Tendency and coincidence evidence

27.170 The following section discusses aspects of the law of evidence concerning the admissibility of ‘tendency’ and ‘coincidence’ evidence, as defined under the uniform Evidence Acts, and ‘propensity’ or ‘similar fact’ evidence at common law. These forms of evidence may include, significantly, the evidence of other complainants who have allegedly been sexually assaulted by the same defendant.

27.171 For example, evidence may be adduced to show that, because the defendant engaged in sexual activity with one child in his or her family, the defendant has a tendency to commit such acts. This evidence may have probative value in relation to allegations of sexual assault against other children in the family. Similarly, evidence may be adduced to show that a defendant engaged in sexual activity with two children in similar circumstances—for example, when another parent was absent. The evidence about one allegation may have probative value in relation to the other because it is improbable that the events were coincidental.

27.172 Tendency and coincidence evidence may also include, for example, evidence of prior convictions for sexual offences or other prior illegal sexual conduct—often referred to as ‘uncharged acts’—such as ‘grooming’ behaviours.

Current law

27.173 Under the uniform Evidence Acts—applicable in NSW, Victoria, Tasmania and the ACT—evidence about a defendant’s prior illegal sexual conduct may be characterised as ‘tendency’ or ‘coincidence’ evidence.[212]

27.174 Tendency evidence is evidence ‘of the character, reputation or conduct of a person, or a tendency that a person has or had’, adduced to prove that the person ‘has or had a tendency (whether because of the person’s character or otherwise) to act in a particular way, or to have a particular state of mind’.[213]

27.175 Coincidence evidence is evidence that ‘2 or more events occurred’ that is adduced to prove that ‘a person did a particular act or had a particular state of mind on the basis that, having regard to any similarities in the events … it is improbable that the events occurred coincidentally’.[214]

27.176 The test for admissibility under the uniform Evidence Acts is whether the tendency or coincidence evidence has significant probative value which substantially outweighs any prejudicial effect it may have on the defendant.[215]

27.177 At common law, evidence of these kinds is referred to as either ‘propensity’ or ‘similar fact’ evidence, depending on the purpose for which it is admitted. In jurisdictions that do not apply the uniform Evidence Acts, the admissibility of these forms of evidence is governed by the common law, as modified by statute in some jurisdictions.

27.178 At common law, as a matter of general principle, evidence of past criminal conduct, including convictions (for sexual offences or otherwise) is not admissible unless there is a ‘striking similarity’ or ‘underlying unity’ between the similar facts.[216] Admissibility is governed by the test in Pfennig v The Queen—that is, propensity or similar fact evidence is admissible if its probative value is such that there is no rational view of the evidence that is consistent with the innocence of the accused.[217]

27.179 Some common law evidence jurisdictions have enacted legislative provisions modifying common law rules relating to propensity or similar fact evidence:

  • in Queensland, similar fact evidence must not be ruled inadmissible on the ground ‘that it may be the result of collusion or suggestion, and the weight of that evidence is a question for the jury’;[218]

  • in Western Australia, the Evidence Act 1906 (WA) governs the admissibility of propensity evidence by incorporating a ‘significant probative value’ test and a public interest test;[219]

  • in South Australia, propensity evidence relating to a count of sexual assault is only admissible in a joint trial if it has relevance beyond mere propensity. However, in deciding admissibility, the judicial officer is not to have regard to ‘whether or not there is a reasonable explanation in relation to the evidence consistent with the innocence of the accused or whether or not the evidence may be the result of collusion or concoction’.[220]

Impact in sexual assault proceedings

27.180 The impact of these rules of evidence depends on the type of trial and the type of evidence the prosecution seeks to adduce. If there are two or more complainants who have allegedly been sexually assaulted by the same defendant—for example, two daughters of the defendant, the prosecution will often make a pre-trial application to have the counts against the defendant heard in a joint trial, rather than separate trials. As discussed in Chapter 26, the threshold question for holding a joint trial is whether each complainant’s evidence will be cross-admissible in respect of the charges involving the other complainant or complainants.

27.181 If there is only one complainant, the prosecution may have evidence from other witnesses about the defendant’s criminal sexual behaviour with them. Alternatively, it may wish to adduce relationship evidence to explain the nature of the relationship between the complainant and the defendant, and the context in which the sexual assault occurred.[221]

27.182 Three aspects of the law of evidence concerning the admissibility of tendency and coincidence evidence are problematic in sexual assault cases. These are:

  • the ‘striking similarities’ test;

  • the ‘no rational view of the evidence’ test; and

  • excluding ‘a reasonable possibility of concoction’.

The ‘striking similarities’ test

27.183 At common law, the cross-admissibility of the evidence of two or more complainants is, at the outset, dependent on the evidence revealing ‘striking similarities’.[222]The cross-admissibility of evidence frequently arises in the child sexual assault context where it has been argued that the ‘striking similarities’ test ‘has been used to create artificial distinctions in relation to sex offender behaviour’.[223]

27.184 This artificiality is said to arise because ‘when it comes to assessing the probative value of the evidence of child complainants, there will be higher or lower degrees of similarity depending on what a judge knows about sex offending behaviour’.[224] For example, a trial judge who is unaware of the range of grooming and sexual practices of child sex offenders ‘may look for greater degrees of similarity in the evidence compared to a judge who is aware of the variety of ways in which … offenders gain a child’s trust and affection and the different sexual practices a serial offender will engage in’.[225]

27.185 A lower threshold for determining probative value may be ‘appropriate in child sexual assault cases where the identity of the offender is not in issue, in order to capture the range of sexual and grooming behaviours of serial offenders’.[226] Since most cases of sexual assault, especially in a family violence context, involve defendants known to the complainant, rather than strangers, the identity of the accused will not usually be a fact in issue. It is also argued that, when applying the ‘striking similarities’ test, many cases have failed to recognise that any sexual practices with children ‘ought to be the sufficient similarity requirement since sex offenders will engage in different sexual conduct with different children depending upon the age and sex of the child, the passivity or resistance of the child, the degree of grooming of, and the degree of access to, the child, the defendant’s relationship with the child and so on’.[227]

