Director of Public Prosecutions Reference No 1 of 2017

SUPREME COURT OF VICTORIA

COURT OF APPEAL

S APCR 2017 0027

DIRECTOR OF PUBLIC PROSECUTIONS REFERENCE NO 1 OF 2017

---

JUDGES:	MAXWELL P, WEINBERG and BEACH JJA
WHERE HELD:	MELBOURNE
DATE OF HEARING:	26 October 2017
DATE OF JUDGMENT:	23 March 2018
MEDIUM NEUTRAL CITATION:	[2018] VSCA 69	First revision 23 March 2018: para [237]

---

CRIMINAL LAW – Trial – Prasad direction or invitation – Right of jury to acquit after close of Crown case – Power of judge to inform jury of right to acquit – Judge required to assess cogency of Crown case – Reference of point of law by Director of Public Prosecutions – Whether Prasad direction contrary to law – Whether direction constitutes invitation to acquit – Whether interference with division of function between judge and jury – Direction not contrary to law – R v Prasad (1979) 23 SASR 161, R v Pahuja (1987) 49 SASR 191, Doney v The Queen (1990) 171 CLR 207 considered – Criminal Procedure Act 2009 s 308.

---

APPEARANCES:	Counsel	Solicitors
For the Crown	Mr G J C Silbert QC with Mr B L Sonnet	Mr J Cain, Solicitor for Public Prosecutions
For the Acquitted Person	Mr O P Holdenson QC with Mr J P O’Connor	James Dowsley & Associates

MAXWELL P:

Summary

The Director of Public Prosecutions has referred a point of law to this Court under s 308 of the Criminal Procedure Act 2009. The Director invites the Court to declare that

the direction commonly referred to as the ‘Prasad direction’ is contrary to law and should not be administered to a jury determining a criminal trial between the Crown and an accused person.

The referral of the point of law followed the acquittal of a person on a charge of murder, the judge having given the jury a ‘Prasad direction’.

The ‘Prasad direction’ owes its origin to the eponymous case of R v Prasad,[1] in which King CJ said:
[1](1979) 23 SASR 161 (‘Prasad’).

It is, of course, open to the jury at any time after the close of the case for the prosecution to inform the judge that the evidence which they have heard is insufficient to justify a conviction and to bring in a verdict of not guilty without hearing more.[2]

It was within the trial judge’s discretion, his Honour said, to ‘inform the jury of this right’. Subsequently, in R v Pahuja,[3] King CJ said that this course should only be taken when

the judge is of [the] opinion that the evidence for the prosecution, although capable in law of supporting a conviction, is insufficiently cogent to justify a verdict of guilty.[4]

[2]Ibid 163.

[3](1987) 49 SASR 191 (‘Pahuja’).

[4]Ibid 201.

The Prasad direction has become an entrenched part of Australian law. Both at first instance and on appeal, the validity of the procedure has been assumed, without question or criticism. This is, as far as I am aware, the first occasion on which the legal foundation for the practice has been challenged.

There is no statutory provision, and no High Court authority, which would justify a conclusion that the practice is contrary to law. For reasons which follow, however, I consider that the practice should be ‘comprehensively disapproved’, as it has been on a number of occasions by the English Court of Criminal Appeal.[5]
[5]See, eg, R v Collins [2007] EWCA Crim 854 (‘Collins’); see [60] below.

What is described in Prasad as the jury’s ‘right to acquit’, seemingly exercisable of the jury’s own motion, depends in practice on the judge exercising a power to invite the jury to consider an acquittal. It is the judge who assesses the prosecution evidence and decides whether it is ‘sufficiently cogent’ to justify a verdict of guilty. If the evidence is assessed as ‘insufficiently cogent’, the judge then invites (or directs) the jury to consider acquitting the accused without hearing more.

In practice, as the English decisions observe, it is difficult for the judge to avoid giving the impression of inviting the jury to acquit. The very fact that the judge is directing the jury to consider a peremptory acquittal necessarily conveys the judge’s view that the state of the evidence is such as to warrant consideration of that (exceptional) course. Anything said by the judge about what has prompted the giving of the direction only reinforces that impression in the jury’s mind.

Thus understood, the Prasad direction constitutes ‘an interference with the traditional division of functions between judge and jury in a criminal trial’.[6] That division is expressed through the strict limits which Australian law imposes on the judge’s power to uphold a no case submission and direct an acquittal. A Prasad direction is only given in circumstances where there is a case to answer, such that the judge cannot direct an acquittal. Yet the giving of the direction will be seen by the jury as an invitation to do precisely that.
[6]Doney v The Queen (1990) 171 CLR 207, 215 (‘Doney’).

In some Victorian cases, as will appear, the decision to give a Prasad direction has rested on the judge’s adverse assessment of the credibility and reliability of prosecution witnesses. In at least one instance, the judge proceeded to inform the jury — in the course of giving the Prasad direction — of what he considered to be the deficiencies in the evidence of the key witness.[7] It is wholly anomalous for a judge to proceed in this way. As counsel for the Director pointed out on this reference, the High Court in recent years has underlined the role of the jury as ‘the constitutional tribunal for deciding issues of fact’[8] and has emphasised that questions of reliability and credibility are for the jury, not the judge.[9]
[7]R v Butler [Rulings 1–10] [2013] VSC 688 [165] (‘Butler’): see [82]–[83] below.
[8]R v Baden-Clay (2016) 258 CLR 308, 329 [65].
[9]IMM v The Queen (2016) 257 CLR 300; see also DPP v Wearn (a pseudonym) [2018] VSCA 39.

The Prasad practice is anomalous in other important respects. Its founding assumption is that the jury can exercise the ‘right to acquit’ with little or no assistance from the trial judge, and without hearing from either prosecution or defence. Yet the system of jury trial depends upon the jury having the assistance of the judge, in explaining ‘so much of the law’ and identifying ‘so much of the evidence’ as is necessary to enable the jury to determine the issues in the trial.[10]
[10]R v AJS (2005) 12 VR 563, 577 [54]–[55]; Jury Directions Act 2015 s 65; Criminal Procedure Act 2009 ss 234-5, 238.

In practice, of course, judges recognise that juries need assistance of this kind. It is for this reason that applications for Prasad directions are often refused.[11] In other cases, what is intended to be ‘minimal’ assistance can lead the jury into error, precisely because it falls short of what is needed.[12] In still other cases, as illustrated by the present case, the direction becomes quite lengthy and becomes a ‘mini summing up’.
[11]DPP v Gillespie [Ruling No 2] [2012] VSC 533 [12] (‘Gillespie’); DPP v Kocoglu [2012] VSC 184 [12]-[13] (‘Kocoglu’); R v White [No 8] [2012] NSWSC 472 (‘White’); R v Dickson; R v Issakidis [No. 10] [2014] NSWSC 1482 (‘Dickson’).
[12]See Seymour v The Queen (2006) 162 A Crim R 576 (‘Seymour’).

In my opinion, this practice should cease. Its disappearance would in no way diminish the right of an accused person to a fair trial. If anything, the right to a fair trial would be enhanced because — as the English cases point out — there is a real risk that the jury will react adversely to a judge’s ‘invitation to acquit’, and form a view adverse to the accused without having heard the judge’s summing-up or the final address of defence counsel. There is, separately, the importance of ensuring a fair trial for the prosecution who, likewise, have no opportunity under this procedure to address the jury about the evidence they have heard.

The justification for retaining the procedure is said to be that it can save time and expense. How much time is actually saved is difficult to assess, depending as it does on whether (but for the acquittal) the defence would have gone into evidence, and on how long counsel’s addresses and the judge’s summing-up would have been.

Importantly, of course, an early acquittal will ordinarily mean that the accused person is released from custody sooner. But, that apart, senior counsel for the acquitted person (who appeared to assist the Court as a contradictor) expressly disavowed any submission that the Prasad procedure was a protection against injustice.[13]
[13]Cf Natalia Antolak-Saper, ‘The role of directed verdicts in the criminal trial’ (2012) 21 Journal of Judicial Administration 146, 165–6.

Division of function between judge and jury

I referred earlier to the strict limits on the trial judge’s power to direct an acquittal. Axiomatically, there cannot be a directed acquittal if there is evidence which if accepted would sustain a verdict of guilty.[14] The evidence must be assessed ‘at its highest’. It is not for the judge considering a no case submission to form a view about the reliability of the evidence. That is a matter for the jury. Nor is the judge to consider whether a conviction would be unsafe. That question arises, if at all, only on an appeal against conviction.
[14]Doney (1990) 171 CLR 207, 215.

As King CJ made clear, a Prasad direction may be given even though there is evidence ‘capable in law of supporting a conviction’.[15] Indeed, that is the only circumstance in which a Prasad direction is sought. But in order to decide whether to give the direction, the judge must do the very thing which he/she cannot do in considering a no case submission, that is, assess the ‘cogency’ of the prosecution case. Not only does the judge thus intrude into the jury’s exclusive function of evaluating the evidence but, if a Prasad direction is then given, the giving of the direction inevitably conveys — or is perceived by the jury to convey — the judge’s view that the state of the evidence is such as to justify consideration of immediate acquittal.[16]
[15]Prasad (1979) 23 SASR 161, 163.
[16]Cf Pahuja (1987) 49 SASR 191, 201.

The Prasad direction distorts the jury’s assessment of the evidence in two significant respects. First, the jury are inevitably influenced by the (express or implied) indication from the judge that he/she has doubts about the prosecution case. Secondly, the jury will be expected to review the evidence having had, at best, a partial summing-up. They will not have heard submissions from the prosecution about how the evidence should be viewed, nor will they have had the conventional directions from the judge about how to assess witnesses or draw inferences.

Although this is said to be the ‘right’ of a jury to acquit, in reality everything depends on the judge’s exercise of discretion to inform the jury of that right. It is for the judge to decide, typically on application by defence counsel, whether the direction should be given. What routinely occurs (as in the present case) is that counsel seeks to persuade the judge — not the jury — of the weaknesses in the Crown case.

In the present case, defence counsel relied on Pahuja in submitting that the judge should conclude that the evidence for the prosecution was ‘not sufficiently cogent to justify a verdict of guilty’. Counsel relied for this purpose on an earlier decision of the same judge, in which his Honour had given a Prasad direction.[17] There the judge himself cited Pahuja in ruling that he would give the direction:

In Pahuja, King CJ notes that in his opinion the decision whether to inform the jury of its power to bring back a verdict of not guilty must be made by the trial judge in light of his assessment of the case. It is my opinion that the evidence may not be sufficiently cogent to justify a verdict of guilty. I note that that is a test different from the test which would apply had a submission been made on behalf of the accused that there was not a case for him to answer.[18]

[17]R v Smart [Ruling No 5] [2008] VSC 94 (‘Smart’).

[18]Ibid [13] (emphasis added).

