Dyer, Andrew --- "Minister for Home Affairs v Benbrika and the Capacity of Chapter III of the Commonwealth Constitution to Protect Prisoners' Rights" [2022] UNSWLawJl 8; (2022) 45(1) UNSW Law Journal 209

---

MINISTER FOR HOME AFFAIRS V BENBRIKA AND THE CAPACITY OF CHAPTER III OF THE COMMONWEALTH CONSTITUTION TO PROTECT PRISONERS’ RIGHTS

ANDREW DYER*

In recent cases in which prisoners have used Chapter III of the Commonwealth Constitution to challenge draconian legislation, the High Court of Australia (‘HCA’) has deployed formalistic reasoning when rejecting their claims. The latest such case was Minister for Home Affairs v Benbrika (‘Benbrika’), where a majority upheld the continuing detention order scheme created by Division 105A of the Criminal Code Act 1995 (Cth), essentially on the basis that imprisonment is not necessarily punishment. Judges should never use such reasoning to avoid striking down laws that breach Chapter III. When they do so, they fail properly to hold power to account. However, the result in Benbrika seems largely justified. Judges are rightly cautious about using Chapter III to strike down punitive laws; and, as Edelman J showed, the Court in Benbrika could exercise restraint without resorting to formalistic evasion. His Honour correctly acknowledged that the HCA has only a limited ability to protect unpopular minorities.

I INTRODUCTION

In recent decades, governments in Western democracies have passed a large amount of draconian legislation that breaches prisoners’ human rights. And, in recent research, I have considered whether the presence of a charter of rights in a jurisdiction can protect prisoners from such legislative excesses.^[1] At all times, my comparison has been between Australian jurisdictions in which there is no charter of rights in force, on one hand, and jurisdictions in which a charter has been implemented, on the other; and the answer that I have arrived at is that charters can lead to improved protections for prisoners against ‘penal populism’.^[2] Because the ‘very purpose’ of a charter is to ‘constrain ... the activities of majorities’,^[3] judges who read and give effect to charter guarantees in accordance with their spirit are apt to strike down – or declare to be incompatible with human rights^[4] – laws that treat disfavoured groups, such as prisoners, unfairly.

This is not to say that charters of rights necessarily will cause prisoners’ rights to be better protected. For instance, there has been a charter of rights in force in Victoria for over a decade;^[5] and yet, as I noted a couple of years ago in this journal,^[6] ‘that [charter] had no effect whatsoever’ on Victorian laws^[7] that, in substance, impose far harsher punishments^[8] on particular, named offenders than had been imposed upon them by the judges who sentenced them. Because the Victorian Parliament disapplied the Charter for the purposes of the relevant legislation,^[9] the prisoners were forced to rely on Chapter III of the Commonwealth Constitution when challenging it. And, as it has done in other such cases,^[10] the High Court of Australia (‘HCA’) declined the invitation to develop the law relating to Chapter III so as to allow the prisoners’ claims to succeed.^[11] Indeed, their Honours – again, consistently with their approach in certain cases^[12] where prisoners have challenged ‘law and order’ legislation on Chapter III grounds – deployed very dubious and highly formalistic^[13] reasoning when refraining from using Chapter III to strike the impugned laws down.

These last observations draw into focus the questions that I wish to consider in this article. As just suggested, in cases where it has had no authority to apply a charter of rights, the HCA has generally shown a distinct unwillingness to use the legal resources that are available to it to interfere with laws that treat prisoners unfairly. To what extent has the Court been right to be ‘wary’ of invalidating laws that, however contrary to human rights they are, have the ‘fundamental purpose of ensuring the safety and protection of the community’?^[14] On the other hand, to what extent has it failed to do its duty to hold the elected branches to account where harshly punitive laws have been challenged on the basis of Chapter III?^[15] Or, to put these questions in different terms, what is Chapter III’s proper role in checking misuses of government power in this context; and in which circumstances would the HCA be misusing its powers if it were to use Chapter III to strike down laws that violates prisoners’ human rights?

I shall argue here that a close analysis of the HCA’s decision in Minister for Home Affairs v Benbrika (‘Benbrika’)^[16] – the most recent case in which the Court considered, and by majority rejected, a Chapter III-based challenge to a law that breaches prisoners’ human rights – helps us to answer the questions just posed.

Benbrika is an interesting case, partly because some reasoning in it might reflect a judicial willingness to evade their Honours’ responsibility to hold power to account, while other reasoning in it rightly recognises the limits of the HCA’s authority to perform such a role. The plurality’s conclusion that Benbrika is not being punished^[17] (a conclusion accepted by Gageler J^[18] and not challenged by Gordon J^[19]), is nothing short of absurd. ‘Full-time custody’ in prison ‘is punitive’^[20] – even where the sole purpose of such custody is to protect the community.^[21] The plurality’s (and Gageler J’s) unwillingness to consider matters of substance when assessing the punitiveness or otherwise of Benbrika’s continuing detention order, is reminiscent of the formalistic approach taken by the Court in cases such as Crump v State of New South Wales,^[22] Knight v Victoria^[23] and Minogue v Victoria,^[24] and by a majority in Magaming v The Queen.^[25] It is contended here that judges are not justified in deploying ‘obviously and woefully inadequate’^[26] reasoning to avoid their obligation to strike down laws that exceed the limits imposed by Chapter III. Indeed, even if the plurality and Gageler J were unaware of the implausibility of their reasoning, such reasoning is still indefensible. It is indefensible because it is unrigorous – that is, even aside from its facilitation of too lax an approach to separation of powers requirements.

Did the reasoning just noted facilitate too lax an approach to the separation of powers? Or is the result in Benbrika – as opposed to some of the reasoning that produced that result – acceptable? In other words, does that result simply reflect an appropriate acknowledgement that, as already noted, Chapter III cannot protect individual rights as a charter of rights can? This brings into play Edelman J’s judgment in Benbrika, which I shall argue adopts a largely supportable approach to which strictures Chapter III imposes on the state’s ability to enact preventive detention schemes.

Justice Edelman eschewed the plurality’s and Gageler J’s formalism. He accepted that Benbrika is being punished.^[27] But Edelman J also found^[28] that it can be within ‘the judicial power of the Commonwealth’^[29] for a Chapter III court, in proceedings ‘detached from the sentencing process’,^[30] to punish an individual for what s/he might do. In so doing, he rejected Gummow J’s statement to the contrary in Fardon v Attorney-General (Qld),^[31] with which Kirby J had agreed.^[32] Justice Edelman had some basis for this approach, partly because, as the Benbrika plurality suggested,^[33] Gummow and Kirby JJ’s reasoning was as formalistic as their own. For Gummow and Kirby JJ, a Chapter III court may validly make a preventive detention order at sentencing.^[34] This, they appeared to think, was punishment for the accused’s prior breach of the law,^[35] which is of course an exclusively judicial function.^[36] But once we accept, as we must, that such preventive detention is ordered not for ‘past crimes’^[37] but because of ‘apprehended conduct’,^[38] Gummow and Kirby JJ’s approach seems to crumble. If, as Gummow and Kirby JJ assumed,^[39] the judge who imposes such an order under a Commonwealth statute is exercising ‘the judicial power of the Commonwealth’, and if s/he is in substance doing the same thing as the judge who imposes a preventive detention order while an offender is serving his/her sentence, why is the latter doing something that Chapter III prohibits?

It will be argued here, then, that in certain recent cases involving prisoners the HCA has abnegated its responsibility to ensure that those prisoners’ rights and liabilities ‘are determined by a judiciary independent of the parliament and the executive’.^[40] It will also be argued that, at first glance, Benbrika seems another instance of the Court’s deploying highly formalistic and unpersuasive reasoning to avoid giving proper effect to the ‘freedoms and liberties’^[41] that Chapter III guarantees. On analysis, however, the position is more complex. That analysis shows that Edelman J was largely right to baulk at the ‘creation of new constitutional restraints ... upon power to detain a person’.^[42] Benbrika’s continuing detention order undoubtedly breaches his human rights.^[43] But Chapter III cannot defend liberty as a Commonwealth charter of rights might.^[44]

II DIVISION 105A – AND WHY IT BREACHES HUMAN RIGHTS

A The Division 105A Scheme and the Reasons for Its Introduction

We must at this stage briefly discuss Division 105A of the Criminal Code Act 1995 (Cth) (‘Criminal Code’), which authorises ‘continuing detention orders’^[45] of the kind to which Benbrika is now subject. Division 105A was inserted into the Criminal Code in 2016^[46] and aims to ‘protect the community from serious Part 5.3 offences by providing that [certain] terrorist offenders who pose an unacceptable risk of committing such offences are subject to: (a) a continuing detention order; or (b) an extended supervision order’.^[47] The detention for which Division 105A provides is ‘post-sentence preventive detention’: it is first ordered, not at sentencing,^[48] but near the conclusion of an adult offender’s sentence for certain terrorism-related offences.^[49] It can be imposed if a state or territory Supreme Court is satisfied: (i) ‘to a high degree of probability, on the basis of admissible evidence,’ that if the offender is released, s/he ‘pose[s] ... an unacceptable risk’ of committing a terrorism-related offence in Part 5.3 of the Criminal Code with a maximum penalty of at least seven years’ imprisonment; and (ii) ‘that there is no less restrictive measure available ... that would be effective in preventing the unacceptable risk’.^[50] The continuing detention order has a maximum timespan of three years,^[51] but can be renewed.^[52] When such an order is made, the relevant Supreme Court must review it annually^[53] (and it may review it sooner in certain circumstances).^[54] At such reviews, the question is whether the statutory criteria just noted are still satisfied.^[55] If they are, the court may affirm the order.^[56] If they are not, it must revoke it.^[57]

One further feature of the scheme must be noted. As indicated above, the person subject to a continuing detention order serves that detention in prison.^[58] Certainly, subject to particular ‘reasonable requirements’, s/he does have to be ‘treated in a way that is appropriate to his or her status as a person who is not serving a sentence of imprisonment’.^[59] And, unless this is ‘necessary’^[60] or ‘reasonably necessary’^[61] for certain purposes, s/he is not to be detained in the ‘same area or unit of the prison’ as sentenced prisoners.^[62] Nevertheless, as the plurality conceded in Benbrika, conspicuously absent from the Division 105A scheme is any ‘special provision for [the] treatment and rehabilitation of detainees’.^[63]

When Division 105A was adopted, the Criminal Code already provided for control orders to be imposed on (among other persons) certain individuals who had been convicted of particular terrorism-related offences.^[64] Accordingly, a question arises: why did the Turnbull Government decide to supplement such orders – which can impose far-reaching restrictions on those to whom they apply^[65] – with continuing detention orders for certain terrorist offenders? Michael Sukkar MP answered this question during the debate that accompanied the second reading of the Bill that, upon its passage through Parliament, inserted Division 105A into the Criminal Code. After expressing the view that the opponents of this ‘sensible’^[66] legislation constituted a ‘scary segment ... of our society’,^[67] Mr Sukkar noted that there were ‘those who questioned the need for a post-sentence detention regime because we have control orders’.^[68] The Member continued:

Even when a control order is in place ... we cannot provide the community with a 100 per cent assurance of protection. The ... saddest recent example is that of the terrorist offender in France who slit the throat of a priest on the altar ... That person was subject to a French version of a control order. ^[69]

The problem with this is that, as Gageler J suggested in Benbrika,^[70] measures that seek to provide the community with such a ‘100% assurance of protection’ from criminal wrongdoing, all too frequently violate individuals’ human rights.^[71] Certainly, as we shall see, the decisions of the United Nations Human Rights Committee (‘UNHRC’) in its Fardon and Tillman communications^[72] make it clear that the Division 105A scheme breaches the International Covenant on Civil and Political Rights (‘ICCPR’).^[73] Indeed, some scholars and judges have argued, or suggested, that preventive detention will in all circumstances, as a normative matter, ‘deprive the [detainee] ... “of the rights of a human being”’.^[74]

B Arguments that Preventive Detention Is Always Morally Impermissible/Contrary to Human Rights

The precise objection here, voiced with perhaps the greatest conviction in the American case law^[75] and literature,^[76] is based upon what Smilansky describes as ‘the idea of respect for autonomous moral personality’.^[77] As that commentator proceeds to explain:

We must respect people’s choices whether to be moral ... To punish before an offence has been committed ... is to treat the person merely as an object. Respect for her moral personality and choice requires of us to give her a ... chance to remain innocent, and not treat her as guilty before she ... is.^[78]

It is this ‘conception of rational personhood’,^[79] this acknowledgment that ‘human beings are genuinely reason-responsive’,^[80] that underlies the statement in Foucha v Louisiana that, under ‘our present system’, ‘with only narrow exceptions and aside from permissible confinements for mental illness’, the state may detain only ‘those who are proved beyond reasonable doubt to have violated a criminal law’.^[81] It also underlies the very similar statement of three HCA justices in Chu Kheng Lim v Minister for Immigration, Local Government and Ethnic Affairs that, leaving aside ‘exceptional cases’ such as ‘[i]nvoluntary detention in cases of mental illness’,^[82] the:

involuntary detention of a citizen in custody by the State is penal or punitive in character and, under our system of government, exists only as an incident of the exclusively judicial function of adjudging and punishing criminal guilt. ^[83]

In other words, according to the philosophy underpinning these statements, the state may detain those who are mentally ill and dangerous. It is not required to reason with the irrational. But if the actor is a reasoning person, the state must reason with him/her: persuasion is all it may use to prevent him/her from acting criminally. If such an actor ignores the criminal law’s moral appeals and ‘threat[s] of punishment’,^[84] the state may punish him/her. But, even then, the punishment must be a reasoning one if it is to be human rights compliant. As many courts have more or less acknowledged, detention so lengthy as to be disproportionate to the seriousness of the offender’s wrongdoing is just as objectifying as detention of an autonomous actor simply because of his/her apparent dangerousness.^[85]

Is it true that we breach a person’s human rights whenever we detain him/her because of what s/he might do? Or do those who maintain such a position display a ‘scary’^[86] indifference to the human rights of those whom such a person might victimise? It is tempting to answer ‘yes’ to the first question and ‘no’ to the second.

Certainly, we might at first be inclined to dismiss as exaggerated the concern expressed by Marshall J in United States v Salerno that preventive detention, once allowed, might lead to ‘tyranny’ and ‘the police state’.^[87] Surely, if it is used as ‘a last resort for the worst cases of terrorist offenders’,^[88] such concerns are misplaced? But, even if they are, it is easy to agree with Jackson J’s earlier admonition, in Williamson v United States, that ‘[i]mprisonment to protect society from predicted but unconsummated offenses is ... fraught with danger of excesses and injustice’.^[89] Especially once we accept that predictions of future violent offending are ‘more often inaccurate than accurate’,^[90] we can see potential for ‘individual injustice’^[91] going beyond that to which Edelman J referred in Benbrika. As his Honour indicated, a person who poses an ‘unacceptable risk’ of terrorism-related offending has not yet done anything to warrant being detained.^[92] But the problem is also that s/he may never have engaged in such activity if released.

C Preventive Detention Nevertheless Seems Justified in Narrow Circumstances

There are, however, other considerations. Ashworth and Zedner^[93] note the well-established principle that European governments will breach the right to life, guaranteed by article 2 of the Convention for the Protection of Human Rights and Fundamental Freedoms (‘ECHR’),^[94] if they fail to take reasonable measures to prevent lethal criminal violence from materialising, where they should have known that the offender posed a real and immediate risk of inflicting such violence.^[95] While this principle should not be taken too far,^[96] it does make us question whether the state will strike a human rights respecting balance between liberty and security if it refuses ever to detain the merely dangerous. Assuming that there are some individuals whom the state can prove are highly likely to commit serious acts of violence if released, even under supervision, would the preventive detention of such individuals for so long as they remain so dangerous, really breach their human rights? Consistently with the views of many other liberal commentators,^[97] I continue to believe^[98] that, if there are people of this kind, the state would in fact breach the human rights of their possible future victims if it were not to detain the former people during the persistence of their proved dangerousness. That said, I also continue to believe that, if it is to exist, preventive detention should apply only to such persons – that is, ‘the most extreme cases; the human time bombs waiting for the opportunity of exploding on release’.^[99]

Must preventive detention satisfy any other conditions if it is to comply with human rights? There are two such conditions and Deane J identified both in Veen [No 2] v The Queen.^[100] After stating that ‘the protection of the community obviously warrants the introduction of some acceptable statutory system of preventive restraint to deal with ... person[s] who ha[ve] been convicted of a violent crime and who ... might represent a grave threat to the safety of other people ... if ... released’,^[101] his Honour provided some indication of what would be ‘acceptable’. He said that the statutory system should be based on ‘periodic orders for continuing detention in an institution other than a gaol’^[102] and ‘provide a guarantee of regular and thorough review by psychiatric and other experts’^[103] of the continuing need for detention. This comports fully with the European Court of Human Rights’ (‘ECtHR’) approach^[104] to post-sentence preventive detention, though it has taken a slightly more relaxed approach to indefinite detention.^[105] The former type of detention, the Court has held,^[106] will breach article 5(1) of the ECHR^[107] – and, in certain circumstances, article 7(1)^[108] – if served in an ‘ordinary prison’ (even if detainees are in a different area of the prison from other prisoners).^[109] For both forms of detention, article 5(4) of the ECHR requires regular judicial review of its continuing necessity.^[110] Moreover, in James v United Kingdom, the Court held that, while prisoners may be required to serve the preventive part of a sentence of indefinite detention in prison, their detention will not be ‘lawful detention of a person after conviction by a competent court’ (and so will breach article 5(1)) unless the state provides them with ‘reasonable opportunities to undertake courses aimed at helping them to address their offending behaviour and the risks they [pose]’.^[111]

D Two Further Issues

Before returning to the reasoning in the Australian preventive detention cases – and, most especially, that in Benbrika – we must deal briefly with two issues.