27.186 Cases such as R v KP; Ex parte Attorney-General (Qld)—which involved evidence of sexual assaults against two brothers by a school music teacher and friend of the family (and against two of the defendant’s own children)—illustrate how rules on the admissibility of propensity and similar fact evidence can frustrate legislative attempts to increase the number of joint trials.[228] This is because a decision to hold a joint trial is dependent on the cross-admissibility of each complainant’s evidence which, in turn, is dependent on finding sufficient similarities in the evidence. Recently, the culmination of the tension between cross-admissibility and joint or separate trials occurred in the High Court case of Phillips v The Queen. Since this decision, the applicability of the ‘striking similarities’ test has been reinforced as the standard for determining the admissibility of similar fact evidence at common law.[229]

27.187 Although the uniform Evidence Acts create a different regime for admitting tendency and coincidence evidence, it has been argued that the striking similarities test is still used in assessing the probative value of the evidence of two or more complainants about a defendant’s sexual conduct.[230] For example, the NSW Court of Criminal Appeal has noted the similarity between the approach under the Evidence Act 1995 (NSW) and the common law.[231] Common law formulations, including ‘striking similarities’ have been held to ‘guide’ the reasoning process when determining whether tendency or coincidence evidence has significant probative value under ss 97 and 98 of the uniform Evidence Acts.[232]

The ‘no rational view of the evidence’ test

27.188 At common law, even if the evidence of two or more witnesses has ‘striking similarities’, it can still be excluded because of its prejudicial effect. In order to prevent the admission of prejudicial propensity and similar fact evidence, the common law developed the ‘no rational view of the evidence test’,[233] adopted by a majority of the High Court in Hoch v The Queen:

to determine the admissibility of similar fact evidence, the trial judge must apply the same test as a jury must apply in dealing with circumstantial evidence, and ask whether there is a rational view of the evidence that is inconsistent with the guilt of the accused.[234]

27.189 The test was confirmed by a majority of the High Court in Pfennig v The Queen,[235] which held that the probative force of similar fact evidence will outweigh its prejudicial effect only if there is no rational view of the evidence that is consistent with the innocence of the accused.[236]

27.190 This test is now referred to as the ‘Pfennig test’ and is the means for determining how the probative force of the evidence and prejudicial effect should be balanced.[237] Where there is a rational view of the evidence consistent with the defendant’s innocence, the probative force of the evidence is automatically outweighed by its prejudicial effect, thus removing any discretion on the part of the trial judge to admit the evidence.

27.191 Probative force and prejudicial effect must also be considered under s 101(2) of the uniform Evidence Acts. In R v Ellis, the NSW Court of Criminal Appeal considered whether the Pfennig test should continue to be applied.[238] The fact that s 101(2) introduced a legislative formulation for balancing probative value against prejudicial effect led Spigelman CJ to conclude that it did not. He stated that the continued application of a ‘no rational view’ test is not ‘consistent with a statutory test which expressly requires a balancing process’ and that the reasoning in Pfennig is ‘inapplicable to a statutory test that probative value substantially outweighs prejudicial effect’.[239]

A reasonable possibility of concoction

27.192 It has been argued that, since the decision in R v Ellis, ‘the necessity to exclude the possibility of collusion or other influence is questionable’ when applying the balancing test under s 101(2) of the uniform Evidence Acts.[240] Recent case law indicates, however, that R v Ellis has not removed the issue of concoction from the admissibility equation under the uniform Evidence Acts.

27.193 Despite expectations that the courts would continue to develop the law away from Hoch v The Queen,[241]obviating the need for legislative intervention, these expectations may have been ‘an overly optimistic view’.[242] It has been argued that, in sexual assault cases, s 101(2) of the uniform Evidence Acts is only likely to be satisfied ‘if a reasonable possibility of concoction can be eliminated’.[243]

27.194 In R v Ellis, Spigelman CJ accepted that there ‘may well be cases where, on the facts, it would not be open to conclude that the probative value of particular evidence substantially outweighs its prejudicial effect, unless the “no rational explanation” test were satisfied’,[244] and that these cases may include those where a trial judge considers that there is a reasonable possibility of concoction.[245]

27.195 The Tasmania Law Reform Institute (TLRI) has observed that, in practice, where a reasonable possibility of concoction is taken into account under s 101(2), this is no different from an application of the ‘no rational view of the evidence’ test, because concoction ‘weighs so heavily in the balance that the reality is that its existence means that there is no balancing to be undertaken’. The position, in the TLRI’s view, ‘effectively remains the same as that rejected … in Ellis’.[246]

27.196 When a judge takes concoction or contamination into account, he or she can be seen as performing the task of the jury, by assessing the strength of the evidence—in particular, whether or not it is true.[247] The TLRI noted that ‘no other exclusionary rule requires the court to determine the veracity of evidence as a basis for admission’.[248]

27.197 The application of the ‘no rational view of the evidence’ test in sexual assault trials is seen to have ‘particular consequences, since where there are two or more complainants, the probative force of the similar fact evidence is destroyed if there is a reasonable possibility of concoction between the complainants’.[249] The courts do not, however, apply a universal standard for measuring what amounts to a reasonable possibility of concoction.

27.198 Where two or more children give evidence about a defendant’s sexual behaviour with them, the admissibility of their evidence may depend on whether they had the opportunity to concoct their allegations. While the targeting and grooming strategies of serial sex offenders are well documented in the literature,[250] the rules governing the admissibility of propensity evidence are based on the belief that, if two or more complainants know each other, the possibility of concoction must be ruled out for one complainant’s evidence to be admissible in the case of another.

