The High Court has allowed an appeal against a decision of the Northern Territory Court of Criminal Appeal on complaint and tendency evidence and probative value in the context of child sexual assault. The appellant was convicted of sexually abusing his step-grandchild on three occasions. The NTCCA upheld the trial judge’s decision to admit evidence from the complainant’s friend and relatives, to direct the jury that if they were satisfied of that complaint evidence they could use it as ‘some evidence that an offence did occur’, and to admit tendency evidence from the complainant about the appellant’s conduct during a massage that indicated his sexual interest in her. Before the High Court the appellant argued that the NTCCA and the trial judge misconstrued the meaning of tendency evidence and probative value, as contained in ss 97 and 137 of the Evidence (National Uniform Legislation) Act (NT) (the NT enactment of the Uniform Evidence Law) respectively.
The Court unanimously allowed the appeal, quashed the appellant’s conviction and ordered a new trial. The majority (French CJ, Kiefel, Bell and Keane JJ) held that the appellant’s arguments on the error in assessing the tendency evidence for probative value should be accepted, but that the grounds of appeal on the complaint evidence were not made out. The majority held that in addressing probative value, a trial judge must assume that the jury will accept the evidence, and consequently questions about its credibility or reliability do not arise: at [39]:
The question as to the capability of the evidence to rationally affect the assessment of the probability of the existence of a fact in issue is to be determined by a trial judge on the assumption that the jury will accept the evidence. This follows from the words “if it were accepted”, which are expressed to qualify the assessment of the relevance of the evidence. This assumption necessarily denies to the trial judge any consideration as to whether the evidence is credible. Nor will it be necessary for a trial judge to determine whether the evidence is reliable, because the only question is whether it has the capability, rationally, to affect findings of fact. There may of course be a limiting case in which the evidence is so inherently incredible, fanciful or preposterous that it could not be accepted by a rational jury. In such a case its effect on the probability of the existence of a fact in issue would be nil and it would not meet the criterion of relevance.
Both ss 97(1)(b) and 137 require the judge to assess the probative value of the evidence, defined as evidence ‘which could rationally affect … the assessment of the probability of the existence of a fact in issue’ (at [42]). For the purposes of s 97, the question is whether the probative value is ‘significant’, and for s 137, whether its probative value outweighs the danger of unfair prejudice to the defendant: in both cases, it requires taking the evidence at ‘its highest in the effect it could achieve on the assessment of the probability of the existence of the facts in issue’ (at [43]–[49]). The majority preferred the approach of the New South Wales Court of Criminal Appeal (see R v XY and R v Shamouil) based in the reasoning of Gaudron J in Adam v The Queen [2001] HCA 57, over the approach taken by the Victorian Court of Appeal (see Dupas) (at [50]): determining the weight to be given to evidence, such as its credibility or reliability, will depend on the whole picture of the evidence and the examination and cross-examination of witnesses (at [51]–[52]):
Once it is understood that an assumption as to the jury’s acceptance of the evidence must be made, it follows that no question as to credibility of the evidence, or the witness giving it, can arise. For the same reason, no question as to the reliability of the evidence can arise. If the jury are to be taken to accept the evidence, they will be taken to accept it completely in proof of the facts stated. There can be no disaggregation of the two – reliability and credibility – as Dupas v The Queen may imply. They are both subsumed in the jury’s acceptance of the evidence.
As applied in this case, and contrary to the conclusions of the NTCCA, the tendency evidence did not have significant probative value and was not admissible under s 97: the evidence about the massage was capable of showing that the appellant had a sexual interest in the complainant, but here the probative value goes to the evidence’s ability to support the credibility of the complainant’s account, and ‘without more, it is difficult to see how a complainant’s evidence of conduct of a sexual kind from an occasion other than the charged acts can be regarded as having the requisite degree of probative value’ (see [60]–[64]). As to the complaint evidence, the majority held that it was potentially significant, did not have low probative value, and did not create the prejudice in s 137, and also rejected the appellant’s arguments that there was a risk of the jury misusing it or giving it more weight than it deserved (see [65]–[74]).