I turn first to examine the limits on the power of a trial judge to direct an acquittal. As will appear, the decision of the Full Court of the South Australian Supreme Court in Prasad was the first in a series of Australian decisions, culminating in the High Court decision in Doney, which held that a trial judge had no power to direct an acquittal on the basis that, in the judge’s view, it would be unsafe for the jury to convict.

The power to direct an acquittal

In Australia, a trial judge has no power to direct an acquittal on the basis that it would be unsafe for the jury to convict. In England, at least for a time, the position was different. In 1977, in R v Mansfield,[19] the English Court of Criminal Appeal noted that a practice had developed in that jurisdiction during the 1960s of defence counsel:

inviting the judge at the end of the Crown’s case, to say that on the Crown’s evidence it would be unsafe for the jury to convict and accordingly the judge ought to withdraw the case from the jury.[20]

The Court observed that the making of such submissions appeared to have commenced following a change in the criminal appeal provisions.[21]

[19]R v Mansfield [1978] 1 All ER 134 (‘Mansfield’).

[20]Ibid 140.

[21]Ibid. Under the Criminal Appeal Act 1907, the Court of Criminal Appeal would not interfere to quash a conviction if there was ‘evidence upon which a reasonable jury could convict’. Following the enactment of the Criminal Appeal Act 1968, however, the question for the Court of Criminal Appeal became whether the verdict was unsafe or unsatisfactory.

The Court in Mansfield affirmed that defence counsel was entitled to make — and the judge was obliged to consider — a submission that:

some of the evidence was so conflicting as to be unreliable and therefore if the jury did rely upon it the verdict could be unsafe.[22]

If that submission were upheld, the judge should direct an acquittal. At the same time, the Court said, the judge was not entitled to direct an acquittal merely because the judge thought ‘that the main witnesses for the prosecution [were] not telling the truth’.[23] To do so would be ‘to usurp the function of the jury’.[24]

[22]Ibid.

[23]Ibid.

[24]Ibid, citing R v Barker (1975) 65 Cr App R 287, 288. See now R v Galbraith [1981] 1 WLR 1039.

In Prasad, the majority of the South Australian Full Court (King CJ and White J) rejected a submission, based on the English practice described in Mansfield, that the trial judge had a discretion:

to stop the case and direct a verdict of not guilty if he considers the evidence for the prosecution is so unsatisfactory that it would be unsafe to convict upon it.[25]
[25]Prasad (1979) 23 SASR 161, 162.

King CJ said:
It seems to me that to say that a judge can direct a jury to bring in a verdict of not guilty when there is evidence capable in law of supporting a conviction is to infringe one of the basic principles of trial by jury. It is fundamental to trial by jury that the law is for the judge and the facts for the jury. If there is no evidence which would justify a conviction then, as a matter of law, there must be an acquittal. That decision is for the judge and the jury must accept and act on his direction on that question of law. If, however, there is evidence which is capable in law of supporting a conviction, a direction to the jury to acquit would be an attempt to take from them part of their function to adjudicate upon the facts. That, as it seems to me, would be contrary to law.[26]

White J agreed:
I think that the reliability of the evidence rests with the jury, whether the supposed unreliability of the evidence arises from the supposed lies or supposed conflicts.
In this way, the time-honoured separate functions of judge and jury will be preserved. There still remains the duty to direct the jury to acquit, on a submission of no case, if there is no evidence on a particular element of the charge. But once there is some evidence on each element, its weight is for the jury.[27]

[26]Ibid 162–3.

[27]Ibid 171–172. For an analysis supporting the majority’s conclusion, see H. H. Glass, ‘The Insufficiency of Evidence to Raise a Case to Answer’ (1981) 55 Australian Law Journal 842, 845ff. See also R v Towers (1984) 75 FLR 77, 80.

In 1981, in Ling v The Queen,[28] the Tasmanian Court of Criminal Appeal unanimously rejected a contention, likewise based on Mansfield, that a trial judge had power to withdraw a case from the jury on the ground that the evidence before the jury was ‘so unreliable that a verdict based upon it would be unsafe’. Green CJ expressed agreement with what King CJ had said in Prasad, and added the following:

In my view, if a judge in the course of determining a submission that there is no case to answer, makes value judgments or embarks upon an assessment of the weight of the evidence, he is performing functions which by law should be performed by a jury and he is thus unlawfully altering the basic character of a criminal trial. … In a criminal trial the issue which is joined by a plea of not guilty is whether the accused is guilty or not guilty. The determination of that issue involves finding facts, applying the trial judge’s directions as to the law and sometimes, making value judgments. If a verdict is arrived at as the result of a judge instead of a jury performing those functions, it cannot be said that the issues joined have been tried by a jury.[29]

[28](1981) 6 A Crim R 429.

[29]Ibid 432 (emphasis added), see also 455–7 (Crawford J).

In Attorney-General’s Reference (No 1 of 1983),[30] the Full Court of the Victorian Supreme Court came to the same conclusion. The Court (Young CJ, Anderson and Gobbo JJ) said:

When a trial judge is considering at the close of the Crown case whether there is a case to answer, he should not be concerned whether a verdict of guilty based upon such evidence might be set aside by an appellate court as unsafe. He should only consider whether there is evidence upon which the accused could lawfully be convicted. If there is not, he should then ‘take the case away from the jury’ or more accurately, direct the jury that there is no evidence upon which they could convict and that accordingly it is their duty to acquit. Where, however, there is evidence upon which the accused could lawfully be convicted, the trial judge should so rule notwithstanding that he may think that a verdict based upon such evidence would be unsafe.[31]

[30][1983] 2 VR 410.

[31]Ibid 417 (emphasis added).

In 1989, in R v R,[32] the New South Wales Court of Criminal Appeal addressed the following question of law, referred to it by the Director of Public Prosecutions:

Does a trial judge have the power to direct a verdict of acquittal when the trial judge assesses the evidence is such that a verdict of guilty based upon it would be unsafe and unsatisfactory?[33]

The Court concluded that the question was to be answered in the negative. Gleeson CJ (with whom Maxwell and Wood JJ agreed) distinguished the ‘broader view’ (exemplified by Mansfield) from the ‘narrower view’, which holds that:
a judge should only direct an acquittal if he concludes that there is no evidence upon which a jury properly directed could properly convict.[34]

[32](1989) 18 NSWLR 74.

[33]Ibid 75.

[34]Ibid 77.

Another way of expressing the broader view, Gleeson CJ said, would be

to say that the trial judge’s powers extend to enabling the judge to prevent the occurrence of what, in the judge’s view, a Court of Criminal Appeal would regard as a miscarriage of justice and which, by hypothesis, the judge would regard as a miscarriage of justice.

There are attractions in the notion that a trial judge should have such a power. However, if it were so, it is not easy to see why it should, in logic, be confined to directing an acquittal in a case where the verdict would be regarded as unsafe.[35]

[35]Ibid 84–5.

His Honour concluded that the narrower view was correct:

As Dawson J observed [in Whitehorn[36]] there are aspects of the trial procedure by which our criminal justice is administered which operate to circumscribe the extent to which a trial judge may intervene in the interests of preventing injustice. To the aspects mentioned by Dawson J I would add the role of the jury as the tribunal of fact. It is that consideration, rather than any point of legal technicality, which underlies the decisions in cases such as R v Galbraith, R v Prasad, Attorney-General’s Reference (No 1 of 1983) and Mezzo. The legal principles which underpin the narrower view in turn flow from, or are particular manifestations of, that basic aspect of criminal justice. It is one thing to recognise in Courts of Criminal Appeal a power to review a jury’s determination of fact. It is another thing altogether to permit a trial judge to pre-empt such a determination.[37]

[36](1983) 152 CLR 657, 675.

[37]R v R (1989) 18 NSWLR 74, 85 (emphasis added).

In 1990, this question came before the High Court in Doney. The trial judge had rejected a submission by the defence that he should direct a verdict of not guilty on the ground that, although there was evidence sufficient to sustain a conviction, a verdict of guilty would be unsafe and unsatisfactory.[38] The Court of Criminal Appeal held that the conviction was not unsafe or unsatisfactory. The central argument on appeal to the High Court was that the applicant had been

denied the right to have the trial judge who heard and observed the [accomplice] witness, determine whether, in his view, a guilty verdict would be unsafe or unsatisfactory.[39]

[38]Doney (1990) 171 CLR 207.

[39]Ibid 210.

The Court (Deane, Dawson, Toohey, Gaudron and McHugh JJ) unanimously rejected the argument. Their Honours noted that the English practice as described in Mansfield was said to rest on the applicable criminal appeal provisions, and that the appellate courts of South Australia, Victoria and New South Wales had rejected the proposition that ‘a similar power in the trial judge derives from the common criminal appeal provisions in Australia’. The Court pointed out that the power of a court of criminal appeal to set aside a verdict as unsafe did not provide:

any basis for enlarging the powers of a trial judge at the expense of the traditional jury function. The power of a court of criminal appeal to set aside a verdict on the ground that it is unsafe or unsatisfactory, like other appellate powers, is supervisory in nature. Its application to the fact-finding function of a jury does not involve an interference with the traditional division of functions between judge and jury in a criminal trial.[40]

[40]Ibid 215. The then Chief Justice of the Supreme Court of the Northern Territory described this statement as ‘rather casuistical’: Keith J. Austin Asche, ‘The Trial Judge, the Appeal Court and the Unsafe Verdict’ (1991) 15 Criminal Law Journal 416, 421.

Their Honours said:

[T]he purpose and genius of the jury system is that it allows for the ordinary experiences of ordinary people to be brought to bear in the determination of factual matters. It is fundamental to that purpose that the jury be allowed to determine, by inference from its collective experience of ordinary affairs, whether and, in the case of conflict, what evidence is truthful.
…
It follows that, if there is evidence (even if tenuous or inherently weak or vague) which can be taken into account by the jury in its deliberations and that evidence is capable of supporting a verdict of guilty, the matter must be left to the jury for its decision. Or, to put the matter in more usual terms, a verdict of not guilty may be directed only if there is a defect in the evidence such that, taken at its highest, it will not sustain a verdict of guilty.[41]

[41]Ibid 214–5.

Against that background, I turn to examine what was said in Prasad itself about what has come to be known as the ‘Prasad direction’ or ‘Prasad invitation’. It should be noted that what was said on that subject was entirely obiter. The Court in Prasad was not concerned with any question of the jury’s ‘right’ to acquit. Rather, the issue was whether the judge had power to direct an acquittal on the ground that a guilty verdict would be unsafe.

The origin of the practice

When King CJ in Prasad identified the ‘right’ of the jury to acquit, he drew explicitly on English practice.[42] It would seem that the practice of the judge informing the jury of that right developed in England in the early 20^th century. It was not, however, until 1977 that the English Court of Criminal Appeal described its emergence in those terms.[43]
[42]Prasad (1979) 23 SASR 161, 163.
[43]See [35] below.