27.199 This assumption is based on the common law’s traditional belief that children are unreliable witnesses, prone to fantasy and highly suggestible[251]—a view that may still prevail despite the fact that it is not supported by the psychological literature.[252] Recent literature is said to show that ‘children are highly resistant to abuse suggestions and do not readily make up stories of sexual abuse even when given the opportunity to do so’.[253]

27.200 Further, the test for concoction does not take into account the practical effects on complainants, since in assessing the possibility of concoction the court will usually conduct a pre-trial hearing, at which the complainants may be required to give evidence.[254]

Reform options

27.201 Several jurisdictions have enacted legislation to overcome problems caused in sexual assault cases by the ‘no rational view of the evidence’ test and with evidence being excluded on the basis of possible concoction.

27.202 In Queensland, s 132A of the Evidence Act 1977 (Qld) governs the admissibility of similar fact evidence and leaves the question of concoction to the jury:

In a criminal proceeding, similar fact evidence, the probative value of which outweighs its potentially prejudicial effect, must not be ruled inadmissible on the ground that it may be the result of collusion or suggestion, and the weight of that evidence is a question for the jury, if any.

27.203 In addition, s 597A(1AA) of the Criminal Code (Qld) provides that, in determining whether there should be joint or separate trials, ‘the court must not have regard to the possibility that similar fact evidence, the probative value of which outweighs its potentially prejudicial effect, may be the result of collusion or suggestion’.

27.204 In South Australia, reforms concerning the prosecution of sexual assault charges were enacted in 2008,[255] following a review of rape laws.[256] Under s 278(2a) of the Criminal Law Consolidation Act 1935 (SA), issues of joinder, the ‘no rational view of the evidence’ test and concoction are dealt with together. The provision creates a presumption that where there are two or more counts involving complainants, the counts are to be tried together. A judge may only order separate trials if the evidence of two or more complainants is not cross-admissible.[257]

27.205 The ‘no rational view of the evidence’ test has been specifically abrogated in relation to determining whether evidence will be cross-admissible. Section 278(2a)(c) states that in determining admissibility for these purposes:

(i) evidence relating to the count may be admissible in relation to another count concerning a different alleged victim if it has a relevance other than mere propensity; and

(ii) the judge is not to have regard to—

(A) whether or not there is a reasonable explanation in relation to the evidence consistent with the innocence of the defendant; or

(B) whether or not the evidence may be the result of collusion or concoction.

27.206 In Western Australia, s 31A of the Evidence Act 1906 (WA) deals with propensity evidence.[258] Section 31A(2) provides:

(2) Propensity evidence or relationship evidence is admissible in proceedings for an offence if the court considers:

(a) that the evidence would, either by itself or having regard to other evidence adduced or to be adduced, have significant probative value; and

(b) that the probative value of the evidence compared to the degree of risk of an unfair trial, is such that fair-minded people would think that the public interest in adducing all relevant evidence of guilt must have priority over the risk of an unfair trial.

(3) In considering the probative value of evidence for the purposes of subsection (2) it is not open to the court to have regard to the possibility that the evidence may be the result of collusion, concoction or suggestion.

27.207 Like the Queensland reforms, s 31A(3) removes the issue of concoction from the admissibility equation and leaves it to the jury when deciding the weight to be given to the evidence.[259] The possibility of concoction cannot be taken into account when determining whether the propensity evidence has significant probative value.[260] This means that the court has to take the evidence at ‘its highest’.[261] Nor can the possibility of concoction be taken into account when the court is applying the balancing test under s 31A(2)(b), in particular the risk of an unfair trial, because of the ‘quite clearly articulated legislative purpose’ of the provision.[262]

27.208 Section 31A abrogates the common law ‘no rational view of the evidence’ test. The test introduced by s 31A(2)(b) is taken directly from the balancing test suggested by McHugh J, dissenting in Pfennig v The Queen.[263] This test is much less stringent than the Pfennig test.[264] In enacting s 31A, the legislature has accepted that, because of its nature, the admission of propensity evidence will always create the risk of an unfair trial.[265] However, that risk must be balanced against the competing public interests in holding a joint trial in which all relevant evidence is admitted.[266]

27.209 The balancing test, when referring to fair-minded people, clearly envisages that the public interest in the prosecution of sexual offences ought to be taken into consideration when weighing up the risk of prejudice to the accused. These considerations may be important, especially in relation to child sexual assault, since a particular complainant’s evidence will often make more sense when evidence is given of the alleged serial nature of a defendant’s sexual behaviour. For example, in VIM v Western Australia,[267] separate allegations of sexual assault over many years had been made by the two step-daughters of the accused, who was indicted on 44 counts. The Western Australian Court of Appeal considered that this was ‘an example of the very type of case in which the legislature intended the jury to have the benefit of a full evidentiary familial picture’.[268]

27.210 None of the reforms enacted in Australia has addressed expressly the striking similarities test, which may still constitute a major barrier to the admissibility of tendency and propensity evidence and the ability to hold joint trials. The Western Australian reforms represent the most complete break from the common law. By articulating a particular threshold test of admissibility—that the evidence must have significant probative value—the common law striking similarities test may no longer be ‘a necessary criterion for admissibility’.[269]

27.211 The National Child Sexual Assault Reform Committee has reviewed the effectiveness of the reforms discussed above. It concluded that the most successful appear to be those enacted in Western Australia[270]—although in the absence of empirical research it is not possible to determine the extent to which, for example, the number of joint trials has increased in Western Australia as a result of those reforms.

