EDELMAN J.
Introduction
In a statement quoted by Windeyer J shortly after his retirement, Dixon CJ once said that "[b]efore the reform of the law can be done, it is essential that its doctrines should be understood, and that may mean an investigation of the foundation of those that are to be reformed". The significant reform which this Court has been asked to make on these applications, and the novel content which it has been suggested should be given to a very recent constitutional implication, requires a focus upon the doctrinal and historical foundations for the implication.
I have had the benefit of reading the joint reasons for decision of the other members of this Court. I agree, for the reasons they give, that the challenge by Mr Graham ("the plaintiff") and Mr Te Puia ("the applicant") to s 503A(2) of the Migration Act 1958 (Cth), on the basis that it substantially impairs a court's institutional integrity, should be dismissed. The significant remaining constitutional issue concerns the content to be given to an implied constitutional constraint upon the Commonwealth Parliament's ability to restrict judicial review. On this point, with respect, I depart from the reasoning and the conclusions in the joint reasons.
The essence of the case for the plaintiff and the applicant relied upon the constitutional implication of a minimum provision of judicial review for constitutional writs. The existence of that implication was first suggested by this Court in obiter dicta in Plaintiff S157/2002 v The Commonwealth. One member of the Court in that case had earlier described the Constitution as "silent about the circumstances in which the writs may issue", and as entrenching only "the jurisdiction of this Court when the writs are sought, rather than any particular ground for the issue of the writs". Nevertheless, in the present applications it was assumed by all counsel in this Court that the implication proposed in Plaintiff S157/2002 applied both to secure the existence of judicial review and to ensure a minimum content of judicial review, at least in relation to the remedies prescribed in s 75(v) of the Commonwealth Constitution.
This appears to be the first case in any court since Australian Federation in which it has been sought to apply this implication so as to ensure a minimum content to judicial review rather than merely to secure the existence of judicial review. Novelty is not a basis to reject a submission. But novelty invites grave caution where no submission having this effect has ever been made and accepted during a century of legislation, some of which imposed greater constraints on judicial review than the constraints involved in the legislation in this case. That caution is relevant both (i) to the existence of the implication and (ii) to its scope and content.
As to the existence of the implication, in the absence of any submissions to the contrary, and not without doubt, I proceed on the basis that such an implication exists concerning the content of judicial review. All language requires necessary implications. The reasons for this include inadequate or infelicitous expression or, as asserted in this case, underlying assumptions. But, in general, the higher the level of abstraction at which an implication is expressed the less plausible it will be to characterise the implication as one that has direct effect.
Possibly as a design to meet any objection about the existence of the implication, including the response that an implication expressed at a high level of generality invited an application of unrestrained policy in a manner that might cast doubt upon the legitimacy of the implication, the submissions of the plaintiff and the applicant generally applied established techniques of constitutional construction to determine the minimum content of judicial review to be implied rather than to rely upon the broad implication of a minimum standard of judicial review and to assert some content that should be given to it. To adapt the statement from the decision of this Court in Lange v Australian Broadcasting Corporation, the focus of the plaintiff and the applicant was not generally to ask in the abstract "What should be required by a minimum content of judicial review?" It was to ask "What do the terms and structure of the Constitution require?" The latter question first requires interpretation and construction of the constitutional text, in its context, to determine the essential meaning of its expressions and implications. Like the interpretation and construction of any other text, the essential meaning is not necessarily literal and it proceeds by reference to the way that the essential meaning would be understood by a reasonable, legally informed person at the time of utterance, which in this case is Federation. That process, as the submissions implicitly accepted, is avowedly historical.
The submissions of the plaintiff and the applicant focused upon the constraints placed by the legislation upon judicial review on the ground of unreasonableness. These submissions, if accepted, go further than preventing Parliament from removing the unreasonableness ground for judicial review. Their submissions would require an implied constraint upon parliamentary power preventing Parliament from impairing judicial review on the ground of unreasonableness. The impairment upon which they relied was removing material from the record available for judicial review.
There are two, or possibly three, reasons why the submissions of the plaintiff and the applicant should not be accepted.
The first reason is that they are ahistorical. In a statement in a joint judgment in 1993, repeated in 1997, all members of this Court said, citing authority spanning more than a century, that it is "well settled that the interpretation of a constitution such as ours is necessarily influenced by the fact that its provisions are framed in the language of the English common law, and are to be read in the light of the common law's history". Legal history is relevant to understand the essential content of a constitutional implication in the same manner as it is used to establish the essential characteristics of an expression. As Gaudron and Gummow JJ said in relation to the latter, this involves "legal scholarship in preference to intuition or divination".
One respect in which the submissions by the plaintiff and the applicant are ahistorical is that they require recognition of a constitutional constraint on judicial review which would have the effect that the Constitution would invalidate legislation which is considerably less extreme than legislation which had existed for more than 150 years before Federation, and which had become a standardised restriction in the mid-nineteenth century.
Another respect in which the submissions by the plaintiff and the applicant are ahistorical is that they require recognition of a constraint upon legislative power to restrict production to a court of confidential State papers, despite four decades before Federation of unquestioned acceptance by courts of the conclusive nature of a certificate by a Minister that disclosure of a State paper would be prejudicial to the public service. There was a theoretical possibility, which never arose, of an "extreme case" where the condition that the document was a "State paper" might be challenged so that the rule did not apply. But, even then, when considering the condition for application of the conclusive certificate rule, the court would not examine the document.
It is, of course, possible that the Constitution involved a break from these lengthy legal traditions. However, neither the plaintiff nor the applicant made any submission to this effect. This is unsurprising because it would be difficult to see how the Constitution broke from longstanding, clear, and established legal history by introducing contrary content to a generalised and broad implication which is ultimately founded on the concept of the rule of law, itself a concept the precise content of which is hotly disputed and which, on many accounts, includes notions of certainty and clarity.
The essential content of the implication, which can only give "effect to what is inherent in the text and structure of the Constitution", could potentially apply over time to new facts or circumstances that did not exist at the time of Federation. But the type of legislation to which the implication was sought to be applied in this case is not new. It has very old antecedents. On another, far more controversial, view it has been suggested that essential constitutional or legislative meaning can change with changing social attitudes or changing common law so that the same circumstances could have a different constitutional consequence at different times. This approach, if legitimate, would permit a change in social attitudes or a change in the common law to have the effect that the Constitution can have "two contradictory meanings at different times, each of which is correct at one time but not another". Legislation which was valid in light of social attitudes or the common law in, say, 1929 would become invalid at some unknown later time due to a social or common law change altering the essential content of the implication without any relevant amendment to the Constitution. Unless a truly ambulatory textual foundation existed for constitutional meaning to be updated in this way, a curious circumstance would arise where constitutional meaning would change yet, ex hypothesi, that change would not be sanctioned by the text itself. Such judicial change, however well intentioned, would come at a great cost to representative self-government. The plaintiff and the applicant made no submission to the effect that some social change or common law change subsequent to Federation altered the essential content of the implication at some unknown time. Without any relevant post-Federation facts, circumstances, or common law, this therefore leaves pre-Federation history as having a vital constructional role when determining the content of the implication.
One historical constraint upon judicial review, which was imposed by a legislative technique standardised by the Summary Jurisdiction Act 1848 (11 & 12 Vict c 43), is far more extreme than the constraint which is imposed by the challenged s 503A(2) of the Migration Act. Another more extreme historical constraint was imposed by the conclusive certificate from a Minister that disclosure of a State paper would be prejudicial to the public service. It is not necessary in this case to consider whether laws in those terms would be valid today, assuming that they were expressed in clear language complying with the principle for such a law that "Parliament must squarely confront what it is doing and accept the political cost". The reason why it is not necessary to consider such hypothetical laws is that these more extreme constraints are relevant only as important matters of legal history against which the constitutional implication falls to be understood.
The second reason why I do not accept the submissions of the plaintiff and the applicant concerns the lack of "fit" of those submissions with the existing jurisprudence of this Court, including legislation which this Court has upheld despite imposing greater constraints upon judicial review than s 503A(2). This includes legislation upheld by this Court in 2010.
There may be a third reason why the submissions of the plaintiff and the applicant should not be accepted. This is the difficulty of principle which faces a constitutional implication said to prevent Parliament from impairing the unreasonableness ground of judicial review when that ground of review itself arises only by implication from the statute. That issue can be considered briefly because it is not essential to my conclusion and neither the Minister nor any intervener to each Special Case made any submissions on this point.
The result is that s 503A(2) of the Migration Act is not invalid. No separate submissions were made to suggest that s 501(3), which was also alleged to be invalid in the Special Cases, was invalid for any reason independent of s 503A(2). Consequently, that sub-section also should not be held to be invalid.