The key passage from the judgment of King CJ in Prasad begins as follows:

It is, of course, open to the jury at any time after the close of the case for the prosecution to inform the judge that the evidence which they have heard is insufficient to justify a conviction and to bring in a verdict of not guilty without hearing more. It is within the discretion of the judge to inform the jury of this right, and if he decides to do so he usually tells them at the close of the case for the prosecution that they may do so then or at any later stage of the proceedings: Archbold, Criminal Pleading and Practice 39th ed. (1976) p 332.[44]

[44](1979) 23 SASR 161, 163.

The two sentences in this passage were quoted almost verbatim from paragraph 577 of the 39^th edition of Archbold. The remainder of the paragraph, however, was not quoted. It read as follows:

However, in R v Young [1964] 1 WLR 717; 48 Cr App R 292, C.C.A., the court expressed the view that maybe the time had come, though the court did not desire to rule on it, when this practice should be only rarely, if ever, used, and that judges should more often take the responsibility themselves of saying to the jury that there is not satisfactory evidence upon which they could convict, and accordingly direct an acquittal. When a submission is made that a case should not be left to the jury it is a judge’s duty not only to consider whether there is some scintilla of evidence which in law could go to the jury but also whether it would be safe for a jury to convict on the evidence as it then stands: R v Hipson [1969] Crim L R 85, C.A. see post, 1356. These principles were re-affirmed in R v Falconer-Atlee (1974) 58 Cr App R 348, C.A. — ‘If the judge is not prepared to stop the case on his own responsibility, it is wrong for him to try and cast the responsibility of stopping it on to the jury’.

As can be seen, the omitted section sheds a very different light on the status of this ‘practice’ in England as at 1979. First, it reveals that the Court of Criminal Appeal had declared some 15 years earlier, in Young, that ‘this practice should be only rarely, if ever used’.[45] As will appear, that Court had expressed similar sentiments on two other occasions — R v Falconer-Atlee[46] in 1973 and Mansfield in 1977 — in the years before Prasad was decided. In Mansfield, the Court of Appeal had said:

There grew up in the two or three decades before the early sixties, and probably for a short time after the early sixties, a practice of inviting the jury to stop the case. This court, in R v Young, ruled that the practice was bad and should stop.[47]

[45]Young [1964] 2 All ER 480, 482.

[46](1974) 58 Cr App R 348 (‘Falconer-Atlee’).

[47]Mansfield [1978] 1 All ER 134, 140.

Secondly, the omitted passage reveals that, in the period 1964–77, the English Court of Criminal Appeal was proceeding on the basis that the trial judge did have power to uphold a no case submission on the basis that, although there was evidence to sustain a conviction, the judge considered that a guilty verdict would be unsafe. One of the stated reasons for the Court’s deprecation of the practice of inviting the jury to stop the case was that it amounted to an abdication by the judge of his/her responsibility to direct an acquittal on that very basis. As Archbold notes, the Court in Falconer-Atlee had said:

If the judge is not prepared to stop the case on his own responsibility, it is wrong for him to try and cast the responsibility of stopping it on to the jury.[48]

[48]Stephen Mitchell (ed), Archbold’s Pleading, Evidence and Practice in Criminal Cases (Sweet & Maxwell, 39^th ed, 1976) 332 [577].

In Prasad, the Chief Justice continued:

[The judge] may undoubtedly, if he sees fit, advise [the jury] to stop the case and bring in a verdict of not guilty. But a verdict by direction is quite another matter. Where there is evidence which, if accepted, is capable in law of proving the charge, a direction to bring in a verdict of not guilty would be, in my view, a usurpation of the rights and the function of the jury. I think that there is a clear distinction for this purpose between a trial before a magistrate or other court which is the judge of both law and facts and a trial by judge and jury. I have no doubt that a tribunal which is the judge of both law and fact may dismiss a charge at any time after the close of the case for the prosecution, notwithstanding that there is evidence upon which the defendant could lawfully be convicted, if that tribunal considers that the evidence is so lacking in weight and reliability that no reasonable tribunal could safely convict on it. This power is analogous to the power of the jury, as judges of the facts, to bring in a verdict of not guilty at any time after the close of the prosecution’s case. It is part of the tribunal’s function as judge of the facts. It cannot, consistently with principle, exist in a judge whose function does not include adjudication upon the facts.[49]

[49]Prasad (1979) 23 SASR 161, 163 (emphasis added).

As can be seen, King CJ viewed as ‘undoubted’ the power of the judge to advise the jury ‘to stop the case and bring in a verdict of not guilty’. Directing an acquittal, on the other hand, was ‘quite another matter’. If there was evidence capable in law of proving the charge, there could be no directed acquittal. Plainly enough, for a judge to advise (though not direct) the jury to acquit goes well beyond merely informing the jury of its ‘right’ to acquit.

The Chief Justice returned to this issue in 1987, in Pahuja. In that case, a Prasad direction had been sought and given at the conclusion of the prosecution case. Although no question about the direction was raised by the grounds of appeal, both the Chief Justice and Cox J took the opportunity to express their disapproval of what had occurred.

The Chief Justice said:

At the conclusion of the case for the prosecution, the learned trial judge, at the request of counsel for the defence, informed the jury of their right to stop the case and to bring in a verdict of not guilty. He addressed the jury at considerable length by way of explanation of the relevant law and during the course of his remarks conveyed quite clearly that he thought that they should stop the case. The jury decided that the case should proceed. I think that the procedure adopted calls for comment. The undoubted right of a trial judge to inform the jury of its power to bring in a verdict of not guilty at any time after the conclusion of the case for the prosecution, should be used sparingly and only when the judge is of opinion that the evidence for the prosecution, although capable in law of supporting a conviction, is insufficiently cogent to justify a verdict of guilty. Even in such a case, the judge should bear in mind that the evidence called by the defence might strengthen the prosecution’s case. The decision as to whether to inform the jury of its power must be made by the trial judge in the light of his assessment of the case and it would not be helpful to offer general advice as to the circumstances in which it would be proper to adopt that course. These observations, however, can be made. The fact that apparently credible evidence given by an alleged victim in a sexual case is uncorroborated is not of itself sufficient reason for taking that course of action. The judge should not interrupt the trial for detailed argument as to whether he should do so. He should do so as a matter of course or not at all. There should be nothing in the nature of a pre-trial summing up. If the jury cannot properly reach a decision at that stage on the law as explained in the opening, perhaps clarified by a concise correction or explanation if necessary, it is better not to embark upon the course of action at all. A partial summing up at that stage of the trial is a serious departure from the due course of trial and is to be avoided.[50]

[50]Pahuja (1987) 49 SASR 191, 201 (emphasis added).

Three points may be noted about this passage. First, King CJ again described as undoubted the ‘right’ of the trial judge to inform the jury of its ‘power’ to acquit. Secondly, his Honour made plain that it was the judge’s assessment of the ‘cogency’ of the prosecution evidence which must determine whether the judge exercised that ‘right’. Thirdly, his Honour emphasised that it would be a ‘serious departure from the due course of a trial’ for the judge to give any ‘summing up’ at that stage. As will appear, the Prasad procedure as adopted in the present case involved just such a ‘partial summing up’.

In the same case, Cox J said:

The direction, including an addendum that was occasioned by a juror’s question, occupied seventeen pages of the transcript. His Honour explained what his and the jury’s powers were, described briefly the elements of the charge, said something about reasonable doubt and the danger of convicting on the uncorroborated evidence of an alleged victim, and then dealt at considerable length with the Crown evidence and the nature of the case generally. He spoke of ‘conflicts here clearly on the evidence’ but that was to anticipate the defence evidence because there were certainly no critical conflicts between the Crown witnesses themselves. He gave the jury the choice of stopping the trial or having it proceed, but the general tenor of the direction clearly favoured a verdict of not guilty without waiting to hear the defence case.
…
[A] Prasad direction should be put to the jury quite simply and shortly. It is not the occasion for any more than a passing glance at the law and a brief reference to whatever feature of the evidence it is that has led the trial judge to give the direction — usually some serious weakness in the Crown case that has emerged during its presentation. I think, with respect, that this direction was far too long.
…
This is not intended to discourage the use of the Prasad procedure where the circumstances call for it. However, it is no small thing, where there is a case for the accused to answer, to encourage the jury to pre-empt their normal function.[51]

[51]Ibid 217–19 (emphasis added).

In 1995, in Dean v The Queen,[52] Cox J returned to the issue of the giving of a Prasad direction. His Honour said:

This Court has previously expressed its concern about too free a use of the Prasad direction to bring a trial to a stop: see R v Pahuja … In the present case the learned judge found that there was a case for the appellant to answer. There was no suggestion of the police witnesses being patently unsatisfactory or of the appellant being put to a lengthy but inevitably successful defence. Practically everything turned on the appellant’s state of mind and that was a subject on which the jury was likely in the circumstances to be very interested in hearing from the appellant himself, as indeed they showed by their answer. All the indicators, in short, pointed to a continuation of the trial in the normal way. A Prasad direction should not be given merely because the trial judge considers that the Crown case is not a very strong one. That would be to usurp the function of the jury.[53]
[52](1995) 65 SASR 234 (‘Dean’).
[53]Ibid 239 (emphasis added) (citations omitted).

The English Court of Criminal Appeal

The power of a trial judge to ‘invite’ the jury to acquit was considered by the English Court of Criminal Appeal on at least eight occasions between 1964 and 2010. As foreshadowed earlier, a review of those decisions reveals increasingly strong disapproval of the practice, but no suggestion that it is bad in law. On the contrary, as Weinberg and Beach JJA point out, the successive decisions appear to accept the ‘right’ of the jury at common law to acquit after the close of the Crown case.

The first in the series of decisions was Young, decided in 1964. In that case, the trial judge at the conclusion of the evidence had

invited the jury, if they so desired, to return a verdict of not guilty without proceeding any further; in other words to say that they were not satisfied that the prosecution had proved their case.[54]

In the event, the foreman told the judge that the jury were not all of that view, and the case proceeded. The judge summarised the prosecution evidence and then adjourned, without having summed up the defence case. Although, on the resumption, the judge finished his summing up, the jury foreman confirmed — in answer to a question from the judge — that the jury had made up their mind before coming into court and hearing the remainder of the summing up. The judge said to the jury:
You see, members of the jury, it is obvious you have decided … without hearing what I had to say about the [defence] case and after hearing what I said about the prosecution’s case.[55]

[54]Young [1964] 2 All ER 480, 481.

[55]Ibid.