Consultation Paper

27.212 In the Consultation Paper, the Commissions proposed that federal, state and territory legislation should provide that, in sexual assault proceedings, a court should not have regard to the possibility that the evidence of a witness or witnesses is the result of concoction, collusion or suggestion when determining the admissibility of tendency or coincidence evidence.[271]

27.213 The Consultation Paper also stated that further consideration may need to be given to the continued reliance of Australian courts on the striking similarities test for the admission of tendency, coincidence, propensity and similar fact evidence (including in uniform Evidence Acts jurisdictions). The Commissions asked to what extent the ‘striking similarities’ test impedes the ordering of joint trials in relation to sex offences; and whether the Western Australian reforms in relation to the cross-admissibility of evidence should be adopted in other jurisdictions.[272]

Submissions and consultations

27.214 Stakeholders expressed a range of views relating to the need for reform to address the implications of the striking similarities test. The perspectives of legal stakeholders differed, however, about the extent to which the test is being applied, in practice, in uniform Evidence Acts jurisdictions. A number also emphasised the impact of the issue within the family violence context.

27.215 The NSW ODPP submitted that the test is a ‘flawed approach to establishing the modus operandi of most sexual predators, but particularly in a family violence context and other instances where there is no issue as to the identity of the offender’. The NSW ODPP noted that rather than acting with a ‘hallmark’, there is more likely to be a ‘progression or an evolution’ in the offending behaviour—for example, in the case of a father who assaults several of his children.[273] It stated that:

the striking similarities test does impede the ordering of joint trials. The need to satisfy the test for the purposes of relying on tendency or coincidence evidence is one thing, but if the evidence fails that test it does not follow that the evidence is not relevant and probative in other ways.[274]

27.216 Cossins considered that the striking similarities approach focuses on a need to find that ‘the sexual practices of the defendant with one complainant are more or less identical to his sexual practices with another complainant’ before their evidence will be considered to be tendency or coincidence evidence and cross-admissible in a joint trial.[275] In her opinion, in uniform Evidence Acts jurisdictions,

like the common law, it is only where there are sufficient similarities between the accounts given by the complainant and another witness that that witness’ evidence will be considered to have significant probative value under the [uniform Evidence Acts].[276]

27.217 She submitted that the striking similarities test has been applied by the NSW Court of Criminal Appeal in a number of child sexual assault cases and is problematic for three key reasons:

(i) Because the test does not accord with information about child sex offender behaviour described in the psychological literature, it amounts to a false measure for assessing the probative value of evidence about a defendant’s sexual behaviour.

(ii) Corroborative evidence of child sexual abuse is uncommon which is likely to be one of the reasons why low conviction rates are found in child sexual assault trials. Thus the striking similarities test is likely to contribute to low conviction rates where its effect is to exclude relevant corroborative evidence which is judged to be insufficiently similar with the evidence of the complainant.

(iii) If there is a lack of striking similarities between the evidence of two or more complainants, a joint trial is less likely to be held. This, in turn, will result in more separate trials with a decreased likelihood of serial sex offenders being convicted.[277]

27.218 Cossins proposed eliminating the ‘striking similarities’ test from the admissibility equation under the uniform Evidence Acts. She suggested that provisions be inserted stating that, in respect of a sexual offence, if two or more counts charging sexual offences involving different complainants are joined in the same indictment, the court must not have regard to whether that evidence has striking similarities with other evidence about the sexual conduct of the defendant, in considering its probative value.[278]

27.219 The need for any reform in relation to the striking similarities test, at least in NSW, was contested by other stakeholders.[279] The Public Defenders Office NSW submitted that NSW courts are not bound to a ‘striking similarities’ test, but rather a rational assessment of probative value under the Evidence Act 1995 (NSW). Further, it was submitted that the distinction between the positions in NSW and Western Australia with regard to the admissibility of tendency (or propensity evidence) may not be as significant as suggested by the Consultation Paper. In particular, in NSW as in Western Australia,[280] ‘a distinctive system of grooming and exploitation will often have more sway than similarity of the physical acts involved’.[281]

27.220 Stakeholders also held divergent views on the desirability of reforms to evidence laws to ensure that the possibility of concoction does not, by itself, render inadmissible the evidence of witnesses in sexual assault proceedings. Some stakeholders, including legal professionals, supported the Consultation Paper proposal.[282]

27.221 The NSW ODPP stated that allegations of concoction, collusion or suggestion are better assessed by a jury than rendered inadmissible. Otherwise these matters become a significant barrier to the joinder of charges in respect of multiple complainants because:

  • it is very difficult to exclude a reasonable possibility of concoction if the complainants are well known to each other, particularly siblings;

  • this issue arises in all institutional sexual assault allegations such as involving teachers, because invariably the victims know each other as they were all at school together;

  • there is not a great deal of case law in this area, because many prosecutors do not run counts together if they involve siblings because it is almost impossible to exclude the reasonable possibility of concoction …[283]

27.222 Women’s Legal Services NSW submitted:

The unique features of sexual assault offences and child sexual assault offences require, in our view, unique legislative solutions. These solutions must reflect the realities for many sexual assault complainants: complainants, particularly children, often know each other and often have some form of connection or relationship. This connection is most often not indicative of concoction.[284]

27.223 The Law Council stated that it would not oppose a provision that stated that ‘the possibility’ of concoction should not be a basis on which to hold tendency or coincidence evidence inadmissible—as long as the provision did not preclude concoction from being taken into account in determining admissibility under relevant tests.[285]

27.224 Other stakeholders opposed the proposal.[286] National Legal Aid observed that if ‘the possibility of concoction, collusion or suggestion cannot be excluded, the probative value of the evidence is properly diminished’.[287]