The remainder of these reasons is divided as follows:
A. |Outline of the three difficulties with the submissions of the plaintiff and the applicant|[89]|
B.|The entrenched minimum provision of judicial review|[101]|
|Plaintiff S157/2002 v The Commonwealth|[102]|
|Bodruddaza v Minister for Immigration and Multicultural Affairs|[109]|
C. |The limited content of judicial review at Federation|[112]|
D.|Three restrictions upon judicial review prior to, and after, Federation|[116]|
|The first type of restriction: privative clauses|[117]|
|The second type of restriction: broad administrative power|[121]|
|The third type of restriction: reducing the content of the record|[123]|
E.|A fourth restriction: prejudice to the public service|[133]|
|The difficulty in obtaining production of records from the Crown or its officers |[134]|
|The Minister's conclusive certificate in relation to State papers |[136]|
|The decision after Federation in Marconi's Wireless Telegraph Co Ltd v The Commonwealth [No 2]|[150]|
|The decline of the conclusive certificate|[161]|
F.|Section 503A of the Migration Act and the minimum content of judicial review|[168]|
|The first reason why s 503A(2) is not invalid|[170]|
|The second reason why s 503A(2) is not invalid|[174]|
|The third potential reason why s 503A(2) is not invalid|[176]|
G.|Whether the Minister could have been satisfied that cancellation of the visas was in the national interest|[179]|
Conclusion|[185]|
A. Outline of the three difficulties with the submissions of the plaintiff and the applicant
The central provision in question is s 503A of the Migration Act. In the Second Reading Speech of the Bill that introduced this provision, Senator Kemp explained that Australian and international law enforcement agencies were reluctant to provide the Department of Immigration and Multicultural Affairs with criminal intelligence and related, sensitive information unless they could be sure that both the information and its sources could be protected. At its core, s 503A aims to provide those law enforcement agencies with that confidence.
Section 503A is concerned to maintain almost absolute confidentiality over particular information communicated to an authorised migration officer by a gazetted agency on the condition that it be treated as confidential. The gazetted agencies include the Australian Secret Intelligence Service, the Australian Security Intelligence Organisation, the Department of Defence, various crime and corruption commissions, and State and Territory police forces. Also included are foreign law enforcement bodies, or parts of those bodies, from numerous other countries. Information falling within s 503A(2) can only be communicated to a court or tribunal if the Minister makes a declaration that the provision does not prevent disclosure after first consulting with the gazetted agency that provided the information on the condition of confidentiality.
The plaintiff and the applicant submitted that s 503A(2) is invalid due to a constitutional constraint upon legislative power, which requires Parliament not to reduce judicial review in this Court below a minimum standard. That implied constitutional constraint was recognised in obiter dicta of this Court in 2003. Since then, it has only been applied once in this Court, in Bodruddaza v Minister for Immigration and Multicultural Affairs. That was a case where the substance or practical effect of the provision entirely denied the plaintiff, and others in various different circumstances, any right to apply for the relief in this Court guaranteed by s 75(v) of the Constitution. The Court held that it was significant that the legislation in that case had no English comparator at any time before Federation.
The submission of the plaintiff and the applicant was that this Court "should be vigilant in giving real content to the notion of an 'entrenched minimum provision of judicial review'". The plaintiff and the applicant submitted that the "real content" must extend beyond provisions like those in Bodruddaza which entirely deny a person his or her constitutional rights. They submitted that the implied constitutional restraint should encompass a provision, such as s 503A(2), that "stymies" judicial review by removing part of the record which would otherwise be before the reviewing court.
The plaintiff and the applicant submitted that s 503A(2) of the Migration Act was an unconstitutional constraint upon judicial review because the reliance by the Minister upon information that was protected from disclosure under s 503A would mean that it would not be possible for an applicant in a judicial review proceeding to establish that the decision was, "in the relevant administrative law sense, unreasonable". "Unreasonableness" was used to mean reasoning that contains a particular error, gives disproportionate weight to some factor, or is illogical or irrational, including where there is no evidence upon which the Minister can rely. However, even in the case where the Minister asserts that he or she relied upon information that was protected from disclosure under s 503A, it is possible that the remainder of the Minister's reasons might disclose some matter that establishes that the conclusion was unreasonable or irrational. Indeed, the plaintiff and the applicant assumed that this was possible because it was one of the grounds for relief in these matters. The essence of the submission, therefore, was that an implied constitutional minimum provision of judicial review renders invalid legislation that might have the effect of precluding an applicant from knowing the extent of a ground of judicial review that he or she might rely on, namely unreasonableness.
By what process could the "real content" proposed by the plaintiff and the applicant be found to be within the implication of a minimum provision of judicial review? At one point, the submissions of the plaintiff and the applicant appeared to invite a policy choice. For instance, they submitted that the Court should consider "what other courses were available to Parliament". But the submissions of the plaintiff and the applicant more generally conformed with a more legitimate answer, which is that the essential content of the implication is to be found by a consideration of the textual basis for the implication, its function, and its legal and contextual background and history. The conclusion reached should also be assessed in the context of its fit with the established constitutional doctrines developed since Federation.
The parties and interveners were correct to take this historical approach to determine the essential content of the implication. However, it must be emphasised that where the historical approach involves a consideration of common law decisions prior to, and at the time of, Federation, the consideration of these decisions is not for the purpose of assessing whether they are correct or not. Rather, those decisions form part of the context from which the meaning of the Constitution, and the content of its implications, can be derived. As McHugh J said, quoting from Judge Easterbrook, a written constitution "is designed to be an anchor in the past. It creates rules that bind until a supermajority of the living changes them". The identification of the constitutional meaning of the words used at Federation and, a fortiori, the essential nature and content of any implications to be derived from those words is therefore "an essential step in the task of construction".
As explained in the introduction to these reasons, there are at least two obstacles in the path of infusing the implied constitutional restriction with a meaning that precludes legislation which would constrain an unreasonableness ground of judicial review in the manner of s 503A(2). The first obstacle is that considerably more extreme, yet analogous, restrictions were commonplace before Federation and continued subsequently. The second obstacle is the lack of coherence between this implied constitutional restriction and other restrictions upon judicial review that have been held by this Court to be valid.
As for the first obstacle, although there is no precise historical analogue of s 503A(2) of the Migration Act, it has strong historical antecedents. For at least a century and a half before Federation, English legislation had removed far more of the record before a decision maker, and had a considerably greater effect upon the subject matter of judicial review, than s 503A(2). Further, in a compelling analogy relied upon by the Attorney-General of the Commonwealth and the Minister, for nearly half a century before Federation, courts had recognised that a certificate provided to a court by a Minister or head of department was a conclusive basis to refuse production of evidence where the certificate said that disclosure would be prejudicial to the public service. If any document had been provided to a State officer on the condition that it remain confidential, a simple certificate from the Minister that disclosure would cause prejudice to the public service would, without a doubt, have precluded a court from examining that part of the record in any judicial review proceedings. Indeed, the certificate might not even have been needed in judicial review proceedings for various reasons, including an approach which had been taken by some courts that the Crown had a privilege to refuse production.
The second obstacle to a conclusion that s 503A(2) infringes an implied constitutional minimum provision of judicial review is the tension between that conclusion and decisions of this Court that other, more onerous, restrictions upon judicial review are valid. At the time of Federation, there were numerous different techniques by which Parliament could attempt to restrict or constrain judicial review. One of those techniques was a privative clause. A second technique was conferring an extremely broad power upon the decision maker. A third was reducing the content of the material upon which the judicial review was conducted.
The first two techniques have been held by this Court to have potentially significant effects in reducing the content of judicial review. As to the first technique, in R v Hickman; Ex parte Fox and Clinton this Court recognised, as had been the case for centuries, that a privative clause could substantially reduce the circumstances in which judicial review would be permissible, without this leading to invalidity. As to the second technique, in 2010 this Court upheld the validity of the conferral of very broad powers on the Minister, including without any obligation even to consider exercising the power. To find that the third technique is unconstitutional, even though it does not operate as a complete legal or practical constraint upon judicial review, and in circumstances in which it has significant historical analogues at the time of Federation, would cause intolerable inconsistencies with the decisions of this Court, and would involve recognition of a vague and uncertain implication without any clear criterion of application.
The third potential obstacle to accepting the submissions of the plaintiff and the applicant concerns the difficulty of expressing the essential content of the implied constitutional constraint in coherent terms where it involves a restriction upon judicial review on the ground of unreasonableness. The difficulty arises because the basis for judicial review on the ground of unreasonableness is an implication from the statute itself. The statutory implication is based upon limits set by the subject matter, scope and purposes of the legislation. There is, at least, significant tension between (i) the notion that the basis for judicial review of whether a power has been exercised reasonably depends upon a statutory implication from the legislation which creates the power and (ii) the notion that a constitutional implication precludes legislation from unduly constraining judicial review of unreasonable decisions.
B. The entrenched minimum provision of judicial review
The submissions of the plaintiff and the applicant were premised upon a constitutional implication of a minimum provision of judicial review, deriving from s 75(v) of the Constitution and, perhaps more pertinently, its place in the Constitution. The constitutional implication requiring a minimum provision of judicial review has only been recognised relatively recently in this Court.
