Tuesday, 18 May 1999
Administrative Law — Aspects of Pre-Millennial Judicial Review
The 1999 Blackburn Lecture, delivered by The Hon John Lehane, judge of the Federal Court of Australia.
In his 1989 Blackburn Lecture Sir Anthony Mason surveyed the operation of the federal system of administrative law introduced during the 1970s and some of the principles underlying it. Sir Anthony was, perhaps, uniquely qualified to undertake that survey. He had been intimately involved in the process leading to the introduction of that system with its four elements: the Administrative Review Council; review by the Ombudsman; judicial review under the Administrative Decisions (Judicial Review) Act 1977 (Cth) (the ADJR Act); and merits review by the Administrative Appeals Tribunal (the AAT). He had also, of course, as a Justice and then Chief Justice of the High Court, delivered leading judgments in a series of important cases on several aspects of administrative law. Two matters stand out in Sir Anthony’s survey. One is the significance of the provisions (ADJR Act s 13 and Administrative Appeals Tribunal Act 1975 (Cth) s 28) requiring a decision maker, at the request of a person entitled to seek review of a decision, to provide reasons for it: not merely reasons but:
“…a statement in writing setting out the findings on material questions of fact, referring to the evidence or other material on which those findings were based and giving the reasons for the decision.”
Nothing need be added about the importance of those provisions; and whatever initial doubts or controversy there may have been about them, they seem now to be both effective and firmly entrenched. The other matter is the role of the AAT in reviewing the merits of administrative decisions — including policy — and its justification against doubts and objections some of which, perhaps in a muted form, persist in some quarters today. But in whatever precise form the AAT continues into the new millennium, there seems to be no doubt that continue it will. And in several of the States and Territories there is legislation establishing, general merits appeal tribunals, or proposals for their establishment, in some cases following closely, and in others departing in interesting ways from, the Commonwealth model: in the former category, the Administrative Appeals Tribunal Act 1998 (ACT); in the latter, the Victorian Civil and Administrative Tribunal Act 1998 (Vic) and, particularly, the Administrative Decisions Tribunal Act 1997 (NSW).
By way of stark contrast, I am a relative newcomer to administrative law. Reflecting my particular- and limited- experience, as a member of the Federal Court, I propose to concentrate on judicial review: particularly, to consider some aspects of the development of the principles which the courts have applied in considering applications for review under the ADJR Act or by way of prerogative relief or equitable remedies and to discuss, by way of a comparison which I hope may be useful, the particular form of judicial review administered by the Federal Court under Pt 8 of the Migration Act 1958 (Cth). First, some matters to do with aspects of judicial review generally.
Standing
Aronson and Dyer (Judicial Review of Administrative Action 1996 p 707) are, I suggest, clearly right in suggesting that there is taking place, if it has not already occurred, a convergence between the “special interest” test developed in the injunction and declaration cases and the standing requirements in applications for prerogative relief. Aronson and Dyer refer to the authorities and nothing is to be gained by repeating or paraphrasing what they say. But a gloss may be added: there is also substantial convergence between the “special interest” test and, the requirement for example, that an applicant for judicial review be a “person who is aggrieved”, in s 5(1) and s 6(1) of the ADJR Act. That, in a sense, is surprising, because the origins and history of the two tests are quite different.
The “special interest” test is, of course, a refinement of the second limb of the so-called rule in Boyce v Paddington Borough Council an individual has, in the absence of the Attorney-General, standing to seek equitable relief in relief in respect of an infringement of a public right if the individual suffers “special damage peculiar to himself”[1] as a result of the infringement. What is now required is a “special interest in the subject matter of the action”[2] which is not merely the interest of an ordinary member of the public in upholding the law, or a “mere intellectual or emotional concern”[3] The plaintiff “must show that he has been specially affected, that is, in comparison with the public at large he has been affected to a substantially greater degree or in a significantly different manner. It is not necessary to show that the plaintiff is uniquely affected; there may be some others whose interests may be affected in like manner.”[4] The principle is flexible and “the nature and subject matter of the litigation will dictate what amounts to a special interest.”[5] Those principles were recently reaffirmed by the High Court in Bateman’s Bay Local Aboriginal Land Council v Aboriginal Community Benefit Fund Pty Ltd:[6] and in the joint judgment of Gaudron, Gummow and Kirby JJ there is a suggestion, not pursued because it had not been put forward in argument, that a better approach might be “…to dispose of any question of standing to seek injunctive or other equitable relief by asking whether the proceedings should be dismissed because the right or interest of the plaintiff was insufficient to support a justiciable controversy, or should be stayed as otherwise oppressive, vexatious or an abuse of process.”[7]
Standing is to be tested under the ADJR Act simply by construing and applying the test that an applicant for judicial review must be a “person who is aggrieved.” The convergence to which I refer appears plainly enough in one of the formulations frequently cited, that of Lockhart J in Right to Life Association (NSW) Inc v Secretary, Department of Human Services and Health:[8]
“The meaning of ’a person aggrieved’ is not in encased in any technical rules; much depends upon the nature of the particular decision and the extent to which the interest of the applicant rises above that of an ordinary member of the public.