27.225 The Public Defenders Office NSW submitted that there was insufficient cause ‘for taking the radical step of pulling one area of criminal prosecution outside the uniform framework of evidentiary rules’. The Public Defenders noted that specific consideration needs to be given to whether the evidence is being put forward as tendency or coincidence evidence. That is, if the evidence is coincidence evidence ‘the reduction in probative value where there is a real possibility of contamination has nothing to do with assumptions based on the common law’s traditional belief that children are unreliable witnesses, prone to fantasy, and highly suggestible’. Rather, it was submitted that:

it is a realistic assessment of the difference in the force of the argument ‘What are the chances of two people coming up with a similar allegation?’ where the complainants know one another. Logic requires recognition that such an argument has greater force for the Crown where the complainants have never met nor heard of each other’s claims.[288]

Commissions’ views

27.226 The application of rules of evidence applying to tendency and coincidence evidence has an important impact on sexual assault proceedings, including in relation to sexual assault committed in a family violence context. These rules of evidence may render some evidence inadmissible and mean that, where two or more complainants have allegedly been sexually assaulted by the same defendant, a joint trial may not be possible.[289]

27.227 In ALRC 102, the ALRC, NSWLRC and VLRC highlighted the difficult task of formulating appropriate rules to deal with probative but prejudicial evidence. In relation to tendency and coincidence evidence, the Commissions recognised ‘a stark contrast between the policy objectives of receiving all probative evidence, and minimising the risk of wrongful conviction’.[290]

27.228 In child sexual assault cases, however, it seems unfair to victims and their families that:

in so many cases the isolation of one child pitted against an adult alleged to be the perpetrator leads to acquittal of the adult, when at the same time there are other allegations of similar behaviour against the adult from other family members not before the Court, or when a history of such offending is known but excluded, or when the conduct is part of an alleged wider course of conduct, but evidence of which for one reason or another is excluded.[291]

27.229 In common law evidence jurisdictions, such perceived problems with the operation of rules of evidence in sexual assault proceedings have led to legislative reform. In the Commissions’ view, there is no reason to recommend reform in uniform Evidence Acts jurisdictions to respond to the ‘striking similarities’ or ‘no rational view of the evidence’ (Pfennig) tests—as has occurred in Western Australia and South Australia. There is a case, however, for specific reform in relation to the impact on admissibility of evidence of a ‘reasonable possibility of concoction’.

Striking similarities

27.230 Although there have been a number of reforms in different jurisdictions to increase the frequency of joint trials in relation to sex offences, only one common law jurisdiction (Western Australia) has abandoned striking similarities as the test for admissibility of propensity evidence.

27.231 It has been suggested that, without changes in the application of the striking similarities test, including in uniform Evidence Acts jurisdictions, reforms to increase the number of joint trials may be undermined.[292] The striking similarities test may also impede the prosecution’s ability to adduce evidence about the defendant’s sexual conduct from witnesses in trials which involve one complainant.

27.232 In Ellis, the NSW Court of Criminal Appeal confirmed that the tendency and coincidence rules in the uniform Evidence Acts should be construed according to their own terms and excluding prior common law principles (such as the striking similarities test).[293] Rather, the test for the admissibility of tendency and coincidence evidence is whether the evidence has significant probative value substantially outweighing any prejudicial effect.[294]

27.233 Similarity or dissimilarity are still commonly referred to in assessing probative value in applying this statutory balancing test—as are other concepts such as ‘pattern’. For example, in R v Fletcher,[295] the NSW Court of Criminal Appeal, in finding that certain evidence was not admissible, stated that there was ‘in the matter now before the Court, insufficient pattern or underlying unity; there are no striking similarities or unusual features … ’.[296]

27.234 The NSW cases since Ellis confirm the continuing role that the existence of similarities and dissimilarities has in assessing the probative value of tendency and coincidence evidence. They do not show, however, that any strict ‘striking’ similarities test continues to be applied.[297]The use of the concept of similarity seems unavoidable in construing the uniform Evidence Acts tendency and coincidence provisions. The coincidence rule itself refers to ‘similarities’ in the events or circumstances about which evidence is sought to be adduced.[298] This may not, however, prevent development of the law to recognise that evidence of patterns of behaviour or systematic activities (such as those relating to ‘grooming’) may have significant probative value, even where there is no close similarity in the physical acts involved.[299]

27.235 In the Commissions’ view, there is insufficient reason to recommend reform in uniform Evidence Acts jurisdictions to address perceived over-reliance on ‘striking similarities’ as a test for the admissibility of tendency or coincidence evidence. The real question is whether the tendency and coincidence rules should continue to apply to sexual assault proceedings involving multiple complainants or entirely different rules developed for this particular category of evidence. The Commissions are not convinced that a case has been made out for such special rules of evidence applicable only in sexual assault proceedings. Such rules would risk introducing complexity and uncertainty in uniform Evidence Acts jurisdictions.

27.236 Further, the Commissions do not consider it would be appropriate, in the context of this Inquiry, to recommend reforms to evidence law in common law evidence jurisdictions directed to the striking similarities test. Rather, the Commissions consider it is preferable that all Australian jurisdictions join the uniform Evidence Acts scheme.

No rational view of the evidence

27.237 The National Child Sexual Assault Reform Committee suggested reforms to overcome the effect of the ‘no rational view of the evidence’ test in jurisdictions which have not already introduced reforms to do so. The suggested reform would expressly eliminate ‘whether or not there is a reasonable explanation in relation to the evidence consistent with the innocence of the defendant’ as an issue when a court is required to decide whether to order joint or separate trials. [300]

27.238 The test does not apply in uniform Evidence Acts jurisdictions and has been overridden by legislation in Western Australia and South Australia. It may still apply in Queensland and the Northern Territory. The Commissions do not consider it would be appropriate, in the context of this Inquiry, to recommend specific reforms to evidence law in those two jurisdictions. Again, the Commissions consider it is preferable that all Australian jurisdictions join the uniform Evidence Acts scheme.

