Writing from bitter personal experience after spending 1,400 days on remand, STEVEN FENNELL explains why he should have received bail, especially given the High Court’s indignant, virtually instant quashing of his conviction. Perhaps he wasn’t sufficiently high profile – like Ben Roberts-Smith?
In the world of high-stakes criminal law, the mantra “bail is not punitive” is frequently recited. It upholds the presumption of innocence: a person should not be punished by pre-trial detention before guilt is proven beyond reasonable doubt. Yet, as someone who endured nearly four years on remand for a charge I was ultimately acquitted of, I can attest from lived experience that the “barometer of bail” swings dramatically depending on who stands in the dock.
The recent granting of bail to former SAS soldier Ben Roberts-Smith in April 2026 highlights this disparity. Charged with five counts of murder relating to alleged war crimes in Afghanistan, he was released after just a few days in custody. The judge cited “exceptional circumstances” due to the extraordinary complexity of national security information (NSI) evidence and the likelihood of a trial taking years. Strict conditions were imposed: a $250,000 surety, reporting to police three times a week, use of only one monitored phone and computer, travel restrictions, and prohibitions on interfering with witnesses or evidence.¹
This stands in sharp contrast to my own situation. Charged in November 2012 with the murder of 85-year-old Liselotte Watson on Macleay Island, Queensland, I spent approximately 1,400 days (nearly four years) in high-security remand before my 2016 trial, followed by further time until the High Court quashed my conviction in 2019. My case rested on a weak chain of circumstantial evidence that the High Court later described as “extremely weak” and “glaringly improbable”.³
The Illusion of Bail for Serious Offences
Under the Bail Act 1980 (Qld), there is generally a presumption in favour of bail. However, for serious offences like murder, section 16(3) triggers a “show cause” provision. The onus reverses: the accused must demonstrate that detention is not justified and that they pose no “unacceptable risk” of failing to appear, reoffending, or interfering with the justice process. Relevant factors include the strength of the prosecution case, community ties, criminal history, sureties, and proposed conditions.²
In practice, for murder charges, bail is an illusion for most ordinary accused. Courts demand “exceptional circumstances”, and these are rarely found unless the applicant has significant resources, high-profile representation, or unusual case features. High-profile cases often succeed where others fail.
This two-tiered system wastes court time, Legal Aid funds, and taxpayer resources by dangling a right that is largely unattainable for the average person. The same can be said for the theoretical “right” to elect a judge-only trial is available on paper but rarely granted, even with strong justifications.
I am not calling for the complete removal of bail in serious cases. Pre-trial detention serves a legitimate purpose in genuine high-risk scenarios. However, the current framework fosters false hope, drains public resources, and inflicts unnecessary punishment on those presumed innocent. A more transparent, streamlined approach focused on objective risk assessment would better serve justice.
My Case: Substantial Sureties and Highly Restrictive Conditions Ignored
I was not a flight risk. I had deep roots on Macleay Island a family home, a local business, and strong community ties. Two family members were prepared to offer over $300,000 each in mortgage equity as surety, totalling more than $600,000. I willingly proposed the most restrictive conditions imaginable: GPS ankle monitoring, strict curfews, daily reporting to police on the island, and full surrender of my passport. These measures would have significantly mitigated any concerns about flight or reoffending under section 16 of the Bail Act.²
Despite this, I remained in custody for 1,400 days on remand. The primary barrier was not the law or evidence, but my Legal Aid representation’s lack of genuine intent to pursue bail.
I completed extensive legwork myself. I compiled documentation on my jurisdictional ties, prepared detailed outlines of proposed conditions, and identified suitable sureties. I provided clear instructions, including contacts for witnesses who could highlight weaknesses in the police case. My lawyer understood the law but treated the bail process as mere busywork that was busywork for me to stop me giving him instructions on the trial tactics. He never submitted a serious funding application for a full bail hearing. Later instructions to interview key witnesses were also ignored.
This was not simple oversight or lethargy it reflected a broader systemic issue. Some practitioners avoid applications that risk denial, viewing them as potential marks against their professional record.⁵
Why I Should Have Received Bail: Comparisons to Other Cases
My case was exceptionally weak, even among circumstantial prosecutions. In Fennell v The Queen [2019] HCA 37, the High Court unanimously quashed my conviction in record time and entered an acquittal. It described the Crown case on opportunity and motive as “extremely weak”, noted that I was in no different position from many other island residents, and found the key hammer identification evidence “glaringly improbable”. Paragraphs 82–85 emphasise that no reasonable jury could convict on such “palpably weak” evidence.³
Stronger circumstantial cases involving ordinary accused routinely result in bail denials or no serious applications. In R v Fennell [2017] QCA 156, the Court of Appeal recorded the absence of DNA or fingerprint evidence linking me to the crime, alongside inconsistencies in witness testimony and issues with the prosecution theory.⁴
High-profile exceptions like Roberts-Smith where complexity and resources created “exceptional circumstances” prove the bar can shift. My four-year wait in a chronically backlogged system was equally extreme, yet no equivalent leniency applied.
A competent early bail application, emphasising the incomplete and contradictory evidence, the biased witness, my strong ties, substantial surety, and strict proposed conditions, had a strong prospect of success. Instead, the absence of aggressive advocacy sealed the outcome.
