Sydney, New South Wales
Friday 17 March 2017

The Hon Justice Peter McClellan AM
Chair, Royal Commission into Institutional Responses to Child Sexual Abuse

Australian Lawyers Alliance NSW Annual State Conference

Seeking ‘justice for victims’ – Part I

Some of you may be asking why it is that the Royal Commission is considering criminal justice issues. Apart from the many issues concerning the nature, cause and impact of sexual abuse and the response of institutions, our Terms of Reference require us to inquire into:

what institutions and governments should do to address, or alleviate the impact of, past and future sexual abuse and related matters in institutional contexts, including, in particular, in ensuring justice for victims through … processes for referral for investigation and prosecution ….[1]

When asking the Commission to look at criminal justice the executive would have had in mind anecdotal accounts of the problems faced by complainants in sexual assault trials. We have all heard about or experienced them. We now have some statistics which confirm the difficulties faced by sexual assault complainants. I will come to those later. The question they raise is why do the outcomes for sexual assault differ so markedly from the outcomes for other crimes?

A little over thirty years ago in Bromley v The Queen Brennan J, in relation to accomplices, children giving evidence on oath, and complainants in sexual assault cases, said:

The courts have had experience of the reasons why witnesses in the three accepted categories may give untruthful evidence wider than the experience of the general public, and the courts have a sharpened awareness of the danger of acting on the uncorroborated evidence of such witnesses.[2]

These words reflect the long and difficult history of victims of sexual offences, including child sexual offences, in the criminal justice system. They also reflect a view amongst judges that they are possessed with particular insight, and that their role is to formulate the rules for criminal trials which reflect that insight. Some of those rules have had a significant and negative impact on the ability of child sexual abuse survivors to access justice.

Because of their classification as potentially unreliable witnesses, together with the form of directions designed specifically in relation to their evidence, and a judicial attitude that juries may not be able to fairly assess all the evidence, the criminal justice system has created a series of barriers in sexual assault cases not known for other crimes. For many complainants the barriers have been insurmountable. For many others the rules have proved a deterrent to any engagement with the criminal justice system.

A recognition that the system may have gone too far and may be denying justice to complainants first started to emerge about 30 years ago. Since then, largely through the intervention of parliament, significant reforms have been made. The question for the Royal Commission is whether there is more that should be done.

Directing the jury

When directing a jury a judge may in some circumstances be required to give a warning or caution. A warning or caution will typically be required where acting upon particular evidence is believed to involve ‘dangers’ or concerns matters about which the court is said to have ‘special experience’ which the jurors do not possess.[3]

The judicial distrust of sexual assault complainants has a long tradition. 

In the mid-1700s the English jurist Sir William Blackstone discussed the common law approach to assessing the truthfulness of a woman alleging sexual assault. He said:

“if the witness be of good fame; if she presently discovered the offence, and made search for the offender … these and the like are concurring circumstances, which give greater probability to her evidence. But on the other side, if she be of evil fame, and stand unsupported by others; if she concealed the injury for a considerable time after she had the opportunity to complain; if the place, where the act was alleged to be committed, was where it was possible she might be heard, and she made no outcry; these and the like circumstances carry a strong, but not conclusive presumption that her testimony is false or feigned.”[4]

The source of later and erroneous approaches to issues of sexual assault can be seen in these remarks. They emerge in the need for corroboration, the relevance of delay to a complainant’s credibility and the inappropriate use of a complainant’s sexual history. 

It may be suggested that the courts were merely relying upon the understanding of these issues current in the general community. However, we know that the courts were not relying on what was commonly understood. Indeed courts expressly rejected the understanding of the general public. What the quote from Bromley clearly demonstrates is that courts believed that their experience had generated knowledge often described as ‘judicial wisdom’ on these issues. They were relying on the ‘sharpened awareness’ the courts said they had about who jurors should and should not believe and why.

Decisions such as Longman,[5] Doggett,[6] Crampton[7] and Murray[8] have had profound consequences for complainants in sexual assault cases; particularly complainants who were children at the time they were assaulted. They derive from what judges’ thought they knew about how genuine complainants behave and what they thought they knew about how memory works. Those assumptions have turned out, with the benefit of empirical research, to be flawed. However, they became embedded in the fabric of the common law and proved difficult even for Parliament to dislodge.

In 1981 NSW enacted legislation to provide that if delay in complaint was raised in a trial for a prescribed sexual offence the trial judge was required to direct the jury that delayed complaint does not necessarily indicate that the allegation is false. The judge is to inform the jury that there may be good reason why a sexual assault victim may hesitate in complaining.[9]

Similar legislation was enacted in Victoria a decade later. That legislation was considered by the High Court in Crofts v The Queen[10] in 1996. Crofts concerned complaints of child sexual abuse that were made six years after the first alleged incident of abuse and just six months after the last alleged incident. Justices Toohey, Gaudron, Gummow and Kirby stated that ‘by the measure of cases of this kind, that was a substantial delay.’[11] We now know that this characterisation of the delay, if it is intended to diminish the veracity of the allegation, is wrong.

