Accusation as Proof: Uncorroborated Historic Sexual Abuse Allegations

Introduction

In this article, I want to examine the issue of proof raised by the recent appeal to the Supreme Court of Victoria in Pell v The Queen. ¹ Following decades of child sex abuse scandals and the emerging culture of ‘accusation is proof’, some campaigners have argued for legal reforms to increase the number of convictions. The Guardian newspaper has reported that the Children’s Commissioner for England ‘has suggested lowering the burden of proof in cases of child sex abuse to the civil standard of balance of probabilities’. ² A balance of probabilities standard of proof would lead to many wrongful convictions. Given the current culture of accusation is proof, might the law now be powerless to prevent some innocent people from being convicted?

If accusation by convincing storytelling is all that is needed as proof in historic sex offences cases, then criminal justice risks adopting an approach of it is better to convict a few innocent people than let one guilty person go free. A convincing raconteur might convince even the most seasoned professionals of his or her lies. Carl Beech superlatively demonstrated the power of convincing storytelling when he convinced senior police, journalists, MPs, lawyers and others that Sir Edward Heath, Lord Bramall and Harvey Proctor MP, among others, committed sex offences against him when he was a child. Remarkably, the police said his allegations were credible and carried out an investigation over many years ruining the lives and distinguished careers of some of those who were subject to the false allegations. Politicians (including the current Deputy Leader of the Labour Party), ³ journalists, police and other MPs all took his claims as very credible. ⁴

Hysteria-driven policy cannot make our society safer and fairer, if it results in the wrong people (non-dangerous innocent people) being convicted. Academics have conducted research that suggests that large sections of the public embrace a presumption of guilt mentality with alacrity when it comes to historic sex abuse cases. ⁵ Media mobbing (often in contempt of court) and trial by social media have led to a culture of ‘accusation is proof’. This puts enormous pressure on police and prosecutors to try to bring charges in cases where the evidence is not sufficient to warrant action. ⁶ The presumption of guilt hysteria enveloping Associate Justice Kavanaugh’s Senate Judiciary Committee hearing is a manifestation of a broader societal trend that has emerged in some advanced common law jurisdictions.

To make it easy to jail innocent people might provide a proxy form of justice, but it does not give the alleged victim any real justice. It simply uses a scapegoat to create a phantasmagorical imagery that justice has been done. Following a number of acquittals and wrongful convictions of people charged with historic sex offences, it is surely time that England and Australia examine the need for a statute of limitations for criminal cases that rest on the uncorroborated evidence of a solo complainant. ⁷ Some of these cases have been brought to trial 40 plus years after the complainant reached mature adulthood. ⁸ The complainant ought not be allowed to seek justice ad infinitum, if it risks others being wrongly convicted. A line has to be drawn somewhere, since it is nigh impossible to obtain justice for both the accused and complainant so long after the event, if all that is left are the faded memories of the complainant. Minors in contemporary society have an awareness of their rights and the wrongness of child sex abuse in a way that previous generations did not. In the 21st century, timely disclosures are more likely due to the cultural shift around this issue to one of awareness and disclosure. ⁹ A statute of limitations in the current climate of awareness and disclosure would exclude far fewer people from making a complaint, than would have been the case in the 1990s and earlier, when victims tended to blame themselves and feel too ashamed and intimidated to come forward in a timely manner.

Research demonstrates around 95 per cent of the allegations of historic sex abuse are genuine, but trying to distinguish the 5 per cent of false allegations from the genuine allegations is done through an arbitrary process of jury speculation, if the only evidence is the uncorroborated dated memories of the complainant. ¹⁰ In most of these historic cases, no complaint was made at the time and thus no DNA, photographic, CCTV or other incontrovertible forensic evidence was collected in a timely manner or at all. Like Cardinal Pell’s case, these cases hinge on nothing more than the word of the complainant. Little more than a plausible story from an articulate witness has been used to convict people of very serious sex offences resulting in lengthy prison sentences.

Pell v The Queen ¹¹ (heard in the Supreme Court of Victoria) raises important questions about proof of guilt in historic sexual offence cases. In that case, it ought to have been obvious to the fact-finders that the uncorroborated evidence of a single witness (even though not unreliable evidence in itself) 20 years after the alleged offence took place was of insufficient weight to meet the exacting requirements laid down by the criminal standard of proof. I shall try to demonstrate that the judges in the majority (Ferguson, CJ and Maxwell, P) applied a standard of proof that fell well short of the beyond reasonable doubt standard required in criminal cases. If the judges had set the bar higher as far as the quantum of proof was concerned, they would have found the evidence insufficient for supporting a conviction.

Pell v The Queen involved an appeal to the Supreme Court of Victoria against convictions for historic sex offences. The judges relied on probabilistic reasoning that met the civil standard of balance of probabilities rather than the criminal standard of beyond reasonable doubt. It is true that the judges stated repeatedly that they were applying the beyond reasonable doubt standard of proof, but they did not apply it in substance. To apply something in name only is not good enough. It is important that we try to avoid convicting potentially innocent people, even if it means letting some guilty people avoid justice. A criminal conviction resulting in a prison sentence destroys the accused’s life. ¹² In cases involving horrendous crimes such as child sex offences, a wrongful conviction (even an allegation followed by an acquittal can be life-destroying) is certain to lead not only to great stigma and social isolation, but also to severe career damage if not career annihilation. A victim of a wrongful conviction is likely to suffer great mental anguish, if not also post-traumatic stress disorder.

