Roach v Electoral Commissioner [2007] HCA 43 | Case summary

Roach v Electoral Commissioner [2007] HCA 43 | Case summary | Zoe

Sir Owen Dixon found a need to explain to American lawyers the scarcity in the Australian Constitution of formal guarantees of rights and freedoms which they associated with the idea of "constitutional rights". That is not to say that the Constitution contains no guarantees or protections of individual rights, express or implied. Yet it reflects a high level of acceptance of what Barwick CJ called "the notion of the sovereignty of Parliament in the scheme of government". 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 democracy[4].
Important features of our system of representative democracy, such as compulsory voting, election of members of the House of Representatives by preferential voting, and proportional representation in the Senate, are the consequence of legislation, not constitutional provision. One striking example concerns a matter which the framers deliberately left to be dealt with by Parliament: female suffrage. The Constitution, in s 128, refers to States "in which adult suffrage prevails." In 1901, adult suffrage meant the franchise for women as well as men. Quick and Garran, referring to the Convention Debates, noted "the difficulty as to women's suffrage" which was taken into account in the wording of s 128[5]. Another example is voting by Aboriginal people, which remained an issue not fully resolved until the second half of the twentieth century.
The combined effect of ss 51(xxxvi), 8 and 30 is that Parliament may make laws providing for the qualification of electors. That Australia came to have universal adult suffrage was the result of legislative action. Universal suffrage does not exclude the possibility of some exceptions. The Oxford English Dictionary says that the term means "the right of all adults (with minor exceptions) to vote in political elections."[6] Among countries which now have universal suffrage there are observable differences in the exceptions that are accepted, but there is also a broad agreement as to the kinds of exception that would not be tolerated. 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, which require that the senators and members of the House of Representatives be "directly chosen by the people" of the State or the Commonwealth respectively. In 1901, those words did not mandate universal adult suffrage. In 1901, the words "foreign power" in s 44(i) did not include the United Kingdom, yet in Sue v Hill[7] this Court held that, by reason of changes in Australia's relations with the United Kingdom and in national and international circumstances over the intervening period, they had come to include the United Kingdom. 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.
In McKinlay[8], McTiernan and Jacobs JJ said that "the long established universal adult suffrage may now be recognized as a fact". I take "fact" to refer to an historical development of constitutional significance of the same kind as the developments considered in Sue v Hill. Just as the concept of a foreign power is one that is to be applied to different circumstances at different times, McTiernan and Jacobs JJ said that the words "chosen by the people of the Commonwealth" were to be applied to different circumstances at different times. Questions of degree may be involved. They concluded that universal adult suffrage was a long established fact, and that anything less could not now be described as a choice by the people. I respectfully agree. As Gummow J said in McGinty v Western Australia[9], we have reached a stage in the evolution of representative government which produces that consequence. I see no reason to deny that, in this respect, and to this extent, the words of ss 7 and 24, because of changed historical circumstances including legislative history, have come to be a constitutional protection of the right to vote. That, however, leaves open for debate the nature and extent of the exceptions. The Constitution leaves it to Parliament to define those exceptions, but its power to do so is not unconstrained. Because the franchise is critical to representative government, and lies at the centre of our concept of participation in the life of the community, and of citizenship, disenfranchisement of any group of adult citizens on a basis that does not constitute a substantial reason for exclusion from such participation would not be consistent with choice by the people[10]. To say that, of course, raises questions as to what constitutes a substantial reason, and what, if any, limits there are to Parliament's capacity to decide that matter.
It is difficult to accept that Parliament could now disenfranchise people on the ground of adherence to a particular religion. It could not, as it were, reverse Catholic emancipation. Ordinarily there would be no rational connection between religious faith and exclusion from that aspect of community membership involved in participation, by voting, in the electoral process. It is easy to multiply examples of possible forms of disenfranchisement that would be identified readily as inconsistent with choice by the people, but other possible examples might be more doubtful. An arbitrary exception would be inconsistent with choice by the people. There would need to be some rationale for the exception; the definition of the excluded class or group would need to have a rational connection with the identification of community membership or with the capacity to exercise free choice. Citizenship, itself, could be a basis for discriminating between those who will and those who will not be permitted to vote[11]. Citizens, being people who have been recognised as formal members of the community, would, if deprived temporarily of the right to vote, be excluded from the right to participate in the political life of the community in a most basic way. The rational connection between such exclusion and the identification of community membership for the purpose of the franchise might be found in conduct which manifests such a rejection of civic responsibility as to warrant temporary withdrawal of a civic right.
This brings me to the issue in the present case. The facts, the legislation, and the historical background appear from the reasons of Gummow, Kirby and Crennan JJ ("the joint reasons"). Since 1902, when the Commonwealth Parliament first legislated with respect to the franchise, the legislation always provided that, along with persons of unsound mind and persons attainted of treason, prisoners of certain kinds were not entitled to vote. The rationale for excluding persons of unsound mind is obvious, although the application of the criterion of exclusion may be imprecise, and could be contentious in some cases. The rationale is related to the capacity to exercise choice. People who engage in acts of treason may be regarded as having no just claim to participate in the community's self-governance. It will be necessary to return to the rationale for excluding prisoners. First, however, the changes in the exclusion over the years should be noted. Not all people in prison are serving sentences of imprisonment. Some are awaiting trial. They are not covered by any of the exclusions. There was some discussion in argument concerning fine defaulters. It was suggested that, perhaps depending on the precise terms of the orders made against them, they also would not be excluded. It is unnecessary to pursue that question. From 1902 until 1983, the exclusion was of convicted persons under sentence or subject to be sentenced for an offence punishable by imprisonment for one year or longer. From 1983 until 1995, the reference to one year was replaced by five years. From 1995 to 2004, the reference to imprisonment for an offence punishable by imprisonment for five years or longer was altered to serving a sentence of five years or longer. From 2004 to 2006, the period of five years was altered to three years. In 2006, Parliament enacted s 93(8AA) of the Commonwealth Electoral Act 1918 (Cth) which provides that a person who is serving a sentence of imprisonment for an offence against the law of the Commonwealth or of a State or Territory is not entitled to vote at any Senate election or House of Representatives election. The plaintiff's challenge to the validity of s 93(8AA) gives rise to the primary issue in the present case. If it succeeds, there is a question whether the previous (three-year) regime still validly applies.
What is the rationale for the exclusion of prisoners? Two possibilities may be dismissed. First, the mere fact of imprisonment is not of itself the basis of exclusion. According to the Australian Bureau of Statistics, at 30 June 2006 there were 25,790 prisoners (sentenced and unsentenced) in Australian prisons. Unsentenced prisoners (typically persons on remand awaiting trial) comprised 22 per cent (5,581) of the total prisoner population[12]. They have the right to vote. We were informed that they do so either by postal voting or by the visit to prisons of mobile voting booths. Accordingly, there is nothing inherently inconsistent between being in custody and voting; even under the current exclusion, more than one-fifth of prisoners vote. Secondly, exclusion by federal law from voting cannot be justified as an additional punishment. The great majority of prisoners in Australia are people who have been sentenced by State courts for offences against State law. The States bear the principal responsibility for the administration of criminal justice. There would be serious constitutional difficulties involved in seeking to justify a federal law such as s 93(8AA) as an additional punishment upon State offenders; especially upon State offenders who had previously been convicted and sentenced under State law. I do not intend to suggest that there would be no difficulties about treating it as additional punishment for offences against federal or territorial law, but the position of State offenders is sufficient to demonstrate the problem with treating it as punishment at all.
The rationale for the exclusion from the franchise of some prisoners, that is, those who have been convicted and are serving sentences, either of a certain duration or of no particular minimum duration, must lie in the significance of the combined facts of offending and imprisonment, as related to the right to participate in political membership of the community. The combination is important. Just as not all prisoners are excluded, even under s 93(8AA), from voting, not all persons convicted of criminal offences are excluded. Non-custodial sentences do not attract the exclusion. A pecuniary penalty, no matter how heavy, does not lead to loss of the vote. Since it is only offences that attract a custodial sentence that are involved, this must be because of a view that the seriousness of an offence is relevant, and a custodial sentence is at least a method, albeit imperfect, of discriminating between offences for the purpose of marking off those whose offending is so serious as to warrant this form of exclusion from the political rights of citizenship.
Since what is involved is not an additional form of punishment, and since deprivation of the franchise takes away a right associated with citizenship, that is, with full membership of the community, the rationale for the exclusion must be that serious offending represents such a form of civic irresponsibility that it is appropriate for Parliament to mark such behaviour as anti-social and to direct that physical separation from the community will be accompanied by symbolic separation in the form of loss of a fundamental political right. The concept of citizenship has itself evolved in Australian law[13]. The preamble to the Australian Citizenship Act 2007 (Cth) declares that Parliament recognises that Australian citizenship represents full and formal membership of the community of the Commonwealth of Australia, and Australian citizenship is a common bond, involving reciprocal rights and obligations. The reference to the reciprocity of rights and obligations is important in the context of membership of the community. Serious offending may warrant temporary suspension of one of the rights of membership, that is, the right to vote. Emphasis upon civic responsibilities as the corollary of political rights and freedoms, and upon society's legitimate interest in promoting recognition of responsibilities as well as acknowledgment of rights, has been influential in contemporary legal explanation of exclusions from the franchise as consistent with the idea of universal adult suffrage.
In Sauvé v Canada (Chief Electoral Officer)[14], Gonthier J cited a passage in a work of the American constitutional law scholar, Professor Tribe[15], who wrote:

Gonthier J made the point[16] that it is legitimate for society to curtail the vote temporarily of people who have demonstrated a great disrespect for the community by committing serious crimes, on the basis that civic responsibility and respect for the rule of law are prerequisites to democratic participation. This, he said, reinforces the significance of the relationship between individuals and their community when it comes to voting.
The litigation in Sauvé concerned an issue similar to the present, but the issue arose under a different legal regime. The Canadian Charter of Rights and Freedoms, in s 3, guarantees every citizen the right to vote. Section 1, however, permits "such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society." This qualification requires both a rational connection between a constitutionally valid objective and the limitation in question, and also minimum impairment to the guaranteed right[17]. It is this minimum impairment aspect of proportionality that necessitates close attention to the constitutional context in which that term is used. No doubt it is for that reason that the parties in the present case accepted that Sauvé (like the case of Hirst discussed below) turned upon the application of a legal standard that was different from the standard relevant to Australia. The Supreme Court of Canada had previously held that a blanket ban on voting by prisoners, regardless of the length of their sentences, violated the Charter[18]. The legislature changed the law to deny the right to vote to all inmates serving sentences of two years or more. Dividing five-four, the Supreme Court of Canada again held that the legislation violated the Charter. The central issue was whether the s 1 justification (involving the minimum impairment standard) had been made out.
The United Kingdom has for many years had legislation which disenfranchises all convicted prisoners. The European Court of Human Rights, in Hirst v United Kingdom (No 2)[19], by majority, held that the automatic blanket ban imposed on all convicted prisoners violated Art 3 of Protocol 1 to the Convention for the Protection of Human Rights and Fundamental Freedoms. The majority accepted that the United Kingdom law pursued the legitimate aim of enhancing civic responsibility and respect for the rule of law by depriving those who had breached the basic rules of society of the right to have a say in the way such rules were made for the duration of their sentence. However, they concluded that the measure was arbitrary in applying to all prisoners, and lacked proportionality (which in this context also required not only a rational connection between means and ends but also the use of means that were no more than necessary to accomplish the objective), even allowing for the margin of appreciation to be extended to the legislature[20]. We were informed by counsel that the United Kingdom's response to the decision has not yet been decided.
There is a danger that uncritical translation of the concept of proportionality from the legal context of cases such as Sauvé or Hirst to the Australian context could lead to the application in this country of a constitutionally inappropriate standard of judicial review of legislative action. Human rights instruments which declare in general terms a right, such as a right to vote, and then permit legislation in derogation of that right, but only in the case of a legitimate objective pursued by means that are no more than necessary to accomplish that objective, and give a court the power to decide whether a certain derogation is permissible, confer a wider power of judicial review than that ordinarily applied under our Constitution. They create a relationship between legislative and judicial power significantly different from that reflected in the Australian Constitution, and explained at the commencement of these reasons. The difference between the majority and minority opinions in both Sauvé and Hirst turned largely upon the margin of appreciation which the courts thought proper to allow the legislature in deciding the question of proportionality. Neither side in the present litigation suggested that this jurisprudence could be applied directly to the Australian Constitution. Even so, aspects of the reasoning are instructive.
To return to Sauvé, Gonthier J, with whom L'Heureux-Dubé, Major and Bastarache JJ agreed, and who favoured upholding the legislation disenfranchising prisoners serving sentences of two years or more, related the disqualification to the idea of citizenship. He said[21]:

Those observations apply also to Australia. It is consistent with our constitutional concept of choice by the people for Parliament to treat those who have been imprisoned for serious criminal offences as having suffered a temporary suspension of their connection with the community, reflected at the physical level in incarceration, and reflected also in temporary deprivation of the right to participate by voting in the political life of the community. It is also for Parliament, consistently with the rationale for exclusion, to decide the basis upon which to identify incarcerated offenders whose serious criminal wrongdoing warrants temporary suspension of a right of citizenship. I have no doubt that the disenfranchisement of prisoners serving three-year sentences was valid, and I do not suggest that disenfranchisement of prisoners serving sentences of some specified lesser term would necessarily be invalid. The specification of a term reflects a judgment by Parliament which marks off serious criminal offending, and reflects the melancholy fact that not all sentences of imprisonment necessarily result from conduct that falls into that category.
That fact is also reflected in one provision of the Constitution itself. Section 44 deals with the disqualification of senators and members of the House of Representatives. The section disqualifies a person who "has been convicted and is under sentence, or subject to be sentenced, for any offence punishable under the law of the Commonwealth or of a State by imprisonment for one year or longer". I do not suggest that, by implication, this imposes a lower limit on Parliament's capacity to disqualify voters. There is, of course, an incongruity in the fact that the current legislation, in the relevant respect, imposes stricter standards upon eligibility to be a voter than the Constitution imposes upon eligibility to be a senator or a member of the House of Representatives. The point, however, is that s 44 recognises that the mere fact of imprisonment, regardless of the nature of the offence or the length of the term, does not necessarily indicate serious criminal conduct. That was so in 1901, and it remains so today.
One of the major problems currently affecting the administration of criminal justice, in Australia and elsewhere, is that of the short-term prison sentence, an expression which is normally used to refer to sentences of six months or less. In a 2001 report, the New South Wales Legislative Council's Select Committee on the Increase in Prisoner Population recommended that the government consider and initiate public consultation in relation to the abolition of sentences of six months or less[22]. The Bureau of Crime Statistics and Research was asked to estimate the impact on the prison system of such abolition. In 2000-2001, offenders sentenced to less than six months accounted for 65 per cent of all persons sentenced to prison by New South Wales adult criminal courts for that year. They are a much lower percentage of the total prison population but, for obvious reasons, the turnover is greater. According to the Bureau, it was estimated that, if all those who currently received sentences of six months or less were instead given non-custodial penalties, the number of new prisoners received in New South Wales prisons would drop from about 150 per week to about 90 per week[23]. In 2004, the New South Wales Sentencing Council reported on the same topic[24]. Short-term sentences were not abolished. In 2007, the Judicial Commission of New South Wales recorded that "sentences of six months or less, usually imposed by lower courts, have a significant impact on the prison population."[25] Section 5(2) of the Crimes (Sentencing Procedure) Act 1999 (NSW) reflects a legislative concern to attempt to limit the number of short sentences. Western Australian legislation has gone further[26]. In England, short-term sentences were significantly affected by ss 181-195 of the Criminal Justice Act 2003 (UK)[27].
As a matter of sentencing practicality, in the case of short-term sentences the availability of realistic alternatives to custody is of particular importance. If an offence is serious enough to warrant a sentence of imprisonment for a year or more, the likelihood is that the sentencing judicial officer will have formed the view that there was no serious alternative to a custodial sentence. In most Australian jurisdictions, there is a legislative requirement to treat imprisonment as a last resort when imposing a penalty[28]. More than 95 per cent of short-term sentences are imposed by magistrates[29]. The availability, in all the circumstances of a particular case, of other sentencing options such as fines, community service, home detention, or periodic detention may be critical. Relevant circumstances may include the personal situation of the offender, or the locality. In the case of offenders who are indigent, or homeless, or mentally unstable, the range of practical options may be limited. In rural and regional areas, the facilities and resources available to support other options also may be limited. In its June 2004 Report, made pursuant to ss 100J(1)(a) and (d) of the Crimes (Sentencing Procedure) Act 1999 (NSW), the New South Wales Sentencing Council recorded that the Chief Magistrate "acknowledged the unavailability of uniform sentencing options throughout NSW" and "clearly demonstrated that alternatives to sentences of full-time imprisonment are not equally distributed across the State."[30] Practical considerations of this kind give particular meaning to "disadvantaged"[31]. I do not suggest these problems are peculiar to New South Wales. I refer to it because it is the largest jurisdiction. A study published in 2002 examined the types of offence for which people were serving short terms of imprisonment in New South Wales[32]. Theft (excluding robbery) was the most common offence. Then followed breaches of court orders, assault, and driving or traffic offences.
The adoption of the criterion of serving a sentence of imprisonment as the method of identifying serious criminal conduct for the purpose of satisfying the rationale for treating serious offenders as having severed their link with the community, a severance reflected in temporary disenfranchisement, breaks down at the level of short-term prisoners. They include a not insubstantial number of people who, by reason of their personal characteristics (such as poverty, homelessness, or mental problems), or geographical circumstances, do not qualify for, or, do not qualify for a full range of, non-custodial sentencing options. At this level, the method of discriminating between offences, for the purpose of deciding which are so serious as to warrant disenfranchisement and which are not, becomes arbitrary.
The step that was taken by Parliament in 2006 of abandoning any attempt to identify prisoners who have committed serious crimes by reference to either the term of imprisonment imposed or the maximum penalty for the offence broke the rational connection necessary to reconcile the disenfranchisement with the constitutional imperative of choice by the people.
I would uphold the challenge to the validity of s 93(8AA). I have already indicated that in my view the previous legislation was valid. For the reasons given in the joint reasons it continues to apply.
For these reasons, I joined in the order made on 30 August 2007.
GUMMOW, KIRBY AND CRENNAN JJ. Section 28 of the Constitution stipulates that unless sooner dissolved by the Governor-General every House of Representatives shall continue for three years from the first meeting of the House. Part III (with respect to the House) and Part II (with respect to the Senate) make further provision with respect to elections and s 57 deals with double dissolutions.
Part VI (ss 81-92) of the Commonwealth Electoral Act 1918 (Cth) ("the Electoral Act") provides for the establishment and maintenance of a roll of electors for each State and Territory and for Division and Subdivision rolls.
Part VII (ss 93-97) deals with qualifications and disqualifications for enrolment and for voting. In particular, s 93 specifies those entitled to enrolment (persons who have attained 18 years and are citizens or a member of a closed class of British subjects) and, with certain exceptions, provides that an elector whose name is so enrolled is entitled to vote at Senate and House of Representatives elections. The provisions with respect to entitlement represent the culmination of the movement for universal suffrage. Over time the cry "one man one vote" came to include women, Australians of indigenous descent, and those aged at least 18 years. The provision in s 93 for exceptions reflects the notion of disqualification, to protect the integrity of the electoral result from the exercise of the franchise by groups of voters sharing some characteristic considered to affect capacity to vote responsibly and independently.
Two of these groups singled out for exclusion in this way by s 93 are those incapable of understanding the nature and significance of enrolment and voting, by reason of unsoundness of mind, and those convicted of treason and treachery and not pardoned. This litigation concerns a third category, those convicted and serving their sentence, a class which includes the plaintiff.
The issues which arise on the Amended Special Case involve constraints which are said by the plaintiff to be derived from the text and structure of the Constitution and to render invalid certain of the amendments to the Electoral Act made by the Electoral and Referendum Amendment (Electoral Integrity and Other Measures) Act 2006 (Cth) ("the 2006 Act"). The relevant provisions of the 2006 Act commenced on 22 June 2006. If the plaintiff makes good her principal submission respecting the 2006 Act, consequential issues will arise as to the identification and effect of surviving provisions of the Electoral Act in their unamended form.
What follows are our reasons for supporting the order with respect to the Amended Special Case which was made on 30 August 2007.