Plaintiff S157/2002 v The Commonwealth
The implication upon which the plaintiff and the applicant relied was first recognised in obiter dicta in Plaintiff S157/2002 v The Commonwealth. That case concerned s 474 of the Migration Act. Section 474, on one reading, purported to preclude any court, including this Court, from granting relief in relation to the review of many decisions of an administrative character under the Act, including prohibition, mandamus, and injunctions. The Court held that the section did not have that effect. It was only concerned with non-jurisdictional errors. It precluded only remedies for non-jurisdictional error of law on the face of the record. However, in obiter dicta in the joint judgment, the following was said:
"The Act must be read in the context of the operation of s 75 of the Constitution. That section, and specifically s 75(v), introduces into the Constitution of the Commonwealth an entrenched minimum provision of judicial review. There was no precise equivalent to s 75(v) in either of the Constitutions of the United States of America or Canada. The provision of the constitutional writs and the conferral upon this Court of an irremovable jurisdiction to issue them to an officer of the Commonwealth constitutes a textual reinforcement for what Dixon J said about the significance of the rule of law for the Constitution in Australian Communist Party v The Commonwealth. In that case, his Honour stated that the Constitution:
'is an instrument framed in accordance with many traditional conceptions, to some of which it gives effect, as, for example, in separating the judicial power from other functions of government, others of which are simply assumed. Among these I think that it may fairly be said that the rule of law forms an assumption.'
The reservation to this Court by the Constitution of the jurisdiction in all matters in which the named constitutional writs or an injunction are sought against an officer of the Commonwealth is a means of assuring to all people affected that officers of the Commonwealth obey the law and neither exceed nor neglect any jurisdiction which the law confers on them. The centrality, and protective purpose, of the jurisdiction of this Court in that regard places significant barriers in the way of legislative attempts (by privative clauses or otherwise) to impair judicial review of administrative action. Such jurisdiction exists to maintain the federal compact by ensuring that propounded laws are constitutionally valid and ministerial or other official action lawful and within jurisdiction. In any written constitution, where there are disputes over such matters, there must be an authoritative decision-maker. Under the Constitution of the Commonwealth the ultimate decision-maker in all matters where there is a contest, is this Court. The Court must be obedient to its constitutional function. In the end, pursuant to s 75 of the Constitution, this limits the powers of the Parliament or of the Executive to avoid, or confine, judicial review."
On a restricted conception of these two paragraphs, the minimum provision of judicial review introduced by s 75 of the Constitution, and in particular s 75(v), is a requirement that a legislature cannot, in form or in substance, abolish any of the subjects of this Court's jurisdiction in that section, being writs of mandamus or prohibition, or injunctions against an officer of the Commonwealth.
This restricted conception of a minimum provision of judicial review is consistent with the view expressed by one member of the joint judgment little more than two years earlier. It is also consonant with a general understanding of the limited accountability function assigned to s 75(v), including in the Convention Debates. Putting aside other possible purposes for s 75(v), the limited accountability function was not understood as altering the content of judicial review or reducing the powers of Parliament, particularly below those that previously existed. Instead, the limited accountability function of the provision was understood to remove doubt about a source of power to make particular orders. In other words, it was included "for more abundant caution" or as "a matter of safety". The caution was to avoid the conclusion in Marbury v Madison, where it was held that Congress could not confer a power to order mandamus upon the United States Supreme Court. During the Convention Debates, Messrs Barton and Symon both insisted that the proposed provision did not confer upon the High Court any right to interfere with "the political Executive of the Federation". Later, Mr Symon defended the proposed provision by explaining that:
"[t]he right to mandamus or prohibition is not conferred one whit more than at present. The provision merely throws within the ambit of the jurisdiction of the federal tribunal the right to determine the question. That question will be determined by the ordinary law of England - by the principles of constitutional government and the prerogatives of the Crown."
However, read in their full context, the two paragraphs in Plaintiff S157/2002 might be thought to travel beyond this limited conception of an accountability function so that the "minimum provision of judicial review" is not merely a guarantee of jurisdiction in relation to mandamus, prohibition, or injunctions against an officer of the Commonwealth. The larger conception is that s 75, and particularly s 75(v), is also concerned with the content of the power of judicial review, although indirectly as a "textual reinforcement" of a broader implication deriving from the "rule of law" and reflected in the "centrality, and protective purpose, of the jurisdiction of this Court". That indirect implication means that, apart from the s 75(v) guarantee of jurisdiction, there is some other, undefined, implied limit to "the powers of the Parliament or of the Executive to avoid, or confine, judicial review".
There are strong reasons for a cautious approach to the application of this broader conception of an implied limitation upon legislative power to constrain judicial review. As Dr Burton Crawford has observed, the reliance by the joint judgment in Plaintiff S157/2002 upon Dixon J's reference to the rule of law was "a strange reconfiguration of Dixon J's statement" that did "little to clarify the role played by the rule of law in the Australian constitutional framework". If the "rule of law" requires more than a guarantee of judicial review of executive action but instead restrains Parliament from "confining" the content of judicial review, then how great must that confinement be before the legislation will be invalid? Which of the different versions of the "contested concept" of the "rule of law", whether thick or thin, is the basis for the implied constraint? Is it necessary, or possible, to "mine the rule-of-law ideal for particular values which may then be translated into legal principles"? How much weight should be put on the value said to be implicit in the rule of law that "the making of particular laws should be guided by open and relatively stable general rules"? Would this same rule of law conception extend beyond the text of s 75(v), which focuses upon the High Court, and apply also to other federal courts? Would such principles extend to the removal of a ground for judicial review, such as procedural fairness, that parliaments have long assumed that they have power to remove?
The difficulty, perhaps impossibility, of giving satisfactory answers to these questions may have been a reason why two of the members of the joint judgment in Plaintiff S157/2002, McHugh and Gummow JJ, only a week after the delivery of that decision, said in their reasons in another decision that it would be "going much further" than merely recognising the values underlying the rule of law "to give those values an immediate normative operation in applying the Constitution".
Nevertheless, as the parties and interveners assumed, if essential content is to be given to the implied constitutional restriction on parliamentary power beyond legislation that has the effect of abolishing some or all of this Court's jurisdiction in s 75(v), then, at the very least, any analogous historical antecedents to the relevant legislation will be very significant. Shorn from history and the context in which the text and structure of the Constitution was enacted, the application of such an abstract conception to give essential content to a constitutional implication could require the Court simply to make unmediated policy decisions. Such unmediated policy decisions could only be made in two ways. One is by reference to the policy views of the individual members of the Court. The other is by reference to the alleged perceptions of the public or some section of it. Both are illegitimate in this context. As to the former, speaking of the abstract wording in the express Due Process clause in the United States Constitution, the United States Supreme Court has described the caution to avoid "the liberty protected by the Due Process Clause [being] subtly transformed into the policy preferences of the Members of this Court". As to the latter, in Nicholas v The Queen, Brennan CJ said of a criterion of "public perception" that this "would be to assert an uncontrolled and uncontrollable power of judicial veto over the exercise of legislative power".
Bodruddaza v Minister for Immigration and Multicultural Affairs
The decision of this Court in Bodruddaza v Minister for Immigration and Multicultural Affairs is not inconsistent with a narrow conception of the function of s 75(v) of the Constitution, namely a constraint upon the legislature, in form or in substance, abolishing the writs of mandamus or prohibition, or injunctions against an officer of the Commonwealth. In that case, the Court held that the entrenched minimum provision of judicial review had been transgressed by s 486A of the Migration Act, which purported to impose a fixed time limit upon an application to this Court for a writ of mandamus, prohibition or certiorari, or an injunction or declaration, in respect of "a privative clause decision". There was no judicial discretion to extend the time limit. Two points are critical to the understanding of that decision.
First, in a number of circumstances, s 486A of the Migration Act operated to preclude entirely, as a matter of practical effect, any claim based upon s 75(v). The time limit was "of such short duration as to deny access to federal jurisdiction to applicants whose delay might not be the result of gross delay or culpable error". The joint judgment cited circumstances where access to the jurisdiction was effectively denied, including where the applicant could not have been aware of corrupt inducement, or actual or apprehended bias, until after expiry of the time limit. As a matter of substance, therefore, the section purported to deny that which the Constitution guaranteed. As Gageler and Keane JJ observed after s 486A was amended to allow a discretion to extend time, the amended provision regulated the exercise of the s 75(v) jurisdiction but did not, and could not, "impose a condition precedent to the invocation of that jurisdiction".
Secondly, although recognising that the "anterior situation in England has not generally been accepted as a comprehensive guide to the operation of s 75(v)", in the joint judgment in Bodruddaza their Honours described as a matter of "significance" the absence in England before Australian Federation of legislatively fixed time bars to judicial review. Equally, for the reasons I have explained, the position concerning restrictions upon judicial review at the time of Federation is highly significant for the applications currently before this Court.
C. The limited content of judicial review at Federation
Two points should be noted about the nature of judicial review in England and the United States at the time of Federation in order to appreciate its limited content in cases of jurisdictional error.
First, a dominant view of jurisdictional error in the nineteenth century, later rejected in Anisminic Ltd v Foreign Compensation Commission, was that it was concerned only with the question of whether the decision maker had power to enter upon the inquiry. On that view, described by Lord Sumner in 1922 as "never since ... seriously disputed in England", jurisdictional error was not concerned with errors made in the course of the decision making process other than those on the face of the proceedings. The leading case that was associated with this view was R v Bolton. In that case, Lord Denman CJ said that:
"the enquiry before us must be limited to this, whether the magistrates had jurisdiction to inquire and determine, supposing the facts alleged in the information to be true: for it has not been contended that there was any irregularity on the face of their proceedings."
In other words, at Federation, courts had not generally embraced the approach to jurisdictional error which recognised, as Lord Reid did in Anisminic, the "many cases" in which a tribunal's decision was a nullity because of something that it had done or failed to do during the course of the inquiry.
