. To this aspect of the case I briefly turn.
In Fencott v Muller [215] , Mason, Murphy, Brennan and Deane JJ said:
Subject to any contrary provision made by federal law and subject to the limitation upon the capacity of non-federal laws to affect federal courts, non-federal law is part of the single, composite body of law applicable alike to cases determined in the exercise of federal jurisdiction and to cases determined in the exercise of non-federal jurisdiction. 1. (1983) 152 CLR 570 at 607.
Earlier, Sir Owen Dixon had spoken of the Australian judge, federal or State, as administering "the common law as an entire system". Adapting and contrasting a well-known passage in the judgment of Holmes J in Black and White Taxicab & Transfer Co v Brown & Yellow Taxicab & Transfer Co [216] , Sir Owen Dixon had declared that an Australian judge was not bound to resist the impression "that there is one august corpus, to understand which clearly is the only task of any court concerned" [217] .
Upon federation it had become plausible, for the first time, to speak of one Australian judicial system which was a unified structure in that, subject to the certification provisions in s 74 of the Constitution and to the need for leave, all avenues of appeal led ultimately to the Privy Council [218] . The end result of legislation enacted between 1968 and 1986 has been to place this Court at the apex of that unified system.
It was well established, even before the first of these legislative changes, in 1968, that this Court and the Privy Council, by whatever path an appeal came, settled the law for Australia. The law was settled not merely for any particular State and not necessarily in conformity with English law. The decision of their Lordships in 1967 in Australian Consolidated Press Ltd v Uren [219] had emphasised this. Appeals from this Court to the Privy Council were excluded (as to federal matters, save for the theoretical probability of appeals in inter se matters with a s 74 certificate) by the Privy Council (Limitation of Appeals) Act 1968 Cth and (as to the balance) by the Privy Council (Appeals from the High Court) Act 1975 Cth. These laws were a valid exercise of the power conferred on the Parliament by s 74 of the Constitution to make laws limiting the matters in respect of which special leave to appeal from this Court may be asked [220] . Now, since the coming into force of the Australia Acts and the removal by s 11 thereof of the appeal from the Supreme Courts of the States to the Privy Council, s 73 of the Constitution places this Court in final superintendence over the whole of an integrated national court system. This ensures the unity of the common law of Australia.
The 1986 legislation produced the result that there is but one stream of authority in Australia and it flows from this Court throughout the nation. It is unnecessary for present purposes to determine whether the Constitution itself, by reason of the existence of the legislative power in s 51(xxxviii) [221] , from the outset contemplated, without the need for Imperial legislation, what has come to pass with the Australia Acts. This is because in any event the Constitution continues to speak to the present by taking into account the operation of the Australia Act 1986 UK [222] .
The existence of such an integrated system of law and the terms of s 73 itself necessarily imply that there be in each State a body answering the constitutional description of the Supreme Court of that State. Contrary to what appeared to be a tentative submission by the Solicitor-General for New South Wales, it would not be open to the legislature of that State to abolish the Supreme Court and to vest the judicial power of the State in bodies from which there could be no ultimate appeal to this Court.
Further responses by the Director to the case for the appellant are to the effect that (i) the Constitution, in providing for investment of State courts with federal jurisdiction, offers the Parliament of the Commonwealth no more than a facility which it may or may not decide to utilise, and (ii) the State court systems stand outside and have no constitutionally mandated relationship with the federal judicature. Neither proposition should be accepted.
It is, of course, open to the Parliament not to invest State courts in the broad terms found in ss 39 and 68 of the Judiciary Act. That is quite plain. But it does not follow that the Constitution is entirely silent as to the character or quality of the State court system which would provide substitute or alternative tribunals to the creation by the Parliament of a system of federal courts.
