HEYDON J. In the Sussex Peerage Case, Lord Brougham stated:
"To say, if a man should confess a felony for which he would be liable to prosecution, that therefore, the instant the grave closes over him, all that was said by him is to be taken as evidence in every action and prosecution against another person, is one of the most monstrous and untenable propositions that can be advanced."
The proposition to which Lord Brougham applied this soft impeachment was far less extreme than the proposition advocated by the appellant in this appeal. The appellant submitted that an out-of-court statement incriminating its declarant is admissible in an accused person's favour even when the declarant is alive, available, and sitting in the courtroom in which the out-of-court statement is tendered, but has simply chosen not to testify. Hitherto Australian courts, apart from the Supreme Court of Queensland, have held that that submission is not the law. That position was challenged in Bannon v The Queen, but this Court did not overturn it. The appellant now challenges Bannon v The Queen itself.
The issues in this case arose out of an extensive fight which broke out during a large "warehouse party" at 3.00am. During that fight, a young man fell through a glass window. He died from the effects of the fall. The prosecution alleged that the appellant and a minor tried jointly with him, LM, caused the deceased's fall. The prosecution case was that the appellant and LM intended to cause the deceased really serious injury and that one or other or both of the accused pushed the deceased, causing his fall.
The question in this appeal is whether hearsay statements made by LM which allegedly exculpated the appellant were admissible in his favour.
Bannon v The Queen: summary
Bannon v The Queen was heard by six Justices. Four judgments were delivered: one by Brennan CJ, one by Deane J, one by Dawson, Toohey and Gummow JJ, and one by McHugh J. Apart from some dicta of Deane J which are examined below, their Honours did not support abandoning the common law position that the rule against hearsay prohibits receiving statements of one co‑accused favourable to another co-accused and statements of third parties favourable to an accused.
The appellant states the key question in this case as being whether the Court should now "recognize an exception to the rule against hearsay of the kind considered, but not decided, in Bannon v The Queen". That formulation compels reconsideration of the arguments which did not prevail in that case.
The appellant in Bannon v The Queen advanced three arguments for the Court's consideration. One was that the rule against hearsay should be applied flexibly. The second argument was that there was an exception to the rule against hearsay when an out-of-court statement by a person other than the accused was against that person's penal interest and that person is not available to testify. The third argument was that an out-of-court statement of that kind which was reliable should be admitted where its admission was reasonably necessary to prove a fact in issue. Each of the six Justices who heard Bannon v The Queen either denied that the law should be changed in this way or held that even if the common law could be appropriately modified, that modification would not assist the appellant in the circumstances of the case.
It is convenient to deal in turn with each of these arguments, with a difficulty that emerged in the course of judgment in Bannon v The Queen, and with the dicta of Deane J. Finally, the arguments of the present appellant not raised in Bannon v The Queen will be examined. However, it is desirable first to deal with two preliminary points of a factual character.
The appellant submitted that LM's out-of-court statements exculpated him. The appellant also submitted that those statements were reliable. Neither submission is correct.
Did LM's statements exculpate the appellant?
The appellant repeatedly submitted that the admissions of LM which the prosecution relied on tended to implicate LM and exculpate the appellant. Those admissions fell into two groups. It is not correct to say that either group exculpated the appellant.
The first group of statements allegedly exculpating the appellant comprised statements by LM to police officers during a formal interview. Those statements were to the effect that he had pushed the deceased. The problem for the appellant is that LM did not admit pushing the deceased out of the window. He said only that he assumed that that was the case. And he specifically denied two leading questions by a police officer suggesting that he had pushed the deceased out of the window. LM told the police officers that the deceased was one or two metres, or one and a half metres, from the window when he pushed him. LM claimed that he did not see the deceased go out of the window. LM's statements that he had pushed the deceased do not amount to evidence that he pushed the deceased out of the window. And they do not amount to evidence that the appellant did not push the deceased out of the window. In the police interview, LM said that he and the appellant had come to the party in LM's car. The appellant had driven. LM said that they left the party by the same means. They left together. The appellant and LM were dancing at the party when the fighting broke out. LM persistently refused to name his friends or describe their conduct. The trial judge treated LM's denials of recollection and refusal to name his friends with great scepticism.
The second group of statements allegedly exculpating the appellant comprised statements by LM in the car in which he left the party. The appellant, LM and three others left in the car immediately after the deceased had fallen. Mr Asfer testified that LM had said to Mr Faulkner: "Look what you made me do". Mr Morgan testified that LM had said: "See what you've done" and "see what you've put us through". In cross-examination, Mr Morgan accepted that he had told the police that LM had said: "See what you've done, look what you've made me do." These rather vague statements may inculpate LM and Mr Faulkner. However, again, they do not exculpate the appellant.
