• Specific Year
    Any

Hord, Brendan --- "Murphy v Electoral Commissioner: Between Severance and a Hard Place" [2017] SydLawRw 17; (2017) 39(3) Sydney Law Review 395


Case Note

Murphy v Electoral Commissioner: Between Severance and a Hard Place

Brendan Hord[∗]

Abstract

Murphy v Electoral Commissioner is the latest decision of the High Court of Australia addressing the constitutional validity of a federal electoral procedure. The Court upheld longstanding provisions of the Commonwealth Electoral Act 1918 (Cth) that temporarily prohibit persons from enrolling to vote or updating their enrolment before a Federal Election. This case note analyses three issues arising from the case. First, it examines the remedial difficulties posed by invalidating electoral legislation while the Parliament is dissolved. Second, it analyses the Court’s narrow application of the ‘directly chosen by the people’ requirement contained in ss 7 and 24 of the Australian Constitution. Finally, it evaluates the majority’s decision not to apply a structured proportionality test in light of recent developments in the context of the implied freedom of political communication.

I Introduction

Justice Bell delivered the reasons for Murphy v Electoral Commissioner[1] in Sydney’s Queen’s Square Courts on 31 August 2016. The courtroom was deserted, save for a lonely figure in the public gallery. Murphy has greater significance than this reception suggests.

The case concerned provisions of the Commonwealth Electoral Act 1918 (Cth) (‘Electoral Act’) that suspend additions or amendments to the electoral rolls during an election campaign. On the surface, Murphy was an opportunity for the High Court of Australia to address two constitutional issues: the nature of the requirement that the Houses of the Commonwealth Parliament are ‘directly chosen by the people’[2] and the status of ‘structured proportionality testing’ in Australian constitutional jurisprudence.

The importance of the case changed upon the unexpected dissolution of the Commonwealth Parliament in 2016. The Governor-General of Australia, acting on the advice of the Prime Minister, dissolved both Houses of Parliament two days before oral argument in Murphy. If the plaintiffs prevailed on the merits, the Court may have been required to invalidate the entire Electoral Act. Yet the dissolution of the Parliament would have prevented the enactment of a new electoral law. Invalidating the Act would have created an infinite regression: there would have been no Parliament to make electoral laws; nor the electoral laws to make a Parliament.

Rather than thrust the country into a constitutional quagmire, the Court unanimously upheld the validity of the impugned law. It applied a narrow reading of the direct choice requirement. Six Justices refused to apply a structured proportionality test, notwithstanding the controversial decision to extend it to a similar constitutional guarantee less than a year earlier. This case note will examine the treatment of these issues in Murphy.

II The Case

The Electoral Act establishes an electoral roll that is used to identify persons eligible to vote in a federal election. Eligible persons are subject to a legal obligation to enrol and maintain the accuracy of their enrolment.[3]

Historically, changes or additions to the electoral roll were suspended prior to polling day. Applications to amend the electoral roll during the suspension period were not processed until after the election, even if the roll was inaccurate or otherwise eligible persons were excluded from voting. Between 1902 and 1983, the electoral rolls were closed on the day that the Governor-General issued the writ for the election.[4]

In the 1930s, the Federal Executive started announcing elections before the issue of the writ to give electors a final opportunity to duly enrol.[5] The practice was enshrined in the Electoral Act in 1983 with the closure of the rolls fixed seven days after the Governor-General issued the writ for the election.[6] In 2006, the Howard Government reverted to closing the rolls on the same day that the writ was issued.[7] However, the High Court struck down the legislation in Rowe v Electoral Commissioner[8] and the seven-day grace period was restored.[9] The amnesty remained in place for the 2016 Federal Election and the rolls were closed 42 days before polling day.

The plaintiffs in Murphy argued that closing the electoral rolls seven days after the issue of the writ was inconsistent with the constitutional requirement that the Houses of Parliament are ‘directly chosen by the people’.[10] Closing the rolls prevented a substantial number of persons from voting. The exclusion was said to be unjustified given the Australian Electoral Commission possessed the technological and administrative capacity to process new enrolments closer to polling day.[11] The High Court unanimously disagreed.[12]

The franchise cases have enduring practical significance. The cases concerned the regulation of two of the closest elections in Australia’s political history. Rowe enabled 57 732 additional electors to vote in the 2010 Election.[13] The Election resulted in only the second ‘hung parliament’ in Australian history (and the first since 1940), with neither major political party forming a majority in the House of Representatives.[14] Similarly, Murphy would have permitted a substantial number of new voters to enrol for the 2016 Election. For example, during the 2013 Election, 52 694 applications for enrolment were received during the suspension period.[15] The 2016 Election resulted in a one seat majority for the incumbent Liberal and National Party Coalition, with the five most marginal seats being determined by a total of 4 456 votes.[16] It is clear that electoral laws, and their judicial interpretation, matter.

III The Remedy

Remedial difficulties cannot save an unconstitutional law from invalidity. However, the possibility of provoking a constitutional crisis magnified the burden borne by the plaintiffs in Murphy. The High Court had two remedial choices if it found for the plaintiffs. The Court could have severed the invalid provisions from the Electoral Act if severance was consistent with the intention of the Legislature. In this circumstance, the Act would continue to operate without the invalid provisions. If the provisions could not be severed, the entire Electoral Act would have been invalidated.

The first remedial difficulty in Murphy arose from the dissolution of the Commonwealth Parliament in 2016. The dissolution was unusual because it occurred before the ordinary expiration of the House of Representatives. Two days before oral argument in Murphy, and approximately six months before the term of the House of Representatives was due to elapse, the Governor-General acted on the advice of the Prime Minister and dissolved both Houses of Parliament under the seldom used ‘double dissolution’ procedure in s 57 of the Constitution.[17] Ordinarily, the Parliament could ameliorate the invalidation of the Electoral Act by re-enacting an amended version of the law. However, the timing of the dissolution meant that a Parliament did not exist when Murphy was heard. A new electoral law could not be enacted until after an election, but an election could not be held without an electoral law.

Murphy was particularly problematic since the High Court could not rely on the remedial strategy that it used in the earlier franchise cases. In Roach and Rowe, the impugned laws were amendments to a constitutionally valid provision of the Electoral Act. The unconstitutional amendment was invalidated and the principal Act was restored to its prior state.[18] It permitted the Court to find for the plaintiff and preserve the operation of the electoral system without the need for Parliamentary intervention. This option was not available in Murphy. An earlier provision of the Electoral Act could not be reinstated since the Act had never provided an opportunity for enrolment that satisfied the constitutional standard advanced by the plaintiffs.

The Parliament ‘lies at the very heart of the system of government for which the Constitution provides’.[19] It is inconceivable that the High Court would make a decision which indefinitely prevented the election of a Parliament. Accordingly, the question arises as to how the Court could have granted relief in Murphy if the plaintiffs prevailed on the merits. Part IIIA will examine the viability of severance and Part IIIB will assess possible responses to a finding that the entire Electoral Act was invalid.

A Invalidity with Severance

The Court may sever an invalid provision from an Act if severance is consistent with Parliament’s intention. Section 15A of the Acts Interpretation Act 1901 (Cth) creates a statutory presumption in favour of severability. However, severance will not be available if the Parliament evinces a ‘positive indication in the enactment that the legislature intended it to have either a full and complete operation or none at all’.[20] Neither s 15A nor the judicial function permits the Court ‘to redraft a [provision] so as to bring it within power and so preserve its validity’.[21]

The Electoral Act provides that a person who is entitled to enrol may apply to the Electoral Commissioner to have their name added to the electoral roll.[22] The person must notify the Electoral Commissioner within 21 days of becoming entitled to enrol or changing address.[23] Failure to do so is an offence.[24] If the application is in order, the Electoral Commissioner must enrol the person ‘without delay’.[25] The Electoral Commissioner is prohibited from considering claims for enrolment or transfer of enrolment which are received between the closure of the rolls and the election.[26] Section 155 fixes the closure of the rolls on the seventh day after the electoral writ has been issued.

The principal function of the electoral rolls is to determine eligibility to vote. However, the electoral rolls also serve a variety of other purposes. The rolls are used: to enable the accurate identification of electors on polling day;[27] to determine whether a candidate has sufficient support to nominate[28] or a party has sufficient support to be registered;[29] and to define the boundaries of an electoral division.[30] The rolls are available to the public for inspection during ordinary business hours.[31] Candidates for the House of Representatives have a statutory right to receive a copy of the roll for the division ‘as soon as practicable’ after the closure of the rolls.[32]

The Commonwealth advanced the extraordinary argument at the interlocutory stage that the impugned provisions were not severable.[33] This argument was removed in its amended submissions[34] and the Commonwealth joined the plaintiffs in advocating for severance in oral argument.[35] The Parties submitted that a successful challenge would invalidate the provision that closed the rolls — namely, s 155. However, the challenge would not affect the Electoral Commissioner’s obligation to process applications ‘without delay’. The Parties submitted that the Act would impose an obligation on the Commissioner to consider claims for enrolment or transfer of enrolment for as long as he or she deemed ‘practicable’.

A factor militating in favour of severance is the nature of the direct choice requirement. Unlike the other heads of power enumerated in s 51, Parliament is effectively subject to an obligation to design and maintain a system facilitating the popular election of the Parliament. The Parties could have argued that the Parliament should be presumed to comply with its constitutional obligation, and therefore, intended the Electoral Act to continue operating in the absence of the invalid provisions.

Yet the proposed severance radically alters the operation of the Act. The decision to close the rolls requires consideration of a variety of factors, including efficiency, accuracy, mitigation of fraud and cost. For example, the removal of the grace period in 2006 reflected the concern of the Joint Standing Committee on Electoral Matters that the volume of enrolments made during the seven-day period prevented the Australian Electoral Commission from thoroughly reviewing each application.[36] Weighing these considerations, and the outcome in Rowe, the Parliament made a judgment that a seven-day period was appropriate. It codified this judgment in s 155. The proposed severance substitutes the opinion of the Electoral Commissioner for that of the Parliament. The change is of sufficient magnitude to suggest the Court would be entering the impermissible realm of legislative drafting.

Although the Court did not address severance in Murphy, several Justices recognised that the requested relief breached the separation of powers. Justice Gageler characterised the case as an invitation to ‘engage in a process of electoral reform’ in line with the plaintiff’s understanding of best electoral practice.[37] Other Justices asserted that the imposition of an obligation on Parliament to expend ‘not insignificant’ funds to improve the electoral system was inappropriate.[38] These passages support the view that the proposed severance would have impermissibly exceeded the judicial function.

B Invalidity without Severance

The prospect of invalidity without severance is what the Solicitor-General of Australia termed a ‘very unattractive’ proposition.[39] A refusal to sever the impugned provisions would have invalidated the entire Electoral Act at a time when there was no Parliament to pass an alternative. In oral argument, the Commonwealth proposed four solutions to the impasse.[40]

First, the election could be conducted without an electoral law. This option has no utility. The Electoral Act regulates a wide range of issues which are necessary to conduct an accurate and fair election, including the method of election,[41] campaign financing,[42] electoral fraud[43] and procedures for returning electoral results.[44] Invalidating the Act would also abolish the Australian Electoral Commission[45] and withdraw the High Court’s jurisdiction to sit as the Court of Disputed Returns.[46]

Second, state laws could govern the election under s 10 of the Constitution. Section 10 provides that state law will determine the method of electing the Commonwealth Parliament ‘until the Parliament otherwise provides’. However, s 10 is a transitory provision that is exhausted when the Commonwealth enacts its own legislative scheme.[47] State laws ceased to apply in Commonwealth elections when the Commonwealth Electoral Act 1902 (Cth) and Commonwealth Franchise Act 1902 (Cth) commenced operation. In any event, the application of state laws under s 10 remains ‘subject to this constitution’. A state law that did not provide a sufficient opportunity for enrolment could not be used in a federal election for the same reason that the Commonwealth law would have been held invalid.

