Australia’s law students are currently sitting their first semester exams. Some of them might be examining hypotheticals like this one:

D shot and wounded the V, who was a police officer, while V was lawfully executing a search warrant in company with other police officers on premises in close proximity to D’s home. The shot struck V in the arm, thereby causing him a serious but non-fatal gunshot wound. In the course of the fire-fight which ensued, one of the other police officers fired a shot which was intended for D, but which unfortunately instead hit V in the neck, thereby inflicting a wound from which he later died. Assume that when D fired at V, D honestly believed that V was someone posing as a police officer who was intent on robbing the D and might have posed a serious risk to the D’s safety.

That exam question could earn the examiner congratulations for her inventiveness, but the next one would probably earn her a meeting with her Dean:

D moved to stab V in the chest but she asked him not to stab her there and he rolled her over and stabbed her a number of times in the back. V believed her only chance to escape was to calm D and weaken him. She said, “If we are going to do this together, then I should have a turn with the knife.” The tip of the knife D then had had broken and was embedded in V’s back so D grabbed another knife from the kitchen, handed it to V and said “Okay” and lay on his back. V stabbed D in the stomach, giving an extra shove to make sure the knife was in deep and she twisted it. It appeared that some of D’s intestines came out, D remarking “That was a good one.”

Adding further details – that D was a former contest on Australian Survivor and that, during D’s attack, a real estate agent entered the flat to open it for inspection, saw ‘red liquid’ everywhere, and left, seemingly without calling the police – may well leave the lecturer without a job. But, as Mark Twain is reputed to have said, ‘the difference between fact and fiction is that fiction must be believable.’ Or, as the late Han Solo said, ‘it’s true, all of it’.

The victims of crime in the above two cases were Bill Crews, a 26 year-old police constable whose parents recounted to the coroner their surprise and pride when their son told them three years earlier that he had got in to the police force and then sat and watched video of his last moments, and Samantha Holland, aged 23 at the time, who was stabbed at least 26 times before she escaped over a balcony and will suffer physical and psychological scars for life, and who sobbed in court as her ex-boyfriend was sentenced. The defendants were Phillip Nguyen, then 55, whose first wife was murdered a decade earlier and whose second marriage failed while he was on remand for killing Crews, and Joel Betts, then 30, a victim of childhood sexual abuse and violence who faces a lifetime of incontinence due to the bowel injury he incurred while stabbing Holland. Each recently lost their final appeals in the High Court. Because both defendants pleaded guilty, the Court’s task was not the criminal law student’s task of applying the rules of criminal responsibility to these strange facts. Rather, the nation’s top judges faced an infinitely harder task: assessing whether the punishment each offender was given fitted their unusual crimes.

The High Court’s role in sentencing law

Despite the enormous public interest in sentencing, Australia’s apex court has generally taken a hands-off approach to sentencing law. Although the High Court is always happy to rule on the meaning of an ambiguous sentencing statute, it stayed out of the nitty gritty of sentencing itself until 1979, when it overruled a life sentence given to a man who stabbed another to death. The Court held that Bobby Veen’s punishment should fit the crime he was convicted of, which was manslaughter rather than murder. That experience proved to be a bruising one for the Court, which was later accused of having ‘blood on its hands’ when Veen killed again after he was released. (For a contemporary account of these events, see here.) Nguyen’s and Betts’s appeals are safer ground for the Court, as the issues in these cases are about particular sentencing rules.

Nguyen v The Queen [2016] HCA 17, like Veen’s case, concerns sentencing for manslaughter, the complex common law crime that captures people who are guilty of homicide but fall outside of the definition of murder. The particular issue in Nguyen’s case was whether and how the sentencing judge could take account of the fact that Nguyen’s victim was a police officer, an aggravating factor in New South Wales. Justice Fullerton ruled that Nguyen’s offences (including the initial wounding) were ‘aggravated to a significant degree’ by Crews’s status as a police officer exercising his duty, but added:

I am not persuaded, however, that either offence is in the worst category. It would have been otherwise were the offender to have shot at Constable Crews intending to inflict grievous bodily harm knowing or believing he was a police officer…

Although Nguyen’s lack of knowledge that Crews was a police officer was part of the agreed facts settled between the prosecution and the defence, the Court of Criminal Appeal held that Fullerton J was wrong to rely on it, reasoning that, if Nguyen had known Crews was a police officer then he wouldn’t have had any reason to defend himself at all and would have been guilty of murder. The Court of Criminal Appeal accepted the prosecution’s argument that taking account of such a fact is precluded by R v De Simoni, a 1981 High Court ruling that bars sentencing judges from relying on facts that would amount to a more serious crime than the one the offender had been convicted of. But the current High Court ruled that that was an incorrect application of its earlier ruling:

The appellant is correct in submitting that the De Simoni principle operates for the benefit of the offender and does not apply to preclude a sentencing court from taking into account the absence of a factor which, if present, may have rendered the offender guilty of a more serious offence. This is because the De Simoni principle is an aspect of the fundamental principle that no one should be punished for an offence of which the person has not been convicted.

