What happened
The appellant, DL, was 16 years old on 19 July 2005 when he attacked 15-year-old TB as she walked home from school through the car park of the Forresters Beach Resort. He stabbed her 48 times to the upper back, upper chest, face and head. One wound penetrated the heart, causing fatal blood loss. The attack ended only when a passer-by intervened. The appellant sustained a cut to his hand during the assault and gave three inconsistent explanations to witnesses that afternoon: falling on a rock, cutting it on a rose bush, or on barbed wire. He was arrested that evening, declined a police interview, and has remained in custody since. At trial before Hulme J and a jury he did not give evidence and did not raise mental illness or substantial impairment defences, though pre-trial psychiatric assessment by Dr Nielssen had identified an underlying schizophrenic illness in its prodromal phase.
Following conviction for murder on 27 March 2008, a sentence hearing was conducted. The Crown sought a finding that the offence was at the very high end of objective seriousness, premeditated, and accompanied by a specific intention to kill. The primary judge accepted the attack was frenzied but rejected premeditation and intention to kill as “much more doubtful”. His Honour found “much irrationality about what occurred”, noting the absence of any prior relationship, slight or rational motive. Psychiatric evidence was called from Drs Allnutt, Kasinathan and Nielssen. Dr Nielssen maintained the opinion that the appellant was in the prodromal phase of schizophrenia at the time of the offence, supported by family history, psychosocial decline in the preceding 6-12 months, restricted affect and subtle thought disorder. The primary judge preferred Dr Nielssen’s opinion, finding it probable the appellant was acting under the influence of some psychosis. He assessed the offence as “a little below the mid-range” of objective seriousness. At that time a standard non-parole period of 25 years applied because the victim was under 18 and the conviction occurred after 1 January 2008. On 14 November 2008 the appellant was sentenced to 22 years imprisonment with a 17-year non-parole period, back-dated to the date of arrest.
In 2016 the appellant sought leave to appeal to the Court of Criminal Appeal, relying on Muldrock v The Queen error and manifest excess. The Crown conceded the Muldrock error: the primary judge had given primary significance to the standard non-parole period contrary to the correct approach. This concession enlivened the CCA’s power under s 6(3) of the Criminal Appeal Act 1912 (NSW) to re-sentence. Affidavits were tendered by both sides on the “usual basis” described in Betts v The Queen (2016) 258 CLR 420: the appellant’s affidavit summarised case management, Juvenile Justice and Justice Health records from 2005-2015, including reports noting initial concerns of possible psychotic illness later revised to autistic spectrum traits; the Crown’s affidavits detailed disciplinary infractions and Justice Health observations. Neither party submitted the new material was relevant to the appellant’s culpability for the 2005 offence; it was directed only to rehabilitation prospects and conditions of custody if re-sentencing occurred.
Before the CCA neither party challenged the primary judge’s factual findings. The Crown submitted any adjustment to the sentence should be minimal, describing some of the primary judge’s findings as “unduly favourable” and the failure to find intention to kill as “generous”, yet expressly stated it did not take issue with the assessment of criminality. The CCA majority (Leeming JA and Wilson J, Rothman J dissenting) nevertheless rejected the primary judge’s acceptance of Dr Nielssen’s opinion, finding that the passage of time without development of schizophrenia showed the opinion had “not been borne out”. They found the appellant had intended to kill TB, that there was “some degree of premeditation”, and that the primary judge’s finding of unlikelihood of re-offending should be rejected. On that basis they concluded no lesser sentence was warranted in law and dismissed the appeal. Rothman J would have respected the unchallenged findings and re-sentenced to a 12-year non-parole period.
Special leave was granted on 15 December 2017. The High Court (Bell, Keane, Nettle, Gordon and Edelman JJ) allowed the appeal on the ground that the appellant had been denied procedural fairness. The Court held that the CCA majority’s decision to depart from the unchallenged findings and to use evidence tendered on the usual basis to substitute aggravated findings, without notice to the appellant, occasioned a miscarriage of justice. The matter was remitted to the CCA for re-determination of the sentence appeal. The second ground, which sought a broader statement about the power of appellate courts to substitute aggravated factual findings, was not decided.