The final two proposals relied on executive power. The Commonwealth argued that the Governor-General of Australia could recall the Parliament to pass an electoral law. The power to determine the commencement and termination of the sessions of Parliament is part of the Crown’s prerogative powers.[48] Sections 5, 28 and 57 of the Constitution expressly vested this power in the Governor-General.[49] However, these provisions do not include the power to reverse a dissolution. Chief Justice Barwick observed that ‘[t]he dissolution itself is a fact which can neither be void nor be undone’, even if the dissolution failed to comply with constitutional procedures.[50] Irrevocability is the characteristic that distinguishes a dissolution from a prorogation.[51]

Alternatively, the Governor-General could instruct electoral officials to conduct the election under a modified version of the Electoral Act, even though the Act had been invalidated. The assent of the Government and the Opposition would be required since the caretaker period had commenced. Further, the Government and the Opposition would need to agree to retrospectively enact the measures upon the commencement of the new Parliament, irrespective of the outcome of the election. However, the proposal would usurp the separation of powers. The Constitution vested the authority to regulate elections in the legislative branch.[52] The proposal would require the Executive to exercise legislative power without a delegation of authority from Parliament.[53] Practical difficulties also arise since the Electoral Act regulates a range of electoral matters beyond polling day logistics.

Orthodox constitutional principles do not provide a satisfactory remedy in Murphy. The Governor-General might have been required to undertake one of the proposed executive actions under the doctrine of necessity. The doctrine is an executive power to restore ‘constitutional governance when for some reason government slips outside the Constitution and cannot otherwise be restored by orthodox legal routes’.[54] It would enable the Governor-General to temporarily act outside ordinary constitutional principles to reinstate the Parliament.[55] There is no express necessity power in the Constitution. Instead, it is an unwritten ‘reserve power’ supported by the Governor-General’s power to act ‘for the maintenance of this Constitution[56] and High Court obiter dicta acknowledging that ‘the Constitution cannot be so construed as to contemplate its own destruction’.[57]

There is limited historical practice to support using the doctrine of necessity to recall Parliament. Chitty identified two extraordinary instances from the 17th century where dissolved British Parliaments reassembled themselves under the doctrine.[58] He also argued that a dissolved Parliament could be reassembled if the Monarch died between a dissolution and the election of a new Parliament.[59] In each instance, ‘the necessity of the case render[s] it necessary for the Parliament to meet ... and necessity supersedes all law’.[60] Even though the Parliament was dissolved, ‘the members of the last Parliament shall assemble and be again a Parliament’.[61]

The proposed actions in Murphy are less intrusive than the instances identified by Chitty since the Crown, rather than the Parliament, would retain control over the duration of the Parliamentary term. Although overriding orthodox constitutional principles with executive power is a grave step,[62] restoring the democratic election of Parliament is one of the few circumstances that would warrant such drastic action.

C Future Implications

The issue may arise again. It is conceivable that a key piece of electoral machinery could be invalidated after the dissolution of Parliament. A late term amendment to an electoral law may limit the time that is available to hear a constitutional challenge. For instance, the electoral laws challenged in Day v Australian Electoral Officer (SA)[63] were enacted less than 50 days before Parliament was dissolved for the 2016 Election. Further complications may arise due to the requirement of standing. Section 76 of the Constitution requires the existence of a ‘matter’ before a plaintiff may invoke the jurisdiction of the High Court.[64] A person’s standing may not materialise until after the dissolution of the Parliament. For example, the plaintiffs in Rowe did not lose the opportunity to enrol until the electoral rolls were closed; an event that automatically occurred after the dissolution of Parliament.[65]

The possibility of constitutional deadlock could be averted by vesting the Governor-General with a narrow power to make electoral regulations. The Parliament’s ability to delegate its legislative power is almost unrestrained.[66] The Parliament could pass a law empowering the Governor-General to make a regulation replicating the Electoral Act if the Act was struck down after the dissolution of Parliament. It would need to permit the Governor-General to amend the provisions that gave rise to the invalidity. Further, the power would need to be passed as a separate enactment to survive the invalidation of the Electoral Act. A more limited power to make regulations about the mode of elections was used by the Governor of New South Wales to avert a constitutional crisis arising from the infamous ‘table-cloth ballot paper’ in the State’s 1999 Election.[67]

The proposal is extraordinary due to the breadth of the power and the significance of the subject matter. The Governor-General could change fundamental electoral procedures with virtually no Parliamentary oversight. This type of clause is often ‘vituperative[ly]’ described as a ‘Henry VIII clause’ due to its association with autocracy.[68] However, the risk of abuse could be mitigated through the drafting process. First, the power could only come in to effect if a court had invalidated a sufficient portion of the Electoral Act after the dissolution of Parliament and it was not possible to proceed with the election under existing law. Second, amendments could be limited to the constitutionally invalid elements of the former law. Third, the power could only be exercisable upon the assent of the leader of the Government and the Opposition. Finally, the power could include a sunset clause terminating the operation of the regulation three months after the first sitting of Parliament. Imposing a sunset clause would effectively obligate the new Parliament to replace the regulation with a new statute. Although the proposed power is highly unusual, it could avert potential constitutional crises without the need to resort to the doctrine of necessity.

IV The Constitutional Mandate of Popular Suffrage

In Murphy, the plaintiffs argued that the closure of the electoral rolls seven days after the issue of the writ was incompatible with the constitutional mandate of popular suffrage. The High Court first recognised the existence of a universal adult franchise in Roach.[69] The franchise is constitutionally entrenched through the mandate of popular suffrage. The mandate is an implied restriction on Commonwealth legislative power. It is derived from the scheme of representative government established in the text and structure of the Constitution, particularly by the requirement of ‘direct choice by the people’ in ss 7 and 24.[70] The High Court has held that to give effect to the mandate, a law that limits the universal adult franchise will be invalid unless it is for a ‘substantial reason’.[71] Acceptable limitations include the exclusion of minors and the mentally infirm from voting.[72] The mandate was applied in Roach and Rowe to invalidate federal electoral laws that restricted the franchise.

Notwithstanding the passage of time since Roach, the principles that govern the constitutional mandate remain unsettled. The object of this Part is to distil the principles that govern the guarantee in light of the developments in Murphy. Part IVA will outline the constitutional basis for the mandate and Part IVB will examine the underlying theory of representative government. Part IVC will consider the present status of the mandate. The analysis reveals that the scope of the constitutional mandate was restricted at almost every turn in Murphy.

A The Source of the Constitutional Mandate

The Constitution does not provide an express right to vote.[73] Nor does it contain an express right to liberty and equal protection like the United States Constitution.[74] Instead, the constitutional mandate operates as an implied limitation on the Parliament’s power to make laws defining the franchise in federal elections. There are two relevant heads of power. The Parliament may make laws concerning the qualification of voters under a combination of ss 8, 30 and 51(xxxvi) of the Constitution. In addition, the Parliament may make laws concerning the method, or system, for electing the House of Representatives and the Senate under a combination of ss 10, 31 and 51(xxxvi) of the Constitution. These provisions, combined with the remainder of ch I, provide the Commonwealth Parliament with a ‘plenary power over federal elections’.[75]

There are few fetters on the Parliament’s power to make federal electoral laws. The Parliament’s choice of electoral design is subject to the requirement in ss 7 and 24 of the Constitution that the Senate and the House of Representatives must be ‘directly chosen by the people’. The ‘direct’ election requirement is a bar to indirect methods of election, such as the electoral college model used in the United States.[76] The requirement of election by ‘the people’ enshrines popular participation.[77] Other provisions expressly require that ‘each elector shall vote only once’[78] and the method for electing senators must be ‘uniform for all the states’.[79]

Otherwise, the Australian Parliament is vested with a broad discretion to design the electoral system, including the franchise. The constitutional scheme is characterised by a ‘deliberate lack of specificity’[80] that fixes ‘only the broadest outlines’[81] of the electoral system and vests a ‘considerable measure of legislative freedom as to the method of choice of the members of Parliament’.[82] In an oft-cited passage, Reid and Forrest argue that Parliament’s discretion includes determining

the method of voting to elect the members of the respective houses; the question of whether members of the House of Representatives would be elected by single-member or multi-member divisions; ... who would be authorised to vote; the question of voluntary or compulsory registration of voters and of voting itself; the control of electoral rolls; the conduct of the ballot; ... the location of responsibility for the administration of the electoral law ...[83]

In recent years, the High Court has curbed Parliament’s discretion through the implication of constitutional guarantees. The Court observed in Lange that the requirement of direct choice by the people in ss 7 and 24 embraces ‘all that is necessary to effectuate the free election of representatives at periodic elections’.[84] The ‘necessary’ features of the electoral system are explained by the concept of representative government.[85] The Constitution establishes a scheme where the Executive Government is accountable to the Legislature and, in turn, the Legislature is accountable to the people at periodic elections.[86] As noted by Sir Samuel Griffith, the Constitution intends that ‘the actual government of the State is conducted by officers who enjoy the confidence of the people’.[87]

There are two ‘indispensable incidents’ or ‘essential concomitants’ flowing from the scheme of representative government prescribed by the Constitution.[88] First, it requires a freedom to communicate about public affairs since ‘it would be a parody of democracy to confer on the people a power to choose their Parliament but to deny the freedom of public discussion from which the people derive their political judgments’.[89] This guarantee takes the form of the implied freedom of political communication. Second, the scheme of representative government requires a universal adult franchise for Commonwealth elections since ‘[v]oting in elections for the Parliament lies at the very heart of the system of government’.[90] This guarantee takes the form of the constitutional mandate of popular choice. The implication of constitutional guarantees on the basis of representative government has been the topic of considerable controversy.[91]

The content of the constitutional mandate remains contested. It is clear that the concept of a universal franchise is consistent with periodic elections of the Houses of Parliament by ‘the people’. Yet the Constitution does not establish an express franchise or chart the precise ‘metes and bounds’ of representative government.[92] It does not specify who comprises ‘the people’ or on what grounds a person may be legitimately excluded from voting. Further difficulties arise from using an imprecise concept of representative government to assess the validity of technical electoral laws.

B Theory of Representative Government

The High Court of Australia has overcome the absence of an express definition of representative government in the Constitution through theories of constitutional interpretation. The underlying theory of representative government is important since it determines whether a constitutionally protected franchise exists; and, if so, the boundaries of the franchise. The Court initially adopted a narrow interpretation of the scheme of representative government in the Constitution. However, the breadth of the interpretation increased, culminating in the broadest definition in Rowe. Murphy represents a withdrawal from the position in Rowe.

The High Court initially adopted an originalist interpretation of the constitutional scheme of representative government. The requirements of representative government were fixed to the limitations recognised at the time the Constitution was passed.[93] Justice Gibbs argued that the Court’s duty was to ‘declare the law as enacted in the Constitution and not to add to its provisions new doctrines which may happen to conform to our own prepossessions’.[94] Accordingly, the Parliament was free to exclude persons from the franchise who were excluded in 1900. The Constitution does not prescribe a universal adult franchise since no such franchise existed at the time of Federation.[95]

Another approach recognised that the understanding of representative government could change over time. In McKinlay, McTiernan and Jacobs JJ argued that the meaning of the phrase ‘chosen by the people’ ‘depends in part upon the common understanding of the time’.[96] Justice Gummow later argued in McGinty that ‘representative government is a dynamic rather than a static institution and one that has developed in the course of [the 20th] century’.[97] This view accepted that changes in historical facts and social attitudes may alter the requirements of representative government imposed by the Constitution. A majority of the High Court accepted this view in Roach, and consequently, a constitutionally mandated universal adult franchise was born.[98]

Rowe established the high watermark of representative government. Relying in part on Roach and the ‘common understanding’ approach to constitutional interpretation, French CJ adopted an ‘evolutionary’ interpretation of representative government.[99] This theory accepted that the Parliament could expand the franchise. However, once expanded, the Parliament was not competent to ‘wind-back’ or limit the franchise.[100] Justice Crennan adopted a similar proposition, arguing that the ‘evolution’[101] of representative government and democratic principles ‘would constrain any reversion to arbitrary exclusions from the franchise’.[102] This method of constitutional interpretation was subject to extensive criticism following Rowe.[103]

Murphy suggests that the High Court is withdrawing from the evolutionary interpretation of representative government. Justice Kiefel criticised this method of interpretation in her dissent in Rowe.[104] None of the four Justices who joined the Court after Rowe endorsed the evolutionary approach.[105] Instead, Keane, Nettle and Gordon JJ expressly questioned whether evolution is a criterion of constitutional invalidity.[106] These Justices affirmed the importance of avoiding ‘abstract’ notions of representative government and locating the requirements of the constitutional mandate in the text of the Constitution.[107]

While Gageler J did not criticise Rowe, his Honour adopted a narrower understanding of representative government. His Honour accepted that representative government ‘“depends in part upon the common understanding of the time”’.[108] However, his Honour emphasised that ‘[t]he understanding of the time ... is not a concern of the moment or a hope for the future ... [the mandate is informed by a] relatively stable and enduring understanding of the nature of our system of representative government’.[109] There are two noteworthy features of this passage. Justice Gageler adopts the nomenclature of the ‘common understanding of the time’ and emphasises the importance of stability in the interpretation of representative government. Both features are associated with the intermediate view of representative government that preceded Rowe.