The applicant’s interest must not be remote, indirect or fanciful. The interest must be above that of an ordinary member of the public and must not be that of a mere meddler or busybody…
Plainly the applicant need not have a legal, financial or proprietary interest in the subject matter of the proceeding. The applicant must establish that he is a person who has a complaint or grievance which he will suffer as a consequence of the decision beyond that of an ordinary member of the public.”
It is common, where standing is in controversy, to find injunction cases and ADJR Act cases cited indiscriminately where ADJR Act standing is in issue[9] and, though perhaps less frequently, where an injunction or a declaration is sought.[10]
One particular aspect of the convergence may be mentioned. Where what is in question is standing to seek review of a decision made under a statute, in accordance with a test such as the ADJR Act requirements, much will depend on the scope and the purpose or object of the statute in question. So the substantial commercial interest of an “innovator” drug company likely to be affected by a decision to register a “generic” drug under the Therapeutic Goods Act 1989 (Cth) was held insufficient to afford standing where the object of the legislation was “the timely availability of therapeutic goods, after evaluation by an expert body and pursuant to a ‘complex and delicate administrative scheme’”[11] and having regard to the way in which the legislation provided for the process of registration. Dicta in the joint judgment in the Bateman’s Bay case[12] indicate that those considerations are equally relevant where the relief sought is an injunction against a decision maker.
One final point on standing. Discussion of the Boyce principle can easily become confused when it merges with a conceptually quite different subject, the circumstances in which equity will lend its aid in the enforcement of a statute which imposes a penalty for conduct which it proscribes. There, as the joint judgment in Bateman’s Bay[13] points out, “special considerations would apply, with as well as without the fiat”.
“Decision of an administrative character made … under an enactment”
This is a matter peculiar to review under the ADJR Act; the words quoted are the essential elements of the definition of the class of decisions review of which (or of conduct leading to the making of which) is available under the ADJR Act. First, “decision”. A decision is reviewable if it a “substantive”, not merely a “procedural” determination. “That will generally, but not always, entail a decision which is final or operative and determinative, at least in a practical sense, of the issue of fact falling for consideration”[14]
Generally, I think, it may be fairly be said that this test rarely leads to difficulty. Borderline cases sometimes arise, in which there is room for differences of view. For example, in Hutchins v Commissioner of Taxation[15] the question was whether a decision of the Deputy of Commissioner of Taxation to vote against a proposal put to a meeting of creditors under Pt X of the Bankruptcy Act 1966 (Cth) was reviewable. The Full Court held that it was not. Black CJ and Lockhart J each held that the decision to vote against the proposal was not substantive, because it determined nothing but was merely a step along the way to a resolution of the creditors who attended the meeting, in person or by proxy [16] Spender J, however,[17] took the view that even if the decision was not crucial (because the result depended on how others might vote) it was, nevertheless, determinative of the matter before the Deputy Commissioner: and that, in his view, was enough.
A decision, to be reviewable under the ADJR Act, must be “of an administrative character”. It is, therefore, neither legislative nor judicial.[18] The distinction between making law and administering or executing law[19] is not always easy to draw. Indeed, as the decision of Gummow J in Queensland Medical Laboratory v Blewett[20] illustrates, a decision to exercise a particular power one way may have an administrative character whereas, if the decision is exercised another way, it is to be characterised as legislative. The decision under consideration in Queensland Medical Laboratory was one made by the Minister under the Health Insurance Act 1973 (Cth). The effect of the decision, which the Minister was empowered to make by a provision of the Act, was to substitute for a table set out in a schedule to the Act a new table of fees by reference to which Medicare benefits were to be calculated and paid.
