Allan, James --- "The Three 'Rs' of Recent Australian Judicial Activism: Roach, Rowe and (No)'Riginalism" [2012] MelbULawRw 19; (2012) 36(2) Melbourne University Law Review 743

CRITIQUE AND COMMENT

THE THREE ‘RS’ OF RECENT AUSTRALIAN JUDICIAL ACTIVISM: ROACH, ROWE AND (NO)’RIGINALISM

JAMES ALLAN^[*]

[In this article the author argues that two recent High Court of Australia decisions, Roach v Electoral Commissioner and Rowe v Electoral Commissioner, are prime examples of judicial activism, of judges employing interpretive techniques that have the effect of significantly inflating their own discretionary powers at the point of application of the Constitution. Indeed, he argues that these interpretive techniques would almost certainly be rejected by voters pondering a move from parliamentary sovereignty to a written constitution were these techniques, and their effects, spelt out in advance. He considers those two decisions in detail and then concludes by noting several unpalatable implications of the thinking underlying them.]

I INTRODUCTION

Judicial activism is a hotly contested notion or concept,^[1] one that usually carries with it pejorative connotations. At its heart, the label ‘judicial activism’ suggests some degree of illegitimacy. The core charge is that the judges have exceeded their proper role in a democracy. They have moved from the many grey areas, or penumbras of doubt,^[2] involved in interpreting the laid-down statutes and constitutional provisions (where disagreement and diverging answers are only to be expected from individual judges who bring differing values, concerns, emphases and intellectual abilities to the task) into something that no longer looks like interpretation. It looks more like legislating from the bench, otherwise described as point-of-application judges imposing their own first-order moral and political preferences, judgements and sentiments on all the rest of us.

The gist of the judicial activism complaint, then, is a complaint about what the unelected top judges are doing — that they are gainsaying or second-guessing or circumscribing or redirecting the elected branches of government without any legitimate warrant or grounds for doing so.

In that above sense the judicial activism charge is a serious one to make. Notice, however, that it does not necessarily connote bad faith. The gainsaying, second-guessing and circumscribing can be done not only to achieve what are believed to be good substantive outcomes (which can motivate even bad faith judicial activism), but also in the belief the constitutional materials and jurisdiction’s rules of recognition^[3] do allow such actions. The latter belief, in other words, can be honestly held by the judges. It is just that disinterested observers may disagree and think such a belief far-fetched in the particular circumstances. Still, that does not amount to bad faith on the part of the judge.

Accordingly, at least in my sense, judicial activism need not be an exercise in bad faith interpreting. This complaint or gravamen is broad enough also to encompass implausible and unconvincing interpretation, where the legal materials do not support the substantive outcomes (however worthy) that judges believe are possible.

Of course the line between interpretation that constitutes judicial activism and interpretation that does not will be drawn in different places by different people. Almost everyone might recognise the possibility of judicial activism in the abstract, but in any particular case where that charge or allegation is made, you are likely to find smart, well-informed, nice people simply disagreeing about the merits of that charge. I accept that reality up-front.

Nevertheless, in the rest of this paper I will seek to convince the reader that two recent High Court of Australia decisions are prime examples of judicial activism in my above sense; they are rather blatant examples of illegitimate judging techniques or interpretive approaches taken by the majority Justices. The fact the outcomes that are achieved in both instances are likely to be seen by many (me included) as on balance a good call in cost–benefit terms (if one were in the position of legislating on a blank slate) does not in some magical, ineffable way make the illegitimate interpretive approaches of the majority judges thereby acceptable or legitimate. This is still judicial activism at its worst, or so I will argue in what follows.

The two cases I will be discussing are Roach v Electoral Commissioner (‘Roach’)^[4] and Rowe v Electoral Commissioner (‘Rowe’).^[5] The first is a prisoner voting rights case. The second has to do with the entitlement to vote as well, but this time more circuitously: the issue in the case was when the electoral rolls (listing all eligible voters) were to be closed and hence prevent any further applications for enrolment. In both Roach and Rowe, the social policy lines that had been drawn by the democratically elected legislature were invalidated by the top judges of the land. The governing statutory provisions were struck down by majority judgments of the High Court of Australia — four of six of the sitting Justices decided to do so in Roach, while in Rowe it was a 4:3 decision.

Both majority decisions, in my view, rest on the most implausible and far-fetched understanding of the meaning of the Australian Constitution, one that significantly liberates the point-of-application interpreter when it comes to gainsaying, indeed overruling, the elected legislature. This Roach and Rowe understanding of how to give meaning to Australia’s written Constitution allows its judicial exponents to claim — at least implicitly — that legislation can be (and was) constitutionally valid at the time of Federation and the coming into force of that Constitution (and indeed that the legislation remained so up to 1983 and beyond) but that that same legislation is today no longer constitutionally valid.

On top of that, this same Roach and Rowe approach to constitutional interpretation — to giving meaning to that text — also carries with it the clear and undeniable suggestion that if Parliament keeps its hands off and leaves alone old legislation governing, say, when prisoners can vote or when electoral rolls must close, then that old legislation will be and will remain valid. But where a Parliament in the recent past happens to have legislated to liberalise those rules then no Parliament of even more recent vintage will be able to revert back to the older rules. Not ever. The Constitution, or so these Roach and Rowe judges claim, forbids it.

Just think about that for a moment and whether these claims are best characterised as the results of persuasive interpretations of an Australian constitutional text that disavows any US-style bill of rights or, alternatively, as the results of point-of-application majority judges simply reading their own moral and political preferences, sentiments and druthers into that text to achieve outcomes they happen to think are better than the ones chosen by the legislature. And while you are pondering which characterisation is likely to be more persuasive, remember that no relevant part of the text of that Constitution — the one the majority judges say used to allow the legislature to do something but now does not — has changed. The relevant parts of the text being interpreted are exactly as they were. The words have not changed. Only the scope for judges to invalidate democratically enacted legislation has changed. That has grown and expanded, quite considerably in fact. Or so a bare majority of Australia’s top judges tells us.

The rest of this paper comes in four Parts. Part II will be brief and will provide some context. Parts III and IV will then take the reader through the two cases, Roach and Rowe. The final Part of this paper will return to the topic of judicial activism and why both Roach and Rowe are prime examples of this sin.

II CONTEXT

The Roach and Rowe cases cannot be understood in isolation. They need to be seen as the latest incarnation of the so-called implied rights series of cases^[6] dating back from the early 1990s. I have written about those implied rights cases elsewhere^[7] and the very fast-and-loose interpretive approach the majority Justices relied upon in those cases. In brief, these decisions were very much premised on a ‘living tree’^[8] or ‘living constitution’ interpretive approach.

For our purposes in this paper there is no need to re-canvass all that in detail. It will suffice simply to remind the reader of the reasoning of Mason CJ in one of the first, and most important, of those implied rights cases, namely Australian Capital Television Pty Ltd v Commonwealth (‘ACTV’).^[9] Writing with the majority, the Chief Justice arrived at the conclusion that the Constitution — one that explicitly and deliberately left out any US-style bill of rights or First Amendment free speech entitlements and protections, opting, after much debate and discussion amongst the founders, to leave such social policy balancing exercises to the elected Parliament — implicitly created an implied freedom of political communication. His reasoning followed these steps:

1 The Constitution provides that elected Members and Senators of Parliament are to be ‘directly chosen by the people’;^[10]

2 hence those elected are representatives of the people;

3 hence they are accountable to the people;

4 thus they have a responsibility to take account of the views of the people;

5 therefore the judges interpreting this Constitution must be able to, and hereby do, assert that there is an implied freedom of political communication in relation to public affairs and political discussion.^[11]

The practical effect of ‘discovering’ this implied right to freedom of political communication (one that presumably had lain dormant for nine decades on the majority’s reasoning) was that the High Court Justices then could move on to strike down or invalidate parts of a statute putting limits on election broadcasting spending.

As this will be a common refrain of mine throughout this paper, let me here signal that as a strong believer in vigorous free speech who favours US-style scope to speak one’s mind, the fact is that I like the substantive outcome of this ACTV case — if the question is what is one’s preferred policy outcome. But if the question relates to the judicial task of properly interpreting and giving meaning to the laid-down laws of the land — the very task at the heart of most understandings of the rule of law concept^[12] — then I think the majority judgments in ACTV and the implied rights cases generally are implausible and wholly unconvincing. As I have said before, I do not think this implied right was discovered; rather, it was made up by the judges at the point of application.^[13]

However, whether the reader agrees with my characterisation of those cases, or not, really does not matter for what follows. One needs only to be aware of them, together with the fact that the minority Justices never quite managed to overturn these cases and so the implied right the majority Justices established became settled,^[14] after waxing then waning, with no successful attempt to extend this thinking to other notable potential rights that needs mentioning.

That should provide the reader with sufficient context for what follows. We can now move forward a decade and more and consider the two cases at the heart of my ‘recent judicial activism’ allegation.

Indeed that context will allow us to see that these two Roach and Rowe decisions did one more egregious thing as well. They greatly strengthened a judge-made proportionality-type test whereby democratically enacted legislation could be (and was) invalidated or struck down because the majority judges took it to be ‘beyond what is reasonably appropriate and adapted (or “proportionate”) to the maintenance of representative government.’^[15] And this determination, recall, was itself part of the larger question of whether the statute was incompatible with the constitutional requirement that the Australian Parliament be ‘directly chosen by the people’.^[16] Put in this way two things become clear. Firstly, and as I have noted already, these two cases of Roach and Rowe are closely connected to, and simply could not have proceeded without, the implied rights cases and the judicially discovered (or made up) implied freedom. In both Roach and Rowe, the judges proceed on the basis that the need for members of both the House of Representatives and Senate to be ‘directly chosen by the people’ authorises or mandates the High Court Justices to supervise (and potentially strike down) line-drawing choices made by Parliament vis-a-vis who can vote and when the electoral rolls can close, something that orthodox opinion (in my view) did not think was within the purview of the top Australian judges before the implied rights cases.

Secondly, this proportionality-type assessment (be it one asking whether the legislation is ‘appropriate and adapted to serve an end consistent or compatible with the maintenance of the prescribed system of representative government,’^[17] or one asking whether the legislation is ‘arbitrary’,^[18] or one asking about its ‘proportionality’,^[19] or one phrased some other way) clearly compounds the scope for debatable judicial value judgements. In fact, it doubles that scope. First off, the judges have to decide if the legislation impinges on or is incompatible with all that the implied rights edifice itself throws into doubt (based on the massively inflated scope this edifice assigns to determinations of whether Members of Parliament and Senators are ‘directly chosen by the people’). That first judgement or determination will often be nothing if not debatable, contentious and, from the perspective of the outside citizen looking in, highly discretionary. But then on top of that, there is now this second at least equally discretionary value judgement the judges need to make of whether the legislation is proportionate. In fact, Thomas Poole goes further than that and is scathing about the malleability and discretion-enhancing qualities of proportionality-type tests in the hands of the judiciary. Poole argues that

proportionality is plastic and can in principle be applied almost infinitely forcefully or infinitely cautiously, producing an area of discretionary judgment that can be massively broad or incredibly narrow — and anything else between.^[20]

Let us be kind before moving to Part II and put it no higher than this. In the Roach and Rowe cases the majority judgments in no conceivable way articulate clear limits or constraints or boundaries on the judges’ second-guessing or gainsaying or overruling powers over Parliament. Quite the opposite in fact.