The plaintiff challenges the validity of those provisions of the 2006 Act which made the changes to the Electoral Act described above, in particular the inclusion of s 93(8AA). The grounds upon which she asserts invalidity involve the following four alternative propositions: first, whilst ss 8 and 30 of the Constitution speak of the "qualification" of electors they do not speak of provisions for "disqualification" and the consequence of this omission is said to be that any legislation for disqualification must "satisfy the representative government criteria"; secondly, s 93(8AA) punishes persons such as the plaintiff who have been convicted under State laws and the Parliament has no power to legislate in that way; thirdly, it follows from the reasoning in Lange v Australian Broadcasting Corporation[36] that there is an implied freedom of political communication (or of political participation) which protects voting in federal elections, and that this is impermissibly burdened by the 2006 Act; finally, the 2006 Act impermissibly limits the operation of the system of representative (and responsible) government which is mandated by the Constitution.
The first three of these submissions may be considered immediately. As to the first, the distinction between qualification and disqualification, the following is to be said. Section 93 of the Electoral Act deals sequentially with those entitled to enrolment and those entitled to vote, and renders that entitlement to vote subject, among other provisions, to s 93(8AA); the phrase "qualification" when used in ss 8 and 30 of the Constitution is sufficiently broad to allow for reservations or exceptions to a qualification which otherwise is conferred by the law in question.
As to the second submission, respecting federal punishment for State offences, two points are to be made. First, the circumstance that the plaintiff is serving a sentence of imprisonment for offences against the law of Victoria supplies the factum upon which the federal law operates. Secondly, if the federal law otherwise be within power, as a law with respect to the qualification of electors, the nature of that factum does not deny to the law that character.
As to the third, for the reasons to be developed below, 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. It is this consideration which marks out as the appropriate ground for the decision in this case the plaintiff's fourth submission. To consideration of that submission we now turn, beginning with the relevant provisions of the Constitution.

Further, it appeared to be common ground (and correctly so) that these provisions were to be read not in isolation but with an appreciation both of the structure and the text of the Constitution. Reference may first be made to s 128. This requires submission of proposed laws for the alteration of the Constitution to be submitted to the electors qualified to vote for the election of members of the House of Representatives. Section 7 requires the Senate to be composed of Senators "directly chosen by the people of the State" and is to be read with the territories power in s 122[41]. Section 24, which also is to be read with s 122, requires that members of the House of Representatives "be composed of members directly chosen by the people of the Commonwealth".
The 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. However, the Commonwealth emphasised that whether the voting system has been so distorted as not to meet that requirement is a matter of permissible degree. The Commonwealth submitted that that degree was not exceeded by the 2006 Act, but it did not offer any particular criterion for the determination of such questions. However, 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; so also, despite prevalent attitudes in 1900, would be a law which now purported to exclude from the franchise persons of indigenous descent or bankrupts.
For her part, the plaintiff emphasised that a law which stipulates a criterion for disenfranchisement fixing upon service at the election date of any sentence of imprisonment operates in an arbitrary or capricious fashion, with no rational ground for the automatic exclusion from exercise of the popular franchise mandated by the Constitution, and would be invalid. She submitted that the 2006 Act was such a law.
Reference also should be made to s 44 of the Constitution. Among those incapable of being chosen or sitting as a senator or member is, as specified in s 44(ii), any person who:

The Evidence Act 1843 (Imp)[55] changed the common law and stated (s 1) that "no person offered as a witness shall hereafter be excluded by reason of incapacity from crime ...". However, as a ground of disqualification of electors, candidates and sitting members, the notion of "infamous crime" was included in the constitutional provisions made by or under Imperial legislation for the establishment in the 1850s of representative government in the Australasian colonies.
Upon the framing of the legislation respecting New Zealand[56], Tasmania[57], New South Wales[58], Victoria[59] and South Australia[60] two points may be made. The legislation linked qualification of electors with membership of the lower house of the legislature. Secondly, it expressed as a proviso to that qualification the exclusion of those attainted or convicted of treason, felony or other infamous offence or crime in any part of the Queen's dominions, with a saving for those who had received a free pardon or undergone the sentence passed upon them for the offence. Disqualification of sitting members was triggered by attaint of treason, and by conviction of felony or any infamous offence or crime.
Several observations may now be made upon the development in the Australian colonies of the principles respecting disqualification of electors, candidates and legislators which accompanied the growth of representative government. First, those casting the ballot in elections for the Legislative Assemblies and Houses of Assembly (restrictive franchises and nominated systems were continued over some time for Legislative Councils) and those whom they elected to the lower houses now were to be drawn from far broader elements of colonial society than was then the case in the United Kingdom. Secondly, the same notions of attaint for treason and conviction for felony or other infamous crime founded grounds for disqualification of electors, candidates and legislators. Thirdly, these grounds for disqualification manifested an understanding of what was required for participation in the public affairs of the body politic, particularly in polities such as the Australian colonies where the immigrant societies were not underpinned by a class system. Fourthly, that understanding fixed upon considerations of fitness and probity of character which were seen to be lacking in those convicted of crimes which answered the common law description of being "infamous".

The words emphasised differed from the confinement of s 44(ii) of the Constitution to federal and State offences, but were consistent with colonial precedents to which reference is made elsewhere in these reasons. It may be added that a proposal that disqualification extend to those in receipt of charitable relief as an inmate of a public charitable institution was withdrawn. Why, Senator Stewart asked, although in some eyes "to be poor is the greatest crime it is possible for a man to commit", should not an inmate of a charitable institution "be allowed to take an interest in the affairs of his country?"[73].
The 1902 Act was repealed in 1918 by s 3 of the Electoral Act, and provision both for entitlement to vote and disqualification has been made by the latter statute as amended from time to time.
A provision to the effect of the second paragraph of s 4 of the 1902 Act was included in s 39 of the Electoral Act and remained there until wholly removed in 1962[74]. With respect to the first paragraph of that section, in 1983 the period of "imprisonment for one year or longer" then appearing in s 93 was replaced by "5 years or longer"[75]. This was recast by the Electoral and Referendum Amendment Act 1995 (Cth) ("the 1995 Act") so as to disqualify persons "serving a sentence of 5 years or longer for an offence against the law of the Commonwealth or of a State or Territory"[76]. The period of three years was substituted in 2004 by the Electoral and Referendum Amendment (Prisoner Voting and Other Measures) Act 2004 (Cth) ("the 2004 Act")[77]. Section 93(8)(b) then took the form set out earlier in these reasons until the commencement of the 2006 Act. The relevant provisions came into force on 10 August 2004 immediately after the commencement on the same day of provisions of the Electoral and Referendum Amendment (Enrolment Integrity and Other Measures) Act 2004 (Cth) ("the First 2004 Act") which the 2004 Act amended or repealed to produce the state of the statute law respecting the three year regime, as just mentioned.
It may be added that s 80(1)(d) of the Electoral Act 1993 (NZ) disqualifies from registration and thus from voting those detained in prisons under a sentence of preventative detention, of imprisonment for life or for a term of three years or more.
We return to the validity of the 2006 Act.

