Sydney, New South Wales
Thursday 2 June, 2016

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

Legal Aid NSW Criminal Law Conference

A Royal Commission is the most significant method of inquiry available to governments. It may investigate individual and institutional conduct. It is often used to determine whether criminal conduct may have been committed. It also provides an opportunity to examine complex issues in a structured manner with the benefit of detailed research. It enables consultation with people and organisations with the knowledge and experience of issues which the government believes require detailed consideration. Through its forensic work it can identify institutional and individual failures. Its research and policy development programs enable it to bring forward recommendations for change intended to improve future outcomes.
 
The Royal Commission which I chair has been resourced to perform all of these functions. Apart from examining the behaviour of institutions, which is fundamental to our work, we have the responsibility to bring together the learning across many disciplines; law, policing, education, human behaviour, effective regulation, amongst others, to provide a safer environment for our children.
 
Apart from the issue of redress, our terms of reference, though broad, provide us with two fundamental objectives: to expose what has happened in the past and to make recommendations aimed at ensuring, so far as possible, that children are not sexually abused in an institutional context in the future.
 
As many of you would be aware it is typical for Royal Commissions to receive evidence in public. However, conscious of the trauma and suffering associated with child sexual abuse and the difficulties many survivors have in telling their stories the Commonwealth Parliament amended the Royal Commissions Act 1902 to create a process referred to as a private session. A private session with one of the six Commissioners provides an opportunity for a survivor, or a survivor's family member, to tell the story of their abuse in a protected and supportive environment. Without such a process the 'bearing witness' and 'truth telling' component of the Royal Commission's work could never have been achieved. The Commissioners have now conducted over 5,000 private sessions. We expect to conduct more than 7,000 before we complete our work in December 2017.
 
From the information we have gathered, much of it from private sessions, I have referred 1,067 matters to police to investigate with a view to the possible prosecution of an offender.            
 
The Royal Commission has now held 39 public hearings. We will begin our fortieth, into Australian Defence Force institutions, later this month. Public hearings have been held in every state and territory in Australia. Although we have enough information to justify a public hearing into more than 1,000 institutions we will only be able to look at somewhere between 50 and 60 in a public hearing. Decisions about which institutions we choose to examine in a public hearing are informed by whether or not the hearing will advance an understanding of systemic issues and provide an opportunity for institutions to learn from previous mistakes. In many cases the issues explored, and the lessons to be learned, will have relevance beyond the individual institution being examined. Many institutions who have not themselves been the subject of a public hearing have already responded to the problems revealed in similar institutions and have implemented change or reviews to improve the safety of the children in their care. Public hearings are also an important means of enhancing the community's awareness and understanding of the nature, circumstances and, the often devastating, impact of child sexual abuse. To allow time for preparation of our final reports our public hearing program will come to an end in early 2017.
 
In addition to private sessions and public hearings our policy and research program underpins the work of the Royal Commission. This program has the assistance of national and international experts across many disciplines. We are examining a wide range of topics. These include child safe organisations, diversity and vulnerability, information sharing, mandatory reporting, advocacy and support services, record-keeping and prevention and treatment. As part of this program we have released consultation papers in relation to out of home care and complaint handling and response. Last month we released two research reports. The first examining information-sharing frameworks and the second examining best practice in implementation.
 
Although our ultimate report is not due to be delivered to government until December next year the Royal Commission has already completed two “final” reports. Both reports contain recommendations to assist people who have been sexually abused and enhance the protection of children in the future.
 
The first report makes recommendations with respect to Working with Children Checks and was released by the Government on 17 August 2015. The report makes 36 recommendations designed to strengthen Working with Children Check regimes throughout Australia.
 
The second report discusses Redress and Civil Litigation. The Commissioners made 99 recommendations designed to provide justice to victims of child abuse in an institutional context. The report says that 'in order to provide redress under the most effective structure for ensuring justice for survivors, the Australian Government should establish a single national redress scheme.' The civil litigation recommendations focused on four issues: limitation periods, identification of the proper defendant, the duty of care owed by institutions and model litigant approaches.

Criminal Justice Project

The criminal justice issues which the Royal Commission must consider cover a broad reach. They extend from reporting to police and police investigations, through all aspects of the prosecution and trial process. We are also looking at sentencing and post-sentencing measures.
 
There are many aspects of the criminal justice system which present challenges for survivors of child sexual abuse. As many of you would be aware the law has created special rules in relation to the reliability of the evidence of both children and adult sexual assault complainants.
 
