Handsley, Elizabeth --- "Comment" [2008] SydLawRw 6; (2008) 30(1) Sydney Law Review 115

• Comment∗

Comment^[∗]

ELIZABETH HANDSLEY^[+]

I thank my colleague Patrick Keyzer for his scholarly and thought-provoking review of the law that led to the 2004 decision in Fardon v Attorney-General (Qld)^[1] — or perhaps did not lead to Fardon, depending on which way you look at it. I share Patrick’s concern at the excessively formalistic approach that the High Court has taken to the separation of judicial power. Patrick demonstrates his concern about this issue by asking the question in his title: to what end the separation of judicial power? His ultimate answer to that question is: to the end of protecting civil liberties. Both of us would like to see the adoption of a more substantive and principled approach, that would put that goal in relief.

It is worth pausing to articulate exactly how the separation of judicial power can protect civil liberties. Most importantly, as mentioned by Patrick, independent judges can shield individuals and minority groups from the ‘tyranny of the majority’, often expressed through laws or practices of the ‘political branches’ — ie the legislature and the executive. In order to provide that shield, judges need to apply pre-determined law and/or to use strictly defined, fair and balanced procedures. Both of these things not only facilitate the application of the rule of law, but act as inherent limits on judicial power. Therefore they are legitimising factors for the relative lack of accountability applying to judges compared to members of the other branches.

This leads to another important role for separation of powers: even if judges cannot do anything to counter governmental practices that infringe civil liberties, they can at least identify them. The flip side of this proposition is that governments cannot use the judiciary as a smoke screen for their oppressive policies. They cannot require judges to do their ‘dirty work’ so as to avoid the electoral accountability that provides their legitimacy.

I share with Patrick a frustration at the way cases like Kable,^[2] Fardon^[3] and Baker^[4] have avoided a square confrontation with the civil liberties implications of the misuse of judges; but, unlike Patrick, I do not think the way out is to focus on the purpose of incarceration. This is particularly the case if we gauge the ‘purpose’ by looking at the conditions under which a person is incarcerated. Patrick is quite right that the soundest constitutional approach is to look at the reality rather than trying to read the legislator’s mind. However, it is necessary to think about where his logic could lead.

Patrick provides a great deal of convincing factual support for the proposition that Australian prisons are punitive in their effect, and argues that this is the reason why judges should not be allowed to order the incarceration of those who have not done anything wrong (for which they have not already been punished). As a matter of policy I totally agree. As a matter of constitutionality, however, I think that this line of argument risks casting the justices of the High Court in a role for which neither they nor the institution is suited. For the constitutionality or otherwise of the use of judges to order preventive detention is made to depend on how bad prison conditions are. The use of such a test, logically, leaves it open for a government to improve conditions for preventive detainees, and return to the court for its stamp of approval. In other words, Patrick’s approach turns High Court judges into prison inspectors. The questions for the court will always be: when are conditions bad enough to be punitive, and when are they good enough to be non-punitive? Any improvement or deterioration in prison conditions could have very serious constitutional implications; how many times would governments and detainees be allowed to go back to the court to argue about the effect of such changes? In principle, there should be no limit.

High Court litigation is not an appropriate procedure for keeping an eye on the level of suffering inflicted on prison inmates, and the broad range of elements of prison conditions and levels at which they might exist does not lend itself to the kind of clear, yes-or-no answers that the High Court has to give. Therefore, while I agree with the impulses that lead Patrick to put the ‘punitiveness’ test forward, I do not see much future in the test itself.

I would propose an alternative approach, which centres not on punishment but on the fact of deprivation of liberty, which will rarely if ever be a matter of factual dispute. This would pave a much smoother way for constitutional analysis, with the starting point being that deprivation of liberty is a judicial function. Therefore, in principle (ie, subject to possible exceptions), it should be carried out only by judges. Perhaps even more importantly, however, it must be carried out in accordance with proper judicial procedures. This is the point that Gaudron J recognised when she said in Kable that the proceedings in question ‘do not in any way partake of the nature of legal proceedings’.^[5] Legal proceedings normally determine the consequences of things that have happened, they do not aim to predict what might happen. Patrick recognises this in his comment that ‘[t]he focus of judicial power on past events is not accidental. Judicial power is characterized by the application of the law to past events of conduct’.^[6] To this I would add that these observations highlight an important aspect of the logic that supports judicial independence: as mentioned earlier, we do not need to worry about judges’ relative lack of accountability compared to that of the other branches of government, because their power is strictly circumscribed by their procedural limitations and, in particular, by the retrospective nature of the decisions that they make. Allowing judges to look into the future is inconsistent with this logic, and therefore strikes at the heart of the delicate balance we have achieved in Australia with the separation of powers.

I suspect that Patrick would accept the value of the approach I am suggesting because it is consistent with his observation that ‘[i]mprisonment is always punitive in effect’.^[7] The only change I would make to that observation is to extend it to all deprivations of liberty. If this is the case, there is no need to look at the particular conditions of imprisonment (or other incarceration), but rather the general position should be that to deprive a person of his or her liberty is part of judicial power, and can be done only following a proper, retrospective judicial procedure.

Stating the principles in this way addresses the problem that I have identified in my previous writings on Kable, that disallowing judicial involvement in State preventive detention plans does not necessarily help the potential detainees, because the BLF case^[8] leaves it open for State Parliaments to exercise State judicial power. Kable confirmed, if there were any doubt, that it was not derogating from that principle. The State government that is thwarted in its attempt to involve judges in preventive detention need only secure the passage of legislation to the same effect. However, if the principle extends not just to the personnel fulfilling the function, but to the procedure by which they fulfil it, the potential detainee should have the same protection because the procedure is the key to the protection. The exercise of judicial power, either by a court or by Parliament, must be retrospective, and if detention must follow a judicial procedure, it can never be ordered, by a judge or by a Parliament, on the basis of something that has not yet happened.