III ROACH AND ANIMAL FARM JUDGING: FOUR YEARS GOOD, TWO YEARS BAD

Only six High Court Justices heard this case and they split 4:2 in favour of striking down legislation that prevented any person serving a full-time sentence from voting in federal elections. The four majority Justices ended up deciding that the existing legislation that disqualified all prisoners was invalid, but that the preceding legislation that disqualified those serving sentences of three years or more was constitutionally valid. So after Roach, the Commonwealth Parliament is left with the scope to disenfranchise those prisoners serving three or more year sentences (‘four years good’), but not to do so to those serving fewer than three years (‘two years bad’ — though note that there is room to argue about where this ‘restricting what Parliament can decide’ line actually is, given the reasoning of the majority Justices). And that limit on the Commonwealth Parliament’s sovereignty and line-drawing power as regards this area of debatable and contested social policy is a limit contained in the Constitution — or to be rather more specific, it is a limit that the four majority Justices in Roach claim and assert can be found by reading the Constitution and giving that foundational legal text its proper meaning.

Given that these four unelected High Court Justices are overriding and gainsaying the elected Commonwealth legislature, and doing so on the basis of a claim that the Constitution does not allow or empower the Parliament to do what it has done, it may surprise some readers to learn that the core textual basis for this majority judicial assertion — that part of their reasoning that focuses on the words of the Constitution itself, as opposed to that part that focuses on earlier High Court decisions and obiter dicta assertions and glosses on those decisions and dicta, as well as on overseas decisions and international law — is the phrase ‘directly chosen by the people’ in ss 7 and 24. That is the direct textual basis underpinning the majority’s claim that Parliament cannot in almost any imaginable real-life sentencing situation disenfranchise prisoners serving sentences of three years or less but can do so to those serving longer sentences.

The plausibility of that majority Roach claim will in large part depend upon what one takes the point or purpose of a written constitution to be and, concomitantly, whether one thinks that in a modern Western democracy such as Australia — where these sorts of decisions about prisoner voting entitlements will either be made by the elected legislature or by the unelected judges — the words of a written constitution can stay the same but their meaning (and so the restrictions they impose on legislative sovereignty and on democratic decision-making) can alter and change as time goes by, as announced by the judiciary.

But that issue of whether it is an attractive approach to constitutional interpretation to think and assert that the meaning of constitutional words can change and alter because of changing moral, political and social values (as perceived by the point-of-application interpreters of those words in the Constitution) I put aside until Part V of this article. For the remainder of this Part of the article I will outline the majority judgments in this Roach case.

Let us start not with the Chief Justice’s judgment but with the joint judgment of Gummow, Kirby and Crennan JJ. After some introductory paragraphs and a recounting of the facts and how the legislation had changed over the years, we come to paragraph 40 in which this joint judgment sets out the four grounds on which the plaintiff challenges the validity of ss 93(8AA) and 208(2)(c) of the Commonwealth Electoral Act 1918 (Cth) (‘Electoral Act’), which were inserted by the Electoral and Referendum (Electoral Integrity and Other Measures) Act 2006 (Cth) (‘2006 Act’). The first two are rejected and can be ignored for our purposes. The third of these grounds rests on an assertion that there is an implied freedom of political participation tied to the implied freedom of political communication. The fourth is that ‘the 2006 Act impermissibly limits the operation of the system of representative (and responsible) government which is mandated by the Constitution.’^[21]

The three Justices then proceed to bypass or sidestep the third of these grounds, half-heartedly asserting that ‘what is at stake on the plaintiff’s case is not so much a freedom to communicate about political matters but participation as an elector in the central processes of representative government.’^[22] That is that. That is the sole basis for claiming that the decision to invalidate these 2006 amendments to the Electoral Act does not flow from any slight extension of the earlier implied rights cases thinking. Of course what the joint judgment gives on this point is not really any argument or reasons for rejecting the plaintiff’s third ground. Instead it is just an assertion. And as we shall see, it is not an overly persuasive or convincing assertion because the joint judgment Justices need to, but do not, tell us why it is impermissible for the elected Parliament to do what it did in enacting the 2006 Act.

Or put the other way around, when the joint judgment Justices come to tell us why it is that they can strike down and invalidate this statute — what gives these judges the power to do this — they have virtually nothing to point to in the Constitution itself. Indeed they again and again make reference to the earlier implied rights case law, as we shall see.

Perhaps that explains, or partially explains, the rather half-hearted or irresolute nature of their rejection of this third ground for invalidating

ss 93(8AA) and 208(2)(c) as inserted by the 2006 Act. Whether it does or not, the joint judgment then turns to the fourth ground, and why that ground should succeed (namely that the Constitution mandates that the impugned legislation operates so as to impermissibly limit representative government).

Yet the very next paragraph after that half-hearted rejection of ground three, the joint judgment cites, and relies on, the key implied rights case from 1997, Lange v Australian Broadcasting Corporation (‘Lange’).^[23]

After that, the next step in the argument is the assertion that ‘the Constitution makes allowance for the evolutionary nature of representative government as a dynamic rather than purely static institution.’^[24] But that assertion finesses the crucial issue of how exactly representative government, including decisions about who can and cannot vote, will alter or change through time. Will the elected Parliament make these decisions or will the unelected High Court have some sort of supervisory role? More to the point, to whom did the Constitution leave this task?

To answer those questions requires a point-of-application interpreter to adopt a theory of, or approach to, interpretation. There are two main choices. One falls under the label of ‘originalism’ where the meaning of the words used is sought by seeking either their public meaning at the time of adoption (one sub-branch) or their intended meaning by those who drafted and passed them (the other sub-branch).^[25]

The ‘living tree’ or ‘living constitution’ option is the other main approach to attributing meaning to the words of a constitution.^[26] It becomes abundantly clear that the joint judgment rejects originalism in favour of ‘living tree’ thinking, though this is done with little or no argument on why this ‘living constitution’ option — one that has the clear effect of giving judges more input and power because they (the judges), and no one else, become the ones who will say how the constitutional words have altered their meaning as time passes — is to be preferred.

Still, the joint judgment recognises that representative government can be a dynamic institution through time in two ways: either because Parliament itself occasionally changes the rules falling under this aegis without any supervisory role or input from the top judges (which of course is precisely the situation in, say, New Zealand)^[27] or, alternatively, because the top judges do have a supervisory role.^[28]

Indeed, this whole Roach case, and the Rowe one that followed, are simply instances of our High Court answering that question in its own favour, concluding that the top judges have been given a supervisory role by the Constitution, at least by the year 2007 if not before.

Of course the constitutional issue is not a first-order one of whether you believe, think or prefer top judges to have this role. No, the issue is a second-order interpretive one of which alternative (no supervisory role on these particular issues for top judges or yes a supervisory role) was meant by the Constitution, properly interpreted. And we must remember that no theory of how best to interpret a constitution — not a Dworkinian one,^[29] not a ‘living tree’ one, not any other plausible option — sweepingly asserts that the words used can be given any meaning at all that latter-day point-of-application interpreters want them to have or think would achieve the best consequences in today’s world.

So the question in Roach is not whether the Justices think prisoners serving sentences of fewer than three years ought to be able to vote. No, the question is whether our written Constitution ultimately left this decision with the elected Parliament or with the unelected High Court.

The joint judgment concedes that ‘[o]n their face, the laws impugned by the plaintiff are supported by s 51(xxxvi) and by ss 8 and 30 of the Constitution’.^[30] However, the three Justices point out that the first of these is limited by the phrase ‘subject to this Constitution’ in the chapeau of s 51,^[31] while ss 8 and 30 contain ‘specific limitations upon the power of the Parliament to prescribe the franchise.’^[32]

However, those specific limitations have nothing to do with whether prisoners could vote. They have to do with non-plural voting and with the qualification of electors not differing between the two legislative chambers.^[33]

The next step in the argument put forward by the three joint judgment Justices is crucial. They now tell us why the ‘[o]n their face’^[34] outcomes the constitutional provisions appear to dictate — namely that this issue of prisoner voting has been left to Parliament to decide — are wrong. In other words, they argue why the otherwise seemingly intended meaning of the Constitution is misguided or misleading.

That fourth step involves a fifth one, namely calling in aid the Solicitor-General and claiming that ‘it appeared to be common ground (and correctly so) that these provisions were to be read not in isolation but with an appreciation of both the structure and the text of the Constitution.’^[35] More particularly, the constitutional text the joint judgment here has in mind is the five words ‘directly chosen by the people’ found in ss 7 and 24,^[36] precisely the same five words of the Constitution — and only five words — that were used (or relied upon) back when the implied freedom of political communication was discovered (or made up, depending on the theory of constitutional interpretation you bring to the table).

The joint judgment then asserts that

[t]he Commonwealth correctly accepts that ss 7 and 24 place some limits upon the scope of laws prescribing the exercise of the franchise, and that in addition to the specific insistence upon direct choice by those eligible to vote, laws controlling that eligibility must observe a requirement that the electoral system as a whole provide for ultimate control by periodic popular election.^[37]

Stop and notice two things about this passage. Firstly, any limits ‘in addition to’ the explicitly articulated direct choice requirement that one reads in ss 7 and 24 are limits that did not exist before the implied rights cases. Again, depending on one’s interpretive theory and philosophy, these additional limits were either created by the High Court Justices back in ACTV and Lange etc, or they were discovered by those Justices in those same cases — presumably after having lain dormant for some nine decades. Either way, this passage in the joint judgment glosses over the immense weight being put onto the reasoning in those implied rights cases. Without those cases, and the edifice it constructed for additional judicial oversight of Parliament, it simply would not be true that ss 7 and 24 prescribe additional limits — ones above and beyond the explicit direct choice ones — on what electoral laws Parliament can enact.

Secondly, there is an element of ‘reasoning by claiming the Solicitor-General conceded the point’ going on here. Indeed in that same paragraph, the joint judgment carries on in the same vein, claiming that ‘in oral submissions, the Solicitor-General of the Commonwealth readily accepted that a law excluding members of a major political party or residents of a particular area of a state would be invalid’.^[38]

However, that is a highly debatable claim and, in my view, not a concession that ought to have been made. The general point can be made by thinking of a parliamentary sovereignty jurisdiction such as New Zealand. There, the elected legislature has no legal or constitutional constraints — no power in the top judges to pronounce a validly enacted law to be invalid. Rather the constraints are all political and moral, many of them tied to the limits on power that democracy creates.