The Court held that the relevant provisions required a balanced direction. The effect was that while a direction in accordance with the section could be given, so too should a direction consistent with the common law that delay is relevant to the complainant’s credibility and should be taken into account.

Although lengthy it is worth quoting in full the relevant portion of the joint judgment as it provides a clear illustration of how closely the courts can protect judicial assumptions. Perhaps it passed without notice but in this passage the judges both reject and then rely on the court’s assumption as to ordinary human experience.

[The decision in Longman] makes it clear that the purpose of such legislation, properly understood, was to reform the balance of jury instruction not to remove the balance. The purpose was not to convert complainants in sexual misconduct cases into an especially trustworthy class of witnesses. It was simply to correct what had previously been standard practice by which, based on supposed ‘human experience’ and the ‘experience of courts’, judges were required to instruct juries that complainants of sexual misconduct were specially suspect, those complained against specially vulnerable and delay in complaining invariably critical. In restoring the balance, the intention of the legislature was not to ‘sterilise’ complainants from critical comment where the particular facts of the case, and the justice of the circumstances, suggested that the judge should put such comments before the jury for consideration. The overriding duty of the trial judge remains to ensure that the accused secures a fair trial. It would require much clearer language than appears in s 61 of the Act to oblige a judge, in a case otherwise calling for comment, to refrain from drawing to the notice of the jury aspects of the facts which, on ordinary human experience, would be material to the evaluation of those facts.

On one hand acknowledging that legislative intervention was necessary in order to correct flawed directions ostensibly based on ‘human experience’ and then, on the other hand invoking ‘human experience’ as the rationale for substantially diluting a direction required by legislation, appears, with respect, not to have troubled the judges.

A further difficulty arises in that there may not be uniformity amongst judges as to their assumptions about what and how jurors think. Rules are typically rationalised as being necessary because the weight of judicial experience says that it must be so, in order to avoid an injustice. It is this body of ‘judicial experience’ or ‘collective wisdom’ that gives these laws their legitimacy. Collective judicial “wisdom”, however, may be nothing more than the way we have always thought about the issue. Further, that collective wisdom may not be so collective.

For survivors, and for many in the community, it may be difficult to understand why on some issues juries are directed in one way in one state, and differently in another. If judges in New South Wales believe a rule is necessary to avoid an injustice, how is it that judges in Victoria do not?

The Markuleski direction is one example of this lack of uniformity. In Markuleski the NSW Court of Appeal held that, in word against word cases, the trial judge should, as a general rule, direct the jury that a reasonable doubt with respect to the complainant’s evidence on any count ought to be taken into account in its assessment of the complainant’s credibility generally.[12]

In contrast, Victorian courts have held that a Markuleski direction should generally be avoided.[13] Justice Buchanan in R v Trainor stated:

‘I think it unlikely that a jury given a separate consideration direction will be entirely uninfluenced by the impressions they derive from the evidence of a witness taken as a whole; I doubt that such a natural tendency needs judicial encouragement in the form of a Markuleski direction.’[14]

It is apparent that the Victorian court took a different view as to the guidance the jury needed as to how to reason to avoid an injustice to the accused. The need identified in NSW to give guidance is premised on an anterior assessment of whether or not the reasoning process the jury would ordinarily undertake is flawed, inadequate or indeed dangerous. It is interesting that a flaw detected by NSW judges in the reasoning process of jurors in Albury is, according to the law, not one present in the jurors in Wodonga.

The point is not to deny an evidentiary rule or a rule requiring a judicial direction in relation to particular evidence will never be necessary. Rather it is that judicial assumptions about how people reason may not be the most appropriate foundation for these types of rules.

One issue the Royal Commission is considering is the need for jury directions containing educative information about children and their response to child sexual assault. I will refer to them as educative directions.

There is significant research that demonstrates that the way a child behaves and reacts in response to sexual abuse is inconsistent with the expectation of many jurors.[15] The behaviours of sexually abused children are commonly described as ‘counter-intuitive’. These erroneous expectations in relation to children may also be present in beliefs held about their capacity to give evidence. This, of course, is not surprising – it was an assumption of the common law for many, many years.

Justification for these directions may be found not only in the fact that incorrect assumptions about victims of child sexual abuse are common in the general community. It is important that research commissioned by the Commission demonstrates that defence counsel commonly use strategies when questioning complainants in child sexual abuse offence trials which rely on these misconceptions or uncertainties amongst jurors in relation to child complainants.