The High Court’s Verdict and Irreparable Harm
The High Court’s 2019 judgment exposed the prosecution’s foundational failures: prejudicial reliance on my history, an unreliable witness, and an overall unsafe case. I ultimately spent over 2,373 days in custody before full exoneration. Those lost years time with family, business, and freedom cannot be restored.
This selective application of justice undermines public confidence. The presumption of innocence becomes hollow when bail turns on the vigour of legal representation, public profile, or connections rather than objective assessment of risk.
Time for Honest Reform
Bail for serious offences should not persist as an unattainable procedural promise. Either criteria must be applied consistently and transparently, including for those with substantial sureties and restrictive conditions or the process should be streamlined to avoid wasting resources on illusory rights. Exceptional circumstances must not be reserved for the elite.
My lived experience reveals the human toll. The scales of justice should balance on evidence and risk, not on the loudest voice or deepest pockets. Until meaningful reform occurs, they rarely do.
Sources & Footnotes
¹ Bail decision concerning Ben Roberts-Smith, Supreme Court (April 2026).
² Bail Act 1980 (Qld) s 16 — show cause provisions and unacceptable risk framework.
³ Fennell v The Queen [2019] HCA 37 — esp. [82–85].
⁴ R v Fennell [2017] QCA 156 — evidentiary record as presented on appeal.
⁵ Based on practices observed in Legal Aid Queensland proceedings.
My partner got bail for a murder charge, our lawyers said dont get your hopes up, but I believe he was a good judge and knew he was innocent, because he actually read it properly and could see there were other suspects. Etc.But yet he was convicted by a jury, and has lost his appeal, and now he has no money left ,the lawyers say he just has to do his sentence, and that we cant go to the high court to appeal unless we get new evidence.
Only trouble is he is already 72 and with a 24 year sentence he will die in prison for a crime he did not commit.
All based on circumstantial evidence.
Linda, I suggest you check with your lawyer about their advice about needing new evidence to seek leave to appeal to the High Court. I do not think that is correct. The High Court doesn’t examine new evidence but matters of law.
Since you were friends or former friends with some witnesses (people who allegedly gave you the hammer) do you think that you might have been seen as someone who could have influenced witnesses?
Bail refusals are usually based on multiple factors together, such as:
– seriousness of offence
– strength of prosecution case
– risk of reoffending
– risk of failing to appear
– risk to public safety
– risk to witnesses
Nick, you’re right about the framework bail decisions are always said to turn on a combination of factors: seriousness, strength of the case, risk of flight, risk to the public, and risk to witnesses. The issue in my case is how those factors actually played out when you follow the sequence of events.
The hammer was recovered in November 2012. At that point, there was no identified link to me. It was only after a short TV appeal in January 2013 that the witness came came forward and said they had loaned it to me. That is important, because the allegation didn’t arise organically through investigation it was prompted externally, months later.
From there, he claimed he was terrified of me and was relocated off the island at police expense. That relocation is often pointed to as evidence of “risk to a witness”. But when you look at the underlying material, the basis for that fear becomes far less stable.
Both The witness and his wife made three statements each. Those statements didn’t just evolve in minor ways , they conflicted with:
their own earlier versions
each other’s accounts
and their later evidence at committal and trial
On top of that, he made a number of additional claims in his statements that simply did not stand up when tested.
There is also a more fundamental inconsistency. He stated that after a personal falling out in 2008, I was “scum” and that he never spoke to me again not even driving down my street.
Yet the original claim was that he loaned me the hammer before 2008 while I was working on a trailer. The uncontested evidence, however, was that I didn’t even own that trailer until 2009. The timeline was then adjusted after talking with police to 2010–2011, which might superficially fix one problem, but directly contradicts his own claim that there was no contact after 2008.
So when you ask whether I might have been seen as someone capable of influencing witnesses, that’s the point — the “risk” wasn’t grounded in any demonstrated conduct. It arose from an allegation that was itself unstable, evolving, and internally inconsistent.
By the time of any serious consideration of bail, the relevant questions should have been:
Was there any evidence I had attempted to contact or interfere with witnesses? (No.)
Were the witnesses’ accounts stable and reliable? (Objectively, no.)
Were there conditions available to neutralise any perceived risk? (Yes — strict ones were offered.)
Instead, the mere existence of an allegation , regardless of its quality, was effectively treated as enough to sustain the “risk to witnesses” limb.
That’s really the broader point. On paper, the system asks courts to assess risk. In practice, especially in serious matters, it often defaults to accepting the prosecution narrative at face value, even when that narrative is later shown as it was in Fennell v The Queen — to be “extremely weak” and, in parts, “glaringly improbable”.
I understand why, looking at it abstractly, you’d think witness relationships might weigh against bail. But when you trace the actual sequence and test the reliability of what was being asserted, it becomes much harder to justify that conclusion on any objective assessment of risk.
That’s an unnecessarily long reply to my simple question!
Let me ask you another simple question:
Did they ever loan you a hammer?
NO
Steven notes: “The primary barrier was not the law or evidence, but my Legal Aid representation’s lack of genuine intent to pursue bail.” With friends like that … He is not alone in suffering disadvantage thanks to poor representation.
But his view that high profile, cashed up applicants have a better chance getting out of remand on bail is unacceptable if that’s the reality.