Benign irrationality

Juries don’t give reasons, so why should judges?

The latest case on the proviso, the High Court’s decision in AK v Western Australia [2008] HCA 8 (26 March 2008), is another example of when an error at trial amounts to a “substantial” miscarriage of justice. Miscarriages of such magnitude cannot be cured on appeal by application of the proviso. In AK the trial had been by judge alone, and the judge had not given reasons for his verdict. Statute required that reasons be given. This failure was held, by a 3-2 majority, to be a substantial miscarriage of justice and a new trial was ordered.

Two majority judgments were delivered. Gummow and Hayne JJ jointly held that Weiss v R (blogged here 16 January 2006) and Wilde v R [1988] HCA 6 were not exhaustive of the situations that can give rise to a substantial miscarriage of justice. Here, the failure to give reasons for the verdict meant that the trial was not conducted according to law and that the miscarriage was therefore substantial (para 58). It was not to the point to ask whether the evidence supported the verdict.

This does not quite explain why the miscarriage was “substantial” as opposed to one that could be cured by the proviso if inspection of the evidence showed that the verdict was reasonable. Minor errors can mean that a trial was not according to law without it being necessary to quash the conviction. The other majority judgment, by Heydon J, went into the meaning of a substantial miscarriage in more detail.

The judgment of Heydon J is a forceful reminder of the advantages of trial by jury, and the resulting need to compensate for loss of those when trial is by judge alone. Footnote 75 is well worth a glance, for phrases in derogation of juries, eg many jurors are “unaccustomed to severe intellectual exercise or to protracted thought”. But juries bring a “benign irrationality” (para 97) to the proceedings. Quoting from Lord Devlin’s Trial by Jury (revised ed, 1966), Heydon J lists the five advantages of jury trials (para 93 – 97), and holds that it is necessary closely to observe the safeguards provided in relation to judge alone trials (para 98). The present case was one of extreme non-compliance with the requirement for reasons, which went to the root of the proceedings (para 109).

Within this framework, Heydon J mentions the power of juries to return perverse verdicts (para 97, and see blog entries for R v Wang 14 February 2005; R v Wanhalla 25 August 2006; R v Krieger 26 October 2006), the dangers in fact-finding by judges (para 101), the twin safeguards for the accused in the burden and standard of proof and the need for jury unanimity or a very substantial majority (para 102), the mental discipline imposed on the judge by the requirement for reasons (para 103 – 105 and 108), and the advantage that appellate courts have in ascertaining the appropriate inferences from the primary facts that have been determined by the judge in the context of the evidence that has been given (para 106, 107).

This approach to the question of how to identify a miscarriage that goes to the root of the proceedings takes us further than did the joint majority judgment, by saying that an example of this sort of miscarriage is one that, as here, prevents the appellate court from carrying out the protective function that is designed to compensate for the loss of the jury.

But is this case an example of that sort of miscarriage? The dissenting judgment of Gleeson CJ and Kiefel J acknowledges (para 17) that the appellant correctly pointed to the breach of the statutory requirement that the judge must give reasons for the verdict. Nevertheless, the magnitude of this error had to be assessed by its effect on the verdict. Here, the issue was narrow: who had committed the offences, the defendant or his brother? There was no evidence suggesting the brother was involved, and the judge’s finding that the offender was the defendant was supported by the objective circumstances (para 27). The absence of reasons was not an obstacle to application of the proviso here. So the dissenters were able to carry out the function of identifying the appropriate inference from the evidence that had been given, without being hampered by the absence of reasons for the verdict.

It is thus not necessarily persuasive to argue that an appellate court must have the judge’s reasons before it can carry out its role of compensating for the loss of a jury. Also questionable are the advantages that Lord Devlin attributed to juries. His prose was, of course, a fine example of the kind of eloquent rhetoric that now seems rather florid, and each of his propositions needs to be tested by experiment.

This is not to suggest that non-compliance with a statutory requirement to give reasons can be brushed aside. There has been a recent example of a seemingly technical error in procedure rendering a trial void (see blog on R v Clarke 7 February 2008 but contrast with Ayles v R 29 February 2008). The upholding of formal constraints on the exercise of power has its place, but it is fair to ask whether every failure to give reasons will inevitably amount to a substantial miscarriage of justice.

Truth and consistency

R v Stirling [2008] SCC 10 (14 March 2008) is an illustration of the common law’s treatment of the admissibility of a witness’s prior consistent statement. The important point has always been that such statements are not admissible as proof of their contents, but only as proof of consistency, and as such they go to the weight that the fact-finder may give to the witness’s testimony.