Why the court decided this way
The High Court’s reasoning rests on two linked misconceptions it identified in the CCA’s approach. First, the CCA treated the parties’ stance that the primary judge’s “assessment of criminality” was not in issue as inconsistent with their invitation to put the finding of objective seriousness entirely to one side. The High Court explained there was no inconsistency. When the primary judge sentenced in 2008 a 25-year standard non-parole period applied and the judge was required to locate the offence on a continuum of objective seriousness. By the time of the CCA hearing that statutory requirement had been removed for offenders under 18 at the time of the offence (s 54D(3) of the Sentencing Act, inserted 1 January 2009). The parties were therefore addressing how the CCA should assess objective seriousness under the new statutory regime while still respecting the primary factual findings. The prosecutor’s concession was a shorthand acknowledgment that the Crown did not seek to disturb the primary judge’s findings that the appellant was probably acting under psychosis, that intention to kill had not been proved beyond reasonable doubt, and that premeditation had not been proved. Those were findings that directly informed the appellant’s criminality. The Crown’s further descriptions of certain findings as “generous” or “unduly favourable” were advanced only in support of its submission that any reduction should be minimal, not as an invitation to the CCA to re-open and reverse those findings.
Second, the CCA misunderstood the basis on which the fresh evidence had been admitted. All affidavits were tendered on the “usual basis” explained in Betts v The Queen: post-sentence evidence of rehabilitation is received so that, if the appellate court reaches the stage of re-sentencing, it can take account of the offender’s progress in custody. The High Court emphasised that Betts makes clear that only in exceptional cases may such evidence be used to revisit primary factual findings, because the interests of justice are not served by allowing a party to run a new or different case on appeal. Here neither party invited the CCA to do so. The CCA majority nevertheless treated the new material (particularly the absence of a diagnosed schizophrenic illness after 11 years) as directly bearing on the objective seriousness of the 2005 offence and as justifying rejection of Dr Nielssen’s opinion. The High Court noted that the new evidence contained no fresh expert opinion about the appellant’s mental state at the time of the offence. Dr Nielssen himself had acknowledged in 2008 that the illness had not progressed as expected after three years, yet maintained his diagnosis on the basis of family history, psychosocial decline commencing only 6-12 months before the offence, restricted affect and subtle thought disorder. The inference that the passage of further time disproved the opinion was not uncontroversial; had the appellant been on notice that the finding was under challenge he could have sought a further report from Dr Nielssen.
The High Court further held that Leeming JA’s alternative reason for rejecting the psychosis finding—that a temporary psychosis could not logically be reconciled with a verdict of murder based on intention to inflict grievous bodily harm while falling short of intention to kill—failed to respect the onus of proof. Murder having been left to the jury on alternative bases, the primary judge was required to sentence on the basis that the appellant had at least the intention to inflict grievous bodily harm. He was not required to find an intention to kill unless satisfied of that beyond reasonable doubt. Acceptance of Dr Nielssen’s evidence and the irrationality of the attack meant the judge was not so satisfied. The CCA’s approach therefore reversed the onus on a circumstance of aggravation.
Because the CCA did not put the appellant on notice of its intention not to act on the prosecution concession, the appellant was entitled to conduct his case on the understanding that the concession would be accepted. The failure to give that notice was held to be a miscarriage of justice because it could not be said the error could not have made any difference: further evidence or submissions might have altered the outcome. The first ground was made good. The Court expressly declined to address the second ground concerning the circumstances in which an appellate court may substitute aggravated factual findings for unchallenged findings of a sentencing judge.
Before and after state of the law
Before DL v The Queen the law governing sentence appeals after identification of error was settled in several respects. Kentwell v The Queen established that once error is found the CCA does not merely correct the error but exercises an independent sentencing discretion, forming its own view of the appropriate sentence. That discretion is ordinarily exercised on the material before the sentencing judge, the judge’s unchallenged factual findings, and any relevant post-sentence conduct. Muldrock v The Queen had clarified that standard non-parole periods are not to be given primary significance; they are merely one factor. Betts v The Queen had confirmed the “usual basis” limitation on post-sentence evidence: it is admitted for use on re-sentencing if the court reaches that stage, but does not ordinarily permit a party to re-litigate primary facts.
The statutory landscape had also changed between sentence and appeal. At sentence in 2008 the 25-year standard non-parole period for the murder of a child under 18 applied. By the time of the CCA hearing s 54D(3) had removed standard non-parole periods for offenders under 18 at the time of the offence. This altered the framework within which objective seriousness was to be assessed.