As the Commonwealth submissions respecting the impermissible exclusion of sections of society such as bankrupts and those of indigenous descent demonstrate, there are constitutional restraints necessarily implicit in the otherwise broad legislative mandate conferred by the words "until the Parliament otherwise provides". The difficulty, as Gaudron J observed in McGinty[79], lies in the process by which it may be determined that a law impermissibly limits the electoral process and system.
So in Mulholland itself, the Court held that provisions in the Electoral Act respecting the registration of political parties and the requirements of "the 500 rule" did not infringe the Constitutional imperatives respecting representative government. Earlier, in Langer v The Commonwealth[80] the Court upheld the prescription by the Electoral Act of a method of full preferential voting for elections for the House of Representatives. McGinty[81] affirmed that the Constitution contained no implication affecting disparities of voting power upon holders of the franchise for the election of members of a State legislature.
On the other side of the line lies the freedom of communication on matters of government and politics which was identified in Lange[82] as "an indispensable incident" of the system of representative government established and maintained by the Constitution. As remarked earlier in these reasons, disqualification from exercise of the franchise is, if anything, a subject even closer to the central conceptions of representative government. Given the particular Australian experience with the expansion of the franchise in the 19th century, well in advance of that in the United Kingdom, this hardly could be otherwise.
Voting in elections for the Parliament lies at the very heart of the system of government for which the Constitution provides. This central concept is reflected in the detailed provisions for the election of the Parliament of the Commonwealth in what is otherwise a comparatively brief constitutional text.
In McGinty[83] the Court held that what is involved here is a category of indeterminate reference, where the scope for judgment may include matters of legislative and political choice. But that 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.
In 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"[84]. This proposition reflects the understanding that representative government as that notion is understood in the Australian constitutional context comprehends not only the bringing of concerns and grievances to the attention of legislators but also the presence of a voice in the selection of those legislators[85]. Further, in the federal system established and maintained by the Constitution, the exercise of the franchise is the means by which those living under that system of government participate in the selection of both legislative chambers, as one of the people of the relevant State and as one of the people of the Commonwealth. In this way, the existence and exercise of the franchise reflects notions of citizenship and membership of the Australian federal body politic.
Such notions are not extinguished by the mere fact of imprisonment. Prisoners who are citizens and members of the Australian community remain so. Their interest in, and duty to, their society and its governance survives incarceration. Indeed, upon one view, the Constitution envisages their ongoing obligations to the body politic to which, in due course, the overwhelming majority of them will be returned following completion of their sentence.
The question with respect to legislative disqualification from what otherwise is adult suffrage (where 18 is now the age of legal majority throughout Australia) thus becomes a not unfamiliar one. Is the disqualification for a "substantial" reason? A reason will answer that description if it be reasonably appropriate and adapted to serve an end which is consistent or compatible with the maintenance of the constitutionally prescribed system of representative government. When used here the phrase "reasonably appropriate and adapted" does not mean "essential" or "unavoidable"[86]. Rather, as remarked in Lange[87], in this context there is little difference between what is conveyed by that phrase and the notion of "proportionality". What upon close scrutiny is disproportionate or arbitrary may not answer to the description reasonably appropriate and adapted for an end consistent or compatible with observance of the relevant constitutional restraint upon legislative power.
The affinity to what is called the second question in Lange[88] will be apparent. It has been said[89] that the ability to cast a fully informed vote in an election of members of the Parliament depends upon the ability to acquire relevant information and thus upon that freedom of communication seen in Lange as an indispensable incident of the representative government mandated by the Constitution. The present case concerns not the ability to cast a fully informed vote but upon denial of entitlement to cast any vote at all. This case concerns not the existence of an individual right, but rather the extent of the limitation upon legislative power derived from the text and structure of the Constitution and identified in Lange[90].
Some guidance for resolution of the present case is provided by Coleman v Power[91]. There Gummow and Hayne JJ and Kirby J were of the view that in the statutory provision under consideration[92] the proscription of "abusive" and "insulting" words was to be construed as applying to words which, in the circumstances where they are used, are so hurtful as either intended to or be reasonably likely to provoke unlawful physical retaliation[93]. Were that not so, and were a broader meaning given to the area of proscribed communication then the end served by the statute would necessarily be the maintenance of civility of discourse; given the established use of insult and invective in political discourse, that end could not satisfy the second question or test in Lange[94]. McHugh J construed the statute as imposing an unqualified prohibition upon the use of insulting words in a broad sense which thus went beyond what could be regarded as reasonably appropriate and adapted to maintaining the constitutionally prescribed system of representative government[95].
Paragraph (a) of s 93(8) of the Electoral Act disentitles those who are incapable of understanding the nature and significance of enrolment and voting because they are of unsound mind. That provision plainly is valid. It limits the exercise of the franchise, but does so for an end apt to protect the integrity of the electoral process. That end, plainly enough, is consistent and compatible with the maintenance of the system of representative government.
The end served by the denial in s 93(8AA) of the exercise of the franchise by electors then serving a sentence of imprisonment for an offence against federal State or Territory law is further to stigmatise this particular class of prisoner by denying them during the period of imprisonment the exercise of the civic right and responsibility entailed in the franchise. The measurement of that end against the maintenance of the system of representative government first requires a closer examination of the particular class of prisoner which has been singled out in this way.
Section 93(8AA) operates without regard to the nature of the offence committed, the length of the term of imprisonment imposed, or the personal circumstances of the offender. As indicated earlier in these reasons, there is long established law and custom, stemming from the terms of the institution in the Australasian colonies of representative government, whereby disqualification of electors (and candidates) was based upon a view that conviction for certain descriptions of offence evinced an incompatible culpability which rendered those electors unfit (at least until the sentence had been served or a pardon granted) to participate in the electoral process. That tradition is broken by a law in the terms of s 93(8AA) as such a law has no regard to culpability. Moreover, the disqualification imposed by that provision may operate more stringently than that imposed by s 44(ii) of the Constitution upon candidates and members of the Senate and the House, even though the latter seek, or are subject to, unique responsibilities as legislators which are different in kind to those of electors. The disharmony between s 93(8AA) of the Act and s 44(ii) of the Constitution is plain.
Contemporary penal policy sometimes asserts that the imposition of a custodial sentence is to be a last rather than first resort. Things may have stood differently at the time of federation. But with respect to the present state of affairs, several matters to which the Chief Justice refers in his reasons are of particular significance. First, a very substantial proportion of prisoners serve sentences of six months or less. Secondly, when decisions to impose short‑term custodial sentences are made, the range of practical sentencing options (including fines, home or periodic detention and community service orders) may be limited by the facilities and resources available to support them and by the personal situation of those offenders who are indigent, homeless or mentally unstable.
Moreover, s 93(8AA) is not yoked to sentencing laws or practices of any particular description. Rather it picks up the consequences of the administration of those laws as they apply from time to time across the range of Australian jurisdictions. Sentencing policy and, in particular, that regarding mandatory sentencing is notoriously a matter of continuing public debate and variable legislative responses in different Australian jurisdictions. In such matters, statutory provisions and administrative policies and emphases constantly change. However, the Constitution with its central notion of electoral representation and participation endures.
The 2006 Act treats indifferently imprisonment for a token period of days, mandatory sentences, and sentences for offences of strict liability. It does not reflect any assessment of any degree of culpability other than that which can be attributed to prisoners in general as a section of society. In that regard, the plaintiff referred, as examples, to current legislation in several States and Territories whereby, as a last resort, failure to pay fines may result in a term of imprisonment, and to legislation in Victoria[96] and Queensland[97] whereby begging is an offence punishable by a term of imprisonment. (The Commonwealth disputed whether all the current legislation with respect to fine defaulters would produce consequences which answered the definition of "sentence of imprisonment" in s 4(1A) of the Electoral Act but that cannot fully meet the point the plaintiff seeks to make.) Further, in 2006 of the prison population 6.3 per cent was serving a sentence for a public order offence or a road traffic or motor vehicle regulatory offence and 17.6 per cent was serving a sentence of less than one year.
The Solicitor-General of the Commonwealth accepted that, for example, manslaughter is a striking example of an offence which involves an extensive range of moral culpability down to little more than negligence; this may be reflected in the term of the sentence imposed. He responded that the 2006 Act operated with a valid degree of precision by limiting the period of disqualification to that for which the law provided incarceration. The difficulty with that proposition is the scope thereby provided for the particularly capricious denial of the exercise of the franchise.
The legislative pursuit of an end which stigmatises offenders by imposing a civil disability during any term of imprisonment takes s 93(8AA) beyond what is reasonably appropriate and adapted (or "proportionate") to the maintenance of representative government. The net of disqualification is cast too wide by s 93(8AA). The result is that ss 93(8AA) and 208(2)(c) are invalid and question (1) in the Amended Special Case should be answered accordingly.