The Australian Constitution clearly and without doubt, not least in the many references to ‘until the Parliament otherwise provides’^[39] and the deliberately chosen lack of a bill of rights,^[40] places much weight on these parliamentary sovereignty-style political limits on power. Unlike in the United States, our founders and our Constitution were extremely confident in the ultimate good sense and moral bearings of the voters. The scope for judges to invalidate statutes is much less than in the United States and Canada (where a potent bill of rights exists, an instrument Australia’s founders explicitly rejected and which is still shunned).^[41] Indeed, putting aside oversight of the federal distribution of powers,^[42] the Constitution puts great weight on parliamentary sovereignty (admittedly in the context of a written constitution), certainly much more so than in Canada, the US, the European Union, South Africa and almost anywhere else in the democratic world with a written constitution.

My point is that much that in the abstract might today seem distasteful, if enacted into law, does not therefore — simply because of its distastefulness or even because of its perceived egregious nature to many present-day sensibilities — become something over which the top judges have been given a supervisory role by the Constitution. And given that, the concessions attributed to the Solicitor-General are problematic, to put it kindly.^[43]

Nevertheless, the joint judgment makes use of those concessions, together with the aforementioned 1) references to the implied rights decisions; 2) ambiguous claims about the evolutionary nature of representative government; and 3) an implicit adoption of a ‘living tree’ approach to constitutional interpretation, given the rejection of any form of originalist interpretation that gives the words of the Constitution their original public meaning (and hence more circumscribed supervisory role for top judges). Indeed these steps or underpinnings of the argument get repeated and re-used.

So the joint judgment again refers to ‘the evolution of the constitutional requirements’.^[44] It relies, again, on Commonwealth concessions, this time to the effect that ‘there are constitutional restraints necessarily implicit in the otherwise broad legislative mandate conferred by the words “until the Parliament otherwise provides”’,^[45] those constitutional restraints meaning a supervisory role for the judges. Or again, having referred to Mulholland v Australian Electoral Commission^[46] and McGinty v Western Australia (‘McGinty’),^[47] in both of which the impugned law was upheld, it promptly cites Lange and the implied freedom of communication.^[48] And this is a prelude to glossing or refocusing the McGinty decision, claiming that that decision ‘does not deny the existence of a constitutional bedrock when what is at stake is legislative disqualification of some citizens from exercise of the franchise.’^[49]

The basis for that change of focus as regards the McGinty case — moving away from the fact the impugned statute in that case was upheld, over to using McGinty as a support for now invalidating a statutory provision — comes in the next paragraph. The joint judgment picks out and cites an obiter dictum from McGinty by Brennan CJ, one bearing on what ‘chosen by the people’ in ss 7 and 24 means. Not a single other dictum on this point of the many other possibilities on offer by many other Justices in McGinty was considered or cited in the joint judgment.^[50] Worse, the joint judgment omits the tentativeness and qualifications and limiting context present in Brennan CJ’s original obiter observations in McGinty.^[51]

Instead, the joint judgment states that

[i]n McGinty Brennan CJ considered the phrase ‘chosen by the people’ as admitting of a requirement ‘of a franchise that is held generally by all adults or all adult citizens unless there be substantial reasons for excluding them’.^[52]

And with that, the joint judgment is effectively finished as far as providing a ratio for thinking ‘the 2006 Act impermissibly limits the operation of the system of representative (and responsible) government which is mandated by the Constitution.’^[53] The Brennan CJ dictum from McGinty provides just the needed component to complete the reasoning, and that further component is the ‘substantial reasons’ test it is made to articulate. Virtually without anything else at all, this is used to presume that Australia’s top judges do have — or rather have been given by the Constitution — a supervisory role over the elected Parliament on whether, and which, prisoners can vote.

It is simply remarkable, in fact, how rapidly in just two paragraphs the Justices of the joint judgment turn the issue from one of 1) whether the Constitution, when properly interpreted, leaves this matter to the elected Parliament or gives the judiciary a gainsaying, overruling, supervisory role that includes the power to invalidate disfavoured statutes; into one of 2) whether the disqualifications in the 2006 Act are ‘for a “substantial” reason’.^[54] Indeed, this joint judgment provides incredibly thin gruel as far as that first issue is concerned. Yes, there are repeated references to the implied rights jurisprudence. Yes, there is an implicit rejection of any sort of originalist interpretive approach to understanding the meaning of the Constitution. Yes, concessions by the Solicitor-General are called in aid. Yes, rather ambiguous claims about the evolutionary nature of representative government are made. And yes, a single obiter dictum — one plucked out of myriad possibilities and one somewhat refashioned to sound less equivocal — is made to bear an immense amount of weight.

But that is it. The rest of the joint judgment^[55] is simply a form of proportionality analysis, however denominated or articulated. It contains all that extra double dose of discretionary judicial input and potential judicial gainsaying power, all that plastic malleability, that Thomas Poole argues all proportionality analyses share.^[56] Ultimately, these three Justices decided that they (or their understanding of the Constitution) will allow the elected Parliament to disenfranchise prisoners serving sentences of three years or more in accordance with the 2004 amendments to the Electoral Act,^[57] but will not allow the more restrictive 2006 Act regime.

At this point I could note the inherent cherry-picking nature of proportionality-type analyses, and how Sauvé v Attorney-General (Canada) (‘Sauvé’)^[58] and Hirst v United Kingdom [No 2]^[59] are mentioned but not New Zealand’s Re Bennett.^[60] Or I could ask why the 2006 Act is characterised as being about ‘stigmatis[ing]’ prisoners rather than about their character.^[61] Or I could be provocative and note what the joint judgment wholly fails to mention about the Sauvé decision.^[62] However, as all this latter part of the judgment comes after the joint Justices have already concluded that they have been given a supervisory role over these sorts of issues — a conclusion with which I strongly disagree and one that rests on feeble and sometimes elusive reasoning — I turn now from the joint judgment to that of Gleeson CJ.

We can be somewhat briefer here. That is because Gleeson CJ’s reasoning on the core issue of whether the top judges do or do not have a supervisory or ‘able to gainsay and overrule the Parliament’ role when it comes to the details of the franchise — an issue over which there was no binding authority, only obiter dicta, before this Roach case — is so truncated. It takes Gleeson CJ only eight paragraphs to conclude that the judiciary in Australia does have a supervisory role in vetting Commonwealth legislation that disqualifies some citizens from voting, a role that had never been acknowledged in the ratio of any case in the preceding hundred-plus years since Federation and a role that allows those judges potentially to invalidate or strike down that legislation.

Chief Justice Gleeson’s judgment starts with five-and-a-half paragraphs that, in effect, restate the fact that the drafters and ratifiers of the Constitution had a fundamental faith in the good sense of the voters, and in the democratic process, and in political checks on distasteful outcomes rather than court-focused, judge-driven ones.^[63] So he states that ‘[t]he Australian Constitution was not the product of a legal and political culture ... that created expectations of extensive limitations upon legislative power for the purpose of protecting the rights of individuals’^[64] and that ‘the framers ... admired and respected British institutions, including parliamentary sovereignty’^[65] (which, of course, means no ‘gainsaying of the elected Parliament role’ for the judges at all). He quotes Barwick CJ’s comment in Attorney-General (Cth) ex rel McKinlay v Commonwealth (‘McKinlay’) that the

Constitution was federal in nature with consequential limitation on the sovereignty of the Parliament. ... But otherwise there was no antipathy amongst the colonists to the notion of the sovereignty of Parliament in the scheme of government.^[66]

He then notes that the Constitution

reflects a high level of acceptance of [parliamentary sovereignty]. ... Nowhere is this more plainly illustrated than in the extent to which the Constitution left it to Parliament to prescribe the form of our system of representative demo-cracy.^[67]

And Gleeson CJ even observes that the fact that ‘Australia came to have universal adult suffrage was the result of legislative action.’^[68]

Federalists like me might quibble with the suggestion that the High Court of Australia has done even a passable job in upholding federalist constraints on the Commonwealth Parliament, and wonder more so at any reference to this by one of the Justices who was in the majority in New South Wales v Commonwealth (‘Work Choices Case’).^[69] Yet those would be quibbles that miss the point here. Up to the first two or three sentences of paragraph 6 of the Chief Justice’s judgment there is no indication that he will decide for the plaintiff and invalidate the relevant 2006 amendments.

His reasons for doing so are given in the next two-and-a-half paragraphs — after that it is just 17 paragraphs^[70] of what amounts to proportionality analysis and asking not whether judges have this supervisory power but rather whether they ought to use it to gainsay Parliament in this instance,^[71] and I am not here directly interested in that latter endeavour.

Returning to paragraphs 6–8, here is the Chief Justice’s argument. Firstly, after all the aforementioned genuflecting in the direction of how large a role parliamentary sovereignty has played in the thinking of those who drafted and ratified our Constitution, and indeed those who interpreted it in years gone by, his first step is to point to overseas democratic jurisdictions and to suggest that there is ‘a broad agreement as to the kinds of exception [to universal suffrage] that would not be tolerated.’^[72]

Stop at this first step and notice two things. One is that interpretation of a constitutional text by appeal to overseas practice makes it overwhelmingly likely that the interpreter is adopting — without argument — a ‘living constitution’ or ‘living tree’ or ‘pick-your-favourite-metaphor for the idea that the meaning of the words of a constitution can change over time’ approach. Yes, it is possible to imagine an originalist interpreter seeking the most likely public meaning of some phrase at the time of Federation, and in cases of genuine historical uncertainty, looking at comparable jurisdictions with the same phrase at the same (or an earlier) time. But clearly Gleeson CJ is not appealing to overseas practice or consensus in that way. He is saying ‘they do it this way elsewhere today’ with the unspoken premise being ‘and so should, or possibly must, we’. And that sort of claim is only allowable where one’s interpretive theory sees the text as one that did not (contrary to originalist claims)^[73] lock in an answer or outcome in the past.

The second thing to notice about Gleeson CJ’s appeal to some broad overseas agreement as to who can be denied the vote is that it is empirically or factually suspect or debatable. Descend from the Olympian heights of moral abstractions and what we are talking about is whether, and which, convicted prisoners can vote. The implied suggestion that the 2006 Act stands off by itself at the far end of some notional spectrum of how other democracies opt to deal with the issue of prisoner voting is plain-out false. Many of the US states (the issue there being a state one) have a considerably more restrictive legislative regime vis-a-vis prisoner voting^[74] than the 2006 Act enacted. And some of those jurisdictions that are more liberal about prisoner voting are more liberal solely because of judges saying a bill of rights demands as much; they are the result of Parliaments in those jurisdictions being overruled by judges under a bill of rights,^[75] a point the Chief Justice fails to articulate.

The second step in the Chief Justice’s reasoning comes in the form of a rhetorical question followed by a statement of belief. ‘Could Parliament now legislate to remove universal adult suffrage? If the answer to that question is in the negative (as I believe it to be) then the reason must be in the terms of ss 7 and 24 of the Constitution’.^[76]

However, this second step (aside from seeming to reason backwards) neatly finesses or fails to distinguish two important reasons for why the answer to the rhetorical question might be in the negative. One possibility, the one the Chief Justice simply assumes to be correct, is that the answer is ‘no’ because the top judges have been afforded a supervisory role by the Constitution and were the elected parliament to legislate in this way the unelected judges would overrule them and invalidate the statute.