Gummow J held that, although it was true to say that if the Minister decided not to make a determination, he was executing or administering the law and it might well be that the decision would have been of an administrative character, nevertheless in deciding to substitute a new table for that in the schedule he was exercising a power delegated to him by the Parliament to act legislatively. But what Queensland Medical Laboratory principally illustrates is the undoubted proposition that tests generally stated do not always provide a clear answer in a particular case and that context is all important.[21] A recent illustration is the decision of the Full Court of the Federal Court in Federal Airports Corporation v Aerolineas Argentineas[22] In that case the Federal Airports Corporation, exercising an express power under Commonwealth legislation, purported to determine a schedule of particular charges payable by airlines whose aircraft landed at airports operated by the Corporation. It might be said that the decision changed the content of the law because it effected a general imposition of charges not otherwise payable, the charges being recoverable by force of the legislation; and unpaid charges attracted a statutory penalty. But, it was held, in context those circumstances did not give the determination a legislative character: the charges were in the nature of fees (calculated so as to recover costs) for services provided by the Corporation in the course of carrying on an enterprise which, though a government enterprise, was commercial. “It is a determination made by the Corporation in the course and for the purpose of its commercial operation: that is, in the execution or administration of the FAC Act”.[23]
The decision, if it is to be reviewable, must be one made “under an enactment”. “Enactment” includes an instrument, including rules, regulations or by-laws made under an Act.[24] Because it is relevant to a theme to which I shall return, a useful starting point is the following passage in the joint judgment of Lockhart and Morling JJ in Chittick v Ackland:[25]
“Questions of construction of enabling statutes like the Judicial Review Act are rarely solved at the one time. In the continuing solution of such problems usually there is a history of development and sometimes of change. Bearing this in mind in our opinion this is not the time to seek to expound definitively the meaning and ambit of the expression ‘instrument … made under such an Act…’. This must be determined progressively in each case as particular questions arise.”
A decision made “under” an enactment is one “for which provision is made by or under a statute”[26] That is, it is one required or authorised by the statute, including impliedly authorised.[27] It is a decision which is given force or effect by the statute or by a principle of law which applies to it.[28] So if a body constituted under an Act enters into a contract of employment under a general power to appoint employees and then, acting under an express provision of a contract with a particular employee, dismisses that employee, the decision to dismiss is likely to be characterised not as one made under the statute but as made, relevantly, solely under the contract of employment.[29] More broadly, Hutchins suggests that there may be cases where a decision, the sole source of authority for which is to be found in a general power of administration conferred by a statute, is not one given force or effect by the statute or by a principle of law applicable to it.[30]
Later cases have confirmed the tentative conclusion of Lockhart and Morling JJ in Chittick[31] that “…to qualify as an instrument for the purposes of the Judicial Review Act [scan instrument made under an enactment] the document must be of such a kind that it has the capacity to affect legal rights and obligations”. It must additionally, of course, have been made under, that is, under the authority of, the statute. So in Chittick itself a decision to dismiss an employee under conditions of employment determined under a statutory provision which gave them binding force, was held to be reviewable; similarly, a decision refusing an application for promotion, under an instrument determining terms and conditions of employment given, by statute, binding effect (and which might be changed, so as to bind an employee without the employee’s consent).[32] But a decision to refuse promotion, under the terms of a document which described the way in which promotion applications would be dealt with, promulgated to employees but which (apart from the contract) had no binding force, was held in Lewins not to be reviewable. The employer was the Australian National University. It was a body constituted by an Act; the Act gave it power to employ, and it had employed the applicant in exercise of that power. But the decision on his promotion application was not made, relevantly, under the statutory power itself or under an instrument which (by virtue of the statute) gave such decisions binding force.
Grounds of review
This is not the occasion for an attempt to survey in detail the grounds on which judicial review may be available. The grounds were not novel when the ADJR Act became law. In several respects they have been developed, refined and explained by a series of decisions of the High Court. Most notable, perhaps, has been the development of the law concerning procedural fairness, starting in Kioa, in reliance, to a considerable extent, on the change brought about by the statutory requirement to give reasons. So, from Kioa, Haoucher v Minister for Immigration and Ethnic Affairs[33] and Annetts v McCann[34] comes the proposition that:
“It can now be taken as settled that, when a statute confers power upon a public official to destroy, defeat or prejudice a person’s rights, interests or legitimate expectations, the rules of natural justice regulate the exercise of that power unless they are excluded by plain words or necessary intendment.”[35]
But it may be different if, rather than directly affecting peculiarly the rights or interests of a limited number of individuals, a decision (whether of an administrative or legislative character) affects indiscriminately the interests of the members of a substantial section, or substantial sections, of the public. A decision to impose a rate is an example given by Mason J in Kioa at 584.
