The Law Commission is calling for submissions on its second statutory review of the Evidence Act 2006. It released an Issues Paper (Second Review of the Evidence Act 2006 / Te Arotake Tuarua i te Evidence Act 2006: He Puka Kaupapa) at the end of March, canvassing the operational issues that have arisen and suggesting a number of options for reform.

One of the provisions the Commission is examining is section 28, which sets out the “reliability rule”. Where there is an evidential foundation for raising an issue about the reliability of a defendant’s statement, the judge is required to exclude it unless satisfied the circumstances in which it was made were not likely to have affected its reliability.

The current wording of section 28 does not say whether the truth of the statement can be considered when assessing its reliability. What happens, then, if a person prone to delusions makes a statement in which he confesses to murder and identifies the location of the victim’s body, which had previously been unknown to the police? Should the fact that the victim’s body was located be taken into account when assessing the reliability of his statement?

When the Commission developed its Evidence Code (which formed the basis for the Evidence Act) in the 1990s, it proposed a reliability rule that would have made the truth of the statement an irrelevant consideration. The Commission’s rationale was that it would usurp the function of the jury for truth to be established at the preliminary admissibility stage. Although the select committee that considered the Evidence Bill 2005 agreed with this view, an express provision to that effect ultimately was not included in the Act.

Conflicting decisions

Following its enactment, section 28’s silence on whether truth was relevant to the reliability inquiry led to conflicting decisions (compare R v Cameron [2007] NZCA 564 and R v Edmonds [2009] NZCA 303, [2010] 1 NZLR 762 with R v Cameron [2009] NZCA 87), and prompted the Law Commission to consider the operation of the provision during its first review of the Act. In its report, The 2013 Review of the Evidence Act 2006 (NZLC R127, 2013), the Commission recommended an amendment to section 28 that would require truth to be disregarded. This time, the Commission explained it was important to guard against the risk of “diverting the court’s attention from questions of improper police conduct to large volumes of corroborating evidence”. But that recommendation was not accepted by the government. A blanket rule requiring truth to be disregarded was thought to be too restrictive – for example, where a confessional statement revealed an aspect of the crime that no one but the offender could know.

The Supreme Court subsequently considered the issue in R v Wichman [2015] NZSC 198, [2016] 1 NZLR 753, in the context of determining the admissibility of a confession arising out of an undercover operation. Four of the judges (William Young, Arnold and O’Regan JJ; and Glazebrook J) concluded that truth was relevant when determining reliability under section 28. They suggested it would be incongruous for an obviously true confession to be excluded on the basis of a theoretical likelihood that the circumstances in which it was made may have affected it reliability. The Chief Justice disagreed, for reasons similar to those expressed by the Commission in the past.

Does the rationale hold up?

However, the Commission has started to move away from the view that the actual reliability of a statement should be disregarded under the section 28 inquiry. In its recent Issues Paper, the Commission reviews the history of section 28 and questions whether the rationale put forward for not inquiring into truth under the pre-Evidence Act law holds up under the rule that was ultimately enacted.

The Law Commission explains that, under the pre-Act law (a combination of the common law “voluntariness” rule and a limited statutory exception), confessions were excluded if they were obtained by an inducement from a “person in authority”, unless the inducement was not likely to cause an untrue admission of guilt to be made. The inquiry focused on the likelihood that police improprieties may have caused an untrue admission, not on the truth or falsity of the actual confession. The rationale for not inquiring into the truth of the statement was to avoid any suggestion that police impropriety should be encouraged. The focus on external inducements also meant that “internal” factors potentially affecting the reliability of the confession (for example, consumption of alcohol or mental illness) were disregarded (as these were factors the police may not be able to control or be aware of).

The reliability rule proposed by the Commission in its Evidence Code, however, differed from the previous law in two important respects: the rule did not require the statement to have been obtained by a “person in authority”; and it anticipated that “internal” factors could affect reliability. These changes were intended to reflect what the Commission described at the time as the “primary purpose” of the reliability rule: to screen out statements that would be unsafe to put before a jury because of the risk of unreliability. The Commission explained the rule was necessary because of the danger that a jury would give too much weight to a statement that came directly from a defendant (in particular, a confession). That danger justified some judicial oversight.

In short, the pre-Act rationale for disregarding truth (control of police practices) sits somewhat awkwardly with the expanded reliability rule in the Evidence Act – the policy justification for excluding statements under section 28 is because of concerns about reliability, not improper police conduct. Concerns about police misconduct can, of course, be dealt with under other provisions in the Act (section 30 and, in extreme cases, section 29). The Commission acknowledges this in its Issues Paper (recognising it had occasionally described section 28 as promoting additional policies beyond that of reliability in the past), as well as the logic in permitting the actual reliability of a statement to be considered under a provision that is, after all, concerned with “reliability”. It is now inviting submissions on a possible amendment to allow the truth of the statement to be considered.

Formulating a workable rule

If the Commission ultimately makes this recommendation, its challenge over the coming months will be in formulating a workable rule – one that is both easy to apply and does not result in judges conducting a “mini-trial” at the threshold admissibility stage.

For example, the Commission will need to consider whether any brakes should be placed on the judge’s ability to consider the consistency (or inconsistency) of the statement with evidence that forms part of the prosecution’s case:

Should only independent real evidence that is discovered as a consequence of the statement (for example, the location of a murder victim’s previously undiscovered remains) be considered?

What about information for which there is no other way of gaining access other than involvement with the offence (for example, a statement identifying elements of the crime that have not been made public)?

Another important consideration is the permissible scope of questioning during a pre-trial hearing or voir dire that is held to determine admissibility – in particular, whether cross-examination of the defendant on whether his or her statement is true should be permitted. Although section 28 involves only a threshold admissibility question, cross-examination on the truth of the statement at this preliminary stage is not unprecedented – in R v Patten (HC Auckland CRI-2006-004-3200, 8 April 2008) the prosecution was permitted to cross-examine the defendant about the truth of the statement he made to the police.

Of the four judges in Wichman who concluded that truth was relevant in the section 28 inquiry, only Glazebrook J addressed the point. Her view was that cross-examination on the truth of the statement should not be permitted: the prosecution should put to the defendant for comment the matters that would be relied on to indicate the statement was reliable, but resist an extensive challenge to a defendant’s assertion that the statement was false. The exercise was to “assess the contents of the statement and any obvious indications of reliability or unreliability in relation to other aspects of the case and any subsequently discovered evidence” (at [535]). For that reason she specifically disagreed with the approach that had been taken in Patten.

Submissions on the Commission’s Issues Paper can be made until 15 June 2018. The Commission is required to provide its final recommendations to the Government by the end of February 2019.

Yasmin Moinfar-Yong ymoinfar-yong@lawcom.govt.nz is a legal and policy adviser with the Law Commission.