|
|
Home
| Databases
| WorldLII
| Search
| Feedback
Court of Appeal of New Zealand |
![[AustLII]](/images/tiny-austlii.gif)
Court of Appeal of New ZealandLast Updated: 2 February 2018
For a Court ready (fee required) version please follow this link
IN THE COURT OF
APPEAL OF NEW ZEALAND
CA33/07
[2007] NZCA 510
Hearing: 25 June 2007
Court: Glazebrook, Hammond, Chambers, Arnold and Wilson JJ
Counsel: C W J Stevenson for Appellant
M D Downs and D R La Hood for Crown
Judgment: 29 June 2007 at 10am
Reasons for Judgment: 16 November 2007 at 11am
|
JUDGMENT OF THE COURT
|
The appeal is allowed, the conviction is quashed
and a verdict of acquittal is
entered.
____________________________________________________________________
REASONS
Glazebrook, Chambers, Arnold and Wilson JJ [1]
Hammond J [236]
GLAZEBROOK, CHAMBERS, ARNOLD AND WILSON JJ
(Given by Glazebrook
J)
Table of Contents
Para No
A fatal collision [1]
What is the correct test to be
applied under s 385(1)(a)? [4]
Contentions of the parties on the current test [4]
Is Ramage out of line with the test in comparable jurisdictions? [7]
THE DECISION IN RAMAGE [8]
THE AUTHORITIES REFERRED TO IN RAMAGE [10]
CONCLUSION ON THE RAMAGE TEST [21]
THE POSITION IN ENGLAND AND WALES [22]
THE POSITION IN CANADA [26]
THE POSITION IN AUSTRALIA [32]
COMPARISON WITH RAMAGE [41]
SHOULD WE ADOPT THE POSITION IN ANY OF THE ABOVE
JURISDICTIONS? [45]
Have subsequent New Zealand cases narrowed the Ramage test? [47]
Should full appellate review be limited to certain types of cases? [54]
CONSTITUTIONAL CONCERNS [57]
IDENTIFICATION EVIDENCE [61]
EXPERT EVIDENCE [64]
DOCUMENTARY EVIDENCE AND INFERENCES [71]
ADVANTAGES IN HEARING AND SEEING WITNESSES [73]
CREDIBILITY [76]
CONCLUSION [85]
The test to be applied under s 385(1)(a) [86]
Can Mr Munro’s conviction stand when the correct test
is
applied? [91]
The issues [91]
Eyewitness and related evidence [97]
EYEWITNESS EVIDENCE [99]
THE VICTIM’S CANNABIS CONSUMPTION [103]
EVIDENCE OF CELLPHONE USE [106]
ASSESSMENT OF EYEWITNESS AND RELATED EVIDENCE [108]
Expert evidence [118]
SENIOR CONSTABLE MADDAFORD [123]
SERGEANT STABLES [140]
MR BECKETT [151]
Deficiencies in the investigation [168]
Assessment of the expert evidence [180]
Assessment of the remaining Crown evidence [202]
EVIDENCE OF MR MUNRO’S INTOXICATION [203]
EVIDENCE OF BAD DRIVING [207]
MR MUNRO’S STATEMENTS [215]
CONCLUSION ON REMAINING CROWN EVIDENCE [219]
Aspects of the trial process [220]
Ought a jury acting reasonably have had a reasonable doubt? [229]
Result [235]
Appendix: Comparison of expert evidence on key points
A fatal collision
[1] On Wednesday 17 November 2004, Mr Munro was driving a single cab flatbed Isuzu truck on State Highway One south bound towards Wellington. Between about 9.40pm and 9.48pm, about half way between Manakau and Otaki in an area known as the Pukehou Rail Overbridge, he collided with a Honda car, which was travelling north. The driver of the Honda, Ms Carley Aldridge, died at the scene. Mr Munro suffered a broken sternum and was taken to Palmerston North Hospital. Blood tests taken at the hospital showed 92 milligrams of alcohol per 100 millilitres of blood (the legal limit being 80 milligrams).
[2] On 22 September 2006, Mr Munro was convicted in the District Court at Palmerston North, after a trial before Judge Atkins and a jury, of one charge of causing death while driving with excess blood alcohol under s 61(1)(b) of the Land Transport Act 1998. Mr Munro appealed against his conviction on the basis that the verdict of the jury was unreasonable or that it could not be supported having regard to the evidence see s 385(1)(a) of the Crimes Act 1961. On 29 June 2007, we allowed Mr Munro’s appeal, quashed his conviction and entered a verdict of acquittal. These are the reasons for that decision.
[3] There were two issues in the appeal and we deal with each in turn:
(a) What is the correct test to be applied in assessing a ground of appeal under s 385(1)(a) of the Crimes Act?
(b) Can Mr Munro’s conviction stand when the correct test is applied?
What is the correct test to be applied under s 385(1)(a)?
Contentions of the parties on the current test
[4] Mr Stevenson, on behalf of Mr Munro, submitted that we should revisit the leading authority on s 385(1)(a) of the Crimes Act, R v Ramage [1985] 1 NZLR 392 (CA). In his submission, Ramage is out of line with the test in comparable jurisdictions. He referred to cases such as Weiss v R [2005] HCA 81; (2005) 224 CLR 300, R v Graham [1996] EWCA Crim 1211; [1997] 1 Cr App R 302 (CA) and Yebes v R [1987] 2 SCR 168; (1987) 36 CCC (3d) 417. As an alternative, Mr Stevenson submitted that subsequent New Zealand decisions have narrowed the Ramage test unacceptably. The test in Ramage is not merely whether there is some evidence which, if accepted, would support the conviction – see for example R v McDonald CA142/04 29 July 2004 at [18]. The question should be whether, on the basis of all the evidence, a jury acting reasonably ought to have had a reasonable doubt as to the guilt of the appellant. In Mr Stevenson’s submission, juries are not infallible and appellate review in these terms is necessary to prevent substantive miscarriages of justice.
[5] Mr Downs, for the Crown, submitted that Ramage sets out the correct test and should not be revisited. He did, however, accept that the Ramage test requires a consideration of all of the evidence and that it will, in exceptional cases, lead to a verdict being overturned even in a case where there is evidence that would support a conviction. In his submission, however, this will occur only in cases where an appellate court has relevant expertise or experience that juries may lack, such as in the area of identification evidence, and, in some cases, expert evidence. If appellate courts too readily overturn jury verdicts, this could, he submitted, undermine confidence in the jury system and encroach on the constitutional division between the roles of judge and jury.
[6] We propose to deal with this part of the argument under the following headings:
- (a) Is Ramage out of line with the test in comparable jurisdictions?
- (b) Have subsequent New Zealand cases narrowed the test?
- (c) Should full appellate review be limited to certain types of cases?
- (d) Summary of the test to be applied under s 385(1)(a).
Is Ramage out of line with the test in comparable jurisdictions?
[7] Mr Stevenson’s submission is that Ramage is out of line with the test in comparable jurisdictions and should be changed to accord with the tests used in those jurisdictions. We propose to examine this topic under the following headings:
(a) The decision in Ramage;
(b) The authorities referred to in Ramage;
(c) Conclusion on the Ramage test;
(d) The position in England and Wales;
(e) The position in Canada;
(f) The position in Australia;
(g) Comparison of position in other jurisdictions with Ramage;
(h) Should we adopt the test used in any of the above jurisdictions?
THE DECISION IN RAMAGE
[8] Ramage concerned an application on behalf of Mr Ramage and Mrs Thomas for leave to appeal against convictions for setting fire to Mrs Thomas’ house with the intention of defrauding an insurance company. The Crown case was circumstantial and the ground of appeal was under s 385(1)(a) of the Crimes Act. Somers J, for the Court, set out the test to be applied as follows (at 393):
The common ground advanced by the applicants - that the verdict is unreasonable or cannot be supported having regard to the evidence - is that contained in s 385(1)(a) of the Crimes Act 1961. A verdict will be of such a character if the Court is of the opinion that a jury acting reasonably must have entertained a reasonable doubt as to the guilt of the applicant. It is not enough that this Court might simply disagree with the verdict of the jury: see R v Mareo (No 3) [1946] NZLR 660; R v Ross [1948] NZLR 167; R v Kira [1950] NZLR 420; Chamberlain v R [(1984) [1984] HCA 7; 153 CLR 521].
[9] After reviewing the evidence, the Court was satisfied that it was open to the jury to find beyond reasonable doubt that the fire was deliberately lit, that it was the applicants who were responsible and that their intent was to receive the insurance money. The Court remarked (at 395) that a decision as to whether a verdict was unreasonable or cannot be supported having regard to the evidence is not one which lends itself to any extensive elaboration of reasons.
THE AUTHORITIES REFERRED TO IN RAMAGE
[10] As noted above at [8], the Court in Ramage referred, as authority for the test it promulgated, to three New Zealand cases and one Australian authority. What is said in those cases can thus be used to give more guidance with regard to the appropriate approach to s 385(1)(a) than the relatively brief statement of principle in Ramage itself.
[11] The first New Zealand case referred to in Ramage was R v Mareo (No 3). The application in that case was under s 4 of the Criminal Appeal Act 1945, the relevant part of which was materially identical to s 385(1)(a). In that case Myers CJ, for the Court, said that in England verdicts have been set aside on the ground that the verdict or the trial has been “unsatisfactory” and this test has been applied in considering whether a verdict cannot be supported having regard to the evidence. He regarded this as a sound working test. He said, however, that a verdict is not to be regarded as unsatisfactory merely because the trial judge or the judges of this Court would not have come to the same conclusion as the jury (at 670). A miscarriage of justice must be apparent before the Court sets aside a conviction.
[12] The next case referred to was Ross. In that case Smith J, for the Court, said that a verdict would not be found to be unreasonable merely on the basis that the evidence at trial amounted to a weak case against the appellant. If there is evidence to support the conviction it will not be quashed even if the members of the appellate Court feel some doubt about it. A new trial will be ordered where the verdict is such that a jury of 12 reasonable men could not reasonably and properly have found, bearing in mind the higher degree of assurance necessary in criminal cases and on reviewing the whole of the evidence. The onus of satisfying the Court that this test is met lies upon the person convicted. In Kira, the third New Zealand case referred to in Ramage, Northcroft J, for the Court, applied the test set out in Ross to dismiss the appeal but without a detailed discussion of the evidence.
[13] The final case referred to in Ramage was the Australian High Court decision in Chamberlain. Mrs Chamberlain had been convicted of the murder of her baby, Azaria, and her husband had been convicted as an accessory after the fact. Mrs Chamberlain’s defence at trial was that a dingo had taken the baby out of the tent in which she was sleeping. Mr and Mrs Chamberlain appealed unsuccessfully against their convictions to the Full Court of the Federal Court. They then applied for leave to appeal to the High Court. The majority of the High Court granted special leave to appeal (Brennan J dissenting) but by a different majority (Murphy and Deane JJ dissenting) dismissed the appeal and upheld the convictions.
[14] Gibbs CJ and Mason J delivered a joint judgment. Gibbs CJ and Mason J accepted that an appellate court has the power and duty to set aside a verdict which it considers to be unsafe or dangerous (see at 530 531). In their view, an appellate court may interfere with a verdict which is unsafe or unsatisfactory even if there is sufficient evidence to support it as a matter of law and even though there has been no misdirection, erroneous reception or rejection of evidence and no other complaint as to the course of the trial.
[15] The responsibility of deciding upon the verdict, whether a conviction or acquittal, lies with the jury, however, and there is no justification for an appellate tribunal to usurp the function of the jury and disturb a verdict of guilty simply because it disagrees with the jury’s conclusion. The trial is by jury and, absent another source of error, the jury’s verdict should not be interfered with unless an appellate court concludes that a reasonable jury ought to have had a reasonable doubt (at 534). Any review of the evidence by an appellate court is for the purpose of deciding whether it was open to the jury to be satisfied beyond reasonable doubt that the accused was guilty.
[16] In taking the above approach, Gibbs CJ and Mason J followed that of Dawson J in Whitehorn v The Queen [1983] HCA 42; (1983) 152 CLR 657. In that case, Dawson J said that the proper question to ask was whether the appellate court considers that, on the whole of the evidence, it was open to the jury to be satisfied beyond reasonable doubt that the accused was guilty (at 686). Dawson J rejected the approach of Barwick CJ in Ratten v The Queen (1974) 131 CLR 510 at 516. In that case, Barwick CJ had said that it is the reasonable doubt in the mind of the court which is the operative factor. He went on to say:
It is of no practical consequence whether this is expressed as a doubt entertained by the court itself, or as a doubt which the court decides that any reasonable jury ought to entertain. If the court has a doubt, a reasonable jury should be of a like mind. But I see no need for any circumlocution; as I have said it is the doubt in the court’s mind upon its review and assessment of the evidence which is the operative consideration.
[17] Dawson J (at 687 of Whitehorn) recognised that there may be cases where a Court, without regard to the verdict of the jury, may entertain a doubt itself but may properly conclude that the jury might reasonably have reached a verdict of guilty on the evidence at trial. This can apply in particular in cases where a verdict turns wholly or partly on questions of credibility or competing inferences (at 688).
[18] While dissenting on the scope of the powers conferred on the Federal Court in regard to criminal appeals and whether special leave to appeal should be granted, Brennan J agreed in Chamberlain with the result arrived at by Gibbs CJ and Mason J: the convictions were not unsafe. Brennan J said that the law does not permit an appellate court to overturn a jury verdict on the grounds that it is unsafe merely because the court happens to entertain a reasonable doubt (at 603). However he considered that an appellate court may realise, from long curial experience, that certain types of evidence are frequently unsafe. Appellate courts will have superior expertise in assessing the safety of such evidence (at 604 605).
[19] Murphy J (in dissent) did not specifically endorse the test set out in the joint judgment but neither did he reject it. He said that the greatest respect should be given by appellate courts to jury verdicts and any attempt to downgrade the jury to a mere nominal or symbolic role should be restricted (at 569). However, he recognised that inevitably juries sometimes make mistakes. Because of such miscarriages of justice, courts of criminal appeal have been given power to set aside convictions, not only where the Judge wrongly admitted or rejected evidence or misdirected the jury, but also where, although there was evidence which could justify the verdict, the appeal court considered it unsafe. In his view, the appellate system must operate as a further safeguard against the mistaken conviction of the innocent (see 569). Murphy J considered that it was unsafe for the jury to accept certain evidence in the case at hand and that the rational hypothesis advanced by the defence had not been excluded beyond reasonable doubt. For those reasons he would have quashed the convictions (at 576 577).
[20] Deane J also dissented and considered the convictions to be unsafe. He specifically endorsed Barwick CJ’s approach in Ratten and thus did not accept the test promulgated by Gibbs CJ and Mason J. In Deane J’s view, the principle that no person should be convicted of a serious crime except by a jury acting on the evidence has no corollary requiring that every person who is found guilty by a jury’s verdict should remain so convicted. In his view, the safeguard provided by trial by jury is not dependent upon any assumption of the infallibility of the verdict of a jury. The cause of the continued acceptance of trial by jury will not be served by treating a jury’s verdict of guilty as unchallengeable or unexaminable (see at 617). He said (at 617 618):
To the contrary, so to treat a jury’s verdict of guilty could sap and undermine the institution of trial by jury in that it would, in the context of modern views of what is desirable in the administration of criminal justice, be liable to be seen as a potential instrument of entrenched injustice.
CONCLUSION ON THE RAMAGE TEST
[21] It is clear from Ramage and the authorities cited in that case that an appellate court may find a verdict to be unreasonable or unsupported by the evidence even where there is some evidence to support it and there has been no misdirection. This will be the case if, taking into account all of the evidence, a reasonable jury could not be satisfied of guilt to the requisite standard. The concentration is on a reasonable jury and not on whether the appellate court might have differed in its conclusion from that reached by the jury. The approach in Ramage appears to us to accord with the statutory wording. Mr Stevenson, however, submitted that this interpretation is out of line with the test used in other jurisdictions. We examine that submission next.
THE POSITION IN ENGLAND AND WALES
[22] The criminal appeal provisions in England and Wales were recast in 1968 and were further refined in 1995. The concept of unreasonable verdicts was replaced with that of unsafe convictions. After the 1968 amendments, the principle of “lurking doubt” emerged as a means of assessing whether a verdict is unsafe or unsatisfactory. The seminal case with regard to this approach was R v Cooper (Sean) [1969] 1 QB 267 (CA). In that case, Widgery LJ for the Court said (at 271) that, in a case where every issue was before the jury and in which the jury was properly instructed, appellate courts will be reluctant to intervene. This is because of the advantage which the jury has in seeing and hearing the witnesses. However, because an appellate court is charged to allow an appeal against conviction if under all of the circumstances of the case it is unsafe, the court must, in the end, ask itself (at 271):
[W]hether we are content to let the matter stand as it is, or whether there is not some lurking doubt in our minds which makes us wonder whether an injustice has been done. This is a reaction which may not be based strictly on the evidence as such; it is a reaction which can be produced by the general feel of the case as the court experiences it.
[23] The Crown submitted that, while the lurking doubt approach continues to exist in England and Wales, its application has been ad hoc. In the Crown’s submission, the lurking doubt approach is reserved for cases where the court considers that there are powerful reasons for concluding that the conviction is unsafe despite the jury’s verdict, including procedural defects such as an accused’s lack of opportunity to put a defence, delay, or where the case should not have been left to the jury in any event.