The other possibility, the live one in New Zealand to this day and the one in keeping with all the Chief Justice’s earlier genuflecting in the direction of the large role our Constitution reserves to parliamentary sovereignty, is that the answer is ‘no’ because of the democratic and political good sense of the voters and their elected representatives. On this possibility, because of political constraints and indeed the moral good sense of politicians (the only two constraints that exist, as it happens, to keep top judges from forsaking honest constitutional interpretation in favour of lying about what the words mean to reach conclusions they happen to desire), no political party or government would ever legislate to remove the vote from women, or indigenous Australians, or Catholics or however else you understand ‘remov[ing] universal adult suffrage’.^[77]

Not only have such political and moral constraints worked perfectly well in New Zealand and in the United Kingdom (leave aside, if you wish, the period after the latter’s entry into what is now the European Union), and not only did the Chief Justice concede that the intention of the drafters and ratifiers of our Constitution was explicitly to rely on those constraints and overwhelmingly to shun judge-operated ones, the further fact is that

Gleeson CJ misses an important qualifier to attempts to defend a supervisory role for judges by appealing to theoretically possible apocalyptic scenarios like removing the vote from Catholics. It is a qualifier I made at length in an earlier article discussing similar argumentative techniques by some of their Lordships in the House of Lords:

On the other hand, if it be the former scenario of a raw grab for power by the government of the day blatantly attacking democratic institutions, five of their Lordships appear to believe that the unelected House of Lords (in either its legislative or judicial manifestations) could stop them. ... If we are to take such an implausibility even remotely seriously then on its own terms — in my view — neither a handful of top judges nor an assembly of appointed placemen would have the slightest prospect of preventing such an outcome. Resistance would have to come from across society. If it did not, then as Professor Hart has said, commenting on an analogously unlikely scenario, ‘[t]he society in which this [did not happen] might be deplorably sheep like; the sheep might end in the slaughter-house’.^[78]

Put simply, if you imagine a scenario where the elected legislature really runs amok in some beyond the Pale way, then the views of a handful of High Court Justices would do nothing to stop it. On the other hand, if you are really talking about a difference of opinion over where to draw policy lines, a difference over which smart, nice, reasonable people can and do differ — an issue precisely such as the one here over which convicted prisoners ought to get the vote — then the apocalyptic scenario does not apply and there is no need for a judicial supervisory role at all. The founders might opt for one, or they might not (with what the founders actually chose presumably mattering to latter-day judges). Or to put that last claim slightly more accurately, there is no need for the point-of-application top judges to give themselves that supervisory role by adopting an approach to constitutional interpretation that is so externally unconstrained, so dependent on moral and ethical judgements divorced from the task of seeking the text’s intended meaning, and so likely to have the side effect of significantly enhancing the power and discretion of those same judges.

I would normally leave my comments about the Chief Justice’s step two at that. However, as he returns to the point — ‘[i]t is difficult to accept that Parliament could now disenfranchise people on the ground of adherence to a particular religion’^[79] — and as the now retired Justice Michael Kirby (who participated in the Roach joint judgment discussed above) has extra-judicially commented upon Gleeson CJ’s motivations in Roach, I will here repeat those extra-judicial Kirby remarks, take no view whatsoever on them, and then move on:

Some of what Professor Allan has been saying today looks rather similar to the dissenting views of Hayne and Heydon JJ in Roach, the prisoners’ voting case. However, the majority in that case came to the different view. I think the tipping point in Roach, as we call it now, was when I asked a question of the Solicitor-General for the Commonwealth: ‘Does your view of the Australian Constitution mean that Parliament could go back to the laws against voting by Roman Catholics? Could Parliament in Australia take away the vote from Roman Catholics?’ Gleeson CJ immediately pricked up his ears at that question about his co-religionists.^[80]

The third and final step of Gleeson CJ’s reasoning, before moving to his proportionality analysis, amounts to an argument that words in a constitution can remain the same, and yet ‘because of changed historical circumstances including legislative history’,^[81] the power and supervisory role they grant to the courts can expand over time.

Of course Gleeson CJ does not put it quite in those terms. Instead, much like the reasoning in the implied rights cases, he begins with the five words in ss 7 and 24 of the Constitution — ‘directly chosen by the people’. Immediately after Federation, he concedes, ‘those words did not mandate universal adult suffrage.’^[82] That is a more circumspect way of saying that the plaintiff in Roach would have lost, and the legislation being impugned upheld, had this case been decided in the early 1900s.

The third step then calls in aid an analogy, the Sue v Hill^[83] case about the meaning of the words ‘foreign power’ in s 44(i) of the Constitution, and that Court’s decision that the United Kingdom now (but not immediately after Federation or indeed for some time thereafter) fell under the aegis of that phrase.

Gleeson CJ characterises this Sue v Hill outcome in these terms: ‘The meaning of the words “foreign power” did not change, but the facts relevant to the identification of the United Kingdom as being included in or excluded from that meaning had changed.’^[84] To buttress that characterisation he cites a passage of McTiernan and Jacobs JJ in McKinlay that ‘universal adult suffrage may now be recognized as a fact’,^[85] though Gleeson CJ does not mention that this obiter dictum was part of the dissent there. He also points to a dictum by Gummow J in McGinty to similar effect.^[86] And that suffices, in the Chief Justice’s judgment, to conclude by analogy that what is happening in Roach is that the five words of ss 7 and 24 did not change their meaning, but that the relevant facts did — where ‘“fact” [refers] to an historical development of constitutional significance of the same kind as the developments considered in Sue v Hill.’^[87] And so universal adult suffrage is now ‘a long established fact, and that anything less could not now be described as a choice by the people ... because of changed historical circumstances including legislative history’.^[88]

And that is enough for Gleeson CJ to decide this novel issue in Roach, and indeed (after 17 paragraphs of proportionality analysis) to invalidate the amendments made to the Electoral Act by the 2006 Act (‘two years bad’), but not to have to do the same to the Electoral and Referendum Amendment (Prisoner Voting and Other Measures) Act 2004 (Cth) (‘four years good’).

In my view, however, the Chief Justice’s third step — the strongest of the three — simply cannot bear anywhere near the weight the Chief Justice needs it to bear. And to be perfectly clear, I say that on the explicit assumption — a debatable one — that Sue v Hill was correctly decided.

First off, it is just about plausible to assert that the drafters and ratifiers — having specifically rejected a bill of rights and having placed the sort of weight on parliamentary sovereignty that Gleeson CJ earlier noted — intended (and were understood to intend) latter-day judges to decide when Australia had sufficiently broken its ties with the mother country, rather than having this issue decided by way of a s 128 referendum and amendment. That understanding of the meaning of ‘foreign power’ in s 44(i), a sort of stand-alone judgment combining many factual determinations and a few value-laden ones, is just about plausible, and certainly defensible. What is totally implausible is to think the drafters and ratifiers — having thought about, debated and in no uncertain terms rejected a bill of rights — intended (and were understood to intend) latter-day judges to oversee who specifically could vote, and indeed which prisoners, serving which sentences, could vote. And indeed even that those same judges could oversee and decide and overrule Parliament on when the electoral rolls could or could not close (a foreseeable extension once you travel down this ‘judges are there to supervise franchise rights, albeit after applying a proportionality analysis’ road).

That is not just implausible. It verges on the ludicrous.

Accordingly, this Sue v Hill analogy cannot work if any consideration at all is being given, when undertaking the task of constitutional interpretation, to the original understanding and original intentions behind the only constitutional words in play, namely ‘directly chosen by the people’ in ss 7 and 24. When it comes to judges having a supervisory role over legislation today bearing on whether and when prisoners could vote, it is plain that the then understanding and intention was that the judges would have no such role.

Of course one can interpret a constitutional text without paying any attention at all to original intentions or understandings, though Larry Alexander argues this in fact does not amount to seeking meaning at all but rather to making it up at the point of application.^[89] At the very least such an approach as Gleeson CJ’s is a strong implicit endorsement of ‘living tree’ or ‘living constitution’ interpretive methods. Yet even so, and even for those wholly committed to rejecting all aspects of originalism, we can still point to serious flaws in this attempted analogy here to Sue v Hill.

One is that the mixture of factual determinations and evaluative moral sentiments or judgements is quite different when deciding if the mother country is now a ‘foreign power’ as distinct from deciding how many inroads into 100 per cent adult suffrage the legislature will be prevented from making based on the phrase ‘chosen by the people’. In my view, both of these involve a mix of the factual and the evaluative, of ‘is’ and ‘ought’, but the latter involves significantly more of the ‘ought’ and the evaluative.

Gleeson CJ obfuscates this by claiming that ‘“fact” [refers] to an historical development of constitutional significance’^[90] and then ‘of changed historical circumstances including legislative history.’^[91] On examination, however, the changed historical circumstances — the so-called facts — that matter in Roach are past decisions by High Court Justices (most importantly the implied rights cases) as well as past legislative changes. But these ‘facts’ are encapsulations of ‘ought’ judgements by judges and legislators. They are not observations about which country now controls Australia’s defence policy or foreign policy or whether Australia has its own embassies abroad.

Put differently, if Gleeson CJ’s view of what counts as a ‘fact’ is correct, and changes to these sort of ‘facts’ suffice to allow judges to give a different meaning to constitutional words — and so to give themselves a supervisory role over the elected Parliament when none existed before — then this amounts to constitutional amendment by statute or by past decision or by both together. As the dissenting judges make plain, and as I alluded to at the start, it is an odd understanding of how to give meaning to a constitutional text to think past legislation can alter the Constitution’s meaning. Indeed, that opens up the possibility of cynical bootstraps operations by both the Parliament and the judges to alter constitutional meaning (or rather ‘meaning’ in this odd Gleeson CJ sense) without the need for a s 128 referendum.

More bluntly put, the sort of ‘facts’ Gleeson CJ needs to rely on here are all ones that are just ethical and evaluative statutory and case law judgements by other political and judicial players (including a bit of glancing overseas to see what other jurisdictions’ value judgements today are about prisoner voting). These are overwhelmingly all ‘oughts’, some of which are masquerading as ‘is-es’ in the form of past statutes and cases.

So this attempted analogy fails in my view, leaving nothing convincing to support the Chief Justice’s conclusion. As with the joint judgment, this is an unpersuasive piece of reasoning to the conclusion that the Constitution grants a supervisory role to the top judges — one, notice, that seems in principle boundless and could expand further in future on this same reasoning — over these Roach-like issues. In fact, it is hard to see what constraints the majority rationes place on judges’ future supervisory powers other than ones the judges themselves feel inclined to observe.