This has given rise to issues relevant to sexual offending including the process by which survivors give evidence and the laws of evidence themselves. Both these issues were considered in our public hearing into criminal justice in March this year. The first week of the public hearing looked at how the criminal justice system deals with allegations against an individual of sexual offending against more than one child. We considered the admissibility and use of tendency and coincidence evidence and their common law counterparts, propensity and similar fact evidence. We also considered the related issue of whether a joint trial may be held with respect to allegations against an accused made by multiple complainants.
 
The second week of the public hearing focused on the challenges for the criminal justice system when a very young child, or a person with a disability that impacts on their ability to communicate, complains of being sexually abused. The difficulties these two groups face by reason of their limited ability to communicate make them some of the community's most vulnerable members. We considered how the requirements of the criminal justice system, including in relation to oral evidence and cross-examination, affect the investigation and prosecution of allegations of child sexual abuse.
 
In April we held three public roundtables to discuss criminal justice issues. The first roundtable discussed criminal offences for failing to report child sexual abuse, including s 316 of the Crimes Act 1900 (NSW). This included consideration of the circumstances, if any, in which blind reporting of offences might be permissible. A blind report is a report that is made to police in which the alleged victim’s name or identifying details are withheld. As some of you may be aware the practice of blind reporting has been a matter of some controversy in NSW.

The second roundtable discussed adult sex offender treatment programs currently operating in Australia and internationally, and the effectiveness of these programs.

The third roundtable discussed the issue of DPP complaint and oversight mechanisms. One issue the Commissioners are considering is whether there should be avenues for victims to seek a review of decisions not to prosecute. We are also considering whether there should be external oversight of DPPs. In the course of the roundtable we heard from the DPP for England and Wales and the Chief Inspector of the Crown Prosecution Service in relation to the complaints and oversight mechanisms operating in that jurisdiction.

We have also released a number of papers as part of our criminal justice program. These include:

  • “Sentencing for child sexual abuse in institutional contexts” by Arie Freiberg, Hugh Donnelly and Karen Gelb.
  • "The admissibility, and use of tendency, coincidence and relationship evidence in child sexual assault prosecutions in a selection of foreign jurisdictions" by Associate Professor David Hamer.
  • "Tendency, coincidence and joint trials" advice prepared for the Royal Commission by Tim Game SC, Julia Roy and Georgia Huxley.
  • "Specialist prosecution units and courts: A review of the literature" by Patrick Parkinson.
  • "A systematic review of the efficacy of specialist police investigation units in responding to child sexual abuse" by Nina Westera.
  • "The use and effectiveness of restorative justice in criminal justice systems following child sexual abuse or comparable harms" by Dr Jane Bolitho.
  • "Brief review of contemporary sexual offence and child sexual abuse legislation in Australia: 2015 update" by Hayley Boxal and Georgina Fuller of the Australian Institute of Criminology.
  • "A statistical analysis for child sexual abuse in institutional contexts" by Dr Karen Gelb.

Jury Research

Child sexual abuse offences are generally committed in private, with no eyewitnesses. In many cases there will be no medical or scientific evidence capable of confirming the abuse. Unless the perpetrator has retained recorded images of the abuse (and some do), or admits the abuse, it is likely that the only direct evidence will come from the complainant.

Where the only evidence of the abuse is the complainant’s evidence, it can be difficult for the jury to be satisfied beyond reasonable doubt that the alleged offence occurred. There may be evidence that confirms some of the surrounding circumstances, or evidence of first complaint, but the jury is effectively considering the account of one person against the account of another.

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. The question is whether that “other evidence” can be admitted in the trial.

As we all know these issues 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 (1995) 182 CLR 461, at 510, 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.

The discussion in the first week of our public hearing indicates that this question remains vexed, even in those jurisdictions that have adopted the Uniform Evidence Act.

We heard in the public hearing that the rules governing both the admissibility of tendency and coincidence evidence, and the admission of evidence from other complainants appear to operate differently in different jurisdictions, both internationally and within Australia.

Both the common law and statutory rules share a common origin. They have been significantly informed by judges’ assumptions about how juries reason. This is particularly the case in relation to sexual offences, where judges have determined that special care must be taken. It has been assumed that allegations of sexual offending, particularly, but not only, against children, are likely to arouse unfair prejudice in the jury. As a consequence we have accepted that a person accused of sexual offences, including sexual offences against children, needs careful protection against the risk of the jury being influenced by such prejudice.