In short, a sentencing judge cannot say to an offender ‘you did more serious crimes than the verdict shows’, but can say ‘you did no more serious crimes than the verdict shows’. Australian sentencing law is largely built on nuanced (some would say meaningless) distinctions like this one.

The question at the heart of Betts v R [2016] HCA 25 is much more important in practice: if an appeal court overturns an offender’s sentence, what facts must it consider when it resentences the offender? This issue arose because Toner J, who originally sentenced Betts, classified Samantha Holland as ‘vulnerable’ and failed to factor in Betts’s debilitating bowel condition when setting his overall sentence. The Court of Criminal Appeal’s ruling that these findings were wrong – both interesting matters that weren’t challenged in the High Court – meant that the appeal court had to resentence Betts for his crimes. But, in the two-and-a-half years between Toner J’s sentence and the appeal hearing, Betts’s new lawyer had obtained new evidence about Betts. In the resentencing, the Court of Criminal Appeal took account of some of that new evidence – the fact that Betts had since completed some coursework and a psychotherapist’s view that Betts’s prospects had ‘improved’ – but refused to consider a new report from a psychiatrist about what caused Betts to attack his ex-girlfriend in the first place. The appeal court stated that the re-sentencing was not ‘an opportunity for a second bite’ on matters already argued at the original sentencing hearing. At the High Court appeal, Betts argued that offenders who were sentenced in error should be free to raise any evidence they want at their resentencing, just as defendants who are wrongly convicted are free to raise any evidence they want at their new trial. However, the High Court rejected this argument:

The analogy is hardly apt. When the Court of Criminal Appeal quashes a conviction and orders a new trial, the successful appellant is restored to the status of an unconvicted person to whom the presumption of innocence applies. The fact that the accused may choose to adduce evidence at the new trial, on which he or she did not rely at the first trial, does not undermine adversarial criminal justice.

By contrast:

Justice does not miscarry by reason of the refusal to allow an appellant to run a new and different case on the question of re-sentence. Exceptional cases apart, the question of whether some other sentence is warranted in law is answered by consideration of the material that was before the sentencing court and any relevant evidence of post-sentence conduct.

This is another instance of the way Australian sentencing law emphasises form over substance. While institutional arrangements in Australia do indeed sharply distinguish between appeals against conviction and appeals against sentence, Sir James Stephen was surely right to observe over 150 years ago that ‘the sentence is the gist of the proceeding. It is to the trial what the bullet is to the powder’ (or, as the High Court itself put it in 1998: ‘To an offender, the only relevant question may be “how long”‘.) So, it is jarring to see the offender’s argument for similar rights at his new resentencing hearing rejected out of hand by Australia’s top court via a mere declaration, without citations or analysis, that a cherished principle like the presumption of innocence has no role to play in sentencing appeals.

The High Court’s role in sentencing facts

The two bland answers the national Court supplied to these two interesting legal questions are not the only point of interest in Nguyen v R and Betts v R. Rather, the cases’ further importance lies in remarks the apex court made when it came to considering how to determine sentences for the complex facts presented by these two offenders.

Because it found that the NSW Court of Criminal Appeal has misapplied the De Simoni principle (and also the rules governing whether sentences should served simultaneously or consecutively) when it quashed Nguyen’s original sentence of 9 years in prison, the High Court itself now had to decide what sentence should be imposed for the wounding and death of Crews. The Court’s unanimous decision was to give Betts the exact same sentence that the NSW Court of Criminal Appeal had reached by error – 17 years imprisonment, or almost twice the sentence Fullerton J had given. The Court held that, although Fullerton J was not barred by any principle from taking account of Nguyen’s mistaken belief that Crews was a criminal, she was still wrong to do so in practice. According to Gageler, Nettle & Gordon JJ:

It is irrelevant in assessing the objective gravity of an offence of manslaughter to contrast it with what would be an offence of murder. It is erroneous because it is likely to result in an assessment of the relative gravity of the subject offence which ill-accords with its objective gravity relative to other instances of offences of that kind.