The first is whether there is in fact any relevant distinction between post-sentence preventive detention and indefinite detention. This is important because, if no such distinction exists, this has implications for our assessment of certain justices’ reasoning in both Benbrika and Fardon. We have seen that an important step in Gummow and Kirby JJ’s reasoning in Fardon was their Honours’ acceptance that these forms of detention differ qualitatively from one another.^[112] If their Honours were wrong, this calls into question their contention that a Commonwealth statute will breach Chapter III if it authorises a judge to order post-sentence preventive detention,^[113] but will not breach Chapter III simply because it authorises him/her to order indefinite detention.^[114]

As I have argued elsewhere,^[115] post-sentence preventive detention and indefinite detention are in substance almost exactly the same thing as one another. With both, a criminal offender is ordered to stay in detention after his/her retributive sentence has expired, because the state has proved that s/he is ‘a serious danger to the community’.^[116] The only difference is that indefinite detention is ordered at sentencing, whereas post-sentence preventive detention is ordered only once the offender is serving his/her sentence (as in Benbrika). It follows that, as stated above, Gummow and Kirby JJ were wrong in Fardon to treat these two forms of detention as being relevantly distinct from one another.^[117] We shall return to this issue.

This brings us to the other issue with which we must deal before discussing the Australian case law. That issue is whether Division 105A of the Criminal Code is compatible with human rights. It is not. The first human rights problem with the Division is that it does not merely allow for preventive detention where the state can prove that a person is highly likely to act very violently if released (even under supervision). The state merely has to prove that it is highly likely that s/he poses an unacceptable risk of committing one of the Part 5.3 offences with a maximum penalty of at least seven years’ imprisonment. There are in fact two difficulties here. Consistently with Slobogin’s argument,^[118] a high probability of an unacceptable risk of terrorist-related offending differs from a high probability of terrorist-related offending. Further, as Gageler J and Gordon J both emphasised in Benbrika, many of the relevant Part 5.3 offences do not involve the commission, or facilitation, of serious acts of violence. As Gageler J put it, ‘[t]he prophylactic approach taken to the imposition of criminal liability [in Part 5.3]’ means that ‘a serious Pt 5.3 offence can involve conduct many steps removed from doing or supporting or facilitating any terrorist act’.^[119] The second human rights problem with Division 105A is that, while the person detained under it will seemingly often be detained away from sentenced prisoners,^[120] his/her conditions will not differ materially from those of other prisoners; and no extra rehabilitative resources will be directed his/her way.^[121]

Indeed, as noted above,^[122] decisions of the UNHRC make it clear that preventive detention schemes of the kind for which Division 105A provides, breach the ICCPR – and, most particularly, the right not to be arbitrarily detained.^[123] ‘To avoid arbitrariness’, the majority stated in both Fardon v Australia and Tillman v Australia, ‘the State Party should have demonstrated that the author’s rehabilitation could not have been achieved by means less intrusive than continued imprisonment or even detention’.^[124] In other words, consistently with the moral argument just presented, the ‘penal’^[125] character of the applicants’ respective detention was crucial to the majority’s finding that that detention breached positive human rights norms.

We can now consider the central question that this article seeks to address. Given that the Commonwealth Government did act contrary to human rights standards when it inserted Division 105A into the Criminal Code, was the HCA wrong in Benbrika to reject Benbrika’s challenge to that Division? Did it cravenly fail to comply with its obligation to hold power to account? Or does the result in Benbrika amount instead to a sensible and proper acknowledgement of the limits of the Court’s own powers?

III AUSTRALIAN PREVENTIVE DETENTION SCHEMES AND CHAPTER III

If we are properly to assess the reasoning in Benbrika, we must understand two cases – Kable v Director of Public Prosecutions (NSW) (‘Kable’)^[126] and Fardon – in which the HCA has considered the constitutional validity of state preventive detention regimes.

A Kable

The events that led to the litigation in Kable are well known, as is the outcome of that litigation. While Gregory Wayne Kable was in prison for the manslaughter of his wife, the authorities became aware that he had written a series of threatening letters to his sister-in-law, who now had the custody of his children.^[127] ‘In the highly charged pre-election atmosphere of the day’, the New South Wales (‘NSW’) Government ‘asserted that there was an urgent need for legislation’ that provided for Kable’s continuing detention once his sentence expired.^[128] The result was the Community Protection Act 1994 (NSW) (‘Community Protection Act’), which applied to Kable alone. Section 5(1) authorised the NSW Supreme Court (‘NSWSC’), at the conclusion of Kable’s sentence, to detain him for a renewable term of six months if it was ‘satisfied, on reasonable grounds’ that, upon his release, he was ‘more likely than not to commit a serious act of violence’ and that it was ‘appropriate, for the protection of a particular person or persons or the community generally’ that he remain in custody.

According to an HCA majority, the Act was invalid because of its failure to observe a previously undiscovered limitation on state legislative power, imposed by Chapter III of the Commonwealth Constitution. Contrary to what had hitherto been supposed,^[129] their Honours held, Chapter III does not merely require state courts to remain ‘courts.’ Rather, by setting up ‘an integrated Australian court system’ with the HCA at its zenith,^[130] and by permitting state courts to exercise the ‘judicial power of the Commonwealth’,^[131] Chapter III was said to require that state courts remain fit repositories of such power. While they could validly exercise certain non-judicial functions,^[132] the majority held, state courts could not validly exercise non-judicial functions that were ‘repugnant to or incompatible with’ their exercise of federal judicial power.^[133] More particularly, they could not validly exercise the non-judicial function that the Community Protection Act had purportedly conferred on the NSWSC.^[134]

Two aspects of Kable are particularly relevant to the present discussion. The first is that at least three of the majority justices considered that Kable had been punished.

Crucial to Toohey J’s conclusion that the impugned Act was invalid, was that it authorised ‘the Supreme Court [to] ... order the imprisonment of a person although that person ha[d] not been adjudged guilty of any criminal offence’.^[135] Justice Toohey clearly regarded provisions granting judges the power to impose indefinite sentences to be consistent with Chapter III’s requirements.^[136] Such a sentence, he implied, amounts to the punishment of ‘criminal guilt’^[137] and therefore is as clear an instance of judicial power as can be imagined.^[138] But, according to his Honour, detention under the Community Protection Act was not punishment for Kable’s manslaughter offence.^[139] Kable had instead been detained because of concerns about ‘what [he] might do’.^[140] After noting^[141] the principle from Lim^[142] – namely, that involuntary detention imposed otherwise than to punish criminal guilt is only exceptionally ‘non-punitive’ (and generally, therefore, cannot be ordered in the exercise of ‘the judicial power of the Commonwealth’) – Toohey J found that Kable’s detention did not ‘fall within the “exceptional cases” mentioned in Lim, directly or by analogy’.^[143] In other words, Kable had been punished for ‘future crimes’. The Court that so punished him, his Honour concluded, had exercised a non-judicial function that was incompatible with its exercise of the ‘judicial power of the Commonwealth’.^[144]

Justices McHugh and Gummow, too, stated that Kable’s detention had amounted to punishment. For the latter, while the impugned Act had a number of ‘striking features’, ‘the most significant’ was that, ‘whilst imprisonment pursuant to Supreme Court order is punitive in nature, it is not consequent on any adjudgment by the Court of criminal guilt’.^[145] Justice Gummow thought it ‘plain’ that the Commonwealth Parliament could not confer ‘such an authority’ on a Chapter III court.^[146] And he concluded that the function was not only non-judicial; it was also ‘of such an extraordinary nature’ as to be ‘incompatible with the [NSWSC’s] exercise ... of federal jurisdiction conferred pursuant to s 77(iii) of the Constitution’.^[147] Likewise, while McHugh J did not attach determinative significance to this consideration when finding that the Act infringed Chapter III, his Honour observed that it provided for ‘punishment by way of imprisonment for what the appellant is likely to do as opposed to what he has done’.^[148]

The final majority justice, Gaudron J, did not state in terms that Kable had been punished. Nevertheless, her Honour seemingly agreed with Toohey, McHugh and Gummow JJ about this matter.^[149] By allowing for Kable’s detention because ‘an opinion [had been] formed’ in irregular proceedings, that he would probably commit a serious offence in the future, Gaudron J said, the Act provided for a process that was ‘the antithesis of the judicial process’.^[150] ‘[O]ne of the central purposes of [that process]’, her Honour continued, ‘is ... to protect “the individual from arbitrary punishment and the arbitrary abrogation of rights ...”’.^[151]

The second aspect of Kable that is presently relevant is its controversial nature. Many commentators focused on the dubiousness of the reasoning that the majority justices deployed in an effort (those commentators thought) to defend liberty. For Winterton, that reasoning was ‘barely plausible in several respects’.^[152] And he concluded that, if constitutional implications are ‘merely plausible’ and are ‘not solidly based on the Constitution’s text or structure’, judges should exercise restraint – even if this frustrates ‘a result which ... may promote human or civil rights’.^[153] Taylor has made like criticisms, noting, like Winterton, that ‘the decision in Kable was based largely on the understandable but irrelevant judicial distaste for the legislation in question’.^[154] And while, for Goldsworthy, the ‘eloquen[ce]’ of ‘Kable’s ... Chief Counsel, Sir Maurice Byers QC’,^[155] rather than human rights considerations, might have led the majority to take a less rigorous approach than was warranted, he has also argued that Winterton in fact ‘overstated’ the plausibility of their Honours’ reasoning.^[156]

We need not examine the precise deficiencies that these three commentators considered the majority’s reasoning to possess. Instead, the critical point is that, around the time of the Kable decision, there arose a perception in some quarters that, in that case and in others, the HCA had gone beyond applying the law in ‘the books’^[157] and had ‘tended to treat itself as another legislature even though it was not chosen by the people’.^[158] This ultimately led to calls from the then Deputy Prime Minister to appoint a ‘capital C conservative’ to the Court;^[159] and the Howard Government then appointed, in fairly quick succession, Callinan J, Gleeson CJ and Heydon J. Soon after his appointment, the new Chief Justice signalled a change of direction. ‘Our laws were not made to be administered by computers’, he said.^[160] ‘Ultimately, however’, he continued, ‘in the administration of any law, there comes a point beyond which discretion cannot travel’.^[161] ‘Judges whose authority comes from the will of the people, and who exercise authority upon trust that they will administer justice according to law’, Gleeson CJ concluded, ‘have no right to subvert the law because they disagree with a particular rule’.^[162] As we shall see, his Honour made strikingly similar remarks when dismissing the prisoner’s appeal in Fardon,^[163] a case that is best understood as a response to claims that the Mason and Brennan Courts had sometimes ‘overstep[ped] their constitutional mandate’^[164] – and as an assurance that the Gleeson Court had an appetite only for ‘strict and complete legalism’.^[165]

B Fardon

At issue in Fardon was the constitutional validity of the Dangerous Prisoners (Sexual Offenders) Act 2003 (Qld) (‘DPSOA’), which, as Meagher has observed, seemed to share with the Kable law the following characteristic: it ‘imposed punishment for possible rather than proven criminal conduct’.^[166] Similarly to the law challenged in Benbrika, if the state could prove to a high degree of probability that certain sexual offenders would pose an unacceptable risk of committing certain sexual offences if released at the conclusion of their respective sentences, even under supervision, the Queensland Supreme Court could order their indefinite continuing detention in prison.^[167] A majority in both the Queensland Court of Appeal (‘QCA’) and the HCA, however, found the DPSOA scheme to be distinguishable from the law struck down in Kable. That said, there were interesting differences in the reasoning deployed by the various majority justices in the latter tribunal. We must note one of those differences.

In both the QCA^[168] and the HCA,^[169] the then Queensland Solicitor-General, Patrick Keane QC, argued that, if the law had been a Commonwealth enactment, it would not have infringed Chapter III. This was so, he submitted, because the Act provided for a new ‘exceptional case’ of non-punitive detention that, due to its non-punitive nature, could be ordered by a Chapter III court in the exercise of the ‘judicial power of the Commonwealth’.^[170] Acceptance of this contention would have been fatal to Fardon’s argument that the DPSOA infringed Kable. If that law would have been valid if passed by the Commonwealth Parliament, it followed that it did not offend the less stringent limitation on state legislative power recognised in that case.^[171] But only some members of the HCA majority thought that the Solicitor-General’s argument was right.

Justices Callinan and Heydon noted that, as recognised in Lim,^[172] quarantine and mental illness detention are two of the ‘exceptional cases’ of non-punitive detention that either a ‘non-judicial [or] ... judicial bod[y]’ may validly order under a Commonwealth statute.^[173] They also noted^[174] the well-established proposition^[175] that the categories of non-punitive detention are not closed, and that, when determining whether detention is non-punitive, the question is whether the detention is ‘reasonably capable of being seen as necessary for a legitimate non-punitive objective’.^[176] In the QCA, de Jersey CJ^[177] and Williams JA^[178] had found that this test was satisfied: the Act’s principal object, stated in section 3, was to protect the community, not to punish the prisoner further. Moreover, the Chief Justice had held that, because of this protective purpose, involuntary detention of sane, though dangerous ‘criminals’ was analogous to the involuntary detention of the mentally ill.^[179] Justices Callinan and Heydon essentially agreed with this analysis.^[180] After noting that the Act’s ‘objects are stated to be to ensure protection of the community and to facilitate rehabilitation’,^[181] and that a continuing detention order will be issued upon proof that the offender is ‘a serious danger, or an unacceptable risk to the community’,^[182] their Honours concluded that the DPSOA was properly to be characterised as ‘a protective rather than a punitive enactment’.^[183]

Justice McHugh was seemingly of like opinion. ‘[W]hen determining an application under the Act’, his Honour said, ‘the Supreme Court is exercising judicial power’.^[184] Further, like Callinan and Heydon JJ, he noted the Act’s stated objects, and concluded that ‘the Act is not designed to punish the prisoner. It is designed to protect the community ... ’.^[185] Certainly, in Benbrika, Edelman J argued that McHugh J in Fardon ‘did not reach any conclusion’ about whether Fardon was being punished.^[186] And certainly, as Edelman J also pointed out,^[187] any conclusion on McHugh J’s part that detention under the DPSOA was non-punitive, was inconsistent with his Honour’s statement in Kable^[188] that Kable had been punished. Nevertheless, McHugh J does appear to have held that Fardon’s detention was non-punitive. According to his Honour in Re Woolley; Ex parte M276/2003,^[189] so long as a Commonwealth law’s ‘purpose is non-punitive’, that law ‘will not offend the separation of powers doctrine’. Given this view, it seems clear that, when McHugh J said in Fardon that the DPSOA’s ‘design’^[190] was to protect, not to punish, he did mean that the relevant detention was not punishment.

The other four members of the HCA in Fardon, however, either did not express a concluded view about,^[191] or rejected,^[192] the submission that the DPSOA would have been valid had it been a Commonwealth statute. In the QCA, McMurdo P, in dissent, had noted that ‘[t]he object of the [Community Protection Act] was to protect the community by providing for the preventive detention of a single individual, Kable’.^[193] Yet, probably all members of the majority in Kable considered that detention to be punitive. Why should a different conclusion be reached in the case of the DPSOA? Her Honour held that it should not. ‘Despite the stated objects of the Act’, McMurdo P concluded, ‘the effect of the Act is punitive and not within the exceptions referred to in Chu Kheng Lim’.^[194] Crucial to this conclusion was that Fardon was ‘subject to substantially the same regime of detention as if convicted of a criminal offence’.^[195]

In the HCA, Kirby J deployed very similar reasoning. ‘[B]y Australian constitutional law’, his Honour stated, ‘punishment ... is reserved to the judiciary for breaches of the law’.^[196] Here, however, Fardon was being punished ‘because of an estimate of some future offence’.^[197] The objects clause in the DPSOA,^[198] his Honour thought, should not distract attention from the detention’s ‘true character or punitive effect’.^[199] The detainee

remains effectively a prisoner. He or she is retained in a penal custodial institution, even as here the very prison in which the sentences of judicial punishment have been served. ^[200]

On the other hand, Gummow J, while agreeing that a Chapter III court could not order such detention in the exercise of ‘the judicial power of the Commonwealth’,^[201] made no claim that Fardon’s detention was punishment. That is because, although his Honour accepted the Lim statement that, the ‘exceptional cases’ aside, a Commonwealth law may only authorise judicially ordered detention following a finding of past guilt, he thought that that statement should be reformulated so as to remove any reference to punishment.^[202] As his Honour had explained two months previously in Al-Kateb v Godwin, ‘[o]nce it is accepted that many forms of detention’ – including imprisonment ordered by a judge following a criminal trial – ‘involve some non-punitive purpose’,^[203] it followed that ‘the focusing of attention on whether detention is “penal or punitive in character” is apt to mislead’.^[204]

That said, Gummow J’s reformulation would cause no different results from those produced by the Lim principle.^[205] Under it, the question would remain whether the relevant detention was analogous to one of the ‘exceptional cases’ recognised in Lim.^[206] If it was not, Commonwealth legislation could not validly authorise a Chapter III court to order it. We have seen that, in the QCA, de Jersey CJ accepted the Queensland Solicitor-General’s submission that Fardon’s detention was indeed analogous to one of the existing categories of non-punitive detention. The reasoning seemed to be that, like mental illness detention, the detention of a sane but dangerous offender past the expiry of his/her sentence aimed to protect the community from violence.^[207] But Gummow J thought that no analogy existed. His analysis here was terse. ‘[I]t is not suggested’, he said, ‘that ... the detention of the mentally ill for treatment is of the same character as the incarceration of those “likely to” commit certain classes of offence’.^[208] Perhaps his Honour gave some further insight into his thinking, however, when he proceeded also to distinguish the detention at issue from pre-trial detention (one of the other ‘exceptional cases’^[209]). ‘[D]etention by reason of apprehended conduct, even by judicial determination on a quia timet basis’, he argued, ‘is of a different character [from pre-trial custody] and is at odds with the central constitutional conception of detention as a consequence of judicial determination of engagement in past conduct’.^[210] In other words, Gummow J’s reasoning seems to have been predicated on the kind of autonomy respecting analysis noted above.^[211] Mental illness detention, he seems to have said, is qualitatively different from the detention of the merely dangerous. The former is imposed on those who cannot reason. The latter is imposed on those who, because they can reason, may only be detained once they have made a choice to offend.^[212]

What this analysis did not satisfactorily explain, however, was why indefinite detention^[213] is constitutionally acceptable. Like McMurdo P in the QCA,^[214] and like Kirby J in the HCA,^[215] Gummow J stated clearly enough^[216] that ‘[p]reventative detention regimes attached by [Commonwealth] legislation to the curial sentencing process upon conviction’^[217] do not offend Chapter III. Such reasoning was seemingly founded on the view that, as McMurdo P put it, such a power of detention amounts to no more than an exercise of the ‘traditional judicial function of imposing a penalty after a criminal conviction’.^[218] But while Gummow J was right to note that indefinite detention has ‘a long history in common law countries’,^[219] such detention is not relevantly different from post-sentence preventive detention. Once more (and contrary to what Toohey J said in Kable):^[220] ‘prior conduct’ is not ‘the basis’ upon which a sentencing judge orders that a person remain in custody after the expiry of his/her retributive sentence. It is merely of ‘evidentiary’^[221] relevance when the judge answers the crucial question, a question that does not differ from that which s/he answers when determining whether to make an order of post-sentence preventive detention. That question is: is the risk that this person will offend in the future high enough to warrant a preventive detention order? Or, to put the matter as Edelman J did in Minogue,^[222] preventive detention is ‘forward looking’ – whether it is imposed at sentencing or while an offender is still serving his/her sentence. It does not ‘impos[e] ... additional punishment for a past offence’.^[223]

As noted above, two justices in Fardon refrained from stating whether the DPSOA would have been valid had it been a Commonwealth law.^[224] They could afford to do this because the DPSOA was a state law and, as also mentioned above,^[225] state laws will infringe Chapter III of the Commonwealth Constitution in more narrow circumstances than Commonwealth laws will.^[226] Accordingly, it was not contradictory for Gummow J to hold,^[227] with five other justices, that the DPSOA was valid.