Central to this appeal is the idea that if the judge had used the prior consistent statement as evidence of its truth, then he would have made a serious error:

“[7] … it is impermissible to assume that because a witness has made the same statement in the past, he or she is more likely to be telling the truth, and any admitted prior consistent statements should not be assessed for the truth of their contents. As was noted in R. v. Divitaris (2004), 188 C.C.C. (3d) 390 (Ont. C.A.), at para. 28, “a concocted statement, repeated on more than one occasion, remains concocted”…”

Further, para 12, the prior statement goes to the witness’s credibility in a general sense; it would be artificial to try to separate out the witness’s credibility on the topic(s) covered in the prior statement from the witness’s credibility generally.

Personally, I wouldn’t have thought that the distinction was that difficult, but Bastarache J, delivering the judgment of the Court, said that such a separation would be “impractical and artificial” (para 12).

It could also be said that the distinction between use of the prior statement for the fact it was made, as opposed to as proof of the truth of what it asserts, is impractical and artificial. It must be acknowledged that this distinction is well established at common law, and (broadly speaking) it applies to prior statements whether they are consistent or inconsistent with the witness’s testimony. It has also, I must admit, been shown to be workable in practice, although arguably more convincingly so in judge-alone cases than where the distinction has to be explained to a jury. But the artificiality of the distinction is easy to see in respect of inconsistent statements: if the fact-finder rejects the witness’s testimony, there is then no evidence – through the witness – on the point if the prior statement is not evidence of its own truth. The prior inconsistent statement cannot then be combined with other evidence in the determination of the relevant fact, even though it has been used to reject the witness’s testimony. Strictly, it also cannot be used to enhance the weight given to other evidence on the point.

It is more sensible to treat prior statements, once they are admissible, and whether they are consistent or inconsistent with the witness’s testimony, as being like any other statements: admissible to prove the truth of what they assert. Constraints on their admissibility are necessary to confine the evidence within manageable bounds, but once admissible the availability of their maker for cross-examination should remove the need for a rule rejecting them as proof of their truth.

So, what of the point made in the quotation from Divitaris in para 7 of Stirling? If the prior statement is alleged to have been a lie, its maker (the witness) can be cross-examined to show that. If cross-examination cannot undermine the truth of the prior statement, why should it not be admissible to prove its truth?

In New Zealand there is currently some uncertainty over whether s 35 of the Evidence Act 2006 has made prior statements evidence as proof of their truth. The Law Commission apparently intended them to be evidence of their truth, but there is a first instance decision holding that they are only proof of the witness’s consistency. My own view is that the section is clearly intended to make prior consistent statements admissible, in certain situations, as proof of their truth. Subsection (2) allows such statements to be used, inter alia, to respond to a challenge to the witness’s accuracy, and subsection (3) refers to circumstantial reliability of the prior consistent statement.

Dishonesty may not be deceptive

Occasionally, reading a case causes you to think of something almost completely unrelated. This happened to me when I read Norris v United States [2008] UKHL 16 (12 March 2008).

The almost completely unrelated thing
A great risk in writing articles in specialised areas of law is that the arguments advanced in them may be rejected by the highest courts. Why write at all?

The purest motive was put vividly by Emeritus Professor John Burrows QC, formerly of the University of Canterbury and currently of the New Zealand Law Commission, in an interview with Ursula Cheer on 24 January 2007: “… I think I enjoy research probably the most. Because it’s a great thrill in getting the material together in your own mind and producing something on paper and saying it’s yours. It’s really yours.”

Practitioners, as well as academics, may well expect career advancement as a result of their publications. For example, the following has appeared on the web site of one writer:

“As a consequence of his publication in February and March 2005 with Sir Jeremy Lever QC of a two part article in the European Competition Law Review entitled “Cartel agreements, conspiracy to defraud and ‘the statutory offence’ ”, he was instructed by a leading New York law firm in the high profile extradition case ‘The Government of the United States v Norris’, which was the first time that the issue of cartel crime has been tested in an English court. The article formed the basis of the US government’s case as presented by David Perry QC and was extensively quoted by the judge in his judgment; John was led by Alan Jones QC.”

This is the case that the House of Lords has now decided. Unfortunately their Lordships were critical of the article:

“59 The first time it was apparently suggested in any publication that price-fixing might be a common law offence was in an article written in 2005 by Sir Jeremy Lever QC and Mr John Pike, “Cartel Agreements, Criminal Conspiracy and the Statutory ‘Cartel Offence'” [2005] ECLR 90. At p 95, the authors made the point, which has already been touched on, that if, in addition to the price-fixing, something positively misleading is said, such as a dishonest “representation … that offers are being made competitively”, the criminal law will be engaged. More controversially, the authors then went on to suggest that making and operating secret price-fixing agreements could, of itself, operate dishonestly so as to constitute a crime, at least in circumstances where purchasers are acting in the belief, known to the price-fixers, that there is no price-fixing.

