|
|
Home
| Databases
| WorldLII
| Search
| Feedback
New Zealand Criminal Law Review |
Last Updated: 18 February 2023
CASE NOTE: W (SC 38/ 2019) V R [2020] NZSC 93 AND ROIGARD V R
[2020] NZSC 94 — THE ADMISSIBILITY OF PRISON INFORMANT EVIDENCE
LOUIS NORTON*
In September 2020, the Supreme Court handed down twin decisions on prison informant evidence: W v R and Roigard v R.1 The Court divided on the approach to the admissibility of such evidence. The majority, comprising Glazebrook, O’Regan and Ellen France JJ, acknowledged difficulties with prison informant evidence, but placed significant weight on both the constitutional role of the jury as factfinder and the use of judicial directions to ameliorate unfair prejudice risks. The minority, comprising Winkelmann CJ and Williams J, proposed stricter controls on admissibility.
Prison informant evidence is often unreliable.2 Prisoners may be incentivised to invent or exaggerate confessions of other inmates to secure favourable treatment in prison, discounts in their sentences or parole opportunities.3 The Evidence Act 2006 (“the Act”) recognises reliability concerns, requiring trial judges to consider warning the jury of the need for caution in deciding whether to accept the evidence or in deciding the weight it should be given.4 The Act does not contain a standalone reliability threshold for prison informant evidence. Admissibility is instead controlled by the general relevance and unfair prejudice assessment under ss 7 and 8.
A clear tension arises. On one hand, inmates are uniquely positioned to receive information that could prove invaluable in criminal prosecutions. On the other hand, prison informant evidence is tainted with an air of inherent unreliability. New Zealand courts have been alive to this difficulty, noting such evidence requires “careful scrutiny”. But they have declined to adopt a presumption of its inadmissibility.5
There exist two broad jurisprudential approaches to the admissibility of prison informant evidence. The first is to exclude the evidence under ss 7 and 8 of the Act at the pre-trial stage. The second is to leave matters of reliability and weight to the factfinder at trial, with the evidence disciplined by judicial directions or other trial management measures.6 Courts have historically favoured the latter approach.7 The
* BA(Hons)/LLB (Hons), University of Auckland. The views in this paper are entirely the author’s.
1 W (SC 38/2019) v R [2020] NZSC 93, [2020] 1 NZLR 382; and Roigard v R [2020] NZSC 94, [2020]
1 NZLR 338.
2 W v R, above n 1, at [211] per Winkelmann CJ and Williams J. See also Ngarino v R [2011] NZCA 236; Pahau v R [2011] NZCA 147; and Jessica Roth “Informant Witnesses and the Risk of Wrongful Convictions” (2016) 53 Am Crim L Rev 737.
3 Mathew Downs (ed) Cross on Evidence (online ed, LexisNexis) at [EVA122.5].
4 Sections 122(1), 122(2)(c) and 122(2)(d).
5 See Hudson v R [2011] NZSC 51, [2011] 3 NZLR 289 at [33] and [35]–[36].
6 On the distinction between “threshold reliability” and “ultimate reliability”, see Elisabeth McDonald and Scott Optican (eds) Mahoney on Evidence: Act & Analysis (4th ed, Thomson Reuters, Wellington, 2018) at [EV122.01].
7 Scott Optican “Evidence” (2019) 4 NZ L Rev 565 at 584.
dyad of W v R and Roigard v R, however, arguably represents a more rigorous approach to admissibility.8 It countenances exceptions to the traditional rule of deference to the jury on reliability matters.
A. W v R
W v R concerned a challenge to the admission of evidence in the notorious Red Fox Tavern cold case. Police charged Mark Hoggart and a man with name suppression (“Mr W”) with the 1987 aggravated robbery of the tavern and the murder of its publican, Christopher Bush.9 At issue was the admissibility of evidence of several proposed Crown witnesses. Some of those witnesses had previously been detained in prison with Mr W. They were to give testimony about discussions they claimed to have had with Mr W, in which he alluded to, or confessed outright, his involvement in the crimes.10
The Court of Appeal differed from the High Court in its rulings on the admissibility of the various witnesses’ evidence.11 Mr W was granted leave to appeal against the decision of the Court of Appeal ruling some of that evidence admissible. Meanwhile, the Crown was granted leave to cross-appeal against the decision ruling the evidence of certain other witnesses inadmissible.12 The issues came before the Supreme Court for determination.