[24] The Crown’s submission receives support in the article by Leigh “Lurking Doubt and the Safety of Convictions” [2006] Crim LR 809. Leigh notes (at 815) that the Court of Appeal of England and Wales has not, even in the most expansive of modern cases, sought to substitute its view of the facts reached by a properly instructed jury in determining whether the verdict is unsafe. In his view, were a court to do so, it would fall foul of the principles enunciated by the House of Lords in R v Pendleton [2002] 1 WLR 72. In Leigh’s view, Cooper (Sean) should not be construed as permitting a purely visceral reaction to a case and it has not been applied in that way. Leigh emphasises that the Court’s task in reviewing a conviction is purely analytical see Dial v State of Trinidad and Tobago [2005] UKPC 4; [2005] 1 WLR 1660 at [43] (PC). We also refer to the recent decision of the Privy Council in Dookran v The State (Trinidad and Tobago) [2007] UKPC 15 in which that Court said that the review must be based on an overall view of the features of the specific case (see at [29]).
[25] The cases referred to by Leigh and the case of Dookran make it clear that the essential question for an appellate court is simply whether the convictions are unsafe. Where the Court experiences a lurking doubt about the safety of a conviction then it may be that the conviction is unsafe. The Court must not displace the role of the jury as the finder of fact and any inquiry into the safety of a conviction must be conducted in the context of the overall features of the case and the other evidence.
THE POSITION IN CANADA
[26] The Canadian Criminal Code (R.S. 1985, c. C-46) contains a ground of appeal (s 686(1)(a)(i)) that is virtually identical to s 385(1)(a). The leading case on how this ground of appeal is assessed is R v Corbett [1975] 2 SCR 275 (SCC); (1973) 14 CCC (2d) 385 (SCC). The majority of the Supreme Court in Corbett made it clear that the proper test was whether the verdict was one that no reasonable jury could have reached. It held that the function of a reviewing court is not to substitute itself for the jury and to decide guilt or innocence. The task is to decide whether the verdict was one that a properly instructed jury acting judicially could reasonably have rendered (see at 389). There is a presumption that the result achieved at trial is the proper result and it falls to the appellant in a given case to demonstrate that the conviction is a miscarriage of justice. The minority (Laskin and Spence JJ) would have allowed the appeal, and emphasised that the existence of some evidence upon which the jury could convict will not permit an appellate court to abdicate its function, namely, an assessment of whether on the evidence taken as a whole the verdict is unreasonable (at 391). Of course an appellate court will not view a verdict as unreasonable merely because for itself it might have reached a different conclusion than the jury (see at 394).
[27] In Yebes the Supreme Court explained the obligations of an appellate court under this provision in the Criminal Code. The test is very similar to that of the majority in Corbett. The appellate court must look beyond the question of whether there is some evidence to support the conviction. Rather, it must determine whether, on the whole of the evidence, the verdict is one that a properly instructed jury, acting judicially, could reasonably have rendered (at 430). It was recognised that, to some extent, this will require the court on appeal to re-examine and re-weigh the evidence and ascertain whether it is sufficient to justify the verdict.
[28] The test was further elaborated on by another unanimous bench of the Supreme Court in R v W(R) [1992] 2 SCR 122; (1992) 74 CCC (3d) 134 (SCC). McLachlin J, writing for the Court, said (at 140) that the test was not whether the verdict was one that no jury could possibly have come to, but whether it was a verdict no jury could reasonably have come to taking into account the standard of beyond reasonable doubt. An appellate court could overturn a verdict based on findings of credibility where, after considering all the evidence and having due regard to the advantages afforded to the fact finder, it considers the verdict unreasonable (at 142).
[29] We were also referred to R v Biniaris [2000] 1 SCR 381; (2000) 143 CCC (3d) 1. In that case, the Supreme Court unanimously stated that triers of fact, whether juries or judges, have considerable leeway in their appreciation of the evidence and the proper inferences to be drawn from that evidence, in their assessment of the credibility of witnesses and in their ultimate assessment of whether the Crown’s case is made out beyond a reasonable doubt (at [24]). The Court said that any judicial system must tolerate reasonable differences of opinion on factual issues and that all factual findings are open to the trier of fact except unreasonable ones. The test in Yebes requires the appellate court to determine what verdict a reasonable jury, properly instructed, could judicially have arrived at and in doing so, to review, analyse and, within the limits of appellate disadvantage, weigh the evidence (at [36]).
[30] The Court confirmed, at [38], that “lurking doubt” is not a proper basis upon which to interfere with the findings of a jury without further articulation of the basis for such a doubt. While lurking doubt is not, in itself, sufficient to justify interfering with the conviction, however, it may be a useful signal that the verdict was indeed reached in a non-judicial manner. In that case the appellate court must proceed further with its analysis of the evidence. Close scrutiny might reveal that the jury reached its verdict pursuant to an analytical flaw, or judicial experience about the need for special caution in evaluating certain types of evidence might lead an experienced appellate judge to conclude that in a given case the jury’s fact-finding process was flawed and thus the result was unreasonable (at [39]).
[31] In A-G v R [2000] 1 SCR 439; (2000) 143 CCC (3d) 46 (SCC) the comments in Biniaris were endorsed, with the following qualification: the fact that an appeal court would have had a doubt is insufficient to justify the conclusion that the trial judgment was unreasonable (at [29]).
THE POSITION IN AUSTRALIA
[32] The relevant statutes in most Australian states and territories contain a ground of appeal which is similar to s 385(1)(a) in New Zealand – see the discussion in Chamberlain at 530 – 531. The leading case in Australia on unreasonable verdicts, after Chamberlain, is M v R [1994] HCA 63; (1994) 181 CLR 487. The appellant in that case was convicted of indecently assaulting his 13 year old daughter. There were discrepancies in the evidence of the daughter and there was no corroboration. Also of concern was the delay of the complainant in making the allegations. The majority (Mason CJ, Deane, Dawson and Toohey JJ) held, that in the circumstances, the convictions were unsafe and unsatisfactory. Brennan and Gaudron JJ dissented.
[33] The majority in that case observed (at 492) that the criminal appeal provisions in Australia allow a verdict that is unsafe or unsatisfactory to be set aside, notwithstanding that those words do not appear in the legislation. The question is one of fact which the court must decide by making its own independent assessment of the evidence to determine whether, notwithstanding that there is evidence upon which a jury might convict, nonetheless it would be dangerous in all of the circumstances to allow the verdict of guilty to stand. The court must ask itself whether, upon the whole of the evidence, it was open to the jury to be satisfied beyond reasonable doubt that the accused was guilty. In answering that question, full regard must be paid to the consideration that the jury is the body entrusted with the primary responsibility for determining guilt or innocence and the fact that the jury has had the benefit of seeing and hearing the witnesses (see at 493).
[34] The majority said, however, (at 494) that in most cases a doubt experienced by an appellate court will be a doubt which a jury ought also to have experienced. It is only where a jury’s advantage in seeing and hearing the evidence is capable of resolving a doubt experienced by an appellate court that the court may conclude that no miscarriage of justice occurred. Where the evidence lacks credibility for reasons which are not explained by the manner in which it was given, a reasonable doubt experienced by the Court will be a doubt which a reasonable jury ought to have experienced.
[35] If the evidence contains discrepancies, displays inadequacies, is tainted or otherwise lacks probative force in such a way as to lead the court to conclude that, even making full allowances for the advantages enjoyed by the jury, there is a significant possibility that an innocent person has been convicted, then the court is bound to set aside the verdict (see at 494). In doing so, the majority did not consider the court to be substituting trial by the appellate court for trial by jury, for the ultimate question must always be whether the court thinks that, upon the whole of the evidence, it was open to the jury to be satisfied beyond reasonable doubt that the accused was guilty (at 495).
[36] In our view, the test in M v R is less jury-focussed than the test in Chamberlain and moved much closer to the approach taken by Barwick CJ in Ratten than had been the case in Chamberlain. Under the M v R approach, deference to jury verdicts only applies where any discrepancy of view between the Court and the jury can be explained by the manner in which the evidence was given. In other cases, the majority in M v R treated any doubt experienced by the Court as one which in most cases a jury should have had.
[37] Brennan and Gaudron JJ dissented in separate opinions in M v R. Both emphasised the role of the jury in assessing the credibility of a complaint. Brennan J said (at 507) that it is the jury’s composite and broad experience of life that is paramount in our structure of administrative and criminal justice. Words in a printed transcript may tell one story to a critical legal mind and another to those who test a story for truths or falsehood according to a broad experience of life. Inconsistencies which “loom large when painted with the colours of advocacy” may be insignificant minutiae once the witness convinces a jury that he or she is honestly attempting to tell the truth (see at 507). Gaudron J agreed with the test formulated by the majority and considered that, when applied, it led to the conclusion in this case that one of the convictions was unsafe. The other convictions rested on the credibility of the appellant’s evidence, which was not unsatisfactory, and the case had no significant weakness (at 516). She would have upheld the other convictions.
[38] Arguably, the High Court of Australia moved even further towards a return to Ratten in Weiss. Although Weiss was a case relating to the equivalent of the proviso to s 385 of our Crimes Act, the High Court said that the assessment of whether a substantial miscarriage of justice has actually occurred is to be undertaken in the same way an appellate court decides whether a verdict of the jury should be set aside on the ground that it is unreasonable or cannot be supported having regard to the evidence.
[39] The High Court said that assessments based on what the actual jury or a hypothetical jury might have done are liable to distract attention from the task to be undertaken. An appellate court must make its own independent assessment of the evidence and determine whether, making due allowance for the “natural limitations” that exist in the case of an appellate court proceeding wholly or substantially on the record, the accused was proved beyond reasonable doubt to be guilty of the offence on which the jury returned its verdict of guilty. The High Court held that the task is an objective one, the standard of proof is the criminal standard and that it must be undertaken on the whole of the record of the trial, including the fact that the jury returned a guilty verdict.
[40] It must be noted, however, that Weiss concerned wrongly admitted evidence which could have had an effect on the jury’s assessment of the relative credibility of the witnesses in that case. In Weiss therefore, it was not possible to rely on the jury’s assessment of the witnesses. Weiss was a case on the Australian equivalent of the proviso to s 385 and not on unreasonable verdicts – see Libke v R [2007] HCA 30; (2007) 235 ALR 517 at [115] (HCA) per Hayne J. We also note that the latest pronouncement of the High Court endorses the M v R test - see R v Hillier [2007] HCA 13; (2007) 233 ALR 634 at [20] (HCA). This reduces the force in Mr Stevenson’s argument that Weiss mandates a broader test in the assessment of whether convictions are unreasonable or unsupported by the evidence.
COMPARISON WITH RAMAGE
[41] The statutory test in England and Wales is different from that in New Zealand. However, the difference may not be significant – see for example the comments in Mareo (see at [11] above) and Chamberlain (see at [14] above) which equate the statutory test of unreasonableness with “unsafe”, “unsatisfactory” or “dangerous” verdicts. In any event, apart from the “lurking doubt” jurisprudence, we consider that the approach in England and Wales is essentially similar to that employed under the Ramage test. We are not attracted to the “lurking doubt” concept, except as a trigger for a fuller review of the evidence (see above at [30]). If a lurking doubt remains after a review of the evidence then the reason for that doubt must be able to be articulated. The task then should be to assess whether a reasonable jury ought to have experienced such a doubt. We consider that the lurking doubt approach may even risk a court focusing on the wrong inquiry: it might lead to focus on actual guilt or innocence, rather than on whether a jury ought to have had a reasonable doubt as to guilt.
[42] We do not consider that the Canadian approach to the equivalent of s 385(1)(a) is different from the Ramage approach. Indeed, we consider that the statements of principle in the Canadian cases are applicable in New Zealand. We do note, however, that the Canadian courts may in practice pay more deference to jury verdicts than is the case in New Zealand – see Furgiuele “The Self-Limiting Appeal Courts and Section 686” (2007) 52(2) Crim L Q 237, Plaxton “The Biased Juror and Appellate Review: A Reply to Professor Coughlan” (2002) 44 CR (5th) 294 and the comments of Mr Justice Kaufman in his Report of the Commission of Proceedings Involving Guy Paul Morin (Report 1998) (Vol 2, Recommendation 87, at 1189 90).
[43] This leaves the test in Australia as expressed since Chamberlain. As indicated above, the test in Australia has subsequently moved closer to the Ratten test than was the case in Chamberlain. Nevertheless, this is a change in degree rather than kind. The test is still whether a reasonable jury ought to have had a reasonable doubt essentially the Ramage test. As pointed out by the Crown, however, since M v R, the Australian test often leads, as a general rule, to a much more detailed examination of the evidence (even in cases resting primarily on credibility findings) than occurs in New Zealand.
[44] Mr Stevenson put great weight on the comments in the Australian cases that a reasonable doubt experienced by the appellate court would usually mean that a reasonable jury ought to have experienced such a doubt. We do not accept the proposition that a reasonable doubt entertained by the court will necessarily be a reasonable doubt that ought to have been entertained by a jury. This will depend on the type of case (as indeed is recognised in a limited manner in M v R itself). In cases where there is a clear disadvantage for the appellate court in not having seen and heard the witnesses then a doubt felt by the appellate court may well be because of that very fact. In cases calling for an analysis of documentary or technical evidence, it is much more likely that a doubt experienced by an appellate court will be one that ought to have been experienced by a jury. We do not, however, consider it helpful to express the test in that way as it risks misconstruing the task of the appellate court and the court usurping the proper role of the jury.
SHOULD WE ADOPT THE POSITION IN ANY OF THE ABOVE JURISDICTIONS?
[45] There is nothing in the above jurisdictions that would suggest that we should depart from the Ramage test. The approach in Canada is the same as in New Zealand (and, in practice, probably stricter). Apart from the “lurking doubt” test, to which we are not attracted, the approach in England and Wales is similar to that in New Zealand.
[46] The differences between the Australian and New Zealand approaches are, we think, of degree and not of kind. Nevertheless, the move back in M v R towards the Ratten test is contrary to the test promulgated in Chamberlain, which was approved by this Court in Ramage. Any move in the same direction would require Ramage to be overruled in this regard. In our view, the factors in R v Chilton [2005] NZCA 295; [2006] 2 NZLR 341 (CA) that may lead this Court to overrule one of its previous decisions are not met. In addition, the Supreme Court has given leave in O v R [2007] NZSC 49 and will be considering the application and interpretation of the test in s 385(1)(a). It would be inappropriate in these circumstances for us to review Ramage, particularly as it is not necessary in this case to depart from the Ramage test in order to allow Mr Munro’s appeal. In any event, as we said at [21], we consider the Ramage test to accord with the statutory wording.
Have subsequent cases narrowed the Ramage test?
[47] We now analyse how Ramage has been applied in subsequent decisions. We do this in order to assess Mr Stevenson’s submission that Ramage has not been properly understood in subsequent cases and in particular his submission that subsequent cases have sustained convictions if there is some evidence which points to guilt without examining that evidence in the context of all of the evidence at trial.
[48] The first case Mr Stevenson drew our attention to was that of McDonald where this Court stated at [18] that, if there is evidence available to a jury which, if accepted, would support its verdict, then such verdict cannot be impugned on the ground of unreasonableness. That comment, however, has to be read in context. In the previous paragraph, it was stated that an appeal on the grounds contained in s 385(1)(a) can only succeed if the Court is of the view that a jury, acting reasonably, must have entertained a reasonable doubt of the guilt of the appellant. This was the test applied by the Court in McDonald and the Court assessed whether that test was met by examining all of the evidence relevant to identification (the main issue at trial and on appeal). The Court traversed that evidence in some detail before the conclusion was drawn that there was ample foundation for the jury’s verdict. The Court said that the appellant merely sought to reassert propositions which would have been evaluated by the jury and which the jury was entitled to reject.
[49] The next case referred to by Mr Stevenson was R v H CA200/98 28 October 1998. In that case there was a sexual assault on a woman in her home by a masked attacker. There was strong circumstantial evidence pointing to her former partner as the perpetrator. The victim, however, said in her testimony that, despite being well acquainted with her former partner, at no stage did she recognise the attacker as being him. Notwithstanding this evidence, Mr H was convicted.
[50] Heron J, for the Court, said that it was not the law that if this Court considered that there exists a reasonable doubt, so too must the jury. He considered that Ramage encapsulates the view (expressed in other authorities) to the effect that this Court:
does not proceed on such lines as these - “look at the evidence, see what conclusion the Court would have come to and set aside the verdict if it does not correspond with such conclusion.” R v Hancox (1913) 8 Cr App 193, 197.
[51] There is nothing unorthodox in the above statement. A court may be required to consider all of the evidence but the purpose is to assess whether a reasonable jury ought to have come to the view that there was a reasonable doubt about the appellant’s guilt, not whether the court itself had a reasonable doubt. This clearly aligns with the test accepted in Chamberlain (see above at [13]), which was in turn approved in Ramage.
[52] In line with that approach, the Court in R v H went through the whole of the evidence in some detail before concluding that a verdict of guilty was available to the jury. The Court noted that all of the inconsistencies in the circumstantial evidence had been fully placed before the jury. So too had the allegations of the police planting of evidence (a damning piece of evidence allegedly found in the accused’s car) and the jury had also heard the evidence of the complainant (at trial) that she did not recognise her attacker. The Court noted that, when interviewed immediately after the event, the complainant had simply said that it could not have been him. The Court concluded that all of the challenged evidence had been examined in detail at the trial and was for the jury to weigh.