The important matter for present purposes is not the precise reasoning that led the majority justices to hold that ‘the process for which the [impugned] Act provide[d]’^[228] was sufficiently different from that created by the Community Protection Act to allow Kable to be distinguished. Rather, it is that there was nothing inevitable about this decision. As McMurdo P and Kirby J showed, and as many commentators have subsequently suggested,^[229] it was open to the Court to find that, as in Kable, this was a punitive preventive scheme that could not operate compatibly with Chapter III. And this brings us back to the Gleeson Court’s ‘legalism’.

It was noted above^[230] that, in Fardon, Gleeson CJ made some remarks that are strikingly similar to those in the Australian Bar Review article^[231] in which he announced his commitment to what Campbell has approvingly called ‘antipodean positivis[m]’.^[232] After noting that the majority justices in Kable had made some references to ‘the capacity of the legislation there in question to diminish public confidence in the judiciary’, his Honour said:

[N]othing would be more likely to damage public confidence in the integrity and impartiality of courts than a judicial refusal to implement the provisions of a statute upon the ground of an objection to legislative policy. If courts were to set out to defeat the intention of Parliament because of disagreement with the wisdom of a law, then the judiciary’s collective reputation for impartiality would quickly disappear. ^[233]

HCA justices tend to quote this statement when rejecting Chapter III challenges to harshly punitive legislation.^[234] Why? What does it mean? It means, relevantly, that challenges to laws of this kind should rarely succeed – the courts should be ‘wary’^[235] about striking such laws down – because, once the courts do intervene readily in this area, a perception can arise that they are doing so because of their personal distaste for the impugned legislation.^[236] Most HCA justices recognise that they have a law making role.^[237] But they also recognise the importance of ensuring that the Court is seen as ‘an apolitical institution’, the decisions of which do not simply reflect ‘the whims’ of its members.^[238] And they understand that, while the press and public will tolerate judicial creativity – and even the ‘stretch[ing]’ of the law^[239] – where this delivers results consistent ‘with ... present expectations or values’,^[240] they are inclined to be harshly critical of outcomes that, however defensible the reasoning that produces them is, clash with the values of ‘the community or an important section of it’.^[241] This understanding appears to tell us much about why, in Fardon, when choosing how broadly or narrowly Kable should apply, the majority chose to keep it narrow.

Before moving to the reasoning in Benbrika, we must highlight three aspects of Fardon. First, Callinan and Heydon JJ, and also seemingly McHugh J, deployed very formalistic reasoning when rejecting Fardon’s challenge to the DPSOA. This reasoning – that Fardon’s detention in gaol was not punitive – also clashed with what three justices (including McHugh J) had explicitly held in Kable. Secondly, however, in one important respect, Gummow and Kirby JJ’s reasoning was no less questionable. Their Honours contended that detention under the DPSOA was relevantly distinct from preventive detention ordered at sentencing. This contention was of some importance to their conclusion that, if the DPSOA had been a Commonwealth law, it would have been invalid. For the reasons given above, the distinction that their Honours perceived does not exist. Thirdly, as just noted, the majority’s conclusion that the DPSOA was a valid enactment seems to reflect their Honours’ concern that, at least without a charter of rights,^[242] the judicial law-making function must remain highly attuned to the wishes of the community.

All of this must be kept in mind when assessing Benbrika – a case, like Fardon, that raises questions about the extent to which the HCA should use Chapter III to hold the elected branches of government to account and defend individuals against legislative breaches of their human rights. Is the Court rightly slow to develop the law in this area in a manner in which the majority of the community and/or ‘the spokesmen of powerful interests’^[243] might disapprove? If so, have their Honours been right even to use formalistic reasoning to achieve such results? Finally, what implications does Gummow and Kirby JJ’s formalistic reasoning have for their views about the Commonwealth’s ability validly to pass legislation of the kind at issue in Benbrika?

C Some Principles regarding the Commonwealth Constitution and Punishment

In the time between Fardon and Benbrika, the HCA clarified some matters. After some doubt had been introduced about this issue,^[244] the Court accepted that a limitation on power of the kind recognised by Brennan, Deane and Dawson JJ in Lim exists and flows from Chapter III. When stating this principle, their Honours adopted the plurality’s language in Lim: contrary to Gummow J’s reformulation in Fardon,^[245] they chose not to eschew references to punishment. ‘[T]he involuntary detention of a citizen in custody by the State’, the plurality said in Falzon v Minister for Immigration and Border Protection,^[246] ‘is [generally] penal or punitive in character and under our system of government exists only as an incident of the exclusively judicial function of adjudging or punishing criminal guilt’. That said, as noted above,^[247] that principle would seem to have no different scope from that which Gummow J recognised.^[248] The Court also made it clear that, when a court assesses whether detention is non-punitive, it asks what ‘the true purpose of the law’ is.^[249] If the detention is ‘reasonably capable of being seen as necessary for a legitimate non-punitive objective’, the Commonwealth law that authorises it will not breach Chapter III.^[250] But if the detention ‘goes further’^[251] – if it cannot reasonably be seen as necessary to achieve the relevant non-punitive purpose – ‘it may be inferred that the law has a purpose of its own, a purpose to effect punishment’.^[252] The Court has denied that the ‘reasonably capable’ inquiry involves proportionality testing.^[253] But, as many commentators have noted,^[254] it is very difficult to accept that this is, or at least should be, so. As Zines pointed out,^[255] McHugh J, whose views about the irrelevance of proportionality analysis the HCA has now accepted,^[256] thought that a law that capriciously authorised the solitary confinement of asylum seekers would be invalid because ‘it would go beyond’ what could reasonably be seen as necessary to achieve the non-punitive purpose of ‘prevent[ing] ... detainee[s] from entering the Australian community while [their visa applications] ... were being determined’.^[257] As Zines observed, that seems just to be another way of saying that it would be ‘disproportionate in relation to the end’ sought.^[258]

The suspicion exists that the HCA’s reluctance to conduct proportionality assessments in this context owes much to its concern not too readily to invalidate schemes that, however punitive they are, have been provided for by democratically elected governments, with broad public support. For, when a court asks itself whether detention is proportionate to the non-punitive end upon which the government relies, it takes a critical approach to that government’s claim that the detention is justified.^[259] Rather than simply considering the purpose that the government says it is pursuing, taking the government at its word, and leaving it at that, the court will normally also assess the effects of the detention. It is not just in Fardon that members of the HCA have displayed a distinct unwillingness to do this.^[260] Indeed, this was a feature of the plurality’s reasoning in Benbrika.

D Benbrika

In Benbrika, Keane J considered the validity of the argument that he had advanced before the QCA and HCA in Fardon nearly 20 years before.^[261] Writing with three colleagues, he accepted that argument. The plurality noted that, in Fardon, Gummow J had acknowledged^[262] that ‘schemes for preventative detention have a long history in common law countries’.^[263] Their Honours also noted, correctly in my view, that

if the lawful exercise of judicial power admits of the judge assessing the danger an offender poses to the community at the time of sentencing it is curious that it does not admit of the judge making such an assessment at or near the time of imminent release when that danger might be assessed more accurately.^[264]

But the plurality did not argue that a Commonwealth law may validly authorise a Chapter III court to order punitive preventive detention. Rather, because their Honours accepted the Lim principle,^[265] they concluded that Division 105A would only be valid if the detention that it authorised was non-punitive.^[266] Indeed, not only that: it was also necessary that that detention be analogous to an existing category of non-punitive detention.^[267]

Like the QCA majority in Fardon,^[268] and like Callinan and Heydon JJ when that matter reached the HCA,^[269] the plurality thought that the analogy could undoubtedly be drawn and that the detention was non-punitive. Although such detention is served in prison, and although under Division 105A there is no ‘special provision for treatment and rehabilitation of detainees’,^[270] the plurality noted that that Division’s stated^[271] object ‘is plainly directed to the protection of the community from harm’.^[272] And, for their Honours, that was enough. ‘There is no principled reason’, they said, ‘for distinguishing the power of a Chapter III court to order that a mentally ill person be detained in custody for the protection of the community from harm and the power to order that a terrorist offender be detained in custody for the same purpose’.^[273] Such detention was not punitive, their Honours held. While detention in prison ‘prima facie’ is punitive, they conceded, ‘that characterisation may be displaced by an evident non-punitive purpose’.^[274]

Their Honours never mentioned the established rule that, to be non-punitive, detention must be reasonably capable of being seen as necessary to achieve a non-punitive aim.^[275] They therefore refrained from considering whether prison – as opposed to a less punitive form of detention – can reasonably be seen as necessary to protect the community. And they wrongly stated that ‘[t]his Court has consistently held, and most recently in Fardon, that detention that has as its purpose the protection of the community is not punishment’.^[276] As we have seen, at least three justices in Kable stated that the detention at issue there, despite its protective aim, was punishment;^[277] and only a minority in Fardon stated that detention under the DPSOA was not.^[278]

Did this oversight, and did this error, evidence a judicial concern not to challenge Parliament’s will regarding a controversial issue of public policy? Certainly, Edelman J, who, like the plurality, upheld the Division 105A scheme in its entirety, referred repeatedly to the undesirability of using Chapter III too readily to interfere with legislative schemes that struck at individual liberty.^[279] Before reaching his Honour’s judgment, however, it is necessary to note Gageler J’s and Gordon J’s reasoning.

In some cases, including Benbrika, these two judges have taken a broader, more ‘functionalist’^[280] approach to Chapter III: that is, an approach that identifies and gives effect to a constitutional value underlying the separation of powers. That value is individual liberty.^[281] So, in Benbrika, Gageler J announced that we can only understand the Lim principle if we first recognise why it is that the adjudgment and punishment of criminal guilt is an exclusively judicial function.^[282] By granting judges the exclusive power to determine and then punish criminal guilt, his Honour observed, Chapter III protects the individual from arbitrary executive punishment.^[283] Liberty is further protected, he continued, by the constitutional requirement that the judiciary order punitive detention only where the state has first proved that the person contravened ‘positive law’ in the past.^[284] In other words, it would be affront to individual liberty if persons were to be punished for conduct that they had not yet performed. Yet his Honour then in substance held that Chapter III allows for precisely that.

Unlike the plurality,^[285] Gageler J – with whom Gordon J agreed on this point^[286] for essentially the same reasons^[287] – held that Division 105A was invalid insofar as it allowed judges to make continuing detention orders where such orders were not necessary to ‘prevent’^[288] an unacceptable risk of the detainee’s committing, or supporting or facilitating, a terrorist act.^[289] ‘[L]iberty would be subverted’, his Honour thought, if a person could be detained because of the risk that s/he would commit a mere ‘criminal offence’^[290] – as opposed to causing or facilitating a ‘serious harm’.^[291] But his Honour accepted – as Gordon J possibly would^[292] – that continuing detention orders imposed to prevent serious harm are non-punitive and are within ‘the judicial power of the Commonwealth’.^[293] Unlike the plurality, Gageler J noted that whether Benbrika was being punished turned on ‘whether Div 105A is reasonably capable of being seen as necessary for a legitimate non-punitive objective’.^[294] But he did not explain how prison can reasonably be seen as necessary to prevent an unacceptable risk of serious harm, or its support or facilitation.

We have, then, five justices in Benbrika using extraordinarily narrow and formalistic reasoning to uphold a Commonwealth post-sentence preventive detention scheme – either in its entirety or with some modifications. Is such an approach defensible? Alternatively, should these justices have given fuller effect to the constitutional value of liberty^[295] that Gageler and Gordon JJ consider to underlie Chapter III? Justice Edelman’s judgment in Benbrika helps us to answer these questions.

Justice Edelman stated, correctly it is submitted, that ‘transparency and constitutional fidelity require that the true character of a continuing detention order ... be recognised’.^[296] In other words, the detention that it sanctions is punishment and the law should acknowledge this.^[297] Strangely, when justifying this conclusion,^[298] Edelman J never stated that such detention cannot reasonably be regarded as necessary to achieve the non-punitive objective of protecting the community from harm. He made no reference to this established test.^[299] Further, although his Honour rightly stated that the plurality justices had made a ‘category error’ when they ‘reason[ed] that Div 105A is not punitive because it aims to protect the community by preventing the commission of offences’,^[300] he could perhaps have explained more clearly than he did why mental illness detention in an appropriate facility bears a different, non-punitive, character from the detention in Benbrika.^[301] The reason seems to be that such detention, unlike ‘traditional criminal punishment’^[302] and preventive detention in gaol, can reasonably be regarded as necessary to achieve the non-punitive purpose of protecting the community from harm. That said, however, Edelman J was right not to hide behind formalism.^[303] We shall return to this point.

Justice Edelman seems to have thought that Gummow J’s attempt in Fardon to have the Court reformulate the Lim principle amounted also to an attempt to ‘expand’ that principle’s scope.^[304] Consistently with what is argued above,^[305] such a view seems wrong. Further, Edelman J rejected the idea that the constitutional value of liberty should feature as prominently as Gummow, Gageler and Gordon JJ thought it should, when giving effect to Chapter III’s requirements. ‘[C]onstitutional implications to protect liberty’, he urged, ‘must be based upon the text and structure of the Constitution’.^[306] And ‘[t]here is’, he found, ‘insufficient constitutional foundation to expand the Lim principle from one which is concerned with the separation of powers to one which is also founded upon the liberty of the individual’.^[307]

In other words, contrary to what Brennan, Deane and Dawson JJ seemed clearly to say in Lim,^[308] according to Edelman J the judiciary – though not the other branches of government^[309] – may order punitive detention for what a person might do, under a Commonwealth statute. So long as the power is judicial in form^[310] – which the one at issue was, though it created new rights rather than determining existing rights or obligations^[311] – and so long as the power is also ‘exercised judicially’,^[312] it will, his Honour held, fall within the ‘judicial power of the Commonwealth’. Edelman J held that the latter requirement would only not be satisfied if the detention was: (i) never necessary (that is, if a less restrictive measure than detention would always sufficiently achieve the legislative purpose);^[313] or (ii) imposed for ‘slight or trivial’ reasons (say, presumably, to prevent the commission of minor property offences).^[314] In the case of Division 105A, Edelman J concluded that (i) did not apply: a court could only order detention once satisfied that a control order or like measure would not have the necessary preventive effect.^[315] Nor did (ii). Like the plurality, but unlike Gageler and Gordon JJ, Edelman J declined to find the impugned scheme invalid insofar it allowed for detention to prevent prophylactic offending. And, importantly, he justified this refusal with reasoning that drew upon Gleeson CJ’s call in Fardon for judicial restraint. Justice Edelman said that a court should only in an ‘extreme case’ strike down a punitive preventive detention law because the offences that it aimed to prevent were insufficiently serious.^[316] That is because

the very integrity and impartiality of the courts ... would be seriously impaired if the judiciary could generally refuse to implement statutory provisions on the grounds of an objection to legislative policy. ^[317]

Like the plurality,^[318] Edelman J held that Parliament was entitled to find that all of the terrorist-related offences to which Division 105A applied were serious enough to necessitate preventive detention. However remote some of the prohibited conduct was from actual criminal harm, these offences all could be ‘aimed at the very destruction of civilised society’.^[319]

IV WHAT DO WE MAKE OF BENBRIKA?

A Summary of the Argument in this Part

We can now return to the questions posed above, which I said Edelman J’s judgment in Benbrika assists us to answer. To what extent should the HCA use Chapter III to defend liberty and uphold the rights of people such as Benbrika? Is the Court ever justified in using formalistic and highly unpersuasive reasoning to avoid reading Chapter III in such a way? It is submitted that the answer to this second question is ‘no’. It is never justifiable for a court – knowingly, recklessly or otherwise^[320] – to use such reasoning to avoid applying a law that, in fact, requires it to defend individual liberty. The answer to the first question is slightly less definite. I shall argue below that, though the legislation that he upheld in Benbrika is deplorable, Edelman J’s general approach in that case was correct. In other words, in cases, such as Benbrika, where (i) Chapter III does not clearly require a liberty promoting outcome, and (ii) there is a real risk that any such outcome will cause the Court to be seen as ‘activist’, their Honours should be ‘wary’^[321] about reaching such a result. That said, while Edelman J’s conclusion in Benbrika – that the impugned scheme was valid – does seem supportable, there is force in Gageler and Gordon JJ’s insistence that the majority should have given greater weight than it did to the constitutional value of liberty. The idea that Chapter III allows the judiciary to order punitive preventive detention is one thing. Such detention is undesirable and contrary to human rights, yet we cannot expect Chapter III to protect rights in the same way a human rights charter might. The idea that Chapter III allows the judiciary to order punitive preventive detention to prevent prophylactic offences, on the other hand, does seem even harder to square with the hostility to arbitrary detention that animates well established Chapter III principles.^[322]

B Dubious, Formalistic Judicial Reasoning and Its (Im)permissibility

In a number of publications,^[323] Goldsworthy has considered ‘one of the most important but neglected questions in legal philosophy ... under what circumstances are judges morally justified in covertly changing the law, when they have no legal authority to do so, in order (by their lights) to improve it?’.^[324] His focus is on decisions, including Kable, that produce rather more liberal conclusions than those for which the relevant written law seemed to provide; and his response – ‘in exceptional and extreme circumstances’^[325] – is slightly more restrictive than the one that I would give.^[326] That said, I agree with Goldsworthy that, if the Kable majority knowingly deployed implausible reasoning to protect Kable’s human rights,^[327] their Honours were wrong to do so. It is undemocratic for judges to refuse to apply clear law so as to improve human rights protection, unless, in my view, they can be confident that the community would support or be indifferent to their doing so.^[328] Indeed, even if their Honours unknowingly deployed such reasoning, they were still wrong. Such a lack of rigour is not admirable. And, again, in Kable, it seemed to produce an undemocratic outcome (the wrongful striking down of legislation that, however draconian it was, seemed to enjoy much public support).