B. Roigard v R
Roigard v R was an appeal against conviction on eight charges of theft in a special relationship and one charge of murder. The case against the appellant, David Roigard, was mostly circumstantial. The appeal centred on the evidence of two Crown witnesses to whom Mr Roigard had allegedly admitted his responsibility for the crimes. The two men had been in prison with Mr Roigard at the time of the alleged admissions. The question was whether that evidence had been properly admitted at trial.13
The issues in W v R were:14
8 Anna High “The exclusion of prison informant evidence for unreliability in New Zealand” [2021] 25 E&P 217 at 218.
9 W v R, above n 1, at [1].
11 In R v W [2018] NZHC 2457 at [85]–[143], Woolford J ruled admissible the evidence of eight informants, generally finding it to be relevant, probative and not unfairly prejudicial. The Court of Appeal, in R v Hoggart [2019] NZCA 89 at [71]–[153], found that the evidence of only six of those witnesses was properly admissible. It too approached the issue by reference to criteria of relevance and unfair prejudice.
12 W (SC 38/2019) v R [2019] NZSC 64.
13 Roigard v R, above n 1, at [1]–[2]. The Court of Appeal had earlier upheld the admissibility of the evidence: Roigard v R [2019] NZCA 8.
14 W v R, above n 1, at [3] (footnotes omitted).
(a) To what extent may issues of reliability be considered in determining whether, under s 7 of the Evidence Act 2006, the evidence is relevant?
(b) To what extent may issues of reliability be considered in determining whether, under s 8(1) of the Evidence Act, the probative value is outweighed by the risk that the evidence will have an unfairly prejudicial effect on the proceeding?
(c) Was the evidence inadmissible under s 30 of the Evidence Act dealing with improperly obtained evidence?
Similarly, the focus on appeal in Roigard v R was whether the evidence should have been excluded under s 8(1) of the Act because the risk of unfair prejudice it carried outweighed its probative value. As in W v R, there was also a question whether the evidence should have been excluded under s 30 of the Act on the ground that it had been improperly obtained. Both issues stemmed from the appellant’s contention the witnesses had perverse incentives to give evidence. Those incentives impaired the evidence’s reliability.15
The following discussion focuses on the statements of principle and law set out in W v R. There, the majority and the minority articulated their respective analyses of the relevant principles and the approach to be taken to admissibility. They applied those approaches in Roigard v R.
A. Majority decision
On the question of reliability going to relevance, the majority referred to R v Bain.16 R v Bain concerned the admissibility of part of a taped phone call that the appellant, who was later convicted of the murder of his father and other family members, had made to emergency services. Police contended the recording contained an exhalation by the appellant that he had “shot the prick”.17 The majority in W v R noted the R v Bain approach was to treat the question whether the evidence was probative of the matter in issue as “ultimately a matter for the jury or trier of fact”.18 Judges would retain their gatekeeper role as arbiters of the relevance of that evidence. The test for relevance was “whether it would be open to the jury, acting reasonably, to rely on the disputed words as an admission”.19
That approach, however, would not always be conducive to human testimony.20 Where human testimony is at issue, a judge’s pre-trial assessment will be much closer to what the factfinder at trial would have to determine. Cases where reliability issues enter into the relevance inquiry would more likely arise where it would be speculative