The three year provisions (to put the subject matter in short form) of the 2004 Act differ in their nature from the 2006 Act. They operate to deny the exercise of the franchise during one normal electoral cycle but do not operate without regard to the seriousness of the offence committed as an indicium of culpability and temporary unfitness to participate in the electoral process. In that way the three year provisions are reflective of long established law and custom, preceding the adoption of the Constitution, whereby legislative disqualification of electors has been made on the basis of such culpability beyond the bare fact of imprisonment.
The plaintiff seemed to eschew her standing to challenge a disqualification system such as that of five years or longer established by the 1995 Act. But to succeed even with respect to the three year provisions the plaintiff has to make good her original submission. This was that disqualification of persons serving a term of imprisonment could only be a basis of exclusion "rationally connected with representative democracy" if the offence involved an attack on the existence of the federal polity or electoral fraud such as to undermine the integrity of the electoral system.
At a general level of debate there is support for and against reasoning of this kind in the majority and minority reasons given by the Supreme Court of Canada in Sauvé v Canada (Chief Electoral Officer)[102]. However, the Supreme Court there was considering (and held invalid) a two year or more sentence disqualification provision and did so by reference to an express conferral upon citizens by s 3 of the Canadian Charter of Rights and Freedoms of "the right to vote in an election of members of the House of Commons or of a legislative assembly and to be qualified for membership therein". The reasoning of the majority in Sauvé was that the legislation was an unreasonable infringement of the right to vote guaranteed to citizens by s 3 of the Charter. This reasoning was influential in the decision of the European Court of Human Rights in Case of Hirst v The United Kingdom (No 2)[103]. There the question was whether the exclusion imposed by the United Kingdom[104] upon convicted prisoners in detention was disproportionate according to the jurisprudence of that Court. The Grand Chamber by a decision of 12 of the Judges to five held against the United Kingdom. Article 3 of Protocol 1 to the European Convention on Human Rights guarantees "free elections ... under conditions which will ensure the free expression of the opinion of the people in the choice of the legislature" and this has been classified by the European Court as conferring individual rights[105].
The question respecting the three year provision that is presented by the constitutional jurisprudence of this Court differs from that which would arise at Ottawa or Strasbourg. It is whether the 2004 Act is appropriate and adapted to serve an end consistent or compatible with the maintenance of the prescribed system of representative government. The end is the placing of a civil disability upon those serving a sentence of three years or longer for an offence, the disability to continue whilst that sentence is being served.
Given the 19th century colonial history, the development in the 1890s of the drafts of the Constitution, the common assumptions at that time, and the use of the length of sentence as a criterion of culpability founding disqualification, it cannot be said that at federation such a system was necessarily inconsistent, incompatible or disproportionate in the relevant sense. Further, in the light of the legislative development of representative government since federation such an inconsistency or incompatibility has not arisen by reason of subsequent events. Despite the arguments by the plaintiff respecting alleged imperfections of the three year voting disqualification criterion, such a criterion does distinguish between serious lawlessness and less serious but still reprehensible conduct. It reflects the primacy of the electoral cycle for which the Constitution itself provides in s 28. There is, as remarked earlier in these reasons, a permissible area in such matters for legislative choice between various criteria for disqualification. The 2004 Act fell within that area and the attack on its validity fails.

Section 7 of the Constitution provides (so far as now relevant) that "[t]he Senate shall be composed of senators for each State, directly chosen by the people of the State". Section 24 (again so far as now relevant) provides that "[t]he House of Representatives shall be composed of members directly chosen by the people of the Commonwealth".
The plaintiff alleged that the impugned provisions, in their application to her, are invalid because their application would deny the Constitution's requirement that each House of the Parliament is "directly chosen by the people".
The text of the relevant provisions shows that the power given to the Parliament by s 30 (to provide for the qualification of electors) is to be read as limited by the requirements of ss 7 and 24 that the two Houses are "directly chosen by the people". But what limitation on that power is conveyed by those words?
History provides the only certain guide. The drafting history of what became s 30 shows that the Parliament's power under that section was given so that the Parliament itself could determine the franchise upon which it was elected. That is, the purpose of the conferral of legislative power under s 30 was to provide the Parliament with the power to determine which groups should be given the franchise.
Once that is recognised, it follows that the words "directly chosen by the people" are to be understood as an expression of generality, not as an expression of universality. Because the power to delineate the franchise was given to the Parliament, the ambit of exceptions to or disqualifications from the franchise was a matter for the Parliament itself, so long always as the generality of "directly chosen by the people" was preserved.
The scope, or content, of that "generality" cannot be charted by precise metes and bounds. The nature of its content, however, is indicated by the range of provisions made by the several State laws that were "picked up", at federation, by s 30. All of those laws disqualified some prisoners from voting. Excepting prisoners from the franchise did not and does not deny the generality required by "directly chosen by the people".
Competing approaches to the question necessarily begin from a premise that assumes the answer. It will be necessary, later in these reasons, to identify the competing approaches proffered by the plaintiff and the premises from which those approaches were advanced.

The plaintiff is an Australian citizen of indigenous descent. She is aged over 18 years. She is entitled[108] and required[109] to be enrolled to vote and is enrolled to vote in the Division of Kooyong. The plaintiff is serving a sentence of imprisonment for offences against the laws of the State of Victoria. She is not eligible to be released from prison before the latest date by which the next federal election must be held. If the impugned provisions are valid, she will not be entitled to vote at that election.
The plaintiff has commenced proceedings in the original jurisdiction of the Court seeking, among other relief, declarations that the impugned provisions are invalid. The parties joined in stating[110] what were said to be the questions of law arising in the proceeding in the form of a special case for the opinion of the Full Court. Those questions, as ultimately amended, included questions asking whether ss 93(8AA), 208(2)(c), and 221(3) of the Act are invalid.
By amendments to the Special Case, made in the course of argument, the parties sought to raise some further questions predicated upon the Court finding that the impugned provisions of the Act, in its present form, are invalid. Those further questions addressed the validity of two earlier forms of provisions of the Act dealing with the eligibility of prisoners to vote in federal elections: the provisions as they stood before the Electoral and Referendum Amendment (Electoral Integrity and Other Measures) Act 2006 (Cth) ("the 2006 Act"), and the provisions as they stood before the Electoral and Referendum Amendment (Prisoner Voting and Other Measures) Act 2004 (Cth) ("the 2004 Act").
The Act, as it stood before the 2004 Act, disqualified prisoners serving a sentence of 5 years or longer for an offence against a law of the Commonwealth or of a State or Territory[111]. The 2004 Act provided for the disqualification of prisoners serving a sentence of 3 years or longer for an offence against a law of the Commonwealth or of a State or Territory.
The plaintiff alleges that if the Act validly provides that persons serving a sentence of 5 years or longer are disqualified from voting, she would not be subject to that disqualification. She was convicted on five counts for offences of burglary[112], theft[113], conduct endangering persons[114], and negligently causing serious injury[115]. She was not sentenced, in respect of any of those offences, to a term of imprisonment of 5 years or longer. The orders for cumulation and concurrency that were made resulted in a total effective sentence of six years and it was ordered that she was not to be eligible for parole before the expiration of four years. Whether the plaintiff's contention is correct was not explored in argument, and it is neither necessary nor desirable to consider the point. It is a point that does not arise under the questions that the parties, by their Amended Special Case, have joined in presenting for consideration by the Full Court.
No question is asked by the parties which directly invites attention to whether the provisions enacted in 1902 by the Commonwealth Franchise Act 1902 (Cth) ("the 1902 Act") concerning the disqualification of prisoners were valid. Those provisions remained in force until 1983[116]. They were evidently based upon the model provided by s 44(ii) of the Constitution and its prescription of which persons are incapable of being chosen or of sitting as a senator or as a member of the House of Representatives. The 1902 Act, like the provisions of s 44(ii), fastened upon those who were attainted of treason, or had been convicted and were under sentence, or subject to be sentenced, for an offence punishable by imprisonment for one year or longer. Whereas s 44(ii) of the Constitution identified the relevant offences as offences under the law of the Commonwealth or of a State, the 1902 Act cast its net wider by embracing offences under the law of any part of the King's dominions.

