Andrew L. Urban
In the wake of yesterday’s story about the legal system’s reluctant response to appeals, a significant misunderstanding about what constitutes ‘fresh and compelling’ evidence is explained by legal academic Dr Bob Moles.
We wrote yesterday that “Australian legal scholars Bibi Sangha, Robert Moles and others have criticised the test as overly restrictive. They argue that the “freshness” requirement is unrealistic. The evidence must not only be persuasive but also not reasonably discoverable at trial.
The first crucial thing to note about this last requirement is that the defence is entitled to expect that the Crown will be a model litigant. How naive, as it has turned out. Bob Moles dives into the matter:
Part One
The Australian legal system has gone to great lengths to try to avoid any review of potential wrongful conviction cases.
After the statutory right to an appeal against conviction was first introduced in Australia in the early 1900s, the judges said the statute allowed for just one appeal. When we were seeking to obtain a further appeal for Henry Keogh just over 100 years later, we checked the wording of the legislation. It said nothing about only one appeal. It merely said ‘a person may appeal’ and then specified the grounds for bringing an appeal. The judges had just invented the one appeal limit.
However, the appeal court and the High Court regularly referred to ‘the petition procedure’ for exceptional cases. This allowed a person to petition the Attorney-General (AG) to refer the matter back to the court for another appeal. When we checked the cases on that we found the judges had ‘fixed’ that option as well. In Von Einem v Griffin 1998, the court said the statute gave no rights to the applicant and imposed no duty upon the AG. The AG had an ‘unfettered discretion’ and could refuse a petition for any reason or no reason at all. As we explained in our book Miscarriages of Justice and Criminal Appeals (2015) that interpretation of the law was unconstitutional and contrary to well-established legal authority.
Eventually, we managed to persuade the SA parliament that the current law on this issue was contrary to the International Covenant on Civil and Political Rights (ICCPR). So, legislation was passed to create a right to a second or further appeal in South Australia – subsequently adopted with various adjustments in all other jurisdictions apart from NSW and the NT. It was decided that the test for access to the courts for a further appeal had to pass a higher hurdle than that for a first appeal. That requirement overlooked the provision in the ICCPR which said that at each level of appeal all applicants had to be treated ‘equally’. If the Crown withheld crucial information at the first appeal why should the appellant be at a greater disadvantage in bringing that information to the attention of the appeal court?
The new test for leave to appeal required the applicant to provide ‘fresh and compelling evidence’ of wrongful conviction. Some people have interpreted that to mean evidence which only came to light after the trial was concluded – but that is not correct. In R v Drummond, Peek J explained it as follows:
At trial the accused may rely upon the fact that the Crown will not use evidence which is false or misleading. The accused is not required to use ‘due diligence’ at the time of the trial to identify any possible shortcomings in the evidence put forward by the Crown.
However, if after the trial, the now convicted person finds that some of the evidence produced at trial was in fact false or misleading, then that ‘discovery’ will render that evidence ‘fresh’ for the purposes of a further appeal. It will no doubt also be ‘compelling’ according to Peek J.
That is why I have maintained that the errors identified in the prosecution’s case in the case of Susan Neill-Fraser will satisfy the test for fresh and compelling evidence on a further appeal – the fact that luminol (a preliminary screening test) was used to identify blood on the boat and the dinghy without a follow-up confirmatory test, should be quite sufficient for that purpose. R v Keith Smart [2008] VSC. Once there is some fresh and compelling evidence to trigger a further appeal, any further additional evidence or argument to establish a wrongful conviction becomes admissible just as it would on a first appeal – see R v Keogh (No 2) 2014.
Unfortunately, the limitation of the new appeal right to fresh and compelling ‘evidence’, does not extend the appeal right to those cases based upon ‘non-evidential’ errors. A judicial or prosecutorial mis-direction or incorrect submission to a jury is not captured by the new appeal right. It should be observed that a referral by the Criminal Cases Review Commission in the UK refers to a matter involving new evidence or ‘argument’ as a ground for a further appeal. Those issues would still need to be dealt with by the petition procedure despite the existence of the new appeal right.
See Bibi Sangha, Australia’s Criminal Appeal Rights – Heading in the Wrong Direction? (2025) 47 Criminal Law Journal 586 which outlines the new appeal rights as adopted by each jurisdiction.
Part Two
The case of Derek Bromley raises new issues for the adequacy of the new right of appeal.