15 Roigard v R, above n 1, at [3].
16 R v Bain [2009] NZSC 16, [2010] 1 NZLR 1.
19 W v R, above n 1, at [34], citing R v Bain, above n 16, at [53] and [58] per Elias CJ.
20 At [40], citing K (CA26/2014) v R [2014] NZCA 229.
or unreasonable for the factfinder to accept the disputed evidence. Consequently, there would be “relatively few” cases where evidence would be excluded pre-trial for lack of reliability. Exclusion for lack of reliability under s 7 would only be warranted where evidence was so unreliable “it could not be accepted or given any weight at all by a reasonable [factfinder]”.21
2. Reliability and probative value
May issues of reliability be brought into the assessment of probative value and unfair prejudice under s 8? On this question, the majority made three points about the statutory scheme. First, the scheme expressly contemplates considerations of reliability informing the evaluation of a piece of evidence’s probative value.22 Secondly, the Act sets out a series of categories for exclusion of evidence for want of reliability.23 No bespoke provisions exist in relation to prison informant evidence and there is no presumption of that evidence’s inadmissibility.24 Thirdly, s 122 of the Act provides for warnings as to the reliability of particular types of evidence.25
The majority adopted R v Bain’s characterisation of the s 8 weighing exercise as “concerned with whether the connection between the evidence and proof is ‘worth the price to be paid by admitting it in evidence’”.26 Significant to that assessment would be the judge’s view of the reliability of the disputed evidence.27 In the context of prison informant evidence, alleged inculpatory admissions demand “careful scrutiny”. Inmates are incentivised to cooperate with authorities because of anticipated forms of reciprocity — including sentence discounts, parole opportunities, monetary rewards or securing a better relationship with the Police.28 The reliability issue may be dealt with either by excluding the impugned evidence or by qualifying its use with jury directions as to its inherent dangers.29
On that latter point, the majority traversed precedent to the effect that the legislative intention was to leave reliability assessments to a properly cautioned jury,30 and that s 8 proceeds on the basis that any residual reliability risks are to be addressed by general trial management measures.31 Summarising, it concluded the reliability question had to date been relevant in weighing the probative value of evidence against
22 At [48], citing Evidence Act 2006, s 43(3)(e) (“allegations ... may be the result of collusion or suggestibility”).
23 At [49], citing Evidence Act, ss 18(1) (hearsay), 28 (unreliable statements), 29 (oppressive influence), 45 (visual identification), 46 (voice identification), and 90(7)(a) (previous statements).
24 At [49], citing Hudson v R, above n 5.
26 At [54], citing R v Bain, above n 16, at [62] (citing Australian Law Reform Commission Evidence
(ALRC R26 (Interim), 1985) vol 1 at [315]).
27 At [55]–[57], citing Morgan v R [2010] NZSC 23, [2010] 2 NZLR 508 at [45].
28 At [58], citing Hudson v R, above n 5, at [33].
30 At [59]–[60], citing Hudson v R, above n 5, at [36].
31 At [62]–[63], citing R v Wichman [2015] NZSC 198, [2016] 1 NZLR 753 at [70] and [76].
the risk of illegitimate prejudice. It pointed to apex courts having taken similar positions in England and Wales, Canada and Australia.32
From the statutory scheme and earlier authorities, the majority considered prison informant evidence should be evaluated under s 8 in “the usual way”, with judges retaining a gatekeeping role. In weighing the probative value of the evidence against its prejudicial effect, judges could consider its reliability, including by reference to inconsistencies, inherent implausibility and incentives to lie.33 That approach would respect the constitutional role of the jury as factfinder. The jury remains responsible for weighing the evidence and drawing conclusions from it.34
The majority was alive to the possibility of miscarriages of justice. Jurors are susceptible to the “fundamental attribution error”, whereby they attribute witness behaviours to dispositional factors, rather than situational factors like incentives to give evidence.35 Further, the majority noted that incentivised witnesses in the prison context will often have a history of dishonesty and give evidence in circumstances where there is often no possibility of corroboration.36 Finally, the majority pointed to concerns judicial directions or provision of other information to the jury may not have salutary effects on factfinding.37 These considerations supported the requirement that such evidence be carefully scrutinised.38
What, then, is the correct approach to the s 8 evaluation? The majority summarised:39
(a) The concern in undertaking this evaluation is to determine “whether the connection between the evidence and proof is ‘worth the price to be paid by admitting it in evidence’”.
(b) In undertaking the gatekeeping role, reliability may be considered by the judge in balancing, in the usual way, the probative value of the proposed evidence against the risk of illegitimate prejudice. That reliability assessment should be made without applying any artificial limits or presumptions such as taking the evidence at its highest.
(c) The relevant factors will include consideration of the sorts of concerns about this evidence as have been discussed and which might include, for example, that the credibility of the witness in an informant context has previously been doubted, any incentives or expectations of preference at play (including the inability of the prosecution to confirm whether incentives have been offered or given), and the likely weight to be attached to this evidence.
(d) On the other hand the exercise is not that of a mini trial. The judge will be making his or her assessment in the absence of the full picture of the evidence as it may emerge at trial.
(e) Finally, the constitutional role of the jury as fact-finder needs to be respected. As has been indicated, the statutory scheme and the authorities, particularly Hudson, which we are not overruling, envisage that the court will utilise other mechanisms such as clear judicial directions to the jury to address the generic risk of unfair prejudice.