Gageler J agreed with the majority that the tendency evidence was improperly admitted but the complaint evidence properly admitted (at [101]ff). Like Nettle and Gordon JJ, however, Gageler J preferred the approach based in the judgment of McHugh J in Papakosmas v The Queen [1999] HCA 37 over that of Gaudron J in Adam (though not by construing the Evidence Act against the common law background, as Nettle and Gordon JJ did: see [95]). The omission of a reference to the evidence needing to be accepted in the statute was, in McHugh J’s view, significant because it suggested that an assessment of probative value would necessarily involve assessing reliability (see at [85] and [96]ff). For Gageler J, this meant the trial judge applied the wrong approach to probative value (at [100]) and applying what Gageler J considered to be the correct test would have resulted in the tendency evidence being rejected for lack of significant probative value (see at [102]–[108]), but the complaint evidence being accepted (see at [109]–[111]).
Nettle and Gordon JJ also held that the appeal should be allowed, but, contrary to the majority, because of the wrongful admission of the complaint evidence, rather than the tendency evidence, which they considered was correctly admitted. Contra the majority (but like Gageler J), Nettle and Gordon JJ followed the approach of the Victorian Court of Appeal, founded on the approach of McHugh J in Papakosmas that assessing probative value requires an assessment of reliability. That approach was to be preferred because (at [140], and see at [172]):
Evidence cannot affect the assessment of the probability of the existence of a fact in issue unless the evidence is rationally capable of being accepted. Hence, to determine whether evidence has the capacity rationally to affect the assessment of the probability of the existence of a fact in issue requires a determination of whether the evidence is rationally capable of acceptance. And for the court to determine whether it thinks that evidence is rationally capable of acceptance requires the court, among other things, to determine whether it thinks that the degree of reliability which it would be open to the jury rationally to attribute to the evidence is such that it will be open to the jury rationally to accept the evidence. It follows that, according to ordinary principles of statutory construction, there is no warrant for reading s 97 or the definition of ‘probative value’ in the Dictionary to the Act as involving an assumption that evidence will be accepted.
After discussing Shamouil, Dupas and XY (see [149]ff), Nettle and Gordon JJ outlined the arguments for assessing reliability and credibility under the sections (see [155]–[171]), concluding that there was no reason to depart from the natural and ordinary meaning of the sections, which in their Honours’ view required a trial judge to have regard to both reliability and credibility of the evidence and determine the weight that a jury could give to the evidence and hence ‘the extent to which that evidence could rationally affect the assessment of the probability of the existence of a fact in issue’: at [172]. As applied here, the tendency evidence relating to the massage, while it may have been weakened (see [180]), could have had significant probative value, and thus was correctly not excluded (see [173]–[181]). The trial judge erred, however, in not excluding the complaint evidence under s 137 because the judge failed to have regard to its credibility and reliability, which may have been called into question due to the delays and circumstances surrounding the complaint (at [183]–[184]), and that failure may have led to a miscarriage of justice (at [186]). Consequently, Nettle and Gordon JJ also allowed the appeal, quashed the convictions under challenge, and ordered a new trial.
| High Court Judgment | [2016] HCA 14 | 14 April 2016 |
| Result | Appeal allowed | |
| High Court Documents | IMM | |
| Full Court Hearing | [2016] HCATrans 8 | 3 February 2016 |
| Special Leave Hearing | [2015] HCATrans 266 | 16 October 2015 |
| Appeal from NTCCA | [2014] NTCCA 20 | 19 December 2014 |
| Trial Judgments |
[2013] NTSC 45 | 7 August 2013 |
| [2013] NTSC 44 | 3 July 2013 |
Great summary Martin. Putting on my evidence law hat, I’m generally curious about readers’ views on the identification example the majority gives at [50]:
I struggle to see the difference between the two approaches described (see also Gageler J at [92].) If it helps, here is the referenced discussion by Mr Heydon:
Absolutely, if evidence of that kind is to have a diminished probative value because it is ‘unconvincing’ won’t this necessarily be because there are issues with the circumstances of the identification that tend to diminish its reliability? A second issue is the apparent inconsistency between the assumption that the jury will accept the factual correctness of the evidence and the possibility that they might be unconvinced by it.