Secondly, at the time of Federation, the writ of certiorari had limited recognition as a writ to quash a decision of an officer of the Executive. In England, certiorari to remove and quash a decision required that the decision was one in which the decision maker was under a duty to act judicially. In 1894, Wright J maintained that the writ of certiorari was applicable to "judicial proceedings of courts" but that "the county council [was] not a court, and its proceedings [were] not judicial". Other cases construed the concept of a "judicial" act more liberally, "including many acts that would not ordinarily be termed 'judicial'". Although the English approach that drew a distinction between judicial and ministerial acts was not far from the United States "political question doctrine", the United States cases were even more restrictive in relation to certiorari. In 1913, in Degge v Hitchcock, the United States Supreme Court observed that the case was the first time a writ of certiorari had been sought against an officer of the Executive. The Supreme Court held that the writ was unavailable for the purposes of reviewing an administrative order.
D. Three restrictions upon judicial review prior to, and after, Federation
It must have been very well known at the time of Federation that there were numerous techniques by which parliaments had restricted or constrained judicial review. In a later article, De Smith described three of them. One of those techniques was a common privative clause. A second technique was "giving powers to Ministers and other statutory bodies in terms so broad that it becomes difficult for a court ever to hold that they have been exceeded". A third was "prescribing common forms for summary convictions, omitting a recital of the evidence and of the reasons for the decision". An appreciation of these restrictions is necessary to place in context the plaintiff and the applicant's submissions concerning the essential content of the constraint upon parliamentary power to restrict judicial review.
The first type of restriction: privative clauses
There is a fundamental difficulty involved with privative clauses. A court that upholds a privative clause could only do so for one of two reasons. First, the court might conclude that the effect of the clause was to make anything at all that the decision maker did legal. That is impossible. No tribunal or decision maker, outside an absolute dictatorship, has unlimited power. Secondly, but with the same effect, the court might conclude that any unlawful decision by the decision maker fell outside the authority of the courts. As Wade expressed the point, "[i]f a ministry or tribunal can be made a law unto itself, it is made a potential dictator; and for this there can be no place in a constitution founded on the rule of law".
In Hickman, this Court adopted, as a matter of statutory construction, a pre-Federation technique of reconciliation of the literal terms of a privative clause and the impossibility, outside dictatorship, of a complete lack of review that the clause purported to require. The Court considered a regulation that provided that a decision of a Local Reference Board "shall not be challenged, appealed against, quashed or called into question, or be subject to prohibition, mandamus or injunction, in any court on any account whatever". A writ of prohibition was sought, and granted, in relation to an award made by the Board. In his well-known reasons for decision, Dixon J explained why the regulation did not exclude prohibition. Speaking of privative clauses, of which the regulation was an example, his Honour said:
"Such a clause is interpreted as meaning that no decision which is in fact given by the body concerned shall be invalidated on the ground that it has not conformed to the requirements governing its proceedings or the exercise of its authority or has not confined its acts within the limits laid down by the instrument giving it authority, provided always that its decision is a bona fide attempt to exercise its power, that it relates to the subject matter of the legislation, and that it is reasonably capable of reference to the power given to the body."
Although the meaning of a privative clause will ultimately be a matter of construction, it has been said on many occasions that such a clause might be upheld as within the power of Parliament, provided that (i) the decision of the body does not exceed the authority conferred by the legislation; (ii) the decision of the body constitutes a bona fide attempt to exercise the powers in issue; and (iii) the decision relates to the subject matter of the legislation. The third criterion was said to require that the decision must not, on its face, go beyond power. It was said that a statute which confers on an administrator a jurisdiction that complies with the Hickman principle "will be sufficient to satisfy any constitutional minimum that may exist".
However, the Hickman principle was later said to have been placed "in perspective" by the decision of this Court in Plaintiff S157/2002, which explained that the three Hickman factors were necessary but not sufficient. Nevertheless, this Court in Plaintiff S157/2002 did not deny that which had been assumed before the decision, namely that "there is considerable scope for the legislative conferral of jurisdiction on an administrator in terms which by-pass entirely the traditional grounds of judicial review". Indeed, this Court in Plaintiff S157/2002 recognised that the roots of the Hickman principle, which gave substantial but not absolute effect to a privative clause, were longstanding and predated Federation.
The second type of restriction: broad administrative power
It was well settled at the time of Federation that, as the Lord Chancellor said in a well-known passage in Julius v Lord Bishop of Oxford, mandamus would only lie if a power to act was coupled with a duty to act. By removing the duty to act, there was little utility in seeking judicial review of a decision of an official about whether to act.
An extreme example of this restriction was considered by this Court in Plaintiff M61/2010E v The Commonwealth. The provision of the Migration Act that was challenged in that case, s 46A, prohibited visa applications from "offshore entry persons" unless the Minister thought that an application was in the public interest. That power was exercisable by the Minister personally, but the Minister did not have a duty even to consider whether to exercise the power. This Court unanimously held that s 46A was valid. The Court rejected submissions that it was inconsistent with s 75(v) or its place or purpose in the Constitution, or contrary to the rule of law.
The third type of restriction: reducing the content of the record
Notwithstanding the dominant historical approach, which (i) recognised only limited types of jurisdictional error, and (ii) was subject to legislative restrictions, it was not difficult for courts to quash decisions, usually convictions, for jurisdictional error or non-jurisdictional error of law appearing on the face of the record.
Some of the errors that led to the quashing of convictions were purely formal. Formal errors in recording a conviction were easily made. In 1764, the justice of the peace and scholar Richard Burn wrote of the other justices that "there is not one of them in ten who knows how to draw up a conviction in form" and that even "the greatest lawyers have found it difficult enough, to guard a conviction, so that other lawyers could not break into it". Later, speaking in the House of Lords in 1803, Lord Sheffield said that "not one conviction in a thousand, if attacked, could stand". In the nineteenth century, some magistrates would refer convictions to "star barristers" for settlement. In 1824 a clerk of the peace deposed that he sent the particulars of the proceedings to Joseph Chitty with instructions to prepare a draft of the conviction.
Other errors were substantive, involving errors of law that would be evident when the conviction also set out the evidence upon which it had been obtained. For instance, in R v Little, Lord Mansfield quashed a conviction that a person "traded as a hawker, pedlar or petty chapman" when the only evidence was a single act of selling a parcel of silk handkerchiefs.
The perceived extent to which, and ease by which, convictions could be overturned, particularly on formal grounds, was a matter of controversy. In R v The Inhabitants of the Parish of Ruyton of the Eleven Towns, Crompton J spoke of a privative clause in 1848 legislation designed to address "a practice which had introduced lamentable and disgraceful technicalities". In the same case, Hill J described the statute as having a "plain and manifest intention" to "get rid of expensive and useless litigation upon points of mere form, and to facilitate the settlement of all disputes upon contested orders of removal".
In 1848, the English Parliament passed three Acts. They were commonly described as "Jervis' Acts" after the Attorney-General, Sir John Jervis, who secured their passage. One of those Acts, the Summary Jurisdiction Act 1848, provided in s 17 for a standard form conviction. The standard form substantially reduced the risk of formal errors. It also removed much of the record that could be reviewed for the purpose of certiorari, because the standard form did not contain any of the evidence. The standard form that was introduced by the Summary Jurisdiction Act was not new. It merely standardised a routine approach, possibly first created by Sir Joseph Jekyll MR's drafting in the Spirit Duties Act 1735, that had been adopted in various different legislation for more than a century.
In R v Nat Bell Liquors Ltd, the Privy Council said that the Summary Jurisdiction Act "cut down the contents of the record, and so did away with a host of discussions as to error apparent on its face ... [T]he grounds for quashing on certiorari came in practice to be grounds relating to competence and disqualification." The Privy Council also said that the legislation:
"did not stint the jurisdiction of the Queen's Bench, or alter the actual law of certiorari. What it did was to disarm its exercise. The effect was not to make that which had been error, error no longer, but to remove nearly all opportunity for its detection. The face of the record 'spoke' no longer: it was the inscrutable face of a sphinx."
As Costello has observed, this was an exaggeration. It was not the Summary Jurisdiction Act which wrought the decline of certiorari. That Act merely adopted, and standardised, the same approach taken in various legislative provisions over the previous century. Further, the Summary Jurisdiction Act did not remove all opportunity for the detection of error. Although it substantially removed the prospect of establishing any jurisdictional or non-jurisdictional errors of law from the record, it remained possible, as older cases had recognised, for jurisdictional error to be proved by affidavit alleging failure of a jurisdictional precondition, bias, apprehended bias, or a failure to afford procedural fairness.
In summary, originally the writ of certiorari "to remove" only ordered the lower court to send up the indictment or presentment. The Summary Jurisdiction Act standardised an approach that had developed over a century. That approach substantially confined the record before the reviewing court. The provisions of the Summary Jurisdiction Act were not read down. They were not disapplied. This was so even though, by 1848, there was a long-established judicial approach by which courts had refused to apply, or had read down, privative clauses. As Dr Murray said:
"the court could have exploited the ambiguity in their conceptualisation of jurisdiction, widening the conception so as to allow for the breadth of review that had been available on the face of the record before the introduction of statutory short-form convictions. To do so, however, would have been constitutionally difficult: Parliament had clearly intended a restriction of the inherent power of the Queen's Bench to review convictions, and an arrogation of reviewing powers in the face of this legislative intention would have had the effect of undermining this."
