Sentencing and mental disorder: the evolution of the Verdins Principles, strategic interdisciplinary advocacy and evidence-based reform

Abstract

In DPP v O’Neill, the Victorian Court of Appeal excluded personality disorders from the scope of the sentencing principles that apply to offenders with mental health problems around Australia (the ‘Verdins principles’). This decision was based on a fundamental misunderstanding of the nature of personality disorders and had the potential to create serious injustice for many marginalised people. To redress this problem, the authors engaged in a unique process of strategic advocacy, which resulted in the Victorian Court of Appeal overturning O’Neill in the recent case of Brown v The Queen. This article examines the evolution of the Verdins principles, the problems that arose in O’Neill, the collaborative strategy used to address those problems and the successful outcome of that interdisciplinary strategy.

Introduction

Australian courts have frequently noted the difficulties involved in sentencing offenders with mental health problems.¹ These difficulties arise due to the multiple ways in which offenders’ mental health can affect sentencing determinations² and because the relevant sentencing principles often pull in different directions. For example, mental health problems ‘may reduce the moral culpability of the offender and the deliberation which attended his criminal conduct’,³ pointing towards a more lenient sentence, while at the same time they ‘may mark him as a more intractable subject for reform than one who is not so affected, or even as one who is so likely to offend again that he should be removed from society for a lengthy or indeterminate period’.⁴

Courts have received little assistance about the best way to resolve these tensions. Legislatures have largely left the issue untouched, and relevant academic literature is sparse. This has resulted in the ad hoc development of the law. While there are many virtues to this common-law process of legal development,⁵ in a highly interdisciplinary area involving the intersection of law, psychiatry and psychology, it creates a risk that the law will develop without due regard to contemporary non-legal knowledge. Unfortunately, this risk eventuated in Victorian jurisprudence in 2015.

In the case of DPP (Vic) v O’Neill,⁶ the Victorian Court of Appeal excluded personality disorders from the scope of the principles that govern the sentencing of offenders with mental health problems (the ‘Verdins principles’).⁷ This decision was based on a fundamental misunderstanding of the nature of personality disorders.⁸ As this had the potential to create serious injustice for many marginalised people across the country,⁹ strategic efforts were made to reform the law in the area. This recently culminated in the O’Neill decision being overturned by a specially-convened five-member bench of the Victorian Court of Appeal in the case of Brown v The Queen.¹⁰

This article examines the development of the law in this area, the problems that arose, the strategy that was used to address those problems and the outcome of that strategy. It is divided into three parts. Part 1 looks at the previously unexplored history of the Verdins principles. In doing so, it explicates the nature of those principles and shows how things went wrong. Part 2 outlines the process that was adopted for seeking reform of the law in this area: a unique, interdisciplinary collaboration between a legal academic, a forensic psychiatrist and a legal practitioner that could provide a blueprint for future law reform efforts. Part 3 analyses the decision of Brown v The Queen,¹¹ an exemplar of evidence-based judicial decision-making.

Part 1: the development of the Verdins principles

While the history of the insanity defence is a long one,¹² it was not until 1958 that Australian courts first considered the ways in which mental health problems may affect the sentencing determination. Since that time, there have been six phases in the development of the law.

Phase 1 (1958–1979): lenience, danger and culpability

Compassion grounded the first cases in which the court considered the relevance of mental health problems to sentencing. In R v Smith, the offender was in a ‘confusional state of mind’ due to ‘chronic brain damage’ when he committed assault.¹³ While this did not qualify for the insanity defence, the NSW Court of Criminal Appeal held that it could ‘look at his mental condition and the evidence as to the brain injury when assessing the penalty to be imposed’.¹⁴ It considered his brain injury, and consequent lack of control over his behaviour, to be an ‘extenuating circumstance’ and significantly reduced his sentence.¹⁵ A similar conclusion was reached in R v Evans, in which the Court held that the fact that the offender had ‘lost his power of reasoning with composure’ warranted lenience.¹⁶

It was not long, however, before courts noted that mental health problems may also push in the other direction, due to the dangers the offender may pose to the public. This issue had been raised in several English cases in the 1950s, in which the courts accepted that it was appropriate to impose a sentence of preventive detention on a ‘mentally-disturbed offender who was a potential danger to society and whose chances of recovery, within a given period, could not be estimated’.¹⁷ In such cases, the court could ‘cast aside’ the notions of ‘just retribution’ and proportionality that ordinarily govern sentencing, and instead sentence on the basis of ‘social defence’.¹⁸

This approach was not accepted by Australian courts, which were concerned to ensure that sentences always remain proportionate to the offence committed, even when the offender had a severe mental health problem that made them a significant danger to the public. For example, in R v Gascoigne, the offender was ‘psychopathic with periods of depression and withdrawal, suggestive of a schizoid personality’.¹⁹ He was addicted to alcohol, which would ‘precipitate frank schizophrenic delusions resulting in irresponsible behaviour such as … homicidal attacks’.²⁰ It was considered unlikely that his condition would change in the foreseeable future, and the medical opinions of those who examined him were that he should be kept under control and surveillance for an indefinite period. Despite this, the Court held that the trial judge had been wrong to impose a sentence of life imprisonment, as it was disproportionate to the offence committed (an unlawful wounding). Justice Hanger stated:

[T]he sentence imposed is to be in respect of and related to the offence committed; it is a punishment for that offence; only within this field is criminal punishment to be used for the purpose of protecting the community. It cannot be used for this purpose because, by reason of mental illness, the prisoner is a menace when at large. Such protection must be left to other appropriate authorities.²¹

The court was quick to clarify (just two months later) that this did not mean that community protection was unimportant in the sentencing process: ‘protection of the community is one of the most important results that the criminal law is designed to secure’.²² Consequently, where a sentence of life imprisonment is proportionate to the offence committed, the court may take into account the fact that ‘the mental condition of the convicted person would make him a danger if he were at large’ in deciding whether or not to impose that sentence.²³ However, the court may not impose a disproportionately long sentence for the protection of the public.