32 At [65]–[68], principally citing R v H [1995] AC 596 (HL) at 614; R v Hart 2014 SCC 52, [2014] 2
SCR 544 at [94]–[98]; and IMM v R [2016] HCA 14, (2016) 257 CLR 300 at [39], [47]–[48], [58], [96]
and [171]–[172].
34 At [73], citing R v Hart, above n 32, at [98].
35 At [74]–[80], citing a wide range of social sciences literature on wrongful convictions and incentivised evidence.
39 At [88] (footnotes omitted).
The majority then considered other safeguards addressing the risk of illegitimate prejudice, including the Crown’s responsibility not to adduce blatantly unreliable evidence.40 It noted other jurisdictions had adopted further measures, including statutory provisions governing admissibility of prison informant evidence;41 the imposition of further procedural constraints on admissibility;42 the provision of information concerning the incentives and whether the given informant had previously provided reliable information;43 general restrictions on admissibility;44 and other trial management techniques.45 The majority considered New Zealand could profitably emulate those models to establish a principled foundation for the treatment of prison informant evidence and to ensure prosecutors apply a “consistent standard” in calling such evidence.46
3. Improperly obtained evidence
The majority then turned to consider whether the impugned evidence had been improperly obtained. It disposed of the point quickly, saying Mr W’s argument was “premised on a presumption of inadmissibility”. Given its foregoing analysis that such evidence was not presumptively inadmissible, no distinction could be maintained between admissions that were properly obtained and those that were not. The evidence had not been improperly obtained.47
B. Minority decision
The minority agreed with the majority on the extent to which reliability may be considered in determining admissibility issues under s 8. Judges, in conducting their gatekeeping role, may consider reliability issues under the probative value/unfair prejudice rubric. It endorsed the majority’s five-pronged framework for that assessment.48 As to “careful scrutiny”, however, the minority went further. It proposed an additional framework to guide that scrutiny in the peculiar case of prison informant evidence. Applying that framework, the minority reached different conclusions on the admissibility of the evidence of four of the proposed witnesses.49
Their Honours were clear the appeal was confined to the question of Crown witnesses giving evidence alleging a defendant had admitted guilt, in return for some advantage
— what it called “incentivised secondary confession evidence”.50 That reciprocity is predicated on an understanding “mutual benefits will flow if admissions are elicited”.51
40 At [90]–[91], citing Morgan v R, above n 27, at [39]–[40].
47 At [98]–[101]. The minority did not address the argument further, except to agree with the majority’s conclusion at [193].
48 At [191]. For that framework, see the text accompanying above n 39.
Such “criminal justice incentives” are many and varied, including future advantage, monetary payments, reduction of charges, reduction in sentence, improved prison conditions or parole opportunities.52
1. Existing approach to admissibility
The minority confirmed the primacy of ss 7 and 8 as the “twin gateways” for admissibility, observing the Act contains no bespoke mechanism for incentivised evidence. The reliability of incentivised evidence is only explicitly addressed by provisions authorising jury directions.53 But witnesses may have compelling incentives to give evidence, with little or no corresponding disincentives. Moreover, only the Crown is institutionally positioned to offer the prospectively “life-changing” benefits witnesses seek.54 Despite these issues, courts have uniformly allowed such evidence to be admitted on the basis that reliability questions are for the jury, and unfair prejudice can be handled by jury directions.55
The prosecutorial obligation to act fairly and to protect defendants’ fair trial rights is one important control on blatantly unreliable evidence. The minority noted, however, that prosecutors remain subject to human foibles and biases. The prosecutor’s gatekeeping role consequently offers only limited protection. Like the majority, the minority considered additional safeguards were needed, and that assistance could be drawn from comparator models in overseas jurisdictions.56
2. Reliability of incentivised secondary confession evidence
The Act was enacted in 2006. At that time, the social science literature on incentivised evidence and miscarriages of justice was only incipient. The minority noted the literature has much developed in the interim, now suggesting juries give disproportionate weight to incentivised evidence.57 Their Honours surveyed a range of studies demonstrating incentivised evidence’s significant role in wrongful convictions in overseas jurisdictions. The minority observed that, of the four main categories most closely associated with miscarriages of justice (eyewitness misidentification, flawed forensic evidence, false confessions and false informant evidence), only the last is not controlled by safeguard mechanisms in the Act.58
Similarly, the minority considered that relevant studies show juries give similar weight to incentivised secondary confession evidence as they do primary confessions. That appears true even where juries are aware of the incentives, with one mock trial study
53 At [205]–[210], citing Evidence Act, s 122(c) and (d).
56 At [216]–[218]. The safeguards the minority contemplated included further guidance for prosecutors and a central register of prison informants and how their evidence has been treated (including those informants’ criminal histories and the benefits they received for giving evidence).
