Reverse onus in the House of Lords

While we in New Zealand await our Supreme Court’s decision in Hansen v R, which will decide the standard of proof placed on the accused by reverse onus provisions in the light of human rights legislation, we note that the House of Lords now finds the reading down of reverse onus provisions a relatively routine matter: O v Crown Court at Harrow [2006] UKHL 42 (26 July 2006).

Here, the House of Lords was concerned with the meaning of “is satisfied” in a provision that required refusal of bail unless the court was satisfied that exceptional circumstances existed (s 25 of the Criminal Justice and Public Order Act 1994[UK]):

“(1) A person who in any proceedings has been charged with or convicted of an offence to which this section applies in circumstances to which it applies shall be granted bail in those proceedings only if the court or, as the case may be, the constable considering the grant of bail is satisfied that there are exceptional circumstances which justify it.”

Two approaches to this were considered in the Court of Appeal. Kennedy LJ thought “satisfied” means “considers” and that the question is a matter of judgment for the court, not involving any particular standard of proof. Hooper J, on the other hand, thought that “satisfied” here puts the burden on the bail applicant, and is therefore inconsistent with the ECHR, so s 25 has to be “read down” to place only an evidentiary burden on the bail applicant.

Lords Nicholls and Hutton, and Lady Hale, agreed with Lord Brown; Lord Carswell did too, but examined the difference in interpretation in more detail. He referred, rather obliquely, to Re McClean (blogged here 7 July 2005) in which it was decided, inconsistemtly with the present case, that doubts must be resloved against the prisoner. He held that “satisfied” in s 25 means more than just an exercise of judgment, and that it connotes a burden or presumption, so s 25 has to be read down to comply with convention rights. Thus, Lord Carswell agreed with Hooper J’s approach.

Lord Brown, delivering the leading speech, had a “mild preference” for Hooper J’s approach (para 35). He pointed out that in most cases the decision will be clear cut and that the burden of proof will not assume any relevance. Occasionally, however, the court will be left unsure, and in these cases the default position should be that bail should be granted, and s 25 should be read down to make that plain.

Differences in principle and pragmatism

Appellate judges often disagree over whether an error at trial caused a substantial miscarriage of justice. Finding himself in a minority of one in Bounds v R [2006] HCA 39 (20 July 2006), Kirby J observed, para 85:

Conclusion: a basic error: I accept that different conclusions are available on this issue. The nature of the controversy and the breadth of the language of the ‘proviso’ virtually assures the existence of differences of judicial views. Such differences may reflect the diverse values that judges accord to considerations of principle and pragmatism, as they regard them. Those differences appear in many cases in this Court concerned with the ‘proviso’ … [citing Green v The Queen (1971) 126 CLR 28 at 31; Jones v The Queen (1997) 191 CLR 439; Darkan v The Queen [2006] HCA 34]. They are also reflected in the foregoing differences of opinion in the United States Supreme Court [ie Lane v US 747 US 438 (1986)].”

It seems likely that, on the key point in Bounds, the majority were correct. A second count had been wrongly included in an indictment. The error was that the second offence was purely summary, not triable on indictment. Nevertheless, evidence of its commission would probably have been admissible as similar fact evidence in respect of the only count that should have been in the indictment. The majority (Gleeson CJ, Hayne, Callinan and Crennan JJ) at para 26 held

“… The evidence admitted at the appellant’s trial about the downloading of the images the subject of count 2 was very limited. It would have been admissible on the trial of an indictment alleging only count 1.”

Kirby J did not think this was so (para 100):

“Nor do I consider that the tender of such material would have been permitted in a trial if that trial had been limited to the indictable offence of possession of child pornography. A judge guarding the fairness of the conduct of such a trial would be properly careful to restrict extraneous, and possibly prejudicial, evidence. By impermissibly charging the two offences in the one indictment, an inter-mixture necessarily occurred. Descriptions or conceptions of the contents of the images became inevitable. It is that inter-mixture, before the jury, that presents the risk of a substantial miscarriage of justice. It is that risk that withholds the application of the ‘proviso’.”

Apart from the difference of opinion on that issue, Kirby J also differed from the majority on whether the jurisdictional error was, of itself, sufficient to prevent application of the proviso without examination of issues relating to the conduct of the trial. He reasoned (paras 87-88) that the Western Australia legislature had carefully separated the jurisdictions and that the community and the accused are therefore entitled to a trial that conforms to the law and is without jurisdictional flaw.

The majority, however, did focus on what happened at the trial of the count that was properly before the jury (para 13-30). Their conclusion on the trial conduct contrasts with that of Kirby J, who held, para 96:

“… The impermissible inclusion of the second count in the indictment deprived the appellant of a trial that was free from any reference to this extraneous factor [images of bestiality]. It deprived him of the chance of avoiding this added complication in his trial. Specifically, it deprived him of the forensic choices that the separate trial of the offences referred to in the two counts would have entailed.”

It is not clear what those “forensic choices” were, even on the basis that the evidence was not (as Kirby J held) admissible as similar fact. Even the mere fact that these images (whatever they might have depicted) were downloaded at a particular time seems to have been relevant to whether the images which were the subject of count 1 were downloaded by the accused (see majority judgment, para 26).