Reasonable possibility of concoction

27.239 In the Commissions’ view, in sexual assault proceedings it is not appropriate for the possibility of concoction to render evidence inadmissible—for example, when the prosecution seeks to call a witness who, while not a complainant, can give evidence about the defendant’s sexual behaviour with him or her and has had contact with the complainant. This may occur in cases of sexual assault in a family violence context.

27.240 The current law appears to be that a reasonable possibility of concoction can affect the admissibility of propensity and similar fact evidence in common law evidence jurisdictions. At common law, the possibility of concoction will almost certainly render propensity and similar fact evidence inadmissible.

27.241 Under the uniform Evidence Acts, concoction is a factor that the court may consider in assessing probative value and applying the balancing test under ss 97, 98 and 101 respectively. However, it appears that case law may be evolving towards exclusion of evidence of multiple complainants where there is a possibility of concoction.[301]

27.242 In NSW, while some stakeholders submitted that the concoction issue does not dominate consideration of admissibility,[302] others considered that judicial officers tend towards excluding evidence where possible concoction is raised. Prosecutors perceived a resulting high bar to joint trial in such cases.[303] In practice, it is difficult to exclude a reasonable possibility of concoction where, for example, complainants are siblings.

27.243 The National Child Sexual Assault Reform Committee has made recommendations to address this barrier to the admission of tendency and coincidence (propensity and similar fact) evidence.[304] The Commissions agree that there is a case for reform.

27.244 The Commissions recommend that federal, state and territory legislation should provide that, in sexual assault proceedings, tendency or coincidence evidence is not inadmissible only because there is a possibility that the evidence is the result of concoction, collusion or suggestion.

27.245 In uniform Evidence Acts jurisdictions, the appropriate mechanism would be amendment of Part 3.6 of the Acts—for example, by the insertion of a new section following s 101 (which deals with restrictions on tendency evidence and coincidence evidence adduced by the prosecution). This recommendation for reform of the uniform Evidence Acts would need to be considered by the Australian and state and territory governments through SCAG.

27.246 The Commissions’ preference would be for all Australian jurisdictions to join the uniform Evidence Acts scheme. Failing this, however, common law evidence jurisdictions that have not already done so should enact similar amendments in criminal procedure or evidence legislation providing that propensity evidence and similar fact evidence must not be ruled inadmissible only because there is a possibility that the evidence is the result of concoction, collusion or suggestion.

Recommendation 27–13 Federal, state and territory legislation should provide that, in sexual assault proceedings, tendency or coincidence evidence is not inadmissible only because there is a possibility that the evidence is the result of concoction, collusion or suggestion.

[212] Uniform Evidence Acts, pt 3.6.

[213] Ibid s 97.

[214] Ibid s 98.

[215] Ibid ss 97, 98, 101.

[216]Pfennig v The Queen (1995) 182 CLR 461.

[217] Ibid, 485; see also HML v The Queen (2008) 235 CLR 334.

[218]Evidence Act 1977 (Qld) s 132A.

[219]Evidence Act 1906 (WA) s 31A(2).

[220]Criminal Law Consolidation Act 1935 (SA) s 278(2a).

[221] Relationship evidence is discussed later in this chapter.

[222]Phillips v The Queen (2006) 225 CLR 303.

[223] A Cossins, Alternative Models for Prosecuting Child Sex Offences in Australia (2010), prepared for the National Child Sexual Assault Reform Committee, 202.

[224] Ibid.

[225] Ibid.

[226] Ibid.

[227] Ibid, 195. See, eg, the facts in AE v The Queen [2008] NSWCCA (in which the complainant was the defendant’s step-father); R v KP; Ex parte Attorney-General (Qld) [2006] QCA; R v Fletcher (2005) 156 A Crim R 308.

[228]R v KP; Ex parte Attorney-General (Qld) [2006] QCA. The Queensland Court of Appeal held that the evidence of the older brother about being sexually abused by the defendant was not cross-admissible in relation to the counts involving the younger brother because of lack of similarities in the defendant’s alleged sexual behaviour. The Court held that charges relating to all complainants should not have been joined.

[229]Phillips v The Queen (2006) 225 CLR 303.

[230] A Cossins, Striking Similarities between the Common Law and the Uniform Evidence Acts: Protecting Serial Offenders and Putting Children at Risk, unpublished (2010), 14.

[231]R v F [2002] NSWCCA 125, [28].

[232]R v Fletcher (2005) 156 A Crim R 308, [60]. See also AE v The Queen [2008] NSWCCA; R v Milton [2004] NSWCCA; R v Harker [2004] NSWCCA; R v F (2002) 129 A Crim R 126; R v WRC (2002) 130 A Crim R 89.

[233]Sutton v The Queen (1984) 152 CLR 528, 564.

[234]Hoch v The Queen (1988) 165 CLR 292, 296.

[235]Pfennig v The Queen (1995) 182 CLR 461, 482–483.

[236] McHugh J considered that the test enunciated by Mason CJ, Deane and Dawson JJ in Pfennig was too stringent, although His Honour recognised that ‘where the prosecution case depends entirely on propensity reasoning, the evidence will need to be so cogent that, when related to the other evidence, there is no rational explanation of the prosecution case that is consistent with the innocence of the accused’: Ibid, 530.

[237]R v Ellis (2003) 58 NSWLR 700, [85]; HML v The Queen (2008) 235 CLR 334.

[238]R v Ellis (2003) 58 NSWLR 700.

[239] Ibid, [89]. The High Court granted Ellis leave to appeal but later revoked that leave, stating that it agreed with Spigelman CJ’s construction of the Evidence Act 1995 (NSW): Ellis v The Queen [2004] HCA Trans 488.