LM's statements were thus affected by the same weakness as that of the co-accused whose statement was considered in Bannon v The Queen. Professor C R Williams has analysed the appellant's position in that case as follows:
"The argument of the defence involved using the statement of the co-accused as implied hearsay. From the admission 'I killed the deceased' without reference to the accused, it was sought to infer that the co-accused was in substance saying 'I alone killed the deceased', which if the statement had been made, would have been direct hearsay. If a general rule against the admission of hearsay evidence is accepted, then the statement of the co-accused was correctly regarded as not admissible for the purpose of drawing such an inference. No issues of faulty perception or erroneous memory on the part of the co-accused arose. Possible issues of insincerity and ambiguity were however both present and significant. There are any number of reasons why a person who has committed a killing may choose not to implicate another participant in the offence. When considered for the purpose for which it was sought to be used, the statement was highly ambiguous; a statement 'I committed a killing' by no means necessarily means 'I alone committed a killing'."
That reasoning is particularly apposite here. LM's statements that he pushed the deceased, taken in isolation, are not the same as LM saying: "I, and I alone, pushed the deceased." The contexts in which LM made those statements could of course suggest otherwise. But here those contexts, particularly the police interview, do not suggest otherwise. They are neutral or point in the opposite direction.
In the police interview, LM was extremely keen to minimise the involvement of his friends in general and the appellant in particular. Yet he never explicitly said that the appellant had no involvement. In view of his manifest desire to protect the appellant, that is significant. It is not possible to infer from LM's statements that the appellant did not push the deceased.
The statements by LM in the car were made in the presence of the appellant and three other people. At the time they were made, the appellant was driving the car. It is perhaps unlikely that LM would have intended to accuse the appellant in front of the others, especially while the appellant had control of the vehicle in which he was a passenger. In any event, nothing about the circumstances in which this group of statements were made suggests that LM was taking responsibility for the deceased's death to the exclusion of the appellant.
The hearsay statements by LM do not exculpate the appellant. The factual substratum necessary for debate about creating a new hearsay exception for third party confessions favourable to the accused does not exist in this case. That conclusion alone is sufficient ground for dismissing the appeal. However, the appeal should also be dismissed on other grounds set out below. It is relevant to those grounds to consider a second preliminary point of a factual character - whether LM's statements were reliable.
Were LM's statements reliable?
The appellant submitted that LM's statements were reliable for four reasons. First, the prosecution relied on them as inculpatory of LM. This was significant because the burden on the prosecution to prove guilt beyond a reasonable doubt "tends to ensure a relatively high degree of reliability". Secondly, LM did not object to the admissibility of his own admissions. Thirdly, the first group of admissions were made to the police. Fourthly, the evidence of Messrs Asfer and Masonga that the blows were struck by LM corroborated that group of admissions. It is convenient to deal with these four reasons in turn.
As to the first, it is true that the prosecution relied on the admissions. But it relied on them as inculpatory of LM only. The appellant's submission suggests that the burden of proof resting on the prosecution causes it always to call weighty evidence. It suggests that every piece of evidence that the prosecution relies on is valuable. It suggests that the prosecution vouches for or warrants the truth of each piece of evidence. These suggestions overlook the prosecution duty to call all available evidence. In R v Apostilides, Gibbs CJ, Mason, Murphy, Wilson and Dawson JJ said:
"A refusal to call [a] witness will be justified only by reference to the overriding interests of justice. Such occasions are likely to be rare. The unreliability of the evidence will only suffice where there are identifiable circumstances which clearly establish it; it will not be enough that the prosecutor merely has a suspicion about the unreliability of the evidence."
It was for the jury to assess what caused the deceased's death, taking into account the hearsay evidence of LM as well as many other accounts of the evening. The conflicts between the testimony of prosecution witnesses did not mean that the prosecution was warranting one version or another as reliable.
The second reason advanced for thinking that the hearsay evidence was reliable was that LM did not object to the admissibility of his own admissions. How could he? Police officers recorded the first group of admissions. They were made in circumstances not raising any possibility of objection on grounds of involuntariness. The circumstances did not raise any possibility of discretionary exclusion either. LM had consulted his solicitor. The interview took place in his father's presence. He received all necessary warnings about his rights. The second group of admissions were made in circumstances not raising any possibility of objection by LM. In any event, the fact that LM did not object does not of itself make his admissions reliable. The events he had participated in were confused and fast-moving. They had the capacity to affect the accuracy of perception.
The third supposed reason why LM's evidence was reliable was that the admissions in the first group were made to police officers. It is true that the statements were made in a formal setting under police caution. That circumstance supports the conclusion that the admissions were made. It does not render reliable LM's powers of perception of the events he narrated, his recollection of those events or his expression of that recollection. It is a novel proposition that where the police are investigating a death, there is some special solemnity attaching to police interviews of witnesses.