The interaction between these conflicting principles – compassion and community protection – was first addressed in R v Kocan, a case in which a man with a ‘significant and severe personality disturbance of the inadequate and schizoid type’ attempted to assassinate the prime minister, Arthur Calwell.²⁴ In sentencing him, the Court noted that the offender’s dangerousness did not provide a basis for imposing a disproportionately long sentence. Instead, it provided a reason for refusing to decrease the sentence due to any ‘sentiments of compassion’ that might be felt for the offender.²⁵

It was not until the case of R v Kiltie²⁶ in 1974 that the relationship between mental health problems and moral culpability was made explicit. Until then, courts had simply stated (without further explanation) that mental health problems provided a basis for lenience. However, in Kiltie, Chief Justice Bray noted that the offender’s ‘moral guilt’ could be reduced by a mental health condition.²⁷ Where this was the case, the punishment should be reduced, as the penalty must be proportionate to the offender’s culpability. He held that this should occur even if the offender posed a continuing danger to the public, or would receive better treatment if a disproportionate sanction were imposed. The criminal law sanction should be restricted to a proportionate punishment and any further dangers or treatment needs addressed by civil mental health laws.²⁸

The conflict between reduced culpability and increased risk was again addressed by the Federal Court in Channon v The Queen.²⁹ In line with previous decisions, the Court held that imprisonment could not be imposed merely to secure treatment for an abnormality unrelated to the offending, nor simply to serve the psychiatric needs of the offender.³⁰ However, Justice Brennan held that where ‘the abnormality has caused or contributed to crime … the protection of society may require the imposition of punishment according to the gravity of the threat which the abnormality poses, and the seriousness of the conduct to which it has contributed’.³¹ In line with English authorities,³² he held that where there was a ‘grave risk of dangerous crime’, or ‘an almost certain reversion to other forms of serious crime’, it was permissible to impose a lengthy or indefinite sentence of imprisonment, even if that was disproportionate to the offence committed.³³

This preventive phase in Australian jurisprudence was, however, short lived: it was overturned just nine months later by the High Court in Veen v The Queen.³⁴ In a split decision, the majority held that it was impermissible to impose a disproportionate sentence for the purposes of public protection.³⁵ While community protection is an important sentencing consideration, its proper role (as explained in Kocan) is to prevent mitigation rather than extend a sentence.

Throughout this phase of development, the courts never explicitly addressed the question of which mental health problems sentencing judges should take into account. Rather than considering questions of diagnosis, the courts were concerned with three key issues: was the offender deserving of compassion due to their mental health condition; did that condition reduce their culpability for their offending; and did the offender pose a danger to the public. Consequently, the courts interchangeably used various terms such as ‘psychiatric abnormality’,³⁶ ‘mental disturbance’³⁷ and ‘abnormality of mind’³⁸ without attempting to define or limit the scope of these terms. These terms were used inclusively and covered mental illnesses,³⁹ cognitive impairments⁴⁰ and personality disorders.⁴¹

Phase 2 (1978–1987): the rise of general deterrence

Prior to 1978, the offender’s mental health was only linked to the issue of general deterrence on a few occasions, to limited effect.⁴² This changed with the landmark decision of R v Mooney.⁴³ In that case, the offender assaulted two police officers during a state of ‘manic depressive psychosis’.⁴⁴ While counsel urged the imposition of a non-custodial sentence, the trial judge held that it was necessary to send him to prison for deterrent purposes. This was overturned on appeal, with the Court holding that ‘[g]eneral deterrence should often be given very little weight in the case of an offender suffering from a mental disorder or abnormality because such an offender is not an appropriate medium for making an example to others’.⁴⁵ This principle (the ‘Mooney principle’) was based on the view that it was improper to impose a retributive sanction on an offender with mental health problems.⁴⁶ Chief Justice Young also held (contrary to previous cases) that the offender’s reduced culpability should not be a key consideration in these cases: judges should focus on the issue of general deterrence.⁴⁷

The Mooney principle was adopted in many cases throughout the 1980s⁴⁸ and became the court’s primary focus. Discussion of culpability and dangerousness dropped away. Over this period, the Mooney principle was refined in two key ways. First, it was clarified that general deterrence would not always be eliminated as a sentencing consideration: it may remain an important sentencing consideration, or its significance may simply be reduced.⁴⁹ For example, general deterrence may retain importance where there is no evidence that the offender ‘was not aware of what he was doing, was not aware that it was wrong, or was in any way not responsible for what he did’.⁵⁰ In such cases it was seen to be important that compassion for the offender not outweigh the needs of the community. Secondly, an additional justification was added for the principle: that it was inappropriate to use offenders with mental health problems as a vehicle for general deterrence because other people with similar problems would not be deterred by knowledge of the sentence.⁵¹

It was during this period that the issue of scope was first addressed. In R v Hatherley,⁵² the offender had a severe personality disorder, which created a strong tendency to sexual deviation. He was convicted, inter alia, of multiple counts of aggravated rape. On appeal, it was argued that the sentencing judge had failed to give sufficient weight to his personality disorder. While the Court accepted that ‘where an accused is mentally defective or psychiatrically disturbed little weight should be given, on sentence, to general deterrence’,⁵³ it held that personality disorders should not be equated with ‘mental deficiency or psychiatric disturbance’, and so should not provide a basis for reducing the importance of general deterrence as a sentencing consideration.⁵⁴ The Court based this decision on its understanding of the underpinnings of the Mooney principle:

[T]he basis for treating someone who is mentally deficient or the subject of psychiatric disturbance as not an appropriate vehicle for the imposition of a sentence which reflects general deterrence is that such a person may be taken not fully to comprehend the nature and quality of the acts that he or she has done, or not fully to comprehend that those acts are wrong. If that is … the true basis of the decisions to which I have referred, then I think it is plain that what is described as a personality disorder or a personality defect cannot be equated with mental deficiency or psychiatric disturbance.⁵⁵

It is argued, however, that the Court’s understanding of the basis of the Mooney principle was flawed. The foundation for that principle, as originally stated, was that a generally deterrent sentence ‘will not be acceptable if its retributive effect on the offender is felt to be inappropriate to his situation and to the needs of the community’.⁵⁶ This is about the impact the sentence is likely to have on the offender. The court is concerned with the fairness of requiring a person with mental health problems to suffer an additional burden, even if that would serve a generally deterrent purpose. While it is true that later cases had suggested that general deterrence may retain importance where there is no evidence that the offender ‘was not aware of what he was doing, was not aware that it was wrong, or was in any way not responsible for what he did’,⁵⁷ this was simply provided as an example of a circumstance in which the offender’s sentence should not be significantly mitigated. It was not intended to be a statement of the legal test or its underlying basis.