If we return to Benbrika, the question posed above about formalistic reasoning resembles the one that Goldsworthy considers. It differs only in that it concerns the judicial use of implausible reasoning to avoid reaching liberal, human rights promoting decisions, even though the law clearly provides for them. Are judges ever justified in doing that? They are not. For, when they act in this way, they both override the clear requirements of positive law and ignore their duty, under that law, to protect the individual from state power. In other words, while, unlike Goldsworthy, I believe that judges may engage in subterfuge where they can be confident that the result thus produced will accord with community values,^[329] there is a qualification. They may not use dubious reasoning, knowingly or unknowingly, where this produces a result that clearly breaches human rights. Lying to facilitate a human rights breach is dishonourable. Methodological sloppiness that leads to the same result is not something that we should applaud.

The HCA’s decision in Magaming seems a clear example of the phenomenon just noted (though it is unclear whether the majority knew of the flimsiness of the reasoning it employed). As has been noted in this article more than once, ‘adjudging and punishing criminal guilt is an exclusively judicial function’.^[330] So, surely Chapter III has been breached if: (i) Parliament creates a mandatory or a mandatory minimum penalty for an offence; or (ii) the executive charges a person with an offence carrying such a penalty, instead of another offence, covering the same conduct, to which no mandatory penalty applies?^[331] Surely, as Jordan CJ thought, it has ‘dictate[d] to a Court of Justice that at least a certain penalty shall be imposed in the event of conviction’?^[332] In Magaming, six justices denied this. In so doing, as Gageler J noted in dissent, their Honours ‘elevate[d] ... form over substance’^[333] and allowed the Commonwealth Parliament to ‘by-pass the structural requirement of Ch III that punishment of crime occur only as a result of adjudication by a court’.^[334] It is true that, where a mandatory penalty applies, the judge in fact imposes the sentence.^[335] But there has still been an impermissible interference with, or usurpation of, judicial power. That is because, to use Kirby J’s words in Gypsy Jokers Motorcycle Club Inc v Commissioner of Police, ‘the hand that directs the process’^[336] is that of one of the non-judicial branches of government.^[337]

Crump, Knight and Minogue provide us with another example of the HCA’s use of formalistic, unpersuasive reasoning when dismissing Chapter III challenges to laws that clearly breached prisoners’ human rights. In each case, a judge had sentenced the appellant to life imprisonment for murder, but had set a non-parole period – 30 years for Crump,^[338] 28 years for Minogue^[339] and 27 years for Knight.^[340] In each case, near the expiry of the appellant’s non-parole period, a state parliament had passed legislation, the ‘practical effect’ of which was to ‘subject ... [him] ... to ... life [imprisonment] without meaningful prospect of parole’.^[341] So, for example, section 74AB of the Corrections Act 1986 (Vic) provides that the Victorian Adult Parole Board may only grant parole to ‘the prisoner Craig Minogue’^[342] if it is satisfied, ‘amongst other things’, that he ‘is in imminent danger of dying or is seriously incapacitated and that, as a result, he no longer has the physical ability to do harm to any person’.^[343] As some commentators have noted,^[344] such laws resemble the law struck down on Kable grounds by the QCA in Attorney-General v Lawrence.^[345] That is, like that law, the impugned legislation in Crump, Knight and Minogue seemed to reverse judicial orders in particular cases.^[346] Accordingly, if the Lawrence law ‘damaged both the appearance and reality of the “decisional independence” of the Supreme Court’,^[347] contrary to Kable,^[348] how could it seriously be said that the Crump, Knight and Minogue laws did not? The HCA’s response to this question – that, in fact, the impugned laws did ‘nothing to contradict the minimum term[s]’^[349] that the respective sentencing judges had fixed, and therefore did not impose on the relevant prisoners a more severe sentence than had been judicially imposed^[350] – is impossible to accept. As I have argued elsewhere,^[351] parliament went well beyond merely altering the conditions that Crump, Minogue and Knight had to satisfy if they were to be released on parole.^[352] It instead substituted for the life with parole sentences that their sentencing judges had imposed on them, the irreducible life sentences^[353] that their Honours had deliberately refrained^[354] from ordering.

This returns us to the reasoning of five of the justices who decided Benbrika. The plurality’s and Gageler J’s contention in that case that prison is not necessarily punishment, was reminiscent of the highly formalistic reasoning that the Court, or a majority, adopted in Magaming, Crump, Knight and Minogue. If this had been the only way of rejecting (or, in the case of Gageler J, partly rejecting) the Chapter III challenge to Division 105A, then such a result would have been unjustified. For, again, if the law clearly requires judges to reach a conclusion, they may not reach another conclusion that would obviously breach human rights. However unpopular are decisions that uphold the prisoners’ rights, and however understandable is the HCA’s desire to maintain its reputation for impartiality, it does have some ‘margin’ for reaching ‘counter-majoritarian decisions’ that will be respected by ‘political actors’.^[355] In those circumstances in particular, the Court in Magaming, Crump, Knight and Minogue did not do what it should have done to counter ‘the oppressive actions of ill-advised majorities’.^[356] Instead of accepting its responsibility to ensure that Parliament observes the limitations imposed by Chapter III, it allowed the elected branches to sentence individuals themselves and therefore step well beyond the relevant boundaries.

Can the result in Benbrika be criticised on similar grounds? It is submitted that Edelman J’s judgment in that case shows that it cannot. That is, his Honour demonstrated that it was possible to uphold the impugned scheme without ‘elevat[ing] form over substance’.^[357]

C Why the Decision in Benbrika Is Defensible

Leading commentators have noted that, as Stellios has put it, the Convention Debates ‘do not reveal all that much’^[358] about which precise functions the framers considered to fall within ‘the judicial power of the Commonwealth’. Further, as Edelman J suggested in Benbrika, neither the text nor the structure of the Constitution provides any specific guidance concerning whether Tinney J was exercising such power when he imposed Benbrika’s continuing detention order on him.^[359] In such circumstances – and considering also the ‘long history’ of ‘preventative detention regimes attached by legislation to the curial sentencing process’^[360] – Edelman J was entitled to hold, as he did, that the Commonwealth parliament will not breach Chapter III by authorising a judge to order the preventive detention of a person who is currently serving a sentence of imprisonment. As argued above,^[361] if a judge may validly order indefinite detention, it would seem strange to hold that s/he may not validly order post-sentence preventive detention.

Now, the opposite conclusion was also available to his Honour. That is, he could have held that, without clear textual or historical support for either party’s position, the Commonwealth Parliament may not validly authorise a Chapter III court to order either punitive indefinite detention or post-sentence preventive detention. Such reasoning would have avoided treating these two forms of detention as somehow relevantly differing from one another (contrary to Gummow and Kirby JJ’s approach in Fardon).^[362] It would also have upheld the Lim principle, which, as has been noted, holds that the judiciary may only order punitive detention as ‘an incident of ... adjudging and punishing criminal guilt’.^[363] And it would have given effect to a plausible conception of the constitutional value of liberty, one focussed on limiting state power over the individual^[364] by, among other things, protecting him/her from arbitrary detention.^[365] Such a ruling would not necessarily have altogether prohibited preventive detention. It might simply have required Parliament to ensure that such detention was served in conditions that were as non-punitive as possible. Detention of that nature, it could plausibly be argued, is analogous to the ‘exceptional case’ of mental illness detention,^[366] so as to mean that it can be ordered either by the Commonwealth executive or by a Chapter III court. Both forms of detention, it could be said, have the shared aim of protecting society from the dangerous.^[367] Nevertheless, there would seem to be a number of difficulties with such an approach.

First, there is some tension between it and the historical use of punitive indefinite sentencing regimes. That said, it could be argued that, when certain judges assumed that such regimes complied with Chapter III’s requirements, their Honours’ analyses were invalid – based as they were on the erroneous view that the relevant exercise of power was addressed to the accused’s past guilt.^[368] Secondly, and more importantly, however, this approach would have had far-reaching practical consequences; and – partly for this reason – would presumably have been unpopular with the public. Assuming that the Court had held that preventive detention served in non-punitive conditions was constitutionally permissible, it would effectively have been instructing the Commonwealth Government to construct new facilities in which to detain dangerous terrorist offenders if it wanted to maintain a detention scheme for such persons.^[369] It is hard to believe that the press and public would have responded with equanimity to such an instruction. And would it all have been worth it? Such a decision would, to an extent, have had a rights-protective effect. It would have ensured that anyone in Commonwealth preventive detention served that detention in non-punitive conditions. But it would also have allowed the executive to order such detention, because, after all, Lim’s effect is that, so long as detention is non-punitive, any branch of government can order it.^[370]

The Commonwealth Government could of course instead have dispensed with the Division 105A scheme altogether, relying simply on control orders to deal with the problem of unreformed terrorist offenders. But, again, if the Court had caused such an outcome, would it have avoided accusations of ‘judicial activism’?

These comments about ‘judicial activism’ return us to Gleeson CJ’s^[371] and Edelman J’s^[372] evident view that the Court should generally not read Chapter III in such a way as to create the perception that it is willing to strike down legislation simply because of its distaste for it. Consistently with what I have argued elsewhere,^[373] there is much to be said for this approach. Again, where Chapter III clearly provides for a human rights-protective outcome in a particular case, the Court should reach that outcome – no matter how unpopular it is. But the position is different where the law is unclear. In such circumstances, the judiciary is understandably reluctant to decide cases in a way that will clash with ‘the values of ... society’.^[374] This is not merely because of a pragmatic concern to ensure that judicial decisions are respected, so that stable government is maintained.^[375] It is also for reasons of principle. In a democracy, and without a charter of rights,^[376] judicial exercises of public power should generally be consistent with public opinion.^[377] Justice Gordon has recently observed that

The making of good law in Australia is not assumed to depend to any large extent on the activity of the courts. None of this is to say that courts simply defer to all judgments of the political branches in all matters. That is clearly not the case. The Constitution establishes a system in which the High Court has the role of ensuring that legislatures remain within the bounds of their constitutional powers ... The point, rather, is that the constitutional system we have – with representative and responsible government at its heart – is one that assumes that most problems will be dealt with in a responsible manner by Parliament, and not by the courts. ^[378]

Consistently with such an approach, Edelman J seems right in Benbrika to have exercised restraint. As just stated, that is especially so, given that, if his Honour had instead held that the Commonwealth Parliament may only validly authorise a Chapter III court to order truly non-punitive preventive detention, this would not have had an entirely rights-protective effect.^[379]

This is not to argue, however, that the constitutional value of liberty should have had no effect on the Court’s decision. As we have seen, Edelman J stated that the Lim principle^[380] must be concerned ‘with the separation of powers’ not ‘the liberty of the individual’.^[381] This might reflect a view that, while Chapter III is clearly intended to create an independent and impartial judiciary,^[382] it is not so clear that, by providing for such a judiciary, Chapter III’s ultimate object is to place major limitations on ‘the exercise of state power against an individual’s liberty interest’.^[383] Chapter III might instead largely be concerned to ensure that ‘government functions ... [are performed by] decision-makers who are best equipped to perform the task’ (though of course that might lead to incidental liberty protections).^[384] If we come back to the precise exercise of power with which we are concerned, Edelman J might have been saying that Chapter III merely requires that a court order punitive detention, because a court is more likely than the other arms of government to exercise such power ‘fairly’ and ‘impartia[lly]’.^[385] In other words, it is best equipped to perform this task.

Nevertheless, a concern to protect liberty more directly was not entirely absent from Edelman J’s analysis. As we have seen, his Honour did insist that even a judge may not validly make a punitive continuing detention order for ‘slight or trivial’ reasons.^[386] Nor, according to his Honour, could the Commonwealth Parliament validly empower a Chapter III court to order such detention where Parliament’s protective aim could always be achieved by a ‘less restrictive’ alternative.^[387] Such an approach has much to recommend it. Indeed, more than that, Gageler and Gordon JJ seem right to have taken this analysis one stage further. Their Honours, that is, seem right insofar as they held that Division 105A would be constitutionally valid only if it restricted continuing detention orders to those terrorist offenders who pose a risk of perpetrating, or supporting or facilitating the perpetration of, a grave harm if released.^[388]

There are a number of reasons why: (i) Gageler, Gordon and Edelman JJ appear right to have attached weight to the constitutional value of liberty when deciding what limits Chapter III places on the judiciary’s ability validly to make continuing detention orders; and (ii) the former two justices’ approach seems preferable to Edelman J’s. First, while it seems clear that Chapter III is not intended to protect individual rights as a charter of rights can,^[389] it is plausible to argue that its protection of judicial independence and impartiality is partly aimed at safeguarding individual liberty (and not merely at distributing governmental functions).^[390] Secondly, and relatedly, Gageler J was right to argue that his approach was consistent with well-established Chapter III principles.^[391] Once it is accepted that the punishment of criminal guilt is an exclusively judicial function,^[392] and that the state may only exceptionally detain an individual for some other reason,^[393] there is force in the view that Chapter III allows the judiciary only in very limited circumstances to order an individual’s detention because of what s/he might do. Such a rule seems consistent with the hostility to arbitrary detention that underlies the principles just stated.^[394] Thirdly, the Court could seemingly have insisted on a limitation of the type contemplated by Gageler and Gordon JJ, without giving rise to claims of ‘judicial activism’. Such a ruling would not have required the Commonwealth Government to build new, non-punitive preventive detention facilities. It would not have prevented the state from imposing continuing detention orders on terrorist offenders who remain dangerous. It would simply have required parliament to make some adjustments to the Division 105A scheme. Finally, though this is far from a determinative consideration, such a rule is normatively desirable. As suggested above,^[395] if the state imprisons individuals simply because they pose a threat of ‘advanc[ing] ... terrorist ideology’,^[396] it breaches their human rights.

V CONCLUSION

In this article, I have scrutinised the reasoning of the various justices who decided Benbrika. As noted at the outset, I have done so with the following questions in mind. Did the HCA, in that case – as in others where prisoners have challenged punitive legislation on the basis of Chapter III of the Commonwealth Constitution – fail to accept its responsibility to hold power to account? Is this another instance of their Honours failing to provide unpopular litigants with the protections for which Chapter III in fact clearly provides? Or does the Court’s decision in Benbrika instead properly recognise that, while Chapter III to an extent has a ‘counter-majoritarian’ focus^[397] – and the capacity to ‘protect ... individuals and minorities’^[398] – it is nevertheless important not to overstate its capacity to protect human rights?^[399] And would the Court in fact have been misusing its own powers if it had decided Benbrika differently from how it did?

I have concluded that the Benbrika decision is a largely defensible one, but that some of the reasoning that produced that decision is indefensible.

There is no doubt that it is always morally impermissible for the state to imprison an individual because of fears about what that person might do if s/he were to be released. There is also no doubt that, by authorising Benbrika’s imprisonment, the Commonwealth Government has placed itself in breach of its international human rights obligations. I have argued here that, however popular such legislation is, and however unpopular a judicial decision to strike it down would have been, that is what the HCA should have done if the law had clearly mandated such an outcome. Judges should never use formalistic and dubious reasoning to uphold a law that clearly breaches the human rights of those whom it targets. In other words, the popularity of decisions such as Crump, Knight, Magaming and Minogue make those decisions no more acceptable – and that is so whether or not the judges who decided those cases knew of the dubiousness of the reasoning that they were deploying.

It follows that, if the Court in Benbrika had only been able to uphold Division 105A by deploying reasoning of the kind deployed by five justices – namely, that Benbrika’s imprisonment is not punishment – such a decision would have been no more justified than the decisions just noted. But Edelman J showed that that was not the only way in which their Honours could reject Benbrika’s Chapter III challenge – and I have argued here that his Honour’s approach was generally correct. Because Chapter III does not clearly prevent the Commonwealth parliament from authorising a Chapter III court to order punitive preventive detention – and because, without a charter of rights, the judicial law-making function must largely respect public opinion – Edelman J seems right overall to have exercised the restraint that he did. When judges act in such a way, they are not failing to do what they should to hold power to account. They are instead observing the proper limits of their own powers.

---

* Senior Lecturer, University of Sydney Law School. Director, Sydney Institute of Criminology.

¹ Andrew Dyer, ‘Irreducible Life Sentences: What Difference Have the European Convention on Human Rights and the United Kingdom Human Rights Act Made?’ (2016) 16(3) Human Rights Law Review 541; Andrew Dyer, ‘(Grossly) Disproportionate Sentences: Can Charters of Rights Make a Difference?’ [2017] MonashULawRw 6; (2017) 43(1) Monash University Law Review 195 (‘(Grossly) Disproportionate Sentences’); Andrew Dyer, ‘Can Charters of Rights Limit Penal Populism?: The Case of Preventive Detention’ [2018] MonashULawRw 18; (2018) 44(3) Monash University Law Review 520; Andrew Dyer, ‘Irreducible Life Sentences, Craig Minogue and the Capacity of Human Rights Charters to Make a Difference’ (2020) 43(2) University of New South Wales Law Journal 484; Andrew Ronald Dyer, Can Charters of Rights Limit Penal Populism? Irreducible Life Sentences, Disproportionate Sentences and Preventive Detention in Australia and under the Human Rights Act 1998 (UK) & the European Convention on Human Rights (PhD Thesis, University of Sydney, 2021) (‘Can Charters of Rights Limit Penal Populism?’).