Finally, Keane J reserved the possibility of reverting to an originalist interpretation of representative government. His Honour reserved the question of whether ss 7 and 24 ‘describe a standard characteristic of the Houses of Parliament’.[110] His Honour referred to obiter dicta in Roach, in which Hayne J asserted in dissent that the phrase ‘directly 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’.[111] In accordance with an originalist interpretation, this view fixes the ‘standard’ required by representative government at the time of Federation. The adoption of this view would cast doubt on the existence of a constitutional universal adult franchise and the correctness of Roach and Rowe.

The form of representative government required by the Constitution was not resolved in Murphy. Yet, Murphy suggests that the Court is shifting towards a narrower interpretation of representative government, especially when viewed in light of Kiefel J’s dissent in Rowe.

C The State of the Mandate

The generality of the phrase ‘the constitutional mandate’ is apt to mislead. In Murphy, the High Court of Australia affirmed that the constitutional mandate is a limitation on two different types of laws: laws concerning the substantive entitlement to vote and laws regulating electoral procedures. Although the assessment of validity in each class is informed by ‘close analogy’, the laws impose different types of burden on the franchise.[112] In Murphy, the Court narrowed the constitutional mandate’s application in the context of procedural laws, but did not disturb existing jurisprudence on substantive disqualifications.

The first class of laws involves a substantive ‘disqualification’.[113] The Australian Parliament has the power to make laws for the qualification of electors under ss 8, 30, 51(xxxvi) of the Constitution. The power also permits the Parliament to disqualify electors.[114] However, this power is subject to the constitutional mandate of popular choice. A law that removes a person’s substantive entitlement to vote is a ‘disqualification’ that will infringe the constitutional mandate if it is not done for a substantial reason.

The impugned laws in Roach effected a disqualification from voting. The majority struck down a law that excluded any person incarcerated on polling day from voting in a federal election.[115] High Court obiter dicta suggests that the removal of an entitlement to vote on the basis of race,[116] religion,[117] gender,[118] property ownership[119] or another arbitrary reason[120] is invalid because it breaches the constitutional mandate. Fixing a maximum voting age might also breach the mandate.[121] Many of these exclusions were permissible at the time of Federation. However, the development of representative government has been relied upon to assert that an exclusion on these bases would burden the constitutional mandate. With the exception of Keane J’s reservation concerning the ‘standardised characteristic’ of ss 7 and 24, Murphy does not cast doubt on the status of this type of burden under the constitutional mandate.

The second and more controversial class of burden concerns the machinery of the electoral system.[122] The Australian Parliament has the power to make laws defining the method of electing the House of Representatives and the Senate under ss 10, 31 and 51(xxxvi) of the Constitution. The Parliament does not directly regulate the franchise under this head of power. However, it may impose procedural and administrative preconditions that must be satisfied before a person can vote. Exclusion from voting on this basis is not a ‘disqualification’ since the person could have voted if they had complied with the procedure. Instead, these laws are ‘practical impediment[s]’ or limitations on the ‘opportunity’ to qualify to vote.[123] The majority in Rowe accepted that a law that imposed procedural requirements on acquiring the right to vote may impose a burden on the constitutional mandate.[124]

The constitutional mandate was used to strike down a procedural law in Rowe.[125] Although citizens over the age of 18 were entitled to vote,[126] the right to vote did not accrue until the person submitted an application to the Electoral Commissioner to be added to the electoral roll.[127] The impugned provisions prevented the Commissioner from considering a claim for enrolment between the issue of the writ for the election and the polling day.[128] Persons who had not made a claim for enrolment before this period could not vote in the election even if they were otherwise entitled to enrol. The exclusion was the result of a failure to comply with the enrolment procedure. A similar procedural provision was at issue in Murphy.

In Murphy, the High Court narrowed the extent to which the constitutional mandate applied to procedural laws. This can be explained in part by the adoption of a more restrictive theory of representative government. Rowe demonstrates that procedural laws limiting the franchise were at the margin of the constitutional mandate’s sphere of operation. Accordingly, the contraction of the underlying theory of representative government in Murphy is likely to reduce the number of procedural laws that are regulated by constitutional mandate.

Further, three Justices in Murphy rejected the broad interpretation of the constitutional mandate which Gummow and Bell JJ advanced in Rowe. Justices Gummow and Bell argued that the constitutional mandate required the electoral system to be ‘as expressive of the popular choice as practical considerations properly permit’.[129] This interpretation of the constitutional mandate imposed a more onerous burden on Parliament. Instead of merely restricting disqualifications or other impediments to voting, it imposed a positive obligation to maximise the opportunity for potential electors to participate in the electoral process, subject to practical limitations. Justices Keane, Nettle and Gordon expressly rejected this view in Murphy, arguing that the bare scheme of representative government in the Constitution does not require the maximisation of participation in the electoral process.[130] Justice Kiefel expressed a similar view during oral argument in Murphy.[131]

Finally, the Court applied Rowe narrowly in Murphy. A majority of the Court refused to recognise the existence of a burden on the constitutional mandate in Murphy,[132] even though similar legislation was struck down in Rowe. Both enactments involved the temporary suspension of additions or amendments to the electoral rolls prior to the election. The provisions in Rowe closed the rolls on the day the writ was issued whereas the provisions in Murphy closed the rolls seven days after the writ. Rowe was distinguished on the basis that Murphy did not result in a reduction or diminishment of the franchise.[133] The legislation in Rowe involved the removal of an existing opportunity for enrolment, whereas the provisions impugned in Murphy were a longstanding limitation on enrolment.

There are a number of difficulties with the majority distinction. Although the absence of a ‘wind-back’ was a convenient basis for distinguishing Rowe, the distinction is predicated on the evolutionary understanding of representative government. Several Justices expressly questioned or contradicted this reasoning in Murphy.[134] The distinction is inconsistent with the narrower theories of representative government that were advanced in the case. Further, it was not clear that the evolutionary theory of representative government commanded the assent of all members of the majority in Rowe.[135] The use of the distinction in Murphy has the potential to embed the controversial theory in the constitutional mandate.

In any event, the distinction suggests that longevity, or the absence of a constitutional challenge, is a criterion of constitutional validity. It cannot be the case that a law or doctrine is constitutionally valid simply because it has existed for a long time. Otherwise, the doctrine of terra nullius would not have been overturned in Mabo[136] or 37-year-old redistricting laws would not have been invalidated in McKinlay.[137]

Finally, the distinction seems to preference form over substance.[138] It places significance on whether the legislation takes the form of an amending act or principal legislation, rather than considering the practical operation of the Act. The location of the impugned provisions in the principal Electoral Act was relevant insofar as it prevented the Court from relying on the remedial approach it used in Roach and Rowe. However, remedial difficulties alone cannot save an unconstitutional law from being invalidated.

A further consideration identified by Gageler J was the absence of discrimination in Murphy. His Honour characterised the disenfranchisement of prisoners in Roach and the young in Rowe as ‘freez[ing]’ discrete minority interests out of the political process.[139] This type of discrimination was a ‘central concern’ underlying the constitutional mandate.[140] Given the similarity between Rowe and Murphy, it is difficult to see why the provisions in Murphy would not discriminate against the young. Nonetheless, discrimination is an important concept in Australian constitutional jurisprudence[141] that may provide fertile ground for developing the constitutional mandate.

Justice Kiefel observed that ‘Rowe and Roach effected something of a turning in the law’.[142] Murphy represents a turn-back. Even though the Commonwealth did not seek leave to overturn Rowe,[143] several Justices criticised the decision and expressed reservations about the correctness of its reasoning. The Court gave Rowe a narrow reading and refused to apply it to a similar legislative scheme. Murphy demonstrates a shift in High Court’s interpretation of the constitutional mandate of popular suffrage. The burden borne by a plaintiff seeking to invoke the mandate has increased since the decision in Murphy.

V The Proportionality Test

No constitutional guarantee is absolute. The constitutional mandate of popular suffrage is no exception. A law that burdens the constitutional mandate will not be invalid if it is for a ‘substantial reason’.[144] In Murphy, the High Court accepted that a proportionality test should be used to determine whether a reason is ‘substantial’.[145] However, the form of the proportionality test was disputed. Six Justices refused to apply a structured proportionality assessment even though a majority in McCloy had recently applied it the context of the implied freedom of political communication.

A proportionality test is a method of assessing the validity of legislation. It involves an examination of whether the means adopted by the Parliament are ‘proportionate’ to the achievement of a legislative object.[146] Proportionality tests have been applied in a variety of Australian constitutional contexts, including to determine whether a law is within a purposive or incidental head of power, or if a law impermissibly infringes a constitutional guarantee or immunity.[147]

There are different mechanisms for assessing proportionality. Traditionally, Australian courts ask whether the measure is ‘reasonably appropriate and adapted to achieve a legitimate end’. In Lange, the High Court unanimously adopted this test for assessing burdens on the implied freedom of political communication.[148] It was also applied by the plurality in Roach and Rowe to assess the validity of laws that burdened the constitutional mandate.[149] This method does not define the considerations a judge is to take into account when discharging the proportionality analysis. Instead, it vests the judge with a wide discretion to be exercised in accordance with principles derived from analogous cases.[150]

In McCloy, a four-Justice majority of the High Court applied a more prescriptive proportionality test in the context of the implied freedom of political communication. Drawing on international jurisprudence, particularly from the Federal Republic of Germany, the majority devised a series of tests to be applied when the Court assesses proportionality. Instead of applying the Lange test, the Court examined:

1. Suitability — whether the law has a rational connection to the purpose of the provision.

2. Necessity — whether there was an obvious and compelling alternative, reasonably practicable means of achieving the same purpose with a less restrictive effect on the freedom.

3. Adequate in balance — whether the extent of the restriction imposed by the impugned law was outweighed by the importance of the purpose which it served.[151]

The law will not be proportionate unless the Court answers each question affirmatively. Each of these inquiries could have been undertaken under the Lange test. However, the novelty lies in the prescription of a rigid schema of factors to be considered whenever the proportionality analysis is engaged. As Stone has observed, the primary distinction between the approaches is the ‘degree of flexibility or discretion they leave in the hands of the judge applying them’.[152] The remaining Justices in McCloy did not apply a standardised proportionality assessment, instead adhering to the ‘reasonably appropriate and adapted’ formulation.[153]

The merits and demerits of structured proportionality testing are contested. Proponents argue that standardised proportionality analysis inheres transparency in what is otherwise a value-laden and impressionistic judgment.[154] Critics argue that the imposition of standardised criteria is incompatible with the operation of established proportionality principles and may invite the Court to breach the separation of powers.[155] Others downplay the practical effect of structured proportionality, as ‘[t]he adoption of [structured proportionality] in McCloy did not reflect the birth of some exotic jurisprudential pest destructive of the delicate ecology of Australian public law’.[156]

McCloy cast a shadow of uncertainty over the method of assessing the proportionality of laws that burden other constitutional guarantees. The content of each McCloy inquiry was substantially derived from cases on the implied freedom of political communication.[157] Yet in other jurisdictions, including the Federal Republic of Germany, a single generic rubric is applied for the assessment of proportionality.[158] The terms of the McCloy test were not limited to the implied freedom of political communication, extending to determining whether a measure ‘exceed[s] the implied limitation on legislative power’.[159]

Murphy was the first time that the High Court of Australia considered the application of structured proportionality analysis outside the implied freedom of political communication. The plaintiffs’ argument was framed in accordance with the structured proportionality test adopted in McCloy.[160] However, Kiefel J was the only member of the McCloy majority to accept the application of structured proportionality in Murphy.[161] Chief Justice French and Justice Bell joined the minority judges in McCloy to apply the ‘reasonably appropriate and adapted’ test.[162] Justice Keane did not need to apply a proportionality test since his Honour held that the impugned law did not impose a burden on the constitutional mandate.[163]

The application of structured proportionality was said to be warranted by the ‘affinity’ between the implied freedom of political communication and the constitutional mandate.[164] After accepting the application of the ‘reasonably appropriate and adapted’ test, the plurality in Roach observed that ‘[t]he affinity to what is called the second question in Lange will be apparent’.[165] The affinity has two bases. First, the implied freedom of political communication and the constitutional mandate of popular choice are drawn from a common constitutional source. Both guarantees are implications derived from the scheme of representative government established by the Constitution, in particular, ss 7 and 24. Neither of these guarantees take the form of a personal right; instead operating as a restriction on the grants of power in the Constitution.[166] Second, in both cases, the Court applies a proportionality test to assess the validity of a law that infringes a constitutional guarantee. Prior to McCloy, an identical formula was used to assess proportionality under each guarantee.[167] In light of the development in McCloy, the ‘affinity’ with the implied freedom was said to warrant the application of structured proportionality to the mandate of popular choice.