If the court’s concern is with the impact the sentence is likely to have on the offender, there seems to be no valid basis for excluding personality disorders. As is discussed below,⁵⁸ these disorders can be severely impairing, and it may be just as inappropriate to use an offender with a personality disorder as a vehicle for general deterrence as it is to use an offender with a mental illness or cognitive impairment. Moreover, even on the court’s own test there seems to be little basis for excluding all personality disorders, as some personality disorders can at least partly compromise an individual’s ability to fully comprehend that their actions are wrong. For example, a person who has a paranoid personality disorder may feel morally justified in assaulting a victim whom they suspect has done them wrong. Rather than adopting a categorical approach, which seeks to exclude personality disorders as a category, it would have been preferable for the courts to continue to focus on the specific effects of mental health problems on the offender (as they had done in Phase 1).

Phase 3 (1988–1995): combining the principles

In 1988, the High Court again considered the relevance of mental health problems to sentencing.⁵⁹ Rather than solely focusing on general deterrence (as courts had done during Phase 2), it reiterated the position it had taken in Veen (No 1).⁶⁰ It noted the conflicting directions in which sentencing principles pull: reducing culpability and increasing danger. It confirmed that ‘a sentence should not be increased beyond what is proportionate to the crime in order merely to extend the period of protection of society from the risk of recidivism on the part of the offender’.⁶¹ However, within the bounds of proportionality, the need for community protection can properly be taken into account: it can offset any leniency which may otherwise be shown to the offender.

This reminder by the High Court that mental health problems can diminish an offender’s culpability led to a re-emergence of cases in which the offender’s sentence was reduced on this ground.⁶² These cases started to draw out the different ways in which mental health problems can affect culpability: by impairing the offender’s judgment and control over their actions;⁶³ by suspending the offender’s concern for the significance of their actions;⁶⁴ by affecting the offender’s motivation for the offending.⁶⁵ At the same time, courts confirmed the relevance of the Mooney principle,⁶⁶ which was held to apply even in the absence of a causal link between the mental health problems and the offending.⁶⁷

During this period the courts started to integrate the principles developed in Phases 1 and 2, addressing the effects of mental health problems on culpability, public protection and general deterrence.⁶⁸ This integration led courts to resile from the categorical approach adopted in Hatherley,⁶⁹ and the effects of the condition (on the offender’s culpability, dangerousness or the need for general deterrence) became important once more. Personality disorders were thus again treated as potentially mitigating,⁷⁰ due to their possible impact on an offender’s culpability. The principles were also held to apply to mental illnesses and intellectual disabilities, regardless of their severity.⁷¹ Courts were not concerned with precisely defining the scope of the principles and used terms such as ‘mental abnormality’,⁷² ‘mental disorder’⁷³ and ‘mental handicap’⁷⁴ interchangeably.

Phase 4 (1996–2006): expanding the principles

The next significant development occurred in 1996, when the Victorian Court of Appeal provided an overview of five ways in which ‘serious psychiatric illness not amounting to insanity’ could affect the sentencing determination:

First, it may reduce the moral culpability of the offence, as distinct from the prisoner’s legal responsibility. Where that is so, it affects the punishment that is just in all the circumstances and denunciation of the type of conduct in which the offender engaged is less likely to be a relevant sentencing objective. Second, the prisoner’s illness may have a bearing on the kind of sentence that is imposed and the conditions in which it should be served. Third, a prisoner suffering from serious psychiatric illness is not an appropriate vehicle for general deterrence, whether or not the illness played a part in the commission of the offence. The illness may have supervened since that time. Fourth, specific deterrence may be more difficult to achieve and is often not worth pursuing as such. Finally, psychiatric illness may mean that a given sentence will weigh more heavily on the prisoner than it would on a person in normal health.⁷⁵

While sometimes considered to be a mere collation of previously developed principles, this passage refined and expanded on previous jurisprudence in various ways:

• While prior cases had occasionally taken mental health problems into account in determining the type of sanction to impose (for example, imposing a non-custodial rather than a custodial sanction⁷⁶), this had not previously been identified as a separate principle.

• This was the first time a higher court explicitly focused on the relationship between mental health problems and specific deterrence.

• This was the first occasion in which a court confirmed that the principles concerning the impact of ill-health on sentencing⁷⁷ applied to both physical and mental health.

This judgment had a significant effect on jurisprudence in the area and was widely cited by courts around Australia.⁷⁸ Of particular importance was the inclusion of the fifth principle, which related to the possibility that the sanction would impose a disproportionate burden on the offender. In the following years there were a large number of cases which relied on this mitigating factor.⁷⁹ These cases held that the mental health condition did not need to predate the offence to be mitigating, and could even have been caused by the process of conviction and punishment.⁸⁰

Over the next decade, courts continued to refine these ‘Tsiaras principles’ in various ways. For example, courts emphasised that the mere existence of a link between a mental health problem and offending behaviour was not sufficient to reduce the offender’s culpability. The sentencing judge must ‘analyse the nature, severity and effect of the condition’.⁸¹ It was made clear (contrary to previous statements⁸²) that judges must consider this issue separately from the issue of general deterrence.⁸³ It was also made clear that any mitigation may be offset by the danger the offender poses to the community⁸⁴ or by the need for specific deterrence (a sentencing factor that could pull in both directions, depending on the circumstances),⁸⁵ despite the fact that this had not been addressed in Tsiaras.

While the Court had indicated in Tsiaras that general deterrence had no role to play in these cases, it was confirmed that general deterrence may still operate on a ‘sensibly moderated’ basis.⁸⁶ However, there was some confusion about the operation of this principle. Although Tsiaras had made it clear that mitigation may be appropriate even if the offender’s mental health problem had not contributed to the offending behaviour,⁸⁷ courts continued to hold that general deterrence should not be moderated to a great extent where the offender had acted with knowledge of what was being done and with knowledge of the gravity of the offending actions.⁸⁸ They also held that the reduction should not be significant where the offender had a capacity for impulse control.⁸⁹ It is not clear how these decisions fit together.