[53] Neither McDonald nor R v H, therefore, is authority for the proposition that a Court should not have regard to all of the evidence when assessing a ground of appeal under s 385(1)(a). The test remains that set out in Ramage. The question is whether, on all of the evidence, a jury acting reasonably ought to have had a reasonable doubt. Put another way, the question is whether it was open to a jury acting reasonably to consider, taking into account all of the evidence, that the Crown had proved guilt to the requisite standard.
Should full appellate review be limited to certain types of cases?
[54] While the Crown accepted that there may be cases where a jury verdict can be impugned under s 385(1)(a), despite there being some evidence to support the conviction, it submitted that such cases will be rare. This is because cases where there is insufficient evidence will have been weeded out at an earlier stage of the process, for example at depositions, on a s 347 application or through a not guilty verdict.
[55] The Crown submitted that only in truly exceptional cases or where judicial experience indicates that certain categories of evidence may be unreliable will a qualitative analysis of the evidence by the appellate court be warranted. It is proper that cases where a s 385(1)(a) ground of appeal succeeds are rare. If that were not the case, in the Crown’s submission, confidence in the jury system would be undermined and the constitutional divide between the role of judge and jury compromised. The Crown submitted therefore that cases where s 385(1)(a) might succeed should be limited to those cases where this Court might have special expertise.
[56] We examine first whether too ready a recourse to s 385(1)(a) will undermine confidence in the jury system and compromise the constitutional divide between judge and jury. We then move to the issue of whether appellate review should be limited to particular types of cases using two of the examples given by the Crown and two of our own. We do this under the following headings:
(a) Identification evidence;
(b) Expert evidence;
(c) Documentary evidence and inferences;
(d) Advantages in hearing and seeing witnesses;
(e) Credibility; and
(f) Conclusion.
CONSTITUTIONAL CONCERNS
[57] The Crown’s submission that appellate review might undermine confidence in the jury system can be shortly dealt with. It cannot be the case that letting unreasonable verdicts stand could enhance public confidence in the jury system. It would, in our view, have exactly the opposite effect. In this regard, we endorse the remarks of Deane J in Chamberlain (see at [20] above). In addition, to take such a position would not accord with the statutory language. Section 385(1)(a) requires the Court to allow an appeal if the verdict is unreasonable or if a verdict cannot be supported having regard to the evidence.
[58] As to the concern about the constitutional divide between judge and jury, this is clearly more of an issue at the trial level than it is at the appellate level. There is not the same clear constitutional divide between an appellate court and the jury because the appellate court has a reviewing role under the Crimes Act. This review, when considering a ground of appeal under s 385(1)(a), must extend to a review of the facts, to the extent that that is possible on appeal and within the limitations under which an appellate court must operate.
[59] Having said this, we accept the Crown submission that the cases in which the Court is likely to intervene under s 385(1)(a) will be relatively rare. This is because cases where there is insufficient evidence will usually have been weeded out at an earlier stage (either by the judge or by the jury through an acquittal).
[60] We now examine whether the rarity of a successful appeal under s 385(1)(a) is also because review should, as the Crown submitted, be limited to particular types of cases.
IDENTIFICATION EVIDENCE
[61] One category of cases identified by the Crown where an appellate court might have particular expertise is where identification evidence is involved.
[62] There is no doubt that there are major risks with eyewitness evidence see Benton and others “Eyewitness Memory is Still Not Common Sense: Comparing Jurors, Judges and Law Enforcement to Eyewitness Experts” (2006) 20 Applied Cognitive Psychology 115, cited in McClellan “Who is Telling the Truth: Psychology, Common Sense and the Law” (2006) 80 ALJ 655 at 660. This research makes it clear that, as a group, jurors often exhibit deficiencies in their assessment of eyewitness knowledge and reliability. See also Bingham The Business of Judging (2000) at 260, citing the report of the Devlin Committee.
[63] However, warnings have long been required to be given to juries in this regard (for example, under the now repealed s 344D of the Crimes Act) and this position continues under s 126 of the Evidence Act 2006, which now also applies to voice identification evidence. Arguably, if a jury still finds an accused guilty in the face of such warnings there may be less rather than more justification for interfering with that jury’s verdict on the basis of a greater understanding of the pitfalls of identification evidence at appellate level see the comments on matters being fully before the jury in R v H (above at [52]).
[64] The next category of cases identified by the Crown were those involving expert evidence.
[65] The proper scope of appellate review of expert evidence was one of the main issues in Chamberlain. One of the main planks of the Crown case had been scientific evidence that traces of blood found in the Chamberlain’s car contained foetal haemoglobin, which is contained in a baby’s blood up to the age of about six months. Baby Azaria was aged two months. There was conflicting evidence from the Crown and defence experts in that regard.
[66] Gibbs CJ and Mason J accepted that it is the function of the jury to consider which of two bodies of conflicting evidence, technical or otherwise, that they will accept. However, they agreed with Jenkinson J in the Full Federal Court (see [1983] FCA 78; (1983) 72 FLR 1 at 81 82) that the means of evaluating evidence, which the jury enjoys by hearing and watching witnesses and which are denied an appellate tribunal, could not have enabled the jury in that case reasonably to have eliminated the doubt as to the foetal haemoglobin issue which a careful consideration of the transcript of evidence and the exhibits raised in the mind. Gibbs CJ and Mason J said that the most that could be said against the defence experts was that their work was done in the “comparative seclusion of academic surroundings” so that they lacked the day-to-day experience of the forensic scientists, that they exhibited “an unbecoming arrogance” (in the words of Bowen CJ and Forster J in the Full Federal Court - see [1983] FCA 78; 72 FLR 1 at 30) and that one of the defence experts did not fare well in cross-examination.
[67] Gibbs CJ and Mason J said, however, (at 559) that there was no challenge to the knowledge, honesty or impartiality of either defence expert. The criticisms that those experts advanced of the Crown evidence appeared rational and compelling. Even though the Crown witnesses had answers to those criticisms and there was evidence to support a finding that the blood in the car contained foetal haemoglobin, the conflicting evidence should have raised a reasonable doubt in a reasonable mind and there was no other evidence that could have resolved that doubt. The jury therefore could not safely assume as a primary fact that the blood was foetal blood. Under the test promulgated in Chamberlain a primary fact is one required to be proved beyond reasonable doubt: see the discussion of Gibbs CJ and Mason J at 538 9). (As an aside, we note that this is not the position in New Zealand – see R v Thomas [1972] NZLR 34 at 38 (CA)).
[68] Murphy J, at 575, agreed with Gibbs CJ and Mason J that a conclusion that the sample contained foetal haemoglobin could not safely be drawn. Deane J (at 628 9) noted that the expert evidence called by the Crown, including in relation to the alleged foetal haemoglobin, was contradicted by expert evidence called by the defence. He said that it was “scarcely feasible” that a compelling answer to that conflict of evidence was to be found in observing and hearing the expert witnesses as they gave their evidence. It was certainly not to be found in a careful examination of the transcript of evidence.
[69] In Chamberlain, Brennan J in dissent said (at 598), referring to Dixon J’s comments in Hocking v Bell [1945] HCA 16; (1945) 71 CLR 430 at 496, that it is the jury’s function to determine all conflicts of evidence, including scientific evidence. In his view, an appellate court enjoys no special knowledge equipping it to assess conflicts of scientific evidence. Brennan J accepted that a jury may not be the ideal forum for debating and resolving such issues, but it is the constitutional forum appointed to determine the facts. The sufficiency of reasons advanced for impugning a scientific opinion and the sufficiency of the counter to such reasons advanced by opposing experts is a matter for the jury. In his view, the jury were entitled to accept that the samples from the car contained foetal haemoglobin.
[70] It is settled law that mere disagreement between opposing experts will not necessarily give rise to a reasonable doubt in the mind of the jury. A jury is able and entitled to decide between conflicting bodies of expert evidence on highly technical matters by bringing to bear its own independent judgment: see Freckelton and Selby Expert Evidence (3ed 2005) at 332 and the cases cited therein. A jury remains free to weigh and, if necessary, reject, expert testimony in light of the circumstances of the case – R v Tobin [1978] 1 NZLR 423 at 432 (CA), and see also R v Smith (1995) 12 CRNZ 616 at 623 (CA). However, the fact that genuine differences exist between experts may create a reasonable doubt as to guilt: this is particularly the case where the facts or circumstances of the case are of little assistance in reaching a conclusion on the expert evidence – R v Manu’ula [2007] NZCA 82 at [14]; cf. Walton v R [1978] AC 788 at 793 (PC). Where, as was the case in Chamberlain, there is no rational basis on which a jury can choose between conflicting expert opinions, this will be an important indicator on appellate review that the verdict may be unsafe and that the jury ought to have experienced a reasonable doubt.
DOCUMENTARY EVIDENCE AND INFERENCES
[71] The Crown did not specifically mention documentary evidence in its submissions but it seems to us that in cases which rely mainly or solely on documentary evidence there is little advantage a jury might have in hearing and seeing the witnesses. The legal method, the time available to undertake the review and the distance from the cut and thrust of the trial may even give an appellate court an advantage in the assessment of such evidence. The same may apply where the real issue at trial is the inferences which may validly be drawn from established facts. An appellate court may be required to intervene where the facts support two inferences of equal weight and a jury has incorrectly drawn one or the other: this is pure speculation – see, for example, R v Puttick (1985) 1 CRNZ 644 at 647 (CA) where this Court emphasised the importance of correct directions as to the use of inferences by the jury. See also the remarks of Gault J in Nomoi Holdings v Dustin CA79/00 3 September 2001 at [5].
[72] Even in cases involving documentary evidence or inferences from proved facts, however, the inquiry must be whether the view the jury came to is one which a reasonable jury could arrive at or whether the jury ought to have had a reasonable doubt. It is not a process of substitution of the appellate court’s view for that of the jury. In this regard, we approve the remarks of the Supreme Court of Canada in Biniaris (see above at [29]) in which the Court emphasised that a finder of fact has considerable leeway in its appreciation of the evidence and that reasonable differences of opinion are possible and must be respected.
ADVANTAGES IN HEARING AND SEEING WITNESSES
[73] A common theme in the cases is that any review under s 385(1)(a) must take full cognisance of the advantage the jury has had in seeing and hearing the witnesses. This is even the case under the test set out in M v R (see at [33] [34] above, and also the discussion of Chamberlain at [17] [18] above). Hearing and seeing the witnesses is an advantage that is denied to an appellate court working off a written transcript. As was said in R v Bertrand [1867] EngR 20; (1867) LR 1 PC 520 by Sir John Coleridge, giving the decision of the Privy Council, an appellate court has only “the dead body of the evidence, without its spirit; which is supplied, when given open and orally, by the ear and eye of those who receive it” (at 535).
[74] Tone of voice is important in conveying meaning and this cannot be captured in a written transcript. For example, the concession “it’s possible” could, depending on the tone, have a meaning ranging from “that’s quite likely” to “yes but pigs might fly”. Pauses can be significant and are rarely shown in a transcript. Gestures and facial expressions can also add to meaning and are often not recorded in a transcript unless vital to a witness’ answer and even then, when digital recording systems are in operation, only when specifically read into the record by counsel or the judge.
[75] For this reason, cases relying on oral testimony to a significant degree are likely to be ones where it is more difficult to succeed on a s 385(1)(a) ground. This brings us to the vexed question of credibility assessments.
CREDIBILITY
[76] Traditionally, both in civil and criminal cases, credibility findings at first instance have been very difficult to impugn: see, for example, the comments of this Court in Liebherr Export – AG v Ellison Trading Ltd CA174/03 29 June 2004 at [39] and [41], Nalder & Biddle (Nelson) Ltd v C & F Fishing Ltd CA145/05 31 July 2006 at [88], Stemson v AMP General Insurance (NZ) Ltd [2006] UKPC 30; [2007] 1 NZLR 289 at [12] (PC), R v A [2007] NZCA 354 at [19], and R v T(CA11/07) [2007] NZCA 296 at [25]. Recently, however, courts have been taking into account studies that show up the difficulties with credibility assessment.
[77] For example, in Fox v Percy [2003] HCA 22; (2003) 214 CLR 118 the High Court of Australia recognised that appeal hearings were often unduly constrained by “incantations” against interfering with findings of credibility at trial. The majority of the Court (see at [31] per Gleeson CJ, Gummow and Kirby JJ) commented that in recent years, judges have become more aware of scientific research that has cast doubt on the ability of judges (or anyone else) to tell truth from falsehood accurately on the basis of the appearance of witnesses. Accordingly, judges, both at trial and on appeal, have limited their reliance on the appearances of witnesses and base their conclusions, as far as possible, on contemporary materials, objectively established facts and the apparent logic of events. The majority said that this does not eliminate the established principles about witness credibility; but it tends to reduce the occasions where those principles are seen as critical.
[78] Similar opinions were expressed by Kirby J (with whom Gleeson CJ agreed) in his subsequent decision in CSR v Della Maddalena [2006] HCA 1; (2006) 224 ALR 1 at [19] [22] where he warned of the need for appellate intervention in circumstances where “incontrovertible facts or uncontested testimony” demonstrates the findings to be erroneous, or where they are “glaringly improbable” and “contrary to compelling inferences”.
[79] Any assessments of credibility are also hampered by the highly artificial setting of the courtroom where witnesses are questioned rather than being able to tell their stories unimpeded. Even outside the courtroom setting, it is now recognised that behavioural cues popularly thought to be associated with lying, such as posture, head movements, shifty eyes or gesturing do not indicate dishonesty or lack of credibility – see, for example, Stone “Instant Lie Detection? Demeanour and Credibility in Criminal Trials” [1991] Crim LR 821 at 829.
[80] Studies have also highlighted the fact that witnesses who appear confident and open, and have a good memory for peripheral detail, are far more likely to be believed, regardless of whether they are truthful see the McClellan article (cited above at [62]). Unsavoury and unattractive witnesses are less likely to be believed, because there is a general bias in favour of believing that attractive people are honest. Similarly, once a positive or negative impression is formed, this will attach to all of that witness’ evidence. People do not tend to differentiate between “parts” of a witness’s testimony. This has been described as the “Halo Effect” whereby one perceived good or bad quality in a person will tend to colour all judgments pertaining to the person (see McClellan at 662).
[81] Further, everyone, witnesses and jurors included, are prone to “heuristic illusions” which may help to simplify complex and uncertain tasks, but which are widely recognised as giving rise to the risk of systematic errors and biases. In particular, people tend to suffer from illusions such as “hindsight illusion” (overestimating what could have been anticipated to occur); overconfidence regarding one’s own judgement and predictions; and false-consensus bias, whereby one’s own behaviour is viewed as typical and alternatives are seen as odd or inappropriate - see Mason “Unconscious Judicial Prejudice” (2001) 75 ALJ 676 at 684.
[82] These empirical and psychological studies show that what was once considered the privileged domain of the jury cannot be justified by reference to a jury’s special ability in these areas. Errors in assessing credibility may well be common, and it cannot in all cases be presumed that the deliberative process has corrected them. All the above means that it is no longer possible to make blanket assertions about witness credibility being the province of the jury.
[83] However, it is equally true that assessing credibility from a written transcript in many cases will not achieve better results, particularly in the absence of other contemporary evidence. There remain advantages in hearing and seeing the witnesses as outlined in the previous section. Further, appellate courts are often directed only to parts of the evidence and do not obtain a full sense of the whole trial and the dynamics of the trial process. We also consider that the caution expressed by Brennan J in M v R as to the role of advocacy in highlighting inconsistencies that are more apparent than real is justified (see at [37] above). It also remains the case that there is nothing that restricts the type of evidence upon which a jury can convict and appellate courts have to guard against introducing a corroboration requirement by the back door.
[84] Despite the greater recognition by courts of the difficulties of assessing credibility, where the appellate court is in no better position than the jury to assess the evidence, it is not likely to be easy for an appellant to show that a verdict is unreasonable in cases involving credibility. In this regard, the comments in Ross set out at [12] above remain apposite. This, however, does not mean that credibility assessments by juries are sacrosanct. Verdicts based on credibility assessments may be unreasonable in certain circumstances, for example, those described by Kirby J outlined at [78] above.
CONCLUSION
[85] We do not accept the Crown submission that the cases where a ground of appeal s 385(1)(a) will succeed will be limited to cases where there is particular expertise in the appellate court. It is true that in such cases the appellate court is much more likely to intervene but this is not to say that matters of credibility and the weight of the evidence can never be reviewed. It would be an unjustified restriction on the statutory test to limit s 385(1)(a) to certain types of cases.
The test to be applied under s 385(1)(a)
[86] The correct approach to a ground of appeal under s 385(1)(a) is to assess, on the basis of all of the evidence, whether a jury acting reasonably ought to have entertained a reasonable doubt as to the guilt of the appellant. We consider the word “ought” is a better indication of the exercise to be conducted than the word “must” used in Ramage. It emphasises the task that the Court has to perform. This test also, in our view, accords with the statutory wording.
[87] We consider that McLachlin J’s comments in R v W(R) encapsulate the main elements of the test. The test is not whether the verdict is one that no jury could possibly have come to. A verdict will be deemed unreasonable where it is a verdict that, having regard to all the evidence, no jury could reasonably have reached to the standard of beyond reasonable doubt (see above at [28]). The Court must always, however, keep in mind that it is not the arbiter of guilt, and that reasonable minds might disagree on findings of fact – see the comments in Biniaris set out at [29] and those in Mareo set out above at [11].