One of the reasons of the framers of the Constitution in providing provision for investment of State courts with federal jurisdiction was the saving of the expense then seen to have been involved in the immediate creation of a system of federal courts and, true to their expectations, the development of such a system was long delayed [223] . Sir Kenneth Bailey, writing in 1940 [224] , outlined Ch III as follows:
The scheme of the Constitution therefore was to vest "the judicial power of the Commonwealth" in three repositories: (i) the High Court of Australia, which was provided for in the Constitution itself; (ii) such other "federal courts" as the Parliament should think fit to create; (iii) such other courts as the Parliament should invest with federal jurisdiction. The intention was clearly to place a State court exercising federal jurisdiction on the same general footing as the federal courts which should be created by the Parliament. Their jurisdiction was in both cases to be fixed by the Parliament; their decisions were alike to be subject to appeal to the High Court (ss 77(i), 77(iii), 73). For this similarity there was, of course, good reason. The State courts were to be used, at the discretion of the Commonwealth Parliament, instead of additional federal courts: as "substitute tribunals", in the words of Starke J. [225] 1. The Commonwealth v Limerick Steamship Co Ltd and Kidman (1924) 35 CLR 69 at 90 Bailey, The Federal Jurisdiction of State Courts, Res Judicatae, vol 2 (1940) 109 (Pt 1), 184 (Pt 2). 2. "The Federal Jurisdiction of State Courts", Res Judicatae, vol 2 (1940) 109. 3. The Commonwealth v Limerick Steamship Co Ltd and Kidman (1924) 35 CLR 69 at 116
The expedient provided for in s 77(iii) would be frustrated if there were no system of State courts to provide these substitute tribunals as repositories of the judicial power of the Commonwealth. Federal jurisdiction could not be invested in a State body which was not a "court" within the meaning of s 77(iii) [226] .
The Constitution is premised upon the proposition (in covering cl 5) that, of every State (a phrase which includes new States admitted or established pursuant to s 121 (covering cl 6)) there will be "courts" and "judges". Indeed, any other conclusion, to adopt the phrase of Dixon CJ in another context [227] , "rather taxes legal credulity".
The continuing existence of a system of State courts is recognised not only in covering cl 5 but also in ss 73, 77(iii) and 79. Further, s 51(xxiv) provides for the making of laws with respect to the service and execution throughout the Commonwealth of the civil and criminal process and the judgments of "the courts of the States", and s 51(xxv) provides for laws with respect to the recognition throughout the Commonwealth of "the judicial proceedings of the States". Section 118 requires the giving of full faith and credit, throughout the Commonwealth, to "the judicial proceedings of every State".
Moreover, allowance is to be made for the operation of s 106 of the Constitution. This states:
The Constitution of each State of the Commonwealth shall, subject to this Constitution, continue as at the establishment of the Commonwealth, or as at the admission or establishment of the State, as the case may be, until altered in accordance with the Constitution of the State. (Emphasis added.)
There may be some uncertainty as to the range of statutes (Imperial and local), instruments, conventions and practices which together, or only in some limited fashion, comprise the Constitution of a State as it existed at the establishment of the Commonwealth [228] . It is unnecessary to resolve any such uncertainties at this stage. That is because the Constitution, in the relevant sense, of the colony of New South Wales undoubtedly included the Imperial statute, the New South Wales Constitution Act 1855 Imp [229] . Section 1 thereof authorised the Crown to assent to the Bill set out in Sch 1 which had been passed by the then New South Wales Legislative Council. Clause 42 of the scheduled Bill stated:
All the Courts of Civil and Criminal Jurisdiction within the said Colony and all Charters legal Commissions Powers and Authorities and all Officers judicial administrative or ministerial within the said Colony respectively except in so far as the same may be abolished altered or varied by or may be inconsistent with the provisions of this Act or shall be abolished altered or varied by any Act or Acts of the Legislature of the Colony or other competent authority shall continue to subsist in the same form and with the same effect as if this Act had not been made. Section 38 preserved the commissions of the present judges of the Supreme Court of the colony. With the coming of federation, the effect of the new Constitution was to render the Supreme Court as it stood at the establishment of the Commonwealth, the Supreme Court of the State of New South Wales. But that transmutation was effected "subject to the Constitution" [230] . 1. McGinty v Western Australia (1996) 186 CLR 140 at 259-260 2. 18 & 19 Vict c 54. 3. See, eg, Victoria v The Commonwealth (1971) 122 CLR 353 at 371-372, per Barwick CJ; Muldowney v South Australia (1996) 186 CLR 352 at 376-377, per Gaudron J.
No doubt the other components of the State court systems may change from time to time, and s 77(iii) is to be read in an ambulatory fashion. But s 73(ii) puts the Supreme Courts in a distinct position. Section 73(ii) states that the High Court "shall have" appellate jurisdiction in appeals from "the Supreme Court of any State". The last paragraph of s 73 rendered applicable on such appeals to the High Court, until the Parliament otherwise provided, the conditions of and restrictions on appeals to the Privy Council from the Supreme Courts of the several States.