It was during this period that the issue of scope came to the fore. This was due to the Court’s use of the phrase ‘serious psychiatric illness’ in the much-cited Tsiaras judgment.⁹⁰ This led to arguments between counsel about whether the offender’s condition was sufficiently ‘serious’ to qualify, or whether it constituted a ‘psychiatric illness’.⁹¹ Courts once again started to adopt a categorical approach to this issue, refusing to apply the principles where the offender’s condition was found to fall outside these categories.⁹²

Throughout this phase there was some confusion about whether the Tsiaras principles applied to personality disorders. Most cases accepted that these disorders were covered, as long as they were sufficiently serious and affected the offender in a relevant way.⁹³ For example, the Court in R v Chambers⁹⁴ found that the offender’s borderline personality disorder reduced his culpability, would make prison more burdensome for him and reduced the need for general deterrence; and the Court in R v Pyke⁹⁵ held that although personality disorders were not ‘psychiatric illnesses’ as described in Tsiaras, they nevertheless may reduce the offender’s culpability and ameliorate the need for general deterrence. By contrast, the Court in R v Lawrence⁹⁶ queried the application of the principles to personality disorders (but did not need to decide the issue); and the Court in Lauritsen v The Queen⁹⁷ drew a distinction between mental illnesses, to which the principles applied, and ‘mere personality disorders’, which do not fall within their scope.

Phase 5 (2007–2014): refining the principles

It had become clear by 2007 that the use of the phrase ‘serious psychiatric illness’ in Tsiaras had created unnecessary complications. This led the Victorian Court of Appeal to produce an authoritative statement of the relevant principles, and their scope, in the leading case of Verdins v The Queen.⁹⁸ The Court held that the reference in Tsiaras to a ‘serious psychiatric illness’ had arisen because the offender in that case had such an illness (schizophrenia). The considerations identified were not intended to be limited to illnesses, or to be restricted to conditions of a particular level of gravity: they may apply ‘in any case where the offender is shown to have been suffering at the time of the offence (and/or to be suffering at the time of sentencing) from a mental disorder or abnormality or an impairment of mental function, whether or not the condition in question would properly be described as a (serious) mental illness’.⁹⁹ To this end, it was emphasised that it was inappropriate to take a categorical approach to this issue, and that diagnostic labels ‘should be treated as the beginning, not the end, of the inquiry’.¹⁰⁰ The court’s focus should be on the ways in which the offender’s mental health condition affected their mental functioning at the time of the offence, and how it is likely to affect them in the future.¹⁰¹

The Court also provided an updated statement of the relevant legal principles (the ‘Verdins principles’), holding that impaired mental functioning can be relevant to sentencing in six ways. It can:

1. Reduce the offender’s moral culpability;

2. Influence the kind of sentence to be imposed;

3. Moderate or eliminate the need for general deterrence;

4. Moderate or eliminate the need for specific deterrence;

5. Make a sentence weigh more heavily on the offender than on a person in normal health; or

6. Create a serious risk of imprisonment having a significant adverse effect on the offender’s mental health.

To a large extent this was simply a clarification of the Tsiaras principles. However, the sixth principle expanded the law in the area, establishing that courts also needed to take into account the potentially detrimental effects of imprisonment on an offender’s mental health.

The Court also provided an expansive list of the ways in which impaired mental functioning may reduce culpability.¹⁰² This includes cases in which the mental health condition impaired the offender’s ability to exercise appropriate judgment, to appreciate the wrongfulness of their conduct, to make calm and rational choices or to think clearly. It also includes cases in which the condition made the offender disinhibited or obscured their intent to commit the offence. While this list was intended to be non-exhaustive, it has been the focus of subsequent decisions in this area.¹⁰³

The Verdins principles were seen to ‘reflect the general law of Australia as understood by an intermediate court of appeal’¹⁰⁴ and were accepted in all Australian jurisdictions and in New Zealand.¹⁰⁵ Subsequent cases confirmed that although the Court in Verdins had solely focused on mitigating factors, community protection continued to operate as a potentially countervailing factor.¹⁰⁶ They also confirmed that the offender’s mental functioning ‘need only be impaired, not severely compromised’,¹⁰⁷ and the Verdins principles were applied to a wide variety of conditions.¹⁰⁸ The status of personality disorders, however, remained unclear.

When discussing the scope of the relevant principles, the Court in Verdins did not explicitly mention personality disorders. It did, however, cite with approval various cases in which personality disorders had been accepted as mitigating, such as Chambers¹⁰⁹ and Pyke.¹¹⁰ Many later cases accepted that the principles did apply to personality disorders.¹¹¹ However, other cases held that personality disorders did not attract the considerations discussed in Verdins.¹¹² The reasoning behind this exclusion was generally unclear. It was often simply asserted that personality disorders did not fall within the scope of the Verdins principles, without further exploration. While in some cases courts went further and explained that, as personality disorders are not mental illnesses or psychiatric disorders, they did not engage the Verdins principles, they did not explain why this should be the case.¹¹³ The one exception was R v Hayes,¹¹⁴ in which the Queensland Court of Appeal provided the following reasoning for the exclusion:

A prisoner’s mental illness which reduces her capacity for judgment or understanding, or ability to control behaviour, and therefore reduces the moral blameworthiness in the offending is rightly regarded as an important mitigating factor in the sentencing process. It is my understanding that the law has never regarded personality disorders as having this effect. This is for the reason that they are not illnesses which impact upon the capacity of the sufferer to perceive the world around her and respond to it.¹¹⁵

The Court’s reasoning in this passage is based on three key premises:

1. That an offender’s culpability will only be reduced where his or her capacity for judgment or understanding, or ability to control behaviour, is impaired;

2. That personality disorders cannot reduce an offender’s culpability, as they do not affect the capacity of individuals to perceive the world around them or respond to it; and

3. That because personality disorders cannot reduce an offender’s culpability, they are not mitigating.

All three of these premises are flawed.¹¹⁶ The first premise assumes that culpability will only be reduced in the limited circumstances specified above. However, Verdins made it clear that culpability can be reduced in other circumstances, such as where the offender has an impaired ability to make calm and rational choices, or where their intent to commit the offence is obscured.

The second premise assumes that personality disorders do not impair an individual’s capacity to perceive the world around them (their ‘cognitive capacity’), or their ability to respond to it (their ‘volitional capacity’). However, unless the notion of ‘capacity to perceive and respond to the world’ is interpreted in a very narrow sense, as referring to psychotic-level disturbances (delusions or hallucinations), this assumption is deeply flawed from a clinical perspective. As Walvisch and Carroll have noted,¹¹⁷ the intrinsic features of many kinds of personality disorder can (and do) adversely affect a person’s cognitive and volitional capacities.