Blewett provides a further example. Gummow J said this:[36]
“I believe there is much to be said for the view that the making of a recommendation by the committee and the decision of the Minister to make a determination in accordance with the recommendation of the committee, did not affect the rights, interests and expectations of pathologists, other medical practitioners and patients in a sufficiently individual direct and immediate way as to attract with regard to persons in these groups the duty to act fairly…”
Another example may be found in the decisions leading to the reopening of the east-west runway at Sydney Airport, leading to an abatement in noise levels suffered by certain large segments of the public and an increase in noise levels suffered by others.[37]
Similarly, the “improper exercise” ground has been considered by the High Court: particularly in Minister for Aboriginal Affairs v Peko-Wallsend Ltd,[38] the process of construction by which, where a decision is made under a statute, the matters which the decision-maker must take into account are to be ascertained, as well as, where the decision is to be made by a Minister, the “due allowance ... to be made for the taking into account of broader policy considerations which may be relevant to the exercise of a ministerial discretion.”[39] The recent decision of the High Court in Minister for Immigration and Multicultural Affairs v Eshetu[40] is authority which both affirms a reasonably strict view of the ground of so-called “Wedensbury unreasonableness”[41] and suggests, particularly in the judgment of Gummow J, new (or, perhaps, rediscovered) approaches to the doctrinal basis of “unreasonableness” and its place in the law of the judicial review. Notably, there is in the passages in Eshetu to which I have referred, associated with the discussion of Wedensbury unreasonableness, an insistence that judicial review not trespass beyond its proper field and into merits review. That exhortation to judicial restraint by no means stands alone. Particularly explicit examples may be found in Minister for Immigration and Ethnic Affairs v Wu[42] and Minister for Immigration and Ethnic Affairs v Guo.[43]
It may be that the development of doctrine by the High Court during the 1980s and 1990s has been characterised by a degree of expansion followed by a contraction of the boundaries.[44] But whether it is characterised in those or other terms, what is undoubtedly true is that the development or exploration of the grounds of judicial review and the evolving process of construing the ADJR Act have proceeded in an orthodox, incremental and orderly way; unsurprisingly and largely uncontroversially. A similar process may be seen in the cases about other aspects of administrative law.
For example, in Federal Airports Corporation v Aerolineas Argentineas[45] the Full Court held that in the Australian legislative context (particularly the express preservation by s 10 of the ADJR Act of other mechanisms of judicial review), the enactment of the ADJR Act did not mean that “collateral challenge” (in that case by way of an action for repayment of charges paid under an allegedly valid determination) was excluded. It was held that collateral challenge was possible, no more and no less after the enactment of the ADJR Act than previously and an action to recover money paid under an ultra vires determination could proceed whether or not, concurrently or previously, the determination was set aside in judicial review proceedings. The Court did not, however, attempt an examination of boundaries of “ultra vires” or the extent to which decisions which might be set aside on other grounds could be challenged collaterally: those interesting and difficult questions were left for future decision in cases directly raising them. Another example of a similar process is, perhaps, the series of recent decisions on the identification of “jurisdictional facts.[46]
Judicial review under the Migration Act
I shall start with some statistics. Of 819 cases commenced between the beginning of January and the end of April this year in the Sydney Registry of the Federal Court, 156 were migration cases. During the same period in 1998, migration cases numbered 168 of the total of 887 proceedings commenced, in Sydney, in the original jurisdiction. For the same period in 1997, the figures were 96 out of 851 and, in 1996, 69 out of 538. I have taken the first four months of each year simply so that, including this year in the statistics, I can compare like with like. The figures sufficiently show that there was a substantial increase, occurring during 1997 and particularly in 1998, in the number of migration cases commenced and that there is, as yet, no sign that the flood is abating to any significant extent. (I should confess that I suspect — I do not know — that the 1996 figures may be an underestimate: there is a seperate “ADJR” category, and it may be that that category for that year included some migration cases commenced under the ADJR Act before the Migration Reform Act 1992 — of which more later- took effect).