Although the general provisions of the Summary Jurisdiction Act 1848 were substantially ameliorated by the "case stated" procedure that was introduced by the Summary Jurisdiction Act 1857 (20 & 21 Vict c 43), the scope of certiorari was still substantially narrowed by the confinement of the record.
The form prescribed in the Summary Jurisdiction Act in 1848 was still in use in the Criminal Code of Canada in 1922. That Canadian legislation was considered in R v Nat Bell Liquors Ltd, where the Privy Council restored a conviction of a magistrate because the evidence before the magistrate formed no part of the record, and was not "available material on which the superior Court [could] enter on an examination of the proceedings below for the purpose of quashing the conviction".
E. A fourth restriction: prejudice to the public service
A fourth constraint existing at Federation, and for decades subsequently, was relied upon heavily in this case by the Attorney-General of the Commonwealth and the Minister. This constraint was the power of a Minister or head of department to object to the production of records on the basis of prejudice to the public service. The existence, at the time of Federation and for several further decades, of this common law and occasionally statutory bar to production of documents in court is relevant to appreciate the meaning and scope of the implied constitutional restriction imposing a minimum standard of judicial review.
The difficulty in obtaining production of records from the Crown or its officers
It appears that, at the time of Federation, this objection never needed to be raised in any judicial review proceedings. There are a number of possible reasons for this. One reason is that, although in the nineteenth century the Crown was entitled to discovery from a subject, there was a view that a subject was not entitled to discovery from the Crown. So far as it applied, this constraint would extend to a range of Crown officers, including Ministers of State. Even s 50 of the Common Law Procedure Act 1854 (17 & 18 Vict c 125) was held not to extend to conferring upon a subject a right to discovery against the Crown in a petition of right.
Whether or not it might ever have been needed in judicial review proceedings, the principle that permitted a conclusive certificate as to the public interest was pervasive and would have applied if it were needed by the Crown or its officers. Even when, in 1947, English legislation abolished any Crown privilege not to give discovery, that legislation preserved the rule of law that authorised the withholding of any document on the ground that disclosure would be injurious to the public interest.
The Minister's conclusive certificate in relation to State papers
At the time of Federation, it had been a universally held view for 40 years that public interest immunity could be invoked over State papers simply by evidence from a Minister or head of department that there would be prejudice to the public service from disclosure. The authorities held that such evidence could not be questioned by a judge. In an extreme case, where it was clear from the document's description that it was not a State paper, then the immunity did not apply, and the question of the conclusive nature of the evidence from the Executive did not arise.
At the time of Federation, the leading case for 40 years was Beatson v Skene. In that case, the plaintiff's claim of slander was dismissed by a jury. The plaintiff had sought production of various documents prior to trial. But the Secretary of State for War objected to production on the basis that it would be injurious to the public service. Baron Bramwell upheld the objection. The plaintiff applied for a new trial. The application was heard by a powerful Court of Exchequer Chamber comprised of Pollock CB, Martin B, Wilde B, and Bramwell B (the trial judge). One ground was the refusal of the trial judge to compel production of various documents. The Court reiterated the general principle that if production would be injurious to the public service then the general public interest must prevail over the private interest of an individual suitor. However, the Court went on to ask whether the public interest was to be determined by the presiding judge, who would need to ascertain in court what the document was and why its production would be injurious to the public service, or by the responsible servant of the Crown with the custody of the paper. The Court concluded:
"It appears to us, therefore, that the question, whether the production of the documents would be injurious to the public service, must be determined, not by the Judge but by the head of the department having the custody of the paper; and if he is in attendance and states that in his opinion the production of the document would be injurious to the public service, we think the Judge ought not to compel the production of it."
One reason that was given for this view was the danger of disclosure if such an inquiry took place in public, as was assumed would necessarily be the case.
The decision of the Court in Beatson was given by the Chief Baron, who explained that Martin B did not "entirely agree". He explained that the reason for disagreement was the view of Martin B that the judge should compel production, despite objection by the head of the department, if the judge were satisfied that the document could be made public without prejudice to the public service. This view was rejected. The Chief Baron said that:
"perhaps cases might arise where the matter would be so clear that the Judge might well ask for it, in spite of some official scruples as to producing it; but this must be considered rather as an extreme case, and extreme cases throw very little light on the practical rules of life."
Importantly, as a matter of logic, it could only be "so clear" in such an "extreme case" if the description of the document revealed that it was not a State paper. The rule in Beatson, that it was for the responsible servant of the Crown to determine whether production of a State paper was injurious to the public service, was applied many times before Federation. It does not appear to have ever been questioned before Federation.
The rule was applied in HMS Bellerophon. In that case, the Secretary of the Admiralty provided an affidavit that deposed that there would be prejudice to the public service if a report of a collision of one of her Majesty's ships were liable to be inspected in legal proceedings. Counsel for the plaintiffs argued that it was "for the Court and not for the Secretary of the Admiralty to judge whether the inspection would be prejudicial". Sir Robert Phillimore rejected that submission and followed the decision in Beatson.
Another case in which the principle was applied was Latter v Goolden. That decision was not reported other than a brief reference in The Times but it appears to have been printed as a precedent for use by the Home Office. It involved an alleged libel contained in a letter written by the defendant firm of electrical engineers to the Civil Service Commissioners. The head of the department of the Civil Service was called, and refused to produce the document on the ground of public policy. A contemporary text described the head of department as saying only that the document was confidential, that the Civil Service received some 20,000 letters a year, and that the Government would be "much inconvenienced if the production of any of them could be enforced in a Court of Law". The claim of privilege was upheld solely upon this evidence of the head of department, without examining the document. Lord Esher MR is quoted as having said that the cases clearly show that when the head of a public department says that it would be contrary to the public interest to produce a document, "it is for him to say so".
A further case was Hughes v Vargas. That was a libel action heard at first instance by Hawkins J. The libel was alleged to be contained in an official report made to the Board of Inland Revenue. The Secretary to the Board of Inland Revenue gave evidence that the Board had instructed him to object to production on the ground that to do so would be prejudicial and injurious to the public service. The Chairman of the Board also objected on this ground. The plaintiff submitted that the court could "go behind the objection of the witnesses and see whether it was well-founded or not". Justice Hawkins refused production, holding that he "could not inquire into the grounds for the objection of those who had the custody of the document". This decision was upheld by the Court of Appeal. Lord Esher MR referred to the rule in Beatson, and reiterated that:
"It is not for the Judge to decide whether the production of a document would be detrimental to the public service, but if the document is a [S]tate document, that is belonging to a [S]tate office, and is in the hands of a public department of [S]tate, if the head of that department takes the objection that it would be contrary to the public welfare to produce the document in Court, the Judge must act upon that, and he is not to inquire closely as to what are the grounds of the objection, or whether he would take the same objection if he were in the place of the public officer."
The same point was made in 1895 by the editor of Judge Pitt Taylor's A Treatise on the Law of Evidence as Administered in England and Ireland, commonly referred to as Taylor on Evidence. One class of public interest immunity was described as the exclusion of evidence from "motives of public policy", which included matters that concern the administration of government and "the disclosure of which would be prejudicial to the public interest". The text continued:
"According to the Court of Appeal [in Hughes v Vargas], the minister to whose department a document belongs, or the head of a department in whose custody it is, is the exclusive judge as to whether such document is or is not protected from production on grounds of State policy, and if he claims such protection the court will not go behind the claim, or inquire whether the document be or be not one which can properly be the subject of such a claim."
Although the court would not go behind such a claim by a Minister or head of department, the same might not be true if a subordinate officer attended court. Only then, as Taylor on Evidence explained, might the judge examine the documents to decide whether they should be withheld.
In 1898, Phipson also set out this rule, relying on cases including Beatson. He explained that one application of public interest immunity was that witnesses "may not be asked, and will not be allowed, to state facts or to produce documents the disclosure of which would be prejudicial to the public service". He continued, explaining that where "the head of the department objects, the judge will not compel the production, nor decide upon the validity of the objection, unless it is a palpably futile one".
The suggestion by Phipson that a court might decide upon the validity of a "palpably futile" objection was plainly a reference to the "extreme case" to which Pollock CB referred in Beatson. The supposition seems to have been that a court might decide contrary to the objection of the head of department if, without inspecting the document, it was clear from its description that it simply was not a State document. In other words, inspection was possible where it was palpably clear that a document possessed by the State was clearly not a State document in any sense, and therefore did not fall within the circumstances or category in which the head of department could give conclusive evidence. An example of such a case was given by Wright J in In re Joseph Hargreaves Ltd. In that case the Board of Inland Revenue claimed that the production of a company's balance sheets on a misfeasance summons would be injurious to the public interest. The claim was upheld on the basis of the certificate of the Board. As Vaughan Williams LJ put the point on appeal, upholding the decision of Wright J, the communications were documents that came "within the rule which enables the heads of Government departments to object on their own responsibility to their production". The example given by Wright J was that it might be different if the documents were "documents belonging to the company which by some accident had got into the hands of the Inland Revenue officers".
