Wednesday, April 26, 2006

Public policy and degrees of misconduct

Public policy exclusion of evidence is sometimes still spoken of by courts as if it was a means of protecting the accused’s right to a fair trial. This was done yesterday by the Privy Council in Williams v R (Jamaica) [2006] UKPC 21 (25 April 2006). A statement had been obtained from the accused when he was aged 12, in breach of the procedures laid down for the interviewing of young suspects (para 27). The Board concluded, at para 28:

"…the circumstances of the appellant's detention and of the taking of the statement were such as to create a significant amount of unfairness to him. Their Lordships cannot conclude that in all the circumstances of the case it was fair to admit the statement."

While the Board was not expressly referring to trial fairness, the expressions "unfairness to him" and it not being "fair to admit" the statement do suggest that trial fairness is the object of concern. Another point in this extract that is misleading is the expression "a significant amount of unfairness".

My view is that in these situations the courts are not excluding the evidence for reasons that have anything to do with trial fairness. The probative value of the evidence is not relevant to the decision on admissibility in these cases (see, for example, my blog entry for 4 April 2006 concerning "cogency"). It is not the accused’s guilt that matters, it is the objectionable way in which the evidence was obtained that is critical. This is why the discretion, in this area, is best called the public policy discretion. It is true that, historically, this discretion evolved from – and remains part of - the court’s inherent power to prevent an abuse of process. In turn, that power has been referred to as giving rise to a discretion to exclude evidence in the interests of "fairness", or, sometimes, "fairness to the accused". But these are not references to trial fairness.

One of the reasons for making this distinction between the public policy exclusion of evidence, and the exclusion of evidence to ensure trial fairness, is to preserve the concept of the absolute nature of the accused’s right to a fair trial. In the above quotation from para 28 of Williams, the expression "a significant amount of unfairness" could, wrongly, suggest that fairness of trial exists in gradations, and that some forms of trial unfairness are acceptable. I have given examples of misuse of this terminology in "The Duty to Prevent an Abuse of Process by Staying Criminal Proceedings" in Essays on Criminal Law – A Tribute to Professor Gerald Orchard (Brookers Ltd, 2004), 133, 146.

What, it is respectfully suggested, the Privy Council should have said in Williams, is that the breaches of the Directions on the conduct of interviews of young persons that occurred in this case were sufficiently serious that admission of the statement obtained thereby would be an abuse of process. The evidence was excluded to prevent the administration of justice being brought into disrepute, as would occur if the courts appeared to endorse the police misconduct. In that context it is appropriate, if one must use the "fairness" terminology, to speak of degrees of unfairness, because official misconduct comes in degrees.

There may be cases where trial fairness could be relevant to the admission into evidence of a statement that had been obtained wrongfully. But it must be remembered that, at the stage when the ruling on admissibility has to be made, the effects of admission on the defence will not be known: the Judge will not know whether the defendant intends to give or call evidence, let alone what any such evidence would be. There is a difference between using fairness as grounds for excluding evidence, and using fairness in considering, as an appellate court, whether to apply the proviso. It is the appellate court that is best placed to evaluate the effect of an erroneous admission of evidence, and, when the public policy discretion has been exercised in favour of admitting the evidence, it may be only retrospectively that the trial can be said to have been unfair.

Seeing and believing

Eyewitness identification evidence may need to be treated with some circumspection, and juries are usually given a direction on the special need for caution before relying on such evidence. In some jurisdictions, these warnings are required by statute, although, as is the case in New Zealand, the points required to be covered are not spelt out in great detail. The common law antecedent of these directions is known as the Turnbull direction, originating in the English Court of Appeal’s decision R v Turnbull [1977] QB 224.

In Edwards v R (Jamaica) [2006] UKPC 23 (25 April 2006) the Privy Council indicated that there are some practices that should not be permitted at trials where identification is an issue and the prosecution relies on evidence of an eyewitness to the offence. These are:

(1) The eyewitness should not be permitted to identify the accused in the dock as the offender. The prosecution should, in general, adopt other means for establishing that the accused is the person who was arrested (para 22):

"… it is only in the most exceptional circumstances that any form of dock identification is permissible: cf the discussion in the Scottish devolution appeal Holland v HM Advocate [2005] UKPC D1, 2005 SLT 563. It may be borne in mind that this was far from being a first identification and it can fairly be said that the dock identification may have had little impact on the minds of the jury. It is, however, an undesirable practice in general and other means should be adopted of establishing that the defendant in the dock is the man who was arrested for the offence charged."

(2) A police officer should not be permitted to give in evidence his opinion on why an ID parade was not considered to be necessary (para 23).

(3) The police should not give in evidence the fact that a warrant was obtained for the arrest of the accused, or of the information on which the police acted, as this is hearsay and potentially highly prejudicial (para 23).

(4) The police should not give in evidence the fact that a potential witness was unwilling to come forward (para 23).

(5) The police should avoid confronting the eyewitness with the suspect (para 25).

There is, at this point – para 25 – a possibly unintentional suggestion by the Board that hearsay evidence might be given to establish the link between the person described to the police as the offender, and the suspect:

"The arresting officer would have been quite capable of establishing that the appellant was the person pointed out to him by Bailey [the eyewitness] near the Mango Tree Bar, so it was unnecessary to ask Bailey to come to the station to confirm that."

This, however, should be read as referring to "establishing" in an investigatory, pre-trial, sense, and not as "establishing" in evidence at trial.

In this case, the eyewitness to the killing had been standing next to the victim, and the bullet that killed the victim had passed through the eyewitness. The offender had been trying to rob the eyewitness, who suddenly and unsuccessfully tried to grab the gun. The eyewitness was hospitalised for 4 weeks, and it was 2 months after the killing before he saw the accused near the same bar. He claimed that the accused was the offender. In his first description of the offender, given 5 days after the incident while he was in hospital, the eyewitness failed to mention a prominent birthmark on the accused’s face, he was unable to say what sort of trousers the offender was wearing, and he claimed that the time he had to observe the offender was a couple of minutes although it must have been shorter than that. The circumstances in which the offender was observed were good: inside a bar in the morning with good lighting.

The Privy Council was concerned that there could have been an erroneous association of ideas arising from the location of the offence and the subsequent identification being similar, and that the judge had not adequately warned the jury of the dangers in accepting the evidence. The conviction was therefore unsafe.

In New Zealand, the Evidence Bill 2006, clause 122, almost exactly repeats the current provision on the need for judicial warning: Crimes Act 1961, s 344D. The slight difference is that instead of requiring the judge to "include the reason for the warning", the Bill requires the judge to "warn the jury that a mistaken identification can result in a serious miscarriage of justice". The need for a warning arises "In a criminal proceeding tried with a jury in which the case against the defendant depends wholly or substantially on the correctness of 1 or more visual or voice identifications of the defendant or any other person …". The inclusion of voice identification is new to the Bill.

The Bill contains other provisions relating to the admissibility of visual identification evidence. It is important to note that here the concern is with admissibility, not with the way admissible evidence is treated at trial. These provisions, for visual identification, are in clause 41, and they concern the implications of whether or not a formal identification procedure was used at the investigatory stage. The criterion for admissibility is proof, on the balance of probabilities, that the evidence is reliable. The Bill does not say to what extent, if any, this reliability should be assessed by reference to the other evidence in the case. It seems plain that the other evidence should not be included in the assessment of the reliability of the visual identification evidence, and that the focus should be on the circumstances in which the identification was made.

Tuesday, April 04, 2006

Balancing "cogency" of wrongfully obtained evidence

Simmons v R (Bahamas) [2006] UKPC 19 (3 April 2006) gives us an opportunity to highlight the distinction between "fairness" when that term is used in the context of the exercise of the public policy discretion to exclude evidence, and "fairness" in the separate sense of trial fairness for the accused.

The two appellants were convicted of murder. Included in the evidence against them were statements they had made to the police. These statements, referred to as confessions, although they were partly exculpatory (and, one accused who gave evidence adopted what he had said to the police), were obtained in breach of the appellants’ constitutional right to be informed of the availability of legal advice before they spoke to the police.

Breach of that right gives rise to a judicial balancing exercise to determine whether to admit or exclude the statement. This balancing exercise is a public policy discretion, having nothing to do with the fairness of the trial. It arises because of the conflict between, on the one hand, the public’s right to have suspects prosecuted and offenders brought to justice, and, on the other hand, the public’s right to have officials comply with the law in the investigation of offences. The only sanction that courts can impose on officials who act in breach of the law in this context is to exclude evidence that they obtain thereby.

Sometimes, in carrying out this balancing exercise, judges say that the cogency of the evidence is a factor strongly favouring its admission. An important observation on this point was made by the Privy Council in this case, para 26:

"The Board has one other concern about the judge's balancing of the respective interests of the prosecution and the defence on the issue of fairness: the evident importance which she attached to the confession being "very cogent evidence against Simmons." Their Lordships cannot accept that the potency of such evidence is necessarily a factor in favour of its admission. If, by denying a suspect his constitutional right to see a lawyer and perhaps be advised against making a statement, the state's case is thereby strengthened by a confession, it is by no means self-evident that fairness demands its admission rather than its exclusion."

However, in New Zealand the cogency of the evidence is routinely taken into account in this balancing exercise: R v Shaheed [2002] 2 NZLR 377 (CA), especially at paras 151-152. At para 151 the joint judgment (Richardson P, Blanchard and Tipping JJ) states: "A trial is not to be regarded as potentially unfair by reason of the admission of evidence unless that evidence may lead to an unsafe verdict." That, with respect, needs to be read with some caution. A safe verdict is not a cure-all for trial unfairness. That point was made strongly by Lord Steyn (Sir Swinton Thomas concurring) in Ebanks v R (Cayman Is) [2006] UKPC 16 (27 March 2006), blogged 28.3.06, at para 40. The Supreme Court acknowledged the same point in Sungsuwan v R [2005] NZSC 57 (25 August 2005), blogged 26.8.05, per Elias CJ at para 6 (putting as alternatives trial unfairness and unsafe verdicts), Tipping J at 112 (lack of a fair trial is itself a miscarriage of justice without the need to consider its effect on the verdict).