The fourth reason the appellant assigned for the reliability of LM's statements was that the testimony of Messrs Asfer and Masonga corroborated them. That takes no account of the fact that they were contradicted by the evidence of Mr Doig, Mr Arcaro and, to some extent, Mr Stuart. In any event, so far as reliability is relevant to the reception of LM's statements, it must be assessed independently of what other witnesses said.
For all of those reasons, LM's statements were not reliable.
Flexible application of the hearsay rule?
It is now necessary to turn to the three arguments put in Bannon v The Queen on which the appellant relied. The first is that the hearsay rule should be relaxed in the case of highly or sufficiently reliable evidence. In Bannon v The Queen, Brennan CJ attacked and rejected the proposed relaxation. Dawson, Toohey and Gummow JJ (and Deane J, who agreed with their Honours) did not address it. McHugh J did not find it necessary to deal with it. His Honour held that the evidence was not sufficiently reliable to satisfy the exception if it existed. The grounds which Brennan CJ assigned for his Honour's conclusion are, with respect, convincing. One of his Honour's points concerned the parties' inability to cross-examine the makers of out-of-court statements who do not give evidence. Another concerned the fact that neither the judge nor the jury would have seen the out-of-court statement being made or seen the declarant giving evidence. To these points may be added the following considerations.
Suppose a witness, Mr W, testifies to a relevant event which he has himself observed. In those circumstances, the jury is invited to accept that Mr W did perceive the event, that he has remembered correctly what he perceived, that there is no ambiguity in his narration of the event to the court, and that he is sincere in his testimony. Sincerity may be promoted by the making of an oath or affirmation. Mr W's fear of criticism for false evidence, inside the court or outside it, and the prospect of prosecution for perjury may also promote truthfulness. But the principal guarantee and test of the witness's honesty and the accuracy of his perception, memory and narration lie in his capacity to be cross-examined. At the end of the day, the jury must be satisfied both that the witness believes that the event took place, and that his belief is justified.
It is obvious that any danger that the jury wrongly reaches, or fails to reach, that state of satisfaction is accentuated when what is offered to the court is not Mr W's account of what he saw, but his account of what another person, who does not testify - Mr Jones - saw. In that instance, the jury must be satisfied that Mr W believed he heard Mr Jones say that the event happened. It must be satisfied that Mr W's belief is justified. It must be satisfied that Mr Jones believed that the event happened. And it must be satisfied that Mr Jones's belief was justified. The jury must decide whether it is so satisfied without the safeguards which the ordinary conditions of testimony would have provided in relation to Mr Jones had he been called as a witness. It must decide without testing of Mr Jones's account in cross-examination. And it must decide despite the diminished value of these safeguards so far as Mr W is concerned. There is no direct link between Mr W's testimony and the conclusion that the jury is invited to reach as a result of it. The direct witness, who asserts that the event occurred, can be cross-examined much more fruitfully about his perception and memory of that event than the hearsay witness can be cross-examined. The direct witness claims personal experience of the event and can be cross-examined on that experience. The hearsay witness, who merely testifies that he heard the direct witness say "the event took place", can only be cross-examined about what he claims he heard the direct witness say. The direct witness is giving the "best" evidence. The hearsay witness is not.
There is a further danger in the introduction of a relaxed or flexible rule based on "reliability". In Pollitt v The Queen certain Justices in this Court employed the kind of thinking that the appellant advocated in this appeal to create a limited exception permitting parties to telephone conversations to be identified. The differences of opinion that emerged between those Justices illustrate that danger.
Whatever the merits of a flexible application of the hearsay rule based on reliability, it would not be correct to indulge in that process here. LM did not directly exculpate the appellant. And the evidence concerning what LM said is not a very reliable guide to what he saw.
Statement against penal interest?
The second argument advanced in Bannon v The Queen that the appellant relied on was that this Court should recognise an exception to the rule against hearsay for out-of-court statements by persons other than the accused where those statements are against the maker's penal interest and the maker is unavailable to testify. In Bannon v The Queen, Brennan CJ pointed out that Holmes J (dissenting in Donnelly v United States) and Wigmore had attacked the narrowness of the exception to the hearsay rule based on declarations against interest. That exception applied to proprietary and pecuniary interest only. Extending that exception to include declarations against penal interest would depend at least on the declarant's death, as the exception concerning declarations against proprietary or pecuniary interest does. An extension to that extent only would not assist in the present case: LM was not dead at the time of the appellant's trial. Brennan CJ observed that in Canada a hearsay exception developed as a result of these criticisms. It depends on satisfaction of at least two conditions. The first is that the declarant "should have apprehended a vulnerability to penal consequences as a result" of making the statement, which vulnerability was "not remote". The second is that the declarant be unavailable through death, insanity, grave illness or absence in a jurisdiction to which the processes of the court do not extend. The first condition corresponds with an equivalent condition that applies to declarations against proprietary and pecuniary interest. In Ward v H S Pitt & Co Hamilton LJ, speaking for himself, Sir Herbert Cozens-Hardy MR and Buckley LJ, said:
"It is essential that the deceased should have known the fact [declared in his out-of-court statement] to be against his interest when he made [the statement], because it is on the guarantee of truth based on a man's conscious statement of a fact, 'even though it be to his own hindrance', that the whole theory of admissibility depends."