The meaning of the term "Supreme Court" in s 73 is to be determined in the process of construction of the Constitution and is not to be governed merely by legislation of the relevant State [231] . It is, in this sense, a constitutional expression. The phrase identifies the highest court for the time being in the judicial hierarchy of the State and entrenches a right of appeal from that court to this Court [232] .
In any event, to say of s 77(iii) that it offers to the Commonwealth but a facility, so that the Constitution does not bring the courts of the States necessarily into any relationship with the federal judicial power, does not meet the appellant's case. Section 73(ii) indicates that the functions of the Supreme Courts of the States, at least, are intertwined with the exercise of the judicial power of the Commonwealth. This is because decisions of the State courts, whether or not given in the exercise of invested jurisdiction, yield "matters" which found appeals to this Court under s 73(ii). By this means, the judicial power of the Commonwealth is engaged, at least prospectively, across the range of litigation pursued in the courts of the States.
In providing for the appellate jurisdiction of the High Court, s 73 uses the term "any matter". Further, s 74 speaks of "matters". Brennan J pointed out in Mellifont v Attorney-General (Q) [233] that this indicates that appeals lie to this Court under s 73(i) and (ii) from judgments, decrees, orders and sentences in "matters" as that term is understood in Ch III of the Constitution. However, in the absence of any constitutional separation of judicial power in the States, there has always existed the possibility that the Supreme Courts of the States may be entrusted by State law with a jurisdiction which does not involve the exercise of power which has the same character or quality as the judicial power of the Commonwealth [234] .
This proposition may be tested by consideration of an attempt to appeal to this Court from orders made by the Court of Appeal pursuant to s 25 of the Act in an appeal so framed that there was involved no exercise of federal jurisdiction, because, for example, the Act was wholly valid and ss 76(i) and 77 of the Constitution were not enlivened. In those circumstances, there would have been no exercise of judicial power at the State level so as to found an appeal to this Court.
The result, as was accepted in Mellifont , is that (i) if a State court be invested with, in this sense, a non-judicial power, no exercise of that power can found an appeal to this Court because, as Brennan J pointed out [235] , this Court has no power to make a non-judicial order in place of any non-judicial order which the State court ought to have made at first instance; (ii) appeals lie to this Court under s 73 from the Supreme Courts of the States only from judgments, decrees, orders and sentences made with respect to a "matter" which was the subject matter of the legal proceeding at first instance [236] ; and (iii) a Supreme Court may have received by State law jurisdiction which engages and calls for the exercise of judicial power, but nevertheless does not attract s 73 because no "matter" is involved.
Thus, as both a practical consideration and as a conclusion drawn from the structure of the Constitution, the submissions for the appellant accurately emphasise that the institutional impairment of the judicial power of the Commonwealth inflicted by a statute such as the Act upon the judicial power of the Commonwealth is not to be confessed and avoided by an attempt at segregation of the courts of the States into a distinct and self-contained stratum within the Australian judicature. Rather, there is an integrated Australian legal system, with, at its apex, the exercise by this Court of the judicial power of the Commonwealth.
I have referred to the general investment of federal jurisdiction in the Supreme Court of New South Wales. Particular reference may be made to s 39 and s 68 of the Judiciary Act. Section 109 of the Constitution renders invalid any law of the State of New South Wales to the extent of its inconsistency with those provisions [237] .
A State law which, during the operation of an investment of federal jurisdiction in this way, would alter or impair the operation of the Judiciary Act by removing a condition or characteristic of a court such as the Supreme Court, which is essential to the investiture, is an attempt to alter or impair the operation of the federal law. Accordingly, the State law is inconsistent with it and invalid [238] .
However, in my view, the issue in the present case is best resolved by recourse to the proposition that the Constitution itself is rendered, by covering cl 5, binding on the courts, judges and people of every State notwithstanding anything in the laws of any State. The particular characteristics of the Supreme Court against detraction from which, or impairment of which, by the Act the appellant complains, are mandated by the Constitution itself. Of course, the effect of the constitutional mandate is the protection of the Commonwealth judicial power as and when it may be invested. But the vice from which the Act suffers is not removed by the operation of s 109 upon inconsistent laws. It is removed by the operation of the Constitution itself.