The third premise is based on the view that only mental health problems that reduce an offender’s culpability are mitigating. However, the Court in Verdins identified six ways in which mental health problems can mitigate a sentence, only one of which relates directly to culpability. Thus, even if an offender’s culpability cannot be reduced by a personality disorder for the reasons suggested, there appears to be little reason for excluding the operation of the other five Verdins principles. For example, why should an offender with a condition likely to cause them to experience imprisonment as particularly burdensome be precluded from relying on that as a mitigating factor, simply because their culpability was not diminished by that condition?

Although Hayes was decided in 2010, it was not cited by another court throughout this phase. For the remainder of the period, the law concerning personality disorders remained unsettled, with courts adopting varying approaches to this issue.

Phase 6 (2015–2020): limiting the principles

Following the Verdins decision, there was an explosion of cases in which the judgment was cited. In Victoria alone, it was cited in 465 Supreme Court and Court of Appeal cases between 23 May 2007 (when the decision was handed down) and 1 July 2015.¹¹⁸ Courts tried to reduce reliance on the principles, emphasising that the Verdins principles should be regarded as exceptional and should not be invoked in ‘routine cases’.¹¹⁹ They were emphatic that the principles should only be applied ‘after careful scrutiny and assessment, based on cogent evidence, of the relationship between the mental disorder and the offending and other matters’.¹²⁰ Despite these efforts, however, the principles continued to be frequently raised, causing some disquiet amongst judges. This was expressed forcefully by Justice King in R v Miller,¹²¹ in which she commented that Verdins had ‘made the task of sentencing judges more difficult’, and requested that the Court of Appeal ‘re-examine this aspect of sentencing with a view to perhaps reining in the overuse of reliance upon Verdins’.¹²²

The Victorian Court of Appeal was provided with an opportunity to do as Justice King had requested later that year, when the Director of Public Prosecutions (DPP) appealed a decision in which the offender’s personality disorder had been found to be mitigating.¹²³ The DPP argued that moderating a sentence on the basis of a personality disorder was ‘inappropriate and contrary to public policy’ and that personality disorders should be excluded from the scope of the Verdins principles.¹²⁴ The Court agreed, holding that the Verdins principles ‘are enlivened only where the offender suffers from an impairment of mental functioning’,¹²⁵ and that the offender’s personality disorder did not reach that threshold:

It is important to keep in mind that, in Verdins, and in this Court’s subsequent application of Verdins, the Court has consistently stated that the principles in Verdins relate to offenders who suffered from ‘mental impairment’ or ‘impaired mental functioning’, whether at the time of the offending or at the time of sentence. While the Court in Verdins regarded the particular diagnostic label as not being determinative, the principles expressed have always been confined to cases in which the offender suffered an impairment of his or her mental functioning. They do not apply to personality disorders such as those from which the respondent suffered.¹²⁶

The Court in O’Neill based its decision on Hayes,¹²⁷ quoting the passage reproduced above.¹²⁸ In doing so, it accepted the three faulty premises discussed in the previous section. In addition, the Court asserted that personality disorders should be excluded because they do not constitute an ‘impairment of mental functioning’. This demonstrated a further misunderstanding as to their nature. Although personality disorders relate to an individual’s enduring personality traits,¹²⁹ it is only when those traits ‘are inflexible and maladaptive and cause significant functional impairment or subjective distress’ that they are considered to constitute a disorder.¹³⁰ While the level of impairment can vary, it can potentially be very severe. This was clearly seen in the previously mentioned case of Chambers,¹³¹ where evidence was given that the offender’s borderline personality disorder was ‘in many respects, more disabling than many types of actual psychiatric illness’.¹³² It has also been borne out in clinical research.¹³³

The O’Neill decision also seemed to contradict the focus that Verdins placed on the degree of impairment of mental functioning, rather than diagnostic categories, as being determinative. That focus had the advantage of finessing the considerable difficulties inherent in employing diagnostic categories for legal purposes.¹³⁴

Despite these concerns, in the years following the O’Neill decision it was accepted that (at least some) personality disorders were now excluded from the scope of the Verdins principles. While some attempt was made to confine O’Neill to its facts, limiting it to the personality disorder that was present in that case (dependent personality disorder),¹³⁵ it was more commonly accepted that the considerations identified in Verdins no longer applied to any personality disorder.¹³⁶ While evidence that an offender had such a disorder may help to explain their offending behaviour, its existence did not ‘attract the level of mitigation of sentence that must be allowed where Verdins principles are applicable’.¹³⁷ In adopting this position, the courts had returned to the categorical approach which had emerged in Phase 4, but which had been strongly rejected by the Court in Verdins.

Part 2: a collaborative, interdisciplinary law reform strategy

Although borne out of entirely legitimate concerns regarding the poor quality of expert evidence often provided to sentencing judges, the O’Neill decision did not appear to be a satisfactory way forward for sentencing people with mental health problems. As discussed above, its jurisprudential foundation (the decision in Hayes) comprised an erroneous interpretation of the functional implications of psychiatric diagnosis. The consequent assertion that it was appropriate to exclude personality disorders from the Verdins principles, on the grounds that they do not constitute an impairment of mental functioning, was made in the absence of any supporting expert evidence and was inconsistent with contemporary scientific understandings of such disorders.

These concerns extended beyond issues of fairness and moral culpability; they included concerns that the aims of sentencing more broadly, including rehabilitation and community protection, would be compromised by a framework that neglected personality pathology. Given that personality disorders of varying severities and kinds are greatly over-represented amongst criminal offenders,¹³⁸ this had potentially very significant and wide-reaching implications.

Rather than improving the quality of legal discourse around mental health and sentencing, the O’Neill approach risked encouraging a culture whereby legal arguments regarding offenders’ mental health issues at sentencing would become increasingly superficial and etiolated – focused only on the presence or absence of psychosis, mood disorders and cognitive disorders – and devoid of the more nuanced, fine-grained thinking that is required when personality dysfunction is part of the offender’s profile. In such a culture, the basis for judicial deliberations regarding moral culpability, impacts of imprisonment, risk of recidivism and the likely effectiveness of different rehabilitative interventions would inevitably be weakened. In addition, it seemed highly likely that, over time, a criminal justice system that failed to consider the implications of personality disorder at the sentencing stage would promote the development of a correctional rehabilitative system that is similarly superficial in perspective. This is particularly troubling given the very strong evidence base that certain personality factors are important targets for rehabilitative interventions.¹³⁹

As practitioners and scholars working in the area, the authors saw a crucial need to seek reform of the law to ensure that personality disorders could engage the Verdins principles in appropriate cases. This section of the article outlines the collaborative, interdisciplinary approach they adopted to achieve this reform.