At his appeal hearing, Mr Bromley had eight leading and well-credentialed experts who all agreed that the evidence given at his trial was misleading and fundamentally flawed. Two of them (one on each of the two grounds of appeal) were crown-appointed experts. They were both in full agreement with the defence-appointed experts. There is legal authority in Australia which states that where there is unchallenged expert opinion evidence a decision which is contrary to that opinion would not be reasonable. Where the evidence of the expert is only required to raise a reasonable doubt (as on an appeal) then even more is required to enable the jury (or court) to act contrary to that evidence. See Miscarriages of Justice at 9.6 discussing R v Klamo (2008) VR.
The appeal judges did say that they accepted the expert opinion evidence. At that stage, they should have granted leave to appeal, allowed the appeal and entered a verdict of acquittal. Instead they went on to refuse leave to appeal.
The judges said that the opinions of the experts would have to be ‘discounted’ (disregarded) because they had not reviewed all of the other evidence at the trial – and other evidence not produced at trial. As a result, their opinions were incomplete.
It is important to appreciate that upon being requested to provide an opinion for legal proceedings, the expert is given a letter of instruction by the instructing solicitor. They are required to act strictly in accordance with their instructions. They must certify that their opinions have been based ‘wholly or substantially’ upon their expert knowledge which is based upon established scientific principles.
The materials which they are required to examine are provided to them along with the instructions. The forensic pathologist experts on the appeal were given the autopsy report completed by the crown witness at trial along with a transcript of his evidence. They were also provided with copies of the reports of the other experts giving evidence on the appeal.
Neither the DPP nor the defence solicitor asked the experts to examine the statements or evidence of the other witnesses at the trial. If they had been asked to do so the experts would have been obliged to decline the invitation. This is because expert witnesses at trial or on an appeal can only express an opinion on a matter which is beyond the ability of ordinary jury members to work out for themselves.
In this case the forensic pathologists had to deal with circumstances relating to a body submerged in a river for five days before being discovered. What effect did the putrefactive changes have on the ability to determine the manner and cause of death?
The forensic psychologists and psychiatrists had to explain what effect the eye witness’s ‘schizo-affective disorder’ had upon his ability to observe, record and remember what he claimed to have witnessed when the assault was said to have occurred.
Each of those tasks required the expert witnesses to explain to the jury the science behind their reasoning which led to their conclusions. That was obviously beyond the capacity of jury members to work out for themselves. However, when it comes to assessing ‘the other evidence in the case’, that is precisely what the jury are there to do. That means that the expert witnesses cannot do that task. The law says quite clearly that expert witnesses can only give opinions which are beyond the ability of ordinary people, without that expertise, to work out for themselves.
So, when the judge said the experts should have examined the ‘other evidence’ given at trial, the simple answer is that
- They were not asked to do that – so they could not do that;
- It was a task assigned to the jury to do – so they (the experts) could not do that.
It we go back to what Peek J said in R v Drummond:
The prosecution must not lead evidence at trial which is either false or misleading.
If after a trial it is found that they have done so, then that is sufficient for the grant of leave to appeal.
The experts on the appeal all said that the evidence given at Mr Bromley’s trial was false and misleading.
The judges said that they accepted the opinions expressed in the expert reports.
Therefore, it follows, as mentioned above, that they should have granted leave to appeal and overturned the conviction.
To minimise the effect of those expert opinions by asserting that the experts should have done what they were not asked to do – and what the law said they must not do – was what we call a breach of judicial duty. Just as it was when the judges first said there was a right to only one appeal, or when they said that an Attorney-General had an ‘unfettered discretion’ when dealing with a petition.
This situation was particularly disturbing because the forensic pathology experts had given almost identical evidence at the Henry Keogh appeal. The judges there said it was sufficient to warrant the appeal being allowed and the conviction being overturned. Yet when the same experts gave almost identical evidence on the Bromley appeal the judges invented a wholly new explanation for rejecting it. When we say ‘new’ we mean that there had not been any previous case in the law reports where expert opinion evidence had been rejected on the basis of such reasoning. Why would they have done such a thing? Mr Bromley had been in prison for over 40 years and is an aboriginal person. Could either of those factors have influenced the court in arriving at their conclusions?