If I may, I should add that "cogency", which means being convincing or compelling, is always a matter for the jury. It is usually called the "weight" that is to be given to the evidence. Weight is separate from the question of admissibility, except on occasions where it is possible for a judge to conclude that no reasonable jury could give the evidence any weight. The Privy Council is correct to see no reason to link cogency to admissibility. However, this is not the end of the matter. What is being considered is not the admissibility of the evidence, but rather whether, as admissible evidence, it should be excluded because of the objectionable way in which it was obtained. The question whether convincing or compelling admissible evidence should be excluded is, appropriately, part of the weighing of the public interest in bringing suspects to justice. There is, though, a difficulty: the cogency of the evidence is also appropriately considered on the other side of the balance, where weight has to be given to the public interest in prevention of such abuse of process as would bring disrepute to the administration of justice. One might properly object to the inclusion of "cogency" in the weighing process on the basis that it falls on both sides of the scales.

That aside, having said that the exclusion of evidence to prevent the trial being unfair to the accused is separate from the public policy discretion, I should now make the distinction clear. A fair trial for the accused is one where the law is accurately applied and the facts are determined without bias. There may be flow-on effects of a wrongful exercise of the public policy discretion, in the sense that the trial may not be one where the law has been accurately applied. The question, in terms of trial fairness, of the significance of the error, will be determined, not by the strength of the other evidence against the accused, but by whether the error put the accused at a disadvantage in the trial. I made this point in the blog entry on 29 March 2006, discussing Gilbert v R (Grenada) [2006] UKPC 15 (27 March 2006).

Simmons is an example of the error in applying the public policy discretion not affecting the fairness of the trial. The statements in issue were partly exculpatory and the accused who gave evidence adopted what he had said. The judgment does, however, focus on the strength of the other evidence of guilt, concluding, para 31, that acquittals would have defied all reason. The relevance of this point is that the proviso could be applied: the error in admitting the statements did not amount to a "substantial miscarriage of justice." The error caused neither trial unfairness, nor the loss of a real chance of acquittal.

Monday, April 03, 2006

You may not get what you want ...

The crime of attempting to commit a crime has come under scrutiny in L v R [2006] NZSC 18 (30 March 2006), which concerned the state of mind required for liability.

The facts of L were extraordinary. The charge, faced by L, a 49 year old woman, was attempting to sexually violate a 15 year old male. She tried to cause his (presumably erect) penis to penetrate her vagina, being reckless as to whether or not he consented to that. In the circumstances, it seems to have been the case that his lack of consent should have been apparent to her.

The legal question was whether, to be liable for the attempt, L had to intend that he not consent, or was it sufficient that she merely be reckless (ie that she knowingly took a risk) as to his lack of consent. The question arose because the crime of attempting to commit a crime requires the accused to have an intent to commit that crime.

Plainly, L intended that the penis enter her own vagina, and she did acts in an effort to bring that about. Was that intent sufficient? Or, in addition, did she need to be reckless about the male’s lack of consent? Or, did she need to intend both penetration and his lack of consent?

There has been some division among the jurists on this general issue. As Simester and Brookbanks, Principles of Criminal Law, point out at para 6.1.1, the decision is one of public policy. Criminalising attempts is often justified because of the danger posed to society by a person who sets out to cause harm. That is, one whose intention is dangerous, albeit that his (here, her) acts may not constitute actual harm. This sort of deliberation suggests that intention should be given a narrow meaning, so as to exclude people who are reckless about circumstances such as consent. But, that narrow meaning would create anomalies (as has been pointed out by Professor Ashworth, cited ibid): imagine two men who set out to have sexual intercourse with two women. Both men are reckless as to consent, and neither woman consents. One man succeeds in penetration, and is guilty of rape. The other fails to penetrate, and, under this narrow view, is guilty of neither rape nor attempt to rape.

In L the Supreme Court held that recklessness (in the sense of absence of a reasonable belief in consent) as to circumstances is sufficient for liability for the attempt: what has to be intended here is the act of penetration.

All well and good. We are left with a lingering image of the facts of the case. In future, will people charged with sexual violation claim they did not consent, and that the other person should be charged?

Wednesday, March 29, 2006

Good character again

Once again, see Index, the topic of the significance of absence of a good character direction has concerned the Privy Council: Gilbert v R (Grenada) [2006] UKPC 15 (27 March 2006). Their Lordships indicated that a forthcoming case also concerns this subject: Bhola v The State (Trinidad and Tobago) [2006].

Gilbert makes the point that there is a duty on counsel for the defence to ensure that the trial judge understands that a good character direction is sought: para 11, citing Thompson v R [1998] AC 811 (PC). Omission of the direction requires attention to two matters: was the trial thereby rendered unfair to the accused, and did the omission render the verdict unsafe? (citing for these, Teeluck and John v The State [2005] UKPC 14 at para 39, blogged here, 1 April 2005).

Whereas it may have seemed on some readings that the law requires the direction to be given as a matter of course, in Gilbert it was emphasised that the requirement is that the direction should "normally" be given (para 15, citing R v Aziz [1996] 1 AC 41). That is, the issues of fairness and effect on the result are examined in the particular circumstances of each case. That, of course, is obvious for the result point, and Gilbert makes it clear that fairness is also a matter to be determined on the particular circumstances. There are, thus, no hard or inflexible rules about whether fairness requires the direction, rather the nature of the issues in each case must be examined: para 20, citing Lord Bingham in Singh v The State [2005] UKPC 35 at para 14.

Here, then, the circumstantial evidence of guilt was substantial (para 18), and the accused’s good character was wholly outweighed by the nature and coherence of the circumstantial evidence (applying a phrase used by Lord Hope in Balson v The State [2005] UKPC 6, para 37). The absence of the good character direction therefore had no effect on the fairness of the trial or on its result.

One must observe, with respect, that the question of fairness, which is a procedural matter, will be swamped by the question of result, which is an evidential matter, unless the two are kept distinct. Certainly, the Privy Council here, at para 19, noted that the trial was fair in all other respects, even being "unduly fair" [sic] in that the judge refrained from commenting on the accused's failure to give evidence. Nevertheless, the risk remains that the impression will be given that unfair procedure may be cured by strong evidence, which is undoubtedly not what the Privy Council intended to convey.

The difficulty is that a good character direction is aimed at assisting the jury to evaluate the weight to be given to the evidence, and here it was the complainant's evidence. In fairness terms, the direction is concerned with the avoidance of bias. What is relevant to bias is, not the strength of the prosecution evidence, but whether the absence of the direction put the accused at a disadvantage. On the facts of Gilbert, where the accused was a minister of religion, his previous good character was probably taken for granted by the jury. Therefore, it could safely be concluded that there was no unfairness to him occasioned by the absence of a good character direction.

Tuesday, March 28, 2006

Two points of conduct

In Ebanks v R (Cayman Islands) [2006] UKPC 16 (27 March 2006) the Privy Council reminded us of a couple of fundamental points:

- when a client instructs that he will not give evidence, that instruction must be in writing and signed by the client.

- counsel in cross-examining must put his client’s case, including allegations that the witness is lying, even if the client is not going to give, or call, evidence to establish that fact.

In Ebanks, the appeal against conviction for murder was brought on the grounds of failure by counsel to carry out the client’s instructions in the conduct of his defence, and that this resulted in an unfair trial.

Here again, we have the difficult topic of trial fairness, and again the Privy Council divided 3 – 2 on this issue. We saw a similar split in Howse v R (see these blogs for 10 October 2004), in a Board that was differently constituted but with some overlap in membership. In Ebanks, Lord Rodger gave the decision of the majority (Lords Carswell and Mance agreeing) upholding the conviction, and the dissenters were Lord Steyn and Sir Swinton Thomas. In Howse, the majority, upholding the convictions, were Lords Hutton and Carswell, and the dissenters were Lord Rodger and Sir Andrew Leggatt.

In Ebanks, the reason that persuaded the majority that there had not been unfairness emerged from a consideration of counsel’s conduct at trial, when he appeared to be regularly consulting with his client in a way that was consistent with a carefully conducted defence. In those circumstances, it was possible, notwithstanding the absence of a written record, to reject the suggestion that counsel had not followed the client’s instructions.

The minority in Ebanks focused on the conduct of the hearing in the Cayman Islands Court of Appeal. They held that this had involved a failure of due process, because the Court had wrongly refused to allow the appellant to supplement his affidavit with oral evidence. That Court had also been wrong in its understanding of the duty of counsel in cross-examining to put the client’s case that the police witnesses were lying. The Court had inferred that these allegations were not put because Mr Ebanks had instructed that he would not give evidence, and that his post-conviction claim that he wanted to give evidence was wrong (para 42 per Lord Steyn). This, Lord Steyn (with Sir Swinton Thomas agreeing) held, was a material irregularity, potentially prejudicing the appellant. It seems that in his affidavit filed in the Court of Appeal, Mr Ebanks had not addressed the point about what he had instructed counsel about whether he would give evidence. Lord Steyn considered that if he had been allowed to give oral evidence on that, counsel may well have agreed that he had indeed given that instruction. If that had happened, he would have been denied a fair trial. Sir Swinton Thomas agreed, emphasising the significance of the absence of a written record of the client's instructions, despite his waiver of privilege.

Process takes priority over substance. It cannot be said that guilt can be determined reliably at an unfair trial. Lord Steyn applied, mutatis mutandis, a passage from Wade and Forsyth, Administrative Law, 9th Ed (2005) at 506-508 (see para 40 of his judgment - here I have italisised the passage from the textbook):

" "Procedural objections are often raised by unmeritorious parties. Judges may then be temped to refuse relief on the ground that a fair hearing could have made no difference to the result. But in principle it is vital that the procedure and the merits should be kept strictly apart, since otherwise the merits may be prejudged unfairly. Lord Wright once said:
'If the principles of natural justice are violated in respect of any decision it is, indeed, immaterial whether the same decision would have been arrived at in the absence of the departure from the essential principles of justice. The decision must be declared to be no decision.'
The dangers were vividly expressed by Megarry J, criticising the contention that 'the result is obvious from the start':
'As everybody who has anything to do with the law well knows, the path of the law is strewn with examples of open and shut cases which, somehow, were not; of unanswerable charges which, in the event, were completely answered; of inexplicable conduct which was fully explained; of fixed and unalterable determinations that, by discussion, suffered a change.'
. . . Judges are naturally inclined to use their discretion when a plea of breach of natural justice is used as the last refuge of a claimant with a bad case. But that should not be allowed to weaken the basic principle that fair procedure comes first, and that it is only after hearing both sides that the merits can be properly considered."

Although made in the administrative law context, these observations are also germane to the question whether the Court of Appeal could have declined to hear oral evidence on the ground that it would make no difference. To have decided the case on such a basis would have been unfair and contrary to due process. After all, it is entirely possible (and even likely) that Mr St. John Stevens [counsel for Mr Ebanks at trial]would have made the same core concession that Mr McGrath [Attorney for Mr Ebanks]made, viz that the Appellant insisted that he made no confession to the police."

One must, of course, make allowances for the differing circumstances of each case, but it is interesting to observe that a judge (Lord Rodger) who could uphold the trial in Ebanks as being fair, was such a strong dissenter in Howse.

The dicta in this case about counsel’s duty to put the case to witnesses in cross-examination, even where evidence is not being called to establish the points, in effect amplify Rule 10.02 of our Rules of Professional Conduct for Barristers and Solicitors.

Thursday, March 23, 2006

School Rules, OK!

Yesterday, the House of Lords decided Begum v Denbigh High School [2006] UKHL 15 (22 March 2006), a case that has some bearing on how a provision commonly found in Bills of Rights is to be understood.

The relevant provision in New Zealand is s 5 of the Bill of Rights:

5. Justified limitations - Subject to section 4 of this Bill of Rights, the rights and freedoms contained in this Bill of Rights may be subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society.

In Begum, the interpretation of article 9 of the ECHR was in issue, and it can be seen that art 9.2 corresponds to s 5 of NZBORA:

"Freedom of thought, conscience and religion
9.1 Everyone has the right to freedom of thought, conscience and religion; this right includes freedom to change his religion or belief and freedom, either alone or in community with others and in public or private, to manifest his religion or belief, in worship, teaching, practice and observance.
9.2. Freedom to manifest one's religion or beliefs shall be subject only to such limitations as are prescribed by law and are necessary in a democratic society . . . for the protection of the rights and freedoms of others."


The difference in wording is immaterial, because neither s 5 nor art 9 seeks to establish a set of procedural steps that courts must apply in deciding whether limitations on rights are justified.

Lord Bingham, at para 31, emphasised that what matters is the practical outcome (ie whether a rule is an unjustified limitation of a right), not the quality of the decision making process that led to it (ie not, on the facts of this case, whether the school authorities reasoned correctly when they decided upon the rule).

Lord Hoffmann agreed, saying at para 68:

" … article 9 is concerned with substance, not procedure. It confers no right to have a decision made in any particular way. What matters is the result: was the right to manifest a religious belief restricted in a way which is not justified under article 9.2? The fact that the decision-maker is allowed an area of judgment in imposing requirements which may have the effect of restricting the right does not entitle a court to say that a justifiable and proportionate restriction should be struck down because the decision-maker did not approach the question in the structured way in which a judge might have done. Head teachers and governors cannot be expected to make such decisions with textbooks on human rights law at their elbows. The most that can be said is that the way in which the school approached the problem may help to persuade a judge that its answer fell within the area of judgment accorded to it by the law."
 
Begum suggests that a formalistic approach to the decision, about whether a right has been subject to unjustified limitation, is inappropriate, and that it is the practical outcome that matters. In Begum the school rule that proscribed the wearing of religious clothing was held, in the particular circumstances of this case (it is important to recognise that the Law Lords were not usurping Parliament’s power by seeking to rule on the validity of legislation that gave the power to make rules to the governing body of each school), not to be an unjustified limitation of the right to express religious belief. The school in question had carefully considered its rule about uniforms, in the context of the composition of its roll and the views of the community it served. Lord Bingham concluded, para 34:

"On the agreed facts, the school was in my opinion fully justified in acting as it did. It had taken immense pains to devise a uniform policy which respected Muslim beliefs but did so in an inclusive, unthreatening and uncompetitive way. The rules laid down were as far from being mindless as uniform rules could ever be. The school had enjoyed a period of harmony and success to which the uniform policy was thought to contribute. On further enquiry it still appeared that the rules were acceptable to mainstream Muslim opinion. It was feared that acceding to the respondent's request would or might have significant adverse repercussions. It would in my opinion be irresponsible of any court, lacking the experience, background and detailed knowledge of the head teacher, staff and governors, to overrule their judgment on a matter as sensitive as this. The power of decision has been given to them for the compelling reason that they are best placed to exercise it, and I see no reason to disturb their decision."

In Hansen, blogged here 19 September 2005, alternative interpretations of a section that has a limiting effect on the presumption of innocence are under consideration. The Begum approach would look at the practical outcome of each. One would require an accused to establish innocence on the balance of probabilities. This is a serious limitation on the presumption of innocence. In the absence of any evidence that Parliament deliberately intended (ie intended after deliberation) that that limitation should be imposed, the alternative interpretation, that the section requires the accused to raise a reasonable doubt about his guilt, should be preferred. In deciding whether Parliament intended the more limiting meaning, it must be relevant to note that when the section was recently re-enacted, there was no Report from the Attorney-General to the effect that the provision appears to be inconsistent with any of the provisions of NZBORA, as is required by s 7. In the absence of an express parliamentary intention to limit the presumption of innocence, the courts should not impose such a limitation through interpretation.

On this approach, s 5 works as a guide to Parliament when it considers legislating to limit rights. This point was made by the Chief Justice in argument during the Supreme Court hearing in Hansen v R. It also reminds courts engaged with provisions that may impinge on rights to look for indications of legislative intent to limit rights.

Friday, March 17, 2006

Rights and Power

This week’s decision of the European Court of Human Rights (Fourth Section) concerning Application no. 23276/04 by Saddam Hussein draws our attention to some characteristics of rights and jurisdiction.

Ideally, human rights should be enjoyed by everyone, no matter where they happen to be. Rights, however, mean nothing unless they can be enforced. Enforcement involves an assertion and exercise of jurisdiction by an authority with power. Rights ultimately depend on power.

It was argued for the Applicant that he had rights under the European Convention on Human Rights, because he was arrested in an area where jurisdiction was exercised by countries, as part of a coalition, that included countries that are bound by the Convention:

"He maintained that he fell within the jurisdiction of all the respondent States because they were the occupying powers in Iraq, because he was under their direct authority and control or because they were responsible for the acts of their agents abroad. He further argued that he remained within their jurisdiction following the transfer of authority, and his transfer, to the Iraqi authorities in June 2004 because the respondent States remained in de facto control in Iraq."

The Court rejected these submissions because they were not substantiated by evidence of the kind of power and control necessary to establish jurisdiction.

The sort of argument advanced in this case calls to mind the current international expansion of criminal jurisdiction without there being a need for territorial control. There are now numerous examples. One in New Zealand law is s 12C of the Misuse of Drugs Act 1975, which provides that every person commits an offence against this Act who, outside New Zealand, does or omits to do any act that would, if done or omitted in New Zealand, constitute an offence against s 6 (dealing with controlled drugs), s 9 (cultivation of prohibited plants), s 12A (manufacturing, producing, supplying equipment or precursor substances for use in manufacturing or cultivating), s 12AB (knowingly importing or exporting precursor substances for unlawful use), or s 12B (laundering proceeds of drug offences). Consequently, courts in New Zealand have jurisdiction over the proscribed behaviour no matter where in the world it occurs. The offender need not be a New Zealander, and the offence need have nothing to do with New Zealand. Once the offender is in New Zealand territory he can be charged. See also s 7A of the Crimes Act 1961 for other examples of this sort of extension of jurisdiction.

Jurisdiction to punish does not, as these examples show, depend on the exercise of territorial control. Are rights limited by territorial control?

If an Iraqi committed in Iraq what we in New Zealand call a drug dealing offence, and then came to New Zealand, it is likely that he would enjoy the protection of our Bill of Rights, and of the International Covenant on Civil and Political Rights, to which New Zealand is a signatory. So, if evidence had been obtained against him in Iraq by methods that would have contravened his rights if he were in New Zealand, a court in New Zealand would apply New Zealand law to determine the admissibility of that evidence on a charge for the offending against s 12C.

For an analogous illustration, see the rejection by English courts of evidence obtained by torture, including torture in a foreign country, discussed in these blogs on 11 August 2004.

In this sense, then, rights do extend beyond the limits of territorial control of the State in which they may be recognised. Was the European Court wrong to require a demonstration of power and control by the respondent States?

The difficulty faced by the Applicant is that the power to try him is asserted by the Iraqi Government, the existence of which is endorsed by the UN Security Council, and that Government is not a party to the ECHR which establishes an obligation to enforce the rights on which he sought to rely. To enforce those rights, the Applicant would have to be tried by a court in one of the respondent States.

Thursday, March 16, 2006

Legality v Common sense

Should one feel sorry for a person who, having served a sentence of imprisonment, has his convictions quashed on appeal because the evidence against him should not have been ruled admissible?

A concurrent sentence of two years imprisonment for two aggravated robberies was served by the appellant "S", a 15 year-old. Then, on appeal, two High Court judges sitting together held that the evidence of his guilt, which was entirely contained in his statement to the police, should have been excluded, and his convictions were quashed: S v Police 14/3/06, Baragwanath and Heath JJ, HC Auckland CRI 2004-404-515.

In reality, unless his confessions were false, S was guilty and, no doubt, he deserved his sentence. But that is to ignore the rule of law, which in this context means that punishment can only be imposed following due process of law. If by "due process" we mean upon lawful conviction, then S did not receive due process.

The problem arises from practicalities. Sentences take effect before rights of appeal against convictions have been exercised. Trials proceed before rights of appeal against evidential rulings have been exercised. Evidential rulings during a trial must be left for appeal after completion of the trial, otherwise trials would have to adjourn to allow appeals to be heard.

The obvious course, to minimise "false" imprisonment, would be to defer sentencing hearings until appeals had been determined, but that is often impractical: many convicted people would run away, or would have to be denied bail.

If there is no viable alternative to the present procedure, should people like S be compensated? How would compensation be measured? Our assessment of the value of due process may have to be balanced against the harm caused by the offender. The courts may be forced to say that the real remedy is the vindication that follows from the quashing of the conviction: see further, my blogs of 6 and 30 March 2005.

In S v Police, a police officer had told S that he wanted him to point out where burglaries had occurred, so that files could be cleared up, and that he would not be charged with committing them. Unexpectedly, from the officer’s point of view, S said he had committed a couple of aggravated robberies. He was interviewed about those, he confessed to them, and was charged. If his initial mention of them had been a "spontaneous" utterance, his consequential dealings with the police would not have been flawed, as the correct procedures for dealing with young people had been followed thereafter. However, if the initial reference to the robberies was not spontaneous, in the sense that it was made in reliance on the officer’s assurance that he would not be charged, then it was made in circumstances that were in breach of the statutory procedures, and what followed, although correct in itself, could be tainted.

The High Court held that the utterance was not spontaneous, but rather it had been made in reliance on the officer’s assurance. There was an error of procedure surrounding that, so the question became whether the following procedures, which led to the making of the confession, were indeed tainted.

The Court found it necessary, at this point, to consider the nature of causation (para 56 – 67). This came down to asking whether, in the circumstances, the non-spontaneous utterance had been the "effective cause" of the subsequent confession. It had, and because the use of an inadmissible statement to secure a subsequent statement was contrary to the policy of the legislation (para 59), both were inadmissible.

A lot turned, in this case, on how the judges assessed the circumstances. It could easily have been decided that S had, on receiving the advice to which he was entitled, decided to make a clean breast of everything, and that that decision, rather than the earlier indication that he would not be charged, was the effective cause of his confession. In cases like this, where the assessment of the circumstances seems to be finely balanced, it is tempting (but, in law, wrong) to think that the court was influenced by the common sense solution: S had served his sentence and may as well be relieved of the convictions to give him some incentive to obey the law.

Did the absence of any compensation for the "wrongful" imprisonment facilitate this balancing of the circumstances?

Tuesday, March 14, 2006

Inferences of innocence

A commonly advanced ground of appeal against conviction is that the evidence supported an inference of innocence and the judge did not correctly tell the jury how they should deal with inferences.

We have passed through a period when much attention was given to what inferences are and how they should be handled. Things got rather complex as efforts were made to be precise about this. The relationship between the process of drawing inferences, and the process of deciding whether something has been proved, was at the centre of this complexity.

Some facts are proved by direct evidence: for example, a witness says that he saw something happen. Other facts are inferred from direct evidence: the facts that are directly proved suggest that something else is true. In turn, these inferred facts may combine, with other inferred facts, or with directly proved facts, or with a mixture of inferred facts and direct facts, to enable a further inference that something else is true. Wherever an inference occurs, it is usually described as an ordinary process of logical thought.

As to proof, it is relatively easy to see that evidence of directly observed facts may prove those facts, to the necessary standard. If the fact is an element of the alleged offence, then, when the jury considers all the evidence in the case, it must decide whether that element has been proved beyond reasonable doubt. But, in relation to other facts, (and with some exceptions), no particular standard of proof is required. The jury does not have to be instructed by the judge about the standard to which they must be satisfied that inferred facts are proved before they can use them to support other inferences, such as an inference of an element of the offence, as long as, ultimately, the jury tests the proof of the elements of the offence against the standard of beyond reasonable doubt.

So, what about the inferences supporting innocence? In what has been regarded as the leading case on inferences in New Zealand, R v Puttick (1985) 1 CRNZ 644 (CA) the Court summarised the position:

"Inference is simply one of the mental processes which may be used by a jury in carrying out its primary task of assessing the evidence and deciding whether or not it establishes the guilt of the accused beyond reasonable doubt. Where the charge has several essential elements, proof of guilt necessarily involves proof of each of those elements to the same standard. It does not, however, require proof beyond reasonable doubt of every fact which may be relevant to proof of each essential element.
… It must be equally unhelpful to tell jurors that, if proven facts support two inferences of equal weight, they should accept one and reject the other. To draw an inference either way from such facts would be pure speculation. Jurors should not be directed to accept or reject inferences when they have no logical basis for either step."


This could suggest (although we can be sure the Court did not intend this reading) that an inference supporting innocence can neither be accepted nor rejected if it is equal in weight to an inference of guilt.

A clearer account has been given in R v Seekamut 10/7/03, CA82/03:

"If on an objective basis that has regard to all the circumstances, a rational alternative to guilt is not excluded, there must for that reason be a reasonable doubt. But the mere fact that some of the circumstances might arguably permit an inference inconsistent with guilt is not enough. The jury’s function is to assess the whole of the evidence and in so doing may conclude that a suggested alternative is not reasonably tenable."

Similarly, yesterday the Privy Council in Taylor v R (Jamaica) [2006] UKPC 12 (13 March 2006), para 18, held:

"Their Lordships agree with the submission made on behalf of the appellant that in the circumstances of this case it was essential that the judge … spell out the possible inferences to be drawn … and instruct them that they must rule out all inferences consistent with innocence before they could be satisfied that the inference of guilt has been proved correct."

This could be read as suggesting a slightly different approach to that in Seekamut: instead of treating the evidence as all being in a big pool, to be considered in the round to see what ends up being proved, the process alluded to in Taylor seems to be one of deciding first whether inferences of innocence can be ruled out, then, if they are, turning to see whether guilt has been proved. We might, however, reasonably wonder whether this explication is correct, insofar as it seems to place a burden of proof on the defence, and it also seems to make an artificial distinction between stages of the reasoning process. The correct approach would be to tell the jury that, if after considering the evidence, they are left with a reasonable doubt about the accused's guilt, they must find him not guilty.

Because the judge in Taylor failed properly to direct the jury in this regard, the appeal against conviction was allowed and the case remitted to determine whether there should be a retrial.

Monday, March 13, 2006

Metaphysical laws

Can the law have been other than it was understood to have been at the time?

If it can, are we currently operating under a misunderstanding of the law?

Are there points of law that we are mistakenly taking for granted?

If we all think that the law says one thing, can we be wrong?

Can some future court say we were wrong, and that the law said something else?

The first of these questions concerned the Privy Council last week in Bowe v R (Bahamas) [2006] UKPC 9 (8 March 2006). The issue involved detailed analysis of legislation, but in essence the question came down to whether the death penalty was mandatory in the Bahamas in 1973, notwithstanding that cases had been conducted on the basis that it was.

The Privy Council answered the conundrum this way (para 42):

"… it took some time for the legal effect of entrenched human rights guarantees to be appreciated, not because the meaning of the rights changed but because the jurisprudence on human rights and constitutional adjudication was unfamiliar and, by some courts, resisted."

Further, it would be wrong to answer the question by now trying to assess what chance the argument would have had of being accepted in 1973.

Instead,

"The task is to ascertain what the law, correctly understood, was at the relevant time, unaffected by later legal developments, since that is plainly the law which should have been declared had the challenge been presented then." (ibid)

So,

"It matters little what lawyers and judges might have thought in their own minds: in the context of a codified Constitution, what matters is what the Constitution says and what it has been interpreted to mean. In 1973 there was no good authority contrary to the appellants' argument, and much to support it. In the final resort, the most important consideration is that those who are entitled to the protection of human rights guarantees should enjoy that protection. The appellants should not be denied such protection because, a quarter century before they were condemned to death, the law was not fully understood." (ibid)

At this point, it seems to be only a subsidiary consideration that there was no authority to the contrary as at 1973, as the governing matter ("the final resort") is that those who are entitled to human rights protection should be entitled to enjoy that protection. Question-begging, of course, as the issue was whether the people were (not "are") entitled to that protection.

Not to worry, for in Bowe the death sentences were quashed because the law as at 1973 was that the sentence was discretionary, not mandatory.

This interpretation was therefore carried through to the present pursuant to the legislative provisions. The cases of the two appellants were remitted for consideration of the appropriate penalty.

It seems, consequently, that we are governed by laws that we have not yet discovered, and may not be discovered until we are all dead.

Thursday, March 09, 2006

What a "fair trial" means

Today, the High Court of Australia decided Nudd v R [2006] HCA 9 (9 March 2006), a case with helpful dicta on the meaning of "fair trial". I have recently noted cases concerning what is a substantial miscarriage of justice (see Index), and have described the two ways in which miscarriages of justice may be considered "substantial" for the purposes of the proviso: by depriving the accused of a real chance of an acquittal, or by causing the trial to be unfair.

Nudd recognises that, even though the evidence of guilt may be overwhelming, there may be rare cases where there has nevertheless been a miscarriage of justice because the trial was rendered unfair to the accused: see Gleeson CJ at para 6, Kirby J at 87 and 100, Callinan and Heydon JJ at 158. The other judges, Gummow and Hayne JJ jointly decided that no miscarriage of justice had been demonstrated to have occurred (para 25). Gleeson CJ agreed with that conclusion.

Kirby J treated the outcome aspect (the depriving of the accused of a real chance of acquittal) in the same way as the other judges, holding that on any approach, there was no prospect of acquittal as the prosecution case was overwhelming, but he also gave the most detailed consideration to the other aspect, that of the fairness of the trial. He found this the more difficult point to decide (para 106), because the defects at trial (counsel’s serious incompetence, especially in not knowing the law and in failing to take detailed instructions and consequently to advise the client) were so serious. He held that this case was "borderline" on the question of fairness (para 109), but that the decisive point was that the evidence against the accused was conclusive of guilt.

This might be thought to muddy the distinction between the outcome aspect and the fairness aspect of substantial miscarriage of justice. However, I suggest that Kirby J’s reasoning is essentially that, from the point of view of fairness, the errors were inconsequential, because the defence in the trial had (and could only have) been to put the prosecution to proof, and the judge in her summing up had properly directed the jury on the law.

I have previously suggested that a fair trial is one where the law is accurately applied to the facts determined without bias. Applying this to Nudd, we can say that the errors at trial (counsel’s incompetence) did not cause the law to be applied inaccurately, nor did it cause the jury to be biased against the accused.

Thursday, March 02, 2006

Bad technique!

Sometimes, evidence that the accused has been convicted for committing other offences is admissible to prove that he committed the offence with which he is now charged. Or again, allegations that he committed other offences may be used to prove that he committed this one. Or, yet again, allegations of offences for which he has been acquitted may be used as evidence that he is guilty of this one.

This sort of evidence, called similar fact evidence, should only be allowed to be used in exceptional circumstances, for the obvious reason that it will have a tendency to bias the tribunal of fact (usually the jury) against the accused.

It must be said that the current law on similar fact evidence in New Zealand is in need of clarification. This is largely due to the decision of the Court of Appeal in R v Holtz [2003] 1 NZLR 667; (2002) 20 CRNZ 14 (CA), where the criterion for admissibility was held to be whether the probative value of the similar fact evidence outweighs its illegitimately prejudicial effect:

"One of the problems with evidence of past conduct, whether labelled evidence of bad character, propensity evidence or similar fact evidence, is the tendency to look for principles of admissibility applicable to all such evidence in all circumstances. At a very general level there is the broad principle that to be admissible the evidence must be such that its probative value outweighs illegitimate prejudice to the accused in having adduced evidence of past conduct that might be given undue weight or used improperly in reasoning towards guilt of the crime charged. It is in the application of that general principle in the circumstances of particular cases that difficulties arise. Often this is because the approach adopted in one set of circumstances is taken as the rule to be applied in a quite different case. Because of the wide variation in circumstances in which the issue calls for consideration, the preferable approach in a particular case is that which determines and weighs probative value and potential prejudice in the circumstances. It is necessary to consider the nature of the evidence, what it is sought to prove, what other evidence there is and its relationship and whether the evidence relates to prior proved offending or to concurrent charges, all against the underlying dangers inherent in propensity or bad character evidence." (para 35)

The so-called weighing of probative value against prejudicial effect is problematic conceptually, and has been criticised in the leading Australian case on similar fact evidence, Pfennig v R (1995) 182 CLR 461 per McHugh J on the basis that these were "incommensurables", and the Supreme Court of Canada has agreed with his criticism, in R v Handy (2002) 164 CCC (3d) 481 (SCC); 213 DLR (4th) 385, at paras 148-149. And I have written an article on it: "Probative value, illegitimate prejudice and the accused's right to a fair trial" (2006) 29 Crim LJ 8.

In Phillips v R [2006] HCA 4 (1 March 2006) the High Court of Australia has confirmed that Pfennig continues to be the authority on similar fact evidence, binding on Australian courts. It issued a thinly-veiled rebuke to courts that sought to find reasons for departing from Pfennig, at para 60:

"It must be said at once that it is for this Court alone to determine whether one of its previous decisions is to be departed from or overruled …. Of course, in criminal cases it is often necessary for trial judges and Courts of Criminal Appeal to elaborate upon rulings of this Court; to gather together rules expressed in several cases; to apply rules to different facts; and sometimes to reconsider rules affected by later legislation. Within spaces left by the binding determinations of this Court, trial judges and intermediate courts retain their proper functions …. However, these do not extend to varying, qualifying or ignoring a rule established by a decision of this Court. Such a rule is binding on all courts and judges in the Australian Judicature."

One misinterpretation of Pfennig, alluded to briefly in Phillips at para 62, is that it is appropriate and sufficient to regard similar fact evidence as a form of circumstantial evidence. This is what had occurred in Holtz. See Adams on Criminal Law, Ch2.8.19(5), and my blog entry for 14 October 2004. A danger of that view is that it leaves the process of drawing inferences unelaborated by particular directions to the jury, as authorities on inferences tend to avoid such assistance. It also tempts the court to admit the evidence, as circumstantial evidence is just a category of admissible evidence.

Phillips and Pfennig hold that similar fact evidence must pass a high threshold before it can be admitted (para 79):

"Criminal trials in this country are ordinarily focused with high particularity upon specified offences. They are not, as such, a trial of the accused's character or propensity towards criminal conduct. That is why, in order to permit the admission of evidence relevant to several different offences, the common law requires a high threshold to be passed. The evidence must possess particular probative qualities; a strong degree of probative force; a really material bearing on the issues to be decided. That threshold was not met in this case. It was therefore necessary that the allegations, formulated in the charges brought against the appellant, be separately considered by different juries, uncontaminated by knowledge of other complaints. This is what Pfennig and other decisions of this Court require. To the extent that O'Keefe [a Queensland Court of Appeal decision critical of the High Court of Australia's judgments in Pfennig] or other authority suggests otherwise, it does not represent the law. No other outcome would be compatible with the fair trial of the appellant."

Phillips had been charged with 8 counts of sexual offending, and had unsuccessfully sought severance of those. Joinder was permitted on the basis that the evidence on each charge was admissible on the others, as evidence of similar facts. The High Court summarised the evidence and its conclusion in this way (para 56):

"The similarities relied on were not merely not "striking", they were entirely unremarkable. That a male teenager might seek sexual activity with girls about his own age with most of whom he was acquainted, and seek it consensually in the first instance, is not particularly probative. Nor is the appellant's desire for oral sex, his approaches to the complainants on social occasions and after some of them had ingested alcohol or other drugs, his engineering of opportunities for them to be alone with him, and the different degrees of violence he employed in some instances. His recklessness in persisting with this conduct near other people who might be attracted by vocal protests is also unremarkable and not uncommon."

Bad technique, indeed, as it got Phillips to court.

The trial judge, and the Queensland Court of Appeal, had reasoned that so many similar complaints could not have arisen by chance, so together they had credibility. The High Court was not so slap-dash. It identified the issue on which the challenged evidence was relevant as being consent, and it pointed out that the non-consent of one complainant could not possibly prove that another had not consented.

One thing this case illustrates is the way judges can lose their reasoning powers in cases of alleged sexual misconduct. There is, in other words, a grave danger of judicial bias (probably unconscious bias) in such cases. In those circumstances, clear rules, rather than discretions, are needed concerning the admissibility of evidence of other alleged misconduct. The Pfennig rule is that similar fact evidence is not admissible unless there is no reasonable view of it that is consistent with the accused’s innocence. It must have strong probative force, a really material bearing on the issues to be decided; its probative force must clearly transcend its merely prejudicial effect; it must have a sufficient nexus, some specific connection, with the issues; its probative force must be such as to make it just to admit it (Phillips, para 54). None of those criteria were met in Phillips (para 55).

It is sad to note that in New Zealand, the Evidence Bill proposes an approach to similar fact evidence (or, in its terminology, "propensity evidence") that applies the probative value/prejudicial effect weighing, criticism of which is noted above. Clause 39(1) of the Bill provides:

"The prosecution may offer propensity evidence about a defendant in a criminal proceeding only if the evidence has a probative value in relation to an issue in dispute in the proceeding which outweighs the risk that the evidence may have an unfairly prejudicial effect on the defendant."

This proposal does nothing to improve our law on this point.

Monday, February 20, 2006

Overcoming the void

As everyone knows, the law of Scotland has its origins in Roman-Dutch law. Its technical language can look a little strange to English jurists. A recent Privy Council decision on Scottish law illustrates how light can be shed on common issues. Here, in Ruddy and others v Procurator Fiscal [2006] UKPC D2 (6 February 2006), the Judicial Committee leant heavily on the expertise of Lord Rodger, all members agreeing with his opinion.

I will describe the case, insofar as I can, in English.

If a judicial proceeding is fundamentally flawed, so that in oft-used terminology it may be called "null and void", or "a nullity", may its result nevertheless be regarded as final?

These appeals concerned convictions and sentences that had been imposed by a tribunal that had no jurisdiction to act. They therefore, it was agreed, were in breach of the appellants’ rights to a "hearing" under s 6 of the ECHR. They had delayed 2 years before complaining of that by bringing the actions that resulted in these appeals.

Two concepts that operate in the context of errors are waiver and acquiescence. Waiver is active, and acquiescence is passive, acceptance that the error does not matter. They can operate where there is delay in seeking a remedy. In the absence of an applicable time limit on seeking redress, action must be taken within a "reasonable" time, as policy favours finality and stability. Acquiescence may be inferred from delay and the circumstances of the case (see paras 28, 37, 38, 46).

Here, if objection had been taken promptly, re-hearings could have been ordered before properly constituted tribunals and proper convictions would, inevitably, have been obtained (para 41). If a remedy was now allowed, justice might not be done in view of the effect of the passage of time.

Lord Rodger pointed to examples where errors causing the proceedings to be a "nullity" were nevertheless cured by acquiescence (para 48). Central to the acceptability of this is reasoning along these lines: a defect can be waived before the conclusion of the proceeding, so there is no reason why acquiescence after it cannot be effective (para 50). Lord Carswell elaborated this at para 58, "…in either case … [the defendant] is representing, expressly or tacitly, that the court is acceptable to him …".

In essence, a flaw in the proceedings can be waived or acquiesced in, even if it is fundamental.

This is potentially relevant to a proceeding which the Supreme Court of New Zealand will hear this week: Hansen v R (previously noted in these blogs on 19 September 2005). If the Court decides that the reverse onus provision in question, s 6(6) of the Misuse of Drugs Act 1975, puts merely an evidential, not a legal, burden of proof on the defence, questions may arise about convictions that have been incurred in trials where juries have been directed differently. While a misdirection on the standard of proof will normally result in the quashing of a conviction, it is, arguably, not so fundamental an objection as lack of jurisdiction by the tribunal. Nevertheless, given that it was possible to anticipate that the New Zealand law on reverse onuses would be brought into line with English law, can it now be said that people who were convicted before the change (assuming, of course, that Hansen does change the law) acquiesced in the error?

A straightforward solution to this would be to say that Hansen changes (again, assuming that it will change) the law from the date it is decided, and that earlier trials (but not Hansen’s, as these are still his proceedings) where heard under the law as it was then correctly understood to be.

Friday, February 17, 2006

The "overarching requirement" of fairness

The interrelationship between the requirements that statements be made voluntarily, and that trials be fair, was central to the unanimous decision of the Judicial Committee in Shabadine Peart v R (Jamaica) [2006] UKPC 5 (14 February 2006).

Which is the dominant requirement? If a confession was made voluntarily, should it always be admissible? If it was obtained wrongfully, but still given voluntarily, should it always be admissible? Obtaining a statement wrongfully can result in lies being told which unfairly diminish the credibility of the defence. This was the position in Shabadine Peart. Should such a statement be admissible, because it was made voluntarily, even though it could result in unfairness at trial? Is it necessary that the effect of the misconduct in obtaining the statement would be to make the trial unfair, or can exclusion of the wrongfully obtained statement be justified on broader grounds of public policy?

I should note, at this point, that this case focuses on the effect of admission of the wrongfully obtained statement on trial fairness, when that matter would not arise unless the statement should have been excluded on public policy grounds. If there were no public policy grounds to exclude the statement, there could be no trial fairness objection to its admission. Yet, once there were sufficient public policy grounds to exclude it, the appellate court had to deal with the fair trial implications of the error.

Well, aside from that matter, what does Shabadine Peart decide about trial fairness? It follows a line of cases, not cited in the judgment but referred to frequently in these blogs, that puts fairness as the overarching criterion (para 23). The most important facet of fairness is the voluntariness of the statement by the accused, and also relevant are factors such as his youth, and whether he received legal advice before making the statement.

The case involved breaches of Rule 3(b) of the Judges’ Rules, which concerns restrictions on questions that may be put to the accused after he has been arrested. The Privy Council doubted that in the circumstances the statement could be regarded as voluntary, but, even if voluntary, it was unfair to admit the statement in evidence (para 29). This was because it contradicted some of the evidence given or called by the accused at trial, and could therefore be used by the prosecution to diminish the credibility of the defence.

Well, it would only be wrong to admit such a statement if it had been wrongfully obtained in such a way as to give rise to public policy concerns over disrepute to the administration of justice. In my view, as stated above, the Board's comments on trial fairness were unnecessary and misleading.

At this point in the judgment it was necessary to consider the application of the proviso – itself an interesting topic, considered in these blogs. It was held that if the statement had been excluded, the course of the trial may have been very different and the defence may have succeeded (para 30). This is recognisable as the criterion of whether the error at trial had resulted in the loss of a real chance of acquittal. That was sufficient to prevent application of the proviso. It should be noted that the Board is not, here, relying on the trial fairness aspect of application of the proviso, but instead is, properly, determining whether the error at trial had had significance for its result. Inconsequential errors in the application of the public policy discretion do not, by definition, matter.

Some aspects of Rule 3(b) are also mentioned. These include the history and rationale for the Rule, the distinction between the suspicion and accusation phases of the police inquiry, and judicial experience of the tendency of those who have told the truth to tell lies to avoid pressure (paras 18 – 20).

Wednesday, February 15, 2006

Easy does it!

The Supreme Court’s decision in Thompson v R [2006] SCNZ 3 (15 February 2006), in which answers to questions at trial were held not to have caused unfairness, gives us an opportunity to reflect on the art of cross-examination.

In my view, the following ten are fundamental points that must be remembered by counsel faced with the task of cross-examining a witness:

1. Find out beforehand the worst that the witness could say, from your client’s point of view.

2. It is unlikely that the witness will be compelled, by your questioning, to tell lies that suit your client.

3. If you are trying to get the witness to change an assertion, it may be best to start by suggesting reasons the witness may be honestly mistaken. This is because an aggressive approach to cross-examination is less likely to find sympathy with the jury than an approach which assumes the witness will in good faith acknowledge ordinary human failings.

4. An aggressive approach is unlikely to succeed with an aggressive witness. Gently hinting at common sense reasons for the witness’s error is preferable, especially if the jurors are likely to accept them.

5. Where the witness can be contradicted by his own previous statement, make the most of that by getting him to give as much "detail" as possible first, before putting the contradicting statement. This applies as long as the witness is unlikely to have been wrong in the previous statement, and where that earlier statement is more favourable to your client than the witness’s present testimony.

6. There is a duty to put your client’s case to the witness. This can be done point by point, in questions that take the form "Didn’t X?", "Didn’t Y?" etc. This is stronger than the formula that is commonly encountered: "My client will give evidence and say that X, what do you say to that?"

7. Don’t ask unnecessary questions. Don’t ask questions that are open-ended, such as "Why do you say that?" Lead the witness as much as possible, to give your questioning a focus that the jury understands. One of the purposes of cross-examination is to get the witness to say what you want him to say, to score a point and/or to undermine your opponent’s case.

8. Don’t ask questions that you don’t know the answer to, unless the whole tenor of the case points in the direction of the answer you want and any other answer would lack credibility.

9. If you have to attack the witness as a liar, remember that this strategy may result in the jury learning of your client’s own bad character (including previous convictions).

10. If your client is of good character, and if you rely on that as part of your case, take opportunities to use it in cross-examination; this can be tactically advantageous as it should result in the judge directing the jury on your client’s good character.

In Thompson, experienced trial counsel was drawn in to engaging in what the Court of Appeal called a "slugfest" with the witness: see R v Thompson 16/6/05, CA445/04, para 56. This, of itself, was not necessarily an error, as the Court of Appeal noted, at para 66:

"… robust cross-examination is one of the many options open to counsel, who must be accorded wide discretion."

But the Court added, para 69:

"Generally, counsel find it politic to conduct cross-examination with ‘a measure of courtesy’ …."

The conclusion, upheld by the Supreme Court, was that the evidence that was inadmissible but which was given in cross-examination, when seen in context, and in the light of a strong warning by the judge to the jury about its use, did not cause a miscarriage of justice.

Thursday, February 09, 2006

Rushing to judgment

Cutting to the chase in trials can look unfair. Where legal submissions are necessary, judges may be tempted to deal with them peremptorily, risking the appearance of bias. The bounds of proper judicial conduct were crossed at the trial that was the subject of the appeals jointly heard in Antoun v R [2006] HCA 2 (8 February 2006).

The judges of the High Court of Australia all agreed that the trial had been handled in a way that created the appearance that the judge was biased against the defence. This "apprehended bias", is

"the appearance of a possibility of an absence of an impartial mind on the part of the judge"

(per Callinan J at para 83, with whom the other members of the Court agreed). As Kirby J noted, para 28:

"…The manifest observance of fair procedures is necessary to satisfy the requirements not only of fairness to the accused but also of justice before the public so that they may be satisfied, by attendance or from the record, that the process has followed lines observing basic rules of fairness. Excessively telescoping the procedures in such cases can lead to a sense of disquiet on the part of the accused, and of objective observers whose attitudes, where relevant, must be represented, and given effect, by appellate courts."

In this case the judge, on being told at the close of the prosecution case, that the defence counsel for each accused would make submissions that there was no case to answer, immediately said that such submissions would be rejected. This was said before an opportunity to make those submissions was given. It was a trial by judge alone, but that is not material because even in a jury trial such submissions would be heard in the absence of the jury.

Callinan J held that the judge’s conduct gave an appearance of pre-judgment, and therefore, para 86:

"It follows that the apparent strength of the respondent's case, and the weaknesses of the appellants' defence cannot be used as justification or excuse for the trial judge's expressions of a determination to reject submissions foreshadowed, but not yet made and developed."

Gleeson CJ put it this way, para 23:

"…The judge regarded this as a strong case of extortion. He formed the view, with good reason, that the no case to answer submission was likely to be implausible. Yet he should not have decided to reject it without giving counsel an opportunity to put the argument. In the circumstances, that would not have required much time. The way in which the judge dealt with the no case argument, and later with the question of bail, gave rise to an appearance of lack of impartiality. Strong as the case against the appellants appeared to be, they were entitled to a fair hearing."

This case did not include submissions on whether the proviso could be relied upon to dismiss the appeal. However, Kirby J noted, para 49, that the denial of the opportunity to make submissions before decision would probably be a sufficient miscarriage of justice to make the proviso inapplicable, and that, had the case been argued on the proviso, it would have required consideration of Weiss v The Queen [2005] HCA 81 at [45] (see these blogs, 16 January 2006).
We might, by now, anticipate that a case of apprehended bias would be a substantial miscarriage of justice, given the fundamental importance of the appearance of justice, but one must acknowledge that there is, at least, an argument that whereas actual bias would be a substantial miscarriage of justice, apprehended bias falls short of that in cases where there was no loss of a real chance of acquittal.

Consideration of Antoun raises a number of fundamental questions. Is there a difference between so-called "apprehended" bias and "actual" bias? If there is such a thing as apprehended bias, does it fall within the scope of the accused’s right to a fair hearing, or is it an associated right? Current jurisprudence suggests that if it is the former, it is absolute, but if it is only an associated right then it is subject to balancing against competing values. Doesn’t the actual decision in Antoun (that apprehended bias requires a retrial even where the judge’s ruling was correct) mean that the real issue in the case was not whether the hearing had involved bias (whether actual or apprehended), but rather whether there had been a "hearing" on the relevant issue? On analysis, the real reason a retrial was necessary in Antoun is that the accused had been denied the right to a hearing on the issue of whether there had been a case to answer.

Friday, January 27, 2006

Fingered!

What makes fingerprint evidence interesting is the increasing unlikelihood that we all have different fingerprints. This, at least, is what ordinary old you and I might think. Fingerprints are not genetically determined: identical twins have different fingerprints. While their design falls broadly into a few types, at a finer level of detail they seem to be formed randomly.

There are more human fingers than there ever have been. Often, so-called identifications are made from incomplete prints left at the scene of a crime. Part of the print of one finger, it is said, may be sufficient for an expert to say that it is from the accused. However many fingers there are, there are many more "parts of fingers". Yet they are all, so it is said, sufficient to mark us out individually. This compression of information, in a small area of skin, is truly remarkable. Mathematicians can assert that there are something like 10 to the power of 20 different possible patterns in a small area of finger, so, if we accept that, our suspicions about the uniqueness of fingerprints are needlessly alarmist.

Let us assume, however, that the experts are not complacent and that they search with sustained vigour for people who have identical "parts of fingerprints". This amounts to assuming that the experts, who have made a career in reliance on the theory that we all have fingerprints that are unique, are prepared to risk making a discovery that will undermine their livelihoods. I exaggerate, of course: even if many people on earth had the same fingerprints, the finding of one at a crime scene that was, after allowance was made for the distorting effects that occur in real life, not significantly different from the accused's, would, on Bayesean analysis, be highly probative.

Far from being suspicious, the courts are apparently becoming more relaxed about fingerprint evidence. There is judicial acceptance of a trend away from the basing of fingerprint identifications on the number of specified points of identity as between the accused’s print and the sample found at the crime scene. Instead of showing the court enlarged photographs with arrows specifying points on the prints that the court should accept as demonstrating identity, the fingerprint experts are now permitted to base their opinions on the "quality and quantity of information in the images": R v Carter 19/12/05, CA155/05, a phrase used in a technical paper quoted by the Court at para 75.

Important, in this approach, is peer review of the expert witness’s opinion. This review must, in England and Wales, be carried out by three "independent" examiners.

In Carter the Court did not go so far as to say that three independent examiners are needed as peer reviewers before evidence of fingerprint identity will be admissible in New Zealand. Plainly, there would be difficulties finding people who are "independent", let alone people who are sufficiently experienced to qualify as experts, in a pool of talent that is, presumably, rather small. Nevertheless, the Court did reiterate a requirement that Tipping J had made of expert witnesses in the well-known case of R v Calder 12/4/95, HC Christchurch T154/94 (where the accused was ultimately acquitted of poisoning), that the evidence needed to have a "minimum threshold of reliability" to be admissible.

So, the need for the judge to warn the jury to exercise caution before accepting evidence of fingerprint identification where there are relatively few points of comparison (R v Buisson [1990] 2 NZLR 542 (CA)) has apparently been replaced by a more relaxed approach. In Carter, the expert was unwilling or unable to indicate whether there were any points of comparison identified (para 60), but the Court of Appeal held that no particular direction to the jury was needed on this, because the defence, in its closing, had not suggested to the jury that the expert was wrong. Instead, the defence, when the time came to close its case, had merely suggested that there were innocent explanations for the accused’s fingerprint being at the scene.

Incidentally, the real question on which the expert's evidence is relevant, is whether there are any significant differences between the print found at the scene and the accused's fingerprint. As a prosecution witness, the expert will be expected to say there are none, and it is his explanations for apparent differences that are material.

One might wonder whether, in a case where defence counsel had apparently been discouraged (the judge having ruled the expert’s evidence admissible) from a full assault on the credibility of the expert, and where peer review witnesses were not called, the jury were given a proper opportunity to reject the fingerprint evidence.

This raises the question of when a decision made as a matter of trial tactics can be regarded as an error that caused a loss of a real chance of acquittal (assuming that if there was a doubt about the fingerprint identification then there would have been a doubt about guilt); as the Supreme Court held in Sungsuwan (blogged here, 26 August 2005), it is the effect of the error, not its magnitude, that matters.

Update: for a critique of fingerprint evidence, see Gary Edmond, "Fingerprint Evidence in New Zealand's Courts: the Oversight of Overstatement" (2020) 29 NZULR 1. And further: Gary Edmond, Fingerprint Evidence in New Zealand's Courts: the Limits of Legal "Challenge" (2020) 29 NZULR 347.

Tuesday, January 24, 2006

Index to these blog entries

To access my Index to all the entries on this blogsite, up to the end of the last month, click on the link on the right.

Recklessness

What does "reckless" mean? Where a person can be guilty of a crime if he has acted with a state of mind called "recklessness", we need to know whether recklessness is to be given a subjective or an objective meaning.

As all criminal lawyers know, the meaning of recklessness was a matter of huge debate in the latter part of the 20th century. This occurred first in relation to the crime of rape. It was clear that having sexual intercourse intending to do so knowing that the woman did not consent was sufficient for guilt, because "intention" is the usual mental requirement for liability. But what about intending to have sexual intercourse, knowing that there was a risk that the woman was not consenting, but hoping that she was? This state of mind, the deliberate taking of a known risk, is recklessness with a subjective meaning, if the taking of the risk was unreasonable. It is "subjective" because the man knew of the risk.

An illustration of the objective meaning of recklessness is where a man has sexual intercourse, thinking that the woman is consenting, but in circumstances where any reasonable person would have been aware of the risk that she was not. This is "objective" because it refers to what the reasonable person would have known, not what the man himself knew.

Although it was not done with great clarity, it seems that the subjective meaning of recklessness was applied by the House of Lords in R v Morgan [1976] AC 182, where the offence in question was rape. It was felt, however, that policy required an objective interpretation of recklessness for some offences, such as arson: R v Caldwell [1982] AC 341 (HL). This view was changed, and Caldwell was overruled, in R v G [2004] 1 AC 1034 (HL). Lord Bingham put it this way, in para 32:

"conviction of serious crime should depend on proof not simply that the defendant caused (by act or omission) an injurious result to another but that his state of mind when so acting was culpable… . The most obviously culpable state of mind is no doubt an intention to cause the injurious result, but knowing disregard of an appreciated and unacceptable risk of causing an injurious result or a deliberate closing of the mind to such risk would be readily accepted as culpable also. It is clearly blameworthy to take an obvious and significant risk of causing injury to another. But it is not clearly blameworthy to do something involving a risk of injury to another if … one genuinely does not perceive the risk. Such a person may fairly be accused of stupidity or lack of imagination, but neither of those failings should expose him to conviction of serious crime or the risk of punishment."

In New Zealand, the subjective meaning of recklessness is applied: R v Harney [1987] 2 NZLR 576 (CA), R v Tihi [1989] 2 NZLR 29 (CA).

This whole topic was aired recently in the High Court of Australia, considering New South Wales legislation in Banditt v R [2005] HCA 80 (15 December 2005). This was a case of rape, and the majority judges (Gummow, Hayne and Heydon JJ) held that the subjective meaning of recklessness applied: it was necessary, for the accused to be guilty, that he be proved to have been aware of the risk that the complainant was not consenting. Callinan J, although agreeing in the dismissal of the appeal, thought that the legal wrangling over the meaning of recklessness was so complex that the legislature must have intended simply that the ordinary meaning, as determined by the jury, should apply. This, with respect, is plainly wrong, for the ordinary meaning of recklessness can include (COD) "disregarding the consequences or danger, etc; lacking caution; rash." When one remembers that sexual intercourse is often indulged in rashly, it is plain that this approach would cast the net too far.

Monday, January 23, 2006

Illegal acquittals

Objections to the power proposed to be given to the prosecution, in some circumstances, to appeal against acquittals, are usually advanced on the grounds that the finality of an acquittal is a fundamental common law right, a matter of "due process".

The Privy Council has considered this in The State v Boyce (Trinidad and Tobago) [2006] UKPC 1 (17 January 2006). The judgment notes different usages of the term "due process": there is a broad sense, (para 13):

"In this sense, the concept of due process incorporates observance of all the mandatory requirements of criminal procedure, whatever they may be."

And there is a narrow, constitutional, sense (para 14):

"… those fundamental principles which are necessary for a fair system of justice."

The absolute prohibition at common law of a challenge to the finality of an acquittal was not, the Judicial Committee held, part of the narrower meaning of due process (para 15). This conclusion required classifying the finality of an acquittal as a "broad principle":

"…the broad principle that a person who has been finally convicted or acquitted in proceedings which have run their course should not be liable to be tried again for the same offence is a fundamental principle of fairness. It is recognised as such in many constitutions (see, for example, section 20(8) of the Constitution of Jamaica) and in international human rights instruments: see for example article 14.7 of the UN International Covenant on Civil and Political Rights and article 4 of Protocol No 7 to the European Convention for the Protection of Human Rights and Fundamental Freedoms. But [their Lordships] do not think that the principle is entirely without exceptions (see, for example, article 4.2 of Protocol No 7) and they certainly do not think that it is infringed by the prosecution having the right to appeal against an acquittal. The possibility of such an appeal is accommodated in the qualification of the principle ("save upon the order of a superior court made in the course of appeal proceedings relating to the conviction or acquittal") in all the Caribbean constitutions to which their Lordships were referred (Jamaica, Barbados, The Bahamas, Grenada, Dominica, Saint Lucia, Saint Vincent and the Grenadines, Guyana, Antigua and Barbuda, Belize, Anguilla, St Christopher and Nevis, Turks and Caicos Islands, Monserrat) as well as the international instruments which their Lordships have mentioned."

Thus, the proceedings may not have "run their course" until any right that the prosecution may have to appeal against an acquittal has been exercised. Once it has, the broad principle of finality would apply.

It is sensible, of course, to permit an appellate court to correct errors that have occurred in the application of the law during a trial, whichever side they may have favoured. This point was made in para 16:

"…There is nothing particularly unfair or unjust about a statutory rule which enables an appellate court to correct an error of law by which an accused person was wrongly discharged or acquitted and order that the question of his guilt or innocence be properly determined according to law. Such a rule exists in many countries. It is true that in Davern v Messel (1984) 155 CLR 21 the High Court of Australia said that the common law rule was so ancient and well settled that a statute giving a right of appeal in general terms against the decisions of a court would not be construed as allowing the prosecution to appeal against a decision in favour of an accused person. Express language was needed. But the court did not suggest that the absence of a right of appeal formed part of fundamental due process and in the present case the language of section 65E [of the Supreme Court of Judicature Act] is clear enough."

Other interesting points made in this case are:
  • Proceedings on indictment start with the filing of an indictment; where an information is laid indictably, that is just a preliminary to indictable proceedings but does not form part of them (para 22).
  • An "error of law" can arise wherever the judge had made a wrong determination, regardless of whether it concerns pure law, or a mixture of fact and law. It thus can include rulings on admissibility of evidence and on whether there is a case to answer (para 24).
  • In deciding whether a witness qualifies as an expert, it is wrong (and this was one of the errors made by the judge in this case) to focus only on the witness’s paper qualifications. The judge was also wrong to call another witness solely to comment on whether another witness should qualify as an expert (para 25, 26).
  • Where an error of law has occurred, and an acquittal has resulted, the question whether a new trial should be ordered can depend on whether a fair trial is still possible. Here, the error of law was corrected (ie the provision creating the right to challenge an acquittal was held to be constitutional), but the appeal was dismissed because it could not be said that a new trial would be fair, 9 years having elapsed since the events in question (para 26, 27).

Friday, January 20, 2006

Three major themes

Three topics of general interest were touched upon by the Privy Council in Grant v R (Jamaica) [2006] UKPC 2 (17 January 2006): (1) the interrelationship between the right to a fair trial, and the balancing of constituent rights; (2) the discretion to exclude evidence where its prejudicial effect exceeds its probative value, and the relationship of this discretion to the right of the accused to a fair trial; and (3) whether the proviso can be applied to dismiss an appeal when potentially significant evidence was never before the jury (including, on the facts of this case, an illustration of what is "potentially significant" evidence).

Right to a fair trial, and balancing

The Privy Council referred to laws similar to those in effect in Jamaica, including decisions of the European Court of Human Rights. The established model is that the accused’s right to a fair trial is absolute, but rights that are subsidiary to that are subject to balancing against each other. See Mathias, "The Accused’s Right to a Fair Trial: Absolute or Limitable?" [2005] New Zealand Law Review 217.

At para 17 of Grant, the Privy Council put it this way:

"The Strasbourg court has time and again insisted that the admissibility of evidence is governed by national law and that its sole concern is to assess the overall fairness of the criminal proceedings in question: see, for example, Kostovski v Netherlands (1989) 12 EHRR 434, para 39; Windisch v Austria (1990) 13 EHRR 281, para 25; Lüdi v Switzerland (1992) 15 EHRR 173, para 43; Saidi v France (1993) 17 EHRR 251, para 43; Doorson v Netherlands (1996) 22 EHRR 330, para 67; PS v Germany (2001) 36 EHRR 1139, para 19. The specific rights set out in article 6(3) of the European Convention (and thus, by analogy, section 20(6) of the [Jamaican] Constitution [see below]) are "specific aspects of the right to a fair trial" (Kostovski v Netherlands, above, para 39) or "particular aspects of the right to a fair trial" (Doorson v Netherlands, above, para 66), and the right to a fair trial can never be compromised in any circumstances. But the constituent rights in article 6 and section 20(6) are not themselves absolute: Brown v Stott (Procurator Fiscal, Dunfermline) [2003] 1 AC 681, 704. The Strasbourg court has been astute to avoid treating the specific rights set out in article 6 as laying down rules from which no derogation or deviation is possible in any circumstances. What matters is the fairness of the proceedings as a whole."

