Should Keeley be reconsidered?
25 In Transurban City Link Limited v Allan (1999) 95 FCR 553 a Full Court of this Court consisting of five members (Black CJ, Hill, Sundberg, Marshall and Kenny JJ) had to consider whether an earlier Full Court decision be reconsidered. It is instructive that their Honours identified this issue [at 26] as "whether this Court should embark upon a reconsideration of the previous Full Court decision". That this is a preliminary, threshold issue needs to be kept firmly in mind. If the subsequent Full Court immediately sets sail into a detailed examination of the issues considered by the earlier Full Court, it may come to the conclusion that it disagrees with the earlier decision. It would therefore follow that the earlier decision, in the view of the later Full Court, is wrong. It is then but a short and almost irresistible step to conclude that the earlier decision should not be followed. How can it be right to follow a decision now established to be wrong? But such an approach can result in the frequently repeated rule of restraint being given little more than lip service. In Transurban the Full Court continues:
"[27] It is not in doubt that a Full Court of this Court has power to decline to follow the previous decision of a differently constituted Full Court. The Court is not bound to perpetuate error if error there be. Nor is it in doubt that while the Court has that power, it is a power which should be exercised with great care. The doctrine of precedent, which is fundamental to the common law, brings with it the consequence that decisions of an intermediate court of appeal will be binding on single judges within the same court hierarchy. They will be relied upon by the broader community and the profession. Decisions of a Full Court of this Court are entitled to due respect and will not be lightly departed from.
[28] In Nguyen v Nguyen (1990) 169 CLR 245 at 268-269, Dawson, Toohey and McHugh JJ, observed that the extent to which the appellate court of the Supreme Court of a State regards itself as free to depart from its own previous decisions must be a matter of practice for the court to determine for itself, citing the judgment of Bowen CJ and Foster J in Chamberlain v The Queen (1983) 72 FLR 1 at 8-9, and noted also that the Full Court of the Federal Court will depart from a previous decision if convinced that it is wrong. Their Honours then said:
'Where a court of appeal holds itself free to depart from an earlier decision it should do so cautiously and only when compelled to the conclusion that the earlier decision is wrong. The occasions upon which the departure from previous authority is warranted are infrequent and exceptional and pose no real threat to the doctrine of precedent and the predicability of the law …'.
See also Australian Securities Commission v Marlborough Gold Mines Ltd (1993) 177 CLR 485 at 492
[29] A differently constituted Full Court would, of course, decline to follow the decision of another Full Court if it concluded that the previous decision was clearly erroneous. It would be wrong to do this merely because the matter was one on which minds might differ: cf Magman International v Westpac (1991) 32 FCR 1 at 20 per Hill J.
[30] What their Honours said in Nguyen must be read in the context of their previous remarks. The statement of principle in Chamberlain v The Queen, cited with evident approval by their Honours, was qualified by the word 'normally'. The use of expressions of this nature leaves the way open for an approach that is appropriate to the circumstances of a particular case: see La Macchia v Minister for Primary Industries and Energy (1992) 110 ALR 201 at 204. Towards the conclusion of their joint judgment in Nguyen, Dawson, Toohey and McHugh JJ, noting that appeals to the High Court were now by special leave only, and that the appeal courts of the Supreme Courts of the States and of the Federal Court were, in many instances, courts of last resort for all practical purposes, observed (at 269-270):
'In these circumstances, it would seem inappropriate that the appeal courts of the Supreme Courts and of the Federal Court should regard themselves as strictly bound by their own previous decisions. In cases where an appeal is not available or is not taken to this Court, rigid adherence to precedent is likely on occasion to perpetuate error without, as experience has shown, significantly increasing the corresponding advantage of certainty."
26 Subsequently another Full Court in Telstra Corporation Ltd v Treloar (2000) 102 FCR 595 was faced with the same question. The majority (Branson and Finkelstein JJ) gave a valuable reminder that the doctrine of stare decisis rests on policy grounds beyond mere conservatism. Their Honours said (at [23]):
"The doctrine of stare decisis takes its name from the Latin phrase 'stare decisis et non quieta movere' which translates as 'stand by the thing decided and do not disturb the calm'. It is a doctrine based on policy. The rationale for the doctrine can be grouped into four categories: certainty, equality, efficiency and the appearance of justice. Stare decisis promotes certainty because the law is then able to furnish a clear guide for the conduct of individuals. Citizens are able to arrange their affairs with confidence knowing that the law that will be applied to them in future will be the same as is currently applied. The doctrine achieves equality by treating like cases alike. Stare decisis promotes efficiency. Once a court has determined an issue, subsequent courts need not expend the time and resources to reconsider it. Finally, stare decisis promotes the appearance of justice by creating impartial rules of law not dependent upon the personal views or biases of a particular judge. It achieves this result by impersonal and reasoned judgments."