A report on miscarriages of justice observes:

Can an historic abuse conviction ever be unsafe simply due to delay in bringing the prosecution? It is arguable that nobody should be prosecuted—let alone convicted—on the basis of uncorroborated, or barely corroborated, allegations of historic sex abuse.…On the one hand, the serious criminal behaviour, if proved, should be punished. On the other hand, the defendant has probably no means of challenging the truth of the allegations, beyond a bare denial that the events took place. Moreover, to attack the honesty and motives of the complainant—especially if her evidence discloses obvious emotional distress—can simply antagonise the jury. ¹³

In Pell v The Queen, Weinberg, JA (dissentiente) said:

An unusual feature of this case was that it depended entirely upon the complainant being accepted, beyond reasonable doubt, as a credible and reliable witness. Yet the jury were invited to accept his evidence without there being any independent support for it. ¹⁴

Weinberg, JA continues ¹⁵ :

In the past, in cases of alleged sexual abuse, juries were warned, in strong terms, of the dangers of convicting in the absence of corroboration. The law has changed in that regard. That makes the task of intermediate appellate courts charged with having to review the safety of convictions in such cases a particularly important and onerous one.

It will be argued that the judgment of Weinberg, JA is to be preferred to that of the majority in two respects. (1) Weinberg, JA concludes that the evidence of the complainant is sufficiently solid in its entirety to have substantial weight, but falls short of what is needed for a criminal conviction. The complainant’s evidence included some inconsistencies that simply could not be written off as inconsistencies resulting from memory loss. Weinberg, JA also rightly concluded an alibi could not be ruled out beyond reasonable doubt.

I think these points need to be framed far more carefully with respect to what is required by the criminal standard of proof. Firstly, the alibi evidence produced by the defence team was sufficient to make the probability of Cardinal Pell not having an opportunity to perpetrate the crimes a real issue. Once an alibi had been made an issue, the Crown had to prove beyond reasonable doubt that there was no chance that Cardinal Pell had an alibi. The evidence at a minimal demonstrated that the alibi was at least probable: even a conservative estimate would allow a fact-finder to safely conclude that there was 35 per cent probability that Cardinal Pell could not have been alone with the complainant. It might be difficult to argue that it was more probable than not that Cardinal Pell had an alibi, but the evidence shows that the probability of Cardinal Pell having a valid alibi was too high (even if short of a 50 per cent probability) for the reasonable doubt standard of proof to be satisfied. Secondly, there was at least 35 per cent probability that second sexual attack alleged by the complainant could not have been perpetrated in the circumstances described by the complainant. Thirdly, Ferguson, CJ and Maxwell, P did not apply the beyond reasonable doubt standard to these probabilities. Instead, they erroneously held that since what the complainant had alleged could possibility have happened as described by the complainant, the Crown had proved beyond reasonable doubt that these things did happen. This was to misinterpret and misapply the law concerning the quantum of proof required in criminal cases. The fact that there was a real possibility that what the complainant alleged could have happened does not prove that there was an 85 per cent an above probability that it did happen, which is what the beyond reasonable doubt standard requires. It requires such strong evidence that any objective fact-finder reviewing the evidence would 85 times out of 100 conclude that they are sure that the person is guilty.

Evidential Case in Pell v The Queen

Cardinal George Pell was convicted of five sexual offences including two offences against a person who died many years before his trial and who had before he passed away denied having ever been a victim of any sexual offence. The offences were said to have been committed on two occasions in 1996 when he was the Catholic Archbishop of Melbourne. Four convictions were for indecent acts on a child under the age of 16. The fifth conviction was of sexual penetration of a child under 16. The core ground on which Pell sought leave to appeal against conviction was that the guilty verdicts were unreasonable and could not stand given the insufficiency of the supporting evidence. ¹⁶ Pell’s defence counsel submitted that it was not improbable that Pell had an alibi and went as far as asserting that the alibi evidence demonstrated that it would have been impossible for him to have perpetrated the alleged offences. It was submitted that the complainant’s claims were ‘fantasy’, because Pell had never been alone during mass while he was giving religious services as archbishop.

The alleged victims were two choirboys in the St Patrick’s Cathedral choir. Both were aged 13 at the time. The appeal judges refer to the complainants as A and B. It is worth noting that B denied that any sexual attack ever took place, but since B had passed away by the time the matter was brought to trial, it was not possible to seek direct evidence from B or test his evidence in court. Even though B had died, there was hearsay evidence that in 2001 his mother asked if he ‘had ever been “interfered with or touched up” while in the Cathedral choir, B said that he had not’. ¹⁷ The first round of offending was said to have involved both boys and was said to have taken place in the Priests’ Sacristy at St Patrick’s. The Priests’ Sacristy is a private room towards the rear of the Cathedral. It was said that Pell entered the Priests’ Sacristy directly after mass and found the boys drinking wine. Pell allegedly responded by physically grabbing the boys and sexually molesting them while self-abusing himself. It was also said that Pell forced his penis into A’s mouth. The second incident involved only A. On the second occasion, it was alleged that Pell saw A in the procession as it moved into the corridor outside the Archbishop’s and Priests’ Sacristies at the Cathedral and pushed him up against the wall and aggressively groped his genitalia. This was allegedly done in front of everyone in the religious procession without anyone noticing it, but Ferguson, CJ and Maxwell, P upheld the jury finding that Pell was guilty of this charge and in doing so reversed the onus of proof, because it based proof on that being a possibility. ¹⁸ If something is merely a possibility, and there is evidence suggesting there is a high probability (even if that probability is short of 50 per cent) that it could not have happened, then that fact has not been proved beyond reasonable doubt. In this instance, Ferguson, CJ and Maxwell, P applied the beyond reasonable doubt standard in name only, because that standard of proof was not satisfied in substance on the available evidence.