In short, comparisons between what the offender did and other offences he or she might have done are irrelevant and distracting. This ruling (like another one in 2013 banning comparisons between the sentence for the charge laid by the prosecutor and other charges the prosecutor could have laid) runs counter to popular conceptions of sentencing, which are all about such comparisons, say between sentences for theft and insider trading, or for assault and sexual assault, or for murder and manslaughter. In this case, the Court’s approach meant that the very comparison the Court said wasn’t barred by the presumption of innocence was nevertheless almost always forbidden:

The appellant’s perception, that the deceased was a robber and not that he was a police officer, was not material to the assessment of the objective seriousness of the manslaughter. The gravamen of that offence was the taking of a life in self-defence where the act was not a reasonable response to the circumstances as the appellant perceived them.

So, Nguyen was to be sentenced merely as a person who unreasonably defended himself in a way that killed someone (in particular, someone who happened to be a police officer), without regard to who he thought he was defending himself from. This approach sits poorly with the Court’s own famous ruling in 1987 that the punishment of criminals in Australia ‘depends not on the statute defining the offence, but on the facts of the case’. I doubt that many lay people would treat the fact that an offender thought he was defending himself against a criminal, rather than a police officer, as irrelevant to the seriousness of a violent crime.

In Betts v R, the High Court’s sentencing role was narrower, because it didn’t have to resentence Betts. Instead, its task was to assess whether the NSW Court of Criminal Appeal should have chosen to take account of Bett’s new psychiatric evidence when it resentenced Betts. The High Court unanimously held that a fair resentencing of Betts could take place without considering new evidence about why he committed the crime, because of the way Betts had initially sought to be sentenced after his original trial. According to the Court:

In the course of her submissions before Judge Toner, senior counsel for the appellant accepted that the evidence was “effectively silent” on the question of why the appellant acted as he did. Senior counsel invited his Honour to consider, as a possible explanation, that the appellant had been exposed to very serious domestic violence in his youth. Senior counsel also put on the appellant’s behalf that “it can take people a long time coming to terms with the violence that they have committed on people … that they have loved” and that the appellant had reached a position of acceptance of responsibility.

By contrast, the psychological evidence Betts tendered at his appeal concluded:

From the history provided by [the appellant] and the information in the documents provided, I believe his intoxication with a drug with unpredictable mind altering effects, together with an underlying emotional state shaped by violence and sexual abuse, and a pattern of substance use, was a significant contributing factor to his sudden decision to end his life and to his offending behaviour.

In short:

The appellant’s case before Judge Toner was not that his ingestion of DMT had significantly contributed to his offending. The forensic choice that was made was to accept responsibility for the offences. Nothing in the new evidence supports the submission that the Court of Criminal Appeal’s refusal to permit the appellant to run a different case before it has occasioned a miscarriage of justice.

The Court’s unanimous conclusion would be uncontentious if these characterisations were correct. But sentencing pleas are complex and are raised against a background of submitted evidence. The evidence Betts put before Toner J included a different psychiatrist’s report (barely mentioned by the NSW Court of Criminal Appeal), which concluded:

[I]t is likely he was adversely affected by the drug DMT at the relevant time including perceptual disturbances and an altered perception of his environment. Despite that however, I am again not able to indicate that he was suffering from a frank drug induced psychosis which might have relevance to the offence itself. … Mr Betts certainly describes altered experiences arising from his use of the substance DMT, but again I cannot confirm that he developed a psychotic illness as a result of the use of the drug …

When Betts pointed this out in the High Court (and argued that the new report amplified, rather than contradicted, his earlier submission), the Court responded:

This is disingenuous. Based upon largely the same material, save for any difference in the history supplied by the appellant, Dr Westmore and Dr Nielssen came to different conclusions with respect to the causal relation of the drug use to the offences. Indeed, Dr Nielssen’s opinion that the appellant was in a psychotic state, or its equivalent, would appear to traverse the appellant’s pleas.

Maybe the High Court is correct on this, maybe not. But what is significant is that an apex Court was willing to make these very fine judgments about arguments, procedures and evidence itself, rather than sending them to be dealt with by the courts that are all too familiar with these sorts of arguments. In the High Court, Betts’s counsel requested that the question of whether or not to take heed of the new evidence simply be sent back to the District Court for reconsideration. However, the Court, after expressing doubts about whether NSW criminal appeal courts have the power to ‘remit’ matters for resentencing in this way (and urging the NSW parliament to consider that issue), simply opted to decide the matter for itself. This is an unsatisfactory way to do justice in a complex and odd case that will leave a man with severe health issues, both physical and mental, in prison for a minimum of 11 years.