Section 20(6) provides:

"(6) Every person who is charged with a criminal offence—
(a) shall be informed as soon as reasonably practicable, in a language which he understands, of the nature of the offence charged;
(b) shall be given adequate time and facilities for the preparation of his defence;
(c) shall be permitted to defend himself in person or by a legal representative of his own choice;
(d) shall be afforded facilities to examine in person or by his legal representative the witnesses called by the prosecution before any court and to obtain the attendance of witnesses, subject to the payment of their reasonable expenses, and carry out the examination of such witnesses to testify on his behalf before the court on the same conditions as those applying to witnesses called by the prosecution; and
(e) shall be permitted to have without payment the assistance of an interpreter if he cannot understand the English language."


These, then, are capable of being compromised in the balancing exercise, which the Privy Council described in Grant (also in para 17) as follows:

"Just as section 13 of the Constitution recognises that individual rights cannot be enjoyed without regard to the rights of others, so the Strasbourg court has recognised the need for a fair balance between the general interest of the community and the personal rights of the individual, and has described the search for that balance as inherent in the whole Convention: Sporrong and Lönnroth v Sweden (1982) 5 EHRR 35, para 69; Sheffield and Horsham v United Kingdom (1998) 27 EHRR 163, para 52; Brown v Stott, above, p 704. Thus the rights of the individual must be safeguarded, but the interests of the community and the victims of crime must also be respected. An example, not based on the present facts, illustrates the point. In Jamaica, as in England and Wales, as already noted, the statement of a witness may be adduced in evidence if he is shown to have absented himself through fear of the consequences to him if he gives evidence. In the case of a prosecution witness, such fear is likely to have been induced by or on behalf of a defendant wishing to prevent adverse evidence being given. As observed by Potter LJ in R v M(KJ) [2003] EWCA Crim 357, [2003] 2 Cr App R 322, para 59, echoed by Waller LJ in R v Sellick [2005] EWCA Crim 651, [2005] 1 WLR 3257, paras 36, 52-53, it would be intolerable if a defendant shown to have acted in such a way could rely on his human rights under article 6 (or section 20) to prevent the admission of hearsay evidence. Where a witness is unavailable to give evidence in person because he is dead, or too ill to attend, or abroad, or cannot be traced, the argument for admitting hearsay evidence is less irresistible, but there may still be a compelling argument for admitting it, provided always that its admission does not place the defendant at an unfair disadvantage."

Obviously, where the need to ensure fairness to the accused is paramount, as clear a definition, as is possible, of fairness needs to be established. It is on this central point that further elaboration is needed. I have suggested, in the article referred to above, that a fair trial meets two essentials: (1) there is an opportunity for the jury properly to apply the law, and (2) that application of the law is to the facts determined without bias.

The concern in Grant was whether the admission of the hearsay evidence would create bias against the accused by undermining his ability to challenge that evidence; whether, in other words, admission of the hearsay statement created an inequality between the prosecution and the defence.

The Board carefully set out the safeguards that the law provided for accused persons faced with such hearsay evidence (para 21). These included strict criteria for establishing the unavailability of the witness as a precondition for admission of his written statement, enhanced rights of the defence to call evidence to challenge the truthfulness of the evidence, and a need for the judge to give special directions to the jury cautioning them on what weight to give the hearsay evidence. Also among these safeguards is the next topic to be considered here: the discretion to exclude the evidence where its prejudicial effect exceeds its probative value.

Probative value and prejudicial effect

I have recently written on this topic: "Probative value, illegitimate prejudice and the accused’s right to a fair trial" (2005) 29 Crim LJ 8. In essence, my point is that the discretion to reject evidence on this ground, which involves a so-called balancing of probative value against prejudicial effect, is flawed because the balancing, or weighing, notion suggests that only a high level of unfairness to the accused will outweigh highly probative evidence. That in turn suggests that trial fairness doesn’t matter so much if the accused is obviously guilty. The real question, I suggest, is whether the evidence has sufficient probative value to be admissible bearing in mind the law’s reluctance to allow hearsay evidence; if it does have sufficient probative value, then it is admissible, subject to the trial still being fair for the accused.

The Privy Council has taken this approach in Grant, para 21:

"Section 31L acknowledges the discretion of the court to exclude evidence if it judges that the prejudicial effect of the evidence outweighs its probative value. In R v Sang [1980] AC 402, some members of the House of Lords (notably Lord Diplock at pp 434, 437 and Viscount Dilhorne (pp 441-442)) interpreted this discretion narrowly, and in Scott v The Queen [1989] AC 1242, 1256-1257, the Board appears to have accepted that reading. It is not, however, clear that the majority in R v Sang favoured a similarly narrow interpretation (see Lord Salmon at pp 444-445, Lord Fraser of Tullybelton at p 449 and Lord Scarman at pp 453, 454, 457). In any event, it is, in the opinion of the Board, clear that the judge presiding at a criminal trial has an overriding discretion to exclude evidence which is judged to be unfair to the defendant in the sense that it will put him at an unfair disadvantage or deprive him unfairly of the ability to defend himself. Such a discretion has been recognised by the Court of Appeal in R v Donald White (1975) 24 WIR 305, 309, and R v Michael Barrett, above. It has been recognised by the Board in Scott v The Queen, above, pp 1258-1259 and Henriques v The Queen [1991] 1 WLR 242, 247: both these appeals concerned the admission of depositions, but the need for a judicial discretion to exclude is even greater when the evidence in question has never been given on oath at all. In England and Wales, the discretion has been given statutory force: see section 25(1) of the Criminal Justice Act 1988; R v Lockley [1995] 2 Cr App R 554, 559-560; R v Gokal [1997] 2 Cr App R 266, 273; R v Arnold [2004] EWCA Crim 1293, para 30. Conscientiously exercised, this discretion affords the defendant an important safeguard."

Thus, whether one takes a broad or a narrow approach to the so-called balancing, or weighing, of probative value against prejudicial effect, doesn’t matter, because the overriding criterion is whether admission of the evidence would make the trial unfair for the accused.

In para 26 of Grant the Board analysed the circumstances of the trial and gave reasons for their conclusion that the admission of only one of two hearsay statements caused unfairness to the accused. Grant was accused of murder. He had been approached by the victim in a threatening manner while urinating by some bushes at night. He claimed the person had a gun. He took out his own gun and fired at the victim several times. The victim turned and ran. Grant chased him and fired several shots again. The evidence diverged crucially at this point: the hearsay statement that was admitted in evidence was of a witness, Bryant, who said he saw the victim lying face down on the ground when he was approached and shot several times. The victim died from wounds he had received in the back. Grant denied approaching the victim and shooting him like that, saying in evidence that he fired after him because he thought he was still at risk of attack: he relied on self defence. A hearsay statement by another person, Kinglock, was excluded from evidence. It would have confirmed the first part of the episode: someone approaching Grant while he was urinating, then shots being heard and the person running and throwing something (this, if Grant’s evidence had credibility, would have been a gun) away.

In summary, the trial was unfair because (para 26):

"It is, in the Board's opinion, plain that fairness required the admission of Kinglock's statement. If admitted, it might not have been understood to exonerate the appellant. The proliferation of shots to the back of the deceased was a formidable problem for him to overcome. The jury might have convicted anyway, and been entitled to do so. But the jury should have known how, according to Kinglock, in large part corroborating the appellant, the incident began. The appellant was entitled to have his case assessed, and his own evidence evaluated, in the light of all the available evidence. … [the judge] could have invited prosecuting counsel to adduce Kinglock's statement in evidence. Had that invitation, improbably, been declined, the judge could, on grounds of fairness, have declined to admit Bryant's statement unless Kinglock's statement were also admitted or could, in the last resort, have introduced the statement of Kinglock herself (R v Oliva [1965] 1 WLR 1028, 1035-1036). … The Crown having given the defence notice of its intention to adduce the statement of Kinglock as well as the statement of Bryant at the trial, defence counsel appears to have been taken by surprise when on the second day of the trial prosecuting counsel applied to adduce the statement of Bryant, but not that of Kinglock. … It was, however, the responsibility of prosecuting counsel and the trial judge to ensure that the proceedings were fair, and they failed to do so. This failure was compounded by an inadequate direction on Bryant's evidence. The jury were given no encouragement to scrutinise it with particular care, and were not alerted to apparent discrepancies between it and the evidence of Constable Wynter (or, of course, the statement of Kinglock)."

These points illustrate the two aspects of trial fairness: the need to avoid bias in the opportunity the jury has to determine the facts, and the need to ensure the jury has an opportunity correctly to apply the law.

The proviso

Finally, Grant illustrates consideration of whether to apply the proviso. That is, whether to dismiss the appeal against conviction notwithstanding that errors at trial had occurred. The statutory criterion is whether errors had amounted to a "substantial miscarriage of justice". The meaning of this phrase is critical, and it is still being worked out after over a century of consideration by courts in the common law countries.

As is usual, the proviso did not receive detailed treatment by the Board, which simply concluded, para 27:

"It would not be appropriate to apply the proviso in a case where potentially significant evidence was never before the jury."

This means that, in the context of Grant, the evidence of Kinglock was "potentially significant", even though "it might not have been understood to exonerate the appellant" (para 26, above). It was corroboration of Grant’s version of how the incident started that was "potentially significant": "The appellant was entitled to have his case assessed, and his own evidence evaluated, in the light of all the available evidence" (ibid). This was so, even though the partial defence of provocation was not relied on. The second part of the incident, not covered in Kinglock’s evidence, would be most unlikely to be the basis for a successful defence of self defence, as the Board recognised: "The proliferation of shots to the back of the deceased was a formidable problem for him to overcome. The jury might have convicted anyway, and been entitled to do so" (ibid).

Plainly, the rationale for the application of the proviso, not elaborated by the Board, is that the trial was not fair. No matter how strong the case, an unfair trial cannot sustain a conviction.