The complainant came forward and made a complaint to the police in June 2015 close to 20 years after the event. As is often the case in historic sex offence cases, it is the word of the victim against that of the defendant. In this case, both Pell and the complainant presented as credible and reliable witnesses. Confronted with the allegations, Pell’s team argued that he had no ‘a realistic opportunity’ to have carried out these crimes. The Supreme Court of Victoria summed up the legal case as follows:

The prosecution case was that A was a witness of truth, on the basis of evidence the jury could be satisfied beyond reasonable doubt that the events he described had occurred. The defence case was that A’s account was a fabrication or a fantasy and that, in any event, the evidence of the opportunity witnesses, taken as a whole, combined to render A’s account ‘either literally impossible, or so unlikely it’s of no realistic possibility’.

The problem for defence counsel in these cases is it cannot do much more than deny the allegations, because incontrovertible alibi evidence (absence a reliable digital record) dissolves in the mist of time. A number of credible witnesses attested that the normal procedure surrounding the archbishop giving mass meant he would never be left alone and if he were it would be for minutes only. There is no need for me to recount the facts in detail in this commentary other than to point out that the general theme of the defence case was that Pell was always surrounded by people and accompanied by the Master of Ceremonies to the Archbishop (Mr Portelli) before, during and after Sunday solemn mass, because the mass as a well-founded religious ceremony followed a rigid pattern. Additionally, after the mass, it was attested that it was normal for Pell to stand on the cathedral steps, meeting and greeting parishioners—and the probability of this was fairly high given that he was a newly installed archbishop. The witnesses attested that Mr Portelli would almost always accompany the archbishop to the sacristies after mass to help him disrobe. The witnesses also attested that the area around the sacristy would be bustling after mass, with altar servers locking away precious vessels and chalices. More significantly, the room where the alleged offences took place was not for the solo use of the archbishop, but was a common room available to other senior members who ‘always’ used the room before and after mass.

While it was possible that Pell could have by chance ended up alone with the boys on the day in question, the 20-year delay created a significant forensic disadvantage for the defence team. If the offence had been reported soon after it was alleged to have happened, witnesses would have had a good recollection about Pell’s whereabouts on the day in question and whether he departed from normal procedure on the day. The 20-plus-year delay meant the defence team could only present evidence about the usual norm for the archbishop to be accompanied and about the room being a common room that was always used by other senior members before and after mass. This was all then left to the jury to speculate about what might have happened on the day in question.

Ferguson, CJ and Maxwell, P said:

[T]he appeal court reviews the evidence as it was presented to the jury and asks itself whether—on that factual material—it was reasonably open to the jury to convict the accused.…Having reviewed the whole of the evidence, we would answer that question affirmatively. In our view, it was open to the jury to be satisfied beyond reasonable doubt that Cardinal Pell was guilty of the offences charged. That is, there was nothing about A’s evidence, or about the opportunity evidence, which meant that the jury ‘must have had a doubt’ about the truth of A’s account. ¹⁹ (emphasis added)

…

An argument of impossibility, if supported by the evidence, is effectively unanswerable. As we have said, the onus of proof required the prosecution to defeat that argument. It was both necessary and sufficient for that purpose to persuade the jury that the events were not impossible ²⁰ and that there was a realistic opportunity for the offending to occur.

Additionally, Ferguson, CJ and Maxwell, P said that A knew facts that he could have known only if he was telling the truth, ²¹ but this seems a strong finding given that anyone regularly attending the Cathedral choir as a participant could have gained an intimate knowledge of Cathedral’s layout, people and processes. Many people can return to their old preparatory school decades later and know exactly where everything is and remember the location of the Head Master’s office and so forth. It was an important consideration that A knew the internal layout of the Priests’ Sacristy. Despite the trauma of the alleged sexual penetration and the fact that A had allegedly been in the room only once, A was able to describe it well and remember the exact location of the wine cabinet. This seems remarkable if he had only ever visited once 20 years before, but not impossible. It is also possible that it was common knowledge (at the time) that the Priests’ Sacristy was being used by the Archbishop to robe and disrobe while Archbishop’s Sacristy was being renovated—after all the rooms are right next to each other. ²² Furthermore, the boys might have even seen the Archbishop entering the room at times, but that does not mean they needed to be in the room to know that he was using it.

A further issue raised by defence counsel was ‘Only a madman would attempt to sexually abuse two young boys in the Priests’ Sacristy immediately after Sunday solemn Mass’. At the time, Pell was a high-profile public figure in his late 50s and was leading a very public campaign (one of the first by any Catholic leader) against paedophiles in the Catholic Church. The submission that Pell would be unlikely to act like a madman is relevant to his propensity for perpetrating this type of offence. Some powerful people have acted like madman in seeking inappropriate sex in the height of their careers, but that does not prove or disprove anything. Ferguson, CJ and Maxwell, P give the example of the Australian soap opera star, Robert Hughes, who was convicted of a number of historic sexual abuse offences for violating child stars on the set of his television programme. However, Hughes had been reported to the police as long ago as the 1980s and his propensity for this sort of offending was exposed since more than one victim came forward. ²³ The former chief of the International Monetary Fund, Dominique Strauss-Kahn, is another example of a high-profile individual brazenly seeking sex in the peak of his career, but unlike Pell, Dominique Strauss-Kahn had other independent and credible allegations made against him. ²⁴ US Congressman Anthony Weiner is another example of a public figure at the height of his career risking it all to sexually harass minors, but Weiner too was a serial offender whose conduct was not out of the blue. ²⁵

The submission that it was improbable that Pell would act like a madman should have been located in an argument concerning propensity, but the defence team used it too generally and this seemed to result in the judges not giving it any weight at all. What might have been submitted is that Pell was of excellent character and that the expert evidence shows paedophiles are normally recidivists, ²⁶ yet no allegation had ever been made before against Pell after 40 years in the church. We were a year into the ‘Me Too’ movement by the time of Pell’s trial in 2018. Coupled with that, the sex abuse scandal involving the Catholic Church had been exposed for decades by then and many complainants had come forward concerning other priests in that organisation. It would have been extraordinarily brazen for Pell to sexually penetrate a boy in front of another boy in an unlocked ‘common’ room that was always used before and after mass by other senior members of the church, but not impossible even though the window of time for such offending could not have been more than minutes. Just because something is possible does not prove it was probable—so probable that a reasonable jury could be sure it did in fact happen.