The French Court’s role in sentencing

While Australian law students sit their exams, much of the internet has been discussing a recent sentencing decision in the United States, where Stanford student Brock Turner was given a six month sentence by Judge Aaron Persky for raping an unconscious 22 year-old following a frat party. The discussion is refreshing in the near unanimity of shock at the seriousness of the offence and close attention to reports of the details of the statements (notably those of the victim) made at the sentencing hearing. It is a pity, in my view, that most Australian law students (who will be Australia’s future High Court judges) are tested using contrived and impersonal hypothetical fact scenarios, rather than the complex details of real cases. Imagine if law students were asked to read the ‘sentencing packet‘ that went before Persky J and explain in detail what sentence Brock should have got, or to review his stated reasons for potential errors (both recently released documents that raise many interesting issues that the internet debate is yet to reach.) Or were asked to comprehensively address the correct sentence for offenders like Nguyen or Betts?

In public discussion of sentencing, much has been made by legal stakeholders and academics of the importance of judges being able to consider all the facts of an offence and of how public frustration at sentencing can abate when the full facts are known. Unfortunately, a trend in recent High Court sentencing judgments threatens to increase the disconnect between Australian sentences and the facts of offending and offenders.  Since 2010, the High Court has held that sentencing judges must:

• disregard any ‘norm’ in sentencing practice about the period a federal prisoner should serve in prison before eligible for early release
• disregard a statutory ‘midrange’ for an offence when sentencing a ‘mentally retarded’ offender
• have regard (in Crown appeals) to a discrepancy between co-offender’s sentences
• disregard (in Crown appeals) that an offender was initially and erroneously given a lower sentence
• disregard the maximum sentences of any other offence the offender could have been charged with (but wasn’t)
• disregard the mere aborignality of an offender (as opposed to the specific disadvantages suffered by that offender)
• disregard the mere disadvantaged background of an offender. (The Court also questioned whether sentencing judges should take account of the prospect of traditional punishment.)
• disregard any submissions the prosecution may make about sentencing range (and, indeed, ban the prosecutor from making such submissions)
• disregard (when revisiting erroneous sentences) the fact that an earlier sentence was reached by a reasoning process that was contrary to law (although such a finding mandates a resentencing in a regular sentencing appeal).
• have regard (in a Crown appeal) to the prosecutor’s failure to object to a sentence at the original hearing
• disregard significantly mitigating facts that are not proven or disproven
• disregard comparisons with more serious crimes that the offender didn’t commit
• disregard (at resentencing after an erroneous sentence) new evidence that contradicts the offender’s case at the original sentencing hearing

Of course, there are quite a variety of different decisions here and some had much more nuance than others. But the overall trend is clear: a steady reduction in the matters that Australian sentencing judges are permitted or required to consider when determining a sentence.

This trend is especially surprising in light of the rise throughout Australia of sentencing statutes that emphasise the complex and multi-faceted nature of sentencing, for example this NSW provision, which went undiscussed in Nguyen and Betts:

The purposes for which a court may impose a sentence on an offender are as follows:

(a) to ensure that the offender is adequately punished for the offence,
(b) to prevent crime by deterring the offender and other persons from committing similar offences,
(c) to protect the community from the offender,
(d) to promote the rehabilitation of the offender,
(e) to make the offender accountable for his or her actions,
(f) to denounce the conduct of the offender,
(g) to recognise the harm done to the victim of the crime and the community.

Perhaps Phillip Nguyen’s harsh sentence serves the above goals despite his mistaken belief that he was fighting off robbers, but it is hard to see how refusing to revisit Joel Betts’s sentence in light of new psychiatric evidence serves any of them.

Some of this trend pre-dated the French Court (for example, the Gleeson Court’s 2008 holding that sentences for federal drug offences should disregard evidence of the harm caused by particular drugs.) But the trend this decade is nevertheless contrasts with the approach of earlier High Courts. An example is the Court’s 2001 judgment on the sentencing of child sex offenders, where McHugh J, Kirby J and Callinan J held that the offender’s otherwise good character cannot be wholly disregarded even in such cases (a ruling subsequently reversed by the NSW parliament) and members of the Court debated at length whether and how to factor in paedophila into the sentencing discretion, yielding five separate judgments  (including short dissents from Gummow J and Hayne J, who would later dominate the French Court.) The 2001 decision (and many earlier ones) encouraged Australian sentencing judges to identify and attempt to measure the significance of all sentence facts. By contrast, the more recent trend could encourage trial and appeal judges to pare back their discussions of sentencing facts, something I view as a loss for judicial independence, public understanding of the courts and, most importantly,  justice for the many individuals affected by the criminal justice system.