The majority did not accept this argument. The application of structured proportionality to the implied freedom of political communication was not sufficient to warrant its application to the constitutional mandate. The affinity was not an identity. After considering the nature of the constitutional mandate, the majority Justices concluded that structured proportionality was not amenable to the unique characteristics of the guarantee. Justice Gordon observed that:

It should not be assumed that, because a particular test for proportionality has been adopted in one particular constitutional context, it can be uncritically transferred into another context, constitutional or otherwise, even within the same jurisdiction.[168]

Part VA–VD will examine whether the majority’s refusal to extend structured proportionality to the constitutional mandate was justified in light of McCloy. There are four key differences between the constitutional mandate of popular choice and the implied freedom of political communication; namely, the scope, the judicial function, the relevance of alternatives and the role of the Parliament. Analysis reveals that there was sufficient divergence between the constitutional guarantees to warrant the retention of the existing proportionality test. However, this does not preclude the extension of structured proportionality to other constitutional guarantees in future.

A Sphere of Operation

The implied freedom of political communication has a wide sphere of operation.[169] The guarantee is a limitation on laws and actions that infringe the freedom of political communication established under the Constitution. It constrains the Legislature, Executive and common law[170] at the local government, state, territory and Commonwealth level.[171] The implied freedom extends to any subject matter which concerns political communication including parole limitations on speech and movement,[172] regulation of duck hunting,[173] criminalisation of insulting language,[174] and the common law of defamation.[175]

The operation of the constitutional mandate is more confined. The constitutional mandate only constrains two heads of Commonwealth legislative power; namely, the power concerning the qualification of electors derived from ss 8, 30 and 51(xxxvi), and the method of electing the Parliament under ss 10, 31 and 51(xxxvi). The constitutional mandate has no application to a state election unless the State’s Constitution entrenches a requirement that its Legislature is ‘chosen by the people’.[176] The laws impugned by the constitutional mandate are exclusively concerned with the electoral system. Examples include limitations on the registration of political parties,[177] the form of the ballot,[178] distribution of electoral boundaries[179] and disqualifications[180] or other burdens[181] on the universal adult franchise.

The limited operation of the constitutional mandate abrogates many of the advantages of a structured proportionality assessment. The rigid approach is said to be necessary to facilitate consistent decision-making in cases with divergent facts and to provide guidance to lower courts.[182] The approach is effective in the context of the implied freedom, since the guarantee encompasses an array of laws that may arise in trial and intermediate appellate courts.[183] However, additional guidance is unnecessary in the context of the constitutional mandate given its limited scope and the High Court’s role as the primary arbiter of challenges to Commonwealth electoral laws. The constitutional mandate is well suited to the flexibility and analogy based reasoning of the existing proportionality test.

B Balancing the Judicial Function

The implied freedom and the constitutional mandate countenance different levels of judicial intervention. The constitutional mandate is exclusively concerned with the design of the electoral system. Designing a system which accurately and efficiently facilitates a choice by over 15 million people is no easy task. A range of considerations must be balanced, including timeliness, finality, cost, precision, the avoidance of fraud and the maintenance of public confidence. The challenge is compounded by the unique characteristics of Australia’s geography. As Gordon J observed in Murphy, ‘[i]t is not a matter of devising and drafting a single provision or division of an Act. There are countless variations of [an electoral] system.’[184] As a result, the High Court has demonstrated significant deference to the Australian Parliament’s choice of electoral design. Justices have frequently emphasised the substantial ‘room’[185] left in the Constitution’s design of representative government and the ‘wide powers’[186] of the Parliament to legislate for federal elections.

The same institutional challenges are not engaged under the implied freedom of political communication. The application of the implied freedom outside the electoral system makes the guarantee more amenable to judicial intervention. Courts are familiar with the assessment of whether a general legislative or executive action is within a head of power or impermissibly infringes a constitutional guarantee. Unlike electoral laws, the legislation is generally less technical and less likely to give rise to major cost. Questions about the operation of a complex wider scheme, such as an electoral system, rarely arise.[187] As Hayne J observed in dissent in Rowe, under the implied freedom of political communication, ‘[n]o more particular question about the form of representative government, let alone the form of electoral system, need[s] to be considered’.[188] The High Court has shown a greater willingness to intervene in the context of the implied freedom. It has affirmed that the judicial function does not permit the Court to ‘substitute [its] own assessment for that of the legislative decisionmaker’.[189] However, the Court has formally rejected the existence of a ‘margin of appreciation’ for legislative judgment under the implied freedom[190] and frequently affirmed the need for judicial intervention to enforce the guarantee.[191]

A standardised proportionality analysis does not account for the different judicial function under each guarantee. Although the role of the Court is not explicit in the reasonably appropriate and adapted formulation, the heavy reliance on precedent ensured the Court exercised the appropriate level of deference. The adoption of a general ‘adequate in balance’ criteria distracts attention from analogous cases and asks the Court to make a value judgment about the importance of the legislative purpose. Applying a rigid standard risks the development of a uniform standard of intervention that does not account for the unique constitutional characteristics of the mandate.

C Reasonable Necessity

A related concern identified in Murphy was the requirement to consider alternatives.[192] The ‘reasonable necessity’ stage of the McCloy test invites the Court to consider whether there are any ‘obvious and compelling alternative, reasonably practicable means of achieving the same purpose which has a less restrictive effect on the freedom’.[193]

Prior to McCloy, the role of alternatives in the implied freedom of political communication was contested.[194] The High Court unanimously accepted the appropriateness of considering alternatives in Lange by asserting the determinative factor in ACTV was the existence of ‘other less drastic means by which the objectives of the law could be achieved’.[195] However, several Justices subsequently raised concerns about the compatibility of considering alternatives and the judicial function.[196] The concern was resolved in McCloy by requiring the alternative to be ‘obvious and compelling’.[197]

It is questionable whether it is appropriate to consider alternatives under the constitutional mandate given Parliament’s broad discretion to design the electoral system. In any event, judicial assessment of alternatives is difficult in light of the complexity of the electoral system. As Gordon J emphasised in Murphy, the Court is not considering a single provision, but a cog in an ‘entire legislative scheme’.[198] In these circumstances, it is unlikely that an alternative is likely to reach the high threshold of being obvious and compelling. Instead, the consideration of alternatives is more likely to be an attempt to incrementally improve the design of the electoral system.

These difficulties were demonstrated by the plaintiff’s attempted reliance on reasonable necessity in Murphy. As Kiefel J summarised, the plaintiff’s ‘principal submission ... was that there are reasonably practicable alternative means which would provide for a longer period of enrolment’.[199] The plaintiffs adduced evidence from the electoral systems in New South Wales, Victoria and Queensland, which provided a greater opportunity to enrol than the Federal electoral system. The Court comprehensively rejected the argument. It impermissibly required additional expenditure of public funds[200] and did not account for the differing context of a federal and state elections.[201] At its core, the argument was an attempt to improve electoral design.[202]

Even if consideration of alternatives is warranted, the sequencing in the McCloy test places undue emphasis on reasonable necessity. The structured proportionality test requires the court to consider reasonable necessity each time a proportionality analysis is undertaken. Assessment of reasonable necessity will always precede an assessment of ‘balance’. Twomey has criticised the ‘rigidity’ of this schema[203] and Stellios argues that it places a ‘premium’ on reasonable necessity.[204] The schema misled the plaintiffs in Murphy to base their primary submission on the existence of alternatives. The questionable relevance of reasonable necessity under the constitutional mandate justifies the decision not to extend structured proportionality analysis in Murphy.

D The Nature of the Power

Finally, the implied freedom and the constitutional mandate concern different types of powers. The implied freedom of political communication primarily attaches to the Commonwealth’s legislative power in s 51 of the Constitution.[205] These heads of power are plenary in scope and are as ‘wide ... as can be created’.[206] Section 51 powers are discretionary since the Parliament is not obliged to exercise the power.[207]

The constitutional mandate of popular choice conditions Parliament’s power to make laws for federal elections. This power appears to be discretionary on the face of the Constitution. Sections 8, 10, 30 and 31 provide for the operation of state electoral systems and franchise laws in federal elections ‘until the Parliament otherwise provides’. Even though the Parliament could replace these laws under s 51(xxxvi) of the Constitution, it was not obligated to do so. State laws could have operated indefinitely.[208] The framers of the Constitution expected, rather than obliged, the Parliament to design an electoral system.[209]

This discretion was transformed into an obligation once the Commonwealth Parliament extinguished the operation of state electoral laws in 1902. The Parliament enacted the Commonwealth Electoral Act 1902 (Cth) and the Commonwealth Franchise Act 1902 (Cth) under its heads of power to make laws for federal elections. After this time, the operation of state laws could not be revived.[210] Since the free election of the Legislature is an indispensable incident of the scheme of representative government contemplated by the Constitution,[211] the Parliament incurred an obligation to maintain laws for an electoral system and franchise that facilitated the direct choice of the Legislature by the people.

During oral argument in Murphy, the Solicitor-General of Australia asserted that the existence of a positive obligation to maintain electoral laws rendered structured proportionality testing inapposite to the constitutional mandate.[212] Justice Gordon accepted this argument.[213] However, it is difficult to understand how the characterisation of the underlying power as a ‘discretion’ or ‘obligation’ changes the assessment of proportionality. Structured proportionality may be more suitable in the context of constitutional obligations because the High Court might be expected to scrutinise the discharge of obligations more closely than discretionary judgments.[214] The key distinction between the guarantees is Parliament’s broad discretion to design an electoral system rather than the character of the power as an obligation.

E Future Implications

The existence of an ‘affinity’ between the implied freedom of political communication and the constitutional mandate of popular suffrage is undeniable. Both guarantees rely on proportionality tests and are drawn from the same system of representative government prescribed by the Constitution. However, the human experience demonstrates that a progenitor does not produce identical progeny. The majority’s refusal to apply structured proportionality in Murphy was justified and necessary to protect the identity of the constitutional mandate of popular choice.