While there is a debate about recidivism rates for child sex abusers, ²⁷ the evidence suggests that child sex abusers normally act more than once and do not randomly start acting in this way in their late 50s. ²⁸ Pell was famous and sex abuse in the church had been a high-profile issue for decades, so it seems remarkable that no one came forward with similar stories from the 1980s, 1990s and 2000s. It is at least probable that Pell did not randomly behave as alleged for the first time in his mid-50s and it is probable that if he had done this more than once in the past as is usually the case with paedophiles that someone would have come out. Zuckerman convincingly argues, ‘As long as we believe that human behaviour is not entirely arbitrary and unpredictable but, on the contrary, that behaviour is on the whole dictated by the mental make-up of the subject, character has both predictive force and probabilistic significance concerning a person’s past acts or omissions’. ²⁹

Furthermore, complainants now remain completely anonymous and that provides a strong incentive for people to come forward. They also give evidence remotely. A potential compensation settlement from the Catholic Church would be a further incentive for victims to come forward. Propensity surely should be weighed up with all the other evidence. ³⁰ In Victoria, s 110 of the Uniform Evidence Act permits evidence to be presented to prove the accused’s good character. Good character evidence need not be limited to demonstrating the defendant has a clear criminal record, but could also show that no allegation of past wrongdoing exists. This might be buttressed with expert evidence showing that there is a real probability that an offender would not start a life of paedophilia in his or her late 50s and that there is some probability that others would have come forward, ³¹ if he or she had routinely abused children in the past, which is often the case with paedophiles. ³² The good character evidence is of probative value when considering the defendant’s propensity for perpetrating the type of offence charged. In England and Wales in the Court of Appeal (Criminal Division), Burnton, LJ said ³³ :

The good character of the defendant is relevant in relation to credibility and to propensity, and he is entitled to the full direction on both. However, in an historic sex abuse case such as the present, there are two aspects of propensity to be considered. The first is that applicable even where the allegation is of a recent offence: a person of good character is less likely to have committed the alleged offence. The second results from the passage of time since the alleged offence: the fact that the defendant has not committed any offence, let alone one involving sexual abuse, since the date of the alleged offence goes to the likelihood or otherwise of his having committed the offence or offences. Such a direction is particularly apt where delay since the date of the alleged offences renders it more difficult for the defendant to defend himself, as is so often the case.

Denis argues that evidence of good character can have probative value with respect to propensity, because ‘it may yield and inference that the accused is unlikely to have committed the offence charged…‘. ³⁴ It is probable that Pell’s version of events is as true as the complainants, since it is unlikely that he would start offending out of the blue in his late 50s.

Ferguson, CJ and Maxwell, P were also influenced by the fact that no evidence had been produced to suggest A had a motivation for lying. Off course, there is no obligation on the defendant to prove A had a motivation for lying, but as a tactical matter if the defence team could have unearthed evidence of a motivation it could have made the truthfulness of A’s testimony an issue. ³⁵ Motives that have been identified in research into false allegations of sexual abuse include (1) revenge ³⁶ and (2) greed. ³⁷ That there could be many plausible reasons why a person might fabricate such a story is irrelevant, if there is no evidence of fabrication. Following the Carl Beech scandal, Sir Richard Henriques’ report of the matter included the following recommendation ³⁸ :

Investigators should be informed that false complaints are made from time to time and should not be regarded as a remote possibility. They may be malicious, mistaken, designed to support others, financially motivated, or inexplicable. When considering non-recent allegations against prominent people they should give full consideration to all background information.

There is a wealth of research that supports the aforementioned recommendation. False allegations might only be 5 out of 100, but the problem of a false allegation is compounded in a historic sexual offence case where a conviction can rest on uncorroborated evidence from a single complainant decades after the alleged event. Some false allegations will make it to trial and some genuine allegations will not make it to trial. ³⁹ When a false allegation makes it to trial and there is no evidence to put its falseness in issue, the defendant risks a wrongful conviction simply because there is nothing to prove or disprove the veracity of the complainant’s claim. If a person had been molested by a priest and the priest had passed away before he had the wherewithal to report the matter to the police, it would not be improbable that such a person might use the Head of the relevant church as a proxy for justice—a fortiori if he had personal knowledge of that person and recalled seeing him in person at the Cathedral. ⁴⁰ As for greed, it would have to be proved beyond reasonable doubt that the defendant had a plan to succeed in a criminal trial to facilitate a civil action for damages and so forth. There simply was no evidence of that so the complainant’s honesty could not be made an issue. There was no evidence raised to put the claim A was a fantasist in issue—so the prosecution did not have to prove beyond reasonable doubt he was not. Since A was an articulate 35-year-old university-educated man when the matter came to trial, it might not be surprising he presented a fairly consistent story.