Even if the powerful objections that Brennan CJ stated to extending the law in this way were put aside, it would be futile to create the exception in this case. LM did not meet any of the conditions of unavailability. And, at least in respect of the second group of statements, the "apprehended vulnerability to non-remote penal consequences" condition was not met either. Hence adopting this proposed exception would not benefit the appellant.
In Bannon v The Queen, Dawson, Toohey and Gummow JJ (Deane J concurring) did not decide on the correctness of the second postulated exception. Their Honours did not need to decide because they held that the declarant in that case did not appreciate that what she narrated was to her prejudice. McHugh J took the same course. In this appeal, the appellant submitted that the Canadian requirements for the reception of third party or co-accused confessions, such as "reliability" and "corroboration", were satisfied here. For the reasons given above, that is not so.
Necessity and reliability?
The third argument from Bannon v The Queen that the appellant advocated was that there exists a hearsay exception for statements exculpatory of an accused where those statements are both reliable and reasonably necessary to prove a fact in issue. Brennan CJ rejected this submission in Bannon v The Queen. His Honour did so on the ground that this third proposed exception is fundamentally inconsistent with Australian authority.
Dawson, Toohey and Gummow JJ (Deane J concurring) did not deal specifically with this third proposed exception. McHugh J criticised it, but held that it was unnecessary to decide whether it should be adopted in Australia. That was because the evidence did not satisfy the reliability threshold.
Even if the exception existed, which it is unnecessary to decide, LM's statements would not satisfy its terms. They do not exculpate the appellant. Even if they did, they are not reliable. And they were not necessary to prove that fact in issue. The relevant fact, that LM alone pushed the deceased, could have been proved by the appellant deciding to enter the witness box and testifying to that effect.
The inculpation of the accused
In the concluding paragraph of his reasons for judgment in Bannon v The Queen, Brennan CJ observed that "counsel for the appellant did not submit that any exception to the hearsay rule should admit evidence inculpatory of an accused person." It is plain that Brennan CJ thought that any new hearsay exception admitting evidence inculpatory of an accused person was undesirable. That is because, in rejecting the "flexible application" exception, his Honour had said:
"To admit hearsay evidence whenever the judge forms the opinion that the evidence is sufficiently reliable would be to transform the nature of a criminal trial. If the judge's opinion be based on no specific criteria but only on an appreciation of the circumstances generally, the judge would have to exercise a lively discretion to exclude evidence that the judge thought to be reliable in order to prevent undue prejudice to the accused who could not cross-examine the maker of the out-of-court statement."
McHugh J noted that non-inculpation of the accused was a requirement in Canadian law.
In the concluding paragraph of his reasons for judgment in Bannon v The Queen, Brennan CJ also said the out-of-court statements of the co-accused, Kerry Calder, "in so far as they inculpated herself were likely to be understood by the jury as inculpatory of the appellant as an aider and abettor." On Brennan CJ's approach, that made them inadmissible. Hence, said Brennan CJ:
"It follows that, on any view of the scope of an exception to the hearsay rule as contended for, the appellant cannot succeed. If Calder's statements were understood to be exculpatory of the appellant in the sense that the appellant was not a party to Calder's offence, Calder's statements can hardly be taken to be reliable as to that fact. That fact is not asserted and the circumstances in which the statements were made give no assurance that Calder was adverting to the appellant's involvement in the murders."
These passages are germane in this appeal too. Here, too, counsel for the appellant did not submit that any new exception to the hearsay rule should admit evidence inculpatory of an accused person. Here, too, LM's statements to the police officers were likely to be understood as inculpatory of the appellant as an aider and abettor. That is so because the prosecution case at the joint trial of the appellant and LM was put on an alternative basis: either they acted in concert, or each aided and abetted the other. If LM's statements were admitted not only against LM, but in favour of the appellant, it was open to the jury to infer, given LM's statement that the appellant was present, that LM's reluctance to speak fully about the appellant's role was attributable to the fact that they had acted in concert or aided and abetted each other. Finally, like Kerry Calder's statements in Bannon v The Queen, if LM's statements were understood to be exculpatory of the appellant, they were not reliable in that respect.
If out-of-court confessions of crime are admissible in favour of the accused, why are they not admissible against the accused? The difficulties which their admissibility against the accused would present suggest that they should not be admitted either against or in favour of accused persons.
Deane J's dicta in Bannon v The Queen
The appellant relied on several passages in Deane J's reasons in Bannon v The Queen. The appellant stressed three in particular. The first appears after his Honour set out certain examples not comparable to the present facts:
"The point of the examples is simply to demonstrate that, in circumstances where the Crown has seen fit to proceed against two accused persons jointly and to lead particular evidence on the joint trial against one only of them, a situation can arguably arise in which ordinary considerations of fairness would be affronted and the administration of criminal justice mocked if the other accused were precluded from relying upon that evidence if it supported his or her innocence or raised a doubt about his or her guilt."
By "doubt" his Honour meant "reasonable doubt". That characterisation is supported by a later passage in Deane J's reasons on which the appellant also relied:
"The central prescript of our criminal law is that no person should be convicted of a crime unless his or her guilt is established beyond reasonable doubt after a fair trial according to law. The specific content of the requirement of a fair trial may vary with changing circumstances, including contemporary standards and perceptions. When it appears that judge-made rules of evidence or procedure conflict, or are liable to conflict, with the basic requirements of fairness, it is a function of a final appellate court … to address the question whether those rules should be altered or adjusted to avoid such conflict." (footnote omitted)
The third passage on which the appellant relied was:
"it appears to me to be strongly arguable that the basic requirement of fairness dictates that, in circumstances where the Crown has seen fit to bring a person (the first accused) to a joint trial with another accused and to place before the jury material which is tendered only against that other accused but which is supportive of the innocence of the first accused, the trial judge have a discretion to direct that that material, even though otherwise inadmissible in the trial of the first accused, be evidence in that trial at the instance of the first accused if, in all the circumstances of the case, the trial judge considers that fairness to the first accused and the interests of the administration of justice support the conclusion that such a direction be given."
Deane J made it plain that he was not asserting these views as concluded statements of the law, but as suggestions only. The appellant's submissions about them are expressed subject to that caveat. The appellant submitted that "Deane J's suggested exception" was sound in principle. But what is the suggested exception? If it corresponds with the final passage quoted, LM's statements do not support the appellant's innocence or raise a reasonable doubt about his guilt.
The appellant referred to Parke B's aphorism: "what a party himself admits to be true, may reasonably be presumed to be so." Parke B's aphorism justifies the reception of admissions which the maker was aware were against interest. But it cannot justify the reception of all admissions. The reception of admissions, unlike the reception of declarations against proprietary or pecuniary interest, does not depend on the maker's awareness that they are against interest at the time they are made. Some admissions are received even though their makers thought them to be in their interests at the time they were made. E M Morgan has advanced a sounder justification for the reception of admissions:
"The exclusionary rules of evidence and the procedure for enforcing them are not designed to be automatic eliminators of untrustworthy testimony; in the main they rather provide a privilege of protection against such testimony to the party against whom it is offered. A litigant can scarcely complain if the court refuses to take seriously his allegation that his extra-judicial statements are so little worthy of credence that the trier of fact should not even consider them. He can hardly be heard to object that he was not under oath or that he had no opportunity to cross-examine himself."
However that may be, on the strength of Parke B's aphorism, the appellant then made a submission assuming that a confession by B implicating himself and exculpating A was necessarily a reliable confession. The submission also assumed that LM's admissions implicated him and exculpated the appellant. The submission was:
"The same rationale dictates that LM's admission that he pushed the deceased and caused him to fall, in circumstances where there are competing versions as to which of the two accused pushed the deceased, and the appellant disputes that he did so, may reasonably be presumed to be (or possibly be) a true statement that the appellant did not push the deceased."
It may be accepted that LM's admission that he pushed the deceased could reasonably be presumed to be true. But LM did not say that the appellant did not push the deceased. Deane J's statements in Bannon v The Queen depend on a consideration stated fairly early in his reasons for judgment. That consideration is that the hearsay statement be one in which the maker "unambiguously admits that, alone and without assistance", he or she committed the crime. That is not true of LM's statements.
What if it were true of LM's statements? It is convenient to examine what force the argument has. There is a conceptual difficulty in it. On the strength of an approach similar to Parke B's, many confessions and admissions have weight because it is assumed that their makers would not invent their involvement in criminal conduct. But it does not follow from that reasonable assumption that those makers would not lie about the involvement of persons other than themselves. On the one hand, it is a commonplace that one accomplice may falsely exaggerate the guilty role of others. On the other hand, it is possible that an accused person may falsely diminish the guilty role of others - whether out of affection for those others, or out of fear of future retribution. It does not follow from the putative reliability of self-inculpation that the exculpation of others is reliable. Indeed, the view that exculpations of others in these circumstances are reliable is not supported by experience.
Another conceptual difficulty in Deane J's appeal to fairness is this. If an admission is made, the tendering party must accept that what is to be received in evidence comprises not only the parts adverse to the opposing party, but also "everything … which is fairly connected with that admission". That is so even if that which is fairly connected is adverse to the tendering party. That is, the party against whom the evidence is tendered must suffer the adverse consequences of the admission, but may take advantage of everything fairly connected with it that is favourable to that party. The appellant's submission seeks to apply this advantage not only to the party against whom the admission is tendered but to a party against whom the admission is not tendered. The prosecution tendered LM's statements against LM. They were not tendered against the appellant and were not admissible against the appellant. In the result, the proposed extension based on "fairness" is anything but fair. "Fairness" applies to all parties to litigation. It must involve parity and reciprocity. It cannot be limited to the defence. At common law, instances in which the rules of evidence differ as between prosecution and defence are rare, apart from the burden and standard of proof. Suppose B makes an inculpatory statement exculpating A. If the prosecution's tender of the confession against B is said, because of its reliability, to make it "fair" to receive the exculpatory evidence in favour of A, why is the reverse not "fair"?
That is, why is it not "fair" for the prosecution to rely, against A, on a reliable confession by B inculpating A? For the reasons given above, it is unfair to take the latter course. And if that is unfair, it must be equally unfair to use B's confession to exculpate A.
Deane J's suggested exception creates a further unfairness. That unfairness is perhaps only transitory, but it has real effects. The police officers who conducted the interview of LM probably thought that the answers LM gave would be admissible evidence against him, but no-one else. They probed LM about what he did. They did not probe LM about what others did. Making a retrospective change to the common law rule, as the appellant suggests this Court should do, would operate unfairly to the prosecution. It would permit LM's evidence to suggest the appellant's innocence in circumstances where the police officers investigating the deceased's death did not, and the prosecution at the trial could not, examine that possibility by questioning.
There may be cases where unfairness arises because the prosecution tenders an out-of-court statement by one co-accused that points to the innocence of another. But in those cases the solution may not lie in the rules of evidence. It might be more appropriate to grant a separate trial or to turn to doctrines of abuse of process to cure the unfairness.
The risk of perjury
It is common now to downplay the risk of manufactured evidence as a reason for not widening admissibility. But it is a real danger in cases such as the present case. In Re van Beelen, Walters, Wells and Jacobs JJ used the following words, plainly written by Wells J, to describe it:
"The mere knowledge that an extra-judicial confession of crime could, in favourable circumstances, be received to exculpate an alleged offender, would, however, be likely, in our opinion, to tempt the less scrupulous members of our community to undertake clandestine operations of self-help. All that would be required by a guilty accused person would be the services of two or three accomplices and a person, known to all, who had died after the date of the alleged offence and who, theoretically, could have committed it. The accomplices, when called as witnesses, could then simply attribute a 'confession' to the deceased man, and the confession could be given artistic verisimilitude by inserting in it evidence of esoteric knowledge that had, in fact, come from the best of all sources - the offender. If there were not at hand a deceased person into whose mouth the confession could conveniently be put, the unavailability of a living person could, no doubt, be arranged by any one of a number of irregular methods - direct or indirect. Where serious crime was alleged, the motive for making such arrangements would be strong.
If answer be made that the foregoing observations are, mutatis mutandis, equally applicable to the alibi, evidence of which is not excluded, the reply may be offered that the alibi is at least closely and naturally associated with the accused himself and his connection with the case against and for him; it is usually susceptible of being tested and assessed with the aid of, and against, the case as a whole. The extra-judicial confession of a stranger who does not appear in the witness box, however, has the merit, from a guilty accused's point of view, that the circumstances of its making would, more often than not, be largely beyond the natural purview of the principal case on which issue had been joined, and would be likely to raise questions that, except for the connection of the text of the confession with the corpus delicti, related to facts and events not likely to be canvassed by the main body of evidence. Perjurious or lying defences would thus become dangerously easy to fabricate, and correspondingly difficult to expose."
Simpson J described that danger as "obvious" in R v Greatorex. In Bannon v The Queen, Brennan CJ said that if the common law were changed, "false confessions untested by cross-examination would bedevil criminal trials" and "[g]aol-house confessions allegedly made by prisoners who would decline to admit guilt testimonially would be a commonplace." The circumstances those remarks contemplate differ from those in this appeal. But the appellant did not attempt to deal with the risk of fabricated confessions that an exception to the hearsay rule might create. He tried to construct a rule limited to the particular circumstances of this appeal. However, any evidentiary rule which would benefit the appellant in this appeal would have to be structured so as to operate more widely as well.
Australian authority
The appellant acknowledged that most intermediate appellate courts in this country had not gone beyond the principles elucidated in Bannon v The Queen. But he submitted that the opposite was true in Queensland. Like the mother watching her son in the parade, the appellant submitted that Queensland was the only one in step, and that the common law in the other jurisdictions should be made to conform with Queensland.
The first of the Queensland cases is R v Zullo. It preceded Bannon v The Queen. The case concerned a hearsay statement by a person not charged with a killing that he was the killer. That person gave a different version when interviewed by police officers. He said that it was the accused and not he who was responsible. At the trial, he successfully claimed the privilege against self-incrimination in relation to his incriminating statement. The prosecution tendered the hearsay statement. The judge directed the jury:
"Before an out of Court confession by some other person would result in an acquittal of an accused, you would have to be satisfied that it had such substance, after scrutinising all the evidence relevant to it carefully, that you had a reasonable doubt as to the accused's guilt. If after scrutinising all the evidence relevant to that confession you reached the conclusion that there is no substance in it, or that it was not truthful, then you may conclude that nothing relating to that evidence causes you to have a reasonable doubt as to the accused's guilt."
On appeal, the accused claimed that this was a misdirection. The Court of Appeal noted that there was "authority of some strength" against the admissibility of the evidence. But it treated the evidence as admissible. No reasons were given beyond reliance on Holmes J's dissenting judgment in Donnelly v United States. The Court of Appeal said:
"The confession … was … to be considered by the jury for what they thought it was worth, and may very well have inclined the jury towards a 'not guilty' verdict, even if they were by no means convinced that it was truthful."
Their Honours also agreed that there had been a misdirection:
"If the jury were inclined to suspect that the … confession was spurious, that would not necessarily have been a justification for ignoring it when coming to their final conclusion."
But how can an out-of-court confession raise a reasonable doubt if the jury does not believe its truth? The Court of Appeal's assumption that the third party confession was admissible was obiter only. It was inconsistent with the Australian common law as received from England. Leaving aside short-run disagreements, there is but one Australian common law. Whether or not the common law rule in question - the hearsay rule - was of a kind which an intermediate appellate court ought to have departed from is a controversy which can be put to one side.
In R v K; Ex parte Attorney-General (Qld), the Court of Appeal followed R v Zullo as a case which "remains binding in Queensland." Whether it was right to do this, or whether it should not have been followed because it was only an obiter assumption inconsistent with the leading common law cases, is another controversy which may be put to one side. In R v K, the Court of Appeal's attention was not drawn to Bannon v The Queen, notwithstanding that it had been decided more than six years earlier.
However, when R v Martin was decided a few months later, Bannon v The Queen was drawn to the Queensland Court of Appeal's attention. The Court of Appeal held that the "Queensland line of authority" remained binding in Queensland until this Court overruled it. In its view, what this Court said in Bannon v The Queen was not necessarily inconsistent with the Queensland cases. The Court of Appeal did not explain this alleged congruity. In truth, Bannon v The Queen is inconsistent with the Queensland cases. Though no majority of this Court decided unequivocally against any future departure from the traditional law, their Honours made it plain that the law as it then stood contained no exception to the hearsay rule in favour of accused persons for confessions made by a co-accused or a third party. The Queensland cases reveal no basis on which they should be preferred. They should be overruled.
Impact of legislative developments
The appellant submitted that legislation largely corresponding with the Evidence Act 1995 (Cth) ("the Evidence Act") now existed in New South Wales, Tasmania and Victoria, and that the Northern Territory was preparing to follow suit. The Australian Capital Territory, formerly governed in large part by the Evidence Act, now also has its own legislation in very similar form. The appellant submitted that there is now a "consistent pattern of legislative policy to which the common law of Australia can adapt itself." The appellant took the quoted words from Esso Australia Resources Ltd v Federal Commissioner of Taxation. They appear in the discussion by Gleeson CJ, Gaudron and Gummow JJ of the questions whether and how the courts should use legislative developments as a guide for changing the common law. Those are large and significant questions.
Gleeson CJ, Gaudron and Gummow JJ quoted a passage from Lord Diplock's judgment in Warnink v J Townend & Sons (Hull) Ltd:
"Where over a period of years there can be discerned a steady trend in legislation which reflects the view of successive Parliaments as to what the public interest demands in a particular field of law, development of the common law in that part of the same field which has been left to it ought to proceed upon a parallel rather than a diverging course."
Their Honours then pointed out that Lord Diplock was speaking in the context of a nation which at the time had only one Parliament. Australia is a federation with nine Parliaments. It can be difficult to detect a "steady trend" or a "consistent pattern" in the conduct of State and Territory legislatures. Often they seek to maintain, so far as ss 109 and 122 of the Constitution will let them, a sturdy independence. Queensland, South Australia and Western Australia have not adopted the Evidence Act. Those polities are very large in area. More importantly, they are quite large in population. Their legislatures are entitled to arrange their court processes along lines which they perceive to be desirable. In Bannon v The Queen, McHugh J observed that to change the Australian common law by changing the common law of Victoria was not a course which should lightly be taken. Writing in 1995, just after the enactment of the Evidence Act, he said:
"The recent legislative activity in this field provides a sound reason for this Court proceeding cautiously when invited to alter the settled rule against hearsay evidence. If any change is to come about as the result of judicial law-making, it should only occur after the Court has had the benefit of full argument from counsel representing the States and the Commonwealth."
This Court heard no argument in this appeal from any counsel within this category except counsel for the respondent.
The appellant's submission that the common law should adapt itself to the Evidence Act raises the question - which part of it? The appellant mentioned s 65(1), (2)(b), (c) or (d) and (8).
Those provisions permit the reception of "representations". The appellant submits that LM made a "representation" exculpating the appellant. That submission is factually incorrect. And so far as the provisions of s 65(2)(b)-(d) require non-fabrication or reliability, the requirements are not satisfied. Another difficulty is that the Evidence Act is, if not a completely integrated and exhaustive code, at least based on a carefully thought out scheme. Its parts are interdependent to some degree. Any adaptation of the common law to conform to the Evidence Act must be workable. To ensure this, it may be necessary to recognise many more provisions as forming part of the common law than those to which the appellant referred - for example, many definitions in the legislation, and the s 67(1) procedural duties to give notice. It would not be easy to decide how far this recognition should go.
A further difficulty is that minds may well differ on the desirability of importing the Evidence Act into the common law. Tests for admissibility that turn on questions of "fabrication" and "reliability" are alien to common law approaches. In Vocisano v Vocisano, Barwick CJ said that "statements made on an occasion when they are unlikely to be concocted are [not] for that reason admissible." In Pollitt v The Queen, Brennan J said:
"[I]t is not desirable to subsume the res gestae principle under a general principle which would admit hearsay evidence when a trial judge believes that concoction was extremely unlikely".
And in Bannon v The Queen, Brennan CJ said the following position would apply if like requirements were imported into the common law:
"Admissibility would reflect no more than the judge's opinion of the fairness of exposing the accused to the risk of conviction on the hearsay evidence. That is not an appropriate power to vest in a trial judge who has not heard the declarant making the statement and ordinarily would not have seen the declarant. It is one thing for a trial judge to rule on the fairness of admitting an accused's confession when the accused was, ex hypothesi, the actor in the relevant events and is able to contest, by cross-examination and by testimony, the case advanced against him that is based on the confession; it is another thing to rule on the fairness of admitting the statement of another declarant when the accused was not involved in the making of the statement and is unable effectively to contest the case based on the statement." (footnote omitted)
The provisions of s 65(2)(b)-(d) are characterised by more safeguards than the regime Brennan CJ attacked. Nonetheless, those regimes share significant common ground. Sections 135 and 137 of the Evidence Act deal with the exclusion of evidence where its probative value is outweighed by the risk of prejudice to the accused.
Further, s 65(8) of the Evidence Act undercuts the safeguards in s 65(2). And the applicability of s 65(8) to the present circumstances is unclear. It applies where the defence tenders evidence. Here the prosecution tendered it.
The processes by which the common law might "adapt itself" to the Evidence Act are thus neither manifestly desirable nor simple. The judgments the Act makes and the techniques it adopts cannot be incorporated into the common law without a violent act of legislation, this time judicial.
English developments
Finally, the appellant relied on the fact that in England a confession by one co-accused is admissible in favour of another co-accused. This was the result of legislative change. The difficulties of adapting the common law of Australia to meet a foreign legislative change are even greater than adapting it to meet an Australian legislative change.
The English legislation was enacted after the decision in R v Myers. In general, that case favours the appellant's position. Evidently, the legislature thought the decision to be unsatisfactory. R v Myers was an appeal by the accused against the reception of her own out-of-court statements. Those statements were tendered by her co-accused. R v Myers thus did not raise the issue in the present case. Indeed, in numerous respects it is factually distinguishable from the present case. And as Lord Hope of Craighead pointed out, the certified question did "not … sufficiently analyse the issues". The appellant made only two passing references to R v Myers. He took the Court to no part of what their Lordships said. The appellant did not demonstrate, or seek to demonstrate, its suitability as a precedent to be followed in Australia.
Conclusion
The present common law in relation to hearsay exceptions should not be changed in the respects the appellant advocated. LM's evidence was not admissible in the appellant's favour. The trial judge's direction was correct. The appeal should be dismissed.