^[2] For some leading analyses of ‘penal populism’, see David Garland, The Culture of Control: Crime and Social Order in Contemporary Society (Oxford University Press, 2001) 131–7; Julian V Roberts et al, Penal Populism and Public Opinion: Lessons from Five Countries (Oxford University Press, 2003); John Pratt, Penal Populism (Routledge, 2007); Nicola Lacey, The Prisoners’ Dilemma: Political Economy and Punishment in Contemporary Democracies (Cambridge University Press, 2008).

^[3] Lord Mance, ‘Destruction or Metamorphosis of the Legal Order?’ (Speech, World Policy Conference, 14 December 2013).

^[4] See, eg, Human Rights Act 1998 (UK) s 4.

^[5] Charter of Human Rights and Responsibilities Act 2006 (Vic).

^[6] Dyer, ‘Irreducible Life Sentences, Craig Minogue and the Capacity of Human Rights Charters to Make a Difference’ (n 1) 515.

^[7] Corrections Act 1986 (Vic) ss 74AA, 74AB.

^[8] See Minogue v Victoria [2019] HCA 31; (2019) 268 CLR 1, 20 [30] (Gageler J), 23 [40] (Edelman J) (‘Minogue’). Cf 15–18 [13]–[21] (Kiefel CJ, Bell, Keane, Nettle and Gordon JJ); Knight v Victoria [2017] HCA 29; (2017) 261 CLR 306, 323–4 [29] (‘Knight’).

^[9] Corrections Act 1986 (Vic) ss 74AA(4), 74AB(4). See also Charter of Human Rights and Responsibilities Act 2006 (Vic) s 31(1). Nor will ‘strong-form’ human rights charters (see Mark Tushnet, ‘New Forms of Judicial Review and the Persistence of Rights- and Democracy-Based Worries’ (2003) 38(2) Wake Forest Law Review 813, 814) – ie, charters that provide judges with the ‘final word’ about whether primary legislation breaches charter rights (see Stephen Gardbaum, ‘The New Commonwealth Model of Constitutionalism’ (2001) 49(4) The American Journal of Comparative Law 707, 709) – necessarily cause prisoners’ rights to be better protected than they would otherwise have been: see especially Dyer, Can Charters of Rights Limit Penal Populism? (n 1) chs 1, 3, 5.

^[10] That is, in cases where prisoners have challenged penal populist laws on the basis of Chapter III. See, eg, Fardon v A-G (Qld) [2004] HCA 46; (2004) 223 CLR 575 (‘Fardon’); Baker v The Queen [2004] HCA 45; (2004) 223 CLR 513 (‘Baker’); Crump v State of New South Wales [2012] HCA 20; (2012) 247 CLR 1 (‘Crump’); Magaming v The Queen [2013] HCA 40; (2013) 252 CLR 381 (‘Magaming’). Cf Kable v Director of Public Prosecutions (NSW) [1996] HCA 24; (1996) 189 CLR 51 (‘Kable’).

^[11] Knight [2017] HCA 29; (2017) 261 CLR 306; Minogue [2019] HCA 31; (2019) 268 CLR 1.

^[12] Crump [2012] HCA 20; (2012) 247 CLR 1; Magaming [2013] HCA 40; (2013) 252 CLR 381.

^[13] I am grateful to one of the three anonymous referees for essentially asking what is wrong with formalism. After all, on one view, as Dixon has noted, the terms ‘formalist’ and ‘formalistic’ mean simply ‘legalist’ or ‘legalistic’: Rosalind Dixon, ‘The High Court and Dual Citizenship: Zines and Constitutional Method 30 Years On’ in John Griffiths and James Stellios (eds), Current Issues in Australian Constitutional Law: Tributes to Professor Leslie Zines (Federation Press, 2020) 135, 139 (‘The High Court and Dual Citizenship’). And, subject to the argument in Part IV of this article, there is much to be said for the legalist emphasis on ‘the text, history and structure of the Constitution, and prior case law’ – especially where these sources provide ‘specific or concrete guidance’ about an issue of constitutional interpretation: at 139 (and note also Dixon’s suggestion that ‘reliance by judges on ethical or political values’ when interpreting the Constitution should occur only after they have engaged seriously with the sources just noted: at 141). As Dixon indicates, however, the better view is that ‘there are important differences’ between formalism and legalism – differences that make the former open to far more fundamental criticisms than the latter: at 139, 140. Indeed, the reasoning in Minogue, Knight and Crump goes some way to illustrating the point. See below nn 338–54 and accompanying text. Contrary to the decisions in those cases, the relevant laws did alter the plaintiffs’ respective sentences and make them more severe. For while, in form, the impugned laws left the prisoners’ non-parole periods intact, in substance those laws transformed the relevant sentences from life with parole sentences into irreducible life sentences. As suggested below, Crump, Knight and Minogue would have been much more legalistic decisions if their Honours had: (a) acknowledged the obvious fact that the impugned legislation reversed judicial orders; and (b) then struck that legislation down on the basis of the well-established Kable principle that a state or territory law is constitutionally infirm if it damages the appearance or reality of the decisional independence of a court capable of exercising ‘the judicial power of the Commonwealth’: Commonwealth Constitution s 71.

^[14] Minister for Home Affairs v Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185] (Edelman J) (‘Benbrika’).

^[15] This is a convenient place for me to address a point made by the anonymous referee mentioned at n 13. For that referee, the role of the High Court of Australia (‘HCA’) might be to uphold the law without regard to policy. On such a conception of the Court’s role, the referee continues, it is neither here nor there whether legislation breaches human rights or protects the community: the sole question, when the Court deals with a person’s Chapter III challenge to such legislation, is whether it exceeds constitutional limits. As indicated at n 13 and in Part IV of this article, I accept that sometimes it is clear that legislation breaches (or does not breach) Chapter III. Where that is so, then, as I argue in Part IV, the Court should generally decide the case

accordingly. But where, as in Benbrika, the law is unclear, it is inevitable that judges will have regard to

‘ethical values and practical consequences’ (including whether its decision will be accepted by the

community), when deciding whether impugned legislation is constitutionally valid: Dixon, ‘The High Court and Dual Citizenship’ (n 13) 140.

^[16] [2021] HCA 4; (2021) 388 ALR 1.

^[17] Ibid 15–16 [38]–[41] (Kiefel CJ, Bell, Keane and Steward JJ).

^[18] Ibid 29 [87]–[88]. See also Pollentine v Bleijie [2014] HCA 30; (2014) 253 CLR 629, 657 [73] (Gageler J) (‘Pollentine’).

^[19] See Benbrika [2021] HCA 4; (2021) 388 ALR 1, 48 [161]–[162], 52 [177]–[178].

^[20] Muldrock v The Queen [2011] HCA 39; (2011) 244 CLR 120, 140 [57].

^[21] See, eg, Patrick Keyzer, ‘Preserving Due Process or Warehousing the Undesirables: To What End the Separation of Judicial Power of the Commonwealth?’ [2008] SydLawRw 5; (2008) 30(1) Sydney Law Review 101, 108–9; Bernadette McSherry, ‘Indefinite and Preventive Detention Legislation: From Caution to an Open Door’ (2005) 29 Criminal Law Journal 94, 109–10.

^[22] [2012] HCA 20; (2012) 247 CLR 1.

^[23] [2017] HCA 29; (2017) 261 CLR 306.

^[24] [2019] HCA 31; (2019) 268 CLR 1.

^[25] [2013] HCA 40; (2013) 252 CLR 381.

^[26] Jeffrey Goldsworthy, ‘Tom Campbell on Judicial Activism’ (2017) 42(1) Australian Journal of Legal

Philosophy 247, 250.

^[27] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 53–4 [182]–[184], 59–61 [200]–[204], 74 [239].

^[28] Ibid 54 [185], 69 [226].

^[29] Commonwealth Constitution s 71.

^[30] Fardon [2004] HCA 46; (2004) 223 CLR 575, 613 [83] (Gummow J).

^[31] Ibid 608–614 [69]–[85].

^[32] Ibid 631 [145].

^[33] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 13–14 [31]–[34]; cf 15 [38] (Kiefel CJ, Bell, Keane and Steward JJ).

^[34] Fardon [2004] HCA 46; (2004) 223 CLR 575, 609–610 [70]–[72], 613 [83] (Gummow J), 637–8 [165] (Kirby J).

^[35] Ibid 609–610 [70]–[73] (Gummow J), 637–8 [165] (Kirby J).

^[36] See, eg, Falzon v Minister for Immigration and Border Protection [2018] HCA 2; (2018) 262 CLR 333, 340 [14] (Kiefel CJ, Bell, Keane and Edelman JJ) (‘Falzon’); Duncan v New South Wales (2015) 255 CLR 388, 407 [41]; Magaming [2013] HCA 40; (2013) 252 CLR 381, 396 [47] (French CJ, Hayne, Crennan, Kiefel and Bell JJ); Chu Kheng Lim v Minister for Immigration, Local Government and Ethnic Affairs (1992) 176 CLR 1, 27 (Brennan, Deane and Dawson JJ) (‘Lim’).

^[37] Fardon [2004] HCA 46; (2004) 223 CLR 575, 637 [165] (Kirby J).

^[38] Ibid 613 [84] (Gummow J).

^[39] Ibid 609–610 [70]–[73], 613 [83] (Gummow J), 637–8 [165] (Kirby J). Many other appellate judges have made the same assumption: see, eg, Kable [1996] HCA 24; (1996) 189 CLR 51, 97 (Toohey J); A-G (Qld) v Fardon [2003] QCA 416, [78]–[79] (McMurdo P) (‘Fardon QCA’). See also R v Moffatt [1998] 2 VR 229, 251–2 (Hayne JA).

^[40] R v Quinn; Ex parte Consolidated Foods Corporation [1977] HCA 62; (1977) 138 CLR 1, 11 (Jacobs J).

^[41] James Stellios, The Federal Judicature: Chapter III of the Constitution (LexisNexis Butterworths, 2010) 71 [3.43] (‘The Federal Judicature’).

^[42] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 65 [217] (Edelman J).

^[43] See Human Rights Committee, Views: Communication No 1629/2007, 98^th sess, UN Doc CCPR/C/98/D/1629/2007 (18 March 2010) 8 [7.4] (‘Fardon v Australia’); Human Rights Committee, Views: Communication No 1635/2007, 98^th sess, UN Doc CCPR/C/98/D/1635/2007 (18 March 2010) 10 [7.4] (‘Tillman v Australia’).

^[44] See Fardon [2004] HCA 46; (2004) 223 CLR 575, 586 [3], 590 [14] (Gleeson CJ), 601 [41] (McHugh J).

^[45] Criminal Code Act 1995 (Cth) s 105A.7.

^[46] Criminal Code Amendment (High Risk Terrorist Offenders) Act 2016 (Cth).

^[47] Criminal Code Act 1995 (Cth) s 105A.1.

^[48] In this article, I refer to preventive detention ordered at the time of sentencing as ‘indefinite detention’. A classic example of indefinite detention is provided by Penalties and Sentences Act 1992 (Qld) which allows a judge, when sentencing for certain offences, to impose an indeterminate sentence on particular offenders who are proved to be a ‘serious danger to the community’: s 163(3)(b). Upon the expiry of the period of detention that the judge would have imposed but for the finding of dangerousness, the offender will remain imprisoned for so long as s/he is proved to remain dangerous in the relevant sense: see ss 171–2, 173. The terminology that I use in this article – ‘post-sentence preventive detention’ and ‘indefinite detention’ – is used also by Bernadette McSherry, Patrick Keyzer and Arie Freiberg, Preventive Detention for ‘Dangerous’ Offenders in Australia: A Critical Analysis and Proposals for Policy Development (Report, December 2006) 9–11.

^[49] Criminal Code Act 1995 (Cth) s 105A.3(1).

^[50] Ibid s 105A.7(1)(b)–(c). See also s 105A.2 (definition of ‘serious Part 5.3 offence’).

^[51] Ibid s 105A.7(5).

^[52] Ibid s 105A.7(6).

^[53] Ibid s 105A.10. That said, in the unlikely event that the Attorney-General fails within the relevant 12-month period to apply for such a review, the order will immediately cease to be in force: ibid s 105A.10(4).

^[54] Ibid s 105A.11(1)–(2).

^[55] Ibid s 105A.12(4).

^[56] Ibid.

^[57] Ibid s 105A.12(5).

^[58] Ibid ss 105A.3(2), 105A.4(1).

^[59] Ibid s 105A.4(1).

^[60] Ibid s 105A.4(2)(b)–(c).

^[61] Ibid s 105A.4(2)(a).

^[62] Ibid s 105A.4(2). However, if the detainee ‘elects to be so accommodated or detained’, the authorities may accommodate his/her wishes: ibid s 105A.4(2)(d).

^[63] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 15 [39] (Kiefel CJ, Bell, Keane and Steward JJ). As Kirby J noted in Fardon [2004] HCA 46; (2004) 223 CLR 575, 640 [173], the same is true of the law at issue in that case – upon which Division 105A was modelled: Commonwealth, Parliamentary Debates, Senate, 15 September 2016, 1036 (George Brandis, Attorney-General).

^[64] Criminal Code Act 1995 (Cth) s 104.4(1)(c)(iv)–(v). In Thomas v Mowbray [2007] HCA 33; (2007) 233 CLR 307 (‘Thomas’), a majority of the HCA confirmed the constitutional validity of the control order scheme for which Division 104 of the Criminal Code provides: at 335 [32] (Gleeson CJ), 365 [154] (Gummow and Crennan JJ, Callinan J agreeing at 509 [600], Heydon J agreeing at 526 [652]).

^[65] See Criminal Code Act 1995 (Cth) s 104.5(3).

^[66] Commonwealth, Parliamentary Debates, House of Representatives, 1 December 2016, 5158 (‘Criminal Code Amendment Bill 2016 Second Reading Speech’).

^[67] Ibid.

^[68] Ibid 5159.

^[69] Ibid.

^[70] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 25 [74].

^[71] To like effect see, eg, Michael Tonry, ‘Predictions of Dangerousness in Sentencing: Déjà Vu All Over Again’ (2019) 48(1) Crime and Justice 439, 445.

^[72] Fardon v Australia (n 43) 8 [7.4]; Tillman v Australia (n 43) 10 [7.4].

^[73] Opened for signature 16 December 1966, 999 UNTS 171 (entered into force 23 March 1976) (‘ICCPR’).

^[74] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185] (Edelman J), quoting CS Lewis, ‘The Humanitarian Theory of Punishment’ (1953) 6 Res Judicate 224, 225.

^[75] See, eg, Foucha v Louisiana, [1992] USSC 54; 504 US 71, 83 (White J) (1992) (‘Foucha’).

^[76] See, eg, Stephen J Morse, ‘Neither Desert nor Disease’ (1999) 5(3) Legal Theory 265, 269–70; Stephen J Morse, ‘Fear of Danger, Flight from Culpability’ (1998) 4(1– 2) Psychology, Public Policy and Law 250, 258–9, 263–4; Stephen J Schulhofer, ‘Two Systems of Social Protection: Comments on the Civil-Criminal Distinction, with Particular Reference to Sexually Violent Predator Laws’ (1996) 7(1) Journal of Contemporary Legal Issues 69, 90–4. However, note that, like the United States Supreme Court (‘USSC’) (see below n 81), certain commentators have seemingly been more ready to state their belief in the relevant principle than they have been to adhere to that principle. Corrado, for example, argues that the rational actor must be subject to criminal sanctions alone – only those who are mentally ill and dangerous may be made

subject to civil commitment: see generally Michael Louis Corrado, ‘Sex Offenders, Unlawful Combatants, and Preventive Detention’ (2005) 84(1) North Carolina Law Review 77. He includes within the latter category those who ‘cannot control ... [their] behavior’: at 105. But then he is not consistent about this requirement – on other occasions, he defends civil commitment for those who find it ‘difficult to control their behavior’ or ‘have great difficulty in controlling their ... behavior’: at 105, 107 (emphasis added). And he supports (at 110) the decision in Kansas v Crane, [2002] USSC 10; 534 US 407 (2002) (‘Crane’), where the USSC held that the United States Constitution allows for detention where the government proves that the individual has a ‘serious mental illness, abnormality, or disorder’ that causes ‘serious difficulty in controlling behavior’: Crane, [2002] USSC 10; 534 US 407, 413 (2002). In short, while Corrado opposes detaining people simply because they are likely to commit violent crimes, Crane and the ‘mental abnormality’ requirement that Corrado supports, allows for precisely this: Corrado (n 76) 97; see below n 81.

^[77] Saul Smilansky, ‘The Time to Punish’ (1994) 54(1) Analysis 50, 52.

^[78] Ibid 52–3.

^[79] Morse, ‘Neither Desert nor Disease’ (n 76) 268.

^[80] Ibid.

^[81] Foucha, [1992] USSC 54; 504 US 71, 83 (White J) (1992). Note, however, that in Kansas v Hendricks, 521 US 344 (1997) the USSC, while purporting to apply the sacred principle that forbids the indefinite confinement of the merely dangerous, in fact subverted that principle. In upholding a Kansas statute that allows for preventive detention orders to be made against sexual offenders who have a ‘mental abnormality’ or ‘personality disorder’ that makes them dangerous, the Court presented itself as allowing for something akin to standard mental illness detention. ‘The Kansas Act’, Thomas J held, ‘is plainly of a kind with ... other civil commitment statutes’ that condition detention on non-responsible dangerousness: at 358. See also Crane, [2002] USSC 10; 534 US 407, 413 (Breyer J) (2002). The difficulty with this is that, as Steiker has observed, ‘virtually all of those who choose to commit criminal acts, especially [very serious ones] ... can be considered “abnormal”. And the range of

potential “disorders” is likewise extraordinarily broad ...’: Carol S Steiker, ‘Foreword: The Limits of the Preventive State’ (1998) 88(3) Journal of Criminal Law and Criminology 771, 786. In other words, a statute like the Kansas one catches many individuals besides those who, because of a mental illness, are not autonomous actors.

^[82] Lim (1992) 176 CLR 1, 27–8 (Brennan, Deane and Dawson JJ).

^[83] Ibid 27.

^[84] Corrado (n 76) 101.

^[85] See, eg, Vinter v United Kingdom [2013] III Eur Court HR 317, 344 [102]; R v Smith [1987] 1 SCR 1045, 1073 (Lamer J). I say ‘more or less’ because those courts have required the sentence to be grossly disproportionate if it is to amount to a breach of the offender’s human rights.