See Bob Moles, The unfortunate background to Derek Bromley’s Criminal Appeal (2025) 47 Criminal Law Journal 577 for a more detailed discussion of these issues. In that paper, readers will find the following: “Earlier attempts to bring these matters before the courts were blocked for some 13 years by the use of the procedural technicalities of the appeal provisions. The most recent attempt by Mr Bromley, utilising new appeal provisions, was ultimately blocked by a refusal to consider the evidence without further explanation or reason.”
______________________________________________________________
The author investigates the circumstances that led to the trial and conviction of child welfare officer Frank Valentine on charges of historical sexual and physical abuse at the Parramatta Training School for Girls decades earlier. It is his fifth book exposing what he considers to be unsafe convictions.
Kindle $11.99 Paperback $23.20
What is Bob Moles on about with the luminol? Is Sue going to be let down again by those who wilfully misread the trial transcript and try to pull the wool over the eyes of appeal court judges? At Sue’s trial, the jury was told at length that luminol was a screening test for blood and that confirmatory tests for blood were then performed which were negative. Even the expert Max Jones said he did not have any issues with this evidence.
The key to getting Sue exonerated at appeal lay with Meaghan Vass telling the truth (but as if that is ever going to happen!). And if Ellis had not hoodwinked the jury and judge with his gift for storytelling, and if Gunson had seen the prejudicial nature of most of the “evidence” and had it excluded from being presented to the jury, then Sue would likely have been acquitted.. What a train wreck of a case.
Here’s a thought – what if Sue had an accomplice who, on their own, rowed out to the Four Winds on Australia Day and killed Bob? Why is Sue murdering Bob the only reasonable doubt explanation for Bob vanishing into thin air (or deep water)?
Yep, a train wreck…with many train drivers.
Appeals Are Not Enough: A System That Refuses to Learn
The work of Andrew Urban and Dr Bob Moles exposes a legal system that remains deeply resistant to revisiting its own conclusions. Their analysis of the “fresh and compelling evidence” test highlights a fundamental problem in Australian criminal appeals. The system places procedural barriers in front of those seeking to correct wrongful convictions, while simultaneously maintaining a fiction that those barriers are necessary for finality.
Yet even this critique, compelling as it is, only addresses part of the problem. I posit that the deeper failure lies not only in how the system resists correcting its mistakes, but in how it avoids learning from them.
When a conviction is overturned on appeal, the public narrative is one of resolution. An error has been identified, the conviction quashed, and the individual released. In formal terms, justice has been restored. But in practical terms, the moment of acquittal often marks the end of scrutiny rather than the beginning of it.
There is no structured or compulsory process requiring police or prosecutors to revisit the decisions that led to the wrongful conviction. There is no obligation to examine why particular evidence was relied upon, why contradictory material was discounted, or why a case that ultimately proved unsafe was considered viable in the first place. Most strikingly, there is no formal mechanism through which the person who lived through the wrongful conviction can contribute to that analysis.
Instead, the individual is released and left to rebuild their life. Some; as I have pursue compensation. Others withdraw entirely from the system that failed them. What is notably absent is any expectation that the system itself will listen.
This absence reflects a broader structural issue in how criminal cases are initiated and sustained. In general terms, police investigate, form the view that charges should be laid, and assemble the evidentiary brief. That brief is then provided to prosecutors, who assess whether there is a reasonable prospect of conviction and whether proceeding is in the public interest. On paper, this division of responsibility is intended to provide a safeguard; an independent prosecutorial assessment standing between investigation and trial.
In practice, the distinction is less clear. Once a charge has been laid and a narrative established, the momentum of the case can become difficult to arrest. Prosecutors are asked to evaluate a case that has already been framed, often under conditions of public scrutiny and expectation.
In serious matters, media attention, political pressure, and community sentiment can all operate, subtly or overtly, on the decision to proceed. The result is that the theoretical independence of the prosecutorial function may be compromised by the context in which it operates.
Within that environment, evidentiary weaknesses can be reinterpreted rather than rejected. One of the least examined contributors to wrongful convictions is not the presence of entirely false evidence, but the elevation of evidence that is demonstrably unreliable. Cases arise in which key witnesses present accounts marked by internal contradictions, inconsistencies over time, or clear issues with perception and recall.
These weaknesses may be known to investigators, yet the evidence is retained and relied upon because it remains the only available pathway to advancing the case.