Of course it needs to be recognised that there are well-known exceptions to the proposition that only a judge can deprive a person of liberty, such as where detention serves the purpose of maintaining control over an illegal immigrant until such time as his or her refugee status can be determined. It is often said, and Patrick has agreed, that the categories of non-judicial detention are not closed. I have no interest in gainsaying that proposition, but I do think it is time to figure out a principled basis on which new categories might be developed and accepted — or indeed rejected. In such a process it may be found that the preventive detention of dangerous individuals, in limited circumstances, could be defended as an extension of the power to detain the mentally ill for their own protection and that of the community. But this would be a far more principled basis on which to proceed than by stretching and squeezing separation of powers principles, as the High Court did in the 2004 cases. This is even more the case considering the unresolved tensions between Kable and its offspring, on the one hand, and the general acceptance that not all Ch III protections apply to State courts, on the other. We know that the appointment and tenure protections of s 72 do not apply, and we know that something to do with institutional integrity does. But in between, we may never know just which Ch III (or Ch III-like) protections do and do not apply to State courts. In a sense one might see the whole Kable-based run of cases as an object lesson in the dangers that arise when High Court judges find interesting propositions to graft onto the Constitution: because of the nature of their roles they do not have an opportunity to work through the full implications of those propositions. Nor is anything helped by changes over time in the intellectual approach of the bench. Under these conditions it is not so surprising to find what Patrick has rightly identified as the Court’s over-reliance in Fardon on the ‘singling out’ aspect of the Kable legislation, when that did not form an essential part of the ratio in that case. He has also been justly critical of the conclusion that the preventive detention orders in question in Fardon were not punitive. This is the kind of formalism that has no place in constitutional adjudication.

Formalism has also been observable in this run of cases in relation to the role of separation of powers in ensuring that governments are held to account for their policies and their actions. If Kable showed us nothing else, it stands as an example of how, even where there is no formal separation of powers between courts and Parliament, it is possible to use the logic of separation of powers as an intellectual tool in thinking through some other question (for example, suitability as a repository for the judicial power of the Commonwealth). That logic, as we have seen, includes a proposition relating to maintaining the accountability of the ‘political branches’, and there were elements of such a line of reasoning in Kable. It was especially evident in McHugh J’s judgment where his Honour stated that the Act left the Supreme Court judges with no alternative but to order Mr Kable’s continued detention.^[9] If this was true, one could conclude that the court was being used as a tool in a legislative plan, whereas the logic of separation of powers would state that, if the government wanted Mr Kable detained, it should have done so itself. The problem was that, by the time the matter reached the High Court, a Supreme Court judge had in fact ordered Mr Kable’s release. Therefore, McHugh J’s assertion bore one of the key hallmarks of formalism: it ignored the facts. In Fardon, we see the same kind of formalism applied to a different aspect of the facts, allowing the majority to conclude that the legislation is not ad hominem, simply because it does not mention Mr Fardon by name, even though it was introduced some weeks before he was due to be released. Patrick recognises, rightly, that the Fardon legislation was functionally identical to that which was challenged in Kable, and points out, again rightly, how unconvincing the Fardon court’s distinction of Kable was.

There is simply too much at stake in constitutional cases to limit one’s vision the way the High Court has done here. Constitutional issues are by definition fundamental and important, and they deserve to be decided on the basis of the fullest facts and the clearest principles. Formalism avoids both, and I thank Patrick for his role in exposing just how this has occurred.

However, I find myself unable to agree with the key statement in Patrick’s paper that Fardon ‘cuts our constitutional jurisprudence adrift from the rich normative reference points provided by the jurisprudence of liberty and due process’.^[10] The reason I disagree is that I don’t believe our constitutional jurisprudence, at least on these issues, was ever tethered to liberty and due process. The uniquely uncompelling analysis of ‘public confidence’ from Kable was certainly not so tethered; nor is the completely opaque concept of ‘institutional integrity’ with which it has been replaced. The case Patrick advances is for a fundamental re-evaluation of the whole issue of misuse of judges, that would take the norms of liberty and due process as it starting point. This would allow, or indeed require, the kind of steady focus on facts and principles that has been sadly lacking from the constitutional jurisprudence up until now.

---

^[∗] Comment on Patrick Keyzer, ‘Preserving Due Process or Warehousing the Undesirables: To What End the Separation of Judicial Power of the Commonwealth?’ (2008) 30 Sydney Law Review 100.

⁺ Professor of Law, Flinders University.

^[1] Fardon v Attorney-General (Qld) [2004] HCA 46; (2004) 223 CLR 575.

^[2] Kable v Director of Public Prosecutions (NSW) [1996] HCA 24; (1996) 189 CLR 51 (‘Kable’).

^[3] Fardon v Attorney-General (Qld) [2004] HCA 46; (2004) 223 CLR 575.

^[4] Baker v The Queen [2004] HCA 45; (2004) 223 CLR 513.

^[5] Kable [1996] HCA 24; (1996) 189 CLR 51 at 106.

^[6] Patrick Keyzer, ‘Preserving Due Process or Warehousing the Undesirables: To What End the Separation of Judicial Power of the Commonwealth?’ (2008) 30 Sydney Law Review 100 at 105.

^[7] Id at 108 (emphasis in original).

^[8] BLF (NSW) v Minister for Industrial Relations (1986) 7 NSWLR 372 (CA).

^[9] Kable v Director of Public Prosecutions (NSW) [1996] HCA 24; (1996) 189 CLR 51 at 123.

^[10] Keyzer, above n 6 at 112.