The only manifest motivation was that he had been a victim of a serious sexual crime and was reporting it for that reason alone. ⁴¹ Hence, the Crown did not have to address the issue of motivation since there was no evidence to make it an issue. ⁴² Some media commentators made a lot of noise about a high-profile case in the USA with some similar facts such as a priest sexually violating a boy after catching him drinking wine in the church sacristy. ⁴³ The facts differed in that case since the priest groomed the boy and targeted one boy on his own and in a secure place where being caught in the act was very unlikely. In Pell v The Queen, the alleged offending took place in circumstances where it was at least probable on the evidence that there was a hive of activity taking place near the front door of the room where the offending was alleged to have taken place. It also was highly probable that other members of the cathedral were using the room before and after mass, because they always used it during those times. The facts are similar, but it is possible that the same sort of incident happened more than once across the world and the wrongful wine drinking was used to initiate, mask and facilitate sexual abuse against the child. That something is possible does not help a case one way or the other in the way that impossibility does. Speculating about possibilities ⁴⁴ can only lead to wrong conclusions and ought to be avoided. ⁴⁵

The burden of proof is placed upon the prosecution and remains with it throughout the trial. Obviously it does not shift to the defendant merely because the prosecution make a prima facie case. What it comes to is that on a trial on indictment the prosecution must give enough evidence to take the case to the jury, and also give the greater amount of evidence necessary to convince the jury. ⁴⁶ These two burdens are now frequently called the evidential and persuasive burdens, respectively, or the burden of production and the burden of persuasion. Accordingly, defence need not actually prove anything, but instead must only present evidence on each element sufficient to avoid a finding of guilt. Therefore, if the defence gives cogent alibi evidence (or evidence that the complainant had a strong motivation for lying and so forth) so that the jury are left in doubt, the defendant is entitled to an acquittal.

The next two issues discussed by the judges in the majority were the inconsistencies in some of A’s account of what happened and his failure to complain. Ferguson, CJ and Maxwell, P said that a 13-year-old would not be able to remember the details surrounding an incident 20 years before and that the inconsistencies were not about fundamental issues. ⁴⁷ Weinberg, JA (dissentiente) rightly points out that there were some inconsistencies about two fundamental aspects of the complainant’s claim: (1) A could not remember whether Pell had pants on under his robes and thought the robe could be parted, when it could not. (2) A’s description of the second sexual attack was unconvincing and was hard to reconcile with A’s own description of the surrounding circumstances. Weinberg, JA rightly said the second incident, ⁴⁸ where A alleged that Pell groped his genitalia in front of everyone as the procession was leaving the cathedral without it being noticed, contained some serious inconsistencies. ⁴⁹ Ferguson, CJ and Maxwell, P seem to uphold a guilt finding on the grounds that the second incident was possible and not impossible, but fail to apply the beyond reasonable doubt standard of proof to their possibility. ⁵⁰

Ferguson, CJ, Maxwell, P and Weinberg, JA were unanimous in holding that there is a wealth of evidence that demonstrates that children (especially from that era) did not necessarily come forward for decades, if ever. ⁵¹ A fact-finder would not be left in doubt of a child abuse complainant’s credibility as a witness based on the fact that his or her disclosure was delayed.

It is respectfully submitted that Ferguson, CJ and Maxwell, P err by concluding that it is likely that Pell did as alleged simply because the claims were a real possibility. The judges run into era by not addressing the probability of an alibi as raising a reasonable doubt about guilt. A part of the problem in these historic sex offence cases is it comes down to the jury believing one side or the other. ⁵² ‘Putting this either/or proposition to the jury excludes the third alternative; namely, that the jury, without believing the accused, after considering the accused’s evidence in the context of the evidence as a whole, may still have a reasonable doubt as to his guilt’. ⁵³ Alas, in these historic sex abuse cases, the time delay before a trial commences provides a charter for being inconsistent and being able to legitimately haze over questions without looking dishonest. ⁵⁴ For example, Ferguson, CJ and Maxwell, P concluded ⁵⁵ :

It seems unsurprising that his recollection was confined to being in the Priests’ Sacristy…It is unremarkable that, when his attention was drawn to other objective circumstances, A accepted that his recollection was mistaken…That is understandable and consistent with human experience. If there is an unexpected incident, what happened leading up to it, or the route taken to arrive at the site of the incident, may assume little or no importance for the person affected. That person may well ask, ‘What does it matter how I got there? What matters to me is what happened when I got there’. The detail of the events and circumstances before and after the incident may not be retained.

It simply means that credibility cannot be impugned on the grounds that the complainant does not remember all the background of the matter with exactitude 20 years after the event. However, the mist of time excuse allows a lot to be hazed over if a person is lying. Many wrongful convictions have rested on this sort of limited testimony. It is testimony that cannot be showing to lack credibility, but if the event had not been so historic it might be better challenged. Given the weight that has to be given to a single witness statement combined with the fact that Pell was of good character and had produced sufficient evidence to show that his alibi was at least a probability, the Crown fell well short of proving guilt beyond reasonable doubt. More is required than proof that the defendant is probably guilty and accusation alone (a fortiori accusation of a decades-old incident) will rarely be proof of guilt.

The Beyond Reasonable Doubt Standard of Proof

There is a wealth of research demonstrating that both judges and scholars regard the criminal burden as requiring a probability of guilt to be around 90–95 per cent. ⁵⁶ Even a liberal reading of the beyond reasonable doubt standard would not permit a conviction to stand where there is a less than 85 per cent probability of guilt. ‘Blackstone would have put the probability standard for proof “beyond a reasonable doubt” at somewhat more than 90%, for he declared: “It is better that ten guilty persons escape than one innocent suffer.”’ ⁵⁷ In In re As.H. ⁵⁸ Schwelb, JA said: ‘Although the Chicago study alone is not dispositive of this appeal, it reveals that very few judges, if any, would have regarded an 80% probability as sufficient to prove guilt beyond a reasonable doubt, and that all of them would have considered a 70% probability as altogether inadequate’. In England and Wales, Lord Denning said: ‘It need not reach certainty, but it must carry a high degree of probability’. ⁵⁹ High probability here does not mean ‘probably guilty’ or that it is ‘not improbable’ or ‘unlikely’ that the defendant did as alleged, it means that the fact-finder must be sure of guilt. In U.S. v Fatico, ⁶⁰ Weinstein, J said:

The high probability required in criminal cases, however, does not mean that most guilty people who are tried are acquitted. In almost all cases the guilt is so clear or the doubt so great that precise quantification is of no moment. In the few instances which this court would roughly estimate on the basis of experience as no more than one in ten cases it may make a difference whether the trier’s perception of the standard is 80, 90, 95, or 99%. (emphasis added)

Obviously, a historic sex offence case relying on uncorroborated evidence from a solo complainant is one of those one in ten cases that requires a fine-grained analysis of the probability of guilt, because an appellant court reviewing the fact-finder’s initial conclusion has to make sure it was a reasonable conclusion to draw upon all the evidence. It is doubtful that the testimony of the complainant alone could have allowed a jury to be 65 per cent sure that Pell was guilty let alone 85 per cent sure. Even if the complainant’s testimony alone allowed for a reasonable jury to conclude that it was 85 per cent sure of guilt, on the other side of the evidential scales was probative evidence about the probability of Pell having an alibi, the probability of the second incident not being likely in the circumstances described by the complainant and, evidence that it was unlikely that someone with Pell’s good character and clean record had a propensity to so offend, especially as he had been alleged to have perpetrated an offence where recidivism is normally prevalent. All these factors would have raised a sufficient doubt about guilt to make a conviction unreasonable. Lord Denning did not give a number, but made it plain that the criminal standard of proof required a very high probability of guilt. Thinking in terms of probability of guilt is not new to the criminal law. Probability has been used to draw inferences about guilt for more than 900 years. Glanvill wrote ⁶¹ :

[T]he truth of the fact shall, then, be inquired into, by means of many and various inquisitions and interrogations, made in the presence of the justices, and that, by taking into consideration the probable circumstances of the facts, and weighing of each conjecture that tends in favour of the accused, or makes against him…’. (emphasis added)

Plowden, discussing a civil case, said:

But where the matter is so far gone that the parties are at issue, or that the inquest is awarded by default, so that the jury is to give a verdict upon that which appears most probable, and by that same reason that which is most probable shall be good evidence. ⁶² (emphasis added)

Sir Edward Coke, said:

And Bracton said, there is probatio duplex, viz. viva, as by witnesses viva voce; and mortua, as by deeds, writings, and instruments. And many times juries, together with other matter, are much induced by presumptions; whereof there be three sorts, viz. violent, probable, and light or temerārius. Violenta praesumptio is many times plena probatio; as if one be run through the body with a sword in a house, whereof he instantly dies, and a man is seen to come out of that house with a bloody sword, and no other man was at that time in the house. Praesumptio probabilis moves little; but praesumptio levis seu temeraria moves not at all. ⁶³

Coke was asserting that an inference of guilt could be drawn when the facts showed that it was extremely probable that the defendant was the culprit. These old presumptions were maxims that gave the fact-finder some guidance about what evidence might allow for a particular inference. ⁶⁴ There was never in English law a presumption of strict liability resting on the notion that a person intended the natural and probable consequences of her actions. ⁶⁵ The use of the word presumption by Coke and later writes was to identify that it was permissible to draw an inference based on a high probability. This simply meant it was permissible for the jury to infer guilt upon a high probability, but it was not bound to infer guilt. ⁶⁶ Per contra, a legal presumption is a mandatory starting point and it has to be rebutted, if the fact-finder is not to take the presumed fact as proved. Unlike an irrebuttable presumption, a permissive inference does not usurp the role of the fact-finder, it leaves the fact-finder to draw an inference from the evidence as it sees fit. ⁶⁷ ‘[A] permissive presumption is one that allows, but does not require, a jury to infer a presumed fact from proof of the other facts.…It “suggests to the jury a possible conclusion to be drawn if the State proves predicate facts, but does not require the jury to draw that conclusion.”’ ⁶⁸

We need not consider guides for drawing inferences in detail here, because our focus is on the quantum of proof issue. The fact-finder need not be 100 per cent certain because that level of certainty would be impossible to achieve unless the fact-finder witnessed the crime first-hand, but the fact-finder must be sure of guilt. ⁶⁹ Judges are often advised not to try to explain to juries what is required by the ‘beyond reasonable doubt’ standard of proof, when in reality a simple explanation in percentage terms would do more good than harm in the rare cases where it is needed. If an appellate court reviewing the evidence came to the conclusion that it was only probable or possible that the alleged crime took place, that ought to lead to an acquittal.

There must be sufficient evidence for a reasonable jury to infer there was above an 85 per cent probability that the defendant did what was alleged. Each piece of evidence needs to be evaluated separately, but it is the evidence considered as a whole that allows the fact-finder to determine guilt or innocence. The percentages that might be discussed by experts providing expert evidence, such as the defendant’s DNA is a statistically certain match with samples taken from a murder victim, is a completely independent assessment to considering guilt on the whole of the evidence. The defendant’s DNA might be a million to one match, but if the victim was seen alive on CCTV at 18:00 giving a contract law lecture to hundreds of students and the alleged defendant was recorded by face recognition technology as he boarded a flight for New York at 17:00, ⁷⁰ no weight need be given to the DNA evidence—even though it is a 100 per cent probable that the defendant’s DNA found its way onto the murder victim. This is an extreme example to make the point, because if the alibi were that categorical no charges would be brought. ⁷¹ The point is that an individual piece of evidence can be certain, but it has to fit with all the other evidence to prove guilt beyond reasonable doubt.