The Court of Appeal quashed the guilty verdict. Delivering the judgment of the Court, Lord Parker CJ said:

It is quite clear that, although the jury had heard all the evidence, had heard speeches for the prosecution and particularly for the defence, yet, unless they were going to find him not guilty, it was their duty before returning a verdict of guilty to listen to what the judge had to say about the defence.[56]

Relevantly for present purposes, his Lordship added the following:
Before leaving the case, the court would like to say that this appears to be yet another case where difficulties have arisen through a practice whereby judges invite juries to stop a case if they feel that the prosecution case has not been proved. It is quite clear to this court that all the difficulties arose from that … It may be that the time has come — the court does not desire to rule on it — when this practice should be only rarely if ever used, and that judges should more often take the responsibility themselves of saying to the jury that it is not satisfactory evidence on which they could convict, and accordingly, direct an acquittal.[57]

[56]Ibid.

[57]Ibid 481–2 (emphasis added).

The second occasion was Falconer-Atlee, in 1973. There, the defence had submitted at the close of the prosecution case that there was no case to answer. The judge refused the no case submission. Then, as the Court of Appeal noted, the judge

turned to the jury who were still there, having listened to all this. He told them in an address extending over four pages of transcript why it was that he was leaving the case against the [accused] to them. With great respect, that was unwise, to say the least, in the circumstances, because it involved expressing, however tentatively, a view on the facts which it would have been very much better not to do. If he was going to leave the case to the jury, he should have left it saying no more than that there was evidence to go to the jury and it was for them to say whether or not the appellant should be convicted.[58]

[58]Falconer-Atlee (1973) 58 Cr App R 348, 356 (emphasis added).

Delivering the judgment of the Court, Roskill LJ said:

Not content with this address to the jury, the learned judge, having ruled that there was evidence to go to the jury, went on almost to invite the jury to stop the case. This Court has repeatedly said in recent years that this practice should not be followed. If a judge thinks that the case is tenuous, then, even though there is some evidence against the accused person, the judge, if he thinks it would be unsafe or unsatisfactory to allow the case to go to the jury even with a proper direction, should take upon himself the responsibility of stopping it there and then. If the judge is not prepared to stop the case on his own responsibility, it is wrong for him to try and cast the responsibility of stopping it on to the jury.[59]

[59]Ibid 357 (emphasis added).

As noted earlier, the Court was emphatic that the ‘invitation’ practice should not be followed. Instead, the judge should assume the ‘responsibility’ of stopping the case if he/she considered that a guilty verdict would be ‘unsafe or unsatisfactory’. That is, of course, precisely what a judge in Australia cannot do.

The third decision was Mansfield, to which reference has already been made. The Court there noted that the ‘invitation’ practice had ‘grown up’ in the 1930s or 1940s,[60] and that the Court in Young had said it should stop. In short, by the time Prasad was decided, three differently-constituted benches of the Court of Criminal Appeal had said — in increasingly strong terms — that the practice of inviting the jury to stop the case was bad and should cease.
[60]As to which, see Comment, ‘Request to Jury to Stop Case’ (1939) 83 The Solicitors’ Journal 951; and Carleton K Allen, ‘The Layman and the Law in England’ (1959) 2 Journal of the International Commission of Jurists 55, 62.

The next decision was in 1995 in R v Kemp.[61] In that case, the judge intervened during the presentation of defence evidence to ask the prosecutor whether, in the light of the evidence given, there was ‘any point in the prosecution trying to achieve a conviction’.[62] The prosecutor, having taken instructions, informed the judge that the prosecution wanted to continue.
[61](1995) 1 Cr App R 151 (‘Kemp’).
[62]Ibid 153.

The judge then brought the jury back into court and said:

Of course you have heard a good deal of evidence in the case now. In particular, you have heard from the last witness, who you may think was an independent witness in that public house that evening, who gave you an account which, again, you may think was entirely inconsistent with the account that you heard from the three young women.
Members of the jury, it is always open to you in a criminal case to say that you have heard enough of the case at any stage. (In certain cases the judge takes it upon himself to say that the prosecution have not called sufficient or adequate evidence to put before a jury, and a judge will direct an acquittal). That is not one of those cases and, indeed, Mr Farley never suggested it was, and he has called evidence before you, and it would be a case for you to decide.
Members of the jury, what concerns me is that because of the situation that has arisen, we would have to adjourn today and come back tomorrow to hear the rest of the evidence and counsel’s submissions to you and my summing-up, and if it be the case, and of course I do not know and may be you do not know, but you can have a brief discussion amongst yourselves, if it be the case that you have already formed the view that the prosecution are not going to be able to make you sure of what happened and you, of course, have been told and know that you have to be sure before you can convict. If that was the position you had already reached, you would be entitled to say that now. Of course, if you want to hear the further evidence and the rest of the case, then you shall do so, but I do not want unnecessarily to waste time if it be that you have come to that conclusion already.[63]

[63]Ibid.

In the event, the jury informed the judge that there were other matters they wanted to consider and that they wanted the case to proceed. The defence case was completed, and then there were final addresses and the judge’s charge. The jury deliberated and returned a verdict of guilty.

On appeal, counsel for the appellant referred to the statement in Falconer-Atlee that the ‘invitation’ practice should not be followed.[64] Delivering the judgment of the Court, McCowan LJ said:

We were of the impression that this practice criticised by Roskill LJ, once quite prevalent, had died out, but it appears that it is still very occasionally done. The danger, as [counsel for the appellant] pointed out, is that if the jury do not accept the judge’s invitation, something may go wrong with the verdict. He submits that juries are often keen to register their independence and do not like to feel that they are being pushed about by the judge. Indeed if they feel leant upon by a judge in favour of the defence, the result may be positively counter to what the judge intended. He submits that that is what may have occurred here.[65]

[64]See [48] above.

[65]Kemp [1995] 1 Cr App R 151, 154–5.

The Court said:

We have no doubt that the judge’s intentions were of the best. He clearly thought that an acquittal was inevitable and wanted to save time and money. However, he was wrong. An acquittal was not inevitable. Plainly the jury must have taken another view of the witnesses … than he himself took.[66]

The Court was not persuaded, however, that the verdict was unsafe. The judge’s intervention had not prevented the jury from discharging their function of deciding the case on the evidence before them.

[66]Ibid 155.

As had occurred 30 years earlier in Young, the Court in Kemp went on to make ‘a few general remarks on the problem that has arisen’.[67] They then set out an extract from the 1993 edition of Archbold which said, amongst other things, that the statement of Roskill LJ in Falconer-Atlee ‘should be ignored’. The Court in Kemp said:

We, for our part, do not agree with the suggestion that what Roskill LJ had to say in Falconer-Atlee should perhaps be ignored. We do not think it will always be very easy to distinguish between an invitation to acquit and a mere intimation of a right to stop the case. In this particular case we have no difficulty in concluding on the words used by the judge that it was an invitation to acquit, but it may not always be so easy to differentiate between the two.
Moreover, a jury may well use their common sense and read a mere intimation that they have a right to stop a case as an invitation to acquit, on the basis that a judge is not likely to be giving them the intimation unless he thinks that they should acquit. If a judge is going to do anything of this sort, and we do not encourage it, he should clearly, in our judgment, not go beyond a mere intimation of the right to stop, for fear that if he goes further and utters a clear invitation to acquit, the result may be as in the present case, leaving a convicted defendant with a grievance, however unjustified.[68]

[67]Ibid 155.

[68]Ibid 156 (emphasis added).

In 2000, the Court of Criminal Appeal had to consider the following question, referred to it by the Attorney-General following an acquittal:

Whether, in a prosecution which has otherwise been properly brought and where there is evidence to go before a jury, a trial judge has power to prevent the prosecution from calling evidence and direct the jury to acquit on the basis that he thinks a conviction is unlikely.[69]

[69]Attorney-General’s Reference (No 2 of 2000) [2001] 1 Cr App R 503.

The Court held that the question should be answered in the negative. Delivering the judgment, Kennedy LJ said:

Clearly, as it seems to us, the effect of the authorities is that other than in cases which are oppressive or vexatious or an abuse of the process of the court, the prosecution has a right to present its case. At the close of the prosecution case the defence may submit that the evidence does not disclose a case to answer, in respect of any or all of the counts on the indictment. If the submission succeeds the judge will direct the jury to acquit the accused. The test a trial judge should apply in determining whether there is a case to answer is set out in Galbraith. A trial judge is not precluded from entertaining and ruling on a submission of no case to answer, at the close of the defence case, but of course that is at a much later stage in the trial. At common law a jury is entitled to decide, at any stage after the prosecution has closed its case, that it wishes to acquit. The judge may remind the jury that this course is open to them. But he should not go further and issue an invitation to them to acquit (see Kemp and Falconer-Atlee).[70]

[70]Ibid 507 (emphasis added)(citations omitted).

In R v Speechley,[71] one of the grounds of appeal was that the trial judge had prevented defence counsel from reminding the jury of ‘their common law right to return a verdict of not guilty at any time after the close of the prosecution case’. The Court of Appeal concluded that the judge’s ruling was correct. Delivering the judgment of the Court, Kennedy LJ said:

It appears to be accepted that a jury does have a right to acquit after the conclusion of the prosecution case, but we know of no case in which that right has ever been exercised other than at the invitation of the trial judge, and we are satisfied that it can only be exercised if the trial judge invites the jury to consider exercising it. That is because it is the duty of the judge to ensure that the trial is fair, both to the defence and to the prosecution, and he must therefore be in a position to decide when the time has come for the jury to be permitted to reach a decision. In almost every case in order to do justice the jury needs to listen to all of the evidence, the submissions of counsel, and the directions in law of the judge. Otherwise, for example in a case of murder, the jury might acquit without ever realising that a verdict of manslaughter was a possible alternative. So if a jury is invited by counsel, or seeks of its own motion, to return a verdict before being asked by the judge to do so the judge should in our judgment direct the jury that it is his duty to ensure that justice is done, and that it is not open to them to return a verdict until he invites them to do so.[72]

[71][2004] EWCA Crim 3067 (‘Speechley’).

[72]Ibid [51] (emphasis added).

The Court went on to endorse the statement by Roskill LJ in Falconer-Atlee ‘that in practice a judge should not invite a jury to stop the case’.[73] The Court also repeated what had been said in Kemp about the difficulty of distinguishing ‘between an invitation to acquit and a mere intimation of a right to stop the case’.[74] The judgment continued:

Plainly, as the facts of [Kemp] demonstrated, a prudent judge will say nothing at all.
We accept that in some cases judicial silence may mean that a trial lasts longer than it need because, for example, the strengths or weaknesses of the prosecution evidence may depend upon the view to be taken of a witness’s reliability, and the judge cannot therefore accede to a submission of no case to answer simply because he regards the key prosecution witness as unreliable (see R v Galbraith (1981) 73 Cr App R 124), but it is worth remembering that in an exceptional case a judge can consider a submission of no case to answer, or decide of his own motion that there is no case to answer, as late as the close of the defence case, as was pointed out in R v Brown [2002] 1 Cr App R 46. We therefore find it difficult to envisage any circumstance where in reality it will be appropriate in the interests of justice for a judge to invite the jury to acquit. Experience shows that when such invitations have been issued in the past they have all too often led to difficulties.[75]

[73]See [48] above.