[240] Tasmania Law Reform Institute, Evidence Act 2001 Sections 97, 98 & 101 and Hoch’s Case: Admissibility of Tendency and Coincidence Evidence in Sexual Assault Cases with Multiple Complainants, Issues Paper 15 (2009), 2.

[241] As implied, eg, by Australian Law Reform Commission, New South Wales Law Reform Commission and Victorian Law Reform Commission, Uniform Evidence Law, Report 102, NSWLRC Report 112, VLRC FR (2005), [11.65]–[11.68].

[242] Tasmania Law Reform Institute, Evidence Act 2001 Sections 97, 98 & 101 and Hoch’s Case: Admissibility of Tendency and Coincidence Evidence in Sexual Assault Cases with Multiple Complainants, Issues Paper 15 (2009), 38.

[243] ‘[A]nd if there are sufficient similarities (striking or otherwise) between the evidence of two or more witnesses to be able to conclude that the probative value of the evidence outweighs its prejudicial effect’: A Cossins, Striking Similarities between the Common Law and the Uniform Evidence Acts: Protecting Serial Offenders and Putting Children at Risk, unpublished (2010), 19–20.

[244]R v Ellis (2003) 58 NSWLR 700, [96]. In R v Gibbs (2004) 146 A Crim R 503, Gray J followed Spigelman CJ’s reasoning by applying the ‘no rational view of the evidence’ test to coincidence evidence.

[245] A Cossins, Striking Similarities between the Common Law and the Uniform Evidence Acts: Protecting Serial Offenders and Putting Children at Risk, unpublished (2010), 19. See, eg, Tasmania v S [2004] TASSC, [11]; L v Tasmania (2006) 15 Tas R 381, [40]; Tasmania v Y (2007) 178 A Crim R 481, [40].

[246] Tasmania Law Reform Institute, Evidence Act 2001 Sections 97, 98 & 101 and Hoch’s Case: Admissibility of Tendency and Coincidence Evidence in Sexual Assault Cases with Multiple Complainants, Issues Paper 15 (2009), 11.

[247] Ibid, 39. See also, K Arenson, ‘The Propensity Evidence Conundrum: A Search for Doctrinal Consistency’ (2006) 8 University of Notre Dame Law Journal 31, 37.

[248] Tasmania Law Reform Institute, Evidence Act 2001 Sections 97, 98 & 101 and Hoch’s Case: Admissibility of Tendency and Coincidence Evidence in Sexual Assault Cases with Multiple Complainants, Issues Paper 15 (2009), 39. This point was noted by McHugh J in Pfennig v The Queen (1995) 182 CLR 461, 517.

[249] A Cossins, Alternative Models for Prosecuting Child Sex Offences in Australia (2010), prepared for the National Child Sexual Assault Reform Committee, 184.See Hoch v The Queen (1988) 165 CLR 292, 397.

[250] A Cossins, Alternative Models for Prosecuting Child Sex Offences in Australia (2010), prepared for the National Child Sexual Assault Reform Committee, 181–182.

[251] Australian Law Reform Commission and Human Rights and Equal Opportunity Commission, Seen and Heard: Priority for Children in the Legal Process, Report 84 (1997), 304; Hoch v The Queen (1988) 165 CLR 292.

[252] As previously recognised by the ALRC and the Human Rights and Equal Opportunity Commission: Australian Law Reform Commission and Human Rights and Equal Opportunity Commission, Seen and Heard: Priority for Children in the Legal Process, Report 84 (1997), 306–307.

[253] A Cossins, Alternative Models for Prosecuting Child Sex Offences in Australia (2010), prepared for the National Child Sexual Assault Reform Committee, 185.

[254] This is said to conflict with ‘public policy objectives behind preventing child complainants from giving evidence on more than one occasion which has seen some jurisdictions enact provisions to prevent children being required to give evidence at committal proceedings’: Ibid, 185. Complainants are also likely to be cross-examined ‘more aggressively’ at a pre-trial hearing because of the absence of a jury: Tasmania Law Reform Institute, Evidence Act 2001 Sections 97, 98 & 101 and Hoch’s Case: Admissibility of Tendency and Coincidence Evidence in Sexual Assault Cases with Multiple Complainants, Issues Paper 15 (2009), 14.

[255]Criminal Law Consolidation (Rape and Sexual Offences) Amendment Act 2008 (SA).

[256] L Chapman, Review of South Australia Rape and Sexual Assault Law: Discussion Paper (2006), prepared for the Government of South Australia.

[257] In contrast, the Victorian provision (discussed above in relation to a presumption of joint trials) attempts to encourage joint trials even where evidence is not cross-admissible:Criminal Procedure Act 2009 (Vic) s 194.

[258] The definition of propensity evidence under s 31A is more extensive than at common law and includes tendency, relationship and character evidence.

[259]Di Lena v Western Australia (2006) 165 A Crim R 482, [51]; Wood v Western Australia [2005] WASCA, [41]; Donaldson v Western Australia (2005) 31 WAR 122, [108].

[260]Donaldson v Western Australia (2005) 31 WAR 122, [154].

[261] Ibid, [153] (Roberts-Smith JA).

[262] Ibid, [157] (Roberts-Smith JA).

[263]Pfennig v The Queen (1995) 182 CLR 461, 528.

[264]Di Lena v Western Australia (2006) 165 A Crim R 482, [94].

[265] Ibid, [58].

[266] The decision made under s 31A is a question of law, not an exercise of a judicial discretion and, once admitted, there is no room to exclude the evidence under the discretion at common law: Ibid,[60], referring to R v Christie [1914] AC.

[267]VIM v Western Australia (2005) 31 WAR 1.

[268] Ibid, 168.