As indicated earlier in these reasons, I accept the submission for the appellant that his detention was not, and any further purported detention under an order made in reliance upon the Act would not be, supported by a valid law of the State.
My conclusion is that the whole of the provisions of Pt 2 of the Act (ss 5-26) are invalid. The provisions of Pt 1 (ss 1-4) are ancillary to Pt 2 and fall with it, save in so far as they may have any valid operation in relation to Pt 3 (ss 27-31).
I turn to Pt 3. Sections 27, 29, 30 and 31 are ancillary to Pt 2 and have no operation in respect of other persons, subject matters or circumstances within the meaning of s 31(2) of the Interpretation Act 1987 NSW, such as to preserve an operation unaffected by the invalidity of the balance of the statute.
No action lies against any person (including the State) for or in respect of any act or omission done or omitted by the person so long as it was done or omitted in good faith for the purposes of, or in connection with the administration or execution of, this Act. The section operates to confer protection in respect of acts or omissions done or omitted "in good faith". But they must have been done or omitted also for the purposes of, or in connection with the administration or execution of, "this Act". That assumes, contrary to the true situation, the validity of the statute. In truth, there was no law to be administered or executed. I conclude that s 28 falls also [239] . 1. Section 35 of the DPP Act also purports to confer protection from liability upon certain officers. There was no point taken as to invalidity of that provision in any of its operations.
The result is that the appeal should be allowed with costs. The orders of the Court of Appeal dismissing the appeal from the primary judge should be set aside. In place thereof the appeal to that Court should be allowed with costs, the orders made by the Supreme Court at first instance, which were entered on 23 February 1995, should be set aside, and the summons filed by the Director on 13 December 1994 should be dismissed with costs.
Kable
Director of Public Prosecutions
(1986) 161 CLR 88
(1986) 7 NSWLR 372
(1976) 15 SASR 66
(1984) 36 SASR 376
(1975) 135 CLR 337
(1996) 186 CLR 140
(1949) 79 CLR 201
(1953) 87 CLR 144
(1929) 42 CLR 481
(1938) 60 CLR 545
(1943) 67 CLR 25
(1970) 122 CLR 69
(1976) 134 CLR 495
(1982) 150 CLR 49
(1995) 184 CLR 348
(1995) 36 NSWLR 374
(1931) 44 CLR 394
(1988) 166 CLR 1
(1991) 172 CLR 501
(1956) 94 CLR 254
(1990) 169 CLR 307
(1960) 103 CLR 15
(1960) 105 CLR 214
(1918) 25 CLR 434
(1916) 22 CLR 437
(1912) 15 CLR 308
(1986) 160 CLR 171
(1992) 174 CLR 455
(1957) 95 CLR 529
(1954) 90 CLR 353
(1996) 189 CLR 1
(1971) 124 CLR 367
(1992) 176 CLR 1
(1988) 164 CLR 465
(1970) 123 CLR 52
(1983) 152 CLR 570
(1971) 127 CLR 106
(1996) 184 CLR 333
(1988) 169 CLR 41
(1992) 177 CLR 1
(1991) 172 CLR 84
(1975) 134 CLR 298
(1924) 35 CLR 69
(1981) 147 CLR 617
(1987) 163 CLR 140
(1987) 163 CLR 656
(1991) 172 CLR 460
(1988) 166 CLR 186
(1943) 67 CLR 1
(1978) 144 CLR 161
(1992) 175 CLR 1
(1993) 178 CLR 477
(1965) 114 CLR 20
(1968) 122 CLR 556
(1972) 128 CLR 427
(1977) 138 CLR 1
(1970) 123 CLR 361
(1995) 183 CLR 525
(1951) 83 CLR 1
(1989) 166 CLR 518
(1994) 182 CLR 51
(1945) 70 CLR 141
(1992) 38 FCR 248
(1985) 158 CLR 596
(1991) 32 FCR 40
(1991) 173 CLR 167
(1995) 183 CLR 323
(1921) 29 CLR 257
(1984) 155 CLR 72
(1967) 117 CLR 221
(1958) 99 CLR 132
(1971) 122 CLR 353
(1996) 186 CLR 352
(1972) 128 CLR 221
(1986) 160 CLR 572
(1991) 173 CLR 289
(1996) 185 CLR 595
(1980) 145 CLR 457
(1932) 48 CLR 128
(1959) 103 CLR 30