Murphy casts doubt on the appropriate form of other proportionality assessments in Australia. Chief Justice French and Justice Bell indicated that the structured test in McCloy is ‘a mode of analysis applicable to some cases involving the general proportionality criterion, but not necessarily all’.[215] The decision in Murphy raises the question of whether structured proportionality testing should be applied to purposive heads of power and other constitutional guarantees.[216]

The High Court is unlikely to apply structured proportionality to determine whether a law falls within a legislative head of power. A proportionality test has been applied to the assessment of laws under the defence,[217] external affairs (with respect to treaty implementation),[218] incidental[219] and ‘nationhood’ powers.[220] However, the assessment of proportionality in these contexts involves ‘a high level of deference to the law-makers’ selection of an appropriate means’.[221] Unlike the Lange test, the Court asks whether the object of the law is reasonably capable of being seen as appropriate and adapted to the power.[222] The addition of the objective ‘reasonably capable’ element gives Parliament a greater degree of deference. The Court refused to add these words to the Lange test in Coleman v Power, since it amounted to a ‘surrender’[223] or otherwise ‘weaken[ed]’[224] the Court’s capacity to enforce the implied freedom of political communication. Other concerns arise from the limited role for considering alternatives in the context of purposive powers. For instance, in relation to the defence power, Dixon J asserted that ‘the possibility of achieving the same ends by other measures are no concern of the court’.[225] Structured proportionality analysis is not amenable to assessing whether legislation falls within a head of power.[226]

The position is less clear with respect to other constitutional limitations on power. The primary question is whether structured proportionality analysis will be applied in the context of s 92 of the Constitution. Section 92 requires that ‘trade, commerce and intercourse among the States ... shall be absolutely free’. Currently, the High Court examines whether the law imposes a burden of a protectionist kind on interstate trade and commerce.[227] A law that imposes a discriminatory burden will not be invalid if it is ‘reasonably necessary’ for the protection of the public.[228]

The relationship between s 92 and the implied freedom of political communication is well established. The assessment of reasonable necessity involves a proportionality assessment.[229] Section 92 cases were extensively considered in the development of structured proportionality[230] and the Betfair criterion of ‘reasonable necessity’ is embedded in the McCloy test. Kirk has advocated for the adoption of a structured proportionality assessment in the context of s 92, explaining the decision in Castlemaine Tooheys[231] through a schema resembling the McCloy test.[232] Justice Kiefel issued a Roach-like harbinger by stating that ‘[a]n analogy with the test confirmed in Betfair, and in Lange, can be drawn.’[233]

However, there are several key differences between s 92 and the implied freedom of political communication. First, s 92 is an express constitutional limitation. The McCloy test is expressed to apply to ‘implied limitation[s] on legislative power’.[234] Second, the constitutional mandate is arguably more closely connected to the implied freedom of political communication. The High Court’s refusal to extend structured proportionality to the constitutional mandate foreshadows a reluctance to extend the schema to s 92. Finally, the relevance of the ‘suitability’ and ‘balancing’ tests may be in doubt given the Betfair majority limited the assessment of proportionality to ‘reasonable necessity’.

VI Conclusion

A combination of seemingly unrelated factors almost created a constitutional crisis in Murphy. Upholding the plaintiffs claim would have required the High Court to choose between rewriting electoral legislation or creating a situation where a new Australian Parliament could not be elected under orthodox constitutional principles. The Governor-General of Australia could have restored the status quo under the doctrine of necessity. Nonetheless, Murphy raises an important issue about how to preserve the system of representative government and ensure that the people can elect a new Parliament if electoral legislation is invalidated while the Parliament is dissolved.

Murphy also provides insight into the future of the constitutional mandate of popular suffrage. The case foreshadows a narrow interpretation of the constitutional mandate and signals a withdrawal from the high watermark of representative government established in Rowe. The difficulty encountered by plaintiffs seeking to invoke the constitutional mandate has undoubtedly increased after Murphy.

Finally, the decision in Murphy casts doubt on the method of assessing proportionality in Australian constitutional law. In McCloy, a majority of the High Court accepted the application of a structured proportionality analysis after a sustained judicial and academic campaign spanning 25 years. Less than a year later, the McCloy majority split, refusing to apply the structured proportionality rubric to a constitutional guarantee that bore an ‘affinity’ to the implied freedom of political communication. The inappositeness of structured proportionality to the constitutional mandate is a reminder that any proportionality test, rigid or flexible, must remain faithful to its constitutional guarantee.


[∗] BPESS (Syd) JD (Syd) (graduand). My thanks to Professor Anne Twomey for her advice and guidance. My thanks to Professor Joellen Riley and the Sydney Law Review for making this case note possible. Any errors are my own.

[1] (2016) 90 ALJR 1027 (‘Murphy’).

[2] Australian Constitution ss 7, 24 (‘Constitution’).

[3] Electoral Act s 101.

[4] Commonwealth Electoral Act 1902 (Cth) s 64; Electoral Act s 45 (as originally enacted).

[5] Graeme Orr, ‘The Voting Rights Ratchet: Rowe v Electoral Commissioner(2011) 22(2) Public Law Review 83, 84.

[6] Section 102(4) of the Electoral Act suspends additions and amendments to the roll between the closure of the rolls and the election. Section 155 specifies that the rolls close seven days after the issue of the writ. The Act also establishes mirroring suspension periods for specific enrolment categories: ss 94A(4), 95(4), 96(4), 103A(4), 103B(5), 118(5).

[7] Electoral and Referendum Amendment (Electoral Integrity and Other Measures) Act 2006 (Cth).

See also Joint Standing Committee on Electoral Matters, Parliament of Australia, The 2004 Federal Election: Report of the Inquiry into the Conduct of the 2004 Federal Election and Matters Related Thereto (2005) 34–6 [2.112]–[2.126].

[8] (2010) 243 CLR 1 (‘Rowe’).

[9] Electoral and Referendum Amendment (Enrolment and Prisoner Voting) Act 2011 (Cth) s 3, sch 1 items 2–8.

[10] Constitution ss 7, 24.

[11] Anthony Murphy, ‘Plaintiff’s Annotated Submissions’, Submission in Murphy v Electoral Commissioner, Case No M247/2015, 11 April 2016, 1–2 [6]–[8].

[12] The case was dismissed ex tempore: Murphy [2016] HCATrans 111 (12 May 2016) 101–2 [4444]–[4485] (French CJ). The Court subsequently delivered its reasons: Murphy (2016) 90 ALJR 1027, 1040 [43] (French CJ and Bell J), 1041 [44] (Kiefel J), 1047 [83] (Gageler J), 1052 [117] (Keane J), 1066 [228] (Nettle J), 1073 [259] (Gordon J).

[13] Joint Standing Committee on Electoral Matters, Parliament of Australia, The 2010 Federal Election: Report on the Conduct of the Election and Related Matters (2011) 7 [2.10].

[14] Nicholas Horne, ‘Hung Parliaments and Minority Governments’ (Parliamentary Library, Parliament of Australia, 2010).

[15] Anthony Murphy, ‘Plaintiff’s Annotated Submissions’, Submission in Murphy v Electoral Commissioner, Case No M247/2015, 11 April 2016, 9 [34]. Data concerning enrolment in the 2016 Election not available at time of publication.

[16] Australian Electoral Commission, 2016 Federal Election: Divisional Results (8 August 2016) Australian Electoral Commission <http://results.aec.gov.au/20499/Website/HouseDivisionalResults-20499.htm> .

[17] The Houses of Parliament were dissolved on 9 May 2016: Commonwealth, Gazette: Proclamation — Dissolution of the Senate and the House of Representatives Simultaneously at 9:00 am on Monday, 9 May 2016, No C2016G00628, 8 March 2016. The High Court heard oral argument for Murphy on 11 and 12 May 2016. For an overview of the political circumstances surrounding the 2016 dissolution, see: Rosemary Laing (ed), Odgers’ Australian Senate Practice: As Revised by Harry Evans (Commonwealth of Australia, 14th revised ed, 2016) 761–4.

[18] Rowe (2010) 243 CLR 1, 40 [87] (Gummow and Bell JJ); Roach v Electoral Commissioner [2007] HCA 43; (2007) 233 CLR 162, 202 [96]–[97] (Gummow, Kirby and Crennan JJ) (‘Roach’).

[19] Roach [2007] HCA 43; (2007) 233 CLR 162, 198 [81] (Gummow, Kirby and Crennan JJ).

[20] Cam & Sons v Chief Secretary (NSW) [1951] HCA 59; (1951) 84 CLR 442, 454 (Dixon, Williams, Webb, Fullagar and Kitto JJ). See also Monis v The Queen (2013) 249 CLR 92, 208 [327]–[329] (Crennan, Kiefel and Bell JJ) (‘Monis’); Tajjour v New South Wales [2014] HCA 35; (2014) 254 CLR 508, 585–6 [168]–[171] (Gageler J) (‘Tajjour’).

[21] Pidoto v Victoria [1943] HCA 37; (1943) 68 CLR 87, 111 (Latham CJ).

[22] Section 93 of the Electoral Act defines who is entitled to enrol. Sections 94, 94A, 95, 96, 96A and 99 outline the limitations on enrolment and transfer of enrolment.

[23] Ibid s 101.

[24] Ibid.

[25] Ibid s 102.

[26] Ibid ss 94A(4), 95(4), 96(4), 102(4).

[27] Section 208 of the Electoral Act requires the Electoral Commissioner to prepare a ‘certified list’ of enrolled voters for each division and distribute a copy of each division’s list to the presiding officer of every polling place in the division.

[28] Electoral Act s 166.

[29] Ibid ss 123, 126.

[30] Ibid pt IV.

[31] Ibid ss 90A, 90B.

[32] Ibid s 90B(1) item 1(a).

[33] Commonwealth of Australia, ‘Defence of the Second Defendant to the Amended Statement of Claim’, Submission in Murphy v Electoral Commissioner, No M247/2015, 23 March 2016, 27–8 [34].

[34] Commonwealth of Australia, ‘Annotated Submissions of the Second Defendant’, Submission in Murphy v Electoral Commissioner, Case No M247/2015, 26 April 2016, 20 [106].

[35] Anthony Murphy, ‘Plaintiff’s Annotated Submissions’, Submission in Murphy v Electoral Commissioner, Case No M247/2015, 11 April 2016, 19–20 [66]–[70]; Murphy [2016] HCATrans 111 (12 May 2016) 89–90 [3890]–[3955] (Justin Gleeson SC).

[36] Joint Standing Committee on Electoral Matters, above n 7, 11 [2.15].

[37] Murphy (2016) 90 ALJR 1027, 1051 [109] (Gageler J).

[38] Ibid 1045 [73] (Kiefel J), 1060 [188]–[190] (Keane J), 1072 [253] (Nettle J).

[39] Murphy [2016] HCATrans 111 (12 May 2016) 90 [3953] (Justin Gleeson SC).

[40] Ibid 89–90 [3900]–[3956] (Justin Gleeson SC).

[41] Electoral Act pt III.

[42] Ibid pt XX.

[43] Ibid pt XXI.

[44] Ibid pt XIX.

[45] Ibid pt II.

[46] Ibid pt XXII.

[47] Victoria v Commonwealth [1957] HCA 54; (1957) 99 CLR 575, 603–5 (Dixon CJ) (‘Second Uniform Tax Case’).

[48] Joseph Chitty, A Treatise on The Law of the Prerogatives of the Crown; and the Relative Duties and Rights of the Subject (Butterworths, 1820) 68; Laing, above n 17, 604–5.

[49] See Laing, above n 17, 604.

[50] Victoria v Commonwealth [1975] HCA 39; (1975) 134 CLR 81, 120 (Barwick CJ).

[51] ‘To prorogue Parliament means to bring an end a session of Parliament without dissolving the House of Representatives or both Houses’: Laing, above n 17, 186. See also Sir John Quick and Sir Robert Randolph Garran, The Annotated Constitution of the Australian Commonwealth (Angus & Robertson, 1901) 407; Chitty, above n 48, 71–2.

[52] The combination of ss 8, 10, 30, 31 and 51(xxxvi) of the Constitution provide the Legislature with a ‘plenary power over federal elections’: Smith v Oldham [1912] HCA 61; (1912) 15 CLR 355, 363 (Isaacs J).

[53] The Executive cannot exercise legislative power without the authorisation of the Legislature: Victorian Stevedoring & General Contracting Co Pty Ltd v Dignan [1931] HCA 34; (1931) 46 CLR 73, 101–2 (Dixon J) (‘Victorian Stevedoring’).

[54] Anne Twomey, ‘Parliament, the Executive and Vice-Regal Reserve Powers: Heading off Crises in a Closely Tied Parliament’ (2016 Annual Harry Evans Lecture, Parliament of Australia, 4 November 2016) 11. See also Anne Twomey, ‘The Fijian Coup Cases: The Constitution, Reserve Powers and the Doctrine of Necessity’ (2009) 83(5) Australian Law Journal 319.

[55] The necessity power is limited to the duration of the crisis. Lord Erskine argued in 1808 that

The King ... may do various acts growing out of sudden emergencies; but in all these cases the emergency is the avowed cause, and the act done is as temporary as the occasion. The King cannot change by, his prerogative ... either the law of nations or the law of the land, by general and unlimited regulations ...