In September last year, I gave a paper at the Supreme Court of New South Wales Annual Conference. In that paper I examined the means by which courts can use the learning from other disciplines in relation to the sexual abuse of children in both the trial and sentencing process.

Judges have, for centuries, relied on their own understandings of human behaviour to inform the content of the relevant rules. The difficulty is that, in the absence of research evidence as to how people behave, we do not know whether the judges’ assumptions are correct.

In some cases, we know that judges’ assumptions have been far from correct. An obvious example is the issue of delay in complaint in sexual assault matters. For years, judges assumed that victims of sexual offences will complain at the first reasonable opportunity. As a consequence delay was accepted to adversely affect the complainant’s credibility. The common law developed special rules for warning the jury, in accordance with this assumption.

Research discredited this assumption. We now know that delay in complaint of sexual abuse is common rather than unusual, particularly in the context of child sexual abuse. Parliaments have legislated to limit or displace this erroneous assumption and the common law rules that developed from it.

In my paper to the Supreme Court, I asked the question: ‘How is it that judges know juries react in a way that requires them to exercise particular vigilance? Is this an accurate assumption? Is the rationale for the rules that this assumption mandates a valid one?’

The assumptions underlying the common law and legislative rules governing the admissibility of tendency and coincidence evidence and the availability of joint trials have been largely untested.

Last Wednesday in Sydney the Royal Commission launched a major empirical study into how juries reason when deliberating on multiple counts of child sexual abuse. This research, which we commissioned, was conducted by Jane Goodman-Delahunty, Annie Cossins and Natalie Martschuk.

The results are interesting. For some they will be counterintuitive and possibly surprising. They will undoubtedly assist all of us to reflect on whether the current rules are appropriate.

It is the world’s largest experimental study of jury behaviour. I anticipate that it will make a significant contribution to our understanding of the accuracy of judges’ assumptions. I would like to discuss here this morning some of the interesting, and at times surprising, results of the research.

Before looking at the results I should draw attention to some of the limitations of this research identified by the researchers. The most obvious is that the mock jurors were restricted in relation to the time which they had to deliberate.
Accordingly, it possible that the results obtained by the researchers would differ from those in a real trial. In a real trial both the presentation of evidence and the jury deliberation would typically proceed more slowly. In a real trial the jurors would also have the capacity to seek clarification from the judge if they required assistance or needed direction. That opportunity was not available to the mock jurors. The jurors also understood that their decisions would be subject to scrutiny. Despite not being informed of the purpose of the study and the hypotheses sought to be tested it remains a possibility that they modified their reasoning and behaviour because they knew they would be the subject of analysis.

Notwithstanding these matters, the fundamental significance of the study is that it allowed the researchers to observe jurors in discussion of the issues which are the same issues which must be examined in our courts every day. The issues of fact and the weighing of evidence in accordance with the law are the same tasks which all juries must perform in sexual assault trials.

Methodology

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. Actors played the part of the complainants, witnesses and the defendant. Real barristers played the part of counsel and a real judge played the judge. All the demands of a real trial were included: opening and closing arguments, examination and cross-examination of the relevant witness and the Judge’s summing up including the relevant directions.

The mock trial concerned charges brought against the defendant by either one or three complainants who had each made allegations of child sexual abuse. The complainants, now adults, each alleged the defendant, their soccer coach, had abused them at different times in the 1990s. The complainants were not acquainted. One complainant, Timothy, whose allegations comprised two counts against the defendant (one penetrative and one non-penetrative) was common to every trial type. His case was constructed to be a moderately strong one. Two other complainants (Simon and Justin), made allegations which were admitted as either uncharged acts in the separate trial of Timothy’s allegations, or as counts in the joint trial. Simon and Justin had weak and strong cases respectively.

There were four variations of trial type:

  • A separate trial of an adult complainant, Timothy, with moderately strong evidence.
  • A separate trial of an adult complainant, Timothy, 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, Timothy, with moderately strong evidence in which tendency evidence from two other prosecution witnesses, Simon and Justin, was admitted.
  • A joint trial with three adult complainants - Simon, Timothy and Justin - who gave weak, moderately strong and strong evidence.

The trials themselves ranged from 45 to 110 minutes. The mock juries had approximately 90 minutes to reach a unanimous verdict on all counts.