^[86] Criminal Code Amendment Bill 2016 Second Reading Speech (n 66) 5158 (Michael Sukkar).

^[87] [1987] USSC 94; 481 US 739, 755 (Rehnquist CJ) (1987) (‘Salerno’).

^[88] Criminal Code Amendment Bill 2016 Second Reading Speech (n 66) 5157 (Michael Sukkar).

^[89] [1950] USCA2 257; 184 F 2d 280, 282 (2^nd Cir, 1950), quoted in Salerno, [1987] USSC 94; 481 US 739, 766 (Marshall J) (1987).

^[90] Tonry (n 71) 450. See also, eg, Bernadette McSherry, Managing Fear: The Law and Ethics of Preventive

Detention and Risk Assessment (Routledge, 1^st ed, 2013) 34–52.

^[91] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185], 68 [224] (Edelman J).

^[92] Ibid 68 [222].

^[93] Andrew Ashworth and Lucia Zedner, Preventive Justice (Oxford University Press, 2014) 146.

^[94] Opened for signature 4 November 1950, 213 UNTS 221 (entered into force 3 September 1953) (‘ECHR’).

^[95] Mastromatteo v Italy [2002] VIII Eur Court HR 151, 165–6 [68]. I mention this jurisprudence, not because the ECHR applies in Australia – of course, it does not – but because this case law assists us to evaluate the normative argument that we are now considering: namely, whether preventive detention is ever permissible (or whether, alternatively, properly viewed, it is always contrary to human rights). I thank one of the anonymous referees for seeking clarification on this point.

^[96] See the discussion in Alibhe O’Loughlin, ‘Risk Reduction and Redemption: An Interpretive Account of the Right to Rehabilitation in the Jurisprudence of the European Court of Human Rights’ (2021) 41(2) Oxford Journal of Legal Studies 510, 533–7.

^[97] See, eg, Ashworth and Zedner, Preventive Justice (n 93) 168–9; Harry Hobbs and Andrew Trotter, ‘Lessons from History in Dealing with Our Most Dangerous’ [2018] UNSWLawJl 13; (2018) 41(2) University of New South Wales Law Journal 319, 351; CR Williams, ‘Psychopathy, Mental Illness and Preventive Detention: Issues Arising from the David Case’ [1990] MonashULawRw 10; (1990) 16(2) Monash University Law Review 161, 180–2.

^[98] Dyer, ‘Can Charters of Rights Limit Penal Populism?: The Case of Preventive Detention’ (n 1) 530.

^[99] Williams (n 97) 181.

^[100] (1988) 164 CLR 465.

^[101] Ibid 495.

^[102] Ibid (emphasis added).

^[103] Ibid.

^[104] Again, I mention this European Court of Human Rights (‘ECtHR’) jurisprudence, not because such case law has direct binding effect in Australia, but instead because it provides us with some guidance about the precise circumstances in which preventive detention complies with human rights/is morally permissible. In this regard, it can be noted that the EctHR’s approach is consistent with Ashworth and Zedner’s views about the conditions that preventive detention must satisfy if it is to be morally justified: Ashworth and Zedner, Preventive Justice (n 93) 169.

^[105] I have defined the terms ‘post-sentence preventive detention’ and ‘indefinite detention’ above: see above n 48 and the text accompanying above nn 48–9.

^[106] M v Germany [2009] VI Eur Court HR 169, 200–1 [86]–[89] (‘M’); Haidn v Germany (European Court of Human Rights, Chamber, Application No 6587/04, 13 January 2011) [73]–[76].

^[107] Article 5(1) provides that a person may only be deprived of his/her liberty in the narrow circumstances listed in that sub-article.

^[108] Article 7(1) relevantly provides that the state may not impose a heavier penalty on a person than was

applicable at the time of his/her criminal offending.

^[109] M [2009] VI Eur Court HR 169, 214 [127].

^[110] Van Droogenbroeck v Belgium [1982] ECHR 3; (1982) 4 EHRR 443, 461 [48].

^[111] [2012] ECHR 1706; (2013) 56 EHRR 12, [218] (‘James’). See also Brown v Parole Board for Scotland [2017] UKSC 69; [2018] AC 1, 18 [29], 33 [83] (Lord Reed JSC).

^[112] Fardon [2004] HCA 46; (2004) 223 CLR 575, 609–610 [70]–[72], 613 [83] (Gummow J), 637–8 [165] (Kirby J).

^[113] Ibid 608–14 [69]–[85] (Gummow J), 631 [145] (Kirby J).

^[114] Ibid 609–10 [70]–[73] (Gummow J), 637–8 [165] (Kirby J).

^[115] Dyer, ‘Can Charters of Rights Limit Penal Populism?: The Case of Preventive Detention’ (n 1) 525–6.

^[116] To use the formulation that appears in Penalties and Sentences Act 1992 (Qld) s 163(3)(b).

^[117] It also follows that Ashworth and Zedner seem to have been mistaken when they concluded recently –

consistently with the ECtHR’s decision in James [2012] ECHR 1706; (2013) 56 EHRR 12 (see above text accompanying n 111) – that ‘it is doubtful’ whether, during the purely preventive part of a sentence of indefinite detention, detainees should be housed in ‘conditions ... as close as possible to normal life’: Andrew Ashworth and Lucia Zedner, ‘Some Dilemmas of Indeterminate Sentences: Risk and Uncertainty, Dignity and Hope’ in Jan W de Keijser, Julian V Roberts and Jesper Ryberg (eds), Predictive Sentencing: Normative and Empirical Perspectives (Bloomsbury, 2019) 127, 142–3. If a person is not being punished, s/he should be detained in conditions that are as non-punitive as possible. That is so regardless of when the person’s preventive detention was ordered. That said, those commentators seem clearly right to have indicated that such an arrangement is not ‘politically plausible’ in ‘the present penal climate’: at 143. And note that some commentators have

expressed scepticism about whether it is possible to create detention conditions that are significantly less punitive than those in prisons. See, eg, Richard Lippke, ‘No Easy Way Out: Dangerous Offenders and Preventive Detention’ (2008) 27(4) Law and Philosophy 383, 409–13.

^[118] Christopher Slobogin, ‘Legal Limitations on the Scope of Preventive Detention’ in Bernadette McSherry and Patrick Keyzer (eds), Dangerous People: Policy, Prediction and Practice (Routledge, 1^st ed, 2011) 37, 39.

^[119] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 30 [93] (Gageler J). See also 49 [163], where Gordon J lists some of these ‘prophylactic offences’.

^[120] Criminal Code Act 1995 (Cth) s 105A.4(2).

^[121] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 15 [39] (Kiefel CJ, Bell, Keane and Steward JJ).

^[122] See above text accompanying nn 72–3.

^[123] ICCPR art 9(1).

^[124] Fardon v Australia (n 43) 9 [7.4]; Tillman v Australia (n 43) 11 [7.4].

^[125] Fardon v Australia (n 43) 8 [7.4]; Tillman v Australia (n 43)10 [7.4].

^[126] [1996] HCA 24; (1996) 189 CLR 51.

^[127] Ibid 68 (Dawson J); Robert French, ‘The Kable Legacy: Its Impact on the

Australian Judicial System’ in John Griffiths and James Stellios, Current Issues in Australian Constitutional Law: Tributes to Professor Leslie Vines (Federation Press, 2020) 209, 212.

^[128] French (n 127) 212.

^[129] As noted by Dawson J in dissent: Kable [1996] HCA 24; (1996) 189 CLR 51, 83.

^[130] Ibid 101–3 (Gaudron J), 113–15 (McHugh J), 126, 137–9 (Gummow J).

^[131] Ibid 94–6 (Toohey J), 101 (Gaudron J), 114 (McHugh J), 126 (Gummow J).

^[132] Ibid 96 (Toohey J), 103–4, 106 (Gaudron J), 118–19 (McHugh J), 132 (Gummow J).

^[133] Ibid 103 (Gaudron J). See also 96, 98 (Toohey J), 106 (Gaudron J), 117 (McHugh J), 132 (Gummow J).

^[134] Ibid 98 (Toohey J), 106–8 (Gaudron J), 119–24 (McHugh J), 132 (Gummow J).

^[135] Ibid 96.

^[136] Ibid 97.

^[137] Ibid, quoting Lim (1992) 176 CLR 1, 27 (Brennan, Deane and Dawson JJ).

^[138] Leeth v Commonwealth (1992) 174 CLR 455, 470 (Mason CJ, Dawson and McHugh JJ). See also the cases listed in n 36.

^[139] Kable [1996] HCA 24; (1996) 189 CLR 51, 97.

^[140] Ibid.

^[141] Ibid.

^[142] See text accompanying above n 83.

^[143] Kable [1996] HCA 24; (1996) 189 CLR 51, 98.

^[144] Ibid.

^[145] Ibid 132 (emphasis added). See also 134, where Gummow J stated that ‘[t]he Act requires the Supreme Court to inflict punishment without any anterior finding of criminal guilt’.

^[146] Ibid 132.

^[147] Ibid.

^[148] Ibid 122 (emphasis added).

^[149] In Fardon QCA [2003] QCA 416, [88], McMurdo P thought it clear that Gaudron J had reached this

conclusion.

^[150] Kable [1996] HCA 24; (1996) 189 CLR 51, 106.

^[151] Ibid 107, quoting Re Nolan; Ex parte Young [1991] HCA 29; (1991) 172 CLR 460, 497 (Gaudron J) (emphasis added). Her Honour also noted that the impugned Act made it ‘plain that [Kable] ... is to be detained in prison and

subject to substantially the same regime as persons convicted of criminal offences’: at 105.

^[152] George Winterton, ‘Justice Kirby’s Coda in Durham’ (2002) 13(3) Public Law Review 165, 168.

^[153] Ibid 170.

^[154] Greg Taylor, ‘Conceived in Sin, Shaped in Iniquity: The Kable Principle as Breach of the Rule of Law’ [2015] UQLawJl 12; (2015) 34(2) University of Queensland Law Journal 265, 266.

^[155] Jeffrey Goldsworthy, ‘Kable, Kirk and Judicial Statesmanship’ [2014] MonashULawRw 7; (2014) 40(1) Monash University Law Review 75, 109.

^[156] Ibid 75.

^[157] Dyson Heydon, ‘Judicial Activism and the Death of the Rule of Law’ [2004] OtaLawRw 2; (2004) 10(4) Otago Law Review 493, 494.

^[158] Ibid 514.

^[159] Fiona Wheeler and John Williams, ‘“Restrained Activism” in the High Court of Australia’ in Brice Dickson (ed), Judicial Activism in Common Law Supreme Courts (Oxford University Press, 2007) 19, 44.

^[160] Murray Gleeson, ‘Judicial Legitimacy’ (2000) 20 Australian Bar Review 4, 6.

^[161] Ibid.

^[162] Ibid.

^[163] See Fardon [2004] HCA 46; (2004) 223 CLR 575, 593 [23] (Gleeson CJ). Similarly, in Baker [2004] HCA 45; (2004) 223 CLR 513, a Kable challenge that the HCA dismissed on the same day as it disposed of Fardon, Gleeson CJ made it clear that Kable ‘was not an invention of a method by which judges may wash their hands of the responsibility of

applying laws of which they disapprove ... The most basic quality of courts in which the public should have confidence is that they will administer justice according to law’: at 519–20 [6].

^[164] Scott Stephenson, From Dialogue to Disagreement in Comparative Rights Constitutionalism (Federation Press, 2016) 85. Stephenson proceeds to observe that, ‘[a]fter implying new rights into the Constitution ... Australian courts retreated in subsequent cases, preserving these rights but applying them in such a manner as to render them largely ineffectual’. Certainly, this has been so with ‘the Kable principle’, which, it is now clear, will be breached only in ‘extreme’ circumstances: Vella v Commissioner of Police (NSW) (2019) 269 CLR 219, 246 [56] (Bell, Keane, Nettle and Edelman JJ) (‘Vella’).

^[165] Swearing in of Sir Owen Dixon as Chief Justice (1952) 85 CLR xi, xiv. This return to ‘legalism’ was

strenuously resisted by one member of the Gleeson Court. See, eg, Justice Michael Kirby, Judicial Activism: Authority, Principle and Policy in the Judicial Method (Sweet & Maxwell, 2004) 9–11. Justice Kirby’s basic criticism of Dixonian ‘legalism’ was that it can involve judges pretending to perform ‘a mechanical function whilst knowing, when they stop to think about it, that that they play a vital role in making law’: at 10 (emphasis in original). In other words, where the law is unclear, judges of course have a law making role – a role that is sometimes downplayed by those who espouse ‘legalism’. To this it can be added that an

apparently ‘legalistic’ decision can in fact be a formalistic one. At above n 13, and in Part IV of this article, I discuss various decisions in which, though the law seemed clear, the HCA ignored matters of substance and deployed highly dubious reasoning to avoid applying that law. Though this is sometimes not recognised, that is the antithesis of ‘legalism’.

^[166] Dan Meagher, ‘The Status of the Kable Principle in Australian Constitutional Law’ (2005) 16(3) Public Law Review 182, 185.

^[167] Dangerous Prisoners (Sexual Offenders) Act 2003 (Qld) s 13.

^[168] Fardon QCA [2003] QCA 416, [28].

^[169] Fardon [2004] HCA 46; (2004) 223 CLR 575, 579–80.

^[170] Ibid; Fardon QCA [2003] QCA 416, [28].

^[171] See, eg, Kable [1996] HCA 24; (1996) 189 CLR 51, 118–19 (McHugh J); HA Bachrach Pty Ltd v Queensland [1998] HCA 54; (1998) 195 CLR 547, 561–2 [13]–[14]; Baker [2004] HCA 45; (2004) 223 CLR 513, 527 [24] (McHugh, Gummow, Hayne and Heydon JJ). The point here is this: the Commonwealth Parliament may only validly authorise the courts mentioned in section 71 of the Commonwealth Constitution to exercise ‘the judicial power of the Commonwealth’ (and functions ancillary or incidental to it): R v Kirby; Ex parte Boilermakers’ Society of Australia [1956] HCA 10; (1956) 94 CLR 254, 275 (Dixon CJ, McTiernan, Fullagar and Kitto JJ); A-G (Cth) v The Queen [1957] AC 288, 311–14

(Viscount Simonds). But as indicated above – see text accompanying above nn 132–3, 144 and 147 – state and territory parliaments may validly require their courts to exercise a wider range of non-judicial functions than this. As Gummow J implied in Kable [1996] HCA 24; (1996) 189 CLR 51, 132, so long as the non-judicial function is compatible with ‘the exercise by [the relevant Court] ... of federal jurisdiction conferred pursuant to s 77(iii) of the Constitution’, the state or territory law that confers such a function on the Court will withstand Kable scrutiny.

^[172] Lim (1992) 176 CLR 1, 28 (Brennan, Deane and Dawson JJ).

^[173] Fardon [2004] HCA 46; (2004) 223 CLR 575, 653 [214].

^[174] Ibid.

^[175] See, eg, Kruger v Commonwealth [1997] HCA 27; (1997) 190 CLR 1, 162 (Gummow J) (‘Kruger’); Fardon [2004] HCA 46; (2004) 223 CLR 575, 613 [83] (Gummow J), 634 [154] (Kirby J).

^[176] Fardon [2004] HCA 46; (2004) 223 CLR 575, 653–4 [215], quoting Kruger [1997] HCA 27; (1997) 190 CLR 1, 162 (Gummow J).

^[177] Fardon QCA [2003] QCA 416, [30].

^[178] Ibid [101].

^[179] Ibid [42].

^[180] As they expressly noted: Fardon [2004] HCA 46; (2004) 223 CLR 575, 650 [207].

^[181] Ibid 654 [216].

^[182] Ibid.

^[183] Ibid 654 [217].

^[184] Ibid 596 [34].

^[185] Ibid 597 [34].

^[186] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 65 [214].

^[187] Ibid.

^[188] [1996] HCA 24; (1996) 189 CLR 51, 122. See text accompanying above n 148.

^[189] [2004] HCA 49; (2004) 225 CLR 1, 32 [77] (‘Woolley’) (emphasis in original).

^[190] Fardon [2004] HCA 46; (2004) 223 CLR 575, 597 [34].

^[191] Ibid 591 [18] (Gleeson CJ), 647 [196] (Hayne J).

^[192] Ibid 608 [68]–[69] (Gummow J), 631 [145] (Kirby J).

^[193] Fardon QCA [2003] QCA 416, [60].

^[194] Ibid [90].

^[195] Ibid.

^[196] Fardon [2004] HCA 46; (2004) 223 CLR 575, 636 [162].

^[197] Ibid 637 [164] (Kirby J).

^[198] See ibid 640 [173].

^[199] Ibid 637 [165].

^[200] Ibid 640 [173].

^[201] Ibid 608–14 [69]–[85].

^[202] Ibid 612–13 [80]–[81].

^[203] [2004] HCA 37; (2004) 219 CLR 562, 612 [136].

^[204] Ibid 612 [137].

^[205] Cf Benbrika [2021] HCA 4; (2021) 388 ALR 1, 65 [215] (Edelman J).

^[206] Fardon [2004] HCA 46; (2004) 223 CLR 575, 613 [83] (Gummow J).

^[207] Fardon QCA [2003] QCA 416, [42].

^[208] Fardon [2004] HCA 46; (2004) 223 CLR 575, 613 [83] (Gummow J).

^[209] Lim (1992) 176 CLR 1, 28 (Brennan, Deane and Dawson JJ).

^[210] Fardon [2004] HCA 46; (2004) 223 CLR 575, 613 [84] (Gummow J).

^[211] See text accompanying above nn 76–85.

^[212] At least one Australian commentator has deployed similar reasoning to this when rejecting the view that

preventive detention of the dangerous is analogous to the ‘exceptional case’ of mental illness detention:

Jeffrey Steven Gordon, ‘Imprisonment and the Separation of Judicial Power: A Defence of the Categorical Immunity from Non-Criminal Detention’ [2012] MelbULawRw 2; (2012) 36(1) Melbourne University Law Review 41, 101.

^[213] As defined at above n 48.

^[214] Fardon QCA [2003] QCA 416, [78]–[80].

^[215] Fardon [2004] HCA 46; (2004) 223 CLR 575, 637–8 [165].