United Kingdom, Parliamentary Debates, House of Lords, 8 March 1808, vol 10, col 959, quoted in Chitty, above n 48, 50.

[56] Constitution s 61. There is a debate in the defence context as to whether these terms codify existing prerogative powers or provide the Executive with an additional power to quell subversion: see James Stellios, Zines’s the High Court and the Constitution (Federation Press, 6th ed, 2015) 27–35. However, this debate is unlikely to affect the existence of a necessity power since ‘there can be no doubt that the Executive Government of the Commonwealth has an executive power to protect the nation’: Davis v Commonwealth [1988] HCA 63; (1988) 166 CLR 79, 110 (Brennan J) (‘Davis’).

[57] Farey v Burvett [1916] HCA 36; (1916) 21 CLR 433, 454 (Isaacs J).

[58] Chitty, above n 48, 68. The first example arose from the restoration of the British Monarchy in 1660. The Parliament met to invite Charles II to return to the throne and re-establish the Monarchy. The second example concerned the ‘Glorious Revolution’ in 1689. The King, James II, fled the country and the Parliament met to determine his replacement. Both cases involved a dissolved Parliament recalling itself since there was no reigning Monarch to summon the Parliament.

[59] Ibid.

[60] Ibid (emphasis added).

[61] Ibid.

[62] ‘[T]here are dangers in maintaining a structure which lends itself to the concentration of political power in the Executive Government. There is a risk of efficiency turning into tyranny’: Sir Gerard Brennan, ‘The Parliament, the Executive and the Courts: Roles and Immunities’ [1997] BondLawRw 10; (1997) 9(2) Bond Law Review 136, 143. For an example of a recent abuse of the doctrine of necessity abroad,

see Twomey, ‘The Fijian Coup Cases’, above n 54.

[63] [2016] HCA 20; (2016) 90 ALJR 639 (‘Day’).

[64] Croome v Tasmania (1997) 191 CLR 119, 132–3 (Gaudron, McHugh and Gummow JJ).

[65] The difficulties posed by standing are exemplified by the timeline in Rowe. The plaintiffs in the case were Australian citizens who sought to enrol or change their enrolment during the suspension period before the 2010 Election. The electoral rolls were closed on 19 July 2010 and the Special Case was filed on 26 July 2010. The High Court heard the matter on 4 and 5 August 2010 and a majority declared that the impugned legislation was invalid on 6 August 2010, 15 days before the 2010 Election.

[66] Plaintiff S157/2002 v Commonwealth [2003] HCA 2; (2003) 211 CLR 476, 512 (Gaudron, McHugh, Gummow, Kirby and Hayne JJ). However, there is High Court obiter dicta suggesting that a bare attempt to delegate an entire head of legislative power would not be valid: Victorian Stevedoring (1931)

[1931] HCA 34; 46 CLR 73, 119–20 (Evatt J).

[67] An unprecedented number of parties registered to contest the 1999 New South Wales State Election. No printer in Australia could print a ballot paper that accommodated all the parties in the form prescribed by the Parliamentary Electorates and Elections Act 1912 (NSW). There was a possibility that the Election would need to be delayed to source the ballot papers; yet the New South Wales Constitution fixed the date of the Election. The Constitution did not allow the Election to be delayed and the dissolution of the Parliament prevented the amendment of the Constitution or electoral legislation. The situation was resolved through an almost-forgotten regulatory power. The Governor changed the form of the ballot paper under s 176(3) of the Parliamentary Electorates and Elections Act 1912 (NSW) to enable all parties to be listed on a ballot paper capable of being printed:

see Parliamentary Electorates and Elections Amendment (Ballot-Papers) Regulation 1999 (NSW).

[68] John Willis, The Parliamentary Powers of English Government Departments (Harvard University Press, 1933) 17.

[69] Roach [2007] HCA 43; (2007) 233 CLR 162, 174 [7] (Gleeson CJ), 198–9 [80]–[83] (Gummow, Kirby and Crennan JJ).

[70] Ibid.

[71] Ibid 174 [7] (Gleeson CJ), 199 [85] (Gummow, Kirby and Crennan JJ).

[72] Mulholland v Australian Electoral Commission [2004] HCA 41; (2004) 220 CLR 181, 191 [14] (Gleeson CJ) (‘Mulholland’); McGinty v Western Australia [1996] HCA 48; (1996) 186 CLR 140, 166–7 (Brennan CJ) (‘McGinty’).

[73] Section 41 of the Constitution provides that ‘[n]o adult person who has or acquires a right to vote at elections for the more numerous House of the Parliament of a State shall, while the right continues, be prevented by any law of the Commonwealth from voting at elections for either House of the Parliament of the Commonwealth’. A six Justice majority of the High Court held in R v Pearson;

Ex parte Sipka that s 41 only protected the voting rights acquired prior to the enactment of the Commonwealth franchise: [1983] HCA 6; (1983) 152 CLR 254, 260–1 (Gibbs CJ, Mason and Wilson JJ),

279 (Brennan, Deane and Dawson JJ). See Anne Twomey, ‘The Federal Constitutional Right to Vote in Australia’ [2000] FedLawRw 6; (2000) 28(1) Federal Law Review 125.

[74] The United States ‘right to vote’ cases are predicated on a combination of the liberty and equal protection clauses: see Harper v Virginia State Board of Elections, [1966] USSC 58; 383 US 663 (1966); Kramer v Union Free School District No 15[1969] USSC 151; , 395 US 621 (1969); Reynolds v Sims, [1964] USSC 202; 377 US 533 (1964); City of Mobile v Bolden, [1980] USSC 69; 446 US 55 (1980).

[75] Smith v Oldham [1912] HCA 61; (1912) 15 CLR 355, 363 (Isaacs J).

[76] Constitution ss 7, 24; A-G (Cth) ex rel McKinlay v Commonwealth [1975] HCA 53; (1975) 135 CLR 1, 21 (Barwick CJ), 44 (Gibbs J), 56 (Stephen J), 61 (Mason J) (‘McKinlay’); McGinty [1996] HCA 48; (1996) 186 CLR 140, 243 (McHugh J); Murphy (2016) 90 ALJR 1027, 1055 [147] (Keane J).

[77] Constitution ss 7, 24; McKinlay [1975] HCA 53; (1975) 135 CLR 1, 21 (Barwick CJ), 44 (Gibbs J), 56 (Stephen J), 61 (Mason J).

[78] Constitution ss 8, 30.

[79] Ibid s 9.

[80] Mulholland [2004] HCA 41; (2004) 220 CLR 181, 189 [9] (Gleeson CJ).

[81] Rowe (2010) 243 CLR 1, 76 [221] (Hayne J).

[82] Ibid 50 [125] (Gummow and Bell JJ). See further McGinty [1996] HCA 48; (1996) 186 CLR 140, 169 (Brennan CJ), 182–3 (Dawson J), 244 (McHugh J), 270 (Gummow J).

[83] Gordon Stanley Reid and Martyn Forrest, Australia’s Commonwealth Parliament 1901–1988:

Ten Perspectives (Melbourne University Press, 1989) 86–7 cited in McGinty (1996) 196 CLR 140, 283–4 (Gummow J); Rowe (2010) 243 CLR 1, 121 [386] (Kiefel J); Murphy (2016) 90 ALJR 1027, 1059 [182] (Keane J).

[84] Lange v Australian Broadcasting Corporation [1997] HCA 25; (1997) 189 CLR 520, 557 (The Court) (‘Lange’).

[85] Ibid 557 (The Court).

[86] Ibid 559 (The Court).

[87] Sir Samuel Griffith, Notes on the Australian Federation: Its Nature and Probable Effects (Edmund Gregory, 1986) 17.

[88] Lange [1997] HCA 25; (1997) 189 CLR 520, 559 (The Court); Australian Capital Television Pty Ltd v Commonwealth [1992] HCA 45; (1992) 177 CLR 106, 136 (Mason CJ) (‘ACTV’).

[89] Nationwide News v Wills [1992] HCA 46; (1992) 177 CLR 1, 47 (Brennan J) (‘Nationwide News’). See also Lange [1997] HCA 25; (1997) 189 CLR 520, 559 (The Court).

[90] Roach [2007] HCA 43; (2007) 233 CLR 162, 198 [81] (Gummow, Kirby and Crennan JJ). See also 174 [7] (Gleeson CJ).

[91] ACTV [1992] HCA 45; (1992) 177 CLR 106, 186 (Dawson J); Australian Broadcasting Corporation v Lenah Game Meats (2001) 208 CLR 199, 337–8 [347] (Callinan J); Roach [2007] HCA 43; (2007) 233 CLR 162, 219–20 [161]–[162] (Hayne J); Rowe (2010) 243 CLR 1, 97–100 [292]–[304] (Heydon J). See also Jeffrey Goldsworthy, ‘Constitutional Implications Revisited’ (2011) 30(1) Queensland University Law Journal 9; James Allan, ‘The Three ‘Rs’ of Recent Australian Judicial Activism: Roach, Rowe and (No)’riginalism’ [2012] MelbULawRw 19; (2012) 36(2) Melbourne University Law Review 743; James Allan, ‘Implied Rights and Federalism: Inventing Intentions While Ignoring Them’ [2009] UWALawRw 2; (2009) 34(2) University of Western Australia Law Review 228. Cf George Williams, ‘Sounding the Core of Representative Democracy: Implied Freedoms and Electoral Reform’ [1996] MelbULawRw 6; (1996) 20(3) Melbourne University Law Review 848; Stephen Donaghue, ‘The Clamour of Silent Constitutional Principles’ [1996] FedLawRw 5; (1996) 24(1) Federal Law Review 133; Orr, above n 5; Jeremy Kirk, ‘Constitutional Implications (II): Doctrines of Equality and Democracy’ [2001] MelbULawRw 2; (2001) 25(1) Melbourne University Law Review 24.

[92] Roach [2007] HCA 43; (2007) 233 CLR 162, 206 [113] (Hayne J). See also Murphy (2016) 90 ALJR 1027, 1048 [89] (Gageler J): ‘To attempt to pin [the concept of representative government] down more tightly would be to fail to grasp its meaning; it defies being diced or squashed to fit within a judicially constructed box.’

[93] McKinlay [1975] HCA 53; (1975) 135 CLR 1, 21 (Barwick CJ), 44 (Gibbs J), 62 (Mason J); Roach [2007] HCA 43; (2007) 233 CLR 162, 219–20 [161]–[162] (Hayne J); Rowe (2010) 243 CLR 1, 97–100 [292]–[304] (Heydon J).

[94] McKinlay [1975] HCA 53; (1975) 135 CLR 1, 44 (Gibbs J).

[95] For an overview of the state of the franchise at the time of Federation, see Rowe (2010) 243 CLR 1, 108–12 [333]–[347] (Crennan J).

[96] [1975] HCA 53; (1975) 135 CLR 1, 36.

[97] [1996] HCA 48; (1996) 186 CLR 140, 280. See also Mulholland [2004] HCA 41; (2004) 220 CLR 181, 189–90 [10] (Gleeson CJ).

[98] Roach [2007] HCA 43; (2007) 233 CLR 162, 174 [7] (Gleeson CJ), 198–9 [83] (Gummow, Kirby and Crennan JJ).

[99] Rowe (2010) 243 CLR 1, 18–19 [18]–[22] (French CJ).

[100] Ibid.

[101] Ibid 107 [328] (Crennan J).

[102] Ibid 115 [356] (Crennan J).

[103] Anne Twomey, ‘Rowe v Electoral Commissioner — Evolution or Creationism?’ [2012] UQLawJl 9; (2012) 31(2) University of Queensland Law Journal 181, 183–95 cited in Murphy (2016) 90 ALJR 1027, 1078 [291] n 373, 1081 [309] n 389 (Gordon J); James Allan, ‘The Three ‘Rs’ of Recent Australian Judicial Activism’, above n 91. Cf Orr, above n 5.

[104] Rowe (2010) 243 CLR 1, 129–31 [417]–[423] (Kiefel J).

[105] Justice Edelman has also been appointed to the High Court since the decision in Rowe. However, his Honour has not had an opportunity to express an opinion on the issue since Murphy preceded his Honour’s appointment.