Analysis of 120 transcripts of complainant evidence from 94 child sexual abuse trials showed that defence counsel often employed strategies to suggest that there is one typical way a victim responds to abuse. If the complainant had not responded in that way, counsel suggested to the jury that this was an indication that the offence may not have occurred.

The researchers found that the types of misconceptions about child sexual abuse exploited included the lack of resistance by the complainant at the time of the offence, the delay in reporting, the apparent lack of emotion by the complainant at the time of the offence and the continued relationship between the complainant and the accused after the offence.[16]

In their report Family Violence: A national legal response the ALRC and NSWLRC stated that there was a strong case for the use of jury directions which would ‘summarise a consensus of expert opinion drawn from the work of psychiatrists, psychologists and other experts on child behaviour.’[17]

Similarly the National Child Sexual Assault Reform Committee has also recommended that three mandatory judicial directions, summarising the research and containing the same information, should be introduced in all Australian jurisdictions. Those three directions are in relation to children’s abilities as witnesses, very young children’s abilities as witnesses, and children’s responses to child sexual abuse.[18]

Providing the jury, and indeed the judge, with accurate information about the behaviour of children who are sexually abused assists in their credibility being assessed more accurately. This in turn promotes the public interest in securing the convictions of the guilty and the acquittal of the innocent.

Tendency and coincidence evidence

Judge’s assumptions about human behaviour have influenced other aspects of the laws of evidence. One important area is tendency and coincidence evidence or, as these concepts were known to the common law, propensity and similar fact evidence.

I have said on many occasions that this issue – that is, how the criminal justice system deals with allegations against an individual of sexual offending against more than one child – is one of the most significant issues we have identified in our criminal justice work. Before turning to that issue you will be interested in a recent analysis of outcomes in sexual assault trials.

Data recently obtained by the Royal Commission from the NSW Bureau of Crime Statistics and Research confirms the difficulties facing complainants of sexual assault, both children and adults.

In New South Wales between July 2012 and June 2016 the conviction rate for all offences, including matters finalised by a guilty plea, was 89%. For child sexual assault offences the conviction rate was 60%. By comparison the conviction rate for assault was 70%, robbery 73% and illicit drugs 94%. One class of offences with a lower conviction rate was adult sexual assault with a conviction rate of only 50%. One possible reason is that in addition to the fact that these cases are, like child sexual assault cases, commonly word against word, consent will often be an issue.

The conviction rate for assault matters is higher than for child sexual assault matters. That is notwithstanding that in many cases, the identity of the offender may well be at issue. This is rarely the case for child sexual assault matters.

The conviction rates for illicit drug matters is significantly higher than for child sexual assault matters. This is likely because in a substantial number of drug cases the offender will be found to have the drugs on them making proof of the offence much easier than in child sexual assault matters.

The law reports are replete with comments by judges about the extra caution with which juries must approach sexual assault cases. In KRM v The Queen, Kirby J said:

In cases involving accusations of sexual offences, courts and prosecutors must exercise particular vigilance, so far as they can, to ensure that the fairness of the trial is maintained because the circumstances are peculiarly likely to arouse feelings of prejudice and revulsion. This duty imposes special difficulties for judges presiding at such trials where they are conducted before a jury.[19]

This concern – the arousing of prejudice and revulsion in jurors – is not borne out by the data. If it is the case that the mere allegation of a sexual offence against a child is peculiarly likely to arouse feelings of prejudice then we should see high conviction rates in these cases. This is because, self-evidently, every trial in relation to a child sexual offence will involve an allegation that a child has been sexually offended against. What we in fact see is much lower than average conviction rates.

In a statement in relation to the severance of trials concerning multiple sexual offence counts Gibbs CJ in De Jesus v The Queen said:

Sexual cases … are peculiarly likely to arouse prejudice, against which a direction to the jury is unlikely to guard.[20]

Again the data suggests these concerns are overblown. In relation to child sexual assault offences which were finalised at a defended hearing or at trial there were 725 matters in NSW between July 2012 and June 2016. Of those the defendant was convicted of all relevant offences in 32% of matters. The defendant was convicted of some but not all relevant offences in 16% of matters. The defendant was convicted of no relevant offences in 52% of matters.

The capacity for jurors to correctly isolate and apply the evidence in relation to each individual case is not only demonstrated in the data but has been recognised by courts in cases in relation to the issue of ‘inconsistent verdicts.’ In rejecting the ground of appeal that if the jury acquitted on one or more counts it follows they did not believe the complainant in relation to these counts and for that reason they should not have believed the complainant in relation to the other counts, appeal courts have accepted that there may be rational reasons for convicting on some counts and acquitting on others. The courts have recognised that jurors follow the direction to consider each count, and the evidence in relation to it, separately.