Unlike Sydney, the proposal put to the Convention in Adelaide, to allow the federal Parliament to enact a uniform franchise, attracted no serious debate[122] about whether the power over the franchise should rest with the federal Parliament rather than the several State parliaments. Instead, the debate centred upon women's suffrage, and whether the Constitution should conclude that issue by providing for adult suffrage. The outcome of the Adelaide Convention was to adopt a clause which, in relevant respects, was in the form submitted to the Convention. The issue of women's suffrage was left for the new federal Parliament to decide.
This assumption underpinned the whole of the Parliamentary debates about the first Parliamentary specification of the federal franchise in the 1902 Act. Like the debates at the 1897 Adelaide Convention, the debates[123] in the Parliament about what was to become the 1902 Act focused chiefly upon the controversy about whether women should have the vote. The issue was resolved by the 1902 Act. That is, as Mr Barton had foreshadowed, when speaking[124] in support of his (failed) proposal to the 1891 Sydney Convention, the resolution of what, at the time, was seen as a difficult political question was effected by the Parliament in exercise of the power given by s 30: the power of "fixing the franchise". As Mr Barton had said[125]: "we must give the house most directly responsible to the people the right of fixing the franchise. You must not only allow that house, for that is a mere form of words, but the people, to fix their franchise" (emphasis added).
That was what the Parliament did in 1902 by its provision[126] that "[s]ubject to the disqualifications hereafter set out, all persons not under twenty-one years of age whether male or female married or unmarried" who met criteria of residence, being a subject of the King, and being enrolled, were entitled to vote (emphasis added). But the franchise thus granted, although general, was not universal. Section 4 of the 1902 Act provided that:

These matters of history point unambiguously to the conclusions expressed at the outset of these reasons. That is, the words "directly chosen by the people" must be understood as words of generality, not as words of universality. The words were not intended to convey a requirement for universal adult suffrage.
There are some additional textual indications that point in the same direction. It is convenient to deal with those here.
It should go without saying that the provisions of ss 7, 8, 24 and 30 must all be read in the context provided by the whole of the Constitution. Particular attention must be paid to the context provided by Pt 2 of Ch I (ss 7-23, concerning The Senate), Pt 3 of the same chapter (ss 24-40, concerning The House of Representatives) and Pt 4 of that chapter (ss 41-50, concerning Both Houses of the Parliament). But it is also necessary to pay due regard to s 128 concerning Alteration of the Constitution.
What is to be observed from the other provisions of Ch I of the Constitution is the frequency of reference (both by the formula "[u]ntil the Parliament otherwise provides" and otherwise) to the powers of the Parliament to enact laws regulating both elections for and membership of both of the Houses of the Parliament. Section 7 (with its provisions about the division of Queensland into divisions, and its provision for the numbers of senators to be elected in each State), s 9 (concerning the method of election of senators), s 10 (applying certain State laws to the election of senators), s 14 (concerning further provision for the rotation of vacancies in the Senate), and s 22 (concerning the quorum at a meeting of the Senate) are examples of such provisions. The examples can readily be multiplied by reference to Pt 3[128] and Pt 4[129] of Ch I. By these provisions, the Constitution provides power for the Parliament to regulate a number of aspects of how it is to be constituted and how it is to be elected. The conferring of these powers is consistent with the franchise being a matter for the Parliament to determine, subject only to the requirement that each House be "directly chosen by the people".
Two different points emerge from consideration of s 128. First, there is the point that the Constitution provides that it is the "electors qualified to vote for the election of members of the House of Representatives" who are ultimately to decide upon constitutional alteration. This is an important element of the form of representative democracy for which the Constitution provides. The second and more directly relevant point comes from the fourth paragraph of s 128. That provides that:

Several States followed a legislative pattern that derived ultimately from the Australian Constitutions Act 1842 (Imp). That Act had provided[131] for a property qualification for electors but provided for the disqualification of those attainted or convicted of "any treason, felony, or infamous offence within any part of Her Majesty's dominions" unless the person had received a free pardon, or one conditional upon not leaving the colony, or had undergone the sentence or punishment. Some colonies that had followed[132] this pattern had, by the time of federation, altered or abandoned the specification of property qualifications for voting. In some colonies[133] there was adult suffrage; in other colonies there was adult male suffrage. But in the colonies other than New South Wales and Victoria, the specification of the disqualification remained substantially in the form enacted in the Australian Constitutions Act 1842. That is, it was a disqualification that hinged about the currency of a sentence for "Treason, Felony or infamous Offence".
The New South Wales and Victorian disqualification provisions were more extensive. They disqualified several different classes of persons from voting. In New South Wales, s 23(IV) of the Parliamentary Electorates and Elections Act 1893 (NSW) provided that all who were "in prison under any conviction" were disqualified from voting. As well, the section provided for the disqualification of a number of other classes of person: some on account of their being under sentence following conviction for some kinds of offence identified by the severity of the maximum sentence that could be imposed for the offence, others on account of their having been imprisoned for an aggregate period of at least three months within the recent past. Still others were disqualified on account of their recent conviction for certain public order offences: being an habitual drunkard, an incorrigible rogue, or a rogue and vagabond. And any man against whom there was an unsatisfied order for maintenance of wife or children or who had recently been convicted of an aggravated assault upon his wife was disqualified.
The relevant Victorian provision was not identical but it contained provisions that were generally similar to those applying in New South Wales. Unlike New South Wales, there was not the blanket disqualification of anyone "in prison under any conviction". When it is recalled, however, that voting was not compulsory, and could be effected only by the voter attending at a polling place, the absence of a blanket disqualification of those in prison is not surprising. Section 24 of the Purification of Rolls Act 1891 (Vic) did require the removal from the electoral roll of "every person ... who during the last three years has served any term or terms of imprisonment for any period or periods amounting in the aggregate to at least three months such term or terms of imprisonment having been imposed without the option of a fine". It required the removal of persons who during the preceding three years had been found guilty of any of a number of offences concerning the conduct of elections. It required the removal of those who in the previous year had been convicted of being an habitual drunkard, idle and disorderly person, incorrigible rogue, or rogue and vagabond, as well as those who had unsatisfied orders for maintenance of wife or children, or who, in the previous year, had been convicted of committing an aggravated assault on his wife.
Several observations may be made about these different laws, all of them "picked up" by s 30 of the Constitution. First there is the obvious point to be made about their variety. There was no single form of franchise that was seen as necessary to produce the result that the Houses of the federal Parliament would be "directly chosen by the people". The most obvious, and then most controversial, difference was between South Australia and Western Australia (each with adult suffrage) and other States which did not provide for women to vote. But there were marked differences between the ways in which the several States identified those who were to be disqualified from voting.
All States excluded some prisoners from voting. For present purposes, the critical observation is that New South Wales excluded[134] "every person who ... is in prison under any conviction".
This being the state of the law picked up by s 30, persons in prison under sentence were, and now can be, excluded from voting without denying the Houses that are thus elected the constitutional description of "directly chosen by the people".
Moreover, this being the state of the law picked up by s 30, no more refined or precise proposition, whether hinged about length of sentence, quality of offence or otherwise, can now be identified as controlling the content of "directly chosen by the people" in its application to the subject of prisoners voting. The diversity of the relevant State provisions denies that a proposition of that kind can be identified as informing the Constitutional adoption and application of those State laws. State laws operated in some cases by reference to the length of the sentence that was imposed, in some by reference to the length or kind of sentence that could be imposed, in others by reference to the quality of the offence (treason, felony or infamous offence). The differences between the provisions are not to be ignored in favour of now devising, a priori, a criterion drawn either by reference to a particular length of sentence (whether actually imposed or available) or by reference to some quality of the offence for which the person has been imprisoned.
Penological theories that seek to connect any particular form of deprivation of rights or freedoms with the attainment of desired goals of punishment or reformation, may be very important considerations for legislators or other policy-makers. They may affect the way in which a court approaches the fixing of sentence for crime. But they are not relevant to the issues that arise in the present matter. Notions of "infamous crime", like notions of "civil death"[135], find no textual footing in the Constitution. Neither of those notions, nor any other form of penological theory, underpins or informs the content of any of the relevant Constitutional provisions.
Moreover, 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. The form of representative democracy for which the Constitution provides was established with British and American models at the forefront of the framers' consideration. But neither of those models was adopted. The Constitution provided its own form of government: a form of government in which there are elements that evidently draw on the experience of others but which, taken as a whole, is unique. To 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.