[106] Murphy (2016) 90 ALJR 1027, 1065 [222] (Keane J), 1069–70 [243] (Nettle J), 1080–1 [309] (Gordon J).

[107] Ibid 1058–9 [177]–[179] (Keane J), 1069–70 [243] (Nettle J), 1080 [305] (Gordon J).

[108] Ibid 1048 [90] (Gageler J), quoting McKinlay [1975] HCA 53; (1975) 135 CLR 1, 36 (McTiernan and Jacobs JJ).

[109] Ibid 1048–9 [91] (Gageler J).

[110] Ibid 1062 [200] (Keane J).

[111] Roach [2007] HCA 43; (2007) 233 CLR 162, 219–20 [161]–[162] (Hayne J).

[112] Rowe (2010) 243 CLR 1, 20 [23] (French CJ). There is little difference between the Court’s approach to assessing the validity of each type of law. The High Court could have looked to authority in the implied freedom of political communication where the standard of the test will vary according to the ‘precise nature and degree of the restriction which each of the impugned provisions imposes’ on the freedom: McCloy v New South Wales [2015] HCA 34; (2015) 257 CLR 178, 239 [156] (Gageler J) (‘McCloy’).

See also ACTV [1992] HCA 45; (1992) 177 CLR 106, 143–4 (Mason CJ); Tajjour [2014] HCA 35; (2014) 254 CLR 508, 579–581 [149]–[152] (Gageler J). However, French CJ expressly rejected the application of this reasoning to the constitutional mandate in Rowe: (2010) 243 CLR 1, 21 [26].

[113] Murphy (2016) 90 ALJR 1027, 1036 [26] (French CJ and Bell J), 1042 [52] (Kiefel J), 1047–8 [86] (Gageler J), 1066 [224] (Keane J), 1069 [240] (Nettle J), 1080 [307] (Gordon J).

[114] An argument that ss 8, 30 and 51(xxxvi) of the Constitution empower the Parliament to qualify, but not disqualify electors, was not accepted in Roach: [2007] HCA 43; (2007) 233 CLR 162, 174 [7] (Gleeson CJ),

199 [85] (Gummow, Kirby and Crennan JJ), 216 [147]–[149] (Hayne J).

[115] Roach [2007] HCA 43; (2007) 233 CLR 162, 184–5 [35]–[39] (Gleeson CJ).

[116] Rowe (2010) 243 CLR 1, 115 [356] (Crennan J); Mulholland [2004] HCA 41; (2004) 220 CLR 181, 261 [232] (Kirby J).

[117] Roach [2007] HCA 43; (2007) 233 CLR 162, 174–5 [8] (Gleeson CJ).

[118] Rowe (2010) 243 CLR 1, 115 [356] (Crennan J); Mulholland [2004] HCA 41; (2004) 220 CLR 181, 261 [232] (Kirby J).

[119] McGinty [1996] HCA 48; (1996) 186 CLR 140, 166–7 (Brennan CJ).

[120] Rowe (2010) 243 CLR 1, 115 [356] (Crennan J).

[121] Mulholland [2004] HCA 41; (2004) 220 CLR 181, 191 [14] (Gleeson CJ).

[122] Murphy (2016) 90 ALJR 1027, 1036 [26] (French CJ and Bell J), 1042 [52] (Kiefel J), 1047–8 [86] (Gageler J), 1066 [225] (Keane J), 1069 [239] (Nettle J), 1080 [307] (Gordon J).

[123] Ibid 1038 [34] (French CJ and Bell J), 1047–8 [86] (Gageler J).

[124] Rowe (2010) 243 CLR 1, 20 [24] (French CJ), 57 [153] (Gummow and Bell JJ), 119 [381] (Crennan J).

[125] (2010) 243 CLR 1, 24 [36] (French CJ).

[126] Electoral Act s 93 as at 9 August 2010.

[127] Ibid s 101 as at 9 August 2010.

[128] Ibid s 102(4) as at 9 August 2010.

[129] Rowe (2010) 243 CLR 1, 57 [154] (Gummow and Bell JJ).

[130] (2016) 90 ALJR 1027, 1059 [180] (Keane J), 1069 [240] (Nettle J), 1081–2 [316] (Gordon J).

[131] [2016] HCATrans 108 (11 May 2016) 16–17 [644]–[729] (Kiefel J). See also Rowe (2010) 243 CLR 1, 128–9 [412]–[415] (Kiefel J).

[132] (2016) 90 ALJR 1027, 1040 [40] (French CJ and Bell J), 1062 [202] (Keane J), 1080 [308] (Gordon J).

[133] Ibid 1035 [25] (French CJ and Bell J), 1042 [52] (Kiefel J), 1050–1 [105]–[108] (Gageler J), 1069 [239]–[240] (Nettle J), 1080–1 [309] (Gordon J).

[134] Ibid 1065 [222] (Keane J), 1069–70 [243] (Nettle J), 1080–1 [309] (Gordon J). See also Rowe (2010) 243 CLR 1, 129–31 [417]–[423] (Kiefel J).

[135] Justices Gummow and Bell JJ appeared to base their judgment on a requirement that the Parliament must maximise participation in the electoral system rather than an evolutionary interpretation of representative government: Rowe (2010) 243 CLR 1, 57 [154]. See also Murphy (2016) 90 ALJR 1027, 1081 [310] (Gordon J).

[136] Mabo v Queensland (No 2) [1992] HCA 23; (1992) 175 CLR 1 (‘Mabo’).

[137] [1975] HCA 53; (1975) 135 CLR 1.

[138] Ha v New South Wales [1997] HCA 34; (1997) 189 CLR 465, 498 (Brennan CJ, McHugh, Gummow and Kirby JJ).

[139] Murphy (2016) 90 ALJR 1027, 1050–1 [106]–[107] (Gageler J).

[140] Ibid.

[141] See Amelia Simpson, ‘The High Court’s Conception of Discrimination: Origins, Applications and Implications’ [2007] SydLawRw 10; (2007) 29(2) Sydney Law Review 263. Justice Gageler has also identified the relevance of discrimination in the context of the implied freedom of political communication: McCloy [2015] HCA 34; (2015) 257 CLR 178, 233–4 [136]–[138].

[142] Murphy (2016) 90 ALJR 1027, 1042 [53] (Kiefel J).

[143] Ibid 1038 [36] (French CJ and Bell J), 1043 [59] (Kiefel J), 1065 [222] (Keane J); Commonwealth of Australia, ‘Annotated Submissions of the Second Defendant’, Submission in Murphy v Electoral Commissioner, Case No M247/2015, 26 April 2016, 10 [64].

[144] Murphy (2016) 90 ALJR 1027, 1036–7 [28]–[31] (French CJ and Bell J), 1043 [61] (Kiefel J),

1047 [85] (Gageler J), 1063–4 [206]–[210] (Keane J), 1069 [239] (Nettle J), 1078 [291] (Gordon J).

[145] Ibid 1037 [32] (French CJ and Bell J), 1043–4 [61]–[65] (Kiefel J), 1047 [85] (Gageler J), 1070 [244] (Nettle J), 1083 [332] (Gordon J).

[146] Rowe (2010) 243 CLR 1, 131 [425]–[426] (Kiefel J).

[147] McCloy [2015] HCA 34; (2015) 257 CLR 178, 195 [3] (French CJ, Kiefel, Bell and Keane JJ).

[148] [1997] HCA 25; (1997) 189 CLR 520, 562 (The Court).

[149] Roach [2007] HCA 43; (2007) 233 CLR 162, 199 [85] (Gummow, Kirby and Crennan JJ); Rowe (2010) 243 CLR 1, 59 [161] (Gummow and Bell JJ).

[150] Adrienne Stone, ‘The Limits of Constitutional Text and Structure: Standards of Review and the Freedom of Political Communication’ [1999] MelbULawRw 26; (1999) 23(3) Melbourne University Law Review 668, 692.

[151] McCloy [2015] HCA 34; (2015) 257 CLR 178, 194–5 [2] (French CJ, Kiefel, Bell and Keane JJ) as expressed in Murphy (2016) 90 ALJR 1027, 1038–9 [37] (French CJ and Bell J).

[152] Stone, above n 150, 688.

[153] [2015] HCA 34; (2015) 257 CLR 178, 238 [150] (Gageler J), 273 [269] (Nettle J), 282 [311] (Gordon J).

[154] See Murphy (2016) 90 ALJR 1027, 1043–4 [61]–[65] (Kiefel J); McCloy [2015] HCA 34; (2015) 257 CLR 178,

212–13 [66]–[68] (French CJ, Kiefel, Bell and Keane JJ); Tajjour [2014] HCA 35; (2014) 254 CLR 508, 574–5 [127]–[133] (Crennan, Kiefel and Bell JJ); Monis (2013) 249 CLR 92, 213–14 [345]–[346] (Crennan, Kiefel and Bell JJ); Rowe (2010) 243 CLR 1, 131–45 [424]–[478] (Kiefel J); Jeremy Kirk, ‘Constitutional Guarantees, Characterisation and the Concept of Proportionality’ [1997] MelbULawRw 1; (1997) 21(1) Melbourne University Law Review 1, 19–21; Gabrielle Appleby, ‘Proportionality and Federalism: Can Australia Learn from the European Community, the US and Canada?’ [2007] UTasLawRw 1; (2007) 26(1) University of Tasmania Law Review 1; Stone, above n 150, 691.

[155] See Murphy (2016) 90 ALJR 1027, 1050 [101]–[103] (Gageler J), 1079–80 [296]–[304] (Gordon J); McCloy [2015] HCA 34; (2015) 257 CLR 178, 236–9 [145]–[153] (Gageler J), 281–2 [308]–[311], 287–9 [336]–[339] (Gordon J); Anne Twomey, ‘Proportionality and the Constitution’ (Speech delivered at ALRC Freedoms Symposium, Federal Court, Queens Square, Sydney, 8 October 2015) <https://www.alrc.gov.au/proportionality-constitution-anne-twomey>; Stone, above n 150, 689–91; Leslie Zines, ‘Federalism and Administrative Discretion in Australia, with European Comparisons’ [2000] FedLawRw 14; (2000) 28(2) Federal Law Review 291, 302; Stellios, above n 56, 596.

[156] Murphy (2016) 90 ALJR 1027, 1039 [37] (French CJ and Bell J).

[157] McCloy [2015] HCA 34; (2015) 257 CLR 178, 195–6 [4] (French CJ, Kiefel, Bell and Keane JJ).

[158] For an overview, see Patrick Quirk, ‘An Australian Looks at German “Proportionality”’ [1999] UNDAULawRw 5; (1999) 1 University of Notre Dame Australia Law Review 39, 40–7.

[159] [2015] HCA 34; (2015) 257 CLR 178, 194 [2], 195 [2] (French CJ, Kiefel, Bell and Keane JJ).

[160] Anthony Murphy, ‘Plaintiff’s Annotated Submissions’, Submission in Murphy v Electoral Commissioner, Case No M247/2015, 11 April 2016, 8–9 [29].

[161] (2016) 90 ALJR 1027, 1043–4 [61]–[65] (Kiefel J).

[162] Ibid 1037 [32] (French CJ and Bell J), 1047 [85] (Gageler J), 1070 [244] (Nettle J), 1083 [332] (Gordon J).

[163] Ibid 1062 [202] (Keane J).

[164] Ibid 1043–4 [63] (Kiefel J); Anthony Murphy, ‘Plaintiff’s Annotated Submissions’, Submission in Murphy v Electoral Commissioner, Case No M247/2015, 11 April 2016, 8–9 [29].

[165] [2007] HCA 43; (2007) 233 CLR 162, 199 [86] (Gummow, Kirby and Crennan JJ).

[166] With respect to the implied freedom of political communication: Lange [1997] HCA 25; (1997) 189 CLR 520, 560 (The Court); Unions NSW v New South Wales (2013) 252 CLR 530, 554 [36] (French CJ, Hayne, Crennan, Kiefel and Bell JJ) (‘Unions NSW’); McCloy [2015] HCA 34; (2015) 257 CLR 178, 202 [29] (French CJ, Kiefel, Bell and Keane JJ). With respect to the constitutional mandate: Roach [2007] HCA 43; (2007) 233 CLR 162, 199–200 [86] (Gummow, Kirby and Crennan JJ); Rowe (2010) 243 CLR 1, 126–7 [406]–[407] (Kiefel J).