58 At [221]–[232]. See Evidence Act, ss 45(1) (visual identification evidence), 28, 29 and 30(6) (false confession evidence), and 25(1) (expert evidence).
showing the presence of an incentivised witness nearly doubled the risk of wrongful conviction.59 Their Honours considered the literature supported the view that juries are predisposed to believe incentivised evidence, do not discount its reliability, and are likely to mistake the reasons witnesses give evidence.60 Cross-examination may be only an anaemic solution to those difficulties. Inmates are commonly practised liars, and jurors may view Crown witnesses as inherently credible. The minority therefore stressed “a need for caution at the point of admission of the evidence”.61
3. Proposed framework
The minority conceived a novel approach to the s 8 balancing exercise in respect of prison informant evidence. To ensure a defendant receives a fair trial, judges would need to be closely solicitous of such evidence, scrutinising how incentives “can operate to procure false evidence, and how this evidence can be concocted”.62 It considered s 8 provided a good framework within which judges could accommodate both incentivised evidence’s known association with miscarriages of justice, and the countervailing reality that many incentivised witnesses will be telling the truth. The minority emphasised its framework did not represent an “exhaustive list of relevant matters”. A case-by-case analysis would always be necessary.63 In the assessment of probative value, the minority saw the following factors as salient:
(a) the significance of the evidence to a matter at issue, the centrality of the issue, and how probative the evidence is of that issue;64
(b) indications the evidence is unreliable or untrue in its own terms;65
(c) whether the relevant evidence was incentivised (acknowledging the difficulties attendant on that assessment);66
(d) whether the witness giving evidence might have other motives to lie;67
(e) whether the witness giving evidence has a record of lying;68 and
(f) other circumstances relating to the collection of the evidence — for example, any indication the witness obtained the details given in his or her statement from earlier interactions with the Police.69
Finally, the minority turned to discuss its approach to unfair prejudice risks. It cautioned prison informant evidence could create unfair collateral prejudice in revealing or implying morally blameworthy conduct incidental to the issue at trial. It also said that where multiple incentivised witnesses give evidence, a jury might improperly read qualities of credibility and mutual corroboration into that sheer “weight of numbers”. The minority therefore confirmed it is necessary to consider the cumulative prejudicial effect of evidence when deciding whether its admission has
60 At [239]. See also the studies cited at nn 235–246.
caused a miscarriage of justice, as well as at the threshold predictive stage of admissibility. After all, juries do not assess items of evidence in isolation.70
The balance of the W v R decision sets out both the majority and the minority’s application of their respective approaches to the facts of the case. This note is not concerned with those particulars.71 In both W v R and Roigard v R, the Supreme Court dismissed the appeal by a 3–2 majority. It also dismissed the cross appeal in W v R.72
A. New Zealand’s approach to admissibility
The New Zealand position on prison informant evidence is now tolerably clear. On the majority’s approach, a court may consider reliability issues when assessing the probative value of prison informant evidence (as with all forms of evidence), and when weighing that value against the risk of unfair prejudice. Prison informant evidence invites the court’s careful scrutiny. The reliability assessment will be informed by considerations of credibility, the presence of incentives and the anticipated weight the factfinder will attach to the evidence. Judicial directions and other trial management measures can be marshalled to combat unfair prejudice risks. There is no presumption of inadmissibility.
The writer’s view is it is doubtful whether that approach goes far enough.73 The majority’s omission to adopt the minority’s carefully reasoned and detailed framework for incentivised secondary confession evidence was an arguable misstep.74 While the writer acknowledges the constitutional character of the question and the general desirability of weight being the preserve of the factfinder at trial, the distinction between “threshold reliability” and “ultimate reliability” remains useful.75 Prison informant evidence that is patently unreliable should not reach the jury, and judges’ gatekeeping role must be robust and meaningful. That is not to derogate from the principle that ultimate reliability of evidence remains for the factfinder. It is “not the
70 At [267]–[270], citing R v Howse [2003] NZCA 178; [2003] 3 NZLR 767 (CA) at [19]; Pickering v R [2012] NZCA 311,
[2012] NZCA 311; [2012] 3 NZLR 498 at [121]; R v Gooch [2009] NZCA 163 at [37]; and arguably departing from Hudson v R, above n 5, at [44].