After DL v The Queen the law is clearer on the intersection of prosecutorial concessions, procedural fairness and the use of evidence tendered on the usual basis. An appellate court is not bound by a prosecutor’s concession as to factual findings; the public interest in correct sentencing permits the court to form its own view. However, if the court is minded not to act on such a concession it must put the offender on notice and afford an opportunity to meet the new position by evidence or submissions. Failure to do so will ordinarily constitute a miscarriage of justice. The decision reinforces that the “usual basis” rule in Betts is not a mere formality; parties and courts must adhere to the limited purpose for which the evidence is received unless both sides are given a clear opportunity to address any expanded use. The judgment also underlines that the onus of proof for aggravating circumstances remains on the Crown on sentence, and that psychiatric opinions accepted by a sentencing judge cannot be dismissed on appeal by simple inference from the subsequent passage of time without giving the offender the chance to respond with updated expert evidence. The second ground left undecided means the precise limits on an appellate court’s power to make more serious factual findings than those made below, in the absence of any challenge, remain open.
Key passages with plain-English translation
One central passage appears in the Court’s analysis of the prosecutor’s concession: “Where, however, the judge (or the appellate court in the case of re-sentencing) is minded not to act on a concession made by the prosecution, the failure to put the offender on notice of that inclination and give him or her an opportunity to deal with the matter by evidence or submissions will ordinarily be a miscarriage of justice.” In plain English this means that even though a judge or appeal court can reject a deal the prosecutor offers about the facts, they cannot spring that rejection on the defence without warning. The defence is entitled to assume the prosecutor’s position will be accepted unless told otherwise. Without that warning the process is unfair and the outcome cannot stand.
Another key statement addresses the limited use of new evidence: “Betts allows that in an exceptional case new evidence may be received for the purpose of revisiting the findings of primary fact. For the reasons there explained, the interests of justice will generally not be served by permitting either party to make a new or different case on the hearing of the appeal. Here neither party invited the Court of Criminal Appeal to re-sentence the appellant upon a factual basis that differed from the primary judge’s findings, much less to use the new evidence to impugn those findings.” Translated, this says post-sentence prison records are normally only for deciding how the person has changed since sentencing. They cannot be used to re-write history about what happened on the day of the crime unless both sides are told that is on the table and given a chance to call more evidence. In this case no one asked the CCA to do that, so it was wrong to use the records to overturn the psychosis finding.
On the onus of proof the Court observed: “The observation fails to take account of the onus in the case of circumstances of aggravation. Murder was left to the jury on the basis that the appellant possessed the intention either to kill or to do grievous bodily harm. The primary judge was required to sentence upon the basis that, at the time the appellant stabbed TB, he possessed at least the intention to do grievous bodily harm. His Honour was not required to sentence the appellant on the basis that he intended to kill TB unless he was satisfied that that intention had been proved beyond reasonable doubt.” In everyday language, once the jury convicted of murder the judge only had to decide the appellant at least meant to cause really serious injury. The judge did not have to decide he meant to kill unless the Crown proved that higher mental state beyond reasonable doubt. The CCA majority reversed that onus when it said the number of stab wounds “pointed inexorably” to intention to kill.
Finally the Court concluded: “The majority’s decision to depart from the primary judge’s unchallenged factual findings, and to take the new evidence into account in substituting a finding of aggravation – the intention to kill (and in Wilson J’s case the finding of premeditation and the rejection of the finding of unlikelihood of re-offending) – without notice to the appellant, was procedurally unfair and has occasioned a miscarriage of justice.” Plain English: the appeal judges changed the facts to make the crime look worse, used prison records they were not supposed to use for that purpose, and never warned the appellant they were thinking of doing so. That unfairness meant the whole appeal process had to start again.
What fact patterns trigger this precedent
This precedent is triggered in sentence appeals where three elements coincide. First, the sentencing judge has made explicit factual findings favourable to the offender on matters such as mental state at the time of the offence, absence of proven intention to kill, absence of premeditation, or prospects of rehabilitation. Second, on appeal the prosecution expressly or by clear implication concedes that it does not take issue with those findings, even while arguing that the ultimate sentence should not be reduced or should be reduced only modestly. Third, fresh evidence is tendered by consent on the “usual basis” explained in Betts—typically affidavits summarising custodial progress, disciplinary records, psychological reports or medical notes accumulated over years in prison—and neither party suggests that evidence is relevant to re-litigating the original facts.