The plaintiff submitted that "disqualification must be reasonably appropriate and adapted to achieve a legitimate end that is consistent with the constitutional system of representative and responsible government in order to be valid". She further submitted that "the validity of the impugned provisions falls to be determined by reference to the representative democracy criteria" and that, however those criteria are formulated, the impugned provisions do not meet them "because they operate in an arbitrary and discriminatory manner and are both over- and under‑inclusive".
The plaintiff identified four paths which she submitted lead to the ultimate propositions just identified. It will be convenient to deal with each separately, recognising that each was said to lead to the same end. But one point, which goes to the root of the plaintiff's submissions, must be made at once.
The plaintiff did not give content to the "representative government criteria" which underpinned all of her submissions. Rather, it was submitted that it mattered not how those criteria were formulated; it sufficed to describe the operation of the impugned law as "arbitrary and discriminatory" and as "over‑ and under‑inclusive". But if, as must be the case, the "representative government criteria" include a criterion about qualification of electors, the specification of that criterion concludes the issue that must now be decided. The plaintiff, at least implicitly, 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. This approach is flawed.
The first of the paths identified by the plaintiff began from the proposition that the Constitution provides no express legislative power to provide for the "disqualification" of electors as distinct from their "qualification". Power in relation to "disqualification" was said to lie only in an incidental power (either as an incident to the power to provide for qualification or under s 51(xxxix)). It was submitted that it follows (a) that the power to provide for disqualification "is purposive in nature: it can be exercised only for the purpose of effectuating the main power"; and (b) that the power to provide for disqualification is subject to ss 7 and 24 (with their references to "directly chosen by the people") and "the other sections of the Constitution providing for representative and responsible government". The result of this analysis was said to be that "any disqualification of persons from voting must satisfy the representative government criteria".
The premise for this aspect of the plaintiff's argument should not be accepted. Section 30 should not be read as drawing a distinction between "qualification" of electors and their "disqualification". When s 30 of the Constitution speaks, as it does, of "the qualification of electors of members of the House of Representatives" and "the qualification of electors of the more numerous House of Parliament of [a] State" it is not to be read as confined to the delineation of a class of persons by inclusion. Rather, in the context of s 30, "qualification" must be read as extending to delineation of the class of those who are "electors of members of the House of Representatives" by inclusion, exclusion, or both. And the reference in s 8 to "[t]he qualification of electors of senators" must be read in the same way.
If s 30 is not read in the way just described, the validity of a particular legislative prescription of who may be an elector of members of the House of Representatives would turn upon the form of the provision, not its substantive operation. The valid engagement of s 30 is not to be understood as turning upon the Parliament adopting a particular drafting technique. Further, the proposition that the legislative power with respect to "qualification" of electors extends only to the prescription of those who are included within the relevant class would require reading the latter part of s 30, picking up State laws, either as picking up only so much of those State laws as was not cast as a form of disqualification, or as using the word "qualification" in a sense different from its use in the first part of the section. Neither of those readings should be adopted.
It may be accepted that the text of the Constitution provides some footing for distinguishing between questions of "qualification of electors" and their "disqualification". The sidenotes to both ss 8 and 30 are "[q]ualification of electors". By contrast, the sidenote to s 25 is "[p]rovision as to races disqualified from voting". Section 25 provides that, for the purpose of the calculation to be made under s 24 of the number of members of the House of Representatives to be chosen in each of the several States, "if by the law of any State all persons of any race are disqualified from voting at elections for the more numerous House of the Parliament of the State, then, in reckoning the number of the people of the State or of the Commonwealth, persons of that race resident in that State shall not be counted" (emphasis added). In addition, it is to be observed that ss 16 and 34 speak of the "qualifications" of a senator and a member of the House of Representatives, respectively, whereas s 44, with its prescription of which persons are to be incapable of being chosen or of sitting as a senator or a member of the House of Representatives, is given the sidenote "[d]isqualification". And both ss 45 and 46 deal with consequences that follow from disqualification under s 44.
Moreover, it must also be accepted that the several State laws governing the franchise in elections for the more numerous House of the State Parliament that were picked up at federation by operation of s 30 were commonly drafted in a form that prescribed who was entitled to vote by first describing generally the class of persons who were to be entitled (those "qualified") and then providing a series of exceptions (by way of "disqualification") to the general reach of the qualification provisions. Even so, as the Australian Constitutions Act 1842 demonstrates, no clear line was drawn in such legislation between matters of qualification and matters of disqualification. Section 5 of that Act specified those who were qualified by reference to certain property criteria. Section 6, the disqualification provision, then dealt with some matters that might more easily be described as qualifications to vote by providing that:

These observations about the different uses of the words "qualification" and "disqualification" in the Constitution itself, in the laws to which s 30 required reference at federation, and in the law which was the ultimate pattern for some of that State legislation, do not require the conclusion that the references in s 30 to "the qualification of electors of members of the House of Representatives" and "the qualification of electors of the more numerous House of Parliament of [a] State" are to be read as confined to the delineation of a class of persons by inclusion. Rather, as stated earlier, "qualification" must be read, in ss 30 and 8, as permitting delineation of the class of those who are "electors" by inclusion, exclusion, or both.
The premise for the first of the four paths identified by the plaintiff as leading to the conclusions for which she contended should be rejected.
It may be that the first path of the plaintiff's argument is to be understood as making a different, and essentially individual and temporal, point. That is, the argument may be understood as contending that, because the relevant legislative power is expressed as a power with respect to the subject of "qualification of electors", there is no express legislative power to make a law that would "disqualify" a person from voting if that person has, at some earlier time, met the criteria of qualification. For the reasons already given, the argument fails. Moreover, its acceptance would lead to absurd results. The absurdity is illustrated by considering the case of a person, qualified and enrolled as an elector, later becoming of unsound mind. There can be no doubt that the legislative power permits the making of a law which would disqualify that person from voting so long as he or she was of unsound mind.
The second path described by the plaintiff fixed upon the limitation provided by the references in ss 7 and 24 of the Constitution to "directly chosen by the people". It was said that "[t]hese sections ... place a limit on the power to provide for the qualification of electors which precludes the Parliament from winding back the franchise and precludes the Parliament from disqualifying those who are otherwise qualified unless such disqualification is not inconsistent with ss 7 and 24". "Satisfaction of the representative government criteria" was said to be "necessary to ensure the requisite consistency".
As noted at the outset of these reasons, it is clear that the power given to the Parliament by s 30 to provide for the qualification of electors is to be read as limited by the requirements of ss 7 and 24 that the two Houses are "directly chosen by the people". The central question is what limitation on the power is conveyed by those words. Thus when the plaintiff submits that "[s]atisfaction of the representative government criteria is necessary to ensure the requisite consistency" the critical step is to identify what is meant by the "representative government criteria". This the plaintiff sought to do by reference first to statements made in decisions of this Court, and then by reference to some decisions of ultimate courts of other countries and some international materials.
Some particular emphasis was given, in argument, to what was said[136] by McTiernan and Jacobs JJ in Attorney-General (Cth); Ex rel McKinlay v The Commonwealth:

Is "directly chosen by the people" to be understood by reference to "the common understanding of the time"? That is, do what might be called "generally accepted Australian standards" provide a valid premise for consideration of the issues presented in this matter?
There are at least two reasons to reject reference to "common understanding" or "generally accepted Australian standards" as informing the content that is to be given to "directly chosen by the people". First, there is the obvious difficulty of determining what those standards are, and to what extent they are "generally accepted". Does it suffice that they are standards that are reflected in legislation which, by hypothesis, has been passed by a majority of popularly elected representatives in the two Houses of the federal Parliament? If that is sufficient, the limitation has no content; the Parliament may do as it chooses. If that is not sufficient, what is it that will demonstrate either the content of the asserted understanding or its common or general acceptance?
Secondly, and more fundamentally, it is not to be supposed that the ambit of the relevant constitutional power (as distinct from the political capacity to exercise the power) is constrained by what may, from time to time, be identified as politically accepted or acceptable limits to the qualifications that may be made to what now is an otherwise universal adult suffrage. Political acceptance and political acceptability find no footing in accepted doctrines of constitutional construction. The meaning of constitutional standards does not vary with the level of popular acceptance that particular applications of the power might enjoy.
The plaintiff's argument that the franchise cannot be "wound back" amounted to the contention that the Parliament has no legislative power to depart from what now is seen as a commonly understood minimum requirement for the franchise. To the extent to which the argument depends upon the invocation of "common understanding", it must be rejected for the reasons that have been given. To the extent to which it makes the temporal point noted in connection with the plaintiff's first path of argument, it must likewise be rejected.
Further, although it is not necessary to decide the point, it may greatly be doubted that the content of the expression "directly chosen by the people" changes over time. "[D]irectly chosen by the people" expresses a standard. It is not an expression that has a relevantly different application as facts change. The standard expressed is unvarying. It describes an important characteristic that each of the Houses of the Parliament must have. That the meaning of "directly chosen by the people" cannot be charted by metes and bounds does not entail that the meaning changes over time.
The expression "directly chosen by the people" may be seen as standing in sharp contrast with expressions like "foreign power"[137], or "postal, telegraphic, telephonic, and other like services"[138]. The latter expressions must be applied to various facts and circumstances that can and do change over time. In particular, the political or technical facts to which they are applied may require different applications of the relevant expression over time. The better view may well be that "directly chosen by the people" is not an expression of that kind. It is, however, not necessary to decide the point. It suffices to say that its content is not to be found by reference to what is "commonly understood", what is politically accepted, or what is politically acceptable.
The plaintiff sought to give content to the "representative government criteria" by reference to a deal of overseas material. Emphasis was placed, in argument, on the ways in which other nations, operating under different constitutional instruments and arrangements, have dealt with prisoners voting. Particular reference was made to several Canadian decisions[139] about the application of the Canadian Charter of Rights and Freedoms to federal laws disqualifying prisoners from voting, to the decision of the European Court of Human Rights in Hirst v United Kingdom (No 2)[140] concerning the compatibility of s 3 of the Representation of the People Act 1983 (UK)[141] with the First Protocol to the European Convention on Human Rights, and to a decision of the Constitutional Court of South Africa[142] concerning the validity of provisions depriving prisoners, serving a sentence of imprisonment without the option of paying a fine, of the right to participate in elections during the period of their imprisonment. All of these decisions held the legislation in question to be incompatible with an applicable statement of rights and freedoms, or to be constitutionally invalid. It was said that these decisions, or these decisions when read in conjunction with international instruments such as the International Covenant on Civil and Political Rights[143], revealed a generally accepted international standard that could, even should, find application either in the search for the "common understanding" of which McTiernan and Jacobs JJ spoke in McKinlay[144], or otherwise in the construction of "directly chosen by the people". American decisions[145] upholding the validity of statutes providing for the life‑long disenfranchisement of felons were said to be irrelevant on the ground that they depended upon the particular text and history of s 2 of the Fourteenth Amendment to the United States Constitution.
The argument from overseas material, in all of the several forms in which it was advanced by the plaintiff, should be rejected. The reasons given earlier in relation to "common understanding" or "generally accepted Australian standards" require that rejection. But there is a further and fundamental flaw in the plaintiff's argument.
Any appeal to the decisions of other Courts about the operation of other constitutional instruments or general statements of rights and freedoms is an appeal that calls for the closest consideration of whether there are any relevant similarities between the instruments that were examined and applied in those decisions and the particular provisions that this Court must consider. The plaintiff's argument that no useful guidance is to be had from United States' decisions acknowledges the force of this proposition.
There is no similarity between the provisions considered in the cases referred to and relied on by the plaintiff and the provisions of the Constitution that are in issue in the present matter. The only connection between the cases and other international materials upon which the plaintiff relied and the present issues is to be found in the statement of the problem as an issue about the validity of legislative provisions excluding prisoners from voting. That the problem may be stated in generally similar terms does not mean that differences between the governing instruments may be ignored. Yet in essence that is what the appeal made by the plaintiff to "generally accepted international standards" seeks to have the Court do.
The third of the paths identified by the plaintiff assumed (contrary to the submission made as the first path) that the impugned provisions are within the power conferred by s 30 and s 51(xxxvi). She submitted that that power is conferred "subject to this Constitution" and that the power is thus subject to an implied freedom of political communication, participation and association "which protects voting together with the communications required to render the vote an informed choice".
Reference to the implied freedom of political communication does not support the plaintiff's case unless it is first assumed that the freedom that is identified is one that either depends upon or implies a particular kind of franchise. But that is the very question for decision and, in the end, the appeal to the implied freedom is to be seen as no more than a restatement of the premise described as the "representative democracy criteria". For the reasons given earlier, that premise, to have the consequence for which the plaintiff contends, must assume the answer to the question for decision.
The fourth of the paths the plaintiff identified commenced with the proposition that the Parliament has no power to impose punishment for breach of a State law. It was said that because the effect of the impugned laws is punitive, it is to be assumed that their purpose was punitive. This being so, it was said that it was for the Commonwealth to "demonstrate some other, legitimate, purpose the law serves, which purpose has displaced the presumed punitive purpose". The plaintiff submitted that no other legitimate purpose had been or could be identified.
To say that the impugned laws are "punitive in their effect" seeks to characterise the way in which a person affected by the laws may describe the consequence of their application. That has been said[146] to be relevant to questions about the exercise of judicial power but it is neither necessary nor appropriate to consider here the utility of such a characterisation to questions arising under Ch III. The point which the plaintiff made was not a point about the exercise of judicial power, it was that the impugned provisions had not only a punitive effect but also a punitive purpose.
This branch of the plaintiff's submissions depended upon melding a number of disparate ideas into the single proposition that because the law "is punitive in nature" it is beyond the power of the Parliament. First, much of this aspect of the plaintiff's argument proceeded from the premise that the "representative government criteria" include a criterion about the franchise that supports her contentions. Thus it was said that