^[216] Ibid 609–10 [70]–[72], 613 [83] (Gummow J).

^[217] Ibid 613 [83] (Gummow J).

^[218] Fardon QCA [2003] QCA 416, [79].

^[219] Fardon [2004] HCA 46; (2004) 223 CLR 575, 613 [83].

^[220] [1996] HCA 24; (1996) 189 CLR 51, 97.

^[221] Ibid.

^[222] [2019] HCA 31; (2019) 268 CLR 1, 27 [48].

^[223] Ibid.

^[224] [2004] HCA 46; (2004) 223 CLR 575, 591 [18] (Gleeson CJ), 647 [196] (Hayne J).

^[225] See above n 171 and accompanying text.

^[226] See, eg, Fardon [2004] HCA 46; (2004) 223 CLR 575, 614 [86]–[87] (Gummow J), 630–1 [144]–[146] (Kirby J), 656 [219] (Callinan and Heydon JJ).

^[227] Ibid 621 [119].

^[228] Ibid 614 [90] (Gummow J).

^[229] See, eg, Meagher (n 166) 185; Keyzer (n 21) 105–6; Fiona Wheeler, ‘The Kable Doctrine and State Legislative Power Over State Courts’ (2005) 20(2) Australasian Parliamentary Review 15, 25; Anthony Gray, ‘Standard of Proof, Unpredictable Behaviour and the High Court of Australia’s Verdict on Preventive Detention Laws’ [2005] DeakinLawRw 9; (2005) 10(1) Deakin Law Review 177, 189; Jeremy Gans, ‘Current Experiments in Australian Constitutional Criminal Law’ (Australian Association of Constitutional Law, 9 September 2014) 1; Gabrielle J Appleby and John M Williams, ‘A New Coat of Paint: Law and Order and the Refurbishment of Kable’ [2012] FedLawRw 1; (2012) 40(1) Federal Law Review 1, 8.

^[230] See text accompanying above n 163.

^[231] See text accompanying above nn 160–2.

^[232] Tom Campbell, ‘Judicial Activism: Justice or Treason?’ [2003] OtaLawRw 2; (2003) 10(3) Otago Law Review 307, 326.

^[233] Fardon [2004] HCA 46; (2004) 223 CLR 575, 593 [23]. McHugh J made similar remarks at 601 [41]–[42].

^[234] See, eg, Vella (2019) 269 CLR 219, 235 [24] (Bell, Keane, Nettle and Edelman JJ); Benbrika [2021] HCA 4; (2021) 388 ALR 1, 69 [226] (Edelman J). The same goes for Brennan CJ’s statement in Nicholas v The Queen [1998] HCA 9; (1998) 193 CLR 173, 197 [37] that ‘[i]ntegrity is the fidelity to legal duty, not a refusal to accept as binding a law which the court takes to be contrary to its opinion as to the proper balance to be struck between competing interests’. See, eg, Vella (2019) 269 CLR 219, 256 [80] (Bell, Keane, Nettle and Edelman JJ).

^[235] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185] (Edelman J).

^[236] As the anonymous referee mentioned at above n 13 has suggested, it could alternatively be argued that Gleeson CJ’s remarks should be taken at face value. On such a view, his Honour was not concerned with

perceptions, but was simply stating that the HCA must apply the law whether it likes that law or not – and he decided the case before him accordingly. The difficulty with such a view is that, as I have just argued – see text accompanying above n 229 – the law to be applied in Fardon provided no clear answer to the question that the Court had to resolve. Another difficulty with it is that, insofar as the applicable law did point to an answer, that law did not favour the state’s submission. In other words, as I have also suggested above – see text accompanying above nn 166, 229 – the unchallenged decision in Kable tended to support the

prisoner’s argument.

^[237] See, eg, Justice Virginia Bell, ‘Keeping the Criminal Law in “Serviceable Condition”: A Task for the Courts or the Parliament?’ (2016) 27(3) Current Issues in Criminal Justice 335, 336 (‘Keeping the Criminal Law in “Serviceable Condition”’); Chief Justice Susan Kiefel, ‘Social Values and the Criminal Law’s Adaptability to Change’ (Speech, International Criminal Law Congress, 6 October 2018).

^[238] Justice Michelle Gordon, ‘The Integrity of Courts: Political Culture and a Culture of Politics’ [2021] MelbULawRw 3; (2021) 44(3) Melbourne University Law Review 863, 866 (‘The Integrity of Courts’). See also, eg, Chief Justice Susan Kiefel, ‘Judicial Independence: From What and to What End?’ (Speech, Austin Asche Oration, 27 March 2021) 12–13.

^[239] Justice MH McHugh, ‘The Judicial Method’ (1999) 73(1) Australian Law Journal 37, 43.

^[240] James Stellios, Zines’s The High Court and the Constitution (Federation Press, 6^th ed, 2015) 696.

^[241] McHugh, ‘The Judicial Method’ (n 239) 43. A very good recent example of this is Love v Commonwealth (2020) 270 CLR 152, in which the HCA by bare majority held that Aboriginal Australians are not capable of being regarded as ‘aliens’ within the meaning of Commonwealth Constitution s 51(xix). As Gerangelos has

persuasively argued, this was not a case where there was a clear answer: Peter Gerangelos, ‘Reflections upon Constitutional Interpretation and the “Aliens Power”: Love v Commonwealth’ (2021) 95(2) Australian Law Journal 109, 110, 116–18. As he also notes, however, this did not stop certain commentators from seeking to ‘impugn the very legitimacy of the decision’: at 109. It is submitted that attacks such as this go a long way towards explaining why, since the appointment of Gleeson CJ in 1998, it has been far more usual for the Court to develop or change the law only where their Honours are reasonably sure that the development or change will accord with ‘contemporary societal values’ (Justice Virginia Bell, ‘Judicial Activists or

Champions of Self-Restraint: What Counts for Leadership in the Judiciary?’ (Speech, The General Sir John Monash Leadership Oration, 4 August 2016) 4) and, therefore, will give rise to few claims that they are

deciding cases on the basis of what they perceive to be ‘desirable policy’ (to use the words of Kiefel CJ in dissent in Love at 172 [8]).

^[242] See Fardon [2004] HCA 46; (2004) 223 CLR 575, 586 [3], 590 [14] (Gleeson CJ), 601 [41] (McHugh J).

^[243] Kirby (n 165) 10.

^[244] See, eg, Woolley [2004] HCA 49; (2004) 225 CLR 1, 24 [57] (McHugh J).

^[245] See text accompanying above nn 202–4.

^[246] [2018] HCA 2; (2018) 262 CLR 333, 341 [16] (Kiefel CJ, Bell, Keane and Edelman JJ). See also, eg, North Australian

Aboriginal Justice Agency Ltd v Northern Territory [2015] HCA 41; (2015) 256 CLR 569, 592–3 [37] (French CJ, Kiefel and Bell JJ), 610 [94] (Gageler J), 651–2 [236] (Nettle and Gordon JJ) (‘NAAJA’).

^[247] See text accompanying above nn 205–6.

^[248] See Benbrika [2021] HCA 4; (2021) 388 ALR 1, 28 [84] (Gageler J); cf 65 [215] (Edelman J).

^[249] Falzon [2018] HCA 2; (2018) 262 CLR 333, 344 [31] (Kiefel CJ, Bell, Keane and Edelman JJ).

^[250] Ibid 343 [27] (Kiefel CJ, Bell, Keane and Edelman JJ), quoting Kruger [1997] HCA 27; (1997) 190 CLR 1, 162 (Gummow J).

^[251] Ibid 344 [29] (Kiefel CJ, Bell, Keane and Edelman JJ).

^[252] Ibid.

^[253] Ibid 344 [32] (Kiefel CJ, Bell, Keane and Edelman JJ). See also 343 [28] (Kiefel CJ, Bell, Keane and

Edelman JJ).

^[254] See, eg, Gabrielle Appleby and Stephen McDonald, ‘Punishment and Chapter III of the Constitution’ in John Griffths and James Stellios (eds), Current Issues in Australian Constitutional Law: Tributes to Professor Leslie Zines (Federation Press, 2020) 64, 77–8; Stephen McDonald, ‘Involuntary Detention and the

Separation of Judicial Power’ (2007) 35 Federal Law Review 26, 47–8; Stellios, Zines’s The High Court and the Constitution (n 240) 317.

^[255] Leslie Zines, The High Court and the Constitution (Federation Press, 5^th ed, 2008) 289. See also Stellios, Zines’s The High Court and the Constitution (n 240) 317.

^[256] Falzon [2018] HCA 2; (2018) 262 CLR 333, 344 [32] (Kiefel CJ, Bell, Keane and Edelman JJ).

^[257] Woolley [2004] HCA 49; (2004) 225 CLR 1, 33 [78].

^[258] Zines (n 255) 289.

^[259] See, eg, Lucia Zedner, ‘Penal Subversions: When Is a Punishment Not Punishment, Who Decides and on What Grounds?’ (2016) 20(1) Theoretical Criminology 3, 11–14.

^[260] See, eg, Behrooz v Secretary, Department of Immigration and Multicultural and Indigenous Affairs [2004] HCA 36; (2004) 219 CLR 486, 499 [21] (Gleeson CJ), 507 [53] (McHugh, Gummow and Heydon JJ), 541–3 [167]–[176] (Hayne J), 559 [218] (Callinan J). Cf 526–9 [117]–[124] (Kirby J).

^[261] See text accompanying above nn 168–70.

^[262] Fardon [2004] HCA 46; (2004) 223 CLR 575, 613 [83].

^[263] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 13 [33] (Kiefel CJ, Bell, Keane and Steward JJ).

^[264] Ibid 14 [34] (emphasis added).

^[265] Ibid 9 [18]. See also, more recently, Commonwealth v AJL20 (2021) 95 ALJR 567, 576 [22] (Kiefel CJ, Gageler, Keane and Steward JJ).

^[266] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 14 [36] (Kiefel CJ, Bell, Keane and Steward JJ).

^[267] Ibid.

^[268] Fardon QCA [2003] QCA 416, [42] (de Jersey CJ, Williams JA agreeing at [94]).

^[269] Fardon [2004] HCA 46; (2004) 223 CLR 575, 653–4 [214]–[217].

^[270] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 15 [39] (Kiefel CJ, Bell, Keane and Steward JJ).

^[271] Criminal Code Act 1995 (Cth) s 105A.1. See also Benbrika [2021] HCA 4; (2021) 388 ALR 1, 4–5 [7] (Kiefel CJ, Bell, Keane and Steward JJ).

^[272] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 15 [39] (Kiefel CJ, Bell, Keane and Steward JJ).

^[273] Ibid 14 [36]. Yet, of course, there is a principled basis for distinguishing mental illness detention from the detention of the merely dangerous: the mentally ill person is non-responsible; the merely dangerous person is an autonomous actor. See text accompanying above nn 76–85. That is not to say, however, that their Honours were necessarily wrong to hold that Chapter III permits both mental illness detention and the detention of the merely dangerous. If the latter kind of detention is served in truly non-punitive conditions (not a gaol), it seems possible to argue that an analogy between it and mental illness detention should be drawn (even if the opposite conclusion can be justified in principle). And even if such detention is punitive in character, it might well be that, in some circumstances at least, the power to order it falls within ‘the judicial power of the

Commonwealth’. Certainly, Edelman J thinks so (Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185]), and we shall soon consider his reasoning on this point.

^[274] Ibid 16 [40].

^[275] See text accompanying above n 250.

^[276] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 16 [41].

^[277] See text accompanying above nn 135–48. Cf Kable [1996] HCA 24; (1996) 189 CLR 51, 88 (Dawson J).

^[278] See text accompanying above nn 172–99. See also on this point Benbrika [2021] HCA 4; (2021) 388 ALR 1, 64–5 [214] (Edelman J). The plurality in Benbrika stated that McHugh J and Callinan and Heydon JJ regarded Fardon’s detention as punishment, which seems right; and that Gleeson CJ made clear his agreement with this analysis, which seems wrong: see Benbrika [2021] HCA 4; (2021) 388 ALR 1, 16 [41] n 88 (Kiefel CJ, Bell, Keane and Steward JJ). The Benbrika plurality justified this view by pointing to Gleeson CJ’s statement, at Fardon [2004] HCA 46; (2004) 223 CLR 575, 592 [20], that ‘[u]nless it can be said that there is something inherent in the making of an order for

preventive, as distinct from punitive, detention that compromises the institutional integrity of a court, then it is hard to see the foundation of the appellant’s argument’. However, as noted above, his Honour also

explicitly refrained from expressing a view about whether the Commonwealth Parliament could have passed the law at issue in Fardon: at 591 [18]. Given that, by so doing, Gleeson CJ necessarily avoided the question of whether the detention was or was not punitive, it seems that, when he used the term ‘punitive ... detention’ in the passage just quoted, he meant detention that was retributive in character. And when he contrasted it with ‘preventive detention’, he seems merely to have been drawing a distinction between such retributive

detention and detention that is purely ‘forward-looking.’ He does not seem to have been expressing a view that ‘preventive detention’ is always non-punitive.

^[279] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185], 65–6 [215]–[217], 69 [226].

^[280] See, eg, Dixon, ‘The High Court and Dual Citizenship’ (n 13), especially 140–5; Rosalind Dixon, ‘Functionalism and Australian Constitutional Values’ in Rosalind Dixon (ed), Australian Constitutional Values (Hart Publishing, 2018) 3, especially 9–10; Rosalind Dixon, ‘The Functional Constitution: Re-Reading the 2014 High Court Constitutional Term’ (2014) 43(3) Federal Law Review 455, especially 461–6.

^[281] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 23–5 [67]–[74], 28 [85] (Gageler J), 32 [103], 39–40 [134]–[136], 41–2 [138]–[142], 52 [178] (Gordon J). See also, eg, Vella (2019) 269 CLR 219, 275–6 [140]–[142] (Gageler J), 292 [190] (Gordon J); NAAJA [2015] HCA 41; (2015) 256 CLR 569, 610–11 [94]–[96] (Gageler J); Magaming [2013] HCA 40; (2013) 252 CLR 381, 400–1 [63]–[67] (Gageler J). Two commentators have recently argued in favour of a similar approach to Chapter III and the Lim principle: Appleby and McDonald (n 254) 95–7.

^[282] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 24 [71].

^[283] Ibid 25 [72].

^[284] Ibid.

^[285] Ibid 16–19 [42]–[48] (Kiefel CJ, Bell, Keane and Steward JJ).

^[286] Ibid 48–9 [163], 51–2 [173]–[178]. Unlike Gageler J (at 28–9 [85]–[88], 30 [93]), Gordon J refrained from stating that the scheme was valid insofar as it allowed for detention orders to be made to ‘prevent’ an

unacceptable risk of terrorist acts, or conduct that supported or facilitated such acts: at 52 [177]–[178].

^[287] In other words, her Honour, too, sought to give effect to what she perceived to be the liberal underpinnings of Chapter III: ibid 40 [135]–[136], 41–2 [138]–[142], 45 [150]–[151], 52 [177]–[178].

^[288] Criminal Code Act 1995 (Cth) s 105A.7(1).

^[289] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 28 [85].

^[290] Ibid.

^[291] Ibid 30 [93].

^[292] See ibid 48 [161]–[162], 52 [177]–[178].

^[293] Ibid 29 [87]–[88]. Justice Gageler’s view that Benbrika is not being punished is fully consistent with the views expressed by his Honour in Pollentine [2014] HCA 30; (2014) 253 CLR 629. Though Pollentine was serving an

indeterminate sentence in prison because of the risk that he would commit serious sexual offences in the

future, Gageler J held that such detention was ‘wholly protective’ and had ‘no punitive element’: at 657 [73].

^[294] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 31 [95].

^[295] As to this value and precisely what it might entail, see James Stellios, ‘Liberty as a Constitutional Value: The Difficulty of Differing Conceptions of “The Relationship of the Individual to the State”’ in Rosalind Dixon (ed), Australian Constitutional Values (Hart Publishing, 2018) 177 (‘Liberty as a Constitutional Value’).

^[296] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 53 [182]. See also 74 [239] (Edelman J).

^[297] Ibid. This approach is reminiscent of his Honour’s rejection, in SZTAL v Minister for Immigration and Border Protection [2017] HCA 34; (2017) 262 CLR 362, of the view that the person who does an act that s/he knows is virtually certain to produce a particular consequence, necessarily intends that consequence to ensue. ‘The law does itself no credit,’ Edelman J said, ‘by deeming one concept [recklessness] to be another [intention]’: at 396 [100].

^[298] See especially Benbrika [2021] HCA 4; (2021) 388 ALR 1, 59–61 [200]–[204].

^[299] At least, he did not do so when defending the view that the detention was punitive: cf ibid 68 [225] (Edelman J).

^[300] Ibid 53 [183].

^[301] See ibid 58 [197].

^[302] Ibid 53 [182] (Edelman J).

^[303] His Honour also seems right to have denied that Fardon is authority for the proposition that preventive

detention in prison is non-punitive: ibid 64–5 [214]; cf 16 [41] (Kiefel CJ, Bell, Keane and Steward JJ). That said, as argued above (see text accompanying above nn 184–90), McHugh J did seem to express such a view: cf ibid 65 [214] (Edelman J).

^[304] Ibid 65 [215].

^[305] See text accompanying above nn 205–6.

^[306] Ibid 65 [217]. Note the similarity of this contention to the statement of Winterton that I have quoted above: see text accompanying above n 153. Note, too, however, that neither Winterton nor Edelman J seem to be saying that the text of the Constitution clearly rules out the impugned implications: cf above nn 15, 165.

^[307] Ibid 65 [215]. For Edelman J’s approach to constitutional interpretation more broadly, see Justice James

Edelman, ‘2018 Winterton Lecture: Constitutional Interpretation’ (2019) 45(1) University of Western

Australia Law Review 1. See also Justice James Edelman, ‘Chief Justice French, Judicial Power and Chapter III of the Commonwealth Constitution’ in Henry Jackson (ed), Essays in Honour of Chief Justice French (Federation Press, 2019) 81, 104–6.

^[308] Lim (1992) 176 CLR 1, 27–8.

^[309] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185] (Edelman J).

^[310] Ibid.