The same pattern has not been evident in Melbourne, where, if anything, there has since 1997 been a slight diminution in the number of migration cases commenced: in the first four months of this year, the figures were 45 out of 577. Generally speaking, proportionately far fewer migration cases are commenced in the other registries, though this year there has been what might be described as a “spike” in Queensland: 12 out of a total of 295.
Over the same period, there has been a substantial increase in the number of Full Court appeals in migration cases. I shall take again the first four months of each year starting in 1996. The total number commenced in all registries was 6 in 1996, 13 in 1997, 18 in 1998 and 28 in 1999. Of those, the number commenced in New South Wales was 2 in 1996, 7 in 1997, 9 in 1998 and 20 in 1999. Those 20 represent just under one half of the 48 appeals commenced in New South Wales this year, up to 30 April1999. Nationally, the 28 migration appeals during the same period were 28 of a total of 115.
One further set of comparative figures may be useful. During the first four months of this year, the statistics record (somewhat to my surprise) that no appeals from the AAT were commenced in New South Wales; ten applications under the ADJR Act were commenced and one under the Judiciary Act 1903. That contrasts with 156 migration cases. In the first four months of 1998 nine ADJR Act cases were commenced and two under the Judiciary Act; that contrasts with 168 migration cases. I cannot provide a subdivision, by category, of the migration cases. What can be said, however, with complete confidence is that overwhelmingly the majority of them, in each year (and probably more overwhelmingly with each passing year), are applications to review decisions of the Refugee Review Tribunal affirming or, far less frequently, reversing a primary decision to refuse a protection visa. I propose to discuss some aspects of those cases. They are interesting principally, I think, because they represent a spate of litigation involving difficult law, usually evidence which provides a less than satisfactory foundation for findings of fact but suggests at least real hardship or deprivation and a distinctive, and in some respects novel, system of judicial review. That is a potent mixture and it is perhaps not surprising that there have been difficulties and, occasionally, controversy.
The legal context is to be found in a combination of domestic and international law. The Migration Act 1958 (Cth) provides, in Div 3 of Pt 2 for the issue of visas to non-citizens. Broadly, provision is made for the prescription, by regulation, of classes of visas. Provision is also made directly, by the Act, for certain particular classes of visas, among them protection visas. Section 36 provides:
“36 (1) There is a class of visas to be known as protection visas.
(2) A criterion for a protection visa is that the applicant for the visa is a non-citizen in Australia to whom Australia has protection obligations under the Refugees Convention as amended by the Refugees Protocol.”
Section 39 authorises regulations empowering the Minister, in effect, to limit the number of visas (other than protection visas) of any class which may be issued in a particular financial year. Section 65 provides that the Minister, after considering a valid application for a visa, must grant it if satisfied that the criteria specified for the visa have been satisfied; if not so satisfied the Minister is to refuse to grant the visa.
For this purpose the criteria include any limitation on the numbers of visas in a particular class fixed for the year in which the application is considered.
The Migration Regulations provide for a class of Refugee and Humanitarian (Migrant) Visas, the primary criteria of which are that, in broad terms, the applicant is subject to persecution or to substantial discrimination, amounting to gross violation of human rights, in the applicant’s home country. But visas within that class are available only to applicants who are outside Australia and it is open to the Minister to restrict the number of visas in that class to be issued in any financial year. By contrast, an applicant for a protection visa must be in Australia; the “quota” provision does not apply; and the primary criterion to be satisfied is defined by reference to Australia’s obligations under the Refugees Convention as amended by the Refugees Protocol: that is, the 1951 Convention relating to the Status of Refugees as amended by the 1967 Protocol relating to the Status of Refugees, to each of which Australia is a party. Thus the legislation (including the Regulations) gives effect to Australia’s international legal obligations.