Secondly, the argument about effects and purposes did not distinguish between the political purposes or effects that may have moved a majority of the members of the two Houses to support a particular proposal and the questions of legal effect[147] that are to be considered when asking whether a law is a law with respect to a head of legislative power. Only the latter kinds of effect ("the rights, powers, liabilities, duties and privileges which [the impugned law] creates", and the "practical as well as the legal operation" of the law[148]) are relevant to the present issues. For the reasons given earlier, the impugned laws have the requisite character of a law made with respect to a matter of the kind described in s 51(xxxvi).
Finally, in support of this fourth path, and her arguments more generally, the plaintiff relied upon a deal of statistical and other material as demonstrating that the impugned laws have an application that is arbitrary or capricious. The statistical material upon which the plaintiff relied may yield a number of conclusions, not all as useful or reliable as others. It may be accepted, however, that not all crime is detected, not all criminals are prosecuted, and sentencing practices vary from jurisdiction to jurisdiction and to some extent may vary within a single jurisdiction. There are those who are in prison who have done far less than some who are at large in the community. There are jurisdictions which provide for mandatory sentences of imprisonment in cases where other jurisdictions do not. The indigenous population of this country is markedly over-represented in the prison population.
All of this may be accepted. But the root question remains: what is the limitation on legislative power that is prescribed by the requirement that the Houses of the Parliament are "directly chosen by the people"? The matters relied on by the plaintiff are relevant to the answer that is to be given if, and only if, some assumption is made about the nature of the representative democracy for which the Constitution provides. But that is the question for decision.

HEYDON J. The responses proposed by Hayne J to the questions asked are correct. His reasons for giving these responses are incontrovertible. Only the following additional points are made.
In the course of argument the Solicitor-General of the Commonwealth, no doubt understandably, made various concessions which were welcomed by the plaintiff. Some were express[149]. Some were implied[150]. Doubtless some are correct, and perhaps, for a variety of possible reasons, they are all correct, but, since they are concessions, they have not been the subject of contested argument, it is not necessary to decide whether they are correct, and anything said to flow from them is to that extent unsupported.
The plaintiff's submissions contained many assumptions as to whether it would be possible now to narrow the franchise on the basis of race, age, gender, religion, educational standards or political beliefs, questions which no Australian legislator has ever dreamed of or is likely to dream of. Resolution of the present case does not call for any of these assumptions to be either made or tested[151]; and certainly none of them were tested. It is enough to say that narrowing the franchise in any of these ways may be highly undesirable; it does not follow that it is unconstitutional.
The plaintiff's key assumption was that it is a necessary but not sufficient condition for the validity of electoral laws that they maintain or widen the franchise: "one cannot wind the clock back". Thus, it was assumed, if an electoral law contracts the franchise it is invalid. Many think that one of the advantages of having a liberal democratic legislature, particularly when the legislators belong to political parties having different opinions on some issues, is its capacity to experiment, to test what does or does not work, to make up for unsatisfactory "advances" by carrying out prudent "retreats". That capacity stands in contrast to the tendency of totalitarian regimes to become gerontocratic and ossified, faithful to only one technique of government. It 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.
The plaintiff relied on the terms of, and various decisions about and commentaries on, certain foreign and international instruments - the International Covenant on Civil and Political Rights, the First Protocol of the European Convention on Human Rights, the Canadian Charter of Rights and Freedoms and the Constitution of South Africa. The plaintiff's primary arguments were fixed, as they had to be, on ss 7, 8, 24, 30 and 51(xxxvi) of the Constitution, and on implications from these provisions. It is thus surprising that the plaintiff submitted that those arguments were "strongly supported" by decisions under the last three instruments "which found that prisoner disenfranchisement provisions were invalid". It is surprising because these instruments can have nothing whatever to do with the construction of the Australian Constitution. These instruments did not influence the framers of the Constitution, for they all postdate it by many years. It is highly improbable that it had any influence on them. The language they employ is radically different. One of the instruments is a treaty to which Australia is not and could not be a party. Another of the instruments relied on by the plaintiff is a treaty to which Australia is a party, but the plaintiff relied for its construction on comments by the United Nations Human Rights Committee. If Australian law permitted reference to materials of that kind as an aid to construing the Constitution, it might be thought that the process of assessing the significance of what the Committee did would be assisted by knowing which countries were on the Committee at the relevant times, what the names and standing of the representatives of these countries were, what influence (if any) Australia had on the Committee's deliberations, and indeed whether Australia was given any significant opportunity to be heard. The plaintiff's submissions did not deal with these points. But the fact is that our law does not permit recourse to these materials. The proposition that the legislative power of the Commonwealth is affected or limited by developments in international law since 1900 is denied by most[152], though not all[153], of the relevant authorities - that is, denied by 21 of the Justices of this Court who have considered the matter, and affirmed by only one.
An aspect of the plaintiff's argument about arbitrariness was that a large proportion of prisoners serve a sentence of two years or less, and whether these prisoners lose the vote depends on the length of time they spend in prison and where that period falls in "the three year federal electoral cycle". In practice the cycle is much less than three years. Many federal elections within living memory have been held less than three years after the previous one, and the plaintiff asserted, plausibly, that over the whole history of Federation they have been held on average about every two years and four months. It would be strange if the constitutional validity of a restriction on the franchise rose and fell with executive decisions about the duration of parliaments.
Finally, the plaintiff submitted that it was not necessary for her to argue that any of the legislation in force before 2004 was invalid. But, despite the plaintiff's refusal to admit this unconditionally[154], the following conclusions flow if her contention is sound. One is that if legislation in the form of the 1902 Act came up for consideration now, it would be declared void. Another is that if federal legislation was enacted in the form of that which existed in New South Wales and Victoria in 1900 and came up for consideration now, it would be declared void. On the assumption (which it is appreciated not everyone shares) that, leaving aside special circumstances capable of satisfactory explanation[155], legislation which would be declared void in 2007 would also have been declared void in 1902 or at any time between those two dates, it would follow that federal statutes in the two forms just described would also have been declared void in 1902, and in any year since that date in which they were challenged. That in turn would mean that every federal election in our history apart from the first one would have been held under invalid electoral laws. These conclusions are so highly improbable that the contentions of the plaintiff which lead to them must be incorrect.

Roach v Electoral Commissioner

At a glance

Court

Decision date

Before

Catchwords

Source

Catchwords

Judgment (252 paragraphs)

Parties

Legislation Cited (24)

Cases Cited (36)

Cited By (16)