In my opinion the dissenting judgments of Hayne J and Heydon J are far superior. Justice Hayne is correct when he asserts that

the Constitution does not establish a form of representative democracy in which the limits to the legislative power of the Parliament with respect to the franchise are to be found in a democratic theory which exists and has its content independent of the constitutional text.^[92]

He is correct when he observes that

[t]o impose upon the text and structure that was adopted a priori assumptions about what is now thought to be a desirable form of government or would conform to a pleasingly symmetrical theory of government is to do no more than assert the desirability of a particular answer to the issue that now arises.^[93]

He is correct that the plaintiff’s argument, accepted by the majority,

makes an assertion that the representative government criterion governing the qualification of electors must have a particular content. That assertion is not based on constitutional text or history and the argument thus becomes circular. The assertion of content determines the answer.^[94]

And Hayne J is correct that ‘[t]he meaning of constitutional standards does not vary with the level of popular acceptance that particular applications of the power might enjoy.’^[95]

On the oddity of Parliament being free to leave things as they are, but not free to liberalise and then later recant, Heydon J is correct that

[i]t would be surprising if the Australian Constitution operated so as to inhibit the capacity of the legislature, having changed the electoral laws in a particular way, to restore them to their earlier form if that change was found wanting in the light of experience.^[96]

And Heydon J is correct that

[t]he proposition that the legislative power of the Commonwealth is affected or limited by developments in international law since 1900 is denied by most, though not all, of the relevant authorities — that is, denied by twenty-one of the Justices of this Court who have considered the matter, and affirmed by only one.^[97]

Perhaps the worst aspect of all as regards these majority judgments is how potentially limitless and unconstrained they leave the supervisory role of the top judges. Here, today, it is which prisoners can vote. Next, on the reasoning of the majority, it could be almost anything else — however trifling — over which the judges can gainsay and overrule our elected Parliament.

And so it proved to be.

IV ROWE AND ROWING GENTLY DOWN THIS STREAM

Rowe is a 2010 4:3 High Court of Australia decision in which the majority invalidated further aspects of the 2006 Howard government Electoral Act reforms, this time relating to when the electoral rolls must close after the calling of an election. Between 1983 and 2006, the Electoral Act had provided a seven-day grace period to those people who were entitled to be enrolled — indeed those same people who had been legally obliged to lodge a claim for enrolment but, despite a penalty for failure to do so, had not done so before the issue of the election writs.^[98] The 2006 Act removed this seven-day grace period.^[99]

Put somewhat differently, only three years after the 2007 Roach decision, and Gleeson CJ’s preliminary paragraphs about how much faith the drafters and ratifiers of the Constitution had placed in the voters and the elected representatives of the people to decide contentious and debatable issues (including those articulated and framed in the language of rights), and about how much scope and room that Constitution had left for parliamentary sovereignty, and we now have in Rowe a decision in which the majority says the top unelected judges get to supervise when the electoral rolls will close. Or to be rather more accurate, the majority in Rowe asserts that the Constitution itself gives the High Court Justices a second-guessing or gainsaying or overruling power over the elected legislature as regards seven days and all the other minutiae surrounding the many competing incentives and disincentives involved in trying to get voters to enrol in a timely fashion. That is the meaning, supposedly, of a constitutional text which explicitly and clearly shunned a bill of rights, one that makes repeated references to ‘until the Parliament otherwise provides’^[100] and one where this claim about it supposedly meaning this rests only on four words of text, ‘chosen by the people’.^[101]

In my opinion Rowe is one of the worst decisions by the High Court of Australia in years, and by worst I mean most feebly reasoned and most reliant on implicit assumptions (including those about how to give meaning to a written text) that can never be explicitly cashed out in any convincing or persuasive way. And just to make myself clear, let me remind the reader what I said at the start, that were I able to legislate on a blank slate I would give the seven-day grace period. My criticism is directed towards the interpretation of the law, not one’s druthers if he or she could make it.

Whether you agree with that evaluation of mine, or not, notice that there are two distinct ways or bases on which to criticise the majority judgments in Rowe. They are quite distinct. One involves playing on the majority’s home field as it were. So it involves accepting that Roach was correctly decided, unlike the second path I will come to which does not. If you opt for the first path, however, you concede that Australia’s top judges have, or now have, a supervisory role — meaning, to be blunt, that they now have the power to invalidate or strike down Parliament’s legislation — over a host of voting-related issues. What falls under this judicial supervisory power, or new supervisory power, is of course somewhat uncertain as the majority’s reasoning in Roach has what might be thought of as a huge ‘penumbra of doubt’^[102] and a small ‘core of settled meaning’^[103] as far as indicating to where these newly enunciated judicial supervisory powers extend. In other words, why you reject the majority decision in Rowe (should you opt for this first path) is not because you say the Constitution gives no supervisory role to the top judges on these matters. For you, that pass has already been sold. Instead, you disavow Rowe simply because you say the majority Justices erred in performing that supervisory function. In particular, you say their proportionality analysis, denominated in whatever terms you prefer,^[104] misfired. The legislation under consideration in Rowe — after having been vetted and checked and supervised by the judges — ought to have been found acceptable.

This first basis for criticising the decision in Rowe is nothing other than an argument about how the proportionality analysis ought to have turned out. It amounts to arguing that Parliament (and the 2006 Act) ought to have been granted or afforded greater judicial dispensation or more of a margin of appreciation than was done by the majority.^[105]

Of course, travel down this first path for criticising the majority in Rowe, and not only do you accept Roach and its creation (or discovery) of a supervisory role for top judges in this area, you also have to ignore or gloss over or finesse the fundamentally unbounded or unconstrained or massive judicial discretion-enhancing nature inherent in all such proportionality analyses, as pointed out by Thomas Poole.^[106] And on top of that you are also forced — implicitly if not explicitly — to adopt an approach to constitutional interpretation that disavows completely all forms of originalism — of giving the words in the Constitution the meaning they were intended to have by the drafters (sub-branch one) or the meaning they would have been understood to have by the ratifiers at the time (sub-branch two). The Constitution, for you, becomes this metaphorical ‘living tree’ or ‘living constitution’ whose meaning changes over time (and so potentially locks in nothing) as determined by — and only by — a majority of High Court Justices at any point in time.

Yet even that is not all. For travel down this first path for criticising the majority and you also make it very likely that any and all future proportionality analyses will involve some looking overseas at other jurisdictions, almost all of which (as it happens) will be ones with a bill of rights. You may make some perfunctory remarks about how proportionality analyses here in Australia without a bill of rights fundamentally differ from ones in jurisdictions with such instruments,^[107] but that will not prevent you from citing and arguably relying on those jurisdictions and the judicial conclusions reached there. (As a strictly empirical matter, though, it may be that American case law and American resolutions of such things as when prisoners can vote will be quietly ignored.)

To put it bluntly, this first basis for criticising the majority in Rowe involves conceding so much justificatory and theoretical turf to them, that you end up playing all your games on the majority’s home ground. You lose before the kick-off (even if you are occasionally allowed to score the odd try or touchdown).

The other basis for rejecting the majority decision in Rowe is more principled, more coherent, and the only one with any long-term attraction or prospects. This second path is founded on an explicit rejection of Roach, on an assertion that Roach is bad law — for all the reasons given above and given in the two dissents there.

And if you have any doubts about how dependent upon Roach the majority judgments in Rowe are, consider this. The majority judgments cite Roach in twenty-seven different paragraphs.^[108] By my rough reckoning that means that over 10 per cent of all of the paragraphs in the majority judgments in Rowe cite or refer to Roach.

Of course if, like me, you believe that any persuasive criticism of the majority decision in Rowe presupposes — indeed demands — the concomitant assertion that Roach was and remains bad law, then you will be extremely critical of the Solicitor-General’s conduct of this Rowe case. That is because, to quote Heydon J:

The Solicitor-General of the Commonwealth assumed the correctness of the test advocated by the plaintiffs. It had two elements. The first turned on whether the impugned provisions amounted to legislative disqualification from adult suffrage. If so, then according to the second element, the disqualification could only be constitutionally valid if, in the words of three Justices in Roach v Electoral Commissioner, it were for a ‘substantial’ reason ... Even if the Solicitor-General was correct in assuming that the second element of this test is applicable to cases of the present kind, this is not a case of disqualification.^[109]

But the Solicitor-General ought not to have relinquished, permanently, home-field advantage by assuming the correctness of the plaintiff’s test, and hence of Roach itself (and hence should not have accepted the almost inevitable proportionality analysis — framed in terms of the need for the legislature to have ‘substantial’ reasons (as evaluated by the judges) for its statutory provisions — that will come with accepting Roach as good law).

Justice Heydon, though, dissents from the decision in Rowe and so his reference to what the Solicitor-General accepts does not constitute a plank or foundation in his reasoning. The same thing cannot be said when Justices in the majority in Rowe point to the Solicitor-General’s acceptance of Roach.^[110] All three majority judgments use that acceptance to arrive at their conclusion. Indeed the reliance on Roach is such that we can be brief in outlining those majority judgments.

Chief Justice French spends one paragraph and the first sentence of the next paragraph at the very start of his judgment^[111] answering in the affirmative the core question of whether the top judges have a supervisory role over Parliament when it comes to ‘[i]ndividual voting rights and the duties to enrol and vote.’^[112] A recital of the constitutional words ‘directly chosen by the people’ from ss 7 and 24, a citation to Roach, a bald assertion that Parliament needs to justify its decisions to the judges in this area,^[113] and the rest is really just proportionality analysis and deciding ‘[i]f the law’s adverse legal or practical effect upon the exercise of the entitlement to vote is disproportionate to its advancement of the constitutional mandate’.^[114]

Seventy-seven or so paragraphs later and the Chief Justice has decided that removing the seven-day grace period is disproportionate or too lacking in a rational connection to some legitimate governmental aim or however you want to phrase the ‘discretionary judgment that can be massively broad or incredibly narrow — and anything else between’^[115] that Thomas Poole argues lies at the heart of all such determinations.

Along the way the Chief Justice asserts:

1 that the ‘content of the constitutional concept of “chosen by the people” has evolved since 1901 and is now informed by the universal adult-citizen franchise’;^[116]

2 that the ‘common understanding of the time’^[117] is ‘not to be equated to judicial understanding’;^[118] and

3 that ‘Parliament has a considerable discretion as to the means which it chooses to regulate elections and to ensure that persons claiming an entitlement to be enrolled are so entitled.’^[119]

In response, of course, one might note:

1 that the claimed evolving nature of constitutional concepts is all the result of High Court case law that gradually has afforded our top judges ever more gainsaying and overruling and supervisory powers over Parliament;

2 that any time the High Court asserts that Parliament, having passed a law to liberalise some state of affairs, is sometimes constitutionally barred from later recanting and passing a new law to take things back to the way they were — precisely what the majority held in both Roach and Rowe — then it comes close to being disingenuous to assert that judicial understandings are not determinative; and

3 that the considerable discretion the Chief Justice assures us that Parliament has is not considerable enough to leave it to Parliament to decide if the electoral rolls close immediately after the calling of an election, or seven days later.