The Commission is aware of comment amongst members of the legal profession that the Commission’s work has had such a transformative effect on the mind of the general community so that it has now become much more difficult, perhaps, it has been suggested, dangerously difficult, to secure an acquittal in relation to child sexual offences. The data clearly suggests otherwise.

In 2012-2013, 73 child sexual assault matters were finalised at a defended hearing or trial in the District Court. In the District Court, of course, charges in defended hearings are most likely to be heard before a jury. Of those 73 matters the defendant was convicted of all relevant offences in 29% of matters, convicted of some but not all relevant offences in 27% of matters, and convicted of no relevant offences in 44% of matters.

By contrast, in 2015-2016, 142 matters were finalised at a defended hearing.

That is almost double the number of matters.  However, in relation to conviction rates the defendant was convicted of all relevant offences in 24% of matters, convicted of some but not all relevant offences in 23% of matters, and convicted of no relevant offences in 54% of matters.

It appears that, although many more complainants are coming forward, the chances of an offender being acquitted have risen rather than fallen.

Issues in relation to tendency and coincidence evidence have troubled our courts for many years. The recent Victorian report into jury directions suggests they have caused problems for more than 100 years. In Pfennig v The Queen, Justice McHugh spoke of:

the vexed question as to the circumstances in which the prosecution may prove a criminal charge by tendering evidence that the accused has engaged in criminal conduct on occasions other than that which is the subject of the charge before the court.[21]

Our work to date indicates this question remains vexed.

Tendency and coincidence evidence has particular significance in relation to the work of the Royal Commission. This is because, in the absence of such evidence, child sexual abuse cases often reduce to word against word. Child sexual offences are generally committed in private. In many cases there will be no medical or forensic evidence. Unless the perpetrator has retained images of the abuse or makes an admission it is likely that the only direct evidence will come from the complainant. Satisfying the jury beyond reasonable doubt in these circumstances can be very difficult. In institutional contexts, a perpetrator may have access to a number of vulnerable children. In these cases there may be evidence available from other complainants or witnesses who allege that the accused also sexually abused them. In these cases issues of tendency and coincidence arise.

The courts have traditionally been cautious about allowing tendency and coincidence evidence, or their common law counterparts, to be put before the jury. The reason for that caution was articulated by Lord Cross of Chelsea in Boardman v DPP. It is:

Not that the law regards such evidence as inherently irrelevant but that it is believed that if it were generally admitted jurors would in many cases think that it was more relevant than it was, so that … its prejudicial effect would outweigh its probative value.[22]

The importance of the role of the judge as the gatekeeper for evidence of this type was recently reaffirmed by the High Court in IMM v The Queen.[23] Justices Nettle and Gordon commented on the requirement that the judge should make a preliminary assessment of evidence that falls within a particular category the common law and/or statute have identified as potentially unreliable:

Such an assessment is not in any sense a usurpation of the jury’s function. It is the discharge of the long recognised duty of a trial judge to exclude evidence that, because of its nature or inherent frailties, could cause a jury to act irrationally either in the sense of attributing greater weight to the evidence than it is rationally capable of bearing, or because its admission would otherwise be productive of unfair prejudice which exceeds its probative value.[24]

In a 1933 article published in the Harvard Law Review Professor Julius Stone traced the history of the rule of exclusion of similar fact evidence in England from its earliest origins until Makin’s Case in 1894. Stone identifies the starting point for the rule as the case of Rex v Cole first referred to in an 1814 text. 

By 1850, Stone argues, the rule in relation to similar fact evidence was understood to be that ‘whenever the evidence of similar facts offered was relevant to any specific act or issue upon which the jury has to make its mind up it was admitted.’[25] He says that Rex v Cole was ‘authority merely for the proposition that if the evidence offered is relevant only by an argument which proceeds from the other crimes to the disposition of the prisoner to commit such crimes’ (ie mere propensity) ‘and thence to the probability of his having committed the crime charged, it is not admissible.’[26]

Stone goes on to discuss the way the law developed between 1850 and Makin’s case in 1894. His general thesis is that prosecutors generally succeeded in attaching the evidence to some issue before the jury other than mere propensity, for example to rebut a defence, with the consequence that the rule in Rex v Cole would not apply.

Over time the issues to which the evidence was attached came to be considered to be heads of exception to a general rule of exclusion. Stone states ‘as so easily occurs in the common law … the tendency to crystallise particular determinations of relevance into categories of admission appeared.’ [27]

This began to cause difficulties for courts. An example Stone offers is the New Zealand case of Regina v Hall, involving similar fact evidence, in relation to which he states:[28]

This remarkable opinion closes with a classic declaration of the impatience of judges before the tyranny of absurd rules of their own creation.