“60 This article was not only published after the 2002 Act, but a number of years after the activities complained of in count 1 had ended. So it is not as if even an astute reader of legal articles in this area of law could have informed himself at the relevant time of the possibility of his price-fixing activities attracting criminal sanctions. In any event, although the Divisional Court was impressed with the article, there are problems with the notion that mere secrecy can of itself render the price-fixing agreement criminal. It is not as if secrecy is always necessary for a price-fixing agreement to be effective, or that it is the secrecy which causes a purchaser loss. As already mentioned, in order to establish a criminal offence along the lines suggested by Lever and Pike, it would be necessary to show that it was the secrecy which caused the purchaser’s loss, since it must be the alleged dishonesty which causes the loss.

“61 Quite apart from this, it would be dangerous and impractical, particularly for the judges, to introduce a general principle that there is some sort of implied representation that the price at which goods are offered has been arrived at on a certain basis. Finally, the very fact that it was not until 2005 that it was first suggested that secret price-fixing could of itself constitute a common law offence underlines the difficulty faced by the argument that it would have been a common law offence in the 1990s, especially when one considers the material which was available on the topic from Parliament and the courts.”

It is to be hoped that the writer of the impugned article will bounce back from this blow and appreciate the light that the House of Lords has cast on the subject. Or, he will wait for the next case and hope that the Law Lords change their minds.

The next case, as it happened, came the same day: R v GG plc [2008] UKHL 17 (12 March 2008). Here, Norris was applied and it was held that in relation to conspiracy to defraud by price fixing, more than mere secrecy is required to constitute the element of fraudulence. Misrepresentation and deception have been found to be necessary (para 16, citing Norris at para 19).

Points mentioned in these cases

  • In the UK there was no common law or statutory offence corresponding to the strict liability offence created in the USA by s 1 of what is known as the Sherman Act (making agreements in restraint of trade illegal; here the relevant agreement was price fixing): Norris para 23, 52.
  • Even if there had been a change in public perceptions of price fixing, it would be for Parliament, not the courts, to create a relevant criminal offence: Norris, para 57.
  • Strictly obiter, but nevertheless determinative: in considering extradition, it is the overall conduct alleged against the defendant that must be considered, and in the UK the question is whether the conduct is against the law of the requesting country (whatever the particular offence may be) and whether it is against UK law (again, whatever the offence – not necessarily an identical offence to the foreign one): Norris para 90, 91.
  • Conspiracy to defraud requires proof of an agreement to make false or misleading statements or to otherwise engage in actively fraudulent behaviour. Mere secrecy and deception is insufficient (GG plc, para 12, 18).

As can be seen from this last point, the concepts here can be rather subtle. It helps to focus on the causal requirement: the relevant deception is the one intended (agreed) to cause loss. Keeping the agreement to fix prices secret, even to the extent of telling lies to conceal the agreement from the authorities, does not cause customers to lose the opportunity of a better price. The prices would still be the same if the customers were told that the prices had been fixed by agreement between suppliers. The necessary deception would occur if it had been represented to the customers that the prices had not been fixed, when the person making that representation knew that that was untrue, and if that deception was (as, of course it would be) designed to induce the customers to make purchases, and if that representation caused the customer to part with more money than would otherwise have been the case.

In New Zealand we have repealed the offence of conspiracy to defraud, and currently the offence of obtaining by deception (or causing loss by deception) – s 240 Crimes Act 1961 – could in appropriate circumstances be charged as a conspiracy against s 310 of that Act, as conspiracy to obtain (or cause loss) by deception. “Deception” is defined in s 240(2) in terms that are expanded from those recommended in the Report of the Crimes Consultative Committee on the Crimes Bill 1989 (April, 1991), so even at that stage of revision of the law the matter was unsettled. The House of Lords decisions mentioned here would be relevant under these provisions.

Insufficiently bad is good enough

Four topics deserve mention here in connection with Pitman v The State (Trinidad and Tobago) [2008] UKPC 15 (3 March 2008). The first three of these are not new law, just reminders of the best way of conducting trials. The fourth, also not new,  is the ground on which the appeal was allowed.

Avoiding the ironic platitude
In commenting on the accused’s failure to give or to call evidence, the judge observed that much cross-examination of apparently reliable witnesses had included imputations against character that were not supported by evidence. The Board regarded the particular comments here as “ill-chosen and of a nature which should be avoided in a criminal trial” (para 20), and quoted a passage from the summing up which concluded “However, it is [the accused’s] right to have elected to remain silent, as he did, and you cannot hold it against him”, and held (para 19):

“if a judge makes unduly unfavourable remarks about a defendant or witness, which may be prejudicial, they cannot be sufficiently neutralised by resorting to the mantra in the last sentence of the passage quoted: cf Mears v The Queen [1993] 1 WLR 818.”