Some will object to any mention of the term probability when directing criminal jurors, but such tests have long existed in criminal law jurisprudence. Glanville Williams points out: ‘The Romans had the maxim that it is better for a guilty person to go unpunished than for an innocent one to be condemned; and Fortescue turned it into the sentiment that twenty guilty men should escape [punishment] through mercy rather than one just man be unjustly condemned’. ⁷² Dworkin argues: ‘People have a profound right not to be convicted of crimes of which they are innocent’. ⁷³ This important common law principle is buttressed with other common law doctrines such as in the case of doubt a criminal statute or doctrine is to be ‘strictly construed’ in favour of the defendant. ⁷⁴

It is a fundamental human right that a person be presumed innocent until proven guilty beyond reasonable doubt. ⁷⁵ This means that a defendant is entitled to the benefit of a reasonable doubt even though it means many acquittals are in dubio pro reo. The Crown does not have to prove that the defendant did what is alleged, but it has to remove any reasonable doubt. It might be crude to put it in percentages, but the ancient doctrine going back to Fortescue and later endorsed by Blackstone and Lord Denning supports the contention that if there was even a 20 per cent probability that the defendant was innocent, that would raise sufficient doubt to warrant an acquittal. Judges are reluctant to instruct juries about what ‘beyond reasonable doubt’ requires, but leaving a jury confused about where the civil liability standard of proof ends and the criminal liability standard of proof begins is not good enough. The American cases are replete with examples of fact-finders resorting to percentages to emphasise that something substantially greater than ‘on balance’ is required to justify a criminal conviction.

Beyond reasonable doubt requires sureness, not certainty, but sureness is a very high standard to meet. ⁷⁶ An event can be said to be likely if there is more than a 50 per cent chance of it occurring. To say that it is likely that a priest would sexually abuse a child is a nonsensical conclusion, because the starting point is to look at the empirical reality and then extrapolate from that to the individual defendant after factoring in all the relevant circumstances. Research has ‘found that 3.5 per cent of the clergy offenders were responsible for approximately 26 per cent of all allegations of sexual abuse’ in the Catholic Church. ⁷⁷ Even if these figures are conservative, there is no evidence to suggest that more than 5 per cent of the population of the Catholic clergy sexually abused children. Consequently, the finding that it was ‘not improbable’ that a senior member of the clergy would have been brazen enough to sexually abuse a child in an unlocked room in a busy Cathedral is the wrong way to approach the question. The fact-finder has to be able to conclude that this particular priest did in fact perpetrate sexual abuse. It would be no different to holding that it is not improbable that the complainant is lying, because ‘[r]eviews of the more rigorous international studies suggest a false allegation rate of 2-8% and 2-10% of rape offences initially recorded as crimes by the police’. ⁷⁸ The fact that a small percentage of priests molest children does not mean it is probable this priest molested a child and the fact that a small percentage of complainants make false sexual abuse claims does not mean it is probable that this complainant lied.

For a person to be convicted of a serious crime, the Crown should present sufficient evidence to make it reasonable ⁷⁹ for a jury to assert it is ‘sure’ that the defendant did what is alleged. In England and Wales, the Court of Criminal Appeal has held that it would be better to ask the jury whether they ‘must be sure that the defendant is guilty of a charge on the indictment’. ⁸⁰

Judges are advised by the Judicial Studies Board, as they have been for many years, to direct the jury that before they can return a verdict of guilty, they must be sure that the defendant is guilty. ⁸¹

Sure does not mean certain. ⁸² Only the victim or the defendant or the complainant will know if the other party is lying in historic sex cases when it is the defendant’s word against complainant’s word. ⁸³ Juries have to try to infer from the evidence that the defendant did in fact do X. If defendant was caught on CCTV forcing his penis into a minor’s mouth, the jury might be sure the defendant penetrated the child, but even here it cannot be certain. The video might be a deep fake video even though the odds of that are extremely low. It is difficult to see how anyone could be sure beyond reasonable doubt, unless they can infer from the evidence that there is an 85 per cent plus probability that the defendant is guilty. In other words, in a hundred cases with the exact same facts as those described in Pell v The Queen, would 85 juries return guilty verdicts? It is very doubtful even 50 juries out of 100 would return guilty verdicts on the evidence that was presented in Pell v The Queen. Pell was retried because the first trial resulted in a hung-jury.

It is irresistible for most humans not to reason and deduct with some percentage or probability in mind, but the courts try to avoid directions that involve fixed percentages or probabilities. Lord Denning simply said: ‘It…it must carry a high degree of probability’. ⁸⁴ In Pell v The Queen, defence counsel argued that it was improbable that someone of Cardinal Pell’s character, fame and clean record would be so brazen to sexually penetrate two boys in an unlocked room that was used by many other clergy at a busy time in the Cathedral and that it was impossible for Pell to have done as alleged because he was never alone. In response to these arguments, Ferguson, CJ and Maxwell, P said ⁸⁵ :

The defence wanted to persuade the jury that, even if they rejected the contentions of fabrication and fantasy, and even if they rejected the arguments about improbable brazenness, they should nevertheless accept that the events simply could have not happened as A alleged. To make good that argument it was necessary to persuade the jury that the opportunity evidence established ‘practical’ impossibility, thereby excluding any realistic opportunity for the offending to occur.

Ferguson, CJ and Maxwell, P in effect reverse the onus of proof here, even though they repeatedly emphasise the burden of proof is with the Crown. It is not necessary for the defence counsel to prove the alibi beyond reasonable doubt, but for the Crown to prove beyond reasonable doubt that there was no chance that an alibi existed. Defence counsel put it too strong when it argued that it was absolutely impossible for Pell to have perpetrated the crimes, because the alibi evidence did not support such a submission. There was a possibility (even if a low one) that Pell by chance found himself alone with the boys and did as alleged. The evidence produced by the defence team was enough to demonstrate that an alibi was a probability of at least 35 per cent. The 20-year time delay meant that the defence team was unable to produce evidence to support its submission that Pell had a watertight alibi. Nonetheless, the Crown failed to prove beyond reasonable doubt that there was no probability of an alibi existing.