‘Viewed in light of science, philosophy or common sense there is without a doubt a nexus between the two events …

We are aware that in cases like the present the application of the rigid principle of the common law must often result in what the public may regard as a failure of justice. That is really not our concern. It is our duty to see that justice be administered according to law.’[29]

In relation to the expression of the rule in Makin’s case, Stone states that it has been cited ‘sometimes with understanding of its meaning, sometimes not.’[30]


Stone’s view is that Makin’s case should be understood as holding that there is ‘no broad rule of exclusion with exceptions, but a broad rule of admissibility where there is relevance, except where the only relevance is via disposition’[31] (that is, mere propensity).

Stone concludes with his suggestion of how the rule should be formulated: ‘It is that the trial judge should be recognised to have a discretion to decide whether the probative weight of the proffered evidence outweighs its mere prejudice. [32]

Two issues stand out in relation to Professor Stone’s conclusion: the first is that he accepts that the evidence is troublesome. Earlier in the article he states that similar fact evidence ‘presents the possibility of undue prejudice in an extreme form’.  [33] The foundation on which this statement rests is not explored in detail other than to state that ‘the policy of avoiding undue prejudice is based on weaknesses of human nature which are today as obvious as ever.’ [34]  The second is in relation to his suggested remedy. Notable is the absence of any requirement for the probative value of the evidence to substantially outweigh the prejudicial effect as required under the uniform evidence law.

The Royal Commission commissioned research designed to interrogate whether the assumption that the evidence ‘presents the possibility of undue prejudice in an extreme form’ is correct.

Using mock juries and a trial involving charges in relation to child sexual abuse in an institutional context the researchers examined the effect of joinder and the admission of tendency evidence upon the jury’s reasoning and their decisions.

1,029 jury eligible citizens, 580 women and 449 men, participated in the study. They were randomly allocated to one of 90 mock juries.

The juries each watched a video of one of ten versions of a trial. Variations of the trial included:

  • A separate trial of an adult complainant with moderately strong evidence.
  • A separate trial of an adult complainant with moderately strong evidence in which relationship evidence comprised of uncharged acts and grooming behaviour was also admitted.
  • A separate trial of an adult complainant with moderately strong evidence in which tendency evidence from two other prosecution witnesses was admitted.
  • A joint trial with three adult complainants who gave weak, moderately strong and strong evidence.
  • The deliberations of each of the juries were audiotaped and videotaped. Transcriptions of each jury deliberation were qualitatively and quantitatively analysed.

The trial was designed to interrogate three types of unfair prejudice:

  1. Inter-case conflation prejudice – the idea that juries will confuse or conflate the evidence led to support different charges in a joint trial and wrongly use evidence related to one charge in considering another charge.
  2. Accumulation prejudice – the idea that juries will assume the accused is guilty because of the number of charges against him or the number of prosecution witnesses regardless of the strength of the evidence.
  3. Character prejudice – the idea that juries will use evidence about the accused’s other criminal misconduct and find guilt by reasoning that because he did it once he has done it again.[35]

The qualitative and quantitative analysis of the deliberations is an important feature of this research. A key criticism of a number of previous studies was their over reliance on conviction rates as the relevant measure. Verdicts on their face reveal nothing about the reasoning process that lies behind them. They do not tell us whether any increase in conviction rates is the result of permissible reasoning arising from the availability of more logically probative evidence such as tendency evidence, or whether they are due to impermissible reasoning based on unfair prejudice.

With respect to the issue of inter-case conflation the researchers found that jurors were more likely to make errors within a case, rather than between cases, suggesting little inter-case conflation prejudice. They found that these errors were corrected by other jurors in the course of deliberation. The researchers found that no verdicts were based on persistent uncorrected or inter-case conflation of the evidence.[36]

In relation to the risk of unfair prejudice from accumulation prejudice the researchers found that conviction rates for the weakest case did not increase significantly with extra witnesses or extra charges, indicating no accumulation prejudice.[37]

Perhaps most significantly, in relation to character prejudice evaluation of the jury deliberations revealed that no juries in either the tendency evidence or joint trials impermissibly used the tendency evidence to conclude that the defendant was guilty because of the number of allegations of prior misconduct that were made. The researchers found no evidence of verdicts motivated by emotional reactions to the severity of the allegations, such as a sense of horror regarding the allegations, or a desire to punish the defendant.[38]

Conscious of the fact that some jurors may have been inhibited from revealing the true basis for their decision to convict in a group setting, as part of the post-trial questionnaire the jurors were asked to identify the main reason for their verdict. This questionnaire was completed anonymously. The responses revealed that only three percent of jurors gave reasons that were identified as displaying character prejudice.[39]

The researchers ultimately concluded that:

the low frequency and isolated examples of reasoning in deliberations involving inter-case conflation of the evidence, accumulation prejudice, or character prejudice suggest that the likelihood of impermissible reasoning, whether in joint trials or separate trials, is exceedingly low. This low probability suggests that there was negligible unfair prejudice to the defendant in joint trials where tendency evidence was admitted.[40]

The law in Australia at present

The law in relation to the admissibility of tendency and coincidence evidence has developed in different ways across Australia.