The joint majority judgment of Gummow and Bell JJ is likewise preoccupied with asking the ‘is this reasonably appropriate or rationally connected to a legitimate aim?’ question rather than the ‘is this any of our constitutionally allocated business?’ question.^[120] The main work in that joint judgment is done from paragraph 150 on and begins with the by now unsurprising reference to what the Commonwealth accepts (meaning to what the Solicitor-General has conceded), and the reference to determining this case ‘in accordance with the reasoning in Roach.’^[121]

Again, and after placing huge reasoning-weight on that Roach decision, the joint judgment follows the Chief Justice in deciding that the 2006 Act fails the ‘rational connection’ or ‘substantial reason’ or ‘reasonably appropriate and adapted’ or ‘proportionality’ test.^[122]

The last of the majority judgments is Justice Crennan’s. As Graeme Orr notes,

she devote[s] six pages [of her judgment] to a thesis that 19^th century Australian colonial practice represented a triumph of egalitarian political values over an oligarchic British inheritance. (This thesis reprised a speech she gave to the 2008 Constitutional Law Dinner in Sydney ... )^[123]

All of this will strike many readers as pretty rambling stuff, and worse aimed at

supporting the living tree approach to constitutional rights ... [T]he problem with such historicism is that the late 19^th century Australian settlement of the franchise only extended to universal male suffrage. It was ... not necessarily free of gender discrimination and certainly not free of racial exclusions.^[124]

Other than that, plus a seeming concatenation of the ‘impolitic’^[125] into the ‘unconstitutional’ together with a digression about the Senate,^[126] and this is all much of a muchness with the other majority judgments — lots of Roach, mentions of what the Solicitor-General conceded,^[127] and the inevitable proportionality analysis coming down against the 2006 amendments.

As for the dissents, I will not linger over them other than to note three things. Firstly, Kiefel J’s is highly unsatisfactory to the extent it is understood as immersing itself in proportionality analysis and so playing the game on the majority’s home turf.

Secondly, Heydon J’s judgment is not only powerful, it betrays real anger at where the majority judgments are taking constitutional interpretation in Australia. We even get a subtle questioning from Heydon J of whether the older implied rights cases like Lange were rightly decided,^[128] though I would prefer this to be explicit and vigorous in the light of the foundations they are now providing for cases such as Roach and Rowe. And Heydon J also laments, self-mockingly, the lack of support for his (and my) preferred originalist approach to constitutional interpretation.^[129]

Thirdly, Hayne J’s dissent — while remaining true to what he argued in Roach — distinguishes what is involved in Rowe from that earlier case. At the same time Hayne J counters the majority by arguing that the evolutionary expansion of the franchise in Australia was democratically-driven, not judicially mandated or overseen,^[130] with the majority too easily ignoring Gibbs J’s warning in McKinlay not ‘to add to the Constitution’s provisions “new doctrines which may happen to conform to our own prepossessions”.’^[131] Hayne J even manages to finish with a powerful final paragraph, repeating his position from Roach, that

[t]he ambit of relevant constitutional powers is not set by the political mood of the time, or by what legislation may have been enacted in the exercise of the powers. Political acceptance and political acceptability have no footing in established doctrines of constitutional interpretation.^[132]

I agree. And with that I have now said enough to move to the final Part of this paper, and my concluding remarks on judicial activism.

V CONCLUDING REMARKS ON JUDICIAL ACTIVISM

As I conceded at the very start of this paper, there is no agreement as to what does or does not constitute judicial activism. There is considerably more ‘penumbra of doubt’ than ‘core of settled meaning’^[133] when it comes to what sorts of judicial behaviour do and do not fall under the aegis of this sin.

That said, a willingness to adopt an interpretive approach to giving meaning to our Constitution that puts few — if any — external constraints on what outcome the judges can reach is as solid a candidate for attracting the label ‘judicial activism’ as any going. And this is precisely what the High Court Justices have done in Roach and Rowe. The interpretive constraints inherent in the reasoning of the majority decisions in those two cases are almost all in the nature of ‘this is an evolving document whose changing meaning we seven judges (or a majority of us) will announce as time goes by, based on whether we consider the challenged legislation to be based on “substantial” or “proportional” or “non-arbitrary” reasons.’ Put more bluntly, the constraints on what the top judges can do are almost wholly self-imposed; they have next to no connection to external factors such as the actual words and text of the Constitution or (just as importantly) the meaning those words had for the real life people who ratified them and drafted them.

I will limit myself to making four related points that I think are all relevant to this charge of mine that the High Court has been indulging in judicial activism in both the Roach and Rowe cases.

The first amounts to a qualification or refinement of that critique of judicial activism. And it is this: An activist judge’s decisions might well be ‘principled’, in the sense of fitting into a coherent political philosophy. But that is not the sin that is being attributed to that judge. No, what is being alleged is that the Constitution — the real-life one at hand, not some preferred alternative — does not authorise the judge to implement this coherent political philosophy of his or hers. It is irrelevant, then, how coherent, principled, attractive or desirable the activist judge’s political philosophy may or may not be. The sin is not one of being unprincipled; the sin is that the laid-down legal instrument did not authorise what the judge decided.

The second point is related to that first one. Judicial activism, as just set out, undermines the rule of law in its old-fashioned procedural sense of the desirability of being governed by the use of general rules, known in advance, that allow citizens to shape their expectations and so guide their behaviour.^[134] But that understanding of the rule of law requires the rules to have some pre-existing content. They have to lock in the point-of-application judicial interpreter as well as the other 99.99 per cent of us. Metaphors about constitutions being ‘living trees’ merely obscure the fundamental choice we have when we forswear New Zealand-style parliamentary sovereignty and opt for a written constitution. Either we will be locked in by the understandings of the words and text at the time of adoption (subject to s 128 constitutional amendment) or we will be locked in by the decisions of our present-day top judges as they, from time to time, change the meanings they attribute to words that have remained the same.

That latter option, the one that permeates Roach and Rowe and the earlier implied rights cases, not only presupposes that a handful of top judges will be good at identifying the changing social values and mores that drive this ‘living tree’ type of interpretation, it assumes they are better at it than the elected representatives of the people who would otherwise be deciding when prisoners can vote or electoral rolls can close.

And that is without even taking into account the potential further negatives that this sort of interpretive approach has the potential to politicise the judiciary (and to be seen to do so by citizens) and too often to circumvent or make redundant the s 128 amending machinery.

My third point involves asking the reader momentarily to shift his or her perspective from a citizen in a democracy with a well-established written constitution (like Australia) to that of a citizen in a successful democracy based on parliamentary sovereignty (like New Zealand). I have made a chapter-length argument to this effect in a recently published book,^[135] but it amounts to this. Would any New Zealander today sign up for an Australian-style written constitution if you told that Kiwi that the words of that constitutional text — the ones that were being debated, refined and fought over — would be given new and shifting meanings, over time, by the unelected top judges? Why would the New Zealander (who today has a vote to choose someone to resolve prisoner voting and the like, and who thinks his or her newly mooted written constitution — by forswearing a bill of rights — leaves such issues to future voters) opt for such an instrument when he or she learns that eight or nine decades down the road certain implied rights will be ‘discovered’ somewhere in the structure of that constitution, and then two decades later on after that some new, further judicial powers and supervisory functions related to voting matters also will be discovered?

My view is that if the Roach and Rowe interpretive approaches were spelt out, clearly and in advance, to people living in a parliamentary sovereignty democracy who were considering whether to adopt a written constitution, they would overwhelmingly reject it and many would do so precisely because of this externally unconstrained interpretive approach you were spelling out in advance (which may be why it never is spelt out in advance).

If you doubt that, or if you think that my mooted change of perspective exercise is irrelevant to present-day Australians, consider the possibility that a rewritten preamble to our Constitution may soon be put to the electors, perhaps to recognise the role of Indigenous Australians. After Roach and Rowe what form of words — however clear — would ever leave you confident latter-day judges might not inflate them or redirect them or apply them to some purpose neither you nor any other people voting ‘yes’ in a s 128 referendum (indeed none of those involved in drafting the words either) intended?

After Roach and Rowe I simply do not see how anyone could be remotely confident that a newly worded preamble, however circumscribed and limited and clearly bounded its intent, might not be used by future judges to give themselves yet further supervisory powers over the elected Parliament.

And returning to present-day New Zealanders, here is a further seeming anomaly. In 2010 the New Zealand Parliament enacted the Electoral (Disqualification of Sentenced Prisoners) Amendment Act 2010 (NZ). This statute had to do with prisoner voting. Prior to this 2010 Act, prisoners in New Zealand were unable to vote if they were in jail serving sentences of three years or more. The 2010 Act removes (prospectively only) the vote, the franchise, from all prisoners (however long their sentences) provided they have been convicted and are serving that sentence at the time of the election.

In broad terms, this New Zealand Act is similar to the Howard government’s 2006 amendments that the High Court struck down or invalidated in Roach. And the implication, the clear implication, must be that the majority Justices in Roach would not consider the New Zealand Parliament of today to be one that has been ‘chosen by the people’. (And if anyone is tempted to counter this claim by arguing that the Roach case was specific to Australia and Australia’s constitutional text, the clear rejoinder I would make is that virtually all of the majority reasoning in Roach relies on a free-standing ethical or political argument about what is politically acceptable or rights-respecting, the text playing little determinative role compared to reliance on the implied rights cases, cherry-picked overseas cases, and the just-mentioned free-standing ethical arguments.)

As I mentioned above in relation to the Canadian Chief Justice in Sauvé and her comments about ‘self-proclaimed democracies’,^[136] (meaning all the then democracies like the US, UK, New Zealand and Australia that put restrictions on prisoner voting), there is a danger that judges who draw these sorts of contestable and debatable policy lines can get a tad puffed up. They can implicitly be read as thinking they have superior moral antennae to voters and politicians. Their reasoning can come close to implying such absurdities as that New Zealand’s Parliament, post-2010 elections, has not been chosen by the people.

My fourth and final point simply raises a hypothetical question. Could Parliament today — post Roach and Rowe — reverse itself and take away, say, Senate representation from the territories? Western Australia v Commonwealth (‘First Territory Senators Case’),^[137] you will recall, was a 4:3 decision in which the High Court, weighing the seemingly conflicting demands of ss 7 and 122, decided that the Constitution had left the question of Senate representation for the territories to Parliament. It was a matter for the political process.

However, after Roach and Rowe, and given the reasoning employed by the majorities in those two cases, it seems quite likely to me that our High Court would now assert (on no textual basis whatsoever other than the all-purpose ‘chosen by the people’ passage) that they — the judges — had a supervisory role over that issue too. What would prevent removal of this territory representation was not the political good sense of the people but the keen supervisory eye of the top judges. If so, this hypothetical would see us moving from the plausible position that the Constitution gave Parliament no power to give such Senate representation to the territories (the position of three of seven High Court Justices in that First Territory Senators Case), through the position that it was a matter that had been left to Parliament and the political process to decide (the position of the four majority Justices there), on to a new position that our High Court would never let Parliament change its mind on this matter (a clear possibility after Roach and Rowe).