71 Shortly stated, the majority, at [184] reached the same conclusion as had the Court of Appeal, finding admissible the evidence of six of the proposed witnesses. The minority, at [417], considered the evidence of only two of the proposed witnesses was admissible.
72 At [184] and [304]; and Roigard v R, above n 1, at [79].
73 Compare, however, Anna High’s view that the majority approach, as compared to earlier authority, shows the Supreme Court “intended to embolden judges to be more rigorous gatekeepers, at least with respect to [prison informant evidence]”: High, above n 8, at 229.
74 See W v R, above n 1, at [87]; and Roigard v R, above n 1, at [54]. The majority’s reluctance to
adopt the narrower framework was predicated on its assessment that some of the matters traversed in that framework were properly for trial and cross-examination.
75 High, above n 8, at 229.
law of evidence that commitment to the judge/jury divide requires complete and blind deference to the jury on all questions of reliability”.76
It is true that the majority approach recognises this principle, and arguably represents a more rigorous treatment of prison informant evidence than the adherence to orthodoxy seen in cognate jurisdictions. But it may say something about this nominally more rigorous approach that the majority in W v R reached identical decisions on the admissibility of the eight witnesses’ evidence to the Court of Appeal — the lower court having cleaved to the traditional approach of leaving reliability assessments to the jury.77 The minority, meanwhile, came to different decisions on the admissibility of the evidence of half of those witnesses.78 Is the majority approach, then, a substantive refinement of the existing law, or just a distinction without a difference? While it may be too soon to say definitively, there is some reason to doubt the newly settled approach proceeds on a materially more principled basis than did the orthodox approach.79
Would the minority approach represent a substantive departure from the existing position? One could argue that the majority’s broadly framed and nonprescriptive approach tacitly admits of the narrower considerations on which the minority focused. While the majority and minority drew different conclusions as to the admissibility of some of the inmate evidence, that fact may simply reflect a difference in the majority and minority’s weighting of the relevant considerations. On that interpretation, it is not obvious that there would be a significant difference in practice between the respective approaches.
The writer’s view, however, is that there are reasons to be sceptical of such a reading. The majority was plain that its approach to the s 8 balancing exercise differed to that of the minority. It expressly disclaimed the minority’s narrower framework: “we do not consider that the further formulation of considerations set out in the reasons of Winkelmann CJ and Williams J should be used”.80 The majority reiterated this disclaimer in Roigard v R, saying the effect of the minority approach in that case was to require independent corroboration of the disputed evidence. It saw the question whether the confession had been constructed by the witness to cohere with facts gained from other sources as a matter for trial and cross-examination, and properly to be put to the jury. It was unwilling to elevate independent corroboration to a requirement for admission.81
As well, in the writer’s view, the minority approach’s close prescription and particularisation of salient factors would discipline and focus the judicial gatekeeping role, allowing crisp and principled distinctions to be drawn between reliable and
77 R v Hoggart, above n 11, at [79]–[80].
78 W v R, above n 1, at [417].
79 See High, above n 8, at 229: “it remains to be seen whether this shift in emphasis, from jury assessment to gate-keeping judicial assessment of reliability, will translate to different outcomes in admissibility decisions”.
80 W v R, above n 1, at [87].
81 Roigard v R, above n 1, at [54].
unreliable evidence at the pre-trial stage. Its application would likely produce more consistent and certain outcomes in challenges to the admissibility of inmate evidence than would the majority’s wider approach to the s 8 balancing exercise. The more plausible view, then, is that the majority approach in W v R and Roigard v R is more permissive than the minority approach and sticks more closely to the orthodox position.
For those reasons, the writer sees the majority approach as placing too much faith in the ability of trial procedures and cross-examination to ameliorate the unique unfair prejudice risks prison informant evidence presents. It is not at all clear such measures are sufficiently responsive to those risks. To that extent, a searching examination of the evidence at the point of admission is preferable. It is regrettable that for all its evident utility, the minority’s stricter approach to admissibility forms no part of New Zealand law.