Selecting a law reform approach

The first task was to determine the best way to secure reform, given the crowded law reform landscape that currently exists in Australia.¹⁴⁰ That landscape includes various institutional bodies, such as law reform commissions, parliamentary committees, royal commissions and ministerial councils.¹⁴¹ While these bodies play a vital role in advising the government about potential law reform issues, they seemed ill-suited to this task for two reasons. First, they generally provide advice on broad, multifactorial issues,¹⁴² whereas this was a single-issue matter (albeit one with potentially far-reaching consequences). Consequently, it was unlikely to be considered sufficiently significant to refer to one of these bodies for consideration. Secondly, even if the matter was addressed by one of these bodies, legislative implementation of any recommended reforms seemed unlikely. In this regard, Justice Kirby has noted that ‘[t]he difficulty of reliance on Parliament for law reform is that many problems in need of reform are either too hot or too cold to secure Parliament’s attention’,¹⁴³ and this issue appeared to fall into both camps: as a relatively minor issue, it seemed unlikely to make it onto a crowded legislative agenda; and as a reform that may mitigate offenders’ sentences, it was unlikely to appeal to politicians operating in a climate of ‘popular punitiveness’ which favours a ‘tough on crime’ approach.¹⁴⁴

By contrast, a process of strategic litigation seemed well-suited. While the definition of ‘strategic litigation’ is contested,¹⁴⁵ it broadly refers to the use of legal proceedings to seek progressive change to laws, policies or practices.¹⁴⁶ It is litigation that is concerned not just with addressing any injustice caused to the parties involved, but also with addressing broader injustices: it seeks ‘to affect greater, lasting change by winning cases that have an impact beyond the case concerned’.¹⁴⁷

Strategic litigation seemed well-suited to this matter for various reasons:

• It was not a matter that required legislative intervention. The O’Neill decision was handed down by the Victorian Court of Appeal, and could be overturned by that Court or by the High Court.

• There were both legal and policy reasons for overruling the O’Neill decision, which could be relied on by a court if it were inclined to do so.

• The legal arguments had a strong evidentiary foundation, grounded in current psychiatric literature.

• The reforms sought were relatively minor in nature; they did not require a substantial change to the Verdins principles, simply a clarification of their scope.

• Given the detrimental effect that O’Neill was having on many offenders, Victoria Legal Aid (VLA) had indicated a willingness to fund a suitable test case to challenge the decision.¹⁴⁸

• As VLA would be funding the case, the client would not be adversely affected by an unsuccessful outcome. However, a successful outcome would be beneficial to both the client and the broader community.

The long-term nature of the strategic litigation process would also allow the authors to collaborate productively, using their different interdisciplinary skillsets to full potential.¹⁴⁹ Dr Walvisch would be able to use his expertise as a legal scholar, with a detailed knowledge of the Verdins principles,¹⁵⁰ to develop the strongest legal arguments possible and to promulgate those arguments in various fora. Dr Carroll could use his expertise as a forensic psychiatrist to develop the psychiatric aspects of the arguments, ensuring that any reform recommendations were ‘based on the best available evidence, and informed by expert opinion’.¹⁵¹ He would also be able to conduct a psychiatric examination of the client, provide a report for use in the case and give oral evidence in court. Mr Marsh, who at the time was Chief Counsel at VLA, could use that role to identify and represent an appropriate client and could use his skills as a barrister to present the arguments in court.

A five-stage strategy

There were five stages to the adopted strategy: developing the arguments; raising awareness; identifying an appropriate client; preparing the case; and presenting the case. These are outlined in turn below.

Stage 1: developing the arguments

In the first stage, Drs Walvisch and Carroll developed the arguments that would be used in the case, critiquing the O’Neill decision from both legal and psychiatric perspectives. This culminated in the publication of a peer-reviewed article (the ‘personality disorder article’).¹⁵²

Stage 2: raising awareness

In the second stage, Drs Walvisch and Carroll sought to raise awareness of the issues outlined in the personality disorder article, to lay the groundwork for the future litigation.¹⁵³ As the matter would ultimately be determined by judges, judicial officers were a primary target of these efforts. Consequently, Dr Walvisch presented a paper on the topic at the 2019 County Court of Victoria Annual Conference in Melbourne, distributing a copy of the personality disorder article to attending judges. He also made that article available to judicial officers more broadly through the Judicial College of Victoria’s judicial information website.

Drs Walvisch and Carroll also raised the matter for consideration at a meeting of the ‘Forensic Evidence Working Group’ (FEWG), a committee convened by the President of the Victorian Court of Appeal with the aim of developing practice notes to enhance the quality of forensic evidence (including mental health evidence) given in court proceedings. The stakeholders represented on the FEWG included all relevant legal and mental health bodies, including the Supreme and County Court judiciary, the Criminal Bar Association, VLA, the Office of Public Prosecutions, the Faculty of Forensic Psychiatry of the Royal Australian and New Zealand College of Psychiatrists, the Forensic College of the Australian Psychological Society and the Victorian Institute of Forensic Mental Health. Dr Carroll also presented at various forensic psychiatric professional development fora, and Dr Walvisch presented at conferences for law and mental health scholars.

Stage 3: identifying an appropriate client

The third stage required identification of an appropriate client for the test case. This is often a key aspect of strategic litigation and requires consideration to be given to factors such as the nature of the client’s case, the strengths and weaknesses of their claim, their personal characteristics and the effects that success or failure may have on them.¹⁵⁴ In the current context, it was seen to be important that the client’s case show certain features in order to keep the issues at stake clear: relatively serious offending (at County Court or Supreme Court jurisdiction level), stopping short of homicide; the clear presence of a personality disorder of a substantial degree of severity; and the absence of significant confounding factors such as comorbid mental illness, intellectual disability or substance abuse.