At the time of Federation, the principle was therefore clear. A certificate from a Minister or head of department that disclosure would be injurious to the public interest was conclusive. It was "not for the judge to say whether the production of any particular document [was] injurious to the public service or not". There was a theoretical possibility that the immunity would not apply in an "extreme case" where the description of the document was such that it clearly fell outside the class within which protection could attach. However, this was not because the evidence or certificate was inconclusive. It was because the document fell outside the class to which protection could be given. In other words, the conclusive nature of the certificate applied only to State papers. It did not apply in an extreme case where it was clear without examination of the document that the document possessed by the State could not, in any sense, contain a State communication.
Various overlapping reasons were advanced in the cases for the conclusive nature of the certificate. One reason was that something that a judge might think to be innocuous would be thought to be noxious by "the better informed officials of the public department". The reason given in Beatson in 1860 was that the privilege would be defeated if there were argument about the content of the document in open court. Nearly a century later, the House of Lords maintained that a court could not be closed for the purpose of seeing the documents in the absence of the other party because:
"where the Crown is a party to the litigation, this would amount to communicating with one party to the exclusion of the other, and it is a first principle of justice that the judge should have no dealings on the matter in hand with one litigant save in the presence of and to the equal knowledge of the other."
The decision after Federation in Marconi's Wireless Telegraph Co Ltd v The Commonwealth [No 2]
The rule which recognised the conclusive nature of a Minister's certificate that production of a State paper was not in the public interest was considered more than a decade after Federation by this Court in Marconi's Wireless Telegraph Co Ltd v The Commonwealth [No 2]. In that case, an action was brought by Marconi's Wireless against the Commonwealth for infringement of a patent. The infringing item was said to be an apparatus for wireless telegraphy. The Postmaster-General opposed inspection of the apparatus, alleging that inspection would be prejudicial to the public interest and welfare of the Commonwealth. Marconi's Wireless submitted that it was the duty of the Court to inquire into whether the disclosure would prejudice the public interest, and that a government official could not "by his mere ipse dixit add to the class of State secrets recognized by law". A majority of the Court, Griffith CJ and Barton J (Isaacs J dissenting), rejected the claim for public interest privilege.
An important point about Marconi's Wireless Telegraph is that the Court was unanimous about the conclusive nature of a certificate from a Minister or head of department. The Court only divided on the question of the scope of the extreme case where the communication about which disclosure was sought palpably did not fall within the class of "State papers" to which the privilege applied. If the communication fell outside that class then the view of the Minister or head of department would not be conclusive.
In the majority, Griffith CJ referred to Beatson with approval but distinguished it on the basis that there was no question about the "general character" of the documents in Beatson. The Chief Justice said that courts had "never abdicated the duty of considering whether the documents, in respect of which the claim is made, come within the reason of the rule" (emphasis added). The claim for privilege was dismissed by the Chief Justice because the disclosure fell outside the class, ie State papers, which came within the reason of the rule. He explained that no reason was given on a "preliminary inquiry into the nature of the alleged State secret" to justify a class that could include the apparatus.
Also in the majority, Barton J endorsed the approach in Taylor on Evidence in relation to a class of case where "evidence is excluded from motives of policy". This class "comprise[d] secrets of State, or matters which concern the administration, either of penal justice, or of government, and the disclosure of which would be prejudicial to the public interest" (emphasis in original). Justice Barton continued:
"No doubt, if the Court sees that a document is of the character by reason of which the privilege arises, it will admit the claim to protection, and will allow the Minister or head of the department having the custody of the paper to determine in each case whether production or inspection would be injurious to the public service; for that authority is the better judge of matters of public policy." (emphasis added)
His Honour emphasised that in the "preliminary consideration" of whether a document belongs to a class where a government objection will be conclusive, "production is, of course, not necessary: it is sufficient that there be such a description as will enable the Court to see whether it comes within the class". Like Griffith CJ, his Honour also referred with approval to Beatson, reiterating the rule that it was for the responsible servant of the Crown to determine whether production of a State paper was injurious to the public service, but adding that "before that question arises for determination the Court must be able to see that it is a State Paper". Justice Barton reiterated that after the preliminary inquiry into whether the description of the document indicates that it falls within the protected class, the Court will "refuse production or inspection if the responsible officer who is the custodian determines that such publicity would be injurious to the public interest". Since the apparatus was not in the class, it was not possible for the Minister to say anything that would bring it within the class.
Justice Isaacs dissented. His Honour thought that the case was so clear that all that would have been needed was a short quotation from Beatson. But he considered it necessary to write a lengthy decision to expose what he described as the "misapprehension" of the majority in the Court that would undermine "the rule in a most important branch of the law touching the relative functions of the Executive and the Judiciary". His Honour considered that the relevant category was simply that the documents were "in possession of a government department". Apart from "extreme cases" that were "practically negligible", all documents in the possession of a government department would be State papers, and the assurance by the Minister of public prejudice was conclusive evidence upon which the Court must act. His Honour then applied the same principles that had been developed in relation to documents to the apparatus. It was a State apparatus, and the Court was required to give effect to the view of the Minister.
In summary, by the time of Marconi's Wireless Telegraph, it had been established for half a century that, in cases of State papers, the evidence of a Minister or head of department that disclosure of a document would be injurious to the public service was conclusive. A court was bound to follow it without inspecting the document. The decision in Marconi's Wireless Telegraph shows only that there was doubt, by 1913, about the scope of the category of State papers in which the evidence was conclusive. But there was no doubt that if a document fell within the category then the view of the Minister or head of department was conclusive.
An application was brought in Marconi's Wireless Telegraph for special leave to appeal to the Privy Council. In the course of refusing special leave to appeal, the Lord Chancellor said "[o]f course the Minister's statement or certificate must be conclusive on a particular document. How can it be otherwise? ... If the Minister certifies quite specifically, his certificate is to be taken as conclusive".
In 1914, immediately following Marconi's Wireless Telegraph, the matter arose again in the Supreme Court of Victoria in Spong v Spong. In that case it was argued that the decision in Marconi's Wireless Telegraph meant that the Court was not bound to accept a statement that had been made by the Minister that production upon subpoena would be detrimental to the public interest. The case was an extreme one. Both parties to the litigation wanted production of the documents. The trial judge explained that his view was that the documents alone, from their description, could be produced without injury because they were merely statements about the sale of land that contained information that could be discovered from the Titles Office. Nevertheless, the trial judge refused production, quoting from the principle in Hughes v Vargas set out above, and saying that this principle was applied by this Court in Marconi's Wireless Telegraph.
The same rule was applied by this Court in 1920, in O'Flaherty v McBride, finding that statutory restrictions upon the Court receiving documents from the Crown preserved the power of a Minister to object to production of documents on the ground of prejudice to the public service. The background to O'Flaherty is that in 1918, the Commonwealth had introduced a tax provision to prevent the Government from being compelled to produce tax returns in ordinary civil litigation. In the Second Reading Speech which introduced that provision, the Treasurer said that the view of the Government was "that no taxpayer's return should be produced to a Court unless a taxpayer's assessment is before the Court on appeal or the Department is suing for recovery of tax". The provision introduced, s 9(4) of the Income Tax Assessment Act 1915 (Cth), was described by Dixon CJ as giving protection to the officer against compulsion, and said by this Court to be "of great importance". It has been replicated numerous times since. It provided that:
"An officer shall not be required to produce in any Court any return, assessment, or notice of assessment, or to divulge or communicate to any court any matter or thing coming under his notice in the performance of his duties under this Act, except as may be necessary for the purpose of carrying into effect the provisions of this Act."
In O'Flaherty, this Court considered an appeal against the dismissal of an information preferred by the appellant, an officer of the Income Tax Department, against the respondent. The information had been dismissed because the appellant failed to produce a report made by him and provided to the Department. It was held the information should not have been dismissed. The Court was unable to apply the exception which permitted production to the Court where it was "necessary for the purpose of carrying into effect the provisions of this Act", because the Court considered that it could not order production of the report for the purpose of determining that necessity. The Court said:
"the statutory provision quoted does not weaken, and is not intended to weaken, the rule of common law that evidence of affairs of State is excluded when its admission would be against public policy. That rule in the present case operates to exclude the admission of the report - and, of course, all secondary evidence of its contents - and consequently operates so as to leave the Court unable to say whether its production is or is not 'necessary' for the purposes mentioned in sub-sec 4 of sec 9 of the Act."
The decline of the conclusive certificate
It is unnecessary to trace in detail the decisions that marked the decline in the conclusive nature of a certificate by a Minister or head of department. It suffices to say that the principle was so well established at Federation that, even by 1925 in Griffin v South Australia, the bold and creative counsel, Owen Dixon KC, did not even attempt to argue against the conclusive nature of the Minister's statement that the production of the documents was prejudicial to the public interest. Instead, on appeal from a decision that accepted as conclusive a statement by the Minister, Dixon KC argued that the Minister's statement that the class of document was State papers was not conclusive. He submitted that the Court could consider whether the documents were State papers if there was "positive evidence of mistake or misconception by the Minister of his duty".