[167] Roach [2007] HCA 43; (2007) 233 CLR 162, 199 [85] (Gummow, Kirby and Crennan JJ).

[168] Murphy (2016) 90 ALJR 1027, 1079 [296] (Gordon J).

[169] Lange [1997] HCA 25; (1997) 189 CLR 520, 557 (The Court).

[170] Ibid 566 (The Court).

[171] Unions NSW (2013) 252 CLR 530, 550 [24]–[25] (French CJ, Hayne, Crennan, Kiefel and Bell JJ); ACTV [1992] HCA 45; (1992) 177 CLR 106, 142 (Mason CJ). See further Anne Twomey, ‘The Application of the Implied Freedom of Political Communication to State Electoral Funding Laws’ [2012] UNSWLawJl 26; (2012) 35(3) University of New South Wales Law Journal 625.

[172] Wotton v Queensland (2012) 246 CLR 1.

[173] Levy v Victoria [1997] HCA 31; (1997) 189 CLR 579 (‘Levy’).

[174] Coleman v Power [2004] HCA 39; (2004) 220 CLR 1; Monis (2013) 249 CLR 92.

[175] Lange [1997] HCA 25; (1997) 189 CLR 520.

[176] Stephens v West Australian Newspapers [1994] HCA 45; (1994) 182 CLR 211, 233–4 (Mason CJ, Toohey and Gaudron JJ), 236 (Brennan J).

[177] Mulholland [2004] HCA 41; (2004) 220 CLR 181.

[178] Langer v Commonwealth [1996] HCA 43; (1996) 186 CLR 302 (‘Langer’); Day [2016] HCA 20; (2016) 90 ALJR 639.

[179] McGinty (1996) 196 CLR 140.

[180] Roach [2007] HCA 43; (2007) 233 CLR 162.

[181] Rowe (2010) 243 CLR 1; Murphy (2016) 90 ALJR 1027.

[182] Stone, above n 150, 691–2.

[183] For example, Monis was an appeal from a criminal conviction originating in the New South Wales District Court. At trial, the appellant argued that the implied freedom of political communication rendered invalid the legislation under which he was charged: (2013) 249 CLR 92, 106 [7] (French CJ).

[184] (2016) 90 ALJR 1027, 1080 [301] (Gordon J).

[185] Mulholland [2004] HCA 41; (2004) 220 CLR 181, 237 [154] (Gummow and Hayne JJ).

[186] Langer [1996] HCA 43; (1996) 186 CLR 302, 343 (McHugh J). See also McGinty [1996] HCA 48; (1996) 186 CLR 140, 169 (Brennan CJ), 182–3 (Dawson J), 244 (McHugh J), 270 (Gummow J); Mulholland [2004] HCA 41; (2004) 220 CLR 181,

188 [6] (Gleeson CJ), 207 [64] (McHugh J).

[187] Although, the implied freedom of political communication may apply to electoral laws. For example, the High Court struck down restrictions on political broadcasting in ACTV [1992] HCA 45; (1992) 177 CLR 106.

[188] (2010) 243 CLR 1, 68–9 [195] (Hayne J).

[189] McCloy [2015] HCA 34; (2015) 257 CLR 178, 219 [89] (French CJ, Kiefel, Bell and Keane JJ) (citations omitted).

[190] Unions NSW (2013) 252 CLR 530, 553 [34] (French CJ, Hayne, Crennan, Kiefel and Bell JJ).

Cf ACTV [1992] HCA 45; (1992) 177 CLR 106, 158–9 (Brennan J).

[191] ACTV [1992] HCA 45; (1992) 177 CLR 106, 144 (Mason CJ); Cunliffe v Commonwealth [1994] HCA 44; (1994) 182 CLR 272, 340 (Deane J) (‘Cunliffe’); Levy [1997] HCA 31; (1997) 189 CLR 579, 646–7 (Kirby J); Coleman v Power (2004)

[2004] HCA 39; 220 CLR 1, 48 [87] (McHugh J), 78 [196] (Gummow and Hayne JJ), 82 [212] (Kirby J).

[192] (2016) 90 ALJR 1027, 1039 [39] (French CJ and Bell J), 1051 [109] (Gageler J), 1072 [251]–[254] (Nettle J), 1080 [303] (Gordon J).

[193] McCloy [2015] HCA 34; (2015) 257 CLR 178, 195 [2] (French CJ, Kiefel, Bell and Keane JJ).

[194] Even in McCloy, Nettle J considered there were differing views and uncertainty as to ‘what significance should be attributed to the availability of less restrictive means’: [2015] HCA 34; (2015) 257 CLR 178, 259 [221].

[195] Lange [1997] HCA 25; (1997) 189 CLR 520, 568 (The Court).

[196] Unions NSW (2013) 252 CLR 530, 576 [129] (Keane J); Levy [1997] HCA 31; (1997) 189 CLR 579, 598 (Brennan CJ); Coleman v Power [2004] HCA 39; (2004) 220 CLR 1, 31 [31] (Gleeson CJ), 123–4 [328] (Heydon J).

[197] [2015] HCA 34; (2015) 257 CLR 178, 211 [58] (French CJ, Kiefel, Bell and Keane JJ). See also Monis (2013)

249 CLR 92, 214 [347] (Crennan, Kiefel and Bell JJ).

[198] (2016) 90 ALJR 1027, 1080 [303]–[304] (Gordon J).

[199] Ibid 1042 [51] (Kiefel J).

[200] Ibid 1044 [65] (Kiefel J), 1060 [188]–[190] (Keane J), 1072 [253] (Nettle J).

[201] Ibid 1064 [215] (Keane J).

[202] Ibid 1039 [39] (French CJ and Bell J), 1051 [109]–[110] (Gageler J).

[203] Twomey, above n 155.

[204] Stellios, above n 56, 596.

[205] Nationwide News [1992] HCA 46; (1992) 177 CLR 1, 76 (Deane and Toohey JJ).

[206] Bank of NSW v Commonwealth [1948] HCA 7; (1948) 76 CLR 1, 186 (Latham CJ) (‘Bank Nationalisation Case’). See also: at 332 (Dixon J); Grain Pool of Western Australia v Commonwealth (2000) 202 CLR 479, 492 (Gleeson CJ, Gaudron, McHugh, Gummow, Hayne and Callinan JJ).

[207] The scope of the Parliament’s discretion is so voluminous that the Commonwealth may ‘cover [a] field’ of legislative power to render state legislation inoperative under s 109, even if the Commonwealth has no intention of introducing substantive legislation on the subject: New South Wales v Commonwealth [2006] HCA 52; (2006) 229 CLR 1, 164–9 (Gleeson CJ, Gummow, Hayne, Heydon and Crennan JJ) (‘Work Choices Case’).

[208] Roach [2007] HCA 43; (2007) 233 CLR 162, 212 [132] (Hayne J). For example, Dixon CJ recognised that s 96 of the Constitution empowered the Commonwealth to make tied grants indefinitely, notwithstanding the power is limited ‘until the Parliament otherwise provides’ and the framers intended the provision to have a transitory operation. ‘Either the power terminates or it continues’: Second Uniform Tax Case [1957] HCA 54; (1957) 99 CLR 575, 603–5.

[209] See further Roach [2007] HCA 43; (2007) 233 CLR 162, 208–12 [121]–[133] (Hayne J).

[210] Second Uniform Tax Case [1957] HCA 54; (1957) 99 CLR 575, 603–5 (Dixon CJ).

[211] Mulholland [2004] HCA 41; (2004) 220 CLR 181, 190 [14] (Gleeson CJ); Roach [2007] HCA 43; (2007) 233 CLR 162, 198 [80] (Gummow, Kirby and Crennan JJ).

[212] Murphy [2016] HCATrans 111 (12 May 2016) 86 [3761]–[3765] (Justin Gleeson SC).

[213] Murphy (2016) 90 ALJR 1027, 1079–80 [301]–[302] (Gordon J).

[214] Although one must exercise great caution when using an administrative tail to wag a constitutional dog, the principles of administrative law provide a useful illustration. Courts are reluctant to impose mandatory considerations on an administrative officer who is reposed with a discretion in comparison to an administrative officer who is subject to an obligation: Minister for Aboriginal Affairs v

Peko-Wallsend Ltd (1986) 162 CLR 24, 40–2 (Mason J).

[215] Murphy (2016) 90 ALJR 1027, 1039 [37] (French CJ and Bell J).

[216] Justice Brennan identified these classes of proportionality analysis in Cunliffe: [1994] HCA 44; (1994) 182 CLR 272, 322–6.

[217] Constitution s 51(vi); Stenhouse v Coleman [1944] HCA 36; (1944) 69 CLR 457, 471 (Dixon J). Traditionally, the Court asks whether the object of the law is reasonably capable as being seen as appropriate and adapted to the defence of Australia: Polyukhovich v Commonwealth [1991] HCA 32; (1991) 172 CLR 501, 697 (Gaudron J) (‘Polyukhovich’). However, Thomas v Mowbray [2007] HCA 33; (2007) 233 CLR 307 casts doubt on this test. Justice Kirby was the only Justice to apply the traditional test: at 411 [296]. Three Justices asked whether the law was appropriate and adapted to protecting the public from a terrorist attack: at 325–6 [9] (Gleeson CJ), 459–60 [443]–[444] (Hayne J), 504 [588] (Callinan J). Justices Gummow and Crennan did not apply a proportionality test: at 363 [145].

[218] Constitution s 51(xxix); Victoria v Commonwealth (1996) 187 CLR 416, 487 (Brennan CJ, Toohey, Gaudron, McHugh and Gummow JJ) (‘Industrial Relations Act Case’).

[219] The use of proportionality testing for the incidental power is controversial: Stellios, above n 56, 56–8.

[220] Davis [1988] HCA 63; (1988) 166 CLR 79, 100 (Mason CJ, Deane and Gaudron JJ), 101 (Wilson and Dawson JJ). However, the plurality did not appear to apply a proportionality assessment in Pape v Federal Commissioner of Taxation [2009] HCA 23; (2009) 238 CLR 1, 87 [228] (Gummow, Crennan and Bell JJ). See also Davis [1988] HCA 63; (1988) 166 CLR 79, 111 (Brennan J).

[221] Stone, above n 150, 676.

[222] See, eg, Polyukhovich [1991] HCA 32; (1991) 172 CLR 501, 697 (Gaudron J).

[223] [2004] HCA 39; (2004) 220 CLR 1, 82 [212] (Kirby J).

[224] Ibid 78 [196] (Gummow and Hayne JJ). See also: 48 [87] (McHugh J).

[225] Australian Communist Party v Commonwealth [1951] HCA 5; (1951) 83 CLR 1, 198 (Dixon J).

[226] Cf Kirk, above n 154, 21–5. Kirk outlines a structured proportionality schema which is amenable to characterising laws under purposive heads of power: above n 154, 21–42. This approach has not been adopted.

[227] Cole v Whitfield [1988] HCA 18; (1988) 165 CLR 360, 394 (The Court).

[228] Betfair v Western Australia [2008] HCA 11; (2008) 234 CLR 418, 475–7 (Gleeson CJ, Gummow, Kirby, Hayne, Crennan and Kiefel JJ) (‘Betfair’).

[229] Ibid 476–7 (Gleeson CJ, Gummow, Kirby, Hayne, Crennan and Kiefel JJ).

[230] Rowe (2010) 243 CLR 1, 135 [440]–[442] (Kiefel J); Monis (2013) 249 CLR 92, 190 [268],

193 [280], 214 [347] (Crennan, Kiefel and Bell JJ).

[231] Castlemaine Tooheys v South Australia (1990) 169 CLR 436 (‘Castlemaine Tooheys’).

[232] Kirk, above n 154, 14–15. See also Gonzalo Villalta Puig, ‘A European Saving Test for Section 92 of the Australian Constitution[2008] DeakinLawRw 4; (2008) 13(1) Deakin Law Review 99. Cf Naomi Oreb, ‘Betting across Borders — Betfair Pty Ltd v Western Australia [2