[269] A Cossins, Alternative Models for Prosecuting Child Sex Offences in Australia (2010), prepared for the National Child Sexual Assault Reform Committee, 202. The Western Australian Court of Appeal has accepted that, in a child sexual assault trial, propensity evidence need not show ‘striking similarities’ or sexual interference by the defendant in a particular way: Donaldson v Western Australia (2005) 31 WAR 122, [149]. If the evidence in question reveals an underlying unity, system or pattern, that will be sufficient to establish that the evidence has significant probative value under s 31A(2) ‘irrespective of what physical acts [were] individually involved’. In Donaldson, the evidence of the four complainants showed that the defendant, a swimming coach, had a particular pattern of conduct ‘or a tendency … by essentially similar means, to inveigle young girls under his charge into situations in which he then commits sexual offences upon them’: [149]. This was sufficient for the evidence of the four complainants to be cross-admissible.

[270] A Cossins, Alternative Models for Prosecuting Child Sex Offences in Australia (2010), prepared for the National Child Sexual Assault Reform Committee, 186–203.

[271] Consultation Paper, Proposal 18–10.

[272] Ibid, Questions 18–7, 18–8.

[273] Office of the Director of Public Prosecutions NSW, Submission FV 158, 25 June 2010.

[274] Ibid.

[275] This approach, she considers, has no empirical basis and appears to ‘reflect a subjective belief that sex offenders are highly specialised’ in the type of sexual behaviour they engage in, a view that is not supported by the research literature: A Cossins, Submission FV 112, 9 June 2010.

[276] Ibid.

[277] Ibid.

[278] Ibid.

[279] National Legal Aid, Submission FV 232, 15 July 2010; Public Defenders Office NSW, Submission FV 221, 2 July 2010.

[280] See, eg, Donaldson v Western Australia (2005) 31 WAR 122.

[281] Public Defenders Office NSW, Submission FV 221, 2 July 2010.

[282] National Association of Services Against Sexual Violence, Submission FV 195, 25 June 2010; Canberra Rape Crisis Centre, Submission FV 172, 25 June 2010; Office of the Director of Public Prosecutions NSW, Submission FV 158, 25 June 2010; Commissioner for Children (Tas), Submission FV 62, 1 June 2010. The Law Council stated that it ‘did not oppose’ the suggestion: Law Council of Australia, Submission FV 180, 25 June 2010.

[283] Office of the Director of Public Prosecutions NSW, Submission FV 158, 25 June 2010.

[284] Women’s Legal Services NSW, Submission FV 182, 25 June 2010.

[285] Law Council of Australia, Submission FV 180, 25 June 2010. That is, it should still be inadmissible where ‘the evidence is, for example, overwhelming that two complainants have jointly prepared their complaints’.

[286] National Legal Aid, Submission FV 232, 15 July 2010; Public Defenders Office NSW, Submission FV 221, 2 July 2010; Law Society of New South Wales, Submission FV 205, 30 June 2010; Barrister, Consultation, Sydney, 10 June 2010.

[287] National Legal Aid, Submission FV 232, 15 July 2010.

[288] Public Defenders Office NSW, Submission FV 221, 2 July 2010.

[289] See Ch 26.

[290] Australian Law Reform Commission, New South Wales Law Reform Commission and Victorian Law Reform Commission, Uniform Evidence Law, Report 102, NSWLRC Report 112, VLRC FR (2005), [11.82].

[291] Queensland Law Reform Commission, The Receipt of Evidence by Queensland Courts: The Evidence of Children, Report 55 (Part 2) (2000), 367—quoting a Western Australian judge.

[292] A Cossins, Alternative Models for Prosecuting Child Sex Offences in Australia (2010), prepared for the National Child Sexual Assault Reform Committee, 197.

[293]R v Ellis (2003) 58 NSWLR 700. The case concerned the possible application of the ‘no rational view of the evidence’ test.

[294] Uniform Evidence Acts, ss 97, 98, 101. The TLRI has recently stated that there is no requirement under the Evidence Act 2001 (Tas) for a striking similarity: Tasmania Law Reform Institute, Evidence Act 2001 Sections 97, 98 & 101 and Hoch’s Case: Admissibility of Tendency and Coincidence Evidence in Sexual Assault Cases with Multiple Complainants, Issues Paper 15 (2009), Part 6, 57.

[295] For example, R v Fletcher (2005) 156 A Crim R 308.

[296] Ibid, [165]. In Victoria, under the Evidence Act 2008 (Vic), the Court of Appeal has stated that there must be something ‘distinctive’ about evidence of an accused’s conduct for it not be excluded as coincidence evidence: PNJ v Director of Public Prosecutions (Vic) [2010] VSCA, [20].

[297]R v Milton [2004] NSWCCA; R v GAC [2007] NSWCCA.

[298]Uniform Evidence Acts, s 98.

[299] As suggested by Public Defenders Office NSW, Submission FV 221, 2 July 2010.

[300] A Cossins, Alternative Models for Prosecuting Child Sex Offences in Australia (2010), prepared for the National Child Sexual Assault Reform Committee, Rec 3.7. The recommendation is based on Criminal Procedure Act 2004 (WA) s 133(5)–(6).

[301] Law Society of New South Wales, Submission FV 205, 30 June 2010.

[302] Public Defenders Office NSW, Submission FV 221, 2 July 2010.

[303] Office of the Director of Public Prosecutions NSW, Submission FV 158, 25 June 2010; Public Defenders and Prosecutors, Consultation, Sydney, 7 June 2010.

[304] The Committee proposed amendments to ss 97, 98 and 101 of the uniform Evidence Acts, of general application to tendency and coincidence evidence and that other jurisdictions make similar amendments with respect to the admissibility of propensity and similar fact evidence: A Cossins, Alternative Models for Prosecuting Child Sex Offences in Australia (2010), prepared for the National Child Sexual Assault Reform Committee, Recs 3.8, 3.9.