Most of us may well wish the territories to have Senate representation. Certainly I do. But it is a reductio ad absurdum of our High Court’s recent judicial activism to believe the Constitution gives our top judges a supervisory power to prevent such a legislative change of mind. It is a manifestation of the unbounded, unconstrained approach to constitutional interpretation that today passes for orthodoxy on our High Court.

What is needed, and needed badly in Australia, is a Solicitor-General who does not accept Roach and Rowe to be good law. That, and the hope that future High Court decisions and Justices may reconsider the merits (such as they are) of the reasoning underlying these two cases, are badly needed.

Judicial activism has all sorts of long-term bad consequences not immediately obvious to many people who happen to find congenial the outcomes of particular decisions aligning with their own first-order political preferences. Such short-sightedness can be costly, most obviously in terms of the long-term inroads it makes into Australia’s wonderful democratic traditions.

^[*] BA, LLB (Queen’s), LLM (LSE), PhD (Hong Kong); Garrick Professor of Law, T C Beirne School of Law, The University of Queensland. The author wishes to thank Gabrielle Appleby, Andrew Geddis, Grant Huscroft, Helen Irving, Daniel Meagher and Graeme Orr, and those who attended the 2011 Samuel Griffith Society Conference, where an earlier, less developed version of this paper was presented for their comments, criticisms and suggestions.

^[1] It may even be what W B Gallie coined an ‘essentially contested concept’: W B Gallie, ‘Essentially Contested Concepts’ (1956) 56 Proceedings of the Aristotelian Society 167.

^[2] H L A Hart first coined the term ‘penumbra of doubt’ when discussing a rule’s scope or coverage: H L A Hart, The Concept of Law (Clarendon Press, first published 1961, 1994 ed) 123.

^[3] Again, a term or idea coined by Hart: ibid 94.

^[4] (2007) 233 CLR 162.

^[5] [2016] FCA 1081; (2010) 243 CLR 1.

^[6] These are the cases starting with Australian Capital Television Pty Ltd v Commonwealth (1992) 177 CLR 106 and Nationwide News Pty Ltd v Wills [1992] HCA 46; (1992) 177 CLR 1 and ending with Lange v Australian Broadcasting Corporation [1997] HCA 25; (1997) 189 CLR 520.

^[7] See, eg, James Allan, ‘Implied Rights and Federalism: Inventing Intentions while Ignoring Them’ (2009) 34 University of Western Australia Law Review 228; James Allan, ‘Paying for the Comfort of Dogma’ [2003] SydLawRw 4; (2003) 25 Sydney Law Review 63. See also James Allan, ‘A Defence of the Status Quo’ in Tom Campbell, Jeffrey Goldsworthy and Adrienne Stone (eds), Protecting Human Rights: Instruments and Institutions (Oxford University Press, 2003) 175.

^[8] For a good account of this approach, see Bradley W Miller, ‘Beguiled by Metaphors: The “Living Tree” and Originalist Constitutional Interpretation in Canada’ (2009) 22 Canadian Journal of Law and Jurisprudence 331.

^[9] (1992) 177 CLR 106.

^[10] Constitution ss 7, 24. And note that this reference to actual words in the Constitution does almost all the textual work in the reasoning process, one that ends with an assertion that the document contains an implied right to freedom of political communication. The Chief Justice also refers briefly to representative and responsible government, and so to ss 1, 61–2, and 128: ibid 137–8, but this sort of ‘argument from abstract principles’ is even less dependent on the actual text and words.

^[11] This five-step reasoning process is most clearly seen in ACTV (1992) 177 CLR 106, 138.

^[12] For a discussion of my ‘thin’ understanding of the rule of law concept, which takes issue with a much ‘thicker’, morally pregnant understanding, see James Allan, ‘Reasonable Disagreement and the Diminution of Democracy: Joseph’s Morally Laden Understanding of “the Rule of Law”’ in Richard Ekins (ed), Modern Challenges to the Rule of Law (LexisNexis NZ, 2011) 79.

^[13] See, eg, Allan, ‘Implied Rights and Federalism: Inventing Intentions while Ignoring Them’, above n 7, 230–3.

^[14] The principle was settled by the Court in Lange v Australian Broadcasting Corporation [1997] HCA 25; (1997) 189 CLR 520.

^[15] Roach (2007) 233 CLR 162, 202 [95] (Gummow, Kirby and Crennan JJ). Chief Justice Gleeson uses different terminology based on whether the legislation is deemed by the judges to be ‘arbitrary’: at 182 [23]. However, this is just as much a proportionality-type assessment, though one with a seemingly different fault line for second-guessing the legislature. At least sometimes. At least maybe.

^[16] Constitution ss 7, 24.

^[17] Roach (2007) 233 CLR 162, 204 [101] (Gummow, Kirby and Crennan JJ).

^[18] Ibid 178 [16] (Gleeson CJ). The Chief Justice also frames it in terms of the need for the elected parliament to have a ‘substantial reason’ for disenfranchising people and in terms of a ‘rational connection’: at 174–5 [7]–[8].

^[19] Rowe [2016] FCA 1081; (2010) 243 CLR 1, 131–2 [424]–[430] (Kiefel J).

^[20] Thomas Poole, ‘The Reformation of English Administrative Law’ (2009) 68 Cambridge Law Journal 142, 146. See also Dan Meagher, ‘The Brennan Conception of the Implied Freedom: Theory, Proportionality and Deference’ [2011] UQLawJl 8; (2011) 30 University of Queensland Law Journal 119, 122–6.

^[21] Roach (2007) 233 CLR 162, 186 [40].

^[22] Ibid 186 [43] (emphasis added).

^[23] [1997] HCA 25; (1997) 189 CLR 520, 557 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ), cited in ibid 186 [44].

^[24] Roach (2007) 233 CLR 162, 186–7 [45].

^[25] For a very recent and comprehensive look at originalism from a variety of leading exponents, sceptics and critics, see Grant Huscroft and Bradley W Miller (eds), The Challenge of Originalism: Theories of Constitutional Interpretation (Cambridge University Press, 2011).

^[26] For an Australian defence of ‘living tree’ interpretation, see Justice Michael Kirby, ‘Constitutional Interpretation and Original Intent: A Form of Ancestor Worship?’ [2000] MelbULawRw 1; (2000) 24 Melbourne University Law Review 1.

^[27] See Allan, ‘A Defence of the Status Quo’, above n 7, 181.

^[28] For a recognition of this, see the final sentence of Roach (2007) 233 CLR 162, 186–7 [45] (Gummow, Kirby and Crennan JJ).

^[29] The ‘best fit, Herculean’ interpretive theory was first set out in Ronald Dworkin, Taking Rights Seriously (Duckworth, 4^th revised ed, 1984) and then reworked in Ronald Dworkin, Law’s Empire (Belknap, 1986).

^[30] Roach (2007) 233 CLR 162, 187 [46] (emphasis added). Constitution s 51(xxxvi) gives Parliament power to make laws over ‘matters in respect of which this Constitution makes provision until the Parliament otherwise provides.’

^[31] Roach (2007) 233 CLR 162, 187 [46].

^[32] Ibid 187 [47].

^[33] Ibid.

^[34] Ibid 187 [46].

^[35] Ibid 187 [48] (emphasis added).

^[36] Ibid.

^[37] Ibid 187–8 [49] (emphasis added). The insistence on direct choice may be contrasted with, say, the electoral college system used in presidential elections in the United States.

^[38] Ibid 188 [49].

^[39] See, eg, Constitution ss 3, 7, 10.

^[40] A proposal for a bill of rights was put forward, but did not carry: see Justice Michael Kirby, ‘A Bill of Rights for Australia’ (Speech delivered at the Parliament House, Brisbane, Young Presidents’ Association, 4 October 1994); Nicholas Aroney, ‘A Seductive Plausibility: Freedom of Speech in the Constitution’ [1995] UQLawJl 6; (1995) 18 University of Queensland Law Journal 249, 252.

^[41] Ibid. A recent committee report recommended that ‘Australia adopt a federal Human Rights Act’: National Human Rights Consultation Committee, National Human Rights Consultation — Report (2009) xxxiv <http://www.humanrightsconsultation.gov.au/Report/Documents/Recommendations.pdf> . However, this recommendation was subsequently rejected by the Commonwealth government.

^[42] I have argued that such federalist judicial review is structurally quite distinct from rights-based judicial review. See James Allan, ‘Not in for a Pound — In for a Penny? Must a Majoritarian Democrat Treat All Constitutional Judicial Review as Equally Egregious?’ (2010) 21 King’s Law Journal 233.

^[43] There are larger separation of powers problems or concerns here with a Solicitor-General purporting to concede constitutional meaning, opening up the possibility of the executive in effect colluding with the courts to diminish legislative authority. And if the Solicitor-General’s concessions merely articulate the current executive’s constitutional position or vision, then the judges’ invocation of that position is purely a makeweight.

^[44] Roach (2007) 233 CLR 162, 189 [53].

^[45] Ibid 197 [78].

^[46] [2004] HCA 41; (2004) 220 CLR 181, cited in ibid 197 [77].

^[47] [1996] HCA 48; (1996) 186 CLR 140, discussed in Roach (2007) 233 CLR 162, 197–8 [78]–[83].

^[48] [1997] HCA 25; (1997) 189 CLR 520, cited in Roach (2007) 233 CLR 162, 198 [80].

^[49] Roach (2007) 233 CLR 162, 198 [82].

^[50] Nicholas Aroney lists some of the possible dicta available here — some rejecting an implied right of this sort and some accepting it: Nicholas Aroney, ‘Towards the “Best Explanation” of the Constitution: Text, Structure, History and Principle in Roach v Electoral Commissioner’ [2011] UQLawJl 10; (2011) 30 University of Queensland Law Journal 145, 156 n 59. Not only that, he also points out that ‘the joint judgment was highly selective in its use of the cases. ... [T]his, it needs to be recalled, is only the judgment of one justice in one particular case’: at 155–6, referring to the Brennan CJ judgment cited by the joint judgment in Roach.

^[51] Ibid 155.

^[52] Roach (2007) 233 CLR 162, 198 [83], quoting McGinty [1996] HCA 48; (1996) 186 CLR 140, 170. Again, as Aroney points out, Brennan CJ ‘was not actually affirming that “chosen by the people” requires that the franchise extend to all adult citizens subject only to reasonable exclusions’: ibid 155 (emphasis in original). This was one possibility that might be ascribed to the phrase. The careful language of ‘admitting of a requirement’ in the joint judgment might be thought to obscure that distinction.

^[53] Roach (2007) 233 CLR 162, 186 [40], reciting the plaintiff’s fourth ground for challenging the impugned statute, which was the one accepted by the joint judgment.

^[54] Ibid 199 [85].

^[55] Ibid 199–204 [84]–[102].

^[56] Poole, above n 20, 146.

^[57] Electoral and Referendum Amendment (Prisoner Voting and Other Measures) Act 2004 (Cth).