B. Scope for reform
In W v R, the Court was unanimous that New Zealand would benefit from additional legal controls on prison informant evidence. Commentators have likewise identified a warrant for legislative or policy reform. Bernard Robertson observes “[t]here is clearly potential for plenty of work by the Law Commission or the Ministry of Justice.”82 Anna High, meanwhile, notes “it is reasonable to assume that judges will continue to be influenced by the common law’s deep-rooted tradition of deference to juries on reliability”, meaning express amendment of the Act to mandate judicial screening of prison informant evidence may be needed.83 And Scott Optican states that New Zealand could helpfully introduce a dedicated statute setting out the circumstances in which prison informant testimony would be permitted at trial and the procedures for a pre-trial hearing regarding such evidence’s admissibility.84
What shape might prospective reform take? Patrick Anderson suggests, notwithstanding the usual principle of the rules of evidence that credibility and weight are jury matters, an exception should be made for prison informant evidence. Admissibility of such evidence should be cabined by principles like those governing hearsay evidence. That is, the evidence should be generally admissible, provided the circumstances underlying it provide reasonable assurance of its reliability. Both the W v R majority and minority approaches could be brought to bear on that assessment in pre-trial admissibility hearings.85
High, meanwhile, posits an express legislative exclusionary rule that would set a defined threshold of reliability for admission. That approach, by providing a safeguard mechanism under the Act, would bring prison informant evidence into line with the three other categories of evidence most commonly associated with miscarriages of justice: eyewitness (mis)identification evidence, (flawed) forensic evidence and (false)
82 Bernard Robertson “Cellmate confessions, relevance and probative value” [2020] NZLJ 399 at 399.
83 High, above n 8, at 232.
84 Optican, above n 7, at 585.
85 Patrick Anderson “Snitched on or stitched up? — a review of the law in New Zealand in relation to jailhouse informant evidence” [2021] NZLJ 155 at 161.
confession evidence. Prison informant evidence would be presumptively inadmissible. Admissibility would be determined pre-trial, with the Crown having to prove on the balance of probabilities that the circumstances relating to the evidence provide assurance it is reliable.86
The writer’s view is that prison informant evidence can be sufficiently controlled without resort to a presumption of inadmissibility. Instead, Parliament could amend the Act to require close judicial scrutiny of such evidence at the point of admission. That scrutiny would usefully be guided by express statutory criteria tracking the minority’s multifactorial framework in W v R (or some near equivalent that contemplates the full spectrum of considerations at play when dealing with incentivised evidence). As well, there would be value in New Zealand exploring broader safeguards, including the establishment of an informant register or an independent auditing body to vet the evidence before it comes to court.87
C. Solicitor-General’s Guidelines
There are promising indications the prison informant cases have promoted a more textured treatment of such evidence in New Zealand’s criminal justice system. The Solicitor-General has promulgated a set of guidelines to assist prosecutors in deciding to call inmate evidence.88 The document is a useful (if anodyne) restatement of the key principles emerging from the two Supreme Court decisions. The Guidelines stipulate that “it is appropriate that the Crown takes an especially careful approach to the decision to call [inmate evidence]”.89 Prosecutors should only adduce the evidence if satisfied it is more likely than not reliable. Inmate evidence is likely to be appropriate only in serious cases, and it is unlikely to be in the public interest to prosecute based only on the uncorroborated evidence of a prison informant.90
The Guidelines require prosecutors inquire into the reliability of the evidence.91 Relevant information includes the proposed witness’s previous convictions; current sentence(s) and remaining term; history of giving prison informant evidence; past treatment of that witnesses’ evidence and the reason for that treatment; the significance of the evidence used to the matter(s) then at issue; and any benefit or other preference the witnesses received. Prosecutors should reassess their decisions to call prison informant evidence if information becomes available that affects the original assessment of that evidence.92 The Guidelines also provide a (non-exhaustive) set of relevant factors. In broad terms, they are:93
(a) motive;
(b) circumstances of alleged interactions;
(c) other confirmatory evidence;
86 High, above n 8, at 232–233.
88 Crown Law “Solicitor-General’s Guidelines for Use of Inmate Admissions” (6 August 2021).
(d) opportunity to concoct; and
(e) character and state of the witness.
It is to be hoped the Guidelines promote prosecutors’ judicious use of prison informant evidence. However, recognising the minority’s point in W v R that prosecutors remain susceptible to human biases, the Guidelines seem unlikely, on their own, to mitigate that evidence’s associated risks.