27 Relevantly for present purposes their Honours also discussed the issue in the context of a case, like the present one, where statutory construction is involved. Their Honours said (at [27-28]):
"[27] The problem is very real when what is at issue is the construction of a statute. For one thing, statutory language is often ambiguous. Courts can struggle to determine the legislative intent. It is often impossible to discover any legislative intent. In many instances the generality of the statutory language is deliberate and allows the courts to develop a body of law to fill the gaps. This may lead to disagreement among judges about what the statute means. It would be sound policy that once that intent has been discerned by an appellate court then that should be the end of the matter.
[28] The view which we prefer is that unless an error in construction is patent, or has produced unintended and perhaps irrational consequences not foreseen by the court that created the precedent, the first decision should stand. In other areas of the law a precedent may be reconsidered if its underlying reasoning is outdated or is inconsistent with other legal developments. Perhaps, with some modification, in some instances these factors could also be applied to cases concerned with the construction of statutes. Accordingly, we venture to suggest it would be on a rare occasion that an intermediate appellate court (contrast the position of the High Court, as to which see Babaniaris v Lutony Fashions Pty Ltd (1987) 163 CLR 1) will allow an issue concerning the construction of a statute, past and closed and especially a repealed statute, to be thrown open, producing as it clearly will, uncertainty, disruption to the conduct of affairs, a sense of grievance in those whom may consequently receive treatment less favourable than that received by others under the same statute and additional cost and expense. In this case, the number of individuals who will relevantly be affected by the construction of the statute may be assumed to be, by reason of the passage of time, relatively small."
28 In my opinion, there are compelling reasons for not embarking upon a reconsideration of Keeley.
29 First, Keeley was a considered decision of three judges of the Court continuing a decision at first instance. Senior Counsel for the Commission properly conceded that Keeley was not a decision per incuriam, such as occurs when some relevant statute or authority has been overlooked.
30 Secondly, special leave to appeal to the High Court was refused on 28 November 2000.
31 Thirdly, Keeley has been applied in at least one subsequent Full Court judgment: Repatriation Commission v Thompson (2001) 32 AAR 514. It is reasonable to infer also that in the large volume of administrative decision making under the VE Act Keeley has been applied in practice since the decision of the primary judge on 13 August 1999, that is to say some three years ago.
32 Fourthly, Keeley is an application in a particular legislative context of a familiar construction exercise which applies general principles that are not in doubt. It is a construction task often encountered by court. In Maxwell v Murphy (1957) 96 CLR 261 at 267 Dixon CJ, speaking of the distinction between laws which affect rights and those which merely affect procedures for enforcing those rights, said:
The distinction is clear enough in principle and its foundation in justice is apparent. Difficulties have always attended its application."
33 The fact that there is usually (as there is in the present case) room for argument suggest that it will be that harder to argue that a particular result reached by one Full Court is "clearly wrong" or shows "patent error".
34 Fifthly, Keeley does not produce irrational or unintended consequences. There is nothing irrational or surprising about statutes or regulations expressly providing, or being construed by courts to provide, that pre-existing rights or obligations are not affected. In an administrative context, this produces consistency and predicability rather than the reverse. Assume for the moment that Mrs Keeley's hypothesis was "upheld by" the SoP in force at the time of her claim (an issue which was not necessary for the Full Court to decide). Assume that another claimant widow of a veteran with essentially the same factual history as Mr Keeley made a claim and a different delegate of the Commission (properly) found the hypothesis upheld by the SoP and not disproved beyond reasonable doubt and thus allowed the claim. It seems only fair that Mrs Keeley should not be any the worse off because she had to pursue her claim to the AAT to have it correctly determined. Like cases should be treated alike.
35 Sixthly, the reconsideration and overruling of Keeley would not necessarily provide any greater certainty. It would be open to litigants to contend that Gorton was wrong and that Keeley was right after all and request that a five member bench be convened. It is a long-standing and well known practice in this Court for the Chief Justice to be asked to convene a Full Court of five members where there are conflicting Full Court decisions or where it is sought to have an earlier Full Court decision reconsidered. As already noted, Transurban was such a case. Other recent examples are Minister for Immigration and Multicultural Affairs v Singh (2000) 98 FCR 469 and Australian Steel Company (Operations) Pty Ltd v Lewis [2000] FCA 1915. As far as I am aware, there was no request for a bench of five to be convened in the present case. Indeed the Commission's written submissions filed before the hearing do not address at all the criteria for reconsideration of earlier Full Court decisions.