The trials were also varied in relation to the directions given to the juries. Some juries were given specific directions on how they could use tendency and relationship evidence. Some juries were also provided with a fact-based question trail to assist them in their deliberations.

Each juror was given a pre-trial questionnaire to control for biases and general attitudes. They were also given a post-trial questionnaire in order to assist in ascertaining what factors drove the deliberations of individual jurors to come to the verdicts they did.

The deliberations of each of the juries were audiotaped and videotaped. Transcriptions of each jury deliberation were qualitatively and quantitatively analysed. Trained law graduates coded the deliberations and assessed the juries’ understanding of the evidence, the juries’ understanding of the directions, whether they made factual errors, and whether they demonstrated unfair prejudice against the defendant.

The qualitative and quantitative analysis of the deliberations is an important feature of this research. A key criticism of a number of previous studies examining the effect of joinder was their over reliance on conviction rates as the relevant measure. Verdicts on their face reveal nothing about the reasoning process that lays 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.

A key task for the researchers was identifying what unfair prejudice in jury reasoning might look like. How is it that joinder is said to infect the reasoning process of jurors so as to produce illogical or unfair outcomes? The researchers determined to measure the hypotheses of prejudice identified by an American judge in the case of US v Foutz, 540 F.2d 733 (4th Cir. 1976). These hypotheses are similar to the judicial concerns in relation to unfair prejudice that have been expressed in the Australian case law.

The first hypothesis was that jurors confuse or conflate the evidence adduced to support different charges in a joint trial. Reformulated as a research question the researchers designed their study to answer the question ‘are juries capable of separating the counts against the defendant in reaching their verdicts in a joint trial?’

The second hypothesis was that a defendant in a joint trial is prejudiced because juries assume guilt due to the amount of evidence against them: accumulation prejudice. Reformulated as a question the researchers asked ‘because of the number of complainants against a defendant in a joint trial, will juries deliver similar conviction rates for counts based on weak compared to stronger case evidence?’

The third hypothesis was that jurors use evidence about the defendant’s other criminal misconduct to infer criminality by reasoning ‘if he did it once, he will do it again’. The research question that followed was ‘are juries in joint trials more prone than those in separate trials to convict on the basis that the defendant has a “criminal disposition”?’

The coders analysed the deliberations to determine if any verdicts were motivated by any of these three types of prejudice. Either inter-case conflation of facts, accumulation prejudice or character prejudice.

Outcomes

The major outcomes revealed by the research were:

  • Collectively, mock juries were capable of distinguishing between the counts. They also based their verdicts on the evidence that pertained to each count whether it was presented in a separate or joint trial.
  • The perceived credibility of the complainant predicted the culpability of the defendant.
  • A complainant’s credibility was enhanced when supported by evidence from an independent source.
  • The same benefit to a complainant’s credibility was obtained by admitting tendency evidence in a separate and a joint trial.
  •  Overall, jury reasoning and verdicts were logically related to the probative value of the evidence.
  • The results provided little indication of mock juries being susceptible to a joinder effect, and, even if there was such an effect, there was no evidence that decisions to convict were the result of impermissible propensity reasoning.
  • Given the verdicts were not based on impermissible reasoning, there was no evidence of unfair prejudice to the defendant.


Importantly, more instances of impermissible reasoning were found in the separate trials and relationship evidence trials than in the more complex trials which contained tendency evidence.

I will explore some of these findings, and discuss some further interesting results, in more detail.

Joinder effect

The researchers defined joinder effect as a statistically significant increase in the conviction rate for an offence when it is tried in a joint trial, compared to the conviction rate for the same offence when it is tried in a separate trial. A previous archival study of conviction rates found an average increase of nine percent in joint trials as compared with separate trials. This study was not, however, limited to child sexual abuse trials and included joint trials for a diverse range of offences.

In order to test the joinder effect hypothesis, the researchers measured the convictions rates for two counts (one penetrative and one non-penetrative) for the complainant with the moderately strong case across all four trial types; separate, separate with relationship evidence, separate with tendency evidence and joint.

They found that the admission of relationship evidence did not increase the conviction rate on either count. The admission of tendency evidence in the separate trial did, however, increase the conviction rates for both counts.

Significant increases in the conviction rates for both counts were also found in the joint trial compared to the separate trial. There were, however, no significant differences between the separate trial with tendency evidence and the joint trial.