The common law in relation to propensity and similar fact evidence, with some modification, applies in Queensland. The test for admissibility is that set down in Pfennig v The Queen. Propensity or similar fact evidence will be admitted only if it possesses ‘a particular probative value or cogency such that, if accepted, it bears no reasonable explanation other than the inculpation of the accused in the offence charged.’[41]

The ALRC has described this test as ‘extremely stringent’.[42] It is more stringent than the test that existed at common law in England and Wales before the significant reforms made in those jurisdictions in 2003.

The uniform evidence legislation applies in NSW, Victoria, Tasmania, the ACT and the NT. In criminal proceedings tendency or coincidence is admissible only if notice is given, the court considers the evidence has significant probative value and the probative value of the evidence substantially outweighs any prejudicial effect it may have on the accused.[43]

Despite this legislative uniformity, differences have emerged between NSW and Victoria in relation to how the provisions for the admissibility of tendency and coincidence should be interpreted and applied. There are three notable differences of significance in the context of institutional child sexual abuse. The first is that, particularly in relation to tendency evidence, there are differences as to whether and to what extent similarity in the manner of offending is required in order for the evidence to be admissible as tendency evidence. In NSW, similarity is not essential, although it may assist in establishing probative value. In Victoria common or similar features or an underlying unity in the offending is required.

The second is in relation to the question of whether features of the institutional context are relevant in determining similarity and the probative value of the evidence. NSW courts have found similarities in the circumstances of institutional offending that Victorian courts have regarded as unremarkable circumstances that are not within the accused’s control.[44]

Thirdly, differences have emerged as to whether the reliability of the evidence, particularly in relation to issues of possible collusion, concoction and contamination should be determined by the judge or the jury. In NSW reliability and credibility generally have no role to play in the assessment of probative value with the result that these are generally matters for the jury. Before IMM v The Queen[45] resolved the issue in favour of the NSW approach, Victoria maintained the common law position that that the reliability of evidence affects its probative value and therefore amongst other matters issues of possible concoction or contamination are to be determined by the judge.

In relation to the first difference it is relevant to ask why we are looking for similarity, when the provisions which, in relation to tendency evidence, do not refer to it. It is because the common law has determined that the probative force of the relevant evidence is tied to the similarities in the behaviour of the accused as alleged by the witnesses. However, in relation to the question of whether sexual offending has in fact occurred, rather than the identity of the offender, it is relevant to ask to what extent is similarity an indicator of greater probative value? If it is the case that sexual offenders typically perform the same sexual acts on victims of the same sex who are the same age then it may be correct to reject as significantly probative, evidence that does not have these similarities. However, the work the Royal Commission confirms that a sexual offender may offend in a variety of ways against a range of victims of varying ages. In case studies 16, 17, 18, 26, 37 and 46 we heard evidence of a single offender offending against both boys and girls.[46] The different types of offence, from touching, fondling and ultimately penetration, will commonly be committed by the same offender. There is a danger that if judges fail to understand how sex offenders actually operate the assessment of the probative value of tendency witnesses will be based on a false premise. Proceeding from a false premise rarely leads to a correct conclusion.

In South Australia new rules for the admissibility of ‘discreditable conduct’ evidence commenced in 2012. Evidence of a defendant’s discreditable conduct may be admitted if its probative value substantially outweighs any prejudicial effect it may have on the defendant. If it is tendered for propensity reasoning, it must have ‘strong probative value’.[47]

The rules in Western Australia are notably different. Propensity evidence is admissible if the court considers that it would have significant probative value and ‘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.’[48]

This formulation of the public interest test adopts McHugh J’s approach in Pfennig. The test for admission in Western Australia is arguably the most liberal in Australia.

This lack of clarity and inconsistency between the states could itself justify a recommendation for reform. It is difficult for the Commissioners to support a situation in which the rules in relation to evidence of such significance are different for a survivor involved in criminal proceedings in one state than for a survivor in another. In an oft-quoted passage, Justice Brennan has observed that ‘[i]nconsistency is not merely inelegant’ but suggests ‘an arbitrariness which is incompatible with commonly accepted notions of justice’.[49]

Considered together the inconsistency across the jurisdictions, the lack of clarity in relation to uniform provisions, the low conviction rates in child sexual offence cases, the fact that juries regularly return different verdicts on different counts and the findings of the jury research suggest that the law in relation to tendency and coincidence evidence may need reform.