If a judge is required to assume the reliability and credibility of the evidence and, therefore, that the jury will accept that evidence, all evidence of positive identification should really have an equal probative value. This is at least before one considers the place of the identification within the body of evidence as a whole, but that seems to me to be another issue. If a person gives evidence that they “saw the Chief justice”, and a judge is required to assume that the jury will accept that evidence, on what basis could a judge find that they might be unconvinced by that evidence?
One of the reasons I suspect that the majority ties themselves in knots trying to answer this question is that it seems so contrary to logic and common sense to hold that an identification that occurs in broad daylight, at a short distance, and between persons who are known to each other could rationally effect the assessment of the probability of the existence of a fact in issue (identity) to no greater extent than evidence of an identification that occurred at some distance, on a foggy night, while the witness was drunk and where the identified party was both a stranger to the witness and a member of a racial group with which they had no previous involvement.
Brett Walker also struggled with this issue in oral submissions when the question was put to him by French CJ (at p 63). What Walker tries to do to resolve the matter is distinguish between objective facts upon which a person might base an identification (“I saw a silhouette that resembled the Chief Justice”) and the ultimate conclusion or opinion (“I saw the Chief justice”). Where the objective facts or observations founding the identification are vague or generalised those facts or observations, even if accepted as factually correct, would have a reduced probative value. Fair enough.
The problem is that the distinction isn’t of consequence on the majority view because the conclusion itself is a piece of direct evidence. In the end, if in assessing the probative value of the identification a judge is required to assume that a jury will accept the factual correctness of the witness’ evidence that they ‘saw the Chief Justice’, what does it matter that they had a poor foundation for that conclusion: the judge assumes that the jury will accept it warts and all.
I mostly agree with Patrick – That example is very troubling and it does not sit well with the rest of the majority judgment.
Where I disagree is with his third paragraph. I don’t find it troubling at all that for this purpose, an identification made in ideal circumstances is of equal probative value to one made in dire circumstances. But that is because the exercise is an artificial one.
The question, on the majority’s construction, is the extent to which the evidence, if accepted, is capable of affecting the determination of the fact in issue. The difficulty lies in whether you consider that juries engage in Bayesian reasoning or a two-stage binary approach, where the jury considers whether to accept evidence as a precondition to deciding the extent to whether the evidence makes a difference. On a Bayesian approach, the defects in the evidence reduce the extent to which the evidence can rationally affect the answer, whereas the binary approach treats the defects as controlling whether the jury accepts the evidence at all.
As Patrick identifies, the problem is that the example, and the possible solution posed by reference to Bayes’ theorem, is at odds with the rest of the majority’s conclusion – that the evidence is only probative if accepted, and that requires discarding issues of credibility of reliability as matters for the jury to consider in determining whether to accept the evidence at all. And this means that while circumstantial evidence can have varying levels of probative value based on the extent to which it supports the conclusion contended for, direct evidence of a relevant fact in issue will, it seems to me, universally have very high levels of probative value because the probative value attaches to the evidence if objectively correct, rather than the value of the evidence as a claim from a particular individual.
The other aspect which troubles me is the application of this line of reasoning to the tendency evidence. If evidence is taken as true for the purpose of measuring probative value, and questions of credit and reliability are discarded in this process, it seems that there is a tension in saying that the evidence has little probative value because it can’t rise higher than the credit of the complainant. If you took the so-called narrow approach seriously, surely the question is whether the facts asserted (the other acts which constitute the tendency evidence) are capable of substantially affecting the probability that the accused committed the charged acts? The identity of who asserted those facts (in this case, the same person who is asserted the charged acts) would be irrelevant, as would the fact that there is no external supporting evidence of those other acts. It would simply measure the degree of relevance of the facts asserted.
I wonder whether this application of the majority’s approach to excluding considerations of reliability and credibility sheds some light on just what that means. But it may take several articles or appellate decisions to quite unearth the distinction the majority draws between factors that do affect probative value and factors which are relevant to credibility or reliability which do not.