The observable pattern was that convictions rates increased when tendency evidence was admitted. Because the increase in conviction rates was observed in both the separate trial with tendency evidence and the joint trial the researchers concluded that it is not the mere fact of joinder itself that increases conviction rates. Rather it is the availability of more independent inculpatory evidence.

Interestingly, further consideration of the conviction rates indicates that both mock jurors individually, and mock juries collectively, were more reticent to convict the defendant on the more serious penetrative account when prosecuted in a separate trial. There were zero convictions for the penetrative offence when prosecuted as a basic separate trial or a relationship evidence trial. In the separate trial with tendency evidence and the joint trial, however, the conviction rates were 62.5 percent and 75 percent respectively. The researchers suggest that multiple witnesses or multiple complainants appear to make an allegation of penetration more credible.

Unfair prejudice

The next step for the researchers was to identify if the increased conviction rates in trials with tendency evidence, both joint and separate, were the result of impermissible reasoning based on any of the three forms of unfair prejudice.

The researchers found that convictions were not based on inter-case conflation of the evidence.

When measuring the capacity of jurors to recall the facts post-trial, jurors in the separate and relationship evidence trial were the most accurate. Jurors in the tendency and joint trials, who had been exposed to two additional complainants, were more likely to confuse the case facts. These findings indicate that it is the complexity of the trial evidence rather than the mere fact of joinder that predicts factual accuracy.

Importantly, however, qualitative analysis of the jury deliberations revealed that errors made by individual jurors were promptly corrected by other members of the jury. The researchers found no evidence that uncorrected errors had any causal effect on the verdicts delivered by juries in trials containing tendency evidence.

A further interesting finding in relation to conflation was that intra-case errors were more common than inter-case errors. Jurors were more likely to make factual errors in relation to the counts pertaining to the same complainant rather than across the complainants.

In relation to accumulation prejudice the results showed that this was not a basis upon which the juries determined their verdicts. The researchers found that there was no significant difference between the conviction rates in the separate trial with tendency evidence compared with the joint trial for the two counts in the moderately strong case. This suggested that the number of counts for which the defendant had been charged was not influencing the conviction rate.

In order to test whether jurors simply accumulated the number of witnesses and whether this was driving conviction rates two versions of the joint trial were run. In one version four witnesses gave evidence for the Crown. In a second version six witnesses gave evidence. The two additional witnesses gave evidence supporting the complainant with the strong case.

The researchers found no differences in the conviction rates for the charges relating to the moderately strong complainant regardless of how many Crown witnesses gave evidence in the trial. Perhaps most importantly, the presence of additional witnesses did not increase the conviction rate for the count with the weakest evidence.

In relation to the third type of prejudice – character prejudice – analysis of the deliberations revealed that only two jurors out of 1,029 made biased comments. These comments were not made in the separate trial with tendency evidence nor the joint trial.

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. The three primary reasons identified were witness consistency, identified by 35 percent of jurors, strong evidence, identified by 34 percent of jurors, and a pattern of grooming, identified by 19 percent of jurors.

Jury Directions

The research also examined the effectiveness of the mechanisms the law has developed to guard against impermissible reasoning – jury directions and fact-based question trails.
In order to ascertain the influence of jury directions the researchers compared jury reasoning in trials with and without a context direction in the relationship evidence trial, and with and without a tendency direction in the separate trial with tendency evidence and the joint trials.

Our researchers report that previous empirical research examining jury directions designed to limit the way in which evidence can be used has suggested that they may be largely ineffective. The results of the present study were in accordance with that finding. When subjected to statistical analysis there were few differences between jury reasoning and decisions in the trials accompanied by the specific directions as to the use of relationship and tendency evidence. The error rates in applying the law on relationship evidence or tendency evidence amongst the juries were unaffected by the presence of these directions.

With respect to the context direction, the conviction rates for the non-penetrative offence were not significantly different as between juries that were given the context direction and those that were not. However, juries that were given the context direction were more likely to convict for the penetrative offence than juries who were not given the direction.

In relation to the separate and joint trials the presence of the tendency direction had no significant influence on the verdict.

The researchers hypothesised that if juries had applied the tendency direction in the manner in which they were instructed they would have expected to find increased ratings by the individual jurors of the culpability and criminal intent of the defendant. They would have expected an increased likelihood of more jurors finding the defendant had a sexual interest in one or more of the complainants which they would use when assessing the defendant’s culpability and criminal intent. This did not occur. The researchers found few examples of explicit permissible tendency reasoning in accordance with the direction. They suggest that this is due to the dense legal language with which the direction is expressed. Because the direction requires juries to make a series of preliminary findings beyond reasonable doubt, juries were concerned that they could not apply it to the first charge by date. This happened to be the weak case. The ambiguity in the direction regarding temporal issues led some juries to ignore tendency evidence rather than misapply it to the first claim.