Conclusion

These sorts of crimes are repulsive and few people would have any sympathy for those found guilty, but the law must protect people from wrongful convictions. It is important that lawyers speak up for those accused of atrocious crimes such as child sex abuse and terrorism and ensure that basic processes are followed. I need not quote in full Niemöller’s prose, ‘First they came for the socialists, and I did not speak out—Because I was not a socialist. Then the unionists…Then they came for the Jews, and I did not speak out—Because I was not a Jew.…Then they came for me—and there was no one left to speak for me’. When campaigners call for the evidential burden of proof to be reduced from beyond reasonable doubt to the balance of probability standard, it is because they assume they will never find themselves subject to a false allegation. They work from the premise of guilty until proven innocent and from the premise it is not possible there could ever be such a thing as a false allegation. It is true that most allegations are credible and true, but as I pointed out above the research shows up to 10 per cent of allegations have been shown to be false as far as some categories of sexual offences are concerned.

The beyond reasonable doubt standard is fundamental in historic sexual offence cases, because there are no statutory limits on the prosecution of crimes in England and Wales or Australia, apart for summary offences: nullum tempus occurrit regi (time does not run against the Crown). Weinberg, JA rightly concluded: ‘Where, however, on reading the transcript and hearing argument the Court of Appeal has a serious doubt about the verdict, it should exercise its power to quash. We do not think that quashing the jury’s verdict where the court believes it to be unsafe ⁸⁶ undermines the system of jury trial’. ⁸⁷ Weinberg, JA was right to conclude that any advantage the jury had was minimal since all the key witnesses had been recorded and the appellate judges had the advantage of watching those recordings. The reality is the jury system is not special. Ninety per cent of the world tries criminal cases without a jury and there is no empirical evidence to suggest trial by jury is more accurate than a trial by a judge. ⁸⁸ It does not matter whether the fact-finder was a judge or a lay juror, what matters is that the fact-finder was able to access the evidence to be in a position to make conclusions about facts from it.

In Pell v The Queen, Weinberg, JA (dissentiente) said ⁸⁹ : ‘where the defence puts forward an alibi, or its equivalent, it is not for the jury to ask themselves whether they accept that alibi to be true. The law requires that they must acquit unless satisfied beyond reasonable doubt that the alibi has been entirely disproved’. Weinberg, JA concluded ⁹⁰ that the alibi was sufficient to raise doubt as to guilt. The room where the offences were said to have taken place was a busy room and the time frame for the offences was so tight that there was a probability that it was not possible for Pell to have done as was alleged. ⁹¹ Weinberg, JA said:

Even a mere ‘reasonable possibility’, unrebutted by the prosecution, that what Portelli and Potter said might be both truthful and accurate, would give rise to a complete defence, and would necessitate an acquittal. Once again, it must be remembered that at trial, the prosecution did not suggest that either man had lied. In these circumstances, I consider that I should proceed on the same basis, though I would have arrived at that conclusion irrespective of the approach taken by the prosecution, at trial. ⁹²

A fortiori the beyond reasonable doubt standard of proof was not met because: (1) there was at least a 35 per cent probability of an alibi existing; (2) there was no evidence to impugn Pell’s good character in the absence of any other complaints; (3) the weight of the uncorroborated testimony of the complainant was not sufficient weight on its own not only because of the inconsistencies concerning whether Pell had trousers on under his robes and whether his robes could be parted, but because it was based on faded memories and was uncorroborated; and (4) there was at least a 35 per cent probability that the second incident could not have taken place in front of many people and go completely unnoticed. For the probability of the alibi existing and the probability of the second incident not having taken place to be completely ignored, the evidence would have to support the inference that those probabilities were so low that they could barely be measured. As I said above, we ought to be careful trying to reason in percentages, but even as a rough guide it would not be unreasonable to claim that there was at least a 35 per cent probability that there was an alibi and a 35 per cent (if not greater) chance that the second incident did not happen. This means the jury ought not have concluded it was sure that there was no alibi or that the second incident did in fact happen, because at the very best the jurors could only have been 65 per cent sure of those factors.

As for the weight of the evidence of Pell’s good character, it ought to have led to a direction and ought to have had some weight. Similarly, the weight of the complainant’s evidence ought to have been discounted not only because it was not corroborated, but because there was some confusion about key facts concerning the actual sexual abuse itself, when one would have expected those facts to have been remembered more clearly, since such a horrific incident would play on the mind of a victim in the weeks, months and years immediately following the event. Even though the complainant presented as a reliable and credible witness, his evidence when weighed with all the other evidence would not allow a reasonable fact-finder to be 85 and above per cent sure of guilt. At best a fact-finder might safely conclude guilt by applying a civil standard of on the balance of probabilities, but that is simply not sufficient for convicting in a criminal case. Ferguson, CJ and Maxwell, P erred in holding that there was not at least a 35 per cent probability that Pell was never alone and that there was not at least a 35 per cent probability that the second incident could not have gone undetected. The defence team raises many other obstacles, but the probability of an alibi and of the second incident alone are strong enough to warrant an acquittal. From the evidence I read, and from my experience of having read thousands of criminal cases, I would not have concluded that he was guilty even on the civil standard of balance of probabilities. I doubt the case should ever have been brought to court, so it is remarkable that it was and that a conviction was upheld. This sort of miscarriage of justice is something one would have expected in the 1980s, not in the 2010s.