The Western Australian provisions have a lower threshold of admissibility than in any other Australian jurisdiction. Yet the Commission has not seen or heard any evidence or indeed any suggestion of injustices flowing from these changes.

And in relation to the position in England and Wales, where the law in relation to admissibility is more liberal still, we have not seen or heard any suggestion of injustices arising as result of the changes occurring there. Those changes were made more than a decade ago.

Whilst reasonably satisfied that the law should change to facilitate greater admissibility of tendency and coincidence evidence in child sexual abuse cases, the Commissioners have not yet identified how this is best achieved. 


A further difficulty for survivors in child sexual abuse cases is the requirement for the prosecution to provide particulars in relation to the offence or offences charged. The requirement to provide particulars, undoubtedly reasonable, is necessitated by the accused being entitled to know the case against him or her.[50]

Victims and survivors of child sexual abuse often find it difficult to provide adequate or accurate details in relation to the offending. There are a number of reasons. The first is that young children may not have a good understanding of dates, times and locations or an ability to describe how different events relate to each other across time. Second, delay in reporting may cause events to be wrongly attributed to a particular time or location when they in fact occurred earlier or later, or at another location. Thirdly, the abuse may have occurred so often and in such similar circumstances, that the victim or survivor is unable to describe specific or distinct occasions in which they were offended against.

The result, particularly in relation to this third reason, is a cruel paradox: the greater the regularity with which a child is offended against, the more difficult it can become to charge and prosecute the offender.

Parliaments have attempted to remedy this situation through the introduction of legislation to create persistent sexual abuse offences. This first occurred in Queensland in 1989. By the end of the 1990s an offence of this type had been introduced in each of the states and territories. There have, however, been challenges in relation to these offences.

The Queensland offence as originally enacted was an offence of ‘maintaining a sexual relationship with a child/young person.’[51] The legislation required the prosecution to prove the sexual relationship by showing three distinct occasions of unlawful sexual conduct each proved beyond reasonable doubt.

The offence was considered by the High Court in 1997 in KBT v The Queen.[52]The Court held that the offence required the jury to be satisfied beyond reasonable doubt as to the commission of the same three acts which constituted relevant sexual offences.[53]

It follows that in order for the jury to agree that the same three acts of sexual offending have been proved, one must be able to identify with some precision each of the three discrete acts or occasions of offending necessary to make up the charge. As a consequence the problem was not removed.

In most Australian jurisdictions the relevant offence continues to require proof of the occurrence of three or more unlawful sexual acts.[54] It is not surprising, given the decision in KBT, that in most, but not all, of these jurisdictions the persistent sexual abuse offence is rarely charged. Data provided to the Royal Commission indicates it is rarely charged in the two largest jurisdictions: New South Wales and Victoria.

In Queensland, however, the persistent sexual abuse offence is regularly charged. In that state the offence has been reframed in order to address the concerns identified in KBT. Under the Queensland legislation the actus reus of the offence is the unlawful sexual relationship, not particular unlawful acts.

The amended section 229B defines an unlawful sexual relationship as one that involves more than one unlawful act over any period. All members of the jury are required to be satisfied beyond reasonable doubt that the evidence established that an unlawful sexual relationship with the child involving unlawful sexual acts existed. Importantly, however, s 229B(4) provides that:

(a) the prosecution is not required to allege the particulars of any unlawful sexual act that would be necessary if the act were charged as a separate offence and;
(b) the jury is not required to be satisfied of the particulars of any unlawful sexual act that it would have to be satisfied of if the act were charged as a separate offence; and
(c) all the members of the jury are not required to be satisfied about the same unlawful sexual acts.

Section 229B has been considered at appellate level in Queensland on a number of occasions. The Court’s decisions establish that the key element of the offence is the unlawful sexual relationship.[55] The indicia of maintaining a relationship include its alleged duration, the number of acts and the nature of the acts alleged.[56] The Court was satisfied that the trial’s judge’s power to ensure a fair trial is maintained under the section,[57] and that the section does not remove the court’s power to set aside a conviction on the grounds that there was a miscarriage of justice where the accused was given so little information about the charge as to render it impractical to prepare a defence.[58] The Court of Appeal has also held that s 229B(4) does not offend Chapter III of the Commonwealth Constitution.[59]

Application for special leave to appeal to the High Court in relation to convictions under s 229B have been twice refused. Once in 2008 and again in 2012.

The Royal Commission is considering the law in relation to persistent sexual offences as part of its criminal justice work. It is unacceptable, in our view, that the criminal justice system should accept a situation in which children who have suffered the most extensive abuse may be those who are less able to receive justice in the criminal courts.


The Royal Commission’s criminal justice project is significant in scale. The issues which we must consider cover a broad reach. In Part II of this speech, which I will deliver at the Modern Prosecutor Conference in Melbourne on 13 April, I will discuss some of the other issues the Commission is examining. These include prosecution responses, including DPP complaints and oversight mechanisms; the evidence of victims and survivors, including the use of special measures such as intermediaries and special hearings; and, sentencing.