Part 8 of the Migration Act provides what is, with a few exceptions, a code in relation to review by the Federal Court of decisions under the Migration Act.[47] Section 475 lists a number of categories of decisions which are “judicially-reviewable decisions” and certain others which are not. Section 486 gives the Federal Court exclusive jurisdiction (apart from the jurisdiction of the High Court under s 75 of the Constitution) with respect to judicially-reviewable decisions; and s 485 excludes all sources of jurisdiction in the Federal Court apart from Pt 8 itself and remitter under s 44 of the Judiciary Act; but, where a matter is remitted under s 44, the Federal Court’s powers are limited to those which it would have had if the remitted matter had been the subject of an application under Pt 8. In Abebe v The Commonwealth[48] the High Court, by majority, upheld the validity of Pt 8. There is, of course, a consequence the precise dimensions of which are yet unascertained, though the Court’s decision in Eshetu may be the beginning of the process of ascertainment: the High Court appears to have under s 75 of the Constitution in some respects wider powers of judicial review in relation to judicially-reviewable decisions than the Federal Court has; wider powers which, because of s 486, the Federal Court cannot exercise on remitter under s 44 of the Judiciary Act.
The permissible grounds of review by the Federal Court of a judicially-reviewable decision are set out in s 476. They are in familiar terms but are in several respects more limited than their ADJR Act counterparts. I shall not quote them in full, but the two most commonly relied on should be noted. They are to be found ins 476(1)(a) and (e):
“(a) that procedures which were required by this Act or the regulations to be observed in connection with the making of the decision were not observed; …
(e) that the decision involved an error of law, being an error involving an incorrect interpretation of the applicable law or an incorrect application of the law to the facts as found by the person who made the decision, whether or not the error appears on the record of the decision; …”
Subsection (2) provides that a breach of the rules of natural justice in connection with the making of the decision, and that the decision involved an exercise of a power that is so unreasonable that no reasonable person could have so exercised the power, are not grounds of review under subs (1). Subsection (3) limits the “improper exercise of power” ground to exercise for a purpose other than one for which the power was conferred, exercise of a personal discretionary power at the direction or behest of another person and exercise of a discretionary power in accordance with the rule or policy without regard to the merits of the particular case; particularly, improper exercise “does not include taking irrelevant considerations, or failing to take relevant considerations, into account, an exercise of a discretionary power in bad faith or other exercises which represent an abuse of power”.
Judicially-reviewable decisions include decisions of the Refugee Review Tribunal. That Tribunal is constituted by s 394 of the Migration Act. Its function is principally (s 411) to review on the merits decisions to refuse to grant, or to cancel, a protection visa. Its proceedings are inquisitorial rather than adversarial. Section 420 exhorts it “to pursue the objective of providing a mechanism of review that is fair, just, economical, informal and quick” and provides that the Tribunal is not bound by technicalities, legal forms or rules of evidence but must act “according to substantial justice and the merits of the case”. Section 430 requires written reasons:
“Where the Tribunal makes its decision on a review, the Tribunal must prepare a written statement that:
(a) sets out the decision of the Tribunal on the review; and
(b) sets out the reasons for the decision; and
(c) sets out the findings on any on material questions of fact; and
(d) refers to the evidence or any other material on which the findings of fact were based.”
The task of the Tribunal is by no means an easy one (nor, for that matter, is the task of the primary decision-maker). The primary criterion — the applicant is a person to whom Australia has protection obligations — is one which requires the construction and application of the Convention. Primarily — though not exclusively — it requires the Tribunal to answer the questions, is the applicant a refugee, as that word is defined by Article 1 of the Convention? That is, does the applicant have a well-founded fear of persecution for one of the Convention reasons? Is the applicant unwilling to return to his or her country of nationality because of that fear? Is the applicant unable to take advantage of protection by his or her country of nationality? If, on the other hand, the applicant is stateless, then the Tribunal must grapple with similar questions in relation to a country of former habitual residence. To answer those questions, the Tribunal, in deciding whether an applicant’s fear of persecution is well-founded, must apply the “real chance” test explained by the High Court in Chan. That involves prediction and speculation, a process which in turn is informed by findings of fact, to be arrived at by applying a standard at least closely akin to the civil standard of balance of probabilities.[49] But, if the Tribunal is less than confident about its findings of fact it must take its lack of confidence into account in assessing “real chance”:
“It is true that, in determining whether there is a real chance that an event will occur or will occur for a particular reason, the degree of probability that similar events have or have not occurred or have or have not occurred for particular reasons in the past is relevant in determining the chance that the event or the reason will occur in the future. If, for example, a Tribunal finds that it is only slightly more probable than not that an applicant has not been punished for a Convention reason, it must take into account the chance that the applicant was so punished when determining whether there is a well-founded fear of future persecution.”[50]
The facts about which the Tribunal must make findings fall broadly into two categories. One relates to the conditions in the applicant’s country of nationality or of former habitual residence. As to that category, the Tribunal usually has available to it a wide range of helpful material, including departmental cables and reports, reports of other countries’ governmental agencies and documents emanating from United Nations agencies and various human rights and aid organisations: as well, of course, as press reports. The other category of facts comprises those relating to the particular applicant: the applicant’s own account of his or her life, associations and treatment, frequently largely if not wholly, uncorroborated and often given in circumstances where the applicant has arrived in Australia on forged documents or without papers of any kind. Much then depends on the Tribunal’s view of the credibility of the applicant and the inherent probability of the applicant’s account, judged against the background of what the Tribunal knows of conditions in the country from which the applicant has come.
