Tendency, relevance and the legacy of the common law

The Supreme Court has recently released a number of rulings in an attempted murder trial from last year - DPP v Hills, NC, Cooper and RAC. Ruling No 6 concerns the operation of the UEA and warrants close consideration. The case alleged by the prosecution was that a group of four people went to the vicitm's house, abducted her, drove her to the Maribyrnong River and attempted to drown her. They desisted when they believed that another car was approaching. The Crown case relied in part on the evidence of one co-offender, Meulenbrock, who agreed to give evidence against the other co-offenders.

Ruling No 6
In Ruling No 6 [2010] VSC 486, one of the accused, Hill, sought to cross-examine Meulenbrock concerning past violent acts involving Meulenbrock and the other accused. The purpose of this was to show that Hill was not involved on that earlier occasion and so was less likely to involved with whatever violent acts her co-accused engaged in on this occasion.

Counsel for Hill sought to characterise this as relationship evidence and restrictions on admitting tendency evidence in s97 of the UEA. The other defendants objected on the ground that the evidence was impermissible tendency evidence, was irrelevant, went to a collateral issue and was unfairly prejudicial.
Kaye J rejected the accused's application to lead the evidence. First, he accepts that the evidence is not tendency evidence:

21 It is well established that evidence of prior criminal conduct, by an accused, is not relevant to establish, generally, a propensity to violence by that accused. Rather, in order to qualify as tendency evidence under s 97 it needs to be demonstrated that there is a relevant degree of similarity between the previous conduct and the events with which the court is concerned; see GBF v The Queen. There is no striking or other similarity between the manner in which the assault of 25 November occurred, and the manner in which the offences at 9 December were carried out. That is, the previous event of 25 November does not disclose any modus operandi, nor was it attended by any particular hallmark, which it relevantly shares with the incident of 9 December. Furthermore, the incident of 25 November was but one single event. In general, the courts are slow to accept that offending, or conduct, on one occasion, is significantly probative of a fact, for the purposes of s 97(1), unless there are significant or remarkable similarities between the previous incident and the incident in question.

The judge then considered whether the evidence was otherwise relevant to a fact in issue, holding:

26 It is not put on behalf of Karen Hills that the three young people were parties in some type of conspiracy, existing exclusively themselves, to commit violent criminal offences. Rather what is sought to be put to Ms Meulenbrock is that in the course of their friendly relationship, the three young people, from time to time, violently assaulted other persons. In my view, it has not been demonstrated, in any way, how such evidence could rationally affect the probabilities of the case between the Crown and Karen Hills. It would seem that it is sought to be demonstrated that there was a relationship between Meulenbrock and the two co-accused, to the exclusion of a relationship with Karen Hills, which involved them in criminal activity. If that is the basis upon which it is put that the evidence is relevant, then that basis is patently flawed. Clearly, the fact that Karen Hills did not participate in the events of 25 November 2008 could not, rationally, affect the probabilities as to whether she was involved in the events of 9 December 2008. As Mr Horgan correctly submitted, any line of reasoning that, because Karen Hills was not present at the incident of 25 November 2008 (or at any other such prior incident) involving Meulenbrock, NC and Cooper, it is less likely that she was present at the events of 9 December 2008, would be a plain non sequitur. Such reasoning would be patently illogical.

27 Indeed, on analysis, it would seem that, notwithstanding the protestations of Ms Randazzo to the contrary, what is sought to be elicited, from that evidence, is the existence of a relationship between the three persons in which, from time to time, they were prone to indulge in violent criminal activity together, but without the involvement, or participation, of an adult. Viewed in that light, the evidence is impermissible tendency evidence which, Ms Randazzo correctly conceded, is not admissible.

Discussion
Several points should be made about this ruling. First, on the facts, it seems plainly sensible that the evidence not be admitted. The reasoning relied on by the accused Hill is tenuous at best and involves the introduction of a form of cut-throat defence that is likely to distract the jury more than the probative value of the evidence warrants. That said, the ruling takes an approach to the question of relevance that appears different to the intention of the drafters of the UEA and the main textbooks on the question.

In the drafting of the UEA, the ALRC sought to separate out the questions of logical and legal relevance that had infused the common law and introduce a binary test of relevance - Evidence either can rationally affect the probability of a fact in issue or it cannot. How much it does so is a matter to be considered at the discretionary exclusion stage. Paragraph 21 (extracted above) appears to confuse logical relevance and legal admissibility.

In my view, the evidence sought to be adduced is logically capable of demonstrating a general tendency towards violence. As a matter of common understanding, a person who has assaulted 5 people in 5 different circumstances is a violent person. No similarity between the 5 assaults (type of victim, weapon used, circumstances of the assault, etc) is necessary before one may draw the conclusion that the person in question has a tendency towards violence. Degree of similarity is a matter that was important to the common law of similar fact evidence (e.g. R v Papamitrou (2004) 7 VR 375), and continues to be relevant under the UEA when measuring whether the evidence has significant probative value (PNJ v R [2010] VSCA 88). But it is not a matter that affects the question of logical relevance. Logical relevance recognises that a tendency to commit particular types of acts may be significantly probative of whether a person committed that kind of act on another occasion (see, e.g. GBF v R [2010] VSCA 135 at [26]). The difficulty arises because Kaye J seeks to ask whether the evidence is tendency evidence before asking whether it is relevant. Instead, he should have asked if the evidence is relevant and, if so, how is it relevant. This would have analysed the existence of any logical connection between the evidence and a fact in issue in the trial.

The problem with analysing whether the evidence is tendency evidence before asking if it is relevant becomes especially apparent at paragraphs 26 and 27. The prosecution's argument that the evidence is a non sequitur seems to me to be unconvincing. An argument that "I had nothing to do with X's criminal escapades. Indeed, here are a bunch of previous occasions where X engaged in criminal conduct without my involvement" isn't a very strong argument for why you couldn't have started now, but I think that it is stretching the point to say that such reasoning is just irrational. Indeed, it seems to me to rely on the same sort of prediction of future conduct from past conduct that is involved in the standard good character direction which asserts, in part, that "I've never committed a crime before, so I'm less likely to start now" (most recently re-endorsed in Braysich v R [2011] HCA 14). To conclude that the evidence is not relevant as tendency evidence at paragraph 21 and then to conclude at paragraph 27 that it is only relevant as inadmissible tendency evidence (because of irrelevance) seems to me to be internally inconsistent.

The case shows the importance of keeping relevance and exclusionary rules separate when applying the UEA. In this case, the defence didn't need to contend with the exclusionary rule in s101 of the UEA that required the probative value to significantly exceed the prejudicial effect. Instead, it only needed to show that the evidence had substantial probative value (a matter that Kaye J seems to treat as an afterthought at the end of paragraph 21). Kaye J's judgment appears heavily influenced by the common law, but overlooks the fact that even the common law recognises that the exclusion of similar fact evidence isn't because the evidence is logically irrelevant, but because of the risk that the evidence would be taken to prove too much (HML v R (2008) 235 CLR 303 per Gleeson CJ at [12]). The result is that he imposed stricter exclusionary rules on the defence than appear necessary and introduced confusion between the capacity of the evidence to rationally influence a decision as a matter of logic, and the legal restrictions on admitting evidence under ss97 and 135.

Is this a cavil at imprecision in the reasoning? Perhaps, but where the Evidence Act introduces small variations compared to the common law, there is a natural tendency for practitioners and judges to assume
that the common law is a safe guide, and that is a tendency that should be guarded against.