36 There is moreover an unusual feature in the present case. The primary argument of the Commission was that s 120A does not oblige the AAT to apply the SoP in force at the time of review or to depart from that obligation where an earlier SoP might be thought to be "more favourable" (submissions par 18); Keeley was correct and means that the AAT had to apply the SoP in force at the time of claim. It was the Commission's alternative argument that Keeley was not correct, and that the 1994 amendments introducing the SoP system displaced any presumption which would protect a "so-called 'accrued right' of an applicant for review to have the review conducted by reference to the SoP in force at the time of the primary decision": see submissions par 38-43 which are headed "An alternative approach: applying the most recent SoP in all cases". In essence the Commission attacked what it regarded as her Honour's "more favourable SoP" approach: submissions pars 16.1 and 16.2. It argued that in all cases the AAT should apply the earlier SoP (Keeley right) or, alternatively, the later SoP (Keeley wrong). So far from contending that Keeley was "manifestly wrong" or "clearly erroneous" (see Queensland v Commonwealth (1977) 139 CLR 585 at 620 et seq), the primary case of the Commission is that this Court should reverse the primary judge on the basis that Keeley is correct.
37 Seventhly, unlike private litigants adversely affected by a Full Court decision on the construction of the Trade Practices Act or the Patents Act, the Commission is in a position to secure legislative amendments if it considers the decision in Keeley operates against the public interest. This could be done either by explicit provision in SoPs themselves or by amendment to VE Act. The former case would simply be doing expressly what s 50 of the AI Act contemplates may be done by necessary implication. The transcript of the special leave application hearing shows that this factor was regarded as significant by the members of the High Court (Gaudron and Hayne JJ).
38 Eighthly, that part of the Commission's case which attacks Keeley centres on the contention that SoPs are evidentiary in character because they reflect "current medico-scientific knowledge". They do not have "immutable authoritative status. When the evidence supporting it is undermined, an SoP ceases to represent proved or known scientific fact and must be repealed" (submissions par 40).
39 I accept that this proposition is arguable and one on which reasonable minds may differ. Nevertheless I do not agree with it. An SoP is sui generis as a form of legislative instrument. Counsel were not able to point to anything similar, either in Australia or overseas. But on first principles an SoP seems to be substantive, something which determines rights rather than a procedural measure relating to the enforcement of those rights. The SoP is a "statute backed declaration of what is proved or known scientific fact": Deledio, supra at 96. True it is that the particular subject matter with which SoPs are concerned is medico-scientific knowledge. But that does not change the legal nature and effect of the instrument. As Kiefel J notes in Keeley, most amending legislation is thought to be an improvement on what went before. It is however an ancient principle that amending legislation, however desirable, ought not to be taken to affect existing rights unless it clearly says so.
40 The SoP regime has the effect that the existence or otherwise of stipulated facts will have legal consequences for VE Act pension claimants. To take the present case as an example, Mr Gorton's material must "raise" (in the sense discussed in the authorities) a hypothesis which includes daily consumption of alcohol. If it does not, no pension. If it does, the Commission may disprove, beyond reasonable doubt, daily consumption of alcohol by Mr Gorton. In that event, no pension.
41 The VE Act provides for pensions and other entitlements as a matter of right, and not discretion, dependent on the claim fitting certain facts (see Keeley at [73]). Often these facts are matters which are not self-evident. For example "operational service", the concept of service in war or war-like operations, is defined in ss 6 to 6F. The definitions are complex but for the most part require the veteran to have been serving in stipulated geographical areas during stipulated periods of time. These are legislatively determined factual criteria which determine rights. Claimants cannot call evidence from military historians to show that hostilities in fact occurred at other places and times. Amendment to the VE Act to restrict the definition would not, unless provided otherwise, affect existing claims. The SoP regime is not relevantly different. The requirement of a reasonable hypothesis is something unique to the VE Act. As Deladio establishes, it is not the same as proof of facts. But there is nevertheless a prescribed relationship between a stipulated fact and a right, whether that fact has to be "raised by material" as part of a hypothesis or disproved beyond reasonable doubt.