Here, the remarks were held to be insufficiently prejudicial to make the conviction unsafe. Plainly, however, the Board is warning against the recitation of the accused’s right to silence as an attempt to rectify improper comment.

The duty of a court in a criminal appeal
“It is the duty of the court in a criminal appeal to take account of all the grounds which could reasonably be advanced on behalf of an appellant, whether or not they have been sufficiently argued, and their Lordships think that it was desirable that this should have been done.” (para 25)

This point was also made in Charles v R (St Vincent and the Grenadines), blogged 20 July 2007, and not cited in Pitman. The point not argued in the court below in Pitman was whether the alleged joint enterprise may have come to an end before the killings, although the Court of Appeal had allowed the co-accused’s appeal on that basis.

Nevertheless, in Pitman the Board did not consider that there had been inadequate directions on joint enterprise at the trial, and this point was not determinative of the appeal. Some criticism was made, however, of the way the judge had approached the subject in his directions to the jury.

The best perspective on complex directions
In covering the subject of liability on the basis of joint enterprise, the judge had chosen the perspective of how the accused could be guilty. The Board considered (para 25):

“…it would have been preferable if the judge had spelled out in the case of each defendant how it might be said that he did not contemplate the murder of the victims ….”

The judge’s approach was not wrong, but it could have been better. It is quite natural for a judge to focus on ways in which an accused may be guilty, because the question is what the prosecution has to prove. However the judge has to show the jury how to relate points of law to particular facts. This can be done by drawing attention to how the prosecution may fail in its task. The Board made these general remarks applicable to allegations of joint enterprise, concerning how an accused may not be a participant (para 25):

“This could be argued in any case of this nature on several bases, according to the facts of the case: that the joint enterprise never went beyond robbery and that the defendant in question did not foresee that his confederate might go beyond that and commit murder; that when he realised that things were getting out of hand and that his confederate was intending murder, the defendant withdrew from the joint enterprise by a sufficiently clearly evinced dissociation; or that the defendant was no longer part of the joint enterprise, which had earlier come to an end. These are in truth facets of the same issue, whether the defendant was part of a joint enterprise which included as one of its elements the possible murder of another person. That depends on what was agreed, expressly or impliedly, by the defendant, and if a murder takes place it may be outside the parameters of the enterprise in a number of ways, including those specified above. It is desirable that a trial judge should tailor his directions to the evidence, so that the jury have put clearly before them the basis on which to decide if the defendant agreed to the commission of the act with which he is charged.”

Fresh evidence
The appeal was allowed in Pitman because fresh evidence had been obtained concerning the appellant’s mental capacity. The accused had not had the facilities before trial to obtain this evidence. It potentially had a bearing on several issues, and the case was remitted to the Court of Appeal for those to be considered.

To qualify as “fresh” the evidence must be capable of belief and there must be a reasonable explanation for its not having been called at trial. At para 31 the Board added:

“These factors are not, however, conclusive of the issue of admission of fresh evidence, and an appellate court has the overriding statutory power to admit it if it is in the interest of justice: see Benedetto v The Queen [2003] UKPC 27, [2003] 1 WLR 1545, and cf Smalling v The Queen [2001] UKPC 12.”

See also Bain v R (New Zealand) [2007] UKPC 33 (blogged here 11 May 2007) at para 34.

In Pitman, the fresh evidence rendered the convictions unsafe because (para 31 – 32) it was both capable of belief and prima facie raised issues which were substantial and such as to require proper investigation by the court. Accordingly, the Court of Appeal would have to approach the matter as follows (para 32):

“They will have to form an opinion on the appellant’s mental capacity, then, depending on the opinion which they form, they may have to decide (i) whether he was fit to plead and stand trial (ii) whether there is a sufficient doubt about his ability to understand and participate in the joint venture (iii) whether his statement should have been admitted and, if necessary, (iv) whether there is sufficient evidence to raise the defences of unsoundness of mind or diminished responsibility. If they find in the appellant’s favour on any of these issues, the conviction will be unsafe and must be set aside.”

In New Zealand many criminal lawyers will be thinking about R v Barlow, in which fresh evidence was, as announced yesterday, insufficient for the Governor-General to refer the case back to the Court of Appeal. That case seems to be heading to the Privy Council. No doubt the question will be whether the fresh evidence there makes the convictions unsafe. They would be unsafe if the jury did not have evidence which was both capable of belief and relevant to assessment of the weight of the evidence it did have.

[Update: for the Privy Council’s decision in Barlow, see note for 9 July 2009.]