This case is interesting for its references to other cases, including English and United States decisions, where differences of opinion have occurred over how to decide whether an error has caused a trial to be unfair. It is now 99 years since England enacted what has been almost universally adopted as the “proviso”, and one would have thought that, by now, appellate judges would have worked out a rational approach to its application, thereby promoting uniformity of result and minimising disagreements. My own view is that the basic problem is that judges have not got to grips with the meaning of the essential requirement of a fair trial. Until judges learn to define the characteristics of trial fairness and to apply them rationally, vagueness over the meaning of “substantial miscarriage of justice” will continue to make the results of appeals difficult to predict.

If one were to apply my own “Fair Trial Analysis”, one would identify the following issues in Bounds. Since a fair trial is one where the law is accurately applied to facts determined without bias, the first question is, what errors of law occurred? It was common ground that count 2 should not have been in the indictment. Was the admission of evidence relevant to that count an error? This requires examination of the admissibility of the evidence as similar fact in relation to count 1. The judges in Bounds did not undertake this inquiry in detail, they simply announced their conclusions, and, as noted, Kirby J dissented on this. If it is correct that the evidence was admissible as similar fact, then the question is whether the judge’s direction to the jury on that topic was adequate to ensure that the evidence was not misused. If the evidence was not admissible as similar fact, then the question is whether the jury were adequately warned to ignore it in relation to count 1. The jury direction was not revealed in the judgments in Bounds. Then, in relation to the unbiased determination of the facts, the appellate court should ask whether the similar fact evidence – even if it was admissible – could have, by reason of its content, prejudiced the rational assessment of the other evidence. This point is addressed in Bounds, but again it was a point of difference. Kirby J thought that it could have caused prejudice, simply because of the risk of an averse reaction by any juror to the mere mention of bestiality. The majority, however, considered that the mention of bestiality (the jury were not shown any images) was of peripheral concern and would not have caused illegitimate prejudice. At the heart of this difference of judicial opinion is the extent to which it should be acknowledged that jurors may not approach their task rationally. This was mentioned as recently as the 19th of this month by the House of Lords in R v Coutts (blogged 21 July 2006), where Lord Bingham (with whom the others agreed), citing a High Court of Australia case, Gilbert v R (2000), said at para 20 that the jury room may not be a place of undeviating intellectual and logical rigour. Coutts and Bounds are cases that complement each other: in Coutts the concern was failure to include an offence in the indictment, whereas in Bounds the concern was improper inclusion of a count. Should the court be less sensitive to the risk of jury prejudice in one than in the other?

Fairer than you may want

The right to a fair trial is a right that the accused cannot waive.

This may seem a strange statement: why would the accused ever want to waive the right to a fair trial? Well, trial tactics are complicated things. If charged with one serious offence, the accused may prefer to hope for an outright acquittal on that rather than have the judge tell the jury that if they acquit on the serious charge they may then consider whether to convict on a lesser charge. A conviction on the lesser charge, and the consequence of imprisonment on that, may be something the accused would see as rendering his victory on the more serious charge Pyrrhic.

A fair trial is a trial conducted according to law. This means, as the House of Lords held this week in R v Coutts [2006] UKHL 39 (19 July 2006), that the jury must be fully informed about the alternatives available to it, whether either side want that or not. In Coutts, both sides agreed that the judge should not instruct the jury about the alternative of manslaughter, when the accused faced a solitary charge of murder. The Crown, confident that it would secure a conviction for murder, did not want the jury to instead convict for manslaughter, so, clothing its stance in the guise of a fairness objection, it argued that the accused was entitled to an acquittal if the jury had a doubt about the way the Crown had sought to prove its case for murder. The defence, having argued that the killing was accidental, did not want the risk of a conviction for manslaughter, because the sentence for that would, in the circumstances, be a lengthy term of imprisonment. Following the conviction for murder, the defence appealed, arguing that the judge should have directed the jury about manslaughter. The House of Lords upheld this argument and quashed the conviction, remitting the case so that a retrial could be considered.

Coutts establishes that fairness requires that the jury be fully informed about the law and the alternatives open for consideration in the particular case. Further, failure to inform the jury of the legal position amounts to a substantial miscarriage of justice. The appellate court does not enter into an inquiry about whether the jury only convicted the accused of the serious charge because it did not want him to walk free. To do that would be to engage in speculation. While a foundation of the system of trial by jury is the assumption that juries apply the directions on the law that judges give them (Lord Rodger, para 87), it is proper to recognise that the jury may be affected in its approach by the choices that it perceives as being available to it (an observation by Callinan J in the High Court of Australia case Gilbert v The Queen (2000) 201 CLR 414 para 101, quoted by Lord Bingham at para 20, Lord Hutton at 54, Lord Rodger at 88 and Lord Mance at 99).