[88] While we have rejected the English “lurking doubt” approach, we consider, like the Canadian Supreme Court, that that concept provides a useful trigger for a fuller review. A lurking doubt or uneasiness experienced by the appellate court may be an important indication that the verdict was not reasonable or unsupportable on the evidence. However, by itself a “lurking doubt” is not sufficient grounds on which an appeal court should deem a conviction to be unsafe. The law in New Zealand has always required an appellate court to recognise that reasonable minds might disagree on findings of fact and that the jury, not the appellate court, is the ultimate arbiter of fact. It is only where a jury’s verdict is unreasonable on all the evidence (in the sense described above at [86] [87]) that an appeal court may properly differ from it.
[89] Finally, we note that an appellant must be able to point to a sufficient foundation for his or her submission that a ground of appeal under s 385(1)(a) exists before the Court is required to embark on the exercise of reviewing all of the evidence. Where the Crown case against an appellant is strong or indeed overwhelming, it will be very difficult to point to a foundation for an appeal under s 385(1)(a). Further, in performing any review of the evidence, regard must also be had to the existence of the jury verdict. The appellate court does not approach the review of the evidence de novo, but rather factors the jury verdict into its decision. It remains for the appellant to make out this ground of appeal: see our discussion of Weiss above at [39] and the comments in Ross, set out at [12].
[90] We also note that, even where a sufficient evidential foundation has been raised, in many cases this ground of appeal can and will be dealt with quite shortly. The contention that a verdict is unreasonable or that it cannot be supported with regard to the evidence will often rest on particular points in the evidence. A judgment in such a case is likely to concentrate on those points rather than review the evidence not in contention. Further, where the Court is of the view that, despite any gaps or inconsistencies in the Crown case, a reasonable jury was entitled to reject contrary evidence and to consider guilt proved beyond reasonable doubt, the case may well not lend itself to much elaboration – see the comments in Ramage referred to above at [9].
Can Mr Munro’s conviction stand when the correct test is applied?
The issues
[91] As indicated above at [2], Mr Munro was convicted of causing death while driving with excess blood alcohol after being involved in a motor accident on State Highway One between about 9.40pm and 9.48pm on 17 November 2004. Mr Munro had been working in Palmerston North that day and was driving south on State Highway One to Waikanae in a white Isuzu truck, which belonged to his employer. Ms Aldridge was travelling north from Wellington in a black Honda car to meet her boyfriend at Ohau. Between Manakau and Otaki, State Highway One crosses the railroad tracks at what is known as the Pukehou Overbridge. The collision between the Isuzu and the Honda occurred just south of the overbridge.
[92] As a result of the collision, the Isuzu spun about 180 degrees and ended up in the middle of the road on its side with its front pointing north. The rear of the Isuzu was in the north bound lane but the cab area was over the centre line in the south bound lane. The Honda also spun nearly 180 degrees and ended up over the fog line in the south bound lane, pointing in a southerly direction. The pattern of damage to the vehicles indicated that the Isuzu had been hit square on, while the Honda had been hit on more of an angle, towards the front passenger side of the vehicle.
[93] The Crown asserted at trial that the collision was caused by Mr Munro’s Isuzu truck crossing the centre line into the path of the Honda and had occurred in the north bound lane as the Isuzu was “tracking back” into the south bound lane. The Crown relied on the analysis of the physical evidence by its crash experts and in particular their analysis of certain marks on the surface of the road. It also called evidence that Mr Munro was intoxicated and evidence, from other drivers on the road that night, as to Mr Munro’s erratic driving shortly before the accident.
[94] The defence case was that the collision had been caused by the Honda crossing the centre line into the Isuzu’s path or, at least, that the Crown had not eliminated the reasonable possibility that this was the case. The physical evidence was, at best, inconclusive as to the exact point of the crash. Mr Munro relied on the evidence of his independent expert and that of the sole eyewitness, Ms Clisby. He also relied on evidence that suggested that the victim’s driving might have been impaired at the relevant time.
[95] The issue for this appeal is whether the jury’s guilty verdict was unreasonable or unable to be supported having regard to the evidence – see s 385(1)(a) of the Crimes Act. Mr Stevenson, for Mr Munro, accepted that there was some evidence which, if accepted by the jury, was capable of sustaining a guilty verdict. However, he submitted that, when regard is had to the whole of the evidence, a jury acting reasonably ought nevertheless to have had a reasonable doubt of Mr Munro’s guilt.
[96] Mr Stevenson was able to point to a sufficient basis for his submission (see above at [89]). His challenge to the conviction was also based on an assessment of the evidence as a whole and not merely on certain parts of the evidence. We thus propose to examine the whole of the evidence at trial and deal with the appeal under the following headings:
(a) The eyewitness and related evidence;
(b) The expert evidence;
(c) Deficiencies in the investigation;
(d) Assessment of the expert evidence;
(e) Assessment of the remaining Crown evidence;
(f) Aspects of the trial process; and
(g) Whether the jury ought to have had a reasonable doubt.
Eyewitness and related evidence
[97] There was one eyewitness to the collision, Ms Clisby. Her evidence is obviously important in assessing the causes of that collision. There was also evidence suggesting the possibility that the victim’s driving might have been impaired at the time of the collision through cannabis consumption and cellphone use.
[98] We first provide a summary of this evidence and then assess its significance.
EYEWITNESS EVIDENCE
[99] The eyewitness, Ms Clisby, was travelling south, from Manakau to Otaki. It was a drive that usually took about seven minutes. She had left Manakau at about 9.40pm. When she came off the Pukehou overbridge (about half way through her journey), she noticed the tail lights of a car ahead of her travelling in the same direction as she was. She could not really make out what sort of car it was, but it looked dark and sporty. Just as she was catching up to that car, she noticed a “white type ute, cab” coming towards them. Then what she started calling “the white car” turned straight in front of the dark coloured car as if it was trying to turn into the side road opposite. She was positive that the white vehicle had its lights on and that she did not see it indicate.
[100] The vehicles then collided in the lane the dark car was travelling in. She confirmed in cross-examination that she saw the brake lights of the car in front of her for a split second before the collision and that she did not have long to observe the cars before the collision occurred. After the collision, both cars went sliding downhill (ie southwards) and the black car ended up facing towards the south, towards Otaki. The white car was facing towards the north, towards Levin, and was a little bit over the centre line.
[101] Ms Clisby had managed to swerve into the side road, which ran off to her left, to avoid the collision. She said that parts of the black car were underneath her car so she must have hit them while she was swerving. Her car had stalled under heavy braking and she then got out and walked up to the dark car. Her explanation for going to help the driver of the Honda was as follows:
Because all I could think was that I hated that person in the white car, I was, I couldn’t believe they had just pulled out, they had just cut someone off, and just done that, that’s why I only, I mean it’s horrible to say that but I only went to help the person in the black car.
[102] It was put to Ms Clisby in examination-in-chief that there was going to be evidence in the trial that the black car was driving north and the white car was driving south. She maintained her evidence that she was following the black car and that the white car was coming towards her. In re-examination, Ms Clisby reiterated that the black car did not deviate from its own lane and that it was the white car that turned across the black car into the incorrect lane.
THE VICTIM’S CANNABIS CONSUMPTION
[103] Blood and urine samples were taken during the post mortem examination of the victim, Ms Aldridge. These were analysed by Dr Poulson, a forensic scientist employed at the Institute of Environmental Science and Research Limited (ESR). She confirmed there was no alcohol in either the blood or urine of Ms Aldridge but said that there were approximately 3 micrograms per litre of tetrahydrocannabinol (THC) in her blood. The THC level in Ms Aldridge’s blood was consistent with smoking a single cannabis cigarette but it could have been more. Dr Poulson’s evidence was that the greatest effect of smoking cannabis is around twenty to thirty minutes after consumption.
[104] The usual effects of cannabis smoking are euphoria and relaxation with some impairment of perception and some loss of motor co-ordination. However, the effects are unique to particular individuals and the particular environment and so it is difficult to assess the effects on a particular individual at any particular time. Moreover, the level of THC in the blood is not a good indicator of the level of cannabis consumption. The most that Dr Poulson could say was that it was possible that Ms Aldridge was affected by the drug at the time of her death but she could not say how she would have been affected.
[105] Dr Poulson agreed in cross-examination that the dangers of driving after using cannabis included taking longer to respond to events, the reduced ability to think clearly and the reduced ability to pay attention. She also agreed that these effects have been commonly observed in studies testing people’s skills after taking cannabis.
EVIDENCE OF CELLPHONE USE
[106] As officer in charge of the investigation of the collision, Detective Constable Townsend obtained Ms Aldridge’s cell phone records. These revealed that, on the evening of the collision, between 8pm (which was probably about the time she left Wellington) and 9.42pm, Ms Aldridge had sent five text messages and received eight. The earlier messages had been from her boyfriend and the pattern of exchanges suggested that she was replying to his text messages. The last text message was at 9.42pm and came from a number other than Mr Wade’s.
[107] The constable also confirmed in evidence that Ms Aldridge was driving on a learner’s licence and that a member of the public had first reported the accident at 9.48pm.
ASSESSMENT OF EYEWITNESS AND RELATED EVIDENCE
[108] Mr La Hood, who argued this part of the case for the Crown, submitted that Ms Clisby’s evidence was essentially neutral and did not assist either party to any significant degree. This is because Ms Clisby was clearly mistaken about the travelling directions of the two vehicles – see at [102]. The Crown said, however, that it is equally clear that, immediately after the accident, Ms Clisby believed that the car that had crossed the centre line was the white truck rather than the black car – see at [99] and [102].
[109] We are unable to accept the submission that Ms Clisby’s evidence was neutral. In our view, Ms Clisby’s mistake as to the travelling direction of the vehicles was understandable and does not necessarily affect the integrity of the rest of her account. It was dark and she only had a short time before the collision to observe the vehicles in front of her. According to Ms Cummins, a Crown witness (see at [211] below), the level of the lights of the Isuzu would not have alerted Ms Clisby to the fact that she was following a truck – see [212] below. Ms Clisby had not, in any event, been following the Isuzu truck for any length of time before the accident. She had, it appears from her evidence, come up behind it as she was coming off the overbridge on the gentle right-hand turn. Ms Clisby must have been travelling slightly faster than the Isuzu truck and, because of the curvature of the road at the overbridge, would not have been able to see it before she came off the bridge and out of the bend.
[110] Further, the position in which the vehicles landed up after the crash could well have caused Ms Clisby’s mistake as to travelling directions. As indicated above at [92], both cars spun 180 degrees and thus were left facing in the opposite direction from the direction in which they had been travelling. In addition, the Honda landed up right over the fog line in the south bound lane and the Isuzu partly in the north bound lane, thus furthering the illusion that they had been travelling in those lanes. Ms Clisby also had debris from the Honda underneath her car after the collision. Her assumption, given the circumstances and her understandably dazed state after the accident (see at [101] above), that she had been following the Honda rather than the Isuzu was perfectly natural.
[111] In our view, there must be a reasonable possibility that Ms Clisby was wrong about the direction of the cars but that everything else she said was correct. Ms Clisby’s evidence was that the vehicle travelling in the opposite direction to her car that had crossed into the path of the vehicle travelling in the same direction that she was. She was adamant that the vehicle travelling in front of her had not deviated from its lane and that the collision had occurred in front of her in the lane in which she was travelling. That this was the case is backed up by her evidence that she had to take evasive action to avoid the collision and that she hit debris from the crash as she was doing so.
[112] Ms Clisby’s evidence as to the way the collision occurred is consistent with much of the physical evidence. Her impression was that the north bound vehicle (ie the Honda) had turned sharply into the south bound lane, across the path of the south bound vehicle (ie the Isuzu), to turn into a side road. There was a side road to Ms Clisby’s left, into which she assumed that the north bound vehicle was trying to turn. There was no side road on the north bound side. The point of impact was straight on for the Isuzu truck but the main force of impact occurred towards the passenger side of the Honda (see [92] above). The pattern of damage to the vehicles is thus consistent with Ms Clisby’s evidence that the south bound vehicle (ie the Isuzu) did not deviate from its own lane and that the north bound vehicle (ie the Honda) turned into its path. Ms Clisby’s description of seeing the brake lights of the car in front of her is also consistent with the defence expert’s theory about a set of tyre marks in the south bound lane – see below at [158]. She cannot have seen the brake lights of the Honda on any analysis as it was travelling in the opposite direction to her.
[113] In addition, there was evidence that raised at least the reasonable possibility that the victim’s driving may have been impaired, which makes it more likely that she may have crossed over to the south bound lane (as Ms Clisby described the path of north bound car) than if she had been an experienced sober driver with her full attention on the road. Ms Aldridge was driving at night on a learner’s licence. It was reasonably possible that she was receiving a text just before the collision occurred. The pattern of earlier texts suggest that she was in the habit of reading and answering texts (see at [106] above) and it must therefore be reasonably possible she was either reading or preparing to reply to the 9.42pm text at the time of the accident. Finally, her reactions and driving co-ordination may well have been impaired through the use of cannabis see at [105] above.
[114] In order for Ms Clisby’s account to be consistent with the Crown theory of the accident, it would need to be assumed that she saw the brake lights of the Isuzu (ie the south bound vehicle) as it collided with the Honda but that she had been mistaken as to the lane the respective vehicles were in at the time and thus also wrong that the south bound vehicle never deviated from its lane. The evidence of the investigating Crown expert, Senior Constable Maddaford, was that tyre marks in the north bound lane could have resulted from a momentary application of brakes by the Isuzu at the time of the collision – see [127] below. Given the constable’s theory that the Isuzu was “tracking back” into the south bound lane at the time of the collision, however, the brake lights would have been at least partly angled away from her and she made no mention of that.
[115] It would also be necessary to assume that, when Ms Clisby saw the Honda (ie the north bound vehicle) turning into the path of the Isuzu, it had been taking evasive action by turning to the right across the path of the Isuzu – see at [132]. This presupposes, however, that her view of the Honda was not obscured by the Isuzu. One might also have thought it more natural for the Honda to have taken evasive action by turning to the left away from the Isuzu and onto the shoulder. This is particularly the case as the Isuzu, on the Crown theory, was “tracking back” to the south bound lane and thus creating, presumably, more space at the left of the north bound lane. We note that there appears, from the photographs, to be a wide grass verge to the left of the north bound lane. Further, it would be necessary to assume that Ms Clisby was mistaken as to the vehicles being in separate lanes until just before the collision and that she was also wrong about the initial collision occurring in front of her in the lane in which she was travelling. That might seem an unlikely mistake to make.
[116] We are mindful of Sergeant Stables evidence as to the fallibility of eyewitnesses – see at [146] below. However, as Ms Clisby’s account accorded with the physical evidence in a number of key ways, we do not think that the jury would have been justified in rejecting her evidence in its entirety. They would therefore not have been justified in deciding that there was no reasonable possibility that the key parts of her account were correct, despite her clearly dazed state after the accident and her mistake as to the travelling directions of the vehicles.
[117] While it cannot be totally ruled out that Ms Clisby’s account might be consistent with the Crown theory of the accident, it is, for the reasons set out above, more likely that her account supports the defence theory of the collision. In any event, it does not appear to have been part of the Crown case at trial or on appeal that Ms Clisby’s account was consistent with Senior Constable Maddaford’s reconstruction of the collision. Rather, the Crown contended that the crucial part of Ms Clisby’s evidence was that after the collision she had gone immediately to assist the driver of the black car, because Ms Clisby was “holding the driver of the white van responsible” for the crash.
Expert evidence
[118] There were two experts called for the Crown – Senior Constable Maddaford, who had been the investigating expert at the scene, and Sergeant Stables, who was the reviewing officer. The defence called one expert, Mr Beckett. We summarise the evidence given by those experts first and deal with some deficiencies in the investigation. We then assess the significance of the expert evidence, including the significance of concessions made by the Crown experts.
[119] Before summarising the evidence, it is necessary to identify four pieces of physical evidence about which there is disagreement between the experts. There are two relevant sets of tyre marks. The first set is in the south bound lane just after the overbridge. On the left hand side (looking south) is a single tyre mark and on the right hand side (apparently almost on the centreline) a double tyre mark from a dual wheel combination (marks ‘A’ and ‘B’ on the diagram at [120] below). The second set of tyre marks is further towards the south. These are two marks which are largely in the north bound lane but they point towards the south bound lane (marks ‘C’ and ‘D’ on the diagram below). Just after the second set of tyre marks is a gouge in the road, near the centreline but just in the south bound lane. There is also what the Crown expert described as a small scrape mark that ran away from the gouge in the direction of the resting position of the Honda. This is not visible on any of the photographs taken at the scene but Senior Constable Maddaford had marked it up on a plan of the scene that he had prepared, which was in evidence (see diagram at [120] below). Mr Beckett’s evidence, however, was that the “scrape mark” was in fact two gouge marks, although nothing really turns on the mark (see at [167] below).
[120] In the appendix to this judgment, we provide, for ease of reference, a table setting out the points of difference between the experts on the main pieces of physical evidence and on aspects of the collision. We also set out below a diagram of the scene, showing the resting positions of the vehicles and the physical evidence. It is based on one of the diagrams placed in evidence by Senior Constable Maddaford but we have corrected the positioning of the gouge mark and added the first set of tyre marks. As there were no measurements of the first set of tyre marks or of their position relative to the second set of tyre marks, their portrayal on this diagram is illustrative only. We have also added the lane to the left but its positioning is approximate only. It can be seen on some of the photographs but its exact position relative to the other physical evidence is unclear. Like the constable’s
diagrams, there are difficulties with scale generally in this diagram.
[121] For completeness, we note that there was evidence from a qualified automotive engineer that he had examined both the Isuzu and the Honda and found no mechanical faults that could have contributed to the accident. It was also common ground that there were no environmental factors, such as an oil spill or frost, that could have contributed to the accident.