^[311] Ibid 66–7 [220]–[221], 71–2 [232]–[233] (Edelman J). See also, eg, Thomas [2007] HCA 33; (2007) 233 CLR 307, 327–9 [15]–[17] (Gleeson CJ), 347–8 [79], 356–7 [114]–[121] (Gummow and Crennan JJ), 507 [594]–[595]

(Callinan J), 526 [651] (Heydon J).

^[312] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185]. See also 67 [222] (Edelman J).

^[313] Ibid 68 [224], 69 [226].

^[314] Ibid 69 [226].

^[315] Ibid 72–3 [235].

^[316] Ibid 69 [226]. See also 54 [185] (Edelman J).

^[317] Ibid 69 [226], citing Fardon [2004] HCA 46; (2004) 223 CLR 575, 593 [23] (Gleeson CJ).

^[318] Ibid 18–19 [46]–[47] (Kiefel CJ, Bell, Keane and Steward JJ).

^[319] Ibid 73 [237] (Edelman J).

^[320] See Goldsworthy, ‘Tom Campbell on Judicial Activism’ (n 26) 248.

^[321] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185] (Edelman J).

^[322] As Gageler J of course argued in Benbrika: ibid 28 [85].

^[323] See, eg, Goldsworthy, ‘Tom Campbell on Judicial Activism’ (n 26); Goldsworthy, ‘Kable, Kirk and Judicial Statesmanship’ (n 155); Jeffrey Goldsworthy, ‘The Limits of Judicial Fidelity to Law: The Coxford Lecture’ (2011) 24(2) Canadian Journal of Law and Jurisprudence 305.

^[324] Goldsworthy, ‘Kable, Kirk and Judicial Statesmanship’ (n 155) 76.

^[325] Ibid 114.

^[326] See Dyer, Can Charters of Rights Limit Penal Populism? (n 1) ch 5.

^[327] And, as noted above, Goldsworthy doubts that their Honours knowingly did this. See text accompanying above nn 155–6.

^[328] I defend this view more fully in Dyer, Can Charters of Rights Limit Penal Populism? (n 1) ch 5; but the old case of Fowler v Padget [1798] 7 TR 508 seems to provide an example of what I am referring to here. In that case, the relevant statute provided that a person would be guilty of a bankruptcy offence if s/he performed certain conduct ‘to the intent or whereby [his/her] creditors may be defeated or delayed’: at 510. The Court held that the word ‘and’ should be read for ‘or’, because otherwise ‘monstrous consequences ... would manifestly ensue’: at 514 (Lord Kenyon CJ). That is, persons would be convicted of this offence without having displayed the slightest fault. Regarding constitutional cases, for Goldsworthy, if there is to be constitutional change, the only way to achieve this is through the procedure contemplated by section 128 of the Constitution: Jeffrey Goldsworthy, ‘Originalism in Constitutional Interpretation’ [1997] FedLawRw 1; (1997) 25(1) Federal Law Review 1, 27; Jeffrey Goldsworthy, ‘Interpreting the Constitution in its Second Century’ (2000) 24(3)

Melbourne University Law Review 677, 683–4. Others, however, have perceived some shortcomings in this argument: see, eg, Jeremy Kirk, ‘Constitutional Interpretation and a Theory of Evolutionary Originalism’ (1999) 27(3) Federal Law Review 323, 352–3.

^[329] A more recent example than the one I provide in above n 328 seems to be the HCA’s decision in Kirk v

Industrial Court of New South Wales (2010) 239 CLR 531. Goldsworthy appears right to argue that the

reasoning in that case was thin and implausible: see Goldsworthy, ‘Kable, Kirk and Judicial Statesmanship’ (n 155) 93–104. But, as he also suggests (at 75–6, 110–11), it additionally tended to shield the individual from state power; and, as Sackville has noted, the decision was popular with the press and public: Ronald Sackville, ‘Bills of Rights: Chapter III of the Constitution and State Charters’ (2011) 18(2) Australian Journal of Administrative Law 67, 77. In my view, it was therefore justified.

^[330] A proposition noted and accepted by the plurality in Magaming [2013] HCA 40; (2013) 252 CLR 381, 396 [47] (French CJ, Hayne, Crennan, Kiefel and Bell JJ).

^[331] A number of commentators have expressed such a view. See, eg, Anthony Gray, Criminal Due Process and Chapter III of the Australian Constitution (Federation Press, 2016) 286–8; Desmond Manderson and Naomi Sharp, ‘Mandatory Sentences and the Constitution: Discretion, Responsibility and Judicial Process’ [2000] SydLawRw 25; (2000) 22(4) Sydney Law Review 585, 604–5; Andrew Trotter and Matt Garozzo, ‘Mandatory Sentencing for People Smuggling: Issues of Law and Policy’ [2012] MelbULawRw 15; (2012) 36(2) Melbourne University Law Review 553, 594–5; Dyer, ‘(Grossly) Disproportionate Sentences’ (n 1) 204–10.

^[332] Ex parte Coorey [1944] NSWStRp 60; (1944) 45 SR (NSW) 287, 300 (‘Coorey’).

^[333] Ibid 407 [81].

^[334] Ibid 408 [82].

^[335] Coorey [1944] NSWStRp 60; (1944) 45 SR (NSW) 287, 314–15 (Davidson J); Palling v Corfield [1970] HCA 53; (1970) 123 CLR 52, 58 (Barwick CJ).

^[336] (2008) 234 CLR 532, 563 [52].

^[337] Neither is it persuasive to argue, as five justices did in Magaming, that, when a prosecutor chooses to charge a person with an offence with a mandatory minimum penalty, instead of an identical offence that lacks such a penalty, his/her choice is ‘no different from the choice which a prosecutor must often make between proceeding summarily against an accused and presenting an indictment’: [2013] HCA 40; (2013) 252 CLR 381, 394 [39] (French CJ, Hayne, Crennan, Kiefel and Bell JJ). On this point, see Dyer, ‘(Grossly) Disproportionate Sentences’ (n 1) 208–9.

^[338] R v Crump (Supreme Court of NSW, McInerney J, 24 April 1997).

^[339] R v Taylor and Minogue (Supreme Court of Victoria, Vincent J, 24 August 1988) 7573.

^[340] R v Knight [1989] VicRp 62; [1989] VR 705, 711 (Hampel J).

^[341] Minogue [2019] HCA 31; (2019) 268 CLR 1, 20 [30] (Gageler J).

^[342] Corrections Act 1986 (Vic) s 74AB(1).

^[343] To use the words of the HCA in Knight [2017] HCA 29; (2017) 261 CLR 306, 317 [3].

^[344] Gans (n 229) 31–4; Dyer, ‘Irreducible Life Sentences, Craig Minogue and the Capacity of Human Rights Charters to Make a Difference’ (n 1) 506–7. Cf Sarah Murray, ‘Ad Hominem Parole Legislation, Chapter III and the High Court’ [2018] UWALawRw 24; (2018) 43(2) University of Western Australia Law Review 275, 282.

^[345] [2014] 2 Qd R 504.

^[346] See ibid 530 [41].

^[347] Ibid 523 [24].

^[348] For cases where the HCA has stated that this is contrary to Kable, see eg, North Australian Aboriginal Legal Aid Service Inc v Bradley [2004] HCA 31; (2004) 218 CLR 146, especially 163 [29] (McHugh, Gummow, Kirby, Hayne, Callinan and Heydon JJ); Assistant Commissioner Condon v Pompano Pty Ltd (2013) 252 CLR 38, 89 [125] (Hayne, Crennan, Kiefel and Bell JJ); NAAJA [2015] HCA 41; (2015) 256 CLR 569, 593–4 [39] (French CJ, Kiefel and Bell JJ).

^[349] Knight [2017] HCA 29; (2017) 261 CLR 306, 323–4. See also Minogue [2019] HCA 31; (2019) 268 CLR 1, 17 [19] (Kiefel CJ, Bell, Keane, Nettle and Gordon JJ).

^[350] Crump [2012] HCA 20; (2012) 247 CLR 1, 18–19 [34] (French CJ), 27 [60] (Gummow, Hayne, Crennan, Kiefel and Bell JJ), 28–9 [70]–[72], 29 [74] (Heydon J); Knight [2017] HCA 29; (2017) 261 CLR 306, 323 [25], 323–4 [29]; Minogue [2019] HCA 31; (2019) 268 CLR 1, 17–18 [19]–[22] (Kiefel CJ, Bell, Keane, Nettle and Gordon JJ), 21 [32] (Gageler J).

^[351] Dyer, ‘Irreducible Life Sentences, Craig Minogue and the Capacity of Human Rights Charters to Make a

Difference’ (n 1) 501–5; Dyer, Can Charters of Rights Limit Penal Populism? (n 1) ch 2.

^[352] Cf Crump [2012] HCA 20; (2012) 247 CLR 1, 19 [35] (French CJ), 29 [72] (Heydon J); Knight [2017] HCA 29; (2017) 261 CLR 306, 323–4 [29]; Minogue [2019] HCA 31; (2019) 268 CLR 1, 17 [17]–[19] (Kiefel CJ, Bell, Keane, Nettle and Gordon JJ).

^[353] Interestingly, the HCA seems to have acknowledged that these prisoners are serving irreducible life

sentences: Minogue v Victoria [2018] HCA 27; (2018) 264 CLR 252, 272–3 [53]–[54] (Kiefel CJ, Bell, Keane, Nettle and Edelman JJ), 276 [72] (Gageler J). See also Minogue [2019] HCA 31; (2019) 268 CLR 1, 20 [30] (Gageler J).

^[354] R v Knight [1989] VicRp 62; [1989] VR 705, 711; R v Taylor and Minogue (Supreme Court of Victoria, Vincent J, 24 August 1988) 7573; R v Crump (Supreme Court of New South Wales, McInerney J, 24 April 1997).

^[355] Dixon, ‘The High Court and Dual Citizenship’ (n 13) 162.

^[356] To use the words of Judge Pinto de Albuquerque in Ilnseher v Germany (European Court of Human Rights, Grand Chamber, Application Nos 10211/12 and 27505/14, 4 December 2018) [130].

^[357] Magaming [2013] HCA 40; (2013) 252 CLR 381, 407 [81] (Gageler J).

^[358] Stellios (n 41) 79 [3.56]. See also Fiona Wheeler, ‘Original Intent and the Doctrine of the Separation of

Powers in Australia’ (1996) 7(2) Public Law Review 96, especially 99, 103–4.

^[359] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 65 [215], 65 [217].

^[360] Fardon [2004] HCA 46; (2004) 223 CLR 575, 613 [83] (Gummow J). See also, eg, Habitual Criminals Act 1905 (NSW).

^[361] See text accompanying above n 264.

^[362] See text accompanying above nn 213–23.

^[363] Lim (1992) 176 CLR 1, 27 (Brennan, Deane and Dawson JJ). As argued above, properly viewed the

preventive part of an order of indefinite detention does not punish guilt; it is ‘forward-looking’.

^[364] See Stellios, ‘Liberty as a Constitutional Value’ (n 295) 190–1.

^[365] Appleby and McDonald (n 254) especially 95–8.

^[366] Cf Gordon, ‘Imprisonment and the Separation of Judicial Power’ (n 212) 101. See also above n 273.

^[367] See Oscar Roos, ‘Commonwealth Legislative Power and Non-Punitive Detention: A Constitutional Roadmap’ (2005) 1(3) High Court Quarterly Review 142, 149; Anthony Gray, ‘Internment of Terrorism

Suspects: Human Rights and Constitutional Issues’ (2018) 24(3) Australian Journal of Human Rights 307, 322. Note, too, Posner’s comments about the influence of policy considerations when judges ‘reason by

analogy’: Richard Posner, How Judges Think (Harvard University Press, 2008) 180–3.

^[368] See, eg, Fardon [2004] HCA 46; (2004) 223 CLR 575, 637 [165] (Kirby J).

^[369] German governments constructed such facilities, at great cost, in response to the ECtHR’s adverse

findings in M [2009] VI Eur Court HR 169 about the previous system of preventive detention in Germany: Ilnseher v Germany (European Court of Human Rights, Grand Chamber, Application Nos 10211/12 and 27505/14, 4 December 2018) [194].

^[370] This is yet another difficulty with the plurality’s reasoning in Benbrika. As we have seen, according to that reasoning, Benbrika’s detention is non-punitive. Therefore, not only can the Commonwealth Parliament

authorise a Chapter III court to order such detention, it can also validly authorise the executive to order it. Justice Edelman avoided that difficulty. As noted above, he made it clear that, because Benbrika’s detention is punitive, the Commonwealth Parliament may only authorise a Chapter III court to order it: Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185].

^[371] Fardon [2004] HCA 46; (2004) 223 CLR 575, 593 [23].

^[372] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 54 [185], 69 [226].

^[373] See, eg, Dyer, ‘Can Charters of Rights Limit Penal Populism?: The Case of Preventive Detention’ (n 1) 539.

^[374] Bell, ‘Keeping the Criminal Law in “Serviceable Condition”’ (n 237) 339.

^[375] Gordon, ‘The Integrity of Courts’ (n 238) 871–81.

^[376] As indicated at the beginning of this article – see text accompanying above nn 2–4 – the position is different in jurisdictions with a charter of rights. That is because, as Lord Bingham observed, ‘decisions in favour of unpopular minorities tend to be unpopular, but are the essence of human rights protection’: Lord Bingham of Cornhill, ‘The Judges: Active or Passive’ (2006) 139 Proceedings of the British Academy 55, 71. See also, eg, Reyes v The Queen [2002] UKPC 11; [2002] 2 AC 235, 246–7 [26] (Lord Bingham). That said, as also indicated at the

beginning of this article, there are various reasons why prisoners’ rights might end up not being better

protected under a charter of rights than they otherwise would have been: see text accompanying above nn 5–10.

^[377] As many judges have observed. See, eg, Justice Michael McHugh, ‘The Law-Making Function of the Judicial Process: Part II’ (1988) 62(2) Australian Law Journal 116, 122; Chief Justice John Doyle, ‘Do Judges Make Policy? Should They?’ (1998) 57(1) Australian Journal of Public Administration 89, 93; Lord Reid, ‘The Judge as Law Maker’ (1972) 12(1) Journal of the Society of Public Teachers of Law 22, 23; Lord Devlin, ‘Judges and Lawmakers’ (1976) 39(1) Modern Law Review 1, 8; Lord Bingham, The Business of Judging: Selected Essays and Speeches (Oxford University Press, 2000) 27, 31–3.

^[378] Gordon, ‘The Integrity of Courts’ (n 238) 886.

^[379] See text accompanying above n 370. What if Edelman J had found that not even a court could order such detention? Alternatively, what if he had drawn a distinction between indefinite detention and post-sentence detention, holding that only the latter had to be non-punitive before a court could order it? Consistently with the views that I have stated in the text, the first of these decisions, if supported by a majority of the Court, would probably not have been very popular with the press and public. It would have amounted to a judicial statement that, unlike in many other Western democracies, including those with a charter of rights in force, non-punitive preventive detention is entirely off-limits. The second of these decisions, as well as drawing an unprincipled distinction between indefinite and post-sentence preventive detention, would also have been liable to give rise to claims of activism. ‘Does the government really have to construct new non-punitive

facilities to house people like Benbrika?’, it might have been said.

^[380] Although, as argued above, Edelman J, by holding that the judiciary may order punitive detention for ‘future crimes’, in fact altered that principle. This seems unproblematic, because the alteration concerned a point that was not squarely at issue in Lim – the judiciary’s ability validly to order punitive detention for ‘future crimes’ – and the plurality’s statement in Lim that a Commonwealth statute may only validly authorise the judiciary to punish past guilt seems, again, to be based on the incorrect assumption that indefinite detention amounts to such punishment.

^[381] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 65 [215].

^[382] James Stellios and Matthew Zagor, ‘Australian Constitutionalism and the UK-Style Dialogue Model of

Human Rights Protection’ in John Bell and Marie-Luce Paris (eds), Rights-based Constitutional Review: Constitutional Courts in a Changing Landscape (Edward Elgar, 2016) 148, 156–7.

^[383] Stellios, ‘Liberty as a Constitutional Value’ (n 295) 191. See also Stellios and Zagor (n 382) 153–6; Stellios, The Federal Judicature (n 41) 71–2 [3.43].

^[384] Stellios, ‘Liberty as a Constitutional Value’ (n 295) 191.

^[385] Ibid, quoting Nicholas Barber, ‘Prelude to the Separation of Powers’ (2001) 60(1) Cambridge Law Journal 59, 64. See also, eg, Fardon [2004] HCA 46; (2004) 223 CLR 575, 586 [2] (Gleeson CJ); Thomas [2007] HCA 33; (2007) 233 CLR 307, 329 [17] (Gleeson CJ), 506–7 [592] (Callinan J). Cf McGarry v Western Australia [2005] WASCA 252; (2005) 31 WAR 69, 78–80 [34]–[39] (Wheeler JA).

^[386] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 69 [226].

^[387] Ibid.

^[388] See text accompanying above nn 290–3. Justice Gordon of course expressed no final view about whether such a scheme would be valid, though her Honour indicated that it might be: ibid 52 [177]–[178].

^[389] Stellios, The Federal Judicature (n 41) 71 [3.43].

^[390] Ibid.

^[391] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 28 [85].

^[392] Ibid 24 [70], 25 [72] (Gageler J).

^[393] Ibid 25 [73] (Gageler J).

^[394] Of course, the same could be said in support of a rule that Chapter III does not permit the judiciary to order punitive preventive detention under any circumstances. After all, the punitive character of such detention

renders it arbitrary: Fardon v Australia (n 43) 8 [7.4]; Tillman v Australia (n 43) 10 [7.4]. But, as argued above, the Court had good reasons for refraining from stating such a principle: see text accompanying above nn 362–79.

^[395] See text accompanying above nn 99, 119.

^[396] Benbrika [2021] HCA 4; (2021) 388 ALR 1, 18 [46] (Kiefel CJ, Bell, Keane and Steward JJ). See also 70 [229], 73 [237] (Edelman J).

^[397] Elizabeth Handsley, ‘Public Confidence in the Judiciary: A Red Herring for the Separation of Judicial Power’ [1998] SydLawRw 9; (1998) 20(2) Sydney Law Review 183, 187.

^[398] Ibid.

^[399] See Stellios, The Federal Judicature (n 41) 71 [3.43].