[74]See [56] above.

[75]Ibid [52]–[53] (emphasis added).

Once again, the Court here treated as established the existence of the jury’s ‘right to acquit after the conclusion of the prosecution case’. It is striking, nevertheless, that the judgment concluded with the statement that it was

difficult to envisage any circumstance where in reality it will be appropriate in the interests of justice for a judge to invite the jury to acquit.[76]

[76]Ibid [53].

The next decision, in 2007, was to very similar effect. In Collins, the appellant submitted that ‘the common law right’ of a judge to inform a jury that it could stop a trial at the close of the prosecution case could not survive the coming into force of the Human Rights Act 1998 (UK).[77] After referring to Falconer-Atlee, Kemp and Speechley, the Court said:

On the basis of these authorities we find it difficult to hold that the common law right of a jury to stop a case after the close of the prosecution case no longer exists. We think it strongly arguable that it cannot survive Article 6 but it is possible to envisage circumstances in which the jury could be reminded of this right in such a way as not to breach Article 6. However, in our judgment it is clear from the authorities that the practice of inviting a jury to exercise such a right has been comprehensively disapproved. At the very least it could only be exercised in the most exceptional circumstances and certainly not in a multi-handed case of some complexity.[78]

[77]Collins [2007] EWCA Crim 854 [44].

[78]Ibid [48] (emphasis added).

Noting that the earlier decisions did not ‘spell out all the specific dangers involved in a judge telling a jury that it has a right to stop a case’[79], the Court in Collins proceeded to identify eight ‘specific dangers’, which may be summarised as follows:
[79]Ibid [45].

1.The jury will be making a decision without the benefit of final addresses or any legal directions from the judge.
2.Just as the jury may provisionally conclude that the prosecution witnesses are not capable of belief, so they may provisionally conclude that the witnesses ‘are not just capable of belief but they are indeed telling the truth’, a conclusion which may be very difficult to displace.
3.Juries are often keen to register their independence and may react against ‘what might be perceived to be pressure from the judge to acquit.’
4.It may be very difficult for the judge to avoid giving the impression of inviting the jury to acquit, rather than merely informing the jury of its right to acquit.
5.The practice is ‘inherently more dangerous’ when there is more than one defendant and the evidence is complex.
6.The practice is unfair to the prosecution when it is given no opportunity to address either the judge or the jury and ‘correct a mistaken impression of its case.’
7.There may be particular dangers if the defence are contemplating not calling any evidence.
8.The prosecution have no right of appeal against an acquittal following the judge informing the jury that they have a right to stop the case.[80]

[80]Ibid [49].

In conclusion, the Court said:

We wish to emphasise the disapproval expressed by this court in Speechley of the practice of informing a jury of its right to stop the case. We find it very difficult now to envisage any circumstances when it would be appropriate for this practice to be adopted.[81]

[81]Ibid [59].

The last decision in the series is R v H(S), decided in 2010.[82] In that case, after the alleged victim of the criminal conduct had given evidence, the judge told the jury that, if they wished to do so, they could acquit without hearing more. In the course of the case, the judge had expressed strong views to the effect that the case was a ‘waste [of a] judge and jury’s time’.[83]
[82][2011] 1 Cr App R 14.
[83]Ibid 196 [48]

Delivering the judgment of the Court, Leveson LJ referred to the disapproval of the ‘invitation’ practice in Falconer-Atlee, Kemp, Speechley and Collins. The catalogue of ‘specific dangers’ set out in Collins was repeated, in full. Leveson LJ added:

There is also another reason which bites if the jury should stop the case. Although arguments have always been articulated as on the basis that fairness must be visited both on the defence and the prosecution, fairness to the prosecution is now well recognised as requiring a proper focus upon the legitimate rights and interests of victims and witnesses. Once there is a case to answer, they are entitled to know that the jury has heard the case through to its conclusion culminating in a fair analysis of the issues from the judge. The few words offering the jury the opportunity to stop the case do not provide this and can only be approached by the jury on the basis of the broadest of broad brushes.[84]

[84]Ibid 197 [50].

As can be seen, the view expressed by the Court of Criminal Appeal in the decades before Prasad — that the ‘invitation’ practice should stop — has been reinforced emphatically in the decades since. So powerfully, and so often, has the practice been disapproved in the United Kingdom that it must now be regarded as existing in name only in that jurisdiction. Not once but twice, the Court of Criminal Appeal has declared that it would be ‘difficult to envisage any circumstance’ in which it would be appropriate to take this step.[85]
[85]See [60] and [64] above.

In Australia, by contrast, there has been no such disapproval. On the contrary, although appellate concern has been expressed about ‘too free a use’ of the Prasad direction,[86] the legitimacy and utility of the procedure appear never to have been questioned. I turn to examine the Australian experience.
[86]Dean (1995) 65 SASR 234, 239; see [43] above.

The Australian experience

In R v Williams,[87] the Full Court of the Victorian Supreme Court was considering ‘the duty of the trial judge to direct an acquittal in identification cases.’ Relevantly for present purposes, the Court (Gobbo J, with whom Young CJ and Anderson J agreed) said:

Where there is some evidence sufficient to meet a no case submission, a judge may nonetheless have a discretion to invite the jury to acquit the accused.[88]

There was, however, no reference to Prasad, and the question of the ‘precise power’ of the judge to give the ‘invitation’ was left for future consideration.[89]

[87][1983] 2 VR 579.

[88]Ibid 584.

[89]Ibid.

In Antoun v The Queen, Gleeson CJ said:

The question whether there is evidence capable of supporting a verdict at a civil or criminal trial by jury is a question of law. As was explained in Doney, this is a different question from whether a jury ought to be warned about the probative value of evidence. It is different from the question whether a trial judge might properly inform a jury, at any time after the close of the prosecution case, of its power to acquit. And it is different from the question which confronts an appellate court when it has to decide whether a conviction is unreasonable. There is no advantage to be gained by blurring these differences. Keeping them in mind helps to avoid confusion.[90]

[90](2006) 80 ALJR 497, 501 (emphasis added).

In R v Reardon,[91] the New South Wales Court of Criminal Appeal rejected a ground of appeal which complained that the trial judge had failed to give a Prasad direction at the conclusion of the Crown case. Simpson J (with whom Hodgson JA and Barr J agreed) set out what King CJ had said in Prasad and continued:

In my experience, it has long been recognised in NSW that a judge may, in a suitable case, and in the exercise of his or her discretion, take the course outlined by King CJ. There is no rule that in any particular set of circumstances a judge is obliged to take that course or ought to take that course. The decision to do so or not to do so lies entirely within the discretion of the judge. I can think of no circumstances in which a refusal to give such a direction could result in a miscarriage of justice.[92]

[91](2002) 186 FLR 1.

[92]Ibid 32-3 [153]; see also R v Karounos (1995) 63 SASR 451, 484.

In that case, it was said for the appellant that the judge’s failure to give the direction had had the consequence that he was

disentitled to the exercise by the jury of its right to consider whether there was sufficient evidence to justify a conviction — a right which is an incident of a fair trial.[93]

Simpson J rejected this submission, saying:
Firstly, as I have noted above, the power to give a direction is entirely discretionary. It depends, inter alia, upon the assessment of the trial judge of the weight of the evidence against the particular accused. As was pointed out in Prasad, a direction by the judge to the jury as to the weight it should attribute to admissible evidence intrudes upon the jury function. The Prasad direction, in the terms stated in Prasad and as ordinarily given, carefully avoids trespassing upon that function. Nevertheless, to give such a direction can carry with it a suggestion to the jury that admissible evidence should be given little or no weight. A judge giving a Prasad direction has to tread a very fine line to avoid trespassing upon the jury function. A decision on an application for such a direction requires an assessment of the evidence in the Crown case but avoidance of conveying the results of that assessment.[94]

[93]Ibid 33 [156]

[94]Ibid 33 [157].

In Seymour, two accused were charged on the basis of joint criminal enterprise with offences against a female.[95] At the conclusion of the Crown case, counsel for both accused made application for a Prasad direction. Over the Crown’s objection, the judge gave the direction. The jury returned a verdict of not guilty for one accused but indicated that they would like the case against the other to continue.[96] He was subsequently convicted, and appealed.
[95]Seymour (2006) 162 A Crim R 576.
[96]Ibid 581 [15].

Hunt AJA (with whom Simpson and Rothman JJ agreed) said that the judge had been persuaded to give the Prasad direction because of his

strong view that [the complainant’s] evidence lacked cogency because of its inconsistencies in relation to a number of matters, that she had made assumptions rather than seen matters to which she had referred, and that this showed that her evidence was unreliable rather than dishonest’.[97]

[97]Ibid 593-4 [59].

Consistently what had been said in Prasad and Pahuja, there were no addresses, and only short legal directions, before the Prasad direction was given.[98] In his Honour’s view, the cause of the ‘seemingly inconsistent results’ of the Prasad direction was that the jury had been given no sufficient explanation of the nature of joint criminal enterprise.[99] It was clear, his Honour said, that the jury

acquitted the co-accused because it had not been explained to them how the part he played in the various events which took place in the apartment made him criminally responsible for the actions of the appellant.[100]

His Honour went on:
This case demonstrates the danger of giving a Prasad direction in circumstances where there may be a problem for the jury in understanding the real nature of the Crown case. In some cases, it may be possible for the direction to be expanded to make it clear how the Crown put its case, but it seems to me that to do so really negates the whole purpose of this procedure, which is premised on the jury being able, without the assistance of the trial judge or counsel, to judge the cogency of the evidence on which the Crown relies — without addresses and without a summing up.[101]

[98]Ibid 595 [62].

[99]Ibid 595 [64].

[100]Ibid 595 [65].

[101]Ibid 595 [66] (emphasis added).

I turn now to consider a series of first instance rulings by trial judges, in response to defence applications for a Prasad direction. The first is Smart, to which reference was made earlier.[102] In that case, Lasry J referred to Prasad and Pahuja and noted, in particular, the statements by Cox J in Pahuja that:
[102]See [18] above.

·a Prasad direction should be put to the jury ‘simply and shortly’;
·the giving of the direction was not the occasion ‘for any more than a passing glance at the law and a brief reference to whatever feature of the evidence it is that has led the trial judge to give the direction’.[103]

Finally, as noted before, his Honour relied on what was said by King CJ in Pahuja, in stating that he would give the Prasad direction because he had formed
the view that the evidence was not sufficiently cogent to justify a conviction.[104]

[103]Smart [2008] VSC 94 [4].