Being able to draw on the resources of VLA was highly beneficial in this regard. The size and breadth of its Victoria-wide practice meant that a wide net could be cast, increasing the likelihood of locating a client whose circumstances would suit the strategic aim. In addition, VLA’s criminal law practice is centrally administered, with internal conferences held each year that permit the dissemination of information to staff. It was at one such conference, Criminal Connections, that Mr Marsh highlighted the work of Drs Walvisch and Carroll. He noted that any successful challenge to O’Neill would require the identification of an appropriate vehicle at an early stage and a considered approach by the legal team to ensure that comprehensive evidence of the highest quality could be marshalled in support of the argument. He called upon practitioners to advise him if they became aware of a case that met the relevant criteria. Mr Marsh was subsequently approached by a solicitor within VLA’s indictable crime team: she had been assigned a new client, Daylia Brown, whose circumstances appeared to fit the bill.

Ms Brown had been charged with committing multiple arsons of varying seriousness shortly after her 18^th birthday. Ms Brown had a complex personality disorder, with an extensive history of self-harm and multiple hospital admissions.¹⁵⁵ The charged offences had been committed with the express purpose of being returned to Youth Detention, an environment in which Ms Brown felt safer and more accepted than in the general community. At first blush, Ms Brown’s case appeared to have a wealth of collateral information to support a confident diagnosis, and there appeared to be a tight causal connection between her disordered personality and the offending.

Stage 4: preparing the case

While providing a court with appropriate evidence is always the key to success in strategic litigation, it was especially important in this case, given the primary argument was that the decision in O’Neill lacked an evidential basis and demonstrated a fundamental misunderstanding of the nature of personality disorders. To rectify this, it was necessary to provide strong evidence concerning the nature of Ms Brown’s specific personality disorder, as well as evidence about the impairing nature of personality disorders generally. It was thus essential that Dr Carroll examine Ms Brown at the earliest possible opportunity.

Dr Carroll diagnosed Ms Brown as having a severe personality disorder with detachment and a borderline pattern.¹⁵⁶ That disorder had a ‘very significant’ impact on her ability to exercise appropriate judgment, contributed to her making irrational choices and caused her to ‘fail to experience the inhibitory mechanisms that most people would experience’.¹⁵⁷ While she knew her behaviour was legally wrong, her disorder impaired her ‘emotional awareness that the conduct was wrong on a “deeper moral level”’.¹⁵⁸ Dr Carroll concluded that there was a ‘substantial causal link’ between the offending and the personality disorder.¹⁵⁹ In Dr Carroll’s view, Ms Brown had reasonable prospects for rehabilitation, but these were contingent on a ‘long-term, intensive, tailormade, bespoke rehabilitation plan both in custody and subsequently in the community’.¹⁶⁰ He opined that imprisonment would be counterproductive for Ms Brown’s rehabilitation and risk mitigation, and that her personality disorder would cause her to suffer more than the average prisoner.¹⁶¹

Stage 5: presenting the case

The final stage of the process involved two court hearings: Ms Brown’s sentencing hearing in the County Court of Victoria and the subsequent appeal to the Victorian Court of Appeal. In the sentencing hearing, Mr Marsh called Dr Carroll to give evidence about Ms Brown’s personality disorder, as well as about the nature of personality disorders generally. Dr Carroll informed the Court that personality disorders can affect an individual’s perception of the world (contra Hayes),¹⁶² that they can cause as much impairment and distress as the most severe mental illnesses and that diagnostic manuals no longer preserve a rigid categorical distinction between personality disorders and other mental health conditions.¹⁶³ To confirm the veracity of this information, Judge Taft asked the prosecution to call Professor James Ogloff to also give evidence. Professor Ogloff, a forensic psychologist and director of the Centre for Forensic Behavioural Science, had not examined Ms Brown: his evidence was limited to a broader discussion of the phenomenon of personality disorder and was in a high degree of concordance with that of Dr Carroll.¹⁶⁴ In Professor Ogloff’s opinion, ‘[s]eeking to draw a distinction between the impact on an individual of a clinical disorder and a personality disorder is … “quite arbitrary”’.¹⁶⁵

At both the County Court and Court of Appeal hearings, Mr Marsh highlighted the ambiguity in the law that had been identified in the personality disorder article: that it was not clear whether the Court in O’Neill had intended to exclude all personality disorders from the scope of the Verdins principles (the ‘broad interpretation’), or whether it only meant to exclude disorders that were similar in some way to Mr O’Neill’s dependent personality disorder (the ‘narrow interpretation’).¹⁶⁶ His primary argument was that, even if O’Neill was intended to have broad application, it drew its authority from Hayes, which in turn made broad statements about personality disorders unencumbered by any factual basis. This meant that the key finding in O’Neill, that personality disorders did not enliven the Verdins principles, was not a finding of law: it was a finding of fact based on erroneous premises. As such, it was not binding and should not be followed.

In the event that this argument was unsuccessful, Mr Marsh argued that a narrow approach to O’Neill should be adopted, and that it should be confined to the personality disorders that formed the subject of that appeal. As Ms Brown’s personality disorder was of a different nature, she should be permitted to rely on the Verdins principles.

Part 3: an evidence-based solution: Brown v the Queen

In making the original sentencing determination in this case, Judge Taft accepted Dr Carroll’s opinion that ‘Ms Brown has a severe personality disorder which constitutes an impairment of her mental functioning and which has strong causal links to the offending’.¹⁶⁷ He noted that the courts in Hayes and O’Neill had not received the same level of expert psychiatric evidence about personality disorders that he had,¹⁶⁸ and he acknowledged the force of Mr Marsh’s arguments.¹⁶⁹ However, he took the view that O’Neill was intended to have broad application and that he was bound by that decision.¹⁷⁰ Thus, although he was of the view that Ms Brown’s personality disorder considerably reduced her moral culpability, and should have moderated the relevance of general and specific deterrence as sentencing considerations, he was unable to mitigate her sentence on these bases.¹⁷¹

Given the general level of sympathy to both Ms Brown’s situation and the arguments advanced on her behalf, it may be surmised that adopting the broad interpretation of O’Neill was a shrewd decision by Judge Taft to ensure that the matter was resolved by the Court of Appeal, rather than simply becoming another in a series of ambiguous first-instance decisions that adopted a narrow approach.¹⁷² This view is supported by his explicit invitation to counsel to challenge his decision, his statement that he was ‘sympathetic’ to the Court of Appeal revisiting O’Neill¹⁷³ and his giving the Court of Appeal immediate notice of his decision to facilitate an early listing of the anticipated appeal.¹⁷⁴