The claim for production in that case was dismissed. Chief Justice Knox said, in a statement that was echoed by Isaacs J, Higgins J, and Rich J, that the rule in Beatson stood "unaffected by any of the later decisions to which we were referred". He explained that Marconi's Wireless Telegraph had recognised that the rule applied "in cases in which the Court is satisfied that the document in question is within that class [of State documents]". No extreme case had ever arisen in which the rule had been disregarded or modified. Justice Isaacs also said of the extreme case where the document, on its face, did not fall within the relevant class:
"If (which is barely conceivable) there should ever be so transparent a claim by a Minister of the Crown for privilege that the Court without seeking evidence or weighing it can perceive ex facie the impossibility of public prejudice, the Court may well for such an extreme possibility reserve an extreme power."
Only Starke J dissented in Griffin. His Honour considered that the commercial activities of Australian governments had become more and more extensive and the sphere of political and administrative action correspondingly wider. This was said to be a reason for submitting executive government "to the jurisdiction of the Courts" and imposing duties of "discovery and inspection of documents according to the ordinary rules of law and practice".
Despite the existence of the extreme possibility that courts might, after preliminary inquiry without examining the documents, conclude that certain documents did not fall within the class that entitled protection based upon a conclusive certificate, as late as 1929 courts refused to countenance that possibility even in the most stark cases. The English equivalent of the extreme circumstances in Spong v Spong was the 1929 decision in Ankin v London and North Eastern Railway Co. In that case, the English Court of Appeal upheld the refusal to disclose a mere notice of a rail accident sent by a railway company to the Minister. In doing so, Scrutton LJ reiterated that it was the practice of English courts to accept as conclusive a statement of a Minister even if "the Court may doubt whether any harm would be done by producing it". This case was later described by Lord Reid as "a good example of what happens if the courts abandon all control of this matter".
In 1931, the Privy Council in Robinson v State of South Australia [No 2], purporting to restate the view of Griffith CJ in Marconi's Wireless Telegraph, said that the Court has "always had in reserve the power to inquire into the nature of the document for which protection is sought, and to require some indication of the nature of the injury to the State which would follow its production". This decision confused the existence of a power to consider whether the document was in a class for which the certificate was conclusive with a general power to inquire in any case. In either event, however, it appears that in the many cases litigated over half a century, such a general power had never been exercised.
The decision in Robinson was not followed by the House of Lords in Duncan v Cammell, Laird & Co. Viscount Simon LC, with whom the six other Lords of Appeal in Ordinary agreed, said that he could not agree with the view. He added that the Privy Council had been mistaken in regarding a rule permitting a court to inspect the document as having any application to the subject matter. Five years after the decision in Duncan, when giving the Second Reading Speech of the English legislation that removed any Crown privilege to resist discovery, but preserved the common law rule concerning a conclusive certificate from a Minister, the Lord Chancellor said:
"I am quite satisfied in my own mind that the Crown must have the right to say that certain documents or sometimes certain classes of documents shall not be produced either in litigation to which the Crown are a party or litigation between two ordinary parties. ... I say further, having discussed this matter with the Judges, that I am quite satisfied that the Judges would think it most undesirable that they should have the task of deciding whether documents should or should not be protected from disclosure. Very often you cannot tell, merely by looking at a document, whether it should or should not be protected. You must know all the circumstances which led up to the document, and for a Judge to inform himself of all those matters might mean that he would have to be closeted, as it were, with one party to the litigation without the other side being there."
However, in 1968, the decision in Duncan was overruled in Conway v Rimmer. In that case, the House of Lords unanimously held that the decision concerning disclosure was always ultimately one for the court. Lord Reid said, quoting from Vinson CJ in United States v Reynolds, that "[j]udicial control over the evidence in a case cannot be abdicated to the caprice of executive officers". In Sankey v Whitlam, the "new light" from Conway v Rimmer was adopted by this Court. In the words of Gibbs ACJ in Sankey, it became "in all cases the duty of the court, and not the privilege of the executive government, to decide whether a document will be produced or may be withheld". From 1978, therefore, Australian courts no longer treated as conclusive a certificate from a Minister or head of department that production of a State paper would be prejudicial to the public service. Nevertheless, the unquestioned existence of this rule at the time of Federation, and for decades beyond, remains an extremely important matter of historical context. That historical context demonstrates the difficulty for the submissions of the plaintiff and the applicant that the Constitution contains an implication to which content should be given to preclude legislation that achieves a similar effect.
F. Section 503A of the Migration Act and the minimum content of judicial review
As I explained in the introduction to these reasons, s 503A of the Migration Act is a measure designed to ensure confidentiality over particular information communicated to an authorised migration officer by particular law enforcement agencies that insist that the information be treated as confidential. To do this, s 503A(2) regulates the material before a court or tribunal. The technique of regulating the material before a court or tribunal in order to achieve a desired policy outcome is not new. As the Attorney-General of the Commonwealth and the Minister pointed out, a majority of this Court upheld legislation to this effect in Nicholas v The Queen. The regulation in that case involved legislation that required a court to disregard various facts concerning the involvement of a law enforcement officer in the commission of an offence. As Gummow J said, the legislative scheme was designed "to strike a balance between competing interests and to give effect with respect to these prosecutions to a perception of the public interest which differs from that expressed in the common law in Australia. That is a matter for the Parliament."
The plaintiff and the applicant submitted that s 503A of the Migration Act contravened an entrenched minimum provision of judicial review because by "withholding" relevant information from a court the statutory scheme "can stymie judicial review". In other words, the alleged vice of s 503A was that, in some cases, a ground of judicial review that might have succeeded would not succeed. The plaintiff and the applicant pointed to the applicant's case, where, it was submitted, the Minister's conclusion that the applicant failed the "character test" was based entirely on protected information. Perhaps ironically, both the plaintiff and the applicant nevertheless also alleged that the Minister could not have been satisfied on the evidence before him, which included the protected information, that cancellation of their visas was in the national interest. Indeed, even in the most extreme case, where the only information relied upon by the Minister is the protected information, an application for judicial review would still potentially have significant content, including in relation to review of whether the conditions, including confidentiality, are satisfied for the operation of s 503A(2). Section 503A does not, therefore, generally prevent judicial review or oust the jurisdiction in s 75(v) of the Constitution. Its effect, instead, is to make a ground of judicial review, such as unreasonableness, more difficult to establish in some cases if there is protected information involved. As I explain below, on one view this might be less extreme than legislation which makes one ground of judicial review (eg procedural unfairness) impossible in every case.
The first reason why s 503A(2) is not invalid
As I have explained above, the first reason why s 503A(2) is not invalid is that the submission by the plaintiff and the applicant in relation to s 503A(2) is ahistorical in two senses. One sense in which the submission is ahistorical is that s 503A(2) of the Migration Act removes far less of the record than the form of legislation that existed for well over a century before Federation, which was standardised in 1848 by the Summary Jurisdiction Act. There is no suggestion that s 75(v) of the Constitution, or any other constitutional provision, involved a break from this longstanding legislative approach by the English Parliament so that English legislation enacted more than a century and a half earlier than Federation, and persisting since, would have become invalid in Australia. In any event, as I have explained, the Convention Debates would contradict such a suggestion.
The second sense in which the submission is ahistorical is that s 503A(2) achieves by legislation a very similar effect to that which a certificate from the Minister would have achieved in any litigation in the nineteenth century. In the Second Reading Speech of the Bill that introduced s 503A, Senator Kemp said that the Bill increased the level of protection for criminal intelligence and related information that was "critical to assessing the criminal background or associations of non-citizen visa applicants and visa holders". He explained that it had been "difficult for the Department to use such information in making character decisions because its disclosure might be threatened". The difficulty arising from threatened disclosure was that "Australian and international law enforcement agencies are reluctant to provide sensitive information unless they are sure that both the information and its sources can be protected".
A review of the nineteenth century decisions from Beatson onwards shows that any nineteenth century court, presented with a certificate from the Minister in relation to information such as that caught by s 503A(2), would have accepted the conclusive nature of the certificate. Even if it could have been alleged that s 503A(2) information fell within an "extreme case" of information that might not be a State paper, and therefore might not be entitled to the benefit of the conclusive certificate, a recital by the Minister of the matters described in the Second Reading Speech by Senator Kemp would have satisfied any court. No nineteenth century court would ever have considered that it had the power to examine the information if the Minister had deposed that it was provided by a domestic agency or foreign law enforcement agency on the condition that it be treated as confidential. As the Lord Chancellor said in Smith v The East India Company, in relation to a mere commercial transaction but without examining the documents, in words that could easily be applied with even greater force to s 503A(2):
"it is quite obvious that public policy requires, and, looking to the Act of Parliament, it is quite clear that the Legislature intended, that the most unreserved communication should take place ... [I]t is also quite obvious that if, at the suit of a particular individual, those communications should be subject to be produced in a Court of justice, the effect of that would be to restrain the freedom of the communications, and to render them more cautious, guarded, and reserved. I think, therefore, that these communications come within that class of official communications which are privileged, inasmuch as they cannot be subject to be communicated, without infringing the policy of the Act of Parliament and without injury to the public interests."