[122] We also note that both Senior Constable Maddaford and Mr Beckett were asked about what is sometimes called the “snaking effect”. This occurs where one vehicle veers out of its lane. The vehicle that it is going towards it then takes evasive action, which means it in turn veers out of its lane. The vehicle, which originally moved out of its lane, then corrects. Both experts confirmed that, because of the snaking effect, although an accident occurs in a particular lane, this is not conclusive as to which vehicle veered out of its lane to start with and thus the location of the collision may not be determinative of which driver is at fault. Senior Constable Maddaford was not directly asked if there was any evidence of snaking in this case. Mr Beckett confirmed in cross-examination that there was no such evidence. The jury was thus entitled to consider that the snaking effect had been ruled out in this case.
SENIOR CONSTABLE MADDAFORD
[123] Senior Constable Maddaford is attached to the Central Districts Serious Crash Unit. He arrived at the accident scene while the crashed vehicles were still on the scene in their post-crash positions, although the ambulance had already taken Mr Munro away to hospital. The constable remained at the scene for some hours and directed a number of photographs to be taken. In the discussion below we have first summarised all his evidence-in-chief before turning to the cross-examination.
[124] The constable’s evidence was that, in his examination of the scene, he noted three things of significance – the gouge, the second set of tyre marks and the scrape mark. The gouge mark, which he had located near the centreline of the highway, was of significance because it was fresh. His evidence was that such a mark gives a reasonable indication of the area of impact where an object has been forced into the seal. This is likely to occur at or near the point of maximum engagement (ie the point when the forces between the vehicles are equal). This was the only gouge mark that he found.
[125] Senior Constable Maddaford considered that the gouge mark was made by the undercarriage of the Honda on the driver’s side. He portrayed the gouge mark under the Honda in this position in a computer generated diagram, which was put in evidence. This diagram was done using a computer aided design (“CAD”) package which aligned the panel damage on the vehicles to show the position of the vehicles at the point of maximum impact. We reproduce a pictorial representation of that diagram below.
[126] Also of significance for the constable was the second set of tyre marks (marks ‘C’ and ‘D’ in the diagram at [120] above). These were short (a maximum of one metre) and not quite parallel to each other. The constable’s evidence was that they contained striation lines, which indicate a locked wheel. He could not tell the direction in which the locked wheels were travelling. However, his opinion was that, taking into account the rest positions of the vehicles and his analysis of the hypothetical line and points of impact, the second set of tyre marks were made by the Isuzu and not the Honda. He later measured the wheel track (ie the width of one wheel across to the other) of the Isuzu and confirmed that the second set of tyre marks matched that wheel track.
[127] When asked if there was any evidence of pre-impact braking by either vehicle, he at first said that there was none. He then said that the second tyre marks could have resulted from a momentary application of brakes by the driver of the Isuzu just before impact or could have resulted from the locking of the wheels because of the collision. Asked whether there were any other tyre marks at the scene, he said “Not that I could find, sir, no”.
[128] The constable was of the opinion that the Isuzu had tipped onto its side when it came to a halt at the end of the slide following the collision. He rejected the proposition that it had fallen on its side before starting to slide.
[129] In the constable’s view, the two vehicles had pivoted around each other after the point of maximum impact. The Honda had spun clockwise (ie as the driver sits) and the Isuzu had rotated anti-clockwise. The Honda had come to rest in the south bound lane because, at impact, the Isuzu had “swept the Honda aside” and pushed it over to the other side of the road. As the Honda was lighter than the Isuzu it travelled further from the point of impact than the Isuzu and ended up resting in the south bound lane.
[130] The constable confirmed that there was no physical evidence to indicate how the vehicles rotated and slid to their final resting places after the collision, apart from the small scrape mark but that was insufficient in length to draw any firm conclusions. He said, however, that the small scrape mark tracked towards the rest position of the Honda and was thus attributable to the Honda. That supported his theory that the direction of rotation was clockwise for the Honda.
[131] Senior Constable Maddaford examined the vehicles on site and also in the depot in order to establish the angle of engagement, being “the principal direction of force through that vehicle”. The type of damage to the Honda (mainly to the front passenger side) indicated that it was hit at an angle. He said that this might suggest that it had been trying to get out of the way. The damage profile for the Isuzu was pretty well across the entire front area of the truck almost in an even manner, which indicates a more square-on impact to that vehicle.
[132] Taking into account all of the above factors, including the position of the gouge mark (right on the centre line), the tracking of the second tyre marks, the line of impact, the scrape mark, and the resting position of the vehicles, the constable was of the opinion that, at the point of impact, the whole of the Isuzu truck was on the wrong side of the road – ie in the north bound lane. He said that there was nothing, either at the scene or by analysis or by examination of the two vehicles, to suggest anything to lend support to a supposition that the Honda was in the wrong lane at the time of impact.
[133] In cross-examination, the constable confirmed that the side road would have been on the left side of the Isuzu travelling south and on the right of the Honda travelling north. He was asked by defence counsel to rotate his diagram of the reconstruction of the collision (set out at [125] above) slightly clockwise so that the Isuzu at the maximum point of engagement was straight on in its own lane. This left the Honda at a more extreme angle heading towards its right. We produce a pictorial representation of that below. This is for illustrative purposes only as the defence did not put any such diagram in evidence and the defence expert’s opinion was that the angle of engagement was less than that postulated by the constable.
[134] It was then put to the constable that the positions of the vehicles as set out above might be consistent with the Honda heading on an extreme angle to its right across the road as if it was going into the side road on the left of the Isuzu. The constable said that this alternative scenario would not fit in with the physical evidence that he saw on the surface of the road at the collision scene but agreed that it was consistent with the damage to the vehicles.
[135] It was put to Senior Constable Maddaford that the final resting places of the vehicles and the fact that the debris from the accident was in the south bound lane were consistent with the collision having occurred in the south bound lane. He disagreed with that proposition on the basis that the physical evidence confirmed that the impact had occurred in the north bound lane.
[136] The constable was also asked whether, if the collision had occurred in the north bound lane, he would have expected the Honda to rotate anti-clockwise around its centre of mass and remain in the north bound lane, rather than spinning clockwise and landing in the south bound lane. He said that that the direction of rotation was a consequence of the angle of approach which in this case made the Honda spin clockwise and the Isuzu anti-clockwise.
[137] Senior Constable Maddaford was also referred in cross-examination to the first set of tyre marks just after the overbridge. In his view, these had nothing to do with the accident. He said that they were not friction marks made by wheels in the locked or brake position. He expressed the view that the first tyre marks had been left by emergency service vehicles, which had been moved out of the way to allow the photographer to do his job. He confirmed, however, that he had not physically seen another vehicle make those marks. Nor had he surveyed the marks. He also confirmed that the tyre marks on the right were from dual wheels and that the Isuzu has that combination of wheels on its rear axle. In re-examination, he had ruled out the possibility that the Isuzu had made the first set of tyre marks. He answered:
I marked all the skid marks that I found at the scene and if they were sir they would be marked.
[138] In cross-examination, the constable also said that the only tyre friction or skid marks made by locked and sliding wheels that he observed on the night were the second set of tyre marks discussed above (marks ‘C’ and ‘D’ on the diagram above at [120]). He was asked whether these marks could be consistent with the sort of marks a vehicle leaves when it has tipped on its side and is rotating anti-clockwise around the centre of its mass with the back wheels briefly making contact with the road surface. He answered that he believed that the second set of tyre marks were from the front wheels of the Isuzu and not from a dual wheel assembly (ie they were not made by the rear wheels of the Isuzu).
[139] Asked if his conclusion that the impact occurred in the Honda’s lane was based primarily on the striation marks contained in the second tyre marks, he said that this was partly so but that you had to look at the entire picture. You could not focus just on an individual item as that narrows down the inquiry. He said:
- No sir, you would have to look at the entire picture and you can’t focus just on an individual item, because that narrows down your inquiry. You have to look at post-crash rest positions.
Q. Yeah?
SERGEANT STABLES
[140] Sergeant Stables of the Wellington District Serious Crash Unit was called as a further witness for the Crown. He had reviewed the report of and conclusions drawn by Senior Constable Maddaford. He confirmed that the report followed the format used by both the Central and Wellington Police Districts. He also confirmed that, having read the report he concurred with the findings made in it. He was, however, unfortunately somewhat handicapped in giving his evidence as he did not have a copy of his report in front of him and had last reviewed that report some two and a half weeks before giving evidence.
[141] The sergeant had not made a scene visit at the time of his review or at any time before giving evidence. His review was thus limited to the information and conclusions contained in the report, which the sergeant felt “made sense and followed logic”. The sergeant did not look at any primary evidence outside the report. Nor did he attempt to draw his own conclusions on the evidence before reading Senior Constable Maddaford’s analysis.
[142] Sergeant Stables said that the gouging identified in the report indicated an area of impact and probably the area of maximum engagement. He said that it is usually a significant item of the sub-train of the vehicle which causes the gouging, something quite hard and solid like the bottom of the engine or the suspension towers. Before the point of maximum engagement the vehicles will already have crumpled together so that, at the point of maximum engagement, the vehicles are forced downwards into the roadway. After that point which lasts a split second, there is an equal and opposite force and the vehicles are repelled from one another.
[143] The sergeant confirmed that he had, “as far as [he] could with the report”, reviewed the direction of force and the points of contact between the two vehicles as found by Senior Constable Maddaford. He said that the fact that the damage to the vehicles was not square on but more angular could indicate an evasive manoeuvre by one or both of the vehicles. He could not go further than that.
[144] Asked whether the second tyre marks indicated a direction of travel, he answered that a skid mark on the road surface allows you to place the tyre that caused it back onto the mark. It can thus be used to track movement. However, that is best done where there are at least two and preferably three tyre marks to ensure that you orientate the vehicle correctly. He said that when the tandem back wheels of a truck are locked up they would most often leave dual tracks. He did not recall anything in this case to suggest that the second set of tyre marks (marks ‘C’ and ‘D’ on the diagram above at [120]) were left by the dual back wheels of the Isuzu.
[145] He was asked whether there was anything in his review of the material to suggest that the reconstruction in the Senior Constable’s diagram (above at [125]) was incorrect. Based on the alignment of the vehicles, their damage profile and what was reported by Senior Constable Maddaford, he said that
I felt that it [the diagram] made sense, it followed logic and agreed with what the profiles of the cars allowed for the damage to have happened as.
[146] Sergeant Stables was asked whether there was anything in the account of the eyewitness which would alter his view. He replied that there was not. Investigators are trained to focus on the physical evidence. Witness testimony is considered but it has to be balanced against the physical evidence. He repeated the old saying that physical evidence does not lie but people do (or at least people can get things wrong).
[147] In cross-examination, the sergeant was asked about the first set of tyre marks (marks ‘A’ and ‘B’ in the diagram above at [119]). He confirmed that they were tyre marks but said that, if these were locked and sliding tyre friction marks, he would expect to see striations within the mark. There was not enough detail in the relevant photograph to make out the characteristics within those marks.
[148] In relation to the second set of tyre marks, he confirmed that there were lines running within them which appeared to be striation marks. This would indicate that it was a locked tyre that had caused the mark. He was asked if the marks could have been made at some time during the movement of a vehicle as it tipped and where the back wheels, either in a sliding motion or during the process of turning, might make brief contact with the road surface. He replied “They could be made that way, yes”. He agreed that the second tyre marks were very short.
[149] The sergeant was asked whether the debris being in the south bound lane was consistent with the impact occurring in the south bound lane. He replied that the debris field is not conclusive but it can give some assistance. He, however, agreed that the debris being in the south bound lane was arguably consistent with the impact having occurred in the south bound lane.
[150] Although the evidence was somewhat confused, Sergeant Stables appears also to have agreed that, if the impact had been in the north bound lane, he would have expected the Honda to have rotated anti-clockwise. The concession was made, however, in the course of a discussion about hypothetical vehicles rather than specifically related to this case.
MR BECKETT
[151] The defence expert, Mr Beckett, was formerly the officer in charge of the Auckland Police District Serious Crash Unit. This was the first formal serious crash unit in the country and he was involved in setting it up. Since his retirement from the police, he has been operating as a private independent road crash investigator.
[152] Mr Beckett was asked first about the first set of tyre marks. In his view the mark on the right was a skid mark from a rear axle tandem wheel, the same formation as the Isuzu back wheels. Asked to comment on Senior Constable Maddaford’s view that the first set of tyre marks had been caused by emergency vehicles, he indicated that there were no marks before or after those marks and thus nothing to show that they had been left by a rolling wheel. He confirmed that they did not run continuously to where the Isuzu was situated.
[153] Turning to the second set of tyre marks, he said that normal practice with such marks at an accident scene would be to photograph them. He said that he had not seen any clear photographs of those marks. The length and the distance between the marks would also be measured and the exact place they were found mapped. Only if that were done would it be possible to prove conclusively that they were marks from the front wheels of the Isuzu and that the vehicle was on the incorrect side of the road at that point. He noted that Constable Maddaford said that there had been measurements done and that the tyre marks fitted the wheel track of the Isuzu but there were no measurements recorded in Constable Maddaford’s report.
[154] In Mr Beckett’s opinion, if the collision had occurred in the south bound lane, it was entirely possible that the second set of tyre marks had been caused by the Isuzu but not by the front wheels. In his view, the tyre mark closest to the centre of the road (mark ‘C’ on the diagram above at [120]) could have been caused by the right hand steer axle. The tyre mark in the middle of the north bound lane (mark ‘D’) could have been caused by the rear tyre of the Isuzu as the truck rotated after the collision. He said that for two reasons. The first was that the tyre marks were not uniform as he would have expected if they had been left by the front wheels: one mark started well before the other and they were not straight. Secondly, he noted that the tyre mark closest to the centre of the north bound lane (mark ‘D’) had a “crick” in it which might be left where a vehicle was rotating and its tyre was in contact with the road.
[155] As to the gouge mark, Mr Beckett said that the vehicles should have been placed on hoists and their undercarriages inspected to see which vehicle, and what part of the undercarriage of that vehicle, had caused the gouge. In his opinion, to say that the gouge came from the Honda was “pure guesswork”. In cross-examination he refused to speculate and did not accept that the gouge mark was more likely to have been made by the Honda. The most Mr Beckett could say was that the gouge mark was probably indicative of the “loose area” of the point of impact.
[156] Moving to a discussion of the debris from the accident, Mr Beckett said that debris will travel in a direction that it has been accelerated in. As there was no debris in the north bound lane, in his opinion the impact must have taken place either in the south bound lane or very close to the centreline. He was incorporating into that opinion the resting position of the Honda vehicle right over the fog line of the south bound lane and the resting position of the Isuzu about two-thirds into the south bound lane. If the collision had occurred in the north bound lane, he would have expected the vehicles to have ended up in the north bound lane. Wherever the impact occurred he would have expected the Honda to rotate anti-clockwise.
[157] In Mr Beckett’s professional opinion, taking all of the above matters into account, it was not safe to conclude that the collision had occurred in the north bound lane.
[158] In cross-examination, Mr Beckett confirmed that he had not seen the first set of tyre marks on the road in his inspection of the site the day before but was working off the photographs. He said that he had deduced from the resting place of the Isuzu that it was possible that the marks near the centre line (mark ‘B’ on the diagram above at [120]) were from the right hand rear dual wheels of the Isuzu in a locked or braking position. His evidence was that he could see striations in the photograph of that mark. He maintained his evidence on those tyre marks, despite being challenged as to why only one side of the vehicle should have made those marks.
[159] Mr Beckett agreed that, although there was some distance from the first set of marks to the point of collision without further brake marks, it would only have taken about half a second to cover the distance (1520 metres) and so the lack of further brake marks was not significant. Asked whether it was his position that the Isuzu, by reference to the point of maximum impact, was at all times in the south bound lane, Mr Beckett said that he could not say that. This was because the double tyre marks (showing the braking or locked right rear wheels of the Isuzu - mark ‘B’ on the diagram above at [120]) were heading towards the centre line. The Isuzu therefore may at some stage have crossed the centre line, returning to the south bound lane after impact. It would appear, therefore, Mr Beckett was saying that, if the Isuzu did move into the north bound lane, this occurred only after the Isuzu had braked just before the collision. This portion of the cross-examination is, however, not very clear and later Mr Beckett says that any movement into the north bound lane was after impact.
[160] Mr Beckett also said that the mark on the left looking south (mark ‘A’ on the diagram above at [120], and see above at [137] discussing Senior Constable Maddaford’s testimony) could be from the left front wheel of the Isuzu. He was unable to assess the length or the marks from the photographs or their distance from the Isuzu and no measurements had been taken. He also confirmed that he could not exclude the possibility that the marks were on the road before the accident.
[161] Turning to the second set of tyre marks, Mr Beckett, in cross-examination, said that the tyre mark in the middle of the north bound lane (mark ‘D’ on the diagram above at [120]) began very narrowly and widened out before it stopped. In his view this was typical of a “scrub” tyre mark from the rollover of a truck. He pointed out that one of the photographs showed how that scrub mark lined up with the right rear tyres of the Isuzu. He did not consider the mark to be consistent with a tyre friction mark from a single axle locked up and leaving striations on the road surface as postulated by Senior Constable Maddaford.