[104]Ibid [5].

In 2012, in Gillespie, Kaye J upheld a no case submission with respect to two charges but declined to give a Prasad direction on the remaining two charges. Referring again to what Cox J said in Pahuja, his Honour said:

Generally, for those reasons, trial judges are cautious in reaching the conclusion that a case is appropriate to give a Prasad direction.[105]

His Honour concluded that such a direction was not appropriate because the evidence of self-defence in relation to one charge was
not so convincing or straightforward as to make the case an appropriate one for a Prasad direction.[106]

In relation to the other charge, his Honour considered that
the jury would derive some benefit from a short summary of the various witnesses’ evidence in relation to that aspect of the case.[107]

[105]Gillespie [2012] VSC 553 [11].

[106]Ibid [12].

[107]Ibid [15].

In Kocoglu, Kyrou J also declined to give a Prasad direction. Noting that there had been an ‘extensive and effective cross-examination’ of the eyewitness to the alleged murder, his Honour said:

It would be open to the jury to conclude that he was not a witness of truth and to thereby reject all his evidence that implicated the accused.[108]

[108]Kocoglu [2012] VSC 184 [9].

After setting out what Lasry J had said in Smart, however, his Honour referred to a particular piece of evidence which rendered the giving of a Prasad direction inappropriate. His Honour said:

This is because it would be necessary for me to give the jury reasonably detailed directions about, among other matters, admissions, the difference between evidence relevant to credit and evidence relevant to facts in issue, the drawing of inferences and how the jury is to deal with any finding that there was a reasonable possibility that [the eyewitness] was involved in the death of the Deceased. In the absence of such directions, the jury would not have a proper understanding of the legal parameters within which to consider whether to return a verdict of ‘not guilty’ at this stage of the trial.
In other words, I do not believe that I could sensibly give to the jury a short and simple direction that makes only a passing reference to the law and a brief reference to the issue of [the eyewitness’s] credit.[109]

[109]Ibid [13]–[14].

In White,[110] R A Hulme J was presiding over the trial of three co-accused. Counsel for one of the accused sought a Prasad direction on the basis that the case against her was

a weak one and it would be appropriate to inform the jury of their right to return a verdict of not guilty at any stage from now.[111]

His Honour said:
A Prasad invitation is most appropriate in cases where little or no explanation of the law is required; the issues are straightforward; and there is no need for there to be any detailed review of the evidence. Such an invitation to the jury may most commonly be found in cases involving alleged personal or sexual violence, where the only evidence upon which the prosecution relies is that of the alleged victim and where there is a real issue as to whether the victim can be believed.[112]

[110]White [2012] NSWSC 472.

[111]Ibid [5].

[112]Ibid [6] (emphasis added).

Refusing the application, his Honour noted that a considerable body of evidence had been adduced by the prosecution, which had ‘some complexity’. In his Honour’s view:

[I]t would be very difficult for the jury to fully understand, amongst that volume of evidence, what aspects of it are relied upon by the Crown on the one hand and by [the accused] on the other in the case that concerns her. That is to say, it would be difficult for the jury to have that understanding, absent the assistance they would be given by hearing closing addresses by counsel and a summing-up by me.

In Butler, Croucher J rejected a no case submission. Citing Doney, his Honour said:

Whilst it was, in my view, a very weak case, it was a case that passed the ‘no case’ test.[113]

His Honour did, however, accede to the defence request for a Prasad direction. In his published reasons for giving the Prasad invitation, his Honour said that he had formed the view that the deficiencies in the evidence of the key Crown witness
were so compelling that it would not be open to properly instruct a jury to be satisfied beyond reasonable doubt ‘of [her] evidence of the events ...[114]

[113]Butler [2013] VSC 688 [159].

[114]Ibid [163].

Noting the references in Smart and Kocoglu to the requirement that the Prasad ‘invitation’ be given only if it could be put ‘simply and shortly’, his Honour said he had formed the view that he could give the jury ‘a relatively short and simple explanation’ of the relevant aspects of the law and of ‘the potential defects in the evidence.’[115] His Honour continued:
[115]Ibid [164].

Accordingly, in summary, I directed the jury:
a)that the prosecution case was now closed and at its highest point;
b)on the difference between my function as trial judge and theirs as jurors; on the onus and standard of proof; that any verdict must be unanimous; and on the elements of manslaughter (all of which I had explained in preliminary directions at the commencement of the trial);
c)that they could not find the accused guilty of manslaughter unless satisfied beyond reasonable doubt of the truth and reliability of the essential aspects of the account of Ms Harris;
d)that, even if satisfied beyond reasonable doubt of Ms Harris’s account, the accused could not be found guilty unless the jury were prepared to go on and draw certain inferences or conclusions from her evidence and other evidence so as to make out all of the elements of manslaughter;
e)that, because of the ten matters mentioned above (which I set out for the jury), it would be dangerous or unsafe to act on Ms Harris’s evidence;
f)that, as a result, they were entitled to return a verdict of not guilty now rather than continue with this trial;
g)that that was an unusual course but was nevertheless a course that sometimes occurs in trials;
h)that, if they wished to hear more of the case, they may do so
i)that, if the matter went on, I would be explaining these concepts – and many others – to them in more detail …[116]

[116]Ibid [165].

In Dickson, Beech-Jones J rejected both a no case submission and an application for a Prasad direction. Citing Seymour, his Honour noted that such directions were

usually reserved for simple cases in which the critical evidence against an accused appears to lack credibility or reliability.[117]

[117]Dickson [2014] NSWSC 1482 [5].

In the case at hand, his Honour said, the giving of such a direction ‘would occasion a serious injustice to the Crown’.[118] The principal evidence against the accused was documentary and the Crown had not yet had the opportunity to fully put its case about what the documents showed as to the accused’s involvement. His Honour said:

Thus the only possible fair manner in which a Prasad direction could be given in that respect at this point would be to effectively stop the trial and provide a relatively detailed mini summing up. Such a process is implicitly disapproved of in the authorities to which I have referred.[119]

[118]Ibid [6].

[119]Ibid [7].

In Rv Rapovski [Ruling No 3],[120] the accused was charged with attempted murder. Just prior to the close of the prosecution case, Beale J informed the parties that he was considering giving the jury a Prasad direction once the evidence in the case had closed. After inviting submissions as to whether it was appropriate for him to do so, his Honour ruled in favour of giving a Prasad direction

essentially because of issues concerning the credibility and reliability of the key prosecution witnesses.[121]

[120][2015] VSC 356 (‘Rapovski’).

[121]Ibid [2].

In his published reasons, his Honour set out at some length the evidence on which the prosecution had relied, noting the prosecutor’s submission that the case

had not descended to a point where the evidence lacked cogency or consistency so great as to warrant a Prasad direction.[122]

His Honour then set out a detailed analysis of the evidence which supported the defence case that someone other than the accused had fired the fatal shot. His Honour also set out five ‘significant issues’ in respect of the credibility and reliability of the key prosecution witnesses.[123]

[122]Ibid [34].

[123]Ibid [42].

Taking those matters into account, his Honour said, it was clear that

the threshold articulated by King CJ in R v Pahuja, namely that the evidence ‘although capable in law of supporting a conviction, is insufficiently cogent to justify a verdict of guilty’, has been met.[124]

His Honour rejected, however, a defence submission that the Prasad direction should include ‘an itemisation of the unsatisfactory aspects of the evidence’.[125] His Honour said:
In formulating the direction, I am mindful not to usurp the jury’s role as judge of the facts. Were I to itemise the problems with the credibility and reliability of the witnesses, it would not be a short direction and there is a real risk that the jury may be unduly influenced by ‘what the judge thinks’.[126]

[124]Ibid [43].

[125]Ibid [44].

[126]Ibid [46].

In R v Castaneda [No 3],[127] Wilson J upheld a no case submission on a charge of murder but declined to give a Prasad direction with respect to the alternative charge of manslaughter. On the murder charge, the judge concluded that, even taking the Crown case at its highest, there was no evidence capable of negativing self-defence.[128] On the manslaughter charge, however, the judge said:

I do not consider the Crown case to be a compelling one in any sense, but nor is it as frail and lacking in cogency as would militate in favour of the direction (or invitation, as it is sometimes termed) being given.
Additionally, concepts of self-defence and the reasonableness of the degree of force employed in response and the like are, in my opinion, not straightforward or, perhaps more properly, they are made complex by the way in which the law defines and applies them. The jury will in my view benefit from the addresses and summing up, even if there is no further evidence.[129]

[127][2015] NSWSC 1104.

[128]Ibid [58].

[129]Ibid [63]–[64].

R v Gant[130] concerned charges of art fraud with respect to three paintings. At the close of the prosecution case, Croucher J refused no case submissions on behalf of both accused but then took what his Honour described as ‘the rather exceptional course’ of giving the jury a Prasad direction.[131] His Honour explained the decision in his subsequent sentencing reasons as follows:

This option may be given to a jury in circumstances where a criminal case is thought to be so weak that the jury is concerned that a guilty verdict would be unsafe. Having heard all of the evidence to that point in the trial, I formed the view that it was not open to a properly instructed jury to exclude the reasonable possibility that the three paintings the subject of the charges were created by Mr Whiteley in 1988. If that was right, verdicts of not guilty had to follow.[132]

As his Honour noted, however, the jury took a different view and ultimately convicted the two accused. Those convictions were later quashed on appeal.[133]

[130][2016] VSC 662 (‘Gant’).

[131]Ibid [37].

[132]Ibid (citations omitted).

[133]Gant v The Queen [2017] VSCA 104.

In R v Johnson [No 3],[134] Button J rejected an application that he give the jury a Prasad invitation. His Honour said:

An essential practical precondition of me adopting that course would be my own personal assessment that the Crown case is in some way weak.[135]

His Honour said that he was ‘far from being’ of that opinion, regarding the case of murder as strong.[136]

[134][2017] NSWSC 572.

[135]Ibid [2]

[136]Ibid [12].

In R v Maitland [No 5],[137] Adamson J rejected a no case submission and refused to give a Prasad invitation. Citing Seymour and Dickson, his Honour said:

In the present case, the evidence is complex and detailed. The Crown case is a circumstantial one which depends on an assessment of the evidence as a whole. The jury has not been taken to all of the documentary evidence and can be taken not to have read it in its entirety or appreciated its significance to the Crown case.[138]

[137][2017] NSWSC 167.

[138]Ibid [46].

In my respectful opinion, the reasons given in Smart, Butler, Rapovski and Gant — all Victorian cases — expose the fundamental flaw in the Prasad procedure. In each of those cases, the judge ruled — or the defence had conceded — that there was a case to answer. On the authority of Doney, therefore, it was not within the power of the judge to stop the case by directing an acquittal. But, as the reasons reveal, in each of those cases the judge had come to the view — on the basis of his own assessment of the evidence — that the case should stop. The Prasad procedure enabled the judge to invite the jury to acquit, notwithstanding that he could not direct them to do so.

In Smart, the judge considered that the evidence ‘may not be sufficiently cogent to justify a verdict of guilty’.[139] In Butler and again in Gant & Siddique, the judge concluded that it would not be reasonably open to the jury to convict the accused — in other words, that a conviction would be unsafe. And in Rapovski, the judge’s consideration of factors adversely affecting ‘the credibility and reliability’ of the key witness had led him to conclude that the evidence was ‘insufficiently cogent to justify a verdict of guilty’.[140] In that case, his Honour declined to itemise the problems going to credibility and reliability, whereas in Butler the judge set out for the jury all of the matters said to affect the credibility and reliability of the key witness.[141]
[139]Smart [2008] VSC 94 [13].
[140]See [88] above.
[141]See [83] and [88] above.

I turn finally to what occurred in the present case, which further illustrates these difficulties. In this instance, the Court has the transcript of the direction as given to the jury. In the other cases, only the reasons for giving the direction have been published.

The present case

The accused was charged with the murder of her husband. It was alleged that she had caused his death by striking him on the head. At the close of the Crown case, defence counsel asked the judge to give the jury what counsel referred to as the ‘Prasad invitation’.

The judge responded in these terms:

[L]et me just step through it and tell you, as I go, what I think the difficulties are and you can address them. The jury would have to be informed that at this stage they would be entitled to, if they wish to, bring in a verdict of not guilty in relation to the charge of murder. I would have to, I assume, if I was going to do that, give them enough direction to understand that they would have to work their way through the elements of murder. When they got to element 3 they would have to consider whether they were satisfied that there was enough evidence to support proof of intention.

It was submitted that for that reason, the Prasad direction might better be described as a Prasad invitation, it being understood, of course, that it was not an invitation to bring in a particular verdict, but rather to consider the option of bringing in a not guilty verdict at an earlier stage, should the jury wish to do so.

It was submitted that contrary to the Director’s submission, the Prasad direction did not cut across the quintessential fact-finding function of the jury.

Senior counsel for the respondent submitted that, as the Director had conceded, the High Court in Doney v The Queen,[196] despite having referred to Prasad, had said nothing about the correctness or otherwise of King CJ’ dictum.
[196](1990) 171 CLR 207 (‘Doney’).

Senior counsel drew attention to R v Reardon,[197] a decision of the New South Wales Court of Criminal Appeal. He relied, in particular, upon a passage in the judgment of Simpson J where her Honour said, of a Prasad direction:

In my experience, it has long been recognised in NSW that a judge may, in a suitable case, and in the exercise of his or her discretion, take the course outlined by King CJ. There is no rule that in any particular set of circumstances a judge is obliged to take that course or ought to take that course. The decision to do so or not to do so lies entirely within the discretion of the judge. I can think of no circumstances in which a refusal to give such a direction could result in a miscarriage of justice.[198]

[197](2002) 186 FLR 1.

[198]Ibid 32–33 [153].

It was submitted that there was nothing in either of two recent High Court decisions, R v Baden-Clay[199] and IMM v The Queen,[200] that cast any doubt upon the lawfulness of the giving of a Prasad direction.
[199](2016) 334 ALR 234.
[200](2016) 257 CLR 300.

Turning to the English authorities upon which the Director relied, it was submitted that these cases should be approached with great care. Properly understood, they identified a number of reasons why, in many cases, it would be undesirable to give a Prasad direction. However, none of them stated categorically, or even by implication, that the practice was contrary to law, or that no such direction should ever be given.

It was submitted that the decision in Kemp, and the criticisms levelled by the Court of Appeal at the practice, were entirely justified having regard to the inappropriate terms in which the direction given in that case had been couched.

Plainly, the judge had entered into the arena by making his views of the credibility of the prosecution witnesses known, and by indicating that he personally preferred the evidence of a defence witness. The direction was so strongly worded that the Court of Appeal concluded that the jury would inevitably have considered that the judge considered the prosecution case to be worthless. What he said to the jury was not an invitation to consider whether they wished to hear more evidence, but a direct, and impermissibly blunt, incursion into the jury’s own fact finding responsibilities.

It was submitted that Speechley[201] involved a quite different issue to that of whether the practice should be continued. The relevant ground of appeal raised the question of whether defence counsel (and not the judge) could inform the jury of their right to acquit the accused at any time after the close of the prosecution case. Not surprisingly, that ground of appeal had failed. The criticisms levelled at the practice, which largely reiterated what had been said in Kemp, were all dicta.
[201]Speechley [2004] EWCA Crim 3067.

As for Collins,[202] it was submitted although the Court of Appeal spoke of the practice as having been ‘comprehensively disapproved’, it was noteworthy that it did not rule it out absolutely. It was acknowledged by Gage LJ that there might be circumstances in which such a direction would be appropriate when he said

[a]t the very least [the practice of giving the direction] could only be exercised in the most exceptional circumstances and certainly not in a multi-handed case of some complexity.[203]

[202]Collins [2007] EWCA Crim 854.

[203]Ibid [48].

It was submitted that this formulation differed only marginally, if at all, from that adopted in Pahuja which stipulated that a Prasad direction should only ever be given ‘sparingly’, and where otherwise appropriate.

As for the other dangers of giving such a direction mentioned in Collins, including the risk of a jury being ‘keen to register independence’, and the need to avoid being seen to be inviting the jury to acquit, these could easily be overcome by ensuring that any such direction was carefully worded.

It was submitted that the problem with R v H(S),[204] and its analysis of the dangers of giving such a direction was that, in that case the direction was couched in very poor terms. It made clear to the jury that, in the judge’s view, they should acquit. Indeed, he told the jury to retire for ‘two or three minutes’ to consider whether they wanted to hear more evidence.
[204]R v H(S) [2011] 1 Cr App R 14.

If there were any doubt as to the judge’s opinion of the weakness of the prosecution case, it would have been dispelled by his comments after the jury had brought in a verdict of not guilty. The judge described the case as having been a ‘scandalous waste of taxpayer’s money’.

Importantly, the Court of Appeal did not have to consider the validity, or otherwise, of a properly worded direction inviting the jury to consider its position at the close of the prosecution case.

In summary, it was submitted by senior counsel for the respondent that the English cases upon which the Director relied merely expressed strong reservations as to the giving of the Prasad direction, but did not go so far as to say that it would be unlawful to administer it.

Turning then to the Director’s alternative submission, based upon the supposed abrogation of the Prasad direction by recent Victorian legislation, senior counsel for the respondent submitted that this Court should reject that contention. He submitted that there was nothing in either the Criminal Procedure Act 2009 or the Jury Directions Act 2015 that was in any way inconsistent with the giving of a Prasad direction. He pointed to the fact that whenever the Jury Directions Act contemplated a departure from the common law of any kind, it said so specifically, and in terms. There was not the slightest indication in that Act that the legislature intended to abrogate the well-established practice of giving a Prasad direction, whether the power to do so stemmed from the common law, or merely from accepted rules of practice.

On a more positive note, senior counsel submitted that a properly worded Prasad direction could assist in the efficient administration of the criminal justice system by avoiding delay associated with unnecessarily prolonged trials. Moreover, it should not be assumed that there was no legitimate public interest, beyond that of saving time and expense, in bringing to an end at an earlier stage a prosecution that was almost certainly foredoomed to fail. At the very least, the accused should be entitled to have the strain of undergoing a highly stressful experience lifted from his or her shoulders once the jury determined that they did not wish to hear any more evidence. In addition, it should not be assumed that any defence case would necessarily be short, or that there would not be significant inconvenience to a number of witnesses, not to mention the jurors themselves, in allowing such a case to be drawn out.

It was submitted, in addition, that it was curious that the Director did not challenge the jury’s right to return a verdict of not guilty once the prosecution case had been closed. He merely submitted that it would be contrary to law to inform the jury that they had a right to do so. The question was asked, rhetorically, why would it be unlawful for a judge to inform the jury of a right which it was conceded they could legitimately exercise?

Conclusion

For the best part of 40 years, judges in this country have, where considered appropriate, given Prasad directions. As a general observation, such directions are carefully expressed, given sparingly and only in cases where there is a proper basis for that to be done.

The issues in this trial, as in so many cases like it, were relatively straightforward. Given brief, but adequate instruction as to the law, the jury were perfectly capable of deciding whether they wished to acquit the respondent at the end of the prosecution case, or whether they wanted to hear more evidence. Initially they said they wished to hear more. However, after having seen and heard the respondent give evidence, they decided at once to acquit. That was hardly surprising.

The judge decided, in the particular circumstances of this case, to give the jury a rather more detailed Prasad direction than some of the authorities might suggest would be sufficient. In fact, his Honour’s direction encompassed virtually everything that might be expected of a final summation. There is no suggestion that the verdicts of acquittal were anything other than entirely appropriate.

Nonetheless, the Director is aggrieved by the process that was adopted in giving the Prasad direction. He now seeks to have this Court determine that it was contrary to law to have done so.

When, after Galbraith,[205] the law changed in England making it possible for judges to direct acquittals in cases where they considered that any conviction would be likely to be set aside on appeal as being unsafe or unsatisfactory, the need for an equivalent of a Prasad direction was considerably diminished.
[205]Galbraith [1981] 1 WLR 1039.

The Director contends that any practice of giving such a direction is contrary to law because it upsets the balance between the role of a trial judge, and the responsibility of the jury in determining questions of guilt or innocence. The Director’s argument is put as a matter of high principle.

An important counter to the Director’s argument is the need for this Court, in answering the question posed in the reference, to have regard to comity, and the need to respect the views of other intermediate appellate courts unless they are considered to be plainly incorrect. If there is but one common law of Australia, it cannot be that it can legitimately take one form in one State, and another in a different State.

It is interesting to note that in his classic work Trial by Jury,[206] Sir Patrick Devlin wrote of the control which judges properly exercised over juries, throughout the history of English criminal trials. In 1956, when the first edition of that text was published, nothing at all was said about the right of a jury to stop a trial, or to be told of their right to do so. Nor did that ‘right’ find its way into either the second or third editions of that work in 1960 and 1966 respectively.
[206]The Honourable Sir Patrick Devlin, Trial by Jury (Stevens & Sons Limited, first published 1956, 1966 ed).