There are two relevant respects in which the operation of s 503A(2) is, in fact, more liberal than the operation of the nineteenth century conclusive certificates. The first of those is that in the nineteenth century, the prevailing view was that even if both parties supported production, the court was obliged to refuse the production of a document, or not to permit production of a document, where production would be injurious to the public service. In contrast, s 503A(3) permits the Minister to disclose the information to a court or tribunal after consultation with the agency that provided the information on the condition of confidentiality. The second respect in which s 503A(2) is more liberal than the common law conclusive certificate is that a nineteenth century court would only consider whether a document was a "State paper" in a hypothetical "extreme case" that never arose in half a century before Federation. But the conditions under which s 503A(2) applies must be considered in every case. The preconditions to the application of s 503A(2) are that (i) information is communicated to an authorised migration officer; (ii) the information is communicated by a gazetted agency; (iii) the information is communicated on the condition that it be treated as confidential information; and (iv) the information is relevant to an exercise of a power under s 501, s 501A, s 501B, or s 501C of the Migration Act. If s 503A(2) were to be translated in nineteenth century terms, it would be as though the court would scrutinise every case in which the State alleged that a document was a State paper so that disclosure would be refused, rather than reserving this possibility only for hypothetical "extreme" cases that never occurred.
The second reason why s 503A(2) is not invalid
The second reason why s 503A(2) is not invalid is that such a conclusion would create inconsistencies with this Court's constitutional jurisprudence. Section 503A(2) can be contrasted with other privative clauses since Hickman that have constrained judicial review to a greater extent but which have been upheld as valid by this Court. An example contrasting with s 503A(2) is the restrictive legislation that was held to be valid by this Court in 2010 in Plaintiff M61/2010E. As I explained above, that case considered provisions of the Migration Act that had the effect that an "offshore entry person" could not apply for a visa unless the Minister permitted an application. In some circumstances, the Minister could decide that it was in the public interest to grant a visa, whether or not an application had been made. The Minister's powers were expressly constrained only by the Minister's consideration of the public interest. And the Minister was not obliged even to consider the exercise of these powers. This meant, as the Court found, that mandamus would not issue to compel the Minister to consider exercising the power. Since mandamus would not issue to compel any reconsideration, there was no utility in granting certiorari to quash recommendations made by the Minister after consideration.
As Dr Burton Crawford observed, the effect of the legislation in Plaintiff M61/2010E was to "knock out" the remedy of mandamus guaranteed by s 75(v) of the Constitution. More precisely, the power to award mandamus was unaffected but there was no content upon which that power could operate. Nevertheless, in Plaintiff M61/2010E this Court accepted the submission of the Solicitor-General of the Commonwealth that the decision in Plaintiff S157/2002 did not require that the exercise of a statutory power in every case be accompanied by a duty to consider the exercise of the power so as to give content upon which the power to order mandamus could operate. As the Court explained, s 46A did not clash either with s 75(v) "or with its place or purpose in the Constitution". Nor did an abstract appeal to "the rule of law" lead to a different conclusion.
The third potential reason why s 503A(2) is not invalid
The powers to which the s 503A(2) restriction on disclosure applies are powers under (i) s 501, (ii) s 501A, (iii) s 501B, or (iv) s 501C. In very broad terms, these powers are concerned, respectively, with (i) the Minister's refusal to grant a visa, or cancellation of a visa; (ii) the Minister setting aside a "non-adverse" decision by a delegate or by the Administrative Appeals Tribunal and substituting his adverse decision; (iii) the Minister setting aside an "adverse" decision by a delegate and substituting his adverse decision; and (iv) revocation of a decision under s 501(3) or s 501A(3) where, following submissions from the person, the Minister is satisfied that the person passes the character test.
Suppose, to adapt a hypothetical example given by Dr Kirk, that the Migration Act provided, clearly and unambiguously, that in some circumstances each of the four powers above need not be exercised by the Minister reasonably or rationally so that, in those circumstances, the Migration Act purported to exclude entirely judicial review based on unreasonableness, although leaving intact other grounds of jurisdictional error, including jurisdictional preconditions for the exercise of power. If such a scheme were to contravene an implied constitutional constraint on legislative power, the same constraint might also apply, indeed might apply with greater force, to a scheme which purported to remove the duty to observe procedural fairness. Yet many cases have assumed that there is no implied constitutional restraint upon legislation which provides that a person is not entitled to procedural fairness. The entitlement, and extent of the entitlement, to review for unreasonableness, like review on the ground of procedural fairness, has been repeatedly held to arise as a matter of implication from the statute itself, rather than as an imposed ground of review by implication from the Constitution.
It may not have been impossible, purely as a matter of logic, for a term to have been included in the Constitution which expressly constrained legislation from reducing judicial review. For instance, an express term could have been inserted in the Constitution which prohibited Parliament from legislating in such a manner as to reduce the minimum content of judicial review that would ordinarily exist on a ground such as unreasonableness based upon the usual subject matter, scope, and purposes of legislation. But an express term stated in this way would invite many questions concerning the manner and scope of its operation. That may have been a powerful reason not to imbue an implied term with such uncertain content. However, in the absence of any submissions on this point it is unnecessary to decide it.
G. Whether the Minister could have been satisfied that cancellation of the visas was in the national interest
The remaining issue in each Special Case raised by the plaintiff and the applicant is whether the power of the Minister to cancel a visa under s 501(3) of the Migration Act could reasonably have been exercised. The Minister's power arises if (i) the Minister reasonably suspects that the person does not pass the character test, and (ii) the Minister is satisfied that the cancellation is in the national interest.
As to the first of these conditions, in cancelling the visa of each of the plaintiff and the applicant, the Minister relied upon the character test in s 501(6)(b). That paragraph provides that a person does not pass the character test if:
"the Minister reasonably suspects:
(i) that the person has been or is a member of a group or organisation, or has had or has an association with a group, organisation or person; and
(ii) that the group, organisation or person has been or is involved in criminal conduct".
A separate submission by each of the plaintiff and the applicant was that the Minister could not be satisfied that cancellation of either visa was in the national interest without making findings about either or both of (i) their knowledge of, opinion of, support for, or participation in the suspected criminal conduct of the Rebels Outlaw Motorcycle Gang, and (ii) how cancellation of their visas would disrupt, disable and dismantle the criminal activities of outlaw motorcycle gangs.
In relation to the plaintiff, the Minister did not, as the plaintiff submitted, "leap uncritically from suspicion of membership to a conclusion that visa cancellation '[was] in the national interest, in that it [would] contribute to the national effort to disrupt, disable and dismantle the activities of Outlaw Motorcycle Gangs'". In particular, the Minister relied upon numerous matters for his conclusion that cancellation was in the national interest notwithstanding the plaintiff's period of residence in Australia and his ties to Australia. Those matters included: (i) information that is protected under s 503A of the Migration Act; (ii) the National Security Strategy and National Taskforce Operation Morpheus established by the Australian Crime Commission's Serious and Organised Crime Coordination Committee; (iii) the establishment of the Attero National Taskforce in 2012, the purpose of which was to disrupt, disable and dismantle the criminal activities of the Rebels Outlaw Motorcycle Gang, considered to be one of Australia's highest risk criminal threats; (iv) open source materials and submissions by the plaintiff's legal representative that identify the plaintiff as having been or being a member of the Rebels Outlaw Motorcycle Gang; and (v) a history of the plaintiff's criminal convictions dating back to 11 January 1982, including sentences for terms of imprisonment. The offences included convictions for stealing, breaking and entering, unlawfully damaging and destroying property, possession of various prohibited substances, and multiple convictions for firearm and weapons offences. The plaintiff reoffended even after he was sent a formal warning letter dated 4 July 2011 informing him that any further offending may result in his visa being cancelled.
In relation to the applicant, the matters relied upon by the Minister in addition to the protected information included some of the same matters considered in relation to the plaintiff, namely (ii) and (iii) above. It is possible that much weight might have been placed by the Minister on the information that is protected from disclosure under s 503A of the Migration Act, and that the lack of production of this information makes it more difficult for the applicant to establish that the Minister could not have been satisfied that cancellation of the applicant's visa was in the national interest. Nevertheless, the submissions of the applicant essentially invited this Court to conduct a fresh assessment of the merits of whether the Minister could be satisfied that cancellation of the applicant's visa was in the national interest, which is "largely a political question". Those submissions could not have succeeded even if the information had been disclosed.
This ground must also be dismissed.
Conclusion
The submissions of the plaintiff and the applicant on these applications were directed only to whether s 503A(2) invalidly failed to comply with an implied minimum provision of judicial review, which was said to be required to be provided by this Court. It was understandable that the plaintiff and the applicant confined their cases in this way because the primary textual source for the implication upon which they relied, s 75(v) of the Constitution, is a provision concerned only with the original jurisdiction of this Court. The plaintiff and the applicant made no submissions about the manner in which such an implication would extend also to constrain Parliament's power in relation to the exercise of jurisdiction defined under s 77(i) of the Constitution by reference to s 75, or whether the restriction was based upon a wider structural constitutional implication that operated identically in relation to other federal courts or to Supreme Courts. The Commonwealth and the Minister reserved their position to respond to such submissions if, and when, s 503A(2) were to be applied in the Federal Court. The conclusion that I have reached in relation to these applications is a further reason why it is unnecessary to consider this issue.
It is also unnecessary to consider whether, if s 503A(2) were invalid, it could be "read down" and, if so, the consequence of the existence of a number of possible ways in which it could be read down.
I would answer the questions of law in the Special Case which were reserved for consideration of the Full Court in relation to the plaintiff as follows, with identical answers in relation to the applicant: