Elucidating Weiss? Grappling with “substantial miscarriage of justice”

Another look at Weiss (noted here 16 January 2006, 9 February 2006, 25 June 2007, and various dates – search this site for "Weiss" – to 9 July 2009) and the proviso occurred in Baiada Poultry Ltd v R [2012] HCA 14 (30 March 2012).

In Baiada Poultry Ltd the jury had not been instructed on a requirement for commission of the offence, so the conclusion that there had been a substantial miscarriage of justice was unavoidable and the proviso should not have been applied by the lower appellate majority. A retrial was ordered.

Some of us are about to enjoy an appeal criterion that no longer involves the proviso: see s 232 of the Criminal Procedure Act 2011[NZ]. Nevertheless, some dicta in Baiada Poultry Ltd are of interest. The High Court of Australia continued its "back to the words of the legislation" approach to the requirement of a substantial miscarriage of justice: the phrase should not be replaced by judicially created categories of fundamental defects [23], [31]. Well, I don't think that refusal to say what a phrase means is particularly helpful: it invites re-invention of the wheel with each appeal. The reformed New Zealand law will omit the word "substantial" and will include a definition of "miscarriage of justice". Further elaboration of "unfair trial" will be needed, because counsel always have to say why they are submitting something was unfair.

Another point in Baiada Poultry Ltd is that it is unhelpful to describe the appellate court as exercising a discretion when it considers whether to apply the proviso. We New Zealanders will note that the new provision is mandatory in its terms: "must allow … appeal … if satisfied that …". It is unlikely that if an appeal turns on the assessment of fairness this will be considered a discretionary matter: it is plainly one of making a finding as a matter of law.

Again, in Baiada Poultry Ltd, where the fact that the jury convicted the defendant is being considered on appeal, regard must be had to the issues it was left to decide. If, as here, an issue had not been left to the jury, the verdict is irrelevant [28] and per Heydon J at [67].

And finally, a major point in Weiss was repeated: if the appellate court is satisfied that on the evidence properly admitted the defendant was guilty, that is a necessary, but not a sufficient, condition for applying the proviso. That is to say, even if he was obviously guilty, the defendant's appeal cannot be dismissed if his trial had been unfair. The difficulty here is that fairness encompasses errors that affected the result of the trial, that is, errors that affected the jury, and it is immaterial that the appellate court thinks the errors should not have affected the result. The High Court may be linking the proviso jurisprudence to a narrow definition of trial fairness.

I suggest the trial in Baiada Poultry Ltd was unfair because the law was not properly applied to the facts. In any event, the appeal was allowed because the Court of Appeal majority had, in evaluating the evidence, done so without sufficient information to support its conclusions [37]-[39] or had drawn conclusions that did not necessarily follow from the evidence (per Heydon J at [70]).

Propensity evidence: relevance and probative value

The topics du jour in BBH v R [2012] HCA 9 (28 March 2012) were the requirements of relevance and probative value in relation to the admissibility of propensity evidence.

Everyone agrees that the first requirement for evidence to be admissible is relevance. The evidence must have a tendency to prove a matter in issue in the case. This does not mean it must be looked at in isolation, but assessment of its tendency to prove a matter in issue, its relevance, is made in the context of the other evidence in the case.

All well and good. But judges can differ over whether evidence is relevant. In BBH French CJ, Gummow and Hayne JJ held the contested evidence was not relevant. Their narrow view of it is in contrast to that taken by the other members of the court. French CJ was particularly concerned to avoid circular reasoning [58]: it would be wrong to use other evidence of the offence charged to interpret the tendency of the contested evidence. French CJ described the contested evidence as a "snapshot" of an incident that may or may not have been of the kind that would have made it evidence of similar facts.

But Heydon J [102] rejected the snapshot approach and looked more carefully at the context of the contested evidence, and concluded that it was capable of having the required similarity. So did Crennan and Kiefel JJ jointly [152], [159]. Bell J also noted circumstances in relation to the contested evidence that supported similarity [198].

Once relevance is established, propensity evidence must pass another hurdle. It must reach a required level of probative value. The applicable standard in this case was laid down in Pfennig v R [1995] HCA 7, (1995) 182 CLR 461. This standard is by no means accepted widely, as Crennan and Kiefel JJ note at [134]. In any event, and broadly speaking, it involves pretending that all the contested evidence is accepted at its highest from the prosecution point of view and also pretending that there remains a reasonable doubt about guilt on the other evidence for the prosecution. The test then is, is the contested evidence capable of removing a reasonable doubt about the defendant's guilt? If so, it is admissible as propensity evidence, otherwise not.

French CJ said that even if the contested evidence was relevant so that Pfennig had to be applied, the result would be that it was inadmissible [59], although he did not elaborate. Hayne J, with Gummow J agreeing, came to the same conclusion [81]. There remained a rational explanation consistent with innocence. This seems to mean that even considered with all the other prosecution evidence the contested evidence had so little probative value that it would leave a reasonable doubt about the defendant's guilt.

The other judges concluded that under Pfennig the contested evidence was admissible. It was not circular to look at the contested evidence in the light cast by the other prosecution evidence. Crennan and Kiefel JJ at [159] with Bell J agreeing at [199] stressed the independence of the witnesses to the two events (that is, the witness giving the propensity evidence and the complainant giving evidence of various offences), the absence of collusion, and the unlikelihood of the coincidence of both witnesses giving evidence of similar incidents. The difference from circularity in the relevance assessment is that here, when we get to the Pfennig stage, similarity has been established. Probative value reflects how the contested evidence bears upon the evidence supporting the present allegations.

The correctness of the Pfennig test was not in issue in this appeal: Crennan and Kiefel JJ at [134]. Nor was the judge's direction to the jury that the propensity evidence could only be taken into account if it was proved beyond reasonable doubt. This, as Crennan and Kiefel JJ said [165]-[168], was a consequence of the chain of reasoning analysis in Shepherd v R [1990] HCA 56, (1990) 170 CLR 573. It is not a universal requirement that propensity evidence be proved beyond reasonable doubt; instead it can be treated, once admissible, as just another sort of circumstantial evidence and given some probative value even if doubts about its reliability may persist, as long as it is more probably true than not true. See for example R v Holtz [2003] 1 NZLR 667, (2002) 20 CRNZ 14 (CA). Instead of Pfennig, in New Zealand we use the statutory requirement that the probative value of the propensity evidence outweighs its unfairly prejudicial effect, with specific criteria to be considered: Evidence Act 2006, s 43. These include the extent of the similarity, the number of people making the accusations, whether admission of the evidence would unfairly dispose the fact-finder against the defendant, and whether the fact-finder would give the evidence disproportionate weight.

Kettling and tracking

A lecture on Thursday evening by Professor Andrew Ashworth of Oxford reminded me to mention Austin v United Kingdom [2012] ECHR 459 (15 March 2012). This concerns the procedure of "kettling" people by detaining (oops – begging the question there) them when civil unrest breaks out and the police need to maintain order. In London thousands of people where kettled for about 6 hours without food, drink, toilet facilities, and with only room to stand or sit on the pavement. Was this a breach of Article 5 of the European Convention on Human Rights?

Article 5 gives everyone the right to liberty and security of the person, and only permits deprivation of liberty in six specified circumstances (see para 38 of the majority judgment). The government argued that either kettling was not a deprivation of liberty, or, if it was, it was justified by Article 5 as a detention in furtherance of the police's obligation to preserve the peace, or as a detention reasonably necessary to prevent offending or flight of offenders.

The majority interpreted the Convention as a "living instrument" [53] and held that Article 5 cannot be interpreted so as to make it impracticable for the police to fulfil their duties of maintaining order and protecting the public [56]. Further (or alternatively) the context of the kettling must be considered because unavoidable restrictions on movement arising from circumstances beyond the control of the authorities and which are necessary to avert a real risk of serious injury or damage and which are kept to the minimum required for that purpose are not "deprivations of liberty" within the meaning of that article [59], [60].

On the facts here the majority held that the kettling was reasonable and did not amount to a deprivation of liberty within the meaning of Article 5.

The 3 minority judges considered that there had been a deprivation of liberty. They questioned the proposition that if a restrictive measure was necessary for a legitimate public-interest purpose it did not amount to a deprivation of liberty. The reason for the deprivation of liberty should not be relevant to whether it was a deprivation of liberty. The reason is only relevant to whether the deprivation of liberty was justified under Article 5. Also, Article 5 does not warrant a distinction between deprivations of liberty arising from public order considerations and other deprivations of liberty, as the Grand Chamber had held in A v United Kingdom [2009] ECHR 301 (19 February 2009 and noted here 22 February 2009). See also Jendrowiak v Germany ECHR 14 April 2011. The minority also pointed out that the majority were ignoring another decision of the court: Gillan and Quinton v United Kingdom [2010] ECHR 28 (12 January 2010).

My assessment: Basically, the court's jurisprudence follows pragmatism rather than precedent. There is nothing wrong with pragmatic balancing of the values that underlie competing rights, but a lot of guess work is involved. No information was considered concerning the occurrence in groups of people of the kettled size of lost opportunities, risks to health and financial costs, nor was there any consideration of how those impacts should be measured for comparison with the advantages of reducing the costs of rioting. The majority judgment leaves us with the impression that a fuzzy sort of judicial comfortableness was the criterion, with emphasis on the police having to act in the longer-term interests of everyone. That begs the question of there having been no alternative police response.

Another case mentioned by the Vinerian Professor is United States v Jones
2011 USSC No 10-1259 (23 January 2012). This decided that the covert placing of a GPS tracking device on a car was a trespass and information subsequently gathered about the suspects' movements was unlawful search in breach of the Fourth Amendment. Thus the trespass test continues to be relevant notwithstanding the more recent common law development of the reasonable expectation of privacy test. This case did not require consideration of whether the search was reasonable because that argument was not raised in the courts below. A useful commentary by Atli Stannard on the comparative law can be found at http://www.acclawyers.org/?p=2980 .

When to rescue the rescuer

Procedural fairness may have to be sacrificed to protect a litigant's rights. An illustration is W (Algeria) & Anor v Secretary of State for the Home Department [2012] UKSC 8 (7 March 2012).

The moral issue here was, in the broadest possible terms, should a rescuer be protected if the rescue might be without merit? More specifically, when should a tribunal be able to extend the protection of confidentiality to a witness who claims to have information that would assist a litigant?

The appeal concerned extradition proceedings and whether the appellants, if extradited, would be at risk of ill-treatment in breach of their rights under article 3 of the European Convention on Human Rights.

A witness who claimed to be able to substantiate the risk of ill-treatment refused to give evidence unless the Special Immigration Appeals Commission (SIAC) made an order of confidentiality that would prevent the Secretary of State disclosing to anyone his evidence and his identity. Could SIAC make such an order?

The order would prevent the Secretary investigating the credibility and reliability of the witness, or at least would limit the Secretary's ability to carry out those inquiries. Further, the application for the order would have to be made ex parte, and the Secretary could not therefore oppose it. These were the limits on procedural fairness that would arise if the order was made.

Arguing against the making of such orders, the Secretary submitted that policy favoured the ability to pass on information to governments where threats to security would otherwise not be met. For example, the witness may be a terrorist planning an atrocity, and this may be evident from the nature of the evidence he gives in support of the risk of ill-treatment. Confidentiality might have severe diplomatic consequences if the government of a targeted country discovered that the Secretary had not passed on information that may have saved lives.

So, which is to dominate? The interests of the litigant facing extradition and a risk of ill-treatment, or the interests of those vulnerable to terrorism?

The Supreme Court was unanimous. Two judgments were delivered, each agreeing with the other.

Lord Brown found an answer to the Secretary's potential diplomatic embarrassment in the defence of obedience to a court order [14]. It was necessary to maximise SIAC's chances of arriving at the correct decision [18]. However the power to make such confidentiality orders should be "most sparingly used" [19]. If necessary it should be open to the Secretary to try to persuade SIAC to seek a sufficient waiver of confidentiality to address national security concerns, and if that waiver was not forthcoming then SIAC could view the evidence with scepticism or exclude it [21]. And in any event, in deciding whether to make the order SIAC should require a detailed statement of the proposed evidence, why the witness fears reprisals, and how the person challenging extradition learnt of the witness's proposed evidence and what steps were taken to get the witness to give evidence in the normal way subject to the usual safeguards of anonymity orders and private hearings [20].

Lord Dyson said that a confidentiality order should be made if SIAC is satisfied that the witness can give evidence which appears to be capable of belief and which could be decisive or at least highly material on the issue of safety of return, and if SIAC has no reason to doubt that the witness genuinely and reasonably fears reprisals if his identity and evidence were to be disclosed [34].

The issue arose here in connection with article 3 rights (not to be subjected to torture or to inhuman or degrading treatment or punishment), but Lord Dyson added that it would also arise if the case raised a question under article 2 (right to life) [38]. But if the person resisting extradition relied on an alleged risk of breach of some other article of the Convention, "the balance will almost certainly be struck the other way." [38] He gave as an example breach of article 8 rights (respect for private and family life).

Perhaps too a risk in the requesting country of breach of article 6 rights (fair trial, presumption of innocence) might be sufficient to justify a confidentiality order, but that was not addressed in these appeals.

Lord Brown made it clear that his reasoning was based on the balancing of interests and not on the idea that the playing field should be level or that there should be an equality of arms [22] in the extradition hearing. That is to say, the answer did not emerge from fair hearing considerations. Plainly the procedure is unfair to the Secretary who in any event is obliged to act in the public interest and also to disclose information that could assist the person resisting deportation [22].

A right to a fair hearing belongs to the defendant, not to the prosecutor, so indeed it is inappropriate to speak of level playing fields and equality of arms. Inequalities are assumed, and efforts are made to minimise them, but sometimes they serve the public interest and can be tolerated as long as the hearing remains fair to the defendant. See, for example, R v H [2004] UKHL 3.

So, although this issue arose in the context of extradition, it is potentially relevant to criminal proceedings.

Well, we might wonder if international relations are necessarily conducted according to law. Anyone who was brought up on the novels of Ian Fleming and John Le Carre, or who more recently has enjoyed watching "Spooks", will have doubts. It is easy to imagine in the distant and troubled future that some new Secretary might yell at a subordinate (or at a Chief Justice) "Do you think I am going to let thousands of people die just because of an effing confidentiality order?" As far as the law is concerned, the diplomatic argument should not have been raised and an answer to it not given. Whether the policy of sharing information on terrorist threats outweighs the rights of an individual litigant depends on the content of the evidence in respect of which confidentiality is sought in a specific case. General propositions of the kind advanced by the Secretary here are, without that link to the particular evidence, irrelevant. Also, the Court makes an inappropriate link between refusal of waiver and credibility. Refusal should enhance credibility, not diminish it. A person who refuses to waive confidentiality probably has good reasons for doing so and those reasons should support an inference that he knows what he is talking about. The real issues in this sort of case will be the need for the witness to have confidentiality and the risk of the defendant being ill-treated after extradition, and it is unfortunate that the judges did not reach for their copies of Dworkin's "Justice for Hedgehogs".

Honestly paying for illegal gains

For a 3-2 split on statutory interpretation, involving differences over legislative intent and the scope for application of the principle that property rights are not taken away except where the intention to do so is clear, see Re Peacock [2012] UKSC 5.

The subject of the dispute is not of much interest unless your legislation on recovery of the proceeds of crime leaves it unclear whether property acquired honestly after a recovery order is made can be included within the order. In this case the majority held that the legislature did intend that such honestly and after-acquired property should be recoverable.

The problem arose in Peacock because a confiscation order was made on a calculation which included a reduction to recognise the defendant's means to pay. Subsequently his means legitimately increased. In New Zealand under the Criminal Proceeds (Recovery) Act 2009 a profit forfeiture order does not take into account the defendant's means to pay, and any amount over that actually realised remains owing as a debt to and recoverable by the Crown.

Need witness competency be an issue separate from fairness?

The Supreme Court of Canada has split on the requirements for witness competency: R v D.A.I. 2012 SCC 5 (10 February 2012). The issue in Canada is one of statutory interpretation, and here revolved around whether an adult but mentally impaired witness needed to show an appreciation of the significance of a promise to tell the truth. The majority held that such appreciation did not need to be demonstrated because it could require some difficult abstract concepts. The policy of allowing impaired victims access to justice was of great importance.

Plainly, if the witness's response to questioning while giving evidence was such as to deprive the right of the defence to confront the witness, there would be trial fairness issues. However at the threshold stage the issue was whether the witness showed understanding of the oath or affirmation, and was able to communicate the evidence.

We in New Zealand do not have a competency requirement for witnesses. The idea is that "No person, whether on the grounds of age, intellectual disability, or mental disorder, or on any other ground, may be disbarred from giving evidence on the ground of incompetence. ... In the case of witnesses whose testimony is unhelpful – because of incoherence, for example – the judge may still exclude that evidence under the general exclusionary provisions in s 8": NZLC R55 "Evidence" – Vol 2 at C294.

Book review: “Trial by Ambush” by Joe Karam

No objective reader of Joe Karam's "Trial by Ambush" can possibly come to any conclusion other than that Robin Bain committed the murders of his family. It is equally obvious that David Bain must receive compensation for his years of imprisonment which were a direct result of improprieties in the investigation and failures by the authorities including the judiciary to provide a timely remedy.

It is the failings of the judiciary that are of most concern to readers of this site. Three appeal judges sat on what the Privy Council called the Third Court of Appeal in this case. Their single judgment contained, according to submissions to the Privy Council prepared by Karam and reproduced as Appendix B to his book, an astonishingly large list of errors of fact.

One judge might well make the occasional slip in summarising the evidence in a case, but how can so many errors pass by three judges? This calls into question the soundness of a recent proposal by our Law Commission that factual issues should be decided by a small panel of judges. The Commission likens this to the practice in Belgium, but China would also be a relevant point of reference.

Work done by committees tends to be distributed among members, whereas in trials it is brought home to jurors that they are individually responsible for their decision. The Court of Appeal is over-worked and under-resourced, and its judges – all of whom are of high quality by international standards - are encouraged to bring in unanimous decisions in criminal cases. That is hardly an environment that will promote accuracy.

But the errors in the Bain case began much earlier, according to Karam's book. The police decided too quickly to charge David. They then sought evidence against him rather than being open to the alternative that Robin was the murderer. They failed to preserve, record or have analysed evidence that might have supported David's innocence, and at an astonishingly early stage after the first trial they destroyed evidence. Evidence that was disclosed to the defence before both trials was dumped on the defence in huge volume and in a disordered state, without indication of what was significant.

This is the second aspect that is of interest here: how can the prosecution be required to exercise its disclosure obligations fairly? In the adversarial system, where the trial is a contest with a winner and a loser, procedural fairness can be sacrificed for the sake of egotistical stratagems. Trials and appeals become contests between counsel, and between counsel and the bench, to see who is cleverest.