Although each edition contains an extensive discussion of the vital importance of the summing up, as a key component of trial by jury, there is one particular aspect of the treatment of that subject that is of tangential relevance to the issue before this Court. The learned author pointed out that, in England, the tradition was that a judge was permitted to express his opinion on questions of fact freely, if he wished to do so. The only limitation was that the judge must not comment on any point unfairly,[207] and must make it clear to the jury that, on questions of fact, which were entirely for them to determine, they were free to give such weight as they chose to any comments he might make.[208]
[207]Sir Patrick referred to what he described as an ‘extreme example’. In a case that was passed by the Court of Criminal Appeal in 1910, a judge’s comment to the jury that ‘[the accused] practically stands convicted by the evidence of the prosecution. You must do your duty.’
[208]See R v Hepworth (1910) 4 Cr App R 128. See also the extraordinary case of Stephen Arrowsmith, accused of the rape of a young girl, tried in 1678 at the Old Bailey. Chief Justice Scroggs summed up strongly for conviction, but the jury were reluctant and ultimately proffered a verdict of acquittal. The Recorder of London who, for some reason, had taken over the conduct of the trial, questioned the jurors about their thinking, explained to them why they were mistaken and sent them out to deliberate again. Even then, the jury were unconvinced. The Recorder then recalled the key prosecution witnesses and had them repeat their testimony on oath. Eventually, the jury were worn down and convicted.

Comments of an extreme nature of the kind that were sometimes made in favour of conviction would surely not pass muster today. To a lesser degree, the same is true of strong comments in favour of acquittal, though such comments are still sometimes made. It is rarely suggested that comments of this nature intrude impermissibly into the function of the jury, or that they constitute an impermissible usurpation of their role.

It is also interesting to note Professor John H Langbein’s remarkable study of jury trials in England.[209] The author observes that there is ample evidence throughout history of judges having commented both for and against the accused, sometimes in the strongest of terms.
[209]John H Langbein, The Origins of Adversary Criminal Trial (Oxford University Press, 1^st ed, 2003).

Professor Langbein points out that sometime in the second half of the 18^th century, there was a change in attitude among judges regarding the appropriateness of judicial comment on the facts. They continued to direct verdicts of acquittal as matters of law, but became more circumspect regarding advising juries on the merits.

In the United States, where the judiciary had been tainted by its association with imperial administration in late colonial times, judges had from an early stage been stripped of their powers of judicial comment. Wigmore described the United States approach as a departure from the orthodox common law rule which permitted comment. He said that this departure had thereby done more than any other thing to impair the general efficiency of jury trial as an instrument of justice.

Professor Langbein identified a number of cases in which trial judges in England commented forcefully in favour of the accused, sometimes without success. For example, in one case, a judge struggled to force a jury to acquit an accused charged with theft. The judge told the jury at the end of the prosecution case, and before addresses or summing up, that there was, in his view ‘a chasm in the evidence’ such that ‘I do not think it is material to call on the prisoner to make his defence, therefore, I think the prisoner must be acquitted’. Eventually, and after much cajoling, the jury did as the judge wished them to do.

When a jury returned a verdict, that the judge thought was against the weight of the evidence, or with which he otherwise disagreed, the judge could refuse to accept the verdict, question the jurors about it, and give them further instruction before requiring them to reconsider the matter. This could be done at any stage before the verdict was recorded, and irrespective of whether the verdict was one of guilty or not guilty.

From the late 18^th century, as Langbein notes, a trial judge’s authority over a jury verdict was gradually weakened. Judges moderated their use of the power to comment upon the evidence, and the practice of rejecting verdicts became contentious. Professor Langbein argues that this move away from judicial control to greater ‘adversarialism’ led to the development of a new system of jury control whose emphasis was on preventing jury error. This involved tightened control over the admissibility of evidence, resulting in a transformation of trial procedure.

Once again, for what it is worth, Professor Langbein’s study makes no mention at all of the practice of informing jurors of their right to stop a trial after the conclusion of the prosecution case.

The Director would no doubt say that Professor Langbein’s work supports his contention that the right to stop a trial never truly existed. Senior counsel for the respondent would say that the work proves nothing more than that rules of practice sometimes develop into rules of common law, and that the Prasad direction provides a perfect illustration of this.

Prior to Prasad, the position in this country regarding directed acquittals was, in some respects, the very antithesis of what had been taking place in England for some time. The practice in Victoria, for example, was that a verdict of not guilty might be directed if there were ‘but a scintilla of evidence’.[210]
[210]Doney (1990) 171 CLR 207, 213.

After 1966 a more ‘robust’ view developed in England to the effect that a trial judge should stop a trial if, in his or her opinion, a verdict of guilty would be ‘unsafe or unsatisfactory’.[211]
[211]See, eg Falconer-Atlee (1973) 58 Crim App R 348 and Mansfield (1977) 65 Cr App R 267.

It was thought for a time, in this country, that a similar power in the trial judge derived from the common criminal appeal provisions in Australia, though these were expressed in different terms to the 1966 United Kingdom legislation. That approach was expressly rejected in South Australia in Prasad, and in this State in Attorney-General’s Reference (No 1 of 1983).[212]
[212][1983] 2 VR 410.

In Galbraith[213] the English Court of Appeal made it clear that Mansfield at least had been wrongly decided, and should no longer be followed regarding directed acquittals. The Court stated:

(1) If there is no evidence that the crime alleged has been committed by the defendant, there is no difficulty. The judge will of course stop the case. (2) The difficulty arises where there is some evidence but it is of a tenuous character, for example because of inherent weakness or vagueness or because it is inconsistent with other evidence. (a) Where the judge comes to the conclusion that the prosecution evidence, taken at its highest, is such that a jury properly directed could not properly convict upon it, it is his duty, upon a submission being made, to stop the case. (b) Where however the prosecution evidence is such that its strength or weakness depends on the view to be taken of a witness’s reliability, or other matters which are generally speaking within the province of the jury and where on one possible view of the facts there is evidence upon which a jury could properly come to the conclusion that the defendant is guilty, then the judge should allow the matter to be tried by the jury.[214]

[213][1981] 1 WLR 1039.

[214]Ibid 1042.

In Doney,[215] the High Court specifically endorsed proposition (2)(b) in the passage from Galbraith set out above. However, it rejected proposition (2)(a), as involving too great a usurpation of the role of the jury by a trial judge.
[215]Doney (1990) 171 CLR 207, 214–5 (citations omitted).

The High Court stated the relevant principles to be applied on a no case submission in the following terms:

The question whether, in the words used in Galbraith, evidence has a ‘tenuous character’ or an ‘inherent weakness or vagueness’ may raise, but is not restricted to, the question whether the evidence is truthful. Quite apart from any question of truthfulness, there may be something in the nature of the evidence that brings its probative value into question so that the trial judge must consider whether some warning should be given. And, as earlier noted, there may be rare cases in which it will be necessary to consider whether, although the evidence was not initially excluded as a matter of discretion, it should be withdrawn from the jury’s consideration.
Evidence that attracts a warning is evidence that has been adjudged, either generally or in the particular case, as having probative value such that, subject to warning, it can be taken into account by the jury in its deliberations. Assuming an appropriate warning, the weight to be given to that evidence is as much a matter to be determined by inference based on the jury’s collective experience of ordinary affairs as is the question whether evidence is truthful. And, of course, the same is necessarily true of evidence that does not require a warning.

It follows that, if there is evidence (even if tenuous or inherently weak or vague) which can be taken into account by the jury in its deliberations and that evidence is capable of supporting a verdict of guilty, the matter must be left to the jury for its decision. Or, to put the matter in more usual terms, a verdict of not guilty may be directed only if there is a defect in the evidence such that, taken at its highest, it will not sustain a verdict of guilty.
It is necessary only to observe that neither the power of a court of criminal appeal to set aside a verdict that is unsafe or unsatisfactory (as to which see Whitehorn, Chamberlain v The Queen [No 2] and Morris v The Queen) nor the inherent power of a court to prevent an abuse of process (as to which see Jago v District Court (NSW)) provides any basis for enlarging the powers of a trial judge at the expense of the traditional jury function. The power of a court of criminal appeal to set aside a verdict on the ground that it is unsafe or unsatisfactory, like other appellate powers, is supervisory in nature. Its application to the fact-finding function of a jury does not involve an interference with the traditional division of functions between judge and jury in a criminal trial. Nor does the existence in a trial judge or a court of powers to stay process or delay proceedings where the circumstances are such that the trial would be an abuse of process.
The Court of Criminal Appeal was correct in upholding the trial judge’s ruling that he had no power to direct the jury to enter a verdict of not guilty on the ground that, in his view, a verdict of guilty would be unsafe or unsatisfactory.[216]

[216]Ibid.

Even after Galbraith had been decided, when the earlier more highly interventionist approach that had been favoured in Falconer-Atlee and Mansfield had been rejected, there was still ample scope, in England, for trial judges to direct acquittals in cases they regarded as particularly weak.

After Prasad in South Australia, and Attorney-General Reference [No 1 of 1983] in this State, a more rigorous approach to directed acquittals was adopted. Doney stamped the High Court’s imprimatur on that process. Once it became apparent that judges could no longer direct acquittals in what were perceived to be weak cases, the basis for retaining the Prasad direction, as a residual power, might be thought to have become all the more compelling. Certainly, the criticisms levelled at the practice of giving such direction in England became less cogent, so far as this country was concerned.

For that reason senior counsel for the respondent argued that the English authorities upon which the Director relied should be seen in context, and against the background of the very different approach to directed acquittals currently applicable in that country.

In our opinion, senior counsel’s submission should be accepted. There is no reason, in principle, why trial judges should not continue to give appropriately worded Prasad directions, provided that it is understood that they are to be given only rarely, and where the circumstances make it proper to do so.

Before giving such a direction, the trial judge must form the view that the prosecution case, considered as a whole, though sufficient to be left to the jury, is particularly weak. The case must be one where the jury will be able, without the assistance of closing addresses, still less a full judicial charge, to make a sensible assessment of whether, without hearing further evidence, an acquittal is the just and appropriate verdict.

A Prasad direction should not ordinarily be given in a case of any significant complexity. It should almost never be given in a case involving more than one accused. Any such direction should be accompanied by adequate instructions as to the relevant substantive law, and any appropriate and necessary warnings of an evidentiary kind.

If we may say so, the judge in the present case gave an impeccable Prasad direction. The jury were fully armed with all they needed to know in order to make sense of a simple allegation, and a simple response to that allegation. The prosecution was in no way disadvantaged by not having the opportunity to address when the direction was first given. By the time the jury actually came to consider their verdict, they had heard from the accused. And of course, they had been given a more than adequate charge by the judge as to the elements of the offence, the alternative charge, and all relevant matters of evidence.

We should say that we reject the Director’s submission that even if the Prasad direction had been part of the common law for many years, that is now no longer the case. The Director’s submission, that the provisions of the Criminal Procedure Act and/or the Jury Directions Act are inconsistent with the continued use of a Prasad direction, is strained and unconvincing.

The point of law raised by the Director on the reference should be answered as follows:

The giving of a Prasad direction, in appropriate circumstances, is not contrary to law.

- - - - -

[2018] VSCA 69

23 March 2018