The appeal was heard before a specially-constituted bench of five, headed by the President of the Court of Appeal. Due to the strength of the evidence and arguments presented in the sentencing hearing, the DPP conceded in the written case that a blanket exclusion of personality disorders from the Verdins framework could not be sustained.¹⁷⁵ The Court accepted this position, holding:

An offender diagnosed with a personality disorder should be treated as in no different position from any other offender who seeks to rely on an impairment of mental functioning as mitigating sentence in one or other of the ways identified in Verdins. Statements to the contrary in O’Neill should no longer be followed. Whether and to what extent the offender’s mental functioning is (or was) relevantly impaired should be determined on the basis of expert evidence rigorously scrutinised by the sentencing court.¹⁷⁶

In making this decision, the Court of Appeal once again rejected a categorical approach to determining the scope of the relevant principles (as it had done in Phases 3 and 5): ‘any categorical exclusion of personality disorders, whether of such disorders in general or of a particular type of personality disorder, [can] not be sustained’.¹⁷⁷ Instead, as was made clear in Verdins, the sentencing judge needs to consider the nature and extent of the offender’s impairment of mental functioning and its likely impact on the offender at the time of the offending or in the foreseeable future.¹⁷⁸

This approach aligns well with contemporary shifts in how personality disorders are conceptualised in the major psychiatric diagnostic systems. The fifth edition of the Diagnostic and Statistical Manual of Mental Disorders¹⁷⁹ abolished the ‘multiaxial’ framework, first promulgated in its third edition, that placed personality disorders on a separate diagnostic ‘axis’ from mental illnesses. The move back to a non-axial system, removing the conceptual distinction between personality disorders and other mental disorders, is seen to be consistent with the absence of ‘fundamental differences in their conceptualization’.¹⁸⁰ The statement in Brown that ‘[a]n offender diagnosed with a personality disorder should be treated as in no different position from any other offender who seeks to rely on an impairment of mental functioning’¹⁸¹ is consistent with this reasoning.

The ruling in Brown is also aligned with the trend in psychiatric literature away from simple ‘categorical’ approaches to diagnosing personality disorder. There has long been debate as to whether personality disorders are adequately conceptualised as discrete categories (ie, that a person is simply diagnosable with one or more personality disorders, based on whether the diagnostic criteria for each have been met). Although the main text of the DSM-5 has continued to adopt a categorical approach, Section III of the manual describes an ‘Alternative DSM-5 Model for Personality Disorders (AMPD)’ that is in part ‘dimensional’ (assessing severity of general personality functioning).¹⁸² The most recent edition of the World Health Organization’s International Statistical Classification of Diseases and Related Health Problems (the ‘ICD-11’) adopts a similar approach.¹⁸³ The AMPD and ICD-11 both explicitly require diagnostic assessors to determine the severity of a personality disorder prior to describing the specific pathological personality traits¹⁸⁴ relevant to the specific individual.

In Brown, the Court clearly outlined an expectation that the mental health expert will provide cogent evidence regarding severity of any putative personality disorder. It stated, ‘Looking ahead, it would seem from the expert evidence that a personality disorder is likely to engage the Verdins principles only in a case of some severity, but, plainly enough, it is not for this Court to suggest any threshold level of severity which must be reached before those principles would become applicable’.¹⁸⁵ It can be inferred from this that categorical labels alone will not adequately assist the Court: models that include a focus on severity of the disorder will be of more help.

Conclusion

The road to developing the sentencing principles for offenders with mental health problems has been a long and winding one. This article has outlined that the compassion that motivated the initial development of these principles soon gave way to a concern about danger and a focus on culpability. These factors were then overridden by the view that it was improper to seek to deter individuals in the community from committing crimes by imposing retributive sanctions on people with mental health problems. Courts later proceeded to take all of these factors into account, along with the need to consider whether the offender was capable of being deterred by punishment and the impact that punishment was likely to have on him or her. In doing so, they developed a highly ‘sophisticated and subtle’ approach to the relevance of psychiatric symptomatology to sentencing.¹⁸⁶

For most of this time, courts were not concerned with defining the scope of the principles. Instead, they solely focused on the effects of the relevant mental health condition. While there was a slight ‘blip’ in the years following the Tsiaras decision, when some courts confined the principles to ‘serious psychiatric illnesses’, this appeared to be more accidental than designed. The course was corrected by the Court in Verdins, which made it clear that the principles could apply to any impairment of mental functioning.

One consequence of this broad approach was a vast increase in the number of cases raising the principles in the years following the Verdins decision. Regrettably, many of the reports relied on in these cases lacked a ‘rigorous analysis of disorders and their relation to Verdins considerations’,¹⁸⁷ raising concerns amongst the judiciary that the principles were being raised without merit. It was in the context of these concerns that the Victorian Court of Appeal sought to limit the application of the Verdins principles, precluding them from applying to personality disorders.¹⁸⁸ Unfortunately, in doing so, the Court had itself introduced a legal principle that had an insufficient evidentiary foundation, and which had the potential to disadvantage many marginalised people.

To address this problem, the authors undertook a law reform strategy that utilised their different skills as a legal academic, a forensic psychiatrist and a legal practitioner. This strategy was well suited to achieving the desired goals, which were relatively modest in legal terms, but wide-reaching in impact. While each judicial decision is of course based on the unique facts of the relevant case, in its judgment in Brown the Court of Appeal recorded its appreciation ‘of the very constructive way in which this important question’ had been brought forward for its consideration, and of the high-quality evidence provided by the mental health experts involved in the case.¹⁸⁹ It noted that its acceptance of that evidence entailed rejection of its previous views about the nature of personality disorders.¹⁹⁰ In revising its position on this issue, the Court stated that it was affirming ‘one of the law’s most important aspirations, that decision-making should be evidence-based’.¹⁹¹ It is hoped that the collaborative, interdisciplinary approach outlined in this article can be used by others to achieve further evidence-based reforms of this nature.

Ethical standards

Declaration of conflicts of interest

Jamie Walvisch has declared no conflicts of interest.

Andrew Carroll has declared no conflicts of interest.

Tim Marsh has declared no conflicts of interest.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.