If you think Robin's full bladder eliminated him as a suspect, you won't think so after reading this book. Nor will you think that David turned on the computer. Nor that David put the washing on before going on his paper run. Nor that all victims except Robin were killed before David left on his paper run. You will be convinced that Robin left bloodstained footprints in the carpet as he shot the victims – David's feet were too big to have made them. Blood on Robin's hands (not available for analysis, but visible in tardily disclosed photographs) was consistent with coming partly from the bloody gloves he wore, and partly from splash-back from a victim as he held the rifle. Blood on Robin's trousers and on his shoes was consistent with his position as he pulled the trigger committing suicide, as was blood on the curtain by the computer. Robin was the psychological mess, not David. Robin fitted the profile of men who kill their families, and David didn't. The gurgling heard by David coming from one victim was of the kind that can occur after death, and the evidence of that possibility was stronger at the second trial than it had been at the first.

According to this book, there is no reliable evidence that David was the murderer. At one stage I thought that Karam's account of a green towel containing traces of Robin's blood and found in the laundry left open the possibility that David put it there after killing Robin. But Robin had an injury to his hand that would have bled, and that was probably sustained in the course of a fight with his younger son who had not been killed by the first shot Robin fired at him. Robin probably wiped blood from that wound on to the towel and left it in the laundry with his other blood soaked clothes, for David to put in the wash after his paper route.

It is so hugely unlikely that David was the murderer that, as Karam says, anyone suggesting the contrary had better put up compelling evidence. There is none. It would have to be as incontrovertible as a freely given confession by David, or a reliable eyewitness to the killings, or a video recording of them. If evidence of that kind had been obtained the case would not be one of inferences, or probabilities, but one of certainties. As it is, the probabilities are so enormously in favour of David's innocence that they amount to a certainty.

The community must be thankful for people like Joe Karam. We all are entitled to know that if the State makes an error and wrongly punishes us, we will be properly compensated. Unfortunately it is not always the State's own officials who can be relied on to provide that assurance.

Need the punishment fit the crime?

For those of us who do not need to worry ourselves over the intricacies of Australian constitutional law, Bui v DPP (Cth) [2012] HCA 1 (9 February 2012) makes a simple point.

It is that when on a successful prosecution appeal against sentence the appellate court imposes a more lenient sentence than it considers the sentencing court could have properly imposed, it is reflecting a sentiment behind the principle against double jeopardy.

"13 It has been explained that what is referred to as the rule against double jeopardy is a manifestation of the principles expressed in the maxim nemo debet bis vexari pro una et eadem causa (a person shall not be twice vexed for one and the same cause), which is the foundation of the pleas of autrefois acquit and autrefois convict ... .  The underlying idea is that the State should not be allowed to make repeated attempts to convict an individual thereby subjecting him to embarrassment, expense and ordeal and compelling him to live in a continuing state of anxiety and insecurity as well as enhancing the possibility that even though innocent he may be found guilty. The rule is properly understood as a value which underpins the criminal law." [case citations and internal quote marks omitted]

There is an interesting difference between the Criminal Procedure Act 2000 (Vic), ss 289(2), 290(3), and the Criminal Procedure Act 2011 (NZ), ss 251, 257. The former specifically directs the appellate court not to take into account any element of double jeopardy involved in the appellant being sentenced again, whereas the latter makes no mention of this.

I suppose that in assessing the results of the unspoken competition between law reform bodies to come up with the perfect criminal procedure laws, Victoria should be penalised on this point. At least that is so unless you think that, willy nilly, the offender should get the sentence he deserves, later if not sooner.

Kahneman’s Thinking, Fast and Slow

Everyone is reading Daniel Kahneman's "Thinking, Fast and Slow" (2011). A passage in a recent New Zealand Court of Appeal decision, the details of which are currently suppressed, raises questions about the right way to think about propensity evidence.

The case citation is [2011] NZCA 645 and the date of the decision is 14 December 2011. I will call it X v R. I will also adapt the quotation from para [34] of the judgment to comply with the order suppressing identifying particulars of the appellant:

"... it is an unlikely coincidence that Mr X, twice within a year, would be the hapless and innocent victim of being apprehended driving a car with [other people in it and also with evidence of criminal offending in it]. The evidence goes directly and cogently to the key issue: did Mr X know of the [items] found in the car he was driving on [the second occasion]?"

While the conclusion that the evidence had sufficient probative value to be admissible is intuitively correct, this form of reasoning entails several thinking errors of the kind that Kahneman discusses.

There is a tendency to draw strong conclusions from incomplete information (the "what-you-see-is-all-there-is" error). We are not told anything about the frequencies that matter in the above case: how frequently do people who are guilty of the present sort of offending have a previous recent incident of this sort of police apprehension, compared with how frequently do people who are innocent of offending of the present kind have a recent previous such apprehension?

There is a substitution error: we tend to answer difficult questions by answering a much simpler related question. Here it is easy to answer the question about the recent apprehension and to apply that answer to the more difficult question of guilt on the present occasion. This is closely related to another error.

Base-rate neglect is the error of neglecting statistical likelihoods in favour of accepting what could be causally possible. The other evidence in the case, relating directly to guilt on the present occasion, may significantly affect the strength of our tendency to see a causal connection between the first apprehension and the second.

Another error is the halo effect, or in the present context what might be called the devil's horns effect. Having learnt something bad about the defendant's behaviour on an earlier occasion, we are tempted to overemphasise this when we consider his present guilt.

Further, there is the narrative fallacy: we are tempted to accept what we can build into a story that makes sense, although the events may in reality be unconnected. The defendant may have been innocent on the earlier occasion through lack of knowledge of the presence of the things in the car. The coincidence may be real, but it does not suit the story we are tempted to build in which we cast the defendant as a recidivist.

This is similar to another error, the representativeness bias. Where only partial information is available we lean heavily on stereotypes.

For a review of Kahneman's book summarising these and other thinking errors, see the article in the New Zealand Listener, January 21-27, 2012, by David Hall.

In the above case, where the issue was the defendant's knowledge of the presence of the things in the car, it is easy to build a narrative in which the defendant, being ignorant on both occasions, was simply associating with people who were both his friends and offenders. That too would be a combination of thinking errors.

Courts too often make assumptions about likelihoods without inquiring into occurrences in the real world. The correct approach is Bayesian, as Kahneman – a leading psychologist and Nobel laureate – recognises. But that requires the effort of careful analytical thought rather than our preferred instinctive assessment of circumstances.

Admissibility of eyewitness identification evidence

Just as I was sneaking back to the Southern Hemisphere the Supreme Court of the United States delivered its decision in Perry v New Hampshire. I have previously noted the submissions that the Court received in this case on the vulnerability of eyewitness identification to error.

The reasoning is necessarily tied to the constitutional jurisprudence on due process, so I just summarise the effect of the decision in terms that may be of more general interest.

Sotomayor J dissented, but the majority held, in a decision delivered by Ginsberg J, that there is no need for a trial judge to have the power to exclude such evidence on grounds other than unfairness to the defence. Irregularities in the conduct of an identification can be the subject of trial remedies, such as confrontation and cross-examination, judicial caution to the jury about the need for care before accepting this evidence, and the judicial discretion to exclude the evidence if its illegitimately prejudicial effect outweighs its probative value.

That is a conservative approach. In our neck of the woods we have radically reformed the law: s 45 of the Evidence Act 2006 imposes a burden of proof of reliability on the prosecution where specified formal procedures have not, without good reason, been followed. The standard of proof here is beyond reasonable doubt. The defence may still challenge the admissibility of the evidence if formal procedures have been followed, but the burden is on the defence to prove on the balance of probabilities that the evidence is unreliable. These provisions are discussed in R v Edmonds [2009] NZCA 303 at [79]-[128] (sorry keen readers, not currently available online). The issue for the judge is whether the evidence is sufficiently reliable to go to the jury, not whether the evidence establishes identity to the relevant standard. Proof of reliability is distinguished from proof of identity, and the focus for the prosecution's burden is on the circumstances in which the identification was made. It does not include other evidence suggesting guilt, such as a confession.

There are issues concerning the interpretation of s 45 that remain to be explored. Its application in judge-alone trials has been considered by the Supreme Court: Harney v Police.