The Royal Commission intends to provide its report in relation to criminal justice to government in August.

[1] Terms of Reference, para (d).

[2] (1986) 161 CLR 315, 324.

[3] See, eg, R v Stewart [2001] NSWCCA 260 [71]-[83] (Howie J); Australian Law Reform Commission and New South Wales Law Reform Commission, Family Violence – A National Legal Response, ALRC Report 114, NSWLRC Report 128, 2010, 1309-1310.

[4] Blackstone’s Commentaries on the Laws of England, 1st ed, bk 4, c 15, 213-4.

[5] (1989) 168 CLR 79.

[6] (2001) 208 CLR 343.

[7] (2000) 206 CLR 161.

[8] (1987) 11 NSWLR 12.

[9] Crimes Act 1900 (NSW) ss 405B and 405C.

[10] (1996) 186 CLR 427.

[11] (1996) 186 CLR 427, 442.

[12] R v Markuleski (2001) 52 NSWLR 82, 122 (Spigelman J)

[13] See The Queen v Goss [2007] VSCA 116; R v PMT [2003] VSCA 200 [34]. See also Judicial College of Victoria, Criminal Charge Book, [4.18.1].

[14] R v Trainor [2003] VSCA 200 [32].

[15] A Cossins, J Goodman-Delahunty and K O’Brien, ‘Uncertainty and Misconceptions About Child Sexual Abuse: Implications for the Criminal Justice System’, Psychiatry, Psychology and Law, vol 16, no 3, (2000).

[16]M Powell, N Westera, J Goodman-Delahunty, A S Pichler, An Evaluation of how Evidence is Elicited from Complainants of Child Sexual Abuse (August 2016) 197-201.

[17] Australian Law Reform Commission and New South Wales Law Reform Commission, Family Violence – A National Legal Response, ALRC Report 114, NSWLRC 128, 2010 p 1272.

[18] A Cossins, Alternative models for prosecuting child sex offences in Australia, National Child Sexual Assault Reform Committee, 2010, 26-27, 173-175.

[19] (2001) 206 CLR 221, 254.

[20] (1986) 68 ALJR 1 [4].

[21] (1995) 182 CLR 461, 510.

[22] [1975] AC 421, 456

[23] [2016] HCA 14.

[24] [2016] HCA 14 [161].

[25] Julius Stone, ‘The Rule of Exclusion of Similar Fact Evidence: England’ (1933) 46 Harvard Law Review 954, 959.

[26] Ibid 965.

[27] Ibid 965-966.

[28] Ibid 966: ‘as so easily occurs in the common law … the tendency to crystallise particular determinations of relevance into categories of admission appeared.’

[29] Ibid 973.

[30] 5 NZLR 93 (1887)

[31] Makin v Attorney-General of New South Wales [1894] AC 57, 65.

[32] Stone, above n 25, 975.

[33] Ibid 984.

[34] Ibid 954.

[35]  J Goodman-Delahunty, A Cossins and N Martschuk, Jury Reasoning in Joint and Separate Trials of Institutional Child Sexual Abuse: An Empirical Study (May 2016) 45-48.

[36] Ibid 249.

[37] Ibid 28, 249, 259.

[38] Ibid 35.

[39] Ibid 124. 

[40] Ibid 35.

[41] (1995) 182 CLR 461, 481.

[42] ALRC, Uniform Evidence Law (Report No 102, 2005), [11.6]

[43] See, eg, Evidence Act 1999 (NSW) ss 97, 98, 101.

[44] Compare PNJ v DPP (2010) 27 VR 146 and R v PWD (2010) 205 A Crim R 75.

[45] [2016] HCA 14.

[46] See also, for example, Henderson v R [2016] NSWCCA 8.

[47] Evidence Act 1929 (SA) s 34P.

[48] Evidence Act 1906 (WA) s 31A.

[49] Re Drake and Minister for Immigration and Ethnic Affairs (No 2) (1979) 11 FLR 203.

[50] Johnson v Miller (1937) 59 CLR 467, 489 (Dixon J).

[51] Criminal Code Act 1899 (Qld) s 229B.

[52] (1997) 191 CLR 417.

[53] (1997) 191 CLR 417, 422-423.

[54] NSW, Victoria, WA, Tasmania, ACT, NT.

[55] R v LAF [2015] QCA 130 [4].

[56] R v DAT [2009] QCA 181 [17]

[57] R v CAZ [2011] QCA 231 [45], [51].

[58] R v CAZ [2011] QCA 231 [51].

[59] R v CAZ [2011] QCA 231 [52]-[53].