D. Criminal justice incentives
The categories of incentivised evidence are open. In both W v R and Roigard v R, the Court was explicit it was dealing only with the evidence of Crown-incentivised prison inmates. It expressly did not offer a view on other forms of incentivised evidence — such as Police informants or people who respond to Police offers of monetary reward.94 While those matters were without the scope of the two appeals, in the writer’s view they may usefully be considered in future cases. It would also be interesting to see judicial consideration of whether, and the extent to which, non- institutional incentives operate within New Zealand’s criminal justice system. Such incentives could conceivably exist within criminal networks or between inmates.
It may be unsafe to assume criminal justice incentives operate only in favour of the prosecution. Notionally, an inmate could be incentivised to appear as a defence witness and to give fabricated secondary confession evidence tending to exculpate that defendant (for example, by claiming a different inmate had confessed their responsibility for the relevant crime). The incentives operating in that black market would be less tangible and probably less compelling than the corresponding Crown incentives, but could include securing protection in prison, increasing status with a particular element of the prison population or the promise of reciprocal exculpatory testimony. Would evidence given in response to those incentives be susceptible of the same robust gatekeeping approach espoused in W v R and Roigard v R?
The writer’s preliminary view is such evidence should not attract the anxious scrutiny warranted in cases involving Crown-incentivised witnesses. A number of the reliability concerns raised in that context (principally including the strength and breadth of Crown incentives and the perceived inherent credibility of Crown witnesses) would be much attenuated, or non-existent, where it is the defence proposing to adduce the evidence. As well, miscarriage of justice concerns would arguably be less salient where criminal justice incentives operate in the defendant’s favour.95 On that view, resort to standard trial management techniques, judicial direction and cross-examination would seem sufficient safeguards.
As against that observation, however, it is relevant to note the unfair prejudice test concerns the proceeding as a whole. It is not limited to unfair prejudice to the defendant. And, while the incentives for an inmate to give exculpatory evidence are
94 W v R, above n 1, at [87]; and Roigard v R, above n 1, at [55].
95 Recall Blackstone’s ratio: “[i]t is better that ten guilty persons escape than that one innocent suffer”. See William Blackstone Commentaries on the Laws of England in Four Books (George Sharswood (ed), JB Lippincott Co, Philadelphia, 1893) vol 2 at 358.
different from those to give inculpatory evidence, so too are the consequences. An informant giving evidence implicating an accused will do so having weighed the consequences of giving of that evidence against the benefit they get from giving it. While an informant giving exculpatory evidence may have fewer incentives to fabricate, he or she may not suffer the same consequences from doing so — the evidence would not implicate someone on trial. It remains an open question, then, whether the informant giving exculpatory evidence (anticipating no consequences) is just as prone to give unreliable testimony as the informant giving inculpatory evidence. Conclusive determination of this issue is for another day.
W v R and Roigard v R have settled the New Zealand approach to the admissibility of prison informant evidence. The majority approach, while countenancing a strong gatekeeping role for judges, errs on the side of reserving reliability assessments to the factfinder. While there may be constitutional propriety in that approach, the unique and manifest problems with this category of evidence, in the writer’s view, compel circumspection. Even on the nominally more rigorous majority approach, there exists marked potential for unsafe convictions.
The minority’s granular approach to admissibility, meanwhile, has much to commend it. It shows a greater sensitivity to the fair trial considerations implicated where prosecutors seek to adduce prison informant evidence. That the minority approach would exclude evidence pre-trial more readily than the majority approach does not entail judges usurp the constitutional role of the factfinder. The general rule that reliability assessments are best directed to the jury is already hedged about with qualifications and exceptions.96 The prevailing concern should be that good quality evidence is put before the jury, affording it a good foundation on which to come to a principled verdict.
The case for legislative intervention is compelling. There is no obvious reason inmate admissions should not be treated as a bespoke category of evidence, attended by a specific statutory reliability threshold. But there is a need to proceed carefully. There remains much to be established about criminal justice incentives and how jurors treat incentivised evidence in real life criminal trials.97 In particular, there are no relevant New Zealand studies.98 Until more empirical data becomes available, admissibility best practices remain uncertain. What is clear is any reform will need to ensure defendants’ rights to a fair trial without unduly limiting a powerful tool in the prosecutor’s armoury
— a line not easily drawn.
96 See, for example, the bespoke categories of evidence set out at above n 58. Note also that “deference to the jury on reliability issues is not absolute” in comparator jurisdictions: High, above n 8, at 229.
97 See the minority’s observation that studies on juror treatment of prison informant evidence have been based on artificial situations: W v R, above n 1, at [239].
NZLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.nzlii.org/nz/journals/NZCrimLawRw/2020/1.html