The precedent applies with particular force where the statutory sentencing regime has changed between sentence and appeal (for example, removal of a standard non-parole period for juvenile offenders) so that the appeal court must assess objective seriousness afresh, yet the parties have conducted the appeal on the shared understanding that the primary judge’s underlying factual findings remain the platform for that assessment. It is engaged whenever the appellate court forms the view that the new evidence casts doubt on the original psychiatric or factual conclusions and is minded to make more serious findings (intention to kill rather than grievous bodily harm, premeditation, higher objective seriousness, or greater risk of re-offending) without first alerting the offender that the concession is not being accepted and that the factual platform is no longer secure. The principle extends to any situation in which the appellate court proposes to depart from the primary judge’s acceptance of a particular expert opinion (here Dr Nielssen’s prodromal schizophrenia diagnosis) on the basis of the mere passage of time and the absence of subsequent frank psychotic symptoms, without affording the offender the opportunity to obtain an updated report from the original expert or to lead further evidence explaining why the original diagnosis remains valid.
How later courts have treated it
The judgment itself carefully grounds its conclusions in the earlier authorities it cites. It treats Betts v The Queen as settling the limited purpose for which post-sentence evidence is received and applies that limitation strictly. It cites Kentwell v The Queen for the proposition that the CCA exercises an independent sentencing discretion once error is found, but insists that the exercise of that discretion remains subject to procedural fairness constraints. The Muldrock concession is accepted without re-litigation, illustrating that once a Muldrock error is conceded the CCA must proceed to re-sentence according to law as it stands at the date of the appeal, including the removal of standard non-parole periods for juveniles.
The Court’s treatment of the psychiatric evidence shows how later courts should approach contested expert opinion on sentence. The primary judge’s preference for Dr Nielssen’s evidence was treated as open on the material before him; the CCA was not entitled to substitute its own view based on subsequent non-development of schizophrenia without notice. The judgment emphasises that Dr Nielssen had already accounted for the absence of acute symptoms three years post-offence and had explained the variable course of schizophrenic illness. Later courts are therefore directed to treat accepted psychiatric explanations for apparently motiveless or irrational violence with caution before discounting them on the basis of subsequent custodial history alone.
The decision also illustrates the continuing vitality of the principle that the Crown bears the onus of proving aggravating circumstances beyond reasonable doubt on sentence. The CCA majority’s inability to “conceive of a temporary psychosis which left the appellant with an intention to inflict grievous bodily harm while falling short of an intention to kill” was held to overlook that onus. Later courts must therefore sentence on the basis of the least serious mental state consistent with the verdict unless the higher state is proved.
Still-open questions
The High Court expressly left undecided the appellant’s second ground, which invited the Court to state a principle respecting the power of an appellate court, on an appeal against sentence under the common form provision, to substitute aggravated factual findings for the unchallenged findings of the sentencing judge. That question therefore remains open. It is unclear whether there are any circumstances, short of the parties expressly inviting the appellate court to re-open primary facts, in which the CCA may lawfully make more serious findings than those made below. The judgment leaves for future consideration the precise limits on that power and the standard of appellate review that should apply.
Another open question concerns the interaction between the “usual basis” rule and exceptional cases in which new evidence might legitimately be used to revisit primary findings. The Court accepted that Betts contemplates exceptional cases but gave no further guidance on what would constitute such an exception in the context of unchallenged psychiatric findings. Future cases will need to explore whether fresh expert evidence that directly contradicts the original diagnosis, or undisputed objective facts that are irreconcilable with the original findings, could ever justify departure without notice.
The judgment also leaves unresolved how a sentencing judge should express the effect of a finding of “some psychosis” on the offender’s moral culpability when the defence of mental illness was not raised at trial. The primary judge did not elaborate on whether the psychosis substantially impaired the appellant’s capacity to understand events, judge right from wrong or control himself. The High Court noted that, because the prosecution did not object to the tender of Dr Nielssen’s report, the judge should be taken to have found that any impairment did not reach the threshold for the statutory defence. The exact weight to be given to such a finding on sentence in future cases remains a matter for development.
Finally, the decision does not address the position where the prosecution on appeal withdraws or qualifies its concession after the hearing has begun. The Court proceeded on the basis that the concession was maintained throughout. Whether a late change in the Crown’s position would itself require an adjournment and fresh opportunity to call evidence is a practical question left for another day. These open questions mean that while the procedural fairness obligation is now clear, the substantive limits on appellate re-finding of facts in sentence appeals continue to require careful case-by-case analysis.