Cases in which an applicant (or, as occasionally happens the Minister) seeks review of a decision of the Tribunal are usually (not always) relatively short (that is, can be heard within half a day). The view is generally taken that they should be heard and decided promptly, particularly where, as is often the case, the applicant is in detention under the Migration Act. Many applicants do not have legal representation; often, however, the applicant is represented by a solicitor or counsel. Usually, in my experience, the Minister is represented by counsel. It is not surprising that diligent counsel preparing cases should, from time to time, have found in incompletely mapped territory some novel points, which the Court has had to decide. Nor is it surprising, given the difficulty of some of the points, that from time to time judges have taken different views about some aspects of the legislation. Given those circumstances, combined with the number of cases decided and the delay necessarily inherent in the appellate process — particularly where there is an appeal to the High Court — it is not remarkable that the development of the law in this area has been marked by a somewhat lesser degree of the orderliness described earlier in this paper. The point may be demonstrated by a few examples.
The first example illustrates the difficulties that can arise in construing a novel code for merits and judicial review. As we have seen, it is a ground of judicial review under s 476 that procedures that were required by the Act or the regulations to be observed in connection with the making of the decision were not observed; and s 476(2) excludes, as independent grounds of review, the rules of natural justice and “Wednesbury unreasonableness”. But s 420 requires the Tribunal, in carrying out its functions, to pursue the objective of providing a mechanism that is, among other things, fair and just. Furthermore, in reviewing a decision, the Tribunal must act according to substantial justice and the merits of the case.
There were two views of the interrelationship between s 476 and s 420, each of which had support in decided cases. One view was that s 420 prescribed procedures required to be observed: thus, it was said that there was a legal requirement binding the Tribunal to observe procedures which were “fair” and “just” and to act (procedurally) according to substantial justice and the merits of the case, so as to permit review under s 476(a) independently of the grounds excluded by s 476(2). That view was supported, for example, by dicta in Dai v Minister for Immigration and Ethnic Affairs,[51] Singh v Minister for Immigration and Ethnic Affairs,[52] Asrat v Vrachnas[53] Li v Minister for Immigration and Multicultural Affairs[54] and Mohideen v Minister for Immigration and Ethnic Affairs.[55] The other view was that s 420 did not prescribe procedures required to be observed in the sense contemplated by s 476(1)(a). That view had the support of Olney J in Mohideen; it also had support in Thanh Phat Migration Act v Billings,[56] Di v Minister for Immigration and Ethnic Affairs,[57] De Motte v Minister for Immigration and Ethnic Affairs[58] and particularly in an extended analysis by Lindgren J in Sun Zhan Qui v Minister for Immigration and Ethnic Affairs.[59] The former view, however, was preferred by a majority of the Full Court in Eshetu,[60] and affirmed by later Full Court decisions, notably Sun.[61] It is obvious- and the case illustrates- that if that view were taken there was a potentially difficult question, of how far the Court could explore the Tribunal’s compliance with s 420 without entering the forbidden field of merits review: a point that emerges with some force from the High Court’s decision in Eshetu.
The High Court, in Eshetu, reversed the decision of the Full Court, clearly establishing the correctness of the latter of the two views, and particularly of the analysis of Lindgren J in Sun. Meantime, of course, the decision of the Full Court in Eshetu had been followed many times and had itself been the subject of a degree of elaboration, the detail of which it is now, of course, unnecessary to trace.