[162] Mr Beckett was asked about Senior Constable Maddaford’s supposition that the Isuzu had rotated while still on its wheels and then fallen when it reached its resting point. Mr Beckett said that there was no evidence to support this theory. In his view, if the truck had rotated on its wheels, it would have left yaw marks and pressure marks on the road and it would have probably ended up somewhere over in the side road or in a paddock while it was on its wheels before it could fall onto its side. The “scrub” tyre mark (mark ‘D’) was in the north bound lane because at that stage the vehicle had tipped over on its side after the collision and was rotating and sliding to rest. He said that he would not have expected to see any evidence of scrape marks left by the body of the truck as it slid on its side as the weight of the truck is spread over a wide area and thus has a lower coefficient of friction which would mean that any scrape marks would be very light.
[163] Mr Beckett agreed that the gouge mark was a classic sign for establishing where impact occurred. He said that the point of first impact was likely to have been somewhere before the point where the gouge was located on the road. This is because some time is needed for the vehicles to change shape after impact and for whatever made those marks (probably the undercarriage of one of the vehicles) to come into contact with the road. He refused to speculate on which vehicle made the mark and did not agree that it was more likely to have been the Honda than the Isuzu.
[164] Mr Beckett was asked if he agreed with the portrayal of the alignment of the vehicles at the time of maximum engagement as shown in diagram (set out at [125] above). Mr Beckett said that he would have had the rear of the Honda spun anti-clockwise a little so that the vehicles were straighter than shown in the diagram. He said that he would normally have expected the vehicles to have been physically put back together in a workshop and photographed from a height of about five metres rather than “playing around with it on a computer”. This is so that you could line up the panel damage and better identify the parts of the vehicles that had actually collided with one another.
[165] Mr Beckett also said that the gouge mark had not been accurately portrayed. From his observations at the scene and from the photographs, he considered the gouge mark to have been in the south bound lane by some 50 millimetres or so. He admitted that he had not measured the gouge or its position when he visited the scene the day before. (The Judge at this point drew attention to one of the photographs which showed the gouge to the right of the cat’s eye and just inside the south bound lane).
[166] Mr Beckett was asked how he accounted for the second set of tyre marks being on the road before the gouge and thus before the point of maximum engagement. He replied that the truck was driving south bound. The driver saw a car in his lane and braked but was not able to avoid the impact. Impact occurred and the Honda rotated to the south bound lane. The truck rolled on its side because that was the way it got hit and that is the way it fell. It then rotated counter-clockwise to the position that it was lying in. The second set of tyre marks could have been caused when the truck was on its side.
[167] As to the scrape mark, Mr Beckett confirmed that he had found two deep gouge marks in his examination of the scene the day before. They did not track right up to the Honda and were only about 150 to 200 millimetres in length. In his view these scrape marks were not investigated properly and did not assist because it was not possible to ascertain what part of which vehicle had left them. In particular, Mr Beckett noted that they may have been pre-existing on the road surface and thus did not necessarily assist either party. However, Mr Beckett said that, if the scrape marks had been investigated properly, they might have been valuable evidence.
Deficiencies in the investigation
[168] We now highlight a number of deficiencies in the investigation which potentially affect the reliability of the evidence of Senior Constable Maddaford.
[169] The first deficiency is the lack of notes taken at the scene. Senior Constable Maddaford was asked whether, as he conducted his investigation at the scene, he made notes of what he found and the position of matters that were relevant to his inquiry. His answer was, “I completed an analysis report”. He was then asked whether that report was based on earlier notes made and he answered “and observations at the scene too”.
[170] Later, on the subject of notes, he said that he had marked all the evidence that he saw at the scene and directed photographs. He may have made a rough sketch of the scene at the time but, if he did, it was no longer available. He said:
Q. You didn’t make any notes at the time you attended at the scene?
A. When I attended the scene, sir, I marked the evidence I saw.
Q. Direct photographs?
A. And direct photographs.
Q. And work from there?
A. And work from there sir.
Q. So just to confirm you didn’t make any notes at the time per se?
A. Sorry sir I don’t.
[171] Mr Beckett confirmed in re-examination that he had received disclosure of the police file, including a copy of the report by Senior Constable Maddaford, but that he had not received any notes made by Constable Maddaford at the scene. Mr Beckett said that it had been his practice while in the police to have his staff use dictaphones to record notes while on the scene so that more detail could be captured. We would have thought that this should be standard procedure in any investigation. Relying on memory alone, particularly without adequate photographs or measurements, seems inherently unsafe. We also note that Senior Constable Maddaford was at the scene for some hours (see above at [123]) and so he had not been rushed. It seems to us elementary that some kind of detailed drawing would be completed at the scene (to at least a rough scale) which would include any debris, the probable point of impact, the points of rest, skid marks (including length and width), any roadbed damage, and details of the road itself, the shoulder and surrounding area.
[172] This leads to another deficiency in the investigation, which was the lack of photographs of a number of key items of evidence. There were no photographs of the scrape mark at all see at [130]. As to the second set of tyre marks (marks ‘C’ and ‘D’ on the diagram above at [120]), we agree with Mr Beckett (see at [153]) that there were no proper photographs of those marks. Indeed, we are still uncertain as to their exact location on the photographs where they do appear and we have no doubt that the jury must have been left in a similar position. Other marks at the scene were not photographed or catalogued for elimination purposes. The first set of tyre marks (marks ‘A’ and ‘B’), for example, were visible on a photograph but there were no close up photographs. The lack of clear photographs was exacerbated by the varying terminology used by the experts to describe the marks – see at [220] below.
[173] We would have expected there to have been a full set of close up photographs of all marks on the road, in case there was any question later of their possible relevance. Such a record would have enabled Detective Constable Maddaford to remind himself of the physical evidence, particularly as he did not take notes at the scene, apart possibly from a very rough plan. It would also have facilitated any later review of the evidence, both by the Crown reviewing expert and any defence expert. This would have reduced the scope for later disagreement and enabled the evidence to be properly tested. In particular, it would have reduced any disagreement between the various experts about whether particular marks were skid marks or not – with close up photos of the marks, this could easily have been determined. Such photographs should have indicated the beginning, middle and end of each skid, as well as any changes in direction in the mark. It would also have given the jury the means fully to assess the experts’ opinions in the light of their own review of the evidence. It is obviously better, where possible, for an expert witness’ evidence on the facts to be backed up by other evidence, such as, in this case, photographs of the particular marks on the road that the constable relied upon.
[174] Further, although it may in practice be difficult in some expert testimony to disentangle facts from inferences and opinions, the distinction between perceptions and opinions is a real one and, as far as possible, should be made clear. In a criminal context a jury ought to be made aware what evidence is “fact” (such as tyre marks) and what is the expert’s interpretation or opinion of such facts (whether the marks were caused by the back wheels of the truck). This will enable the jury better to evaluate competing theories see Malek and others (eds.) Phipson on Evidence (16ed 2005) at [33-10]. The distinction was often not evident in this case.
[175] We also consider that there is a real problem with the diagrams prepared by Senior Constable Maddaford. The main error was the depiction of the gouge mark over into the north bound lane. Even on his own evidence, it was “on” the centre line (see above at [132]) and it appears from Mr Beckett’s evidence and the photographs of the mark, that it was actually in the south bound lane, albeit just barely see at [165]. The fact this mistake was made, however, raises concerns about the accuracy of the other parts of the diagrams, particularly given the lack of scale on some of them. Certainly on at least one of the diagrams the second set of tyre marks appear much longer than must have been the case. This underscores the importance of scale diagrams. Scale diagrams would have made the relative distance between marks clear, and also indicated the length of skid marks, thus reducing disagreement, for example, about the length of, and distances between, the first and second set of tyre marks. Scales would also have reduced any errors (perceived or actual) as a result of the photocopying process.
[176] We are also concerned that there was no examination of the underside of the vehicles to ascertain which vehicle made the gouge mark and, more importantly, which part of which vehicle, made the gouge mark see at [155]. If that mark had been made by the passenger side of the Honda then, as it was just inside the south bound lane, this may well have backed up the defence theory that it was the Honda which turned across the Isuzu and not the other way round. As there was no examination of the underside of the vehicles, Senior Constable Maddaford’s evidence that the gouge mark was made by the driver’s side of the Honda (see above at [125]) seems speculative. It may be that his evidence is explicable by some of the factors identified by Sergeant Stables (see above at [142]) but that was never made explicit in the constable’s evidence.
[177] Further, the vehicles were not physically placed together to ascertain, as far as possible, the angle of engagement (see above at [155]). This was important as it was his postulated angle of engagement that caused Senior Constable Maddaford to conclude that the Honda had rotated clockwise and therefore explained why it landed up so far over into the south bound lane. However, the only evidence for this particular angle of engagement was the constable’s initial manual rotation on the CAD computer software, based on his assessment of the panel damage – see above at [125]. We also note that there is no evidence that Senior Constable Maddaford ever considered the eyewitness account and whether he should modify his view in light of that evidence. At least, the constable should have explained in detail why Ms Clisby’s evidence did not fit the physical evidence.
[178] We would have expected that the items of physical evidence (debris, tyre marks, gouge marks etc) at the scene would have been numbered and labelled in some way so that the labels were visible in photographs. This would have allowed the particular marks to be easily identified. We would also have expected their position to have been pinpointed by measuring the distance from a particular reference point. Such a system would have greatly simplified this case, when there was often disagreement about which particular mark was in issue, or indeed its true location.
[179] Finally, Mr Beckett in his evidence indicated that it was usual in these crash investigations to undertake a physics-based or mathematical analysis which estimates distances involved, the angle of approach of the vehicles, their likely speed and direction, location of skid marks, the likely point of impact, and then uses that data to extrapolate a range of possibilities as to how the crash occurred. Such modelling can provide a range of possible variations on the speed, direction and distance of the vehicles involved, and enable assessment of the most likely crash scenario, and the events preceding it. Such analysis would require accurate measurements of the evidence at the site.
Assessment of the expert evidence
[180] We discuss each of the key items of physical evidence in turn. Mr Beckett’s evidence was that the first set of tyre marks (marks ‘A’ and ‘B’ on the diagram above at [120]) could have been made by the right hand back wheels of the Isuzu as it braked just before the collision. Mr La Hood, for the Crown, submitted that the jury was entitled to accept the evidence of Senior Constable Maddaford that there were no other skid marks at the scene as he was the only one who had been at the scene straight after the collision. However, it does not appear to us that Senior Constable Maddaford in fact examined the first set of tyre marks at the scene at all. There were no notes made by the constable at the time discounting those marks as having been made by the emergency vehicles and indeed that suggestion seems to have been made for the first time at trial.
[181] Mr La Hood also pointed out that Mr Beckett was relying solely on the depiction of the first set of tyre marks in a photograph and that Sergeant Stables’ evidence was that he could see no striation marks in those photographs that would indicate a locked wheel (see [147] above). The jury would have been entitled to accept Sergeant Stables’ evidence that he could see no striation marks depicted in the photographs (particularly as the photographs were so unclear) but this is different from saying that there were in fact no striation marks. It was just that the photographs were not good enough to see whether or not there were any.
[182] What is clear is that the marks were made by a dual wheel vehicle and the Isuzu had dual back wheels – this was, unsurprisingly, conceded by Senior Constable Maddaford (see above at [137]). We thus consider that the Crown had not excluded the possibility that these tyre marks were brake marks made just before the collision, particularly in light of Ms Clisby’s evidence as to having seen the vehicle in front of her in her lane brake momentarily just before the collision.
[183] Turning to the second set of tyre marks (marks ‘C’ and ‘D’ on our diagram at [120]), we do not consider that the Crown had ruled out the possibility that those marks had been made by the Isuzu after the collision as it rotated on its side. Senior Constable Maddaford placed a great deal of emphasis on those tyre marks as supporting his thesis that the collision occurred in the north bound lane. He, however, gave no explanation as to why the second set of tyre marks were not parallel. He did say that he had measured the distance between them and it linked up with the wheel track of the Isuzu but he did not say at exactly what point he had measured the distance between the marks and thus which part of the non parallel lines matched up with the Isuzu’s wheel track (and why he was justified in concluding that this was the correct place to measure the distance between those lines). As far as we can tell, he does not seem to have measured the distance from the marks to the gouge or to the overbridge. In any event, his measurements do not seem to have survived see above at [137].
[184] Sergeant Stables’ evidence (at [144]) suggests that he would have put a lot less emphasis on the second set of tyre tracks than Senior Constable Maddaford, at least in providing any clues to the direction the vehicles were travelling in before the collision. The sergeant accepted that there was very little evidence about the precollision movements of the vehicles. Indeed, even Senior Constable Maddaford conceded that it was not possible to ascertain direction from those tyre marks (see above at [126]). Further, it was accepted as a reasonable possibility by Sergeant Stables that the second set of tyre marks were caused by the Isuzu in the manner described by the defence expert (see at [148]) and there was no rational reason for the jury to reject that concession.
[185] The Crown submitted that the fact that the tyre marks were behind the gouge mark shows that they cannot have been caused post-impact. We do not accept this. The point of impact was the front of the Isuzu. By definition the back of the Isuzu was behind that point. In order to spin 180 degrees the back of the truck needed to rotate around the front of the truck. The defence expert’s evidence was that the force of the collision tipped the truck over and then it spun round anti-clockwise. That it rotated anti-clockwise was also Constable Maddaford’s theory, although he said that it had rotated on its wheels. If the defence theory was correct, the Isuzu was spinning on an axis around the point of impact with the back of the truck circling round and then the whole truck sliding southwards. It seems entirely consistent with this theory that the tyre marks would be before the gouge mark.
[186] Constable Maddaford considered that the second set of tyre marks (marks ‘C’ and ‘D’ on the diagram at [120]) were made by the front wheels of the Isuzu as it braked prior to the collision: see above at [126] [127], and [138]. However, in our view there was not enough information for the Crown to exclude the possibility that the marks were made by the back wheels as the truck rotated after impact. There are no recorded measurements, for example, which indicate that mark ‘C’ was so far away from the gouge mark that it could not have been caused by the rear wheel of the Isuzu as it rotated after collision. This was suggested by the Crown in closing but the point had certainly not been proved in evidence.
[187] We turn to the gouge mark. It is accepted by the Crown that the photographs show the gouge mark as slightly to the left of the cat’s eye on the road and therefore slightly to the left of the centre line in the south bound lane. The Crown submitted that the position of the gouge does not indicate that the vehicles must have been in the south bound lane prior to impact. However, Mr Munro did not need to prove this. All he needed to raise was a reasonable possibility that this was the case. That the likely point of maximum engagement was just inside the south bound lane must in our view favour the defence theory. Further, there is a risk that Senior Constable Maddaford’s conclusion that the gouge was made by the driver’s side of the Honda was entirely speculative (see at [125] above). The mistake as to the position of the gouge mark and the speculative nature of the conclusion that it was made by the Honda removes a major plank of the physical evidence that Senior Constable Maddaford said pointed clearly toward the collision being in the north bound lane (see at [132] above).
[188] Turning now to the direction and manner of rotation of the vehicles, the most significant area of disagreement between the experts, with regard to the Isuzu, was whether it rotated on its side or on its wheels. Senior Constable Maddaford’s evidence was that there were no marks on the road that showed the rotation of either vehicle. In cross-examination, (see at [162] above) Mr Beckett said that, if the Isuzu had rotated on its wheels he would have expected to have seen yaw marks. The fact that there were none supported his theory that the truck had tipped before sliding. As there was no comment on that proposition from either of the Crown experts, we do not consider that the jury would have been entitled to reject out of hand the possibility that the lack of marks supported the defence theory.
[189] As to the direction of rotation of the Honda, it appears, as noted above (at [136]), that Senior Constable Maddaford’s evidence on this point relied heavily on the angle of engagement of the vehicles. However, Mr Beckett disagreed with the constable’s theory on the exact angle of engagement (and criticised the fact that the vehicles had not been physically placed together to ascertain the exact angle – see at [164] above). Further, albeit hypothetically, Sergeant Stables agreed that, if the angle of impact had been in line with Ms Clisby’s evidence, he would have expected a vehicle in the Honda’s position to have rotated anti-clockwise (see at [150] above). We do not consider it possible for the jury rationally, in light of the criticisms of the defence expert and the concession (albeit a bit confused) of the sergeant, to have preferred the evidence of the constable on this point and rejected the defence theory as a reasonable possibility.
[190] The evidence of the scrape mark, which Senior Constable Maddaford said confirmed the direction of rotation of the Honda, is most unsatisfactory. There were no photographs taken of it and, as far as we can ascertain, no measurements. If there was a rough diagram made at the scene it is no longer available and Senior Constable Maddaford must have been relying on his memory of that mark when giving evidence. It appears from Mr Beckett’s visit to the scene, admittedly some two years later, that in fact the scrape mark may have been two gouge marks. In the circumstances the scrape mark carries little, if any, probative value: after all, there was no physical evidence to support Constable Maddaford’s assertion (at [130] above) that the scrape mark was attributable to the Honda. It certainly does not tend to show that the Honda was likely to have rotated clockwise. We thus do not consider that the jury would have been entitled to rely on the evidence of the scrape mark as supporting the Crown case.
[191] The fact that the debris was in the south bound lane appears to us to raise at least a reasonable possibility that the collision occurred in that lane. This was conceded by Sergeant Stables (see at [149] above) and even Senior Constable Maddaford appears to have seen this point as neutral (absent the other physical evidence). The Crown submitted that it seems to have escaped everybody at trial that one of the diagrams produced by the prosecution indicates that impact occurred as the truck was turning back into the south bound lane from the north bound lane and thus finding debris in the south bound lane is equally consistent with the direction of the motion of the truck as depicted in the diagram. As this was not a point explored at trial and none of the experts had had an opportunity to comment on it, we do not think it is a matter that can now be raised. In any event, at best this means the evidence of the debris field is neutral. It does not rule out the possibility that the point of first impact was in the south bound lane.