Jurors indicated in the post-trial questionnaire that they did not find the judge’s directions helpful or easy to apply. Compared to jurors who did not receive the context direction jurors who received the context direction found the judge’s instructions more confusing, found assessing witness credibility and applying the law more difficult and felt that the instructions made it harder to understand the charges, recall the facts, weigh the evidence and assess the case for the prosecution.

Mock jurors found the tendency evidence directions more difficult to understand than the standard directions and that they required greater cognitive effort. Mock jurors in the joint trial who were given the tendency direction did, however, rate the charges as easier to understand compared to jurors in the joint trial who had not received the tendency direction.

Question Trails

A question trail is designed to assist the jury to identify and address the issues which need to be determined in order to determine whether or not the accused should be convicted of the relevant charge.
Question trails were provided to mock jurors in the separate trial with relationship evidence and to the mock jurors in the joint trial. An identical version of each of these trials was also run without a question trial in order to isolate the effects of the question trail itself.

The results demonstrate that the use of a question trial appears to promote a more efficient reasoning process. In relation to the separate trial with relationship evidence juries with a question trail reached a verdict on average 25 minutes faster than juries without a question trail. The jurors who used a question trial also reported that they required significantly less cognitive effort to reach a unanimous verdict.

In the separate trial with relationship evidence mock jurors who used the question trail found the defendant to be less culpable that the jurors without the question trail. This result was reflected in the jury verdicts. The number of hung juries reduced from 44 per cent without the question trail to 20-30 per cent with question trial. The hung juries shifted generally towards acquittals. Analysis of the jury verdicts indicated that there was no significant difference in the conviction rates as between the juries with the question trail and juries without the question trail for the non-penetrative offence. There was a statistically significant decrease in conviction rates for the penetrative offence for juries using a question trail compared to juries without a question trail. There are limitations in the question trail data, in particular with respect to the jury verdicts, as only a small number of juries were used to examine the influence of a question trail.

In relation to the joint trial the question trail had no significant influence on perceptions of the defendant’s factual culpability or upon conviction rates.

Miscellaneous findings

The post-trial questionnaire asked jurors about their expectations as to what they would have been told about the defendant’s prior criminal history. A majority of the 1,029 jurors reported that they would have expected to be informed of prior charges against the defendant, of evidence of sexual misconduct on other occasions and of any prior convictions for child sexual abuse and other crimes. These expectations were not influenced by the education level of the mock jurors.

And finally, a very unexpected finding. The post-trial questionnaire also asked jurors to do what a judge most definitely should not. It asked them to quantify their understanding of the criminal standard of proof. They were asked what number between zero and 100 represents ‘beyond reasonable doubt’.

Perhaps unsurprisingly there were differences in the numerical representation of the threshold amongst the trial types. However, the threshold was lower in the basic separate and relationship evidence trials (less than 90 percent) than in the joint trial (greater than 90 percent). And whereas the definition in a separate trial without tendency evidence was 85.2 percent, the definition in a separate trial with tendency evidence was 88% and in a joint trial with tendency evidence 92.1%.

One final observation, mine, not the researchers, who did not examine this issue. The size of the study allowed the researchers to run exactly the same trail before multiple juries. That is multiple juries were subjected to exactly the same variant of the mock trial. They heard exactly the same evidence and instructions on the law. Yet the conviction rates varied. For example in the joint trial 25 percent of juries convicted on the count in the weak case. 75 percent did not. And they had all observed exactly the same trial. If you had been on trial you would be hoping that the odds of one in four played out in your favour. Researchers, going forward, might seek to uncover what accounts for this differentiation. One benefit of this research is that the deliberations have all been recorded and will be available for future research by others seeking to explore some of these issues.

The Royal Commission’s criminal justice project will draw together our research, relevant material from public hearings, and the views of survivors expressed in private sessions. We are in the process of, and will continue, to consult extensively. We will be publishing a consultation paper on criminal justice issues in September this year. You are all encouraged to consider and respond to the matters addressed in the paper when it is published. I expect that we will publish our final report on criminal justice issues, including any recommendations for reform, as a separate report during 2017.