At that point, the investigative process can shift in character. Rather than testing competing hypotheses, it begins to consolidate around a chosen conclusion. A person of interest is identified, and efforts are directed toward locating material capable of supporting that identification. This is not a secret it is openly discussed in police investigations logs. This action does not necessarily involve deliberate misconduct; it can arise from institutional pressure to resolve serious offences. Now combined with a belief, sometimes untested, that the correct individual has been identified. However, the effect is the same.
Evidence is interpreted in a manner that favours progression toward charge and prosecution, rather than rigorous exclusion. The treatment of unreliable witnesses illustrates the problem in particularly stark terms. There exists a narrow but significant space between what is legally permissible and what is ethically defensible.
A witness may be treated as credible in the absence of a formal diagnosis of mental illness, even where their own statements reveal contradictions or instability. Legally, such an approach may be sustainable. Morally, it raises more serious questions. If investigators are aware that a witness’s account is inconsistent, yet proceed on the basis that those inconsistencies can be explained away or ignored, the issue is no longer one of evidentiary sufficiency but of institutional integrity.
When such cases collapse on appeal, these underlying decisions are rarely revisited in any meaningful way. The appeal process focuses on whether the conviction can stand, not on how it came to be. Once the conviction is quashed, the system moves on. There is no requirement for police to justify their reliance on particular evidence, nor for prosecutors to explain how the case satisfied the threshold for proceeding.
The individual who has been acquitted may attempt to raise these issues through complaint mechanisms, but such avenues are typically slow, opaque, and limited in their capacity to drive systemic change. The urgency that existed while the person was in custody dissipates once they are released.
The result is a system that corrects errors without absorbing their lessons. Each wrongful conviction is treated as an isolated event rather than as data capable of informing future decision-making. This is a critical omission. If the objective is not only to remedy injustice but to prevent its recurrence, then the system must create mechanisms for structured reflection.
A logical reform would be the introduction of a compulsory post-acquittal review process in cases where convictions are overturned. Such a process would not be punitive in nature, but analytical. It would require police and prosecutors to account for the evidentiary decisions made during the investigation and prosecution, including the treatment of contradictory material and the assessment of witness reliability.
Importantly, it would provide a formal avenue for the acquitted individual to contribute to that review, whether through written submissions, oral evidence, or both.
The value of such a process lies not in attributing blame, but in capturing insight. Those who have experienced wrongful conviction are uniquely positioned to identify the points at which the system failed to test its own assumptions. Without their input, the analysis remains incomplete. Without compulsion, it is unlikely to occur at all.
The emphasis within public discourse tends to rest on high-profile miscarriages of justice, particularly those that remain unresolved. While such cases are important, they can obscure a broader reality. Individuals who have successfully appealed their convictions occupy an ambiguous space: no longer the subject of an injustice requiring correction, but not fully acknowledged as a source of knowledge about how that injustice occurred. Their cases disappear from attention at precisely the moment they could be most instructive.
This reflects a deeper institutional tendency toward closure rather than inquiry. Finality is treated as an endpoint, rather than as an opportunity to examine the path taken. In that sense, the system does not merely fail; it declines to remember.
Urban and Moles are correct to identify the restrictive nature of current appeal mechanisms and the reluctance of courts to engage with fresh challenges to conviction.
However, even a perfectly functioning appeal system would address only the consequences of error, not its causes. If equal attention were directed toward the quality of decision-making at the point of charge toward ensuring that evidence is rigorously tested before a person enters the trial process then the incidence of wrongful convictions would necessarily diminish.
Appeals are, by definition, a remedy of last resort. A system that relies upon them as its primary safeguard has already accepted a significant level of failure. The more difficult task is to prevent unsafe cases from proceeding in the first place, and to ensure that when failures do occur, they are examined in a way that informs future practice.
Until that occurs, wrongful convictions will continue to be corrected in isolation, without altering the conditions that produced them. The system will remain capable of acknowledging error, but unwilling to learn from it.
As a point of record I will revisit this post after my civil case has been concluded as there are matters that I could introduce that are beyond belief. I am bound to confidentially at this point in time for those that were thinking of responding by seeking specific examples.
Yes, you make an excellent point: the system refuses to learn from its mistakes; it doesn’t even try. I especially noted your statement that “One of the least examined contributors to wrongful convictions is not the presence of entirely false evidence, but the elevation of evidence that is demonstrably unreliable.” Many examples in our various reports confirm this as factual.
WE NEED A CCRC ASAP
JURIES NEED TO BE RECORDED
The inherent hubris of the judiciary is gobsmacking.