[192] With regard to the final resting place of the vehicles, as we understand it, whether this supports the prosecution or defence theory depends on the direction of rotation of the vehicles after impact. Sergeant Stables made some concessions on this point in cross-examination (albeit hypothetically). Given the disagreement between the experts, this could not be seen by the jury as a conclusive point. Putting the Crown case at its highest, the resting positions of the vehicles were consistent with the collision occurring at an angle and in the north bound lane; but the rest positions appear to be equally possible if the crash occurred in accordance with the defence hypothesis. The point to us appears equally balanced.
[193] As to the damage to the vehicles, Senior Constable Maddaford agreed that, if his diagram was rotated (at [133] above), this was consistent with the collision having taken place in accordance with the eyewitness account, a telling factor in favour of the defence (see at [117] above). Further, there is, in any event, uncertainty as to the actual angle of engagement (see at [189] above).
[194] It will be clear from the above that we are of the view that a number of conclusions in the Crown case were unsupported by evidence, or the evidence for them was at least equally consistent with the defence case. We turn then to whether the jury would have been entitled to prefer the Crown case theory on the grounds that their experts were more qualified and/or more credible. All three experts were of roughly equal expertise and choosing between them on the basis of differential expertise was not possible. The Crown submitted that the jury was entitled to accept Senior Constable Maddaford’s evidence over that of Mr Beckett because he had the advantage of having been to the scene just after the accident. We agree that this is an advantage but it is one in our view diminished significantly by the lack of any contemporary notes taken at the scene (see, eg, [180] above).
[195] We accept further that the manner in which experts give their evidence can affect the weight a fact finder may legitimately put on their evidence and that this is not something that can be fully captured by reading a transcript of that evidence. In this case, our impression is that Mr Beckett may have come across as somewhat pedantic in cross-examination. For example when asked, in relation to the gouge mark, whether 50 millimetres was about two inches he replied that he did not use imperial measurements. Some of the prosecutor’s criticisms in cross-examination may have reinforced the impression that Mr Beckett was pedantic.
[196] It may be too that the jury was impressed with the certainty with which Senior Constable Maddaford gave his evidence as against the often tentative nature of Mr Beckett’s evidence. Mr Beckett refused to speculate on a number of matters and often couched his evidence in terms of possibilities rather than certainties. This reticence was largely a function, however, of quite proper professional caution and of difficulties created by deficiencies in the investigation, such as the lack of proper photographs and measurements.
[197] While the impression a witness makes is important, and of course a jury is free to reject the opinions of an expert witness, it is incumbent on a jury to assess the evidence of all witnesses (and in particular expert witnesses) clinically and in a detailed manner rather than purely impressionistically. This is especially true in a case such as this, where the qualifications and experience of the expert witnesses were broadly similar. The jury was required to make a decision as to whether particular expert evidence should be accepted in the context of all the evidence in the case. Perhaps most importantly, however, the defence did not need to prove that its alternative explanation was correct but merely that there was a reasonable possibility that it was so.
[198] The question is therefore whether there was a proper basis for the jury to consider that Senior Constable Maddaford’s evidence proved the Crown case to the requisite standard. Senior Constable Maddaford’s evidence was that the physical evidence was at odds with the eyewitness account and that it pointed to the collision having occurred in the north bound lane. He relied on the second set of tyre marks, the gouge mark, the debris trail, the topography of the crash and the post-crash rest positions of the vehicles (see [132] and [139] above).
[199] For the reasons set out above, the debris trail and the post-crash rest positions appear at best neutral. This is particularly the case because the angle of engagement was crucial for the constable’s theory as to the post crash-rest positions and that engagement angle had not been verified by placing the vehicles physically together. With regard to the second set of tyre marks, there was no rational reason for the jury to reject the defence explanation as not being reasonably possible, particularly in light of the concession made by Sergeant Stables. The significance of the gouge mark was inconclusive due largely to the deficiencies in the investigation but its position in the south bound lane, if anything, favoured the defence theory.
[200] It was never explained exactly what the constable meant by the “topography of the crash site” or the effect it may have had on the reconstruction of the crash. We note, in any event, that, if he meant by topography the local detail and layout of the site (including the position of natural and man-made features), the existence of the lane on the left of the south bound lane favoured the eyewitness account. If, on the other hand, he meant the relative positions of all the pieces of physical evidence (tyre marks, gouge etc), in the absence of recorded measurements, the significance of distances between marks was unknown. In any event, we note that, in the absence of an explanation, the jury could not have lent any weight to this factor.
[201] We thus do not consider that the jury was entitled to accept Senior Constable Maddaford’s evidence as by itself proving, to the requisite standard, that the collision occurred in the north bound lane in the manner he postulated. This was not a case in which the jury was entitled simply to reject the evidence of one expert and accept that of another. This is particularly the case as we have already held that we do not consider that the jury was entitled to reject the eyewitness account in its entirety (see [117] above).
Assessment of the remaining Crown evidence
[202] The remaining evidence can be split into three broad categories – the evidence of Mr Munro’s intoxication, the evidence of Mr Munro’s alleged bad driving before the crash and Mr Munro’s statements after the collision. We examine each in turn.
EVIDENCE OF MR MUNRO’S INTOXICATION
[203] A paramedic called to the scene gave evidence that he noticed a distinctive smell of alcohol when he first entered Mr Munro’s vehicle. Mr Munro admitted to the paramedic that he had had a few beers but did not say how many. On being asked what had happened, Mr Munro was quite vague and was slurring his words. He knew that he had been in an accident but the paramedic did not recall Mr Munro explaining exactly what had happened. The purpose of the paramedic’s questioning was to ascertain whether Mr Munro had sustained a head injury in the accident and whether all his faculties were functioning. On the paramedic’s primary survey, Mr Munro appeared to have no head injury.
[204] As indicated above, Mr Munro was then taken to Palmerston North hospital where he was treated for a broken sternum. Dr Murray, who was at the time attached to the hospital’s accident and emergency department, examined Mr Munro when he was brought in. He said that Mr Munro was conscious but drowsy and smelt strongly of alcohol. He had multiple minor bruising and lacerations over his body. An x-ray showed a fracture of his chest bone and an electrocardiogram indicated bruising to his heart. The doctor did not have his medical notes and could not comment on whether Mr Munro had a bump to the head.
[205] Dr Murray, with Mr Munro’s consent, took blood samples. The blood was analysed by Ms McGavin, a science technician employed by the ESR in Porirua. Her evidence was that the result was a reading of 98 milligrams of alcohol per 100 millilitres of blood. This was reported in the certificate of analysis as 92 milligrams per 100 millilitres of blood. The six milligrams deduction is standard ESR procedure for any reading above 36 milligrams to allow for any possible analytical variation. She also confirmed that, based on certain assumptions including an assumed rate of absorption, she would estimate Mr Munro’s blood alcohol concentration at the time of the accident to have been between 128 and 158 milligrams per 100 millilitres of blood. She acknowledged in cross-examination that estimating the blood alcohol level some three hours prior to when the blood was taken is a somewhat inexact science, in particular because a person’s rate of absorption may vary.
[206] The defence did not contest the fact that Mr Munro had been drinking or the evidence of his blood alcohol level some three hours after the accident. The jury would also have been entitled to accept the evidence of Ms McGavin as to the much higher level of alcohol that would have been present in his blood at the time of the accident and to consider that it would have affected his driving adversely. There is, however, a difference between considering that his driving would have been impaired and a jury being entitled to be sure that Mr Munro crossed the centre line to cause the collision.
EVIDENCE OF BAD DRIVING
[207] The Crown also called a number of witnesses of bad driving, allegedly by Mr Munro, just before the crash. The first was Mrs Litten. On the evening of 17 November 2004 she and a friend, Mrs Vincent, were travelling from Levin, where they had been playing cards, to Otaki. They had finished playing cards between 9.15pm and 9.30pm. Mrs Litten was driving and had only just come out onto the main highway when a white truck overtook her, travelling, she thought, well over 100 kilometres per hour, and, in doing so, passed very close to her vehicle. The truck then weaved in and out of traffic, passing three or four cars in front of her. The truck had to cross over double yellow lines to pass her as she was in a single lane but there was a passing lane further on when the truck was overtaking the cars in front of her. In cross-examination Mrs Litten agreed that, in the statement she made to the police before depositions, she did not make any reference to the truck crossing the double yellow lines but maintained her evidence that this is what had occurred.
[208] About ten minutes after that incident, Mrs Litten came across the accident. She saw the white truck on its side straddled across both lanes and stopped the car. There was a young woman (presumably Ms Clisby) walking around in circles in a very disoriented state and Mrs Litten got out to see if she was all right. The young woman said “I saw it all happen. I saw it all happen”.
[209] The passenger in Mrs Litten’s car, Mrs Vincent, also gave evidence. She said that she had seen “a white van” shooting past them. It went past so quickly that she could not really make it out. She could not say how closely the van passed by their car but it was “too close for comfort” and it had passed them on a double yellow line. She did not notice the van passing any other cars.
[210] Mrs Vincent also described coming across the scene of the accident. She said that they crossed the bridge and saw the van lying on its side with sparks coming from underneath. The other car was about nine feet further away. She stayed near her car and rang 111. She assumed that the van on its side was the same van that had overtaken them earlier as she did not think that anyone else had done so.
[211] The next witness was Ms Cummins. She had left a meeting in Otaki at about 9.20pm and filled up her car with petrol. The till receipt from the service station gave the time of payment as 9.33pm. She then began her journey home on State Highway One, north towards Levin. As she approached the Waikawa Stream bridge, she saw headlights coming towards her. She is familiar with the road and she realised that the lights were in the wrong place. They suggested a south bound vehicle moving towards the north bound lane. At the time she saw the angle of the headlights, there were no other cars around that could have justified the vehicle being in the wrong lane.
[212] Ms Cummins knew that she had to take evasive action but was unable to move to the left because she would have hit the side of the bridge and she did not have enough time or power to move to the right. She therefore speeded up off the bridge and the vehicle went past, by then completely in her lane, and missed the back of her car “by about 20 centimetres”. After that, she had to pull over for a few minutes to calm down. Ms Cummins was not able to describe the other vehicle in any detail. All she could say was that it was white and that it was bigger than her Honda Civic. She thought it was probably a small truck, or a 4x4 or a ute or a van. However, it was not a big truck because it did not have “big truck lights”. She said that there was not much traffic on the road that night and, as far as she could recall, there had been only two or three cars going south.
[213] Given the timing of the two incidents described above in relation to the timing of the accident, the fact that the description of the white vehicle involved in those incidents was consistent with the vehicle being driven by Mr Munro and the fact that there were not many vehicles on the road at the time, the jury would have been entitled to infer that Mr Munro had been driving very badly and that he had crossed the centre line on two occasions within some ten minutes of the accident. This was important evidence for the Crown.
[214] However, while the fact that Mr Munro had driven badly before the accident increased the likelihood that he had also crossed the centre line just before the collision, it cannot by itself prove this beyond reasonable doubt. The overtaking evidence (above at [207]) was evidence of bad driving but arguably different in kind. There was no suggestion of dangerous overtaking at the time of the accident. The bad driving described by Ms Cummins (see at [212] above) bore more resemblance to what the Crown alleged occurred at the time of the accident but one incident, even in such close proximity, would not suffice on its own to found a guilty verdict. This is particularly the case as there was, at least, a reasonable possibility that the victim’s driving was impaired at the time of the collision: see [104] above.
MR MUNRO’S STATEMENTS
[215] Sergeant Williams visited Mr Munro in hospital. Mr Munro told him that he had been travelling from Palmerston North to Waikanae and that he could not recall anything about the crash. When he was told that somebody had been hurt, he became extremely upset. He began crying and said “what have I done”.
[216] Mr Munro was also interviewed at the Otaki Police Station on 16 December 2004. In that interview Mr Munro said that he had no memory of the accident or, indeed, of the day leading up to the accident. He thought that this was because of a “bang to the head” that he received in the accident.
[217] In order to accept what Mr Munro said when he was told that someone had been injured in the accident as an admission of fault, the jury would have had to discount the possibility that he had no memory of the accident. The evidence of a bang on the head was not strong – see at [203] above. The jury would thus have been entitled to discount the possibility that any memory loss resulted from head injuries sustained in the crash. On the other hand, Mr Munro had been drinking and he had sustained other injuries in the crash. His comment at the hospital was not an explicit admission of fault and was certainly not an endorsement of the Crown theory of how the collision occurred.
[218] We do not consider therefore that, by itself or when taken in addition to the expert evidence, Mr Munro’s comment at the hospital would have allowed the jury to find Mr Munro guilty.
CONCLUSION ON REMAINING CROWN EVIDENCE
[219] In our view, the evidence of intoxication, the evidence of bad driving and Mr Munro’s possible admission of fault would have been available to weigh into the process of assessing the eyewitness and expert evidence but would not have been sufficient on its own, or in conjunction with the expert evidence, to sustain a guilty verdict. While this other evidence was probative of the fact that Mr Munro was drunk and driving very badly, it could never be relied upon on its own to prove that Mr Munro had in fact caused the collision, particularly in circumstances where the victim’s driving was also likely to have been impaired.
Aspects of the trial process
[220] As can be seen from our discussion above, much of the expert evidence in this case was highly technical in nature. Unfortunately the fashion in which the expert testimony was given would, we think, have been very confusing for the jury. Differing terminology was used by the Crown and the defence, and indeed even the Crown experts varied the terms with which they described marks on the road. By way of example we note that the marks ‘C’ and ‘D’ on our diagram (above at [120]) were referred to variously throughout the trial as “tyre marks”, “striation marks”, “skid marks”, the “relevant markings”, “friction marks”, “tracks”, and “scrub tyre marks”. Indeed varying terms were also used at the appeal hearing. We have no doubt that the jury would have found such differences in terminology confusing. When combined with the lack of photographs or numbering of particular marks (see at [172] and [178] above) there was a real risk of the jury being uncertain which marks were being referred to in any passage of testimony, or the significance of the particular term employed to describe those marks.
[221] In addition, the computer generated diagrams of Senior Constable Maddaford were presented as if they were fact when they were in most cases a reconstruction of what happened based on the constable’s opinion (and in some cases at least were wrong even on matters of fact – see at [175] above). These diagrams should not have been included in the booklet of photographs of the scene as if they were actual maps of what was found at the scene. Some had no scale and the ones that did became meaningless through the photocopying process – see [175] above. The gouge mark was in the wrong place on the diagrams. There were no measurements on the diagrams of the second tyre marks or of the gouge or any measurements of the distance between them and other physical evidence at the scene. Indeed, it is not even clear whether some of those measurements were done (see [170] above). Most importantly, the positioning of the vehicles (apart from in their final resting places) on the diagrams was according to the prosecution’s theory or hypothesis as to the point and angle of collision rather than the actual position of the vehicles at that time.
[222] If these charts were to go to the jury, they should have been kept separate from the photographs and it should have been made very clear that they were not actual representations of what had happened. The jury had no warning from the judge as to the nature of the diagrams or the use to which they could properly be put. They were not actual representations of the scene. They were illustrations of Constable Maddaford’s evidence as to the physical scene and his opinion of how the accident happened. A jury direction to this effect was, in our view, necessary in this case but none was given.
[223] Interestingly, even in submissions before us, the Crown referred to the diagram prepared by Senior Constable Maddaford set out at [125] above as if that was in fact the position of the vehicles at the point of maximum impact, rather than the Constable’s hypothesis as to the position of the vehicles, an hypothesis challenged strenuously by the defence see the diagram at [133] above. We note at this point that in a comparable case in Queensland, the use of a similar diagram containing expert assumptions as to vehicle position, but based on material not in evidence, was found to have wrongly impressed the jury and caused a miscarriage of justice – R v Faulkner [1987] 2 Qd R 263 at 271 (CCA). Such diagrams may be useful but without careful warning are, we think, very likely to be regarded (wrongly) by a jury as an additional source of confirmatory evidence.
[224] We also have some concern about the summing up of the respective cases of the Crown and Mr Munro. We consider this to have been a case where greater assistance should have been given to the jury in identifying the issues on which they needed to focus. It would, for example, have been helpful for the jury to have been told that, if they accepted that it was a reasonable possibility that Ms Clisby was wrong about the direction the cars were travelling in but right about everything else (ie that the north bound vehicle (the Honda) veered into the path of the south bound vehicle (the Isuzu) with the collision occurring in the south bound lane), then Mr Munro should be acquitted. In this regard, the jury should have been reminded of the areas where the physical and expert evidence accorded with Ms Clisby’s evidence – for example, the existence of the side road and the pattern of damage to the cars, which was accepted by all three experts as consistent with her account.
[225] It would also have been helpful if the jury had been directed to the main items of physical evidence relied on by the experts and the main points of similarity and difference between the accounts of the experts. They should, in particular, have been directed to the concessions made by the Crown experts with regard to these key pieces of evidence. They should also have been told to take into account all of the evidence in their assessment of the physical evidence, including the eyewitness account. The Judge could have prepared a chart similar to the one we have prepared in the appendix setting out the crucial areas of agreement and disagreement between the experts, with references to the relevant portions of the transcript. The jury ought to have been reminded at that point that, if they thought there was a reasonable possibility that the defence version was correct, they should acquit. It should have been pointed out to the jury that, on a number of key points, to reject the defence version as a reasonable possibility also involve rejecting Sergeant Stables’ concessions.