My second example relates to a relatively straightforward, but for several years largely unnoticed, question of construction of the Convention. It has to do with the principal criterion for the grant of a protection visa. Before the Migration Reform Act 1992 took effect on 1 September 1994, the question to be decided by the Minister was whether an applicant for recognition was a “refugee” as defined in the Convention. Since 1 September 1994, the question has been, is the applicant a person to whom Australia owes protection obligations under the Convention? The difference made by that change remained unexplored until the Full Court perceived, in Minister for Immigration and Multicultural Affairs v Thiyagarajah[62] the importance of Art 33 of the Convention. That article provides that:
“No Contracting State shall expel or return … a refugee in any manner whatsoever to the frontiers of territories where his life or freedom would be threatened on account of his race, religion, nationality, membership of a particular social group or political opinion.”
It is that provision, the Full Court held in Thiyagarajah, which principally gives content to the expression “protection obligations” (a term not specifically defined either in the Convention or in the Migration Act). The simple point is that if a third country has granted an applicant “effective protection” Australia may, consistently with its international obligations under the Convention, return the applicant to the third country; and if there is a third country to which Australia may return an applicant, then there is no need to embark on a consideration of whether the applicant is indeed a refugee. But the Full Court did not attempt, in Thiyagarajah, “to seek to chart the outer boundaries of the principles of international law which permit a Contracting State to return an asylum seeker to a third country without undertaking an assessment of the substantive merits of the claim for refugee status.”[63]
There has followed a stream of decisions, both of the Full Court and of single judges, in which the boundaries have been explored and some of the territory within them described. The cases include Rajendran v Minister for Immigration and Multicultural Affairs,[64] Minister for Immigration and Multicultural Affairs v Gnanpiragasam,[65] Al-Zafiry v Minister for Immigration and Multicultural Affairs,[66] Minister for Immigration and Multicultural Affairs v Kabiaf,[67] Al-Anezi v Minister for Immigration and Multicultural Affairs[68] and Al-Sallal v Minister for Immigration and Multicultural Affairs.[69] Plainly by no means the last word has been said on this topic; for example, I understand that a notice of appeal has been lodged from the decision of Katz J in Al-Sallal, in which his Honour decided that only those countries which have acceded to the Convention are “safe third countries” offering “effective protection” for the purposes of the Thiyagarajah principle.[70]
A third and final example may be found in the spate of cases dealing with the interaction between s 430 and s 476(1)(a) of the Migration Act. This, again, relates to the construction of the code in the Migration Act, and the difficult territory separating the role of the Court from that of the Tribunal. I already mentioned s 430: it describes the content of the written statement to be prepared by the Tribunal when it makes its decision on a review. Particularly, the statement is to set out the decision, the reasons for it and the findings on any material questions of fact; and it is to refer to the evidence or any other material on which the findings of fact were based. The cases establish that s 430 defines procedures which the Tribunal is required to observe, so that a failure to observe those procedures is a ground of review under s 476(1)(a). The cases also explore the extent of the obligation to set out reasons and the ambit of the expression “any material questions of fact”. The tension that arises is described as follows by Finn J in Kandiah v Minister for Immigration and Multicultural Affairs:[71]
“There is now a considerable body of case law that emphasises variously: (i) the importance to the parties, to the public and to review bodies of adequate reasons for decisions; (ii) the understanding and restraint that courts should demonstrate when reviewing and construing reasons for administrative decisions; and (iii) the content in terms of findings and recitation of evidence that properly and reasonably can be expected of administrative decision makers.”
For example, there will be a failure to comply with s 430 if the Tribunal fails to make findings about episodes of ill treatment, which an applicant to have occurred which are plainly crucial to an assessment of the prospect that the applicant will be ill-treated if returned to his or her country of nationality,[72] or where the Tribunal fails to make findings in relation to material before it plainly relevant to the reasons for ill treatment found to have happened.[73] The topic is one, generally, which has received a great deal of consideration in a number of recent cases but on which, no doubt, much remains to be said.[74]
In summary, the cases on review, under Pt 8 of the Migration Act, of decisions of the Refugee Review Tribunal may be described as an example, perhaps a unique example, of judicial review under a degree of stress. I have attempted to identify the principal causes of that. They arise from the number of cases which must be decided and the inherently difficult nature of the questions that they raise. Particularly, they arise from the combination of factors to which I have referred.[75] The processes involved are the conventional ones of construing an international instrument and a statute and, within the limits set by the grounds in s 476 of the Migration Act, reviewing the legality, not the merits, of decisions. It should not cause surprise, given both novelty and difficulty, that there are differences of opinion and even, occasionally, controversy.