Accordingly, the public interest in the administration of justice is best served if the judge leaves any obvious alternative offence to the jury irrespective of the wishes of counsel. Lord Bingham, with whom the other law lords agreed, put the requirement as follows (para 23):

“The public interest in the administration of justice is, in my opinion, best served if in any trial on indictment the trial judge leaves to the jury, subject to any appropriate caution or warning, but irrespective of the wishes of trial counsel, any obvious alternative offence which there is evidence to support. I would not extend the rule to summary proceedings since, for all their potential importance to individuals, they do not engage the public interest to the same degree. I would also confine the rule to alternative verdicts obviously raised by the evidence: by that I refer to alternatives which should suggest themselves to the mind of any ordinarily knowledgeable and alert criminal judge, excluding alternatives which ingenious counsel may identify through diligent research after the trial. Application of this rule may in some cases benefit the defendant, protecting him against an excessive conviction. In other cases it may benefit the public, by providing for the conviction of a lawbreaker who deserves punishment. A defendant may, quite reasonably from his point of view, choose to roll the dice. But the interests of society should not depend on such a contingency.

“24. It is of course fundamental that the duty to leave lesser verdicts to the jury should not be exercised so as to infringe a defendant’s right to a fair trial. This might be so if it were shown that decisions were made at trial which would not have been made had the possibility of such a verdict been envisaged. But no such infringement has ordinarily been found where there is evidence of provocation not relied on by the defence, nor will it ordinarily be unfair to leave an alternative where a defendant who, resisting conviction of a more serious offence, succeeds in throwing doubt on an ingredient of that offence and is as a result convicted of a lesser offence lacking that ingredient. There may be unfairness if the jury first learn of the alternative from the judge’s summing-up, when counsel have not had the opportunity to address it in their closing speeches. But that risk is met if the proposed direction is indicated to counsel at some stage before they make their closing speeches. They can continue to discount the alternative in their closing speeches, but they can address the jury with knowledge of what the judge will direct. Had this course been followed in the present case there would have been no unfairness to the appellant, and while taking a contrary view the Court of Appeal did not identify the unfairness which it held would arise. It is not unfair to deprive a defendant, timeously alerted to the possibility, of what may be an adventitious acquittal.”

Public privacy

Courts may disagree over whether language is “grossly” offensive. In DPP v Collins [2006] UKHL 40 (19 July 2006) the House of Lords unanimously disagreed with two lower courts on this point. The defendant had left messages by telephone at his MP’s office, employing terms of racial abuse that were unquestionably offensive. The lower courts held that the messages did not cross the line between offensive and grossly offensive. The embarrassment of disagreeing with the lower courts was exacerbated because essentially the question whether the messages were “grossly” offensive was one of fact, to be determined by applying the standards of an open and just multi-racial society taking into account the context and all relevant circumstances (Lord Bingham, para 9) – standards which the lower courts should be particularly well placed to apply.

How had the lower courts gone wrong? Lord Bingham at para 13 concluded:

“Differing from the courts below with reluctance, but ultimately without hesitation, I conclude that the respondent’s messages were grossly offensive and would be found by a reasonable person to be so.”

He agreed with Lord Carswell, who was a little more explicit about this (para 22):

“I felt quite considerable doubt during the argument of this appeal whether the House would be justified in reversing the decision of the magistrates’ court that the reasonable person would not find the terms of the messages to be grossly offensive, bearing in mind that the principle to which I have referred, that a tribunal of fact must be left to exercise its judgment on such matters without undue interference. Two factors have, however, persuaded me that your Lordships would be right to reverse its decision. First, it appears that the justices may have placed some weight on the reaction of the actual listeners to the messages, rather than considering the reactions of reasonable members of society in general. Secondly, it was conceded by the respondent’s counsel in the Divisional Court that a member of a relevant ethnic minority who heard the messages would have found them grossly offensive. If one accepts the correctness of that concession, as I believe one should, then one cannot easily escape the conclusion that the messages would be regarded as grossly offensive by reasonable persons in general, judged by the standards of an open and just multiracial society. The terms used were opprobrious and insulting, and not accidentally so. I am satisfied that reasonable citizens, not only members of the ethnic minorities referred to by the terms, would find them grossly offensive. “

The offence in question was held to require mens rea – an intention that the words be grossly offensive to those to whom they related, or an awareness that they may be taken to be so (Lord Bingham, para 11). But on the approach taken here, that state of mind need not be followed by actual offence taken by the recipient of the message. Indeed, it was held not to be necessary than anyone actually receive the message (para 8).

Consequently, the offence of sending a grossly offensive message could be committed without actually offending anyone. What makes it an offence is the fact that the social standards have been infringed. The offence is, on this interpretation, sending a message that would grossly offend a reasonable person if such a person became aware of it. That seems to be a bit of a stretch from the wording of the relevant legislation: s 127 of the Communications Act 2003[UK]:

127. Improper use of public electronic communications network
(1) A person is guilty of an offence if he—
(a) sends by means of a public electronic communications network a message or other matter that is grossly offensive or of an indecent, obscene or menacing character; or
(b) causes any such message or matter to be so sent.”

The justification for the imposition of community standards is the public nature of the communication network, which would be fine if everyone had ready access to everyone else’s communications. In reality, of course, the network is no more public than words exchanged in a conversation between two people on the street.