^[58] [2002] 3 SCR 519, cited in Roach (2007) 233 CLR 162, 203 [100].

^[59] [2005] IX Eur Court HR 187, cited in Roach (2007) 233 CLR 162, 203–4 [100].

^[60] (1993) 2 HRNZ 358. For more on the cherry-picking nature of rights adjudication, see James Allan, Grant Huscroft and Nessa Lynch, ‘The Citation of Overseas Authority in Rights Litigation in New Zealand: How Much Bark? How Much Bite?’ [2007] OtaLawRw 7; (2007) 11 Otago Law Review 433.

^[61] See Roach (2007) 233 CLR 162, 200 [89], 202 [95]. See also Aroney, above n 50, 157–9.

^[62] The Chief Justice of Canada, writing the 5:4 majority judgment in Sauvé, talked of jurisdictions that disagreed with her view on which prisoners should be able to vote as being ‘self-proclaimed democracies’: [2002] 3 SCR 519, 548 [41] (McLachlin CJ for McLachlin CJ, Iacobucci, Binnie, Arbour and LeBel JJ) referring to the countries discussed in the dissent: at 588 [125], 591–2 [130]–[131] (Gonthier J for L’Heureux-Dubé, Gonthier, Major and Bastarache JJ), which impliedly at the time meant that the Chief Justice was referring to such jurisdictions as Australia, the US, New Zealand and the UK, an astonishingly self-satisfied (and patently wrong) implication or view to hold.

^[63] Roach (2007) 233 CLR 162, 172–3 [1]–[6].

^[64] Ibid 172 [1].

^[65] Ibid.

^[66] [1975] HCA 53; (1975) 135 CLR 1, 24, quoted in ibid 172 [2]. For my argument that federalist-based judicial review is far less democratically objectionable than rights-based judicial review, see Allan, ‘Not in for a Pound — In for a Penny?’, above n 42.

^[67] Roach (2007) 233 CLR 162, 173 [4], citing Mulholland v Australian Electoral Commission [2004] HCA 41; (2004) 220 CLR 181, 188 [6] (Gleeson CJ).

^[68] Roach (2007) 233 CLR 162, 173 [6].

^[69] [2006] HCA 52; (2006) 229 CLR 1. For an argument that the High Court of Australia has a terrible record in upholding federalism constraints, see James Allan and Nicholas Aroney, ‘An Uncommon Court: How the High Court of Australia Has Undermined Australian Federalism’ [2008] SydLawRw 15; (2008) 30 Sydney Law Review 245. For what it is worth, a similar sentiment might apply to Gleeson CJ’s reliance on federalist concerns: ibid 176 [10].

^[70] Roach (2007) 233 CLR 162, 175–82 [9]–[25].

^[71] The same points I make above as regards the joint judgment about the ‘cherry-picking of precedents you will use’ nature of this sort of analysis, about the flexibility one has to characterise legislation in a way that makes it easier to reach a desired conclusion, and about Thomas Poole’s point as to how this is essentially an unconstrained, plastic and undesirable form of reasoning, all apply here as well: see Poole, above n 20, 146.

^[72] Roach (2007) 233 CLR 162, 173 [6].

^[73] Larry Alexander puts this originalist point as clearly as anyone, that the whole point of a written constitution is to lock in certain outcomes — maybe a federalist division of powers, maybe bicameralism, maybe a specifically enumerated list of rights — and so to freeze those outcomes or locked-in political settlements subject only to constitutional amendment. Everything not so locked-in (say as regards rights) falls under the aegis of legislative sovereignty: Larry Alexander, ‘Introduction’ in Larry Alexander (ed), Constitutionalism: Philosophical Foundations (Cambridge University Press, 1998) 1. See also Richard S Kay, ‘American Constitutionalism’ in Larry Alexander (ed), Constitutionalism: Philosophical Foundations (Cambridge University Press, 1998) 16; Larry Alexander, ‘Simple-Minded Originalism’ in Grant Huscroft and Bradley W Miller (eds), The Challenge of Originalism: Theories of Constitutional Interpretation (Cambridge University Press, 2011) 87. The only alternative to being locked in by original intent or original understanding, of course, is to be locked in by present-day judges’ views of what the metaphorically changing constitution means.

^[74] In some US states, convicted felons are barred from voting even after leaving prison, indeed in some states the bar is for life.

^[75] See, eg, Sauvé [2002] 3 SCR 519, discussed at above nn 58–62 and accompanying text.

^[76] Roach (2007) 233 CLR 162, 173 [6].

^[77] Ibid.

^[78] James Allan, ‘The Paradox of Sovereignty: Jackson and the Hunt for a New Rule of Recognition?’ (2007) 18 King’s Law Journal 1, 12, quoting Hart, above n 2, 117.

^[79] Roach (2007) 233 CLR 162, 174 [8].

^[80] James Allan and Michael Kirby, ‘A Public Conversation on Constitutionalism and the Judiciary between Professor James Allan and the Hon Michael Kirby AC CMG’ [2009] MelbULawRw 36; (2009) 33 Melbourne University Law Review 1032, 1051 (citations omitted).

^[81] Roach (2007) 233 CLR 162, 174 [7].

^[82] Ibid 173 [6].

^[83] [1999] HCA 30; (1999) 199 CLR 462.

^[84] Roach (2007) 233 CLR 162, 173–4 [6].

^[85] [1975] HCA 53; (1975) 135 CLR 1, 36, quoted in ibid 174 [7].

^[86] Roach (2007) 233 CLR 162, 174 [7], citing McGinty [1996] HCA 48; (1996) 186 CLR 140, 286–7.

^[87] Roach (2007) 233 CLR 162, 174 [7].

^[88] Ibid (emphasis added).

^[89] Alexander, ‘Simple-Minded Originalism’, above n 73.

^[90] Roach (2007) 233 CLR 162, 174 [7].

^[91] Ibid.

^[92] Ibid 214 [142].

^[93] Ibid 215 [142].

^[94] Ibid 215 [145].

^[95] Ibid 219 [159].

^[96] Ibid 224 [180].

^[97] Ibid 225 [181] (citations omitted).

^[98] There was also a grace period for those who had failed to transfer their enrolment.

^[99] From the 1930s to the commencement of the Commonwealth Electoral Legislation Amendment Act 1983 (Cth) there had been no statutory grace period, but the executive informally did as much by announcing an election date but delaying the issue of the electoral writ. And indeed, despite the Howard government enacting the 2006 Act, Prime Minister Howard allowed several days of non-legislated grace before the 2007 election by announcing the election on a Sunday with the writs to follow on the coming Wednesday.

^[100] See above n 39 and corresponding text.

^[101] Constitution ss 7, 24. The preceding word ‘directly’ plays no role in the majority reasoning. Of course, past case law plays a huge role in getting to this conclusion but constitutional text does not.

^[102] See Hart, above n 2, 123.

^[103] Ibid 144.

^[104] Perhaps in terms of ‘substantial reason[s]’ or ‘rational connections’: Rowe [2016] FCA 1081; (2010) 243 CLR 1, 59 [161] (Gummow and Bell JJ). Or in terms of ‘practical effect’: at 38 [78] (French CJ).

^[105] Although aspects of her dissenting judgment do not travel down this ‘the majority got its proportionality analysis wrong’ path, large parts of Justice Kiefel’s dissent do: see Rowe [2016] FCA 1081; (2010) 243 CLR 1, 131–47 [424]–[489].

^[106] See Poole, above n 20 and accompanying text.

^[107] See, eg, Roach (2007) 233 CLR 162, 178–9 [17] (Gleeson CJ), cited in Rowe [2016] FCA 1081; (2010) 243 CLR 1, 59 [162] (Gummow and Bell JJ).

^[108] See Rowe [2016] FCA 1081; (2010) 243 CLR 1, 12 [1], 18–19 [20], 19–21 [23]–[25], 26 n 99 (French CJ), 40 [86], 46 [117], 48 [123], 57 [151], 57 [154], 58 [157], 58–9 [160]–[162] (Gummow and Bell JJ), 105 [323], 106–7 [325]–[327], 108 [332], 117 [366], 118 [372]–[374], 119 [376], 119 [381], 120 [384] (Crennan J). Nine of these references to Roach have to be looked for in the footnotes.

^[109] Ibid 94 [283] (Heydon J). See also my general points about the role of the Solicitor-General when making concessions to the plaintiffs: above n 43.

^[110] ‘The Commonwealth ... accepts that ... the consistency of [an impugned] law with ss 7 and 24 of the Constitution is to be determined in accordance with the reasoning in Roach’: Rowe [2016] FCA 1081; (2010) 243 CLR 1, 56–7 [151] (Gummow and Bell JJ); see also at 36–7 [73] (French CJ), 59 [162] (Gummow and Bell JJ), 119 [376] (Crennan J).

^[111] Ibid 12 [1]–[2].

^[112] Ibid 12 [1].

^[113] Ibid 12 [1]–[2].

^[114] Ibid 12 [2] (emphasis added).

^[115] Poole, above n 20, 146.

^[116] Rowe [2016] FCA 1081; (2010) 243 CLR 1, 18 [18].

^[117] Ibid, quoting McKinlay [1975] HCA 53; (1975) 135 CLR 1, 36 (McTiernan and Jacobs JJ).

^[118] Rowe [2016] FCA 1081; (2010) 243 CLR 1, 18 [19].

^[119] Ibid 22 [29].

^[120] See, eg, ibid 45 [111], 59 [161].

^[121] Ibid 57 [151].

^[122] All of these variations on a theme: ibid 59 [161] (Gummow and Bell JJ).

^[123] Graeme Orr, ‘The Voting Rights Ratchet: Rowe v Electoral Commissioner’ (2011) 22 Public Law Review 83, 87.

^[124] Ibid (citations omitted).

^[125] Rowe [2016] FCA 1081; (2010) 243 CLR 1, 110 [339].

^[126] Ibid 116 [363].

^[127] Ibid 118–19 [375]–[376].

^[128] ‘Even on the assumption that [the Lange test] operates satisfactorily in that field’: ibid 96 [291].

^[129] Ibid 97–9 [292]–[302]. See especially his Honour’s comments: at 97 [293].

^[130] Ibid 70–2 [201]–[204].

^[131] Ibid 72 [204], quoting McKinlay [1975] HCA 53; (1975) 135 CLR 1, 44 (Gibbs J).

^[132] Rowe [2016] FCA 1081; (2010) 243 CLR 1, 89 [266].

^[133] See above nn 102–3.

^[134] I defend this conception of the rule of law in Allan, ‘Reasonable Disagreement and the Diminution of Democracy’, above n 12.

^[135] See James Allan, ‘The Curious Concept of the “Living Tree” (or Non-Locked-In) Constitution’ in Grant Huscroft and Bradley W Miller (eds), The Challenge of Originalism: Theories of Constitutional Interpretation (Cambridge University Press, 2011) 179.

^[136] See above n 62.

^[137] [1975] HCA 46; (1975) 134 CLR 201.