[226] We realise that this may seem somewhat of a counsel of perfection to a busy trial judge in a very busy trial court. There are, of course, many cases where such elaborate directions are not necessary. However, in this case the bulk of the evidence was technical. Further, the photographs were inadequate and the expert evidence had been presented in a confusing fashion with inconsistent terminology used by the experts in their evidence about key aspects of the physical evidence. Jury confusion was thus highly likely.
[227] There were also concerns about jury competence which made the need for very clear directions vital. The foreman of the jury had failed to turn up at court on the second day of the trial, having checked the court answerphone and heard that the jury panel was not required that day. Her misunderstanding in this regard led to an unsuccessful application by the prosecutor for her discharge. Further, one of the other members of the jury had brought some toy cars to court, raising concerns about inappropriate experimentation and leading to a warning against this from the Judge.
[228] Finally, we note that the jury was (quite properly) given the transcript of evidence to consult during their deliberations but the standard directions on the use of the transcript, such as making sure that reference is made to the cross-examination as well as the examination-in-chief, was not given.
Ought a jury acting reasonably have had a reasonable doubt?
[229] Taking into account all of the evidence, we do not consider there to have been any rational basis for the jury to have rejected the defence version of events as a reasonable possibility.
[230] We accept that the jury would have been entitled to consider that the other Crown evidence supported Constable Maddaford’s theory of the case, but, in our view, this was not enough to prove Mr Munro’s guilt when considered alongside the eyewitness and the other expert evidence. This is particularly the case in light of the evidence that the victim’s driving may also have been impaired through cannabis use earlier in the evening and possible cellphone use at the critical time of the accident.
[231] In order to find Mr Munro guilty, the jury would have to have rejected the eyewitness account in its entirety. This, in our view, was not possible when her evidence accorded with much of the physical evidence and aligned much better with the defence theory than with that of the Crown (see [111] [112] above). In addition, the jury was not asked by the Crown to reject her evidence. Rather, the Crown sought to rely upon Ms Clisby holding the driver of the white van responsible for the collision (see at [117] above).
[232] A finding of guilt would also have involved the jury rejecting, as a reasonable possibility, that Mr Beckett’s theory about the various items of physical evidence was correct. In our view, the jury would not have been entitled to do so, particularly given the deficiencies in the investigation, and the fact that it would also have involved rejecting key concessions made by the reviewing Crown expert, Sergeant Stables. There was no rational reason for those concessions to have been rejected. This was not a case where experts might disagree and the jury was entitled to choose between competing hypotheses: rather, the concessions made by the Crown, and lack of evidence underpinning key elements of the Crown case, mean that the jury, acting properly, could not reasonably find the appellant guilty beyond reasonable doubt.
[233] As we noted above at [96], this was a case where Mr Stevenson was able to point to a sufficient basis for his contention that s 385(1)(a) was engaged. His challenge was based on the evidence as a whole and not, as is often the case (see at [90] above) on particular points in the evidence. This necessitated a detailed review of the evidence. After that review, we concluded that this is a case where, despite there being some evidence which, if accepted, would have supported Mr Munro’s guilt, a jury acting reasonably ought to have had a reasonable doubt about Mr Munro’s guilt. The ground of appeal under s 385(1)(a) was made out.
[234] This Court accordingly had a discretion, conferred by s 385(2), to direct that a verdict of acquittal be entered, to order a new trial, or to make any other order. It is clear that there are no hard and fast rules about the circumstances in which a retrial will be ordered: the principles and cases are summarised in the judgment of this Court in R v E (CA308/06) [2007] NZCA 404 at [145] – [148]. Mr Munro’s case was of a type in which there was insufficient evidence at trial to justify a conviction. Absent truly exceptional circumstances (see R v E (CA308/06) at [146]), it will not be in the interests of justice that the prosecution be given a chance in such cases at a new trial to cure evidential deficiencies. This was not an exceptional case.
Result
[235] For all of the above reasons, the appeal was allowed, the conviction quashed and a verdict of acquittal entered.
APPENDIX
Comparison of Expert Evidence on Key Points
|
|
|
Maddaford |
Beckett
|
|
|---|---|---|---|---|
|
First set of tyre marks in south bound lane (marked ‘A’ and
‘B’ in the diagram at [120] above).
|
From emergency vehicles and not friction marks made by a locked wheel (see
[137] above).
|
Not enough detail in photographs to tell, but if they were skid marks then
striation marks would be expected (see [147] above).
|
Possible that mark ‘B’ was from braking rear tandem wheel of
Isuzu just before collision (see [158]
above).
|
|
|
|
Second set of tyre marks (mostly in north bound lane, and marked
‘C’ and ‘D’ in the diagram at [120] above).
|
From locked and sliding front wheels of Isuzu (either from braking or
collision). Agreed very short and not quite parallel to each
other (see [126] above).
|
||
|
|
Gouge (on centre line or just in south bound lane).
|
Point of maximum engagement. Made by undercarriage of driver’s side
of Honda. On or near the centre line but portrayed in
diagrams as being in
north bound lane (see [125]
above).
|
Likely to be at or near point of maximum engagement, and thus indicate the
area of impact (see [142] above).
|
Indicates loose area of impact. Pure guesswork as to which vehicle made
it. Just inside the south bound lane (this is clear from
photographs which
were pointed to by the Judge, although the mark is not accurately portrayed in
the diagrams – see [165] above).
|
|
|
Direction of rotation of Isuzu after point of maximum engagement.
|
Anti-clockwise and on its wheels (see [129] above).
|
Was not asked, but accepted that second set of tyre marks could be
consistent with vehicle rotating on side (see [148] above).
|
|
|
|
Direction of rotation of Honda after point of maximum engagement.
|
Clockwise (see [130]
above).
|
Agreed in principle that, if one vehicle had veered straight across the
other vehicle, the veering vehicle would spin anti-clockwise
(see [150]).
|
Anti-clockwise, regardless of where on the road the impact occurred (see [156] above).
|
|
|
Scrape mark (in south bound lane)
|
Tracking from the gouge to the resting place of the Honda. Also indicates
that direction of rotation of Honda was clockwise (see [130] above).
|
No photograph. Did not visit scene. Not asked to comment.
|
Actually consists of two gouge marks. Not possible to track to position of
Honda. Not photographed or measured so not enough information
to draw any
conclusions (see [167] above).
|
|
|
Resting positions of vehicles (mostly south bound lane).
|
Angle of engagement would have spun Honda into south bound lane (see [129] above).
|
No specific comment in examination-in-chief. Not asked in
cross-examination.
|
Suggests accident occurred in south bound lane or very close to centre line
(see [156] above).
|
|
|
Debris (in south bound lane).
|
Neutral, or at the least not inconsistent with his theory (see [135] above). Did not explain why.
|
Not conclusive but arguably consistent with an accident in south bound lane
(see [149] above).
|
Suggests accident occurred in south bound lane because debris will travel
in the direction which it has been accelerated in (see [156] above).
|
|
|
Damage to vehicles
|
Was consistent with eyewitness account but physical evidence was not.
Angle of engagement may indicate evasive action by Honda -
see at [131].
|
Not square on but more angled, which could indicate possible evasive action
by one or both vehicles (see [143]
above).
|
|
|
|
Overall conclusion
|
Looking at the entire picture, the collision occurred in the north bound
lane with the Isuzu truck almost wholly in the north bound
lane. The collision
was square-on for the Isuzu and on the front passenger side of the Honda, with
the angle of engagement sweeping
the Honda away to one side of the Isuzu. The
Honda rotated clockwise and came to rest in the south bound lane.
|
The findings of Constable Maddaford’s report made sense and appeared
to follow logic. Limited evidence to indicate vehicle
movement prior to the
collision, although the gouge mark was a good indicator of the area of impact.
The damage profiles to the
vehicles indicated an evasive manoeuvre by one
vehicle.
|
Pre-impact brake marks suggest the collision was in south bound lane or
right on centre line. After the collision the Isuzu rotated
on its side and the
Honda rotated anti-clockwise to its rest point in the south bound lane. Pure
speculation to say the gouge mark
was made by the Honda. Debris also indicates
collision was in south bound lane.
|
HAMMOND J
[236] I have had the advantage of reading the draft judgment of Glazebrook J. In some respects my views differ on the law, or at least the articulation of it. I therefore set out some observations, in relatively short form. I agree that, on the merits, the appeal was properly allowed for the reasons given by Glazebrook J.
[237] My first concern relates to the proper approach to the interpretation of the relevant portion of s 385(1) of the Crimes Act 1961. That subsection provides that, on any appeal to the Court of Appeal or the Supreme Court, the Court “must allow the appeal if it is of opinion that the verdict of the jury should be set aside on the ground that it is unreasonable or cannot be supported having regard to the evidence ...”.
[238] Now those are reasonably obvious words. They express principles; they are not prescriptive “rule” type words, of a technical character. They would be understood by the person in the street. But what has happened around the common law world is that both the judges and the commentators have fallen upon them, bringing in their wake a confusing volume of judgments and commentary.
[239] This brings to mind the insightful address delivered by Lord Shaw of Dunfermline to the American Bar Association in San Francisco on 9 August 1922. His Lordship said that in ancient times interpretation was afflicted by the obstacle of formality, but “in modern times it is authority”. Lord Shaw said this has produced a new obstacle, “thick as the jungle”:
The words have already been in the hands of the judicial commentators; and as is the way with commentators, the one refers to the other, and the third to the preceding two, till the text is obscured, and the vision of the interpreter cannot get through the thicket except at the risk of his being considered a rebel and iconoclast.
Any recent statute forms an illustration ready to hand. Hardly is it born into the world, till judges fall upon it, tearing it analytically to pieces; and unless they called it at least inartistic they would not be in the fashion. But then their turn comes; and their frequent lines of error are produced and reproduced with a touching deference, till by and by the plain English of the Act does not know itself; and only great judges take the liberty to announce that the Act means what it says. (Reproduced as “The Widening Range of Law” in The Law of the Kinsman (1923), at 104 – 105.)
[240] A consequence of the judges falling upon a statute can be that there is the difficulty of trying to produce a seamless web of jurisprudence across dozens of appellate authorities. This bogs down courts in their day-to-day work and all too often deflects judges from their proper endeavour in an appeal of this character. Yet the appropriate principles are stated in the parliamentary language itself!
[241] This leads to a second broad concern: that of context. As matters presently stand in New Zealand, s 385 of the Crimes Act reflects a general policy in favour of at least one level of appellate review of criminal convictions. No doubt this is largely for “fairness” reasons. But also, we do not have the other mechanisms which exist in some other jurisdictions to review the factual soundness of convictions. In New Zealand, the state of the law is that convicted persons have no alternative but to look to the Court of Appeal or the Supreme Court as a safeguard against wrongful convictions and, at least in my view, in protecting the integrity of the system under which they were convicted. (As to the importance of the latter point I have expressed my views extra-judicially in “The New Miscarriages of Justice” [2006] WkoLawRw 1; (2006) 14 Waikato Law Review 1). It is true that there is a very limited safety net in s 406 of the Crimes Act, but even that involves a reference back to the Court of Appeal, by the Governor-General. The short point is that, as a matter of context, in New Zealand the entire burden falls on the Court of Appeal or the Supreme Court of New Zealand, as the case may be. In such a context, the words of the statute should not be read down, because there is no other avenue of redress for the wrongly convicted person.
[242] My third concern is that, in approaching the words of the statute with the two concerns I have already voiced in mind, any constraints which are put on the words should only be constraints or explanations going to the respective roles which the various actors in this enterprise are to undertake. Apart from that, there is no need for any further exegesis on the statutory words.
[243] For that reason, in my view the correct approach to s 385(1)(a) for the appellate court is to ask itself whether the verdict is one that a properly instructed jury, acting judicially, could reasonably have rendered. It follows that a reviewing court must thoroughly re-examine the evidence and also bring to bear the weight of its judicial experience to decide whether, on all the evidence, the verdict was reasonable.
[244] This leads to a fourth concern, which relates to the consequences of the straightforward recognition of such an approach. Some appellate courts – notably the High Court of Australia in Weiss v R [2005] HCA 81; (2005) 224 CLR 300 – have openly recognised the burden this sort of approach potentially poses for intermediate appellate courts. This is diminished in some jurisdictions by the interposition of leave requirements. For instance, in Canada a frontal attack on the facts is distinctly limited by s 675 of the Canadian Criminal Code. To advance a “Ramage” type appeal in Canada, either the certificate of the trial Judge is required, or the leave of the Court of Appeal. Sopinka and Gelowitz The Conduct of an Appeal (2ed 2000) observe at 109, “Although [in Canada] appeals involving a factual element require leave to appeal, the focus on the merits of the proposed appeal in the leave application results in substantive appellate review even where leave is denied”. Even so, leave provisions do operate as a kind of “sieve” and reduce something of the burden on the intermediate appellate court. Where (as in New Zealand) there is no such screening device and the no-sufficient-evidence appeal point is “as of right”, what is the practical outcome going to be?
[245] In this instance, Glazebrook J has undertaken a comprehensive review of all the evidence as to the merits of the case. That review convincingly demonstrates that there was an unsound conviction in this case.
[246] An exercise of that magnitude does however create consequential difficulties. On my own judicial experience, since leave was granted by the Supreme Court of New Zealand in O v R (the leave judgment is [2007] NZSC 49; the appeal is SC25/2007 and is to be heard on 20 November 2007), the first ground of appeal in every case seems now to be what for convenience I will call a “Ramage type” point of appeal. It seems that the criminal defence bar is, somewhat unsurprisingly, reacting to what it anticipates may likely be said elsewhere: that there must be a thorough examination of all the evidence if the appeal point is taken, because (as the High Court of Australia put it at [47] in Weiss) that is what the statute requires. This raises a question of resource allocation which it is inappropriate for me to address in a judgment, other than to say that if every convicted person is to have an appellate review of the length undertaken in this case, the practical effect on the administration of criminal justice will be relatively obvious. In the end, one can only take refuge in the judicial line frequently espoused by the late Lord Cooke of Thorndon, that what is required is what is appropriate to the occasion.
[247] My fifth and final concern is that I maintain, as I have extra-judicially (see [241] above), that there is a proper place for a residual safety valve, of a relatively innominate character, and by whatever name it is called, in our law on miscarriages of justice.
[248] In England, after the change in the legislation to which Glazebrook J has referred, the Court of Appeal created the admittedly controversial concept of a “lurking doubt” for cases in which there are no distinct grounds for overturning a conviction. It is a rarely used provision, and it has been criticised for not requiring an analytical grounding, and for trespassing on the function of the jury. However, the law is not just about analysis.
[249] A good example of this latter point is to be found in the judgment of Lord Woolf CJ in Bell [2009] UKHL 20; [2003] 2 Cr App R 13. Mr Bell was charged with sexually abusing his step-daughter more than 30 years ago. A stay was refused and Bell was convicted by a jury. The appeal was on the footing that the evidence relied upon was unreliable. That was the only ground Bell could appeal on for it was conceded that everything else had been done impeccably by the Court. The Chief Justice (at [27]) said:
However, there remains in this Court a residual discretion to set aside a conviction if we feel it is unsafe or unfair to allow it to stand. This is so even where the trial process itself cannot be faulted. It is a discretion which must be exercised in limited circumstances and with caution. When we exercise that discretion we must be conscious that we are not only involved in deciding where justice lies for the appellant. We must do justice to the prosecution, whose task it is to see that the guilty are brought to justice. We must also do justice to the victim. In this case we are particularly conscious of the position of the victim. If she is right, she was treated in a most disgraceful way by someone whom she should have been entitled to trust: her stepfather. For years, for understandable reasons, as we have already indicated, she felt unable to make public what had happened. She is entitled to justice as well. But we also have to do justice to the appellant. At the heart of our criminal justice system is the principle that while it is important that justice is done to the prosecution and justice is done to the victim, in the final analysis the fact remains that it is even more important that an injustice is not done to a defendant. It is central to the way we administer justice in this country that although it may mean that some guilty people go unpunished, it is more important that the innocent are not wrongly convicted.
[250] As a matter of “pure” analysis one could say that Bell raised a Ramage-point: how could anybody reasonably rely on evidence which just could not be tested properly after all those years? The English Court of Appeal did not trouble itself with that sort of artificiality: it simply said the evidence was too old to ground a conviction on it.
[251] Quite where that sort of case would fit in the New Zealand miscarriage provision is a nice point. Some Judges might say, just treat it as a “miscarriage”, as per s 385(1)(c); others might conceptualise it as a “bad or inadequate evidence” case. My concern is that, however it is achieved, a residual, innominate category for miscarriages of justice is a very important device in the law.
[252] For a very recent New Zealand example of the need for a straightforward concern with unsafe convictions as such, see R v X [2007] NZCA 500. There a young complainant in a sexual assault case gave evidence that the offence had occurred; then recanted under cross-examination; a s 347 application under the Crimes Act was declined; the jury convicted; and the trial Judge even went so far as to file a Minute after the verdict had been returned justifying (to presumably the Court of Appeal) the verdict as one based on evidence on which the jury could safely rely. If there was not a reasonable doubt in that case (how could the jury have been sure?) there was certainly a “lurking doubt” about the conviction.
[253] What cases like Bell and R v X may be thought to demonstrate is that at some point there must be an end to analysis: what is needed on an appellate review of them is detachment, sagacity, and in Hugo Young’s memorable phrase, an “unseducible engagement” with justice (“The Compromising of Lord Hoffmann” in Supping with the Devils (2003) at 214).
Solicitors:
Crown Law Office, Wellington
NZLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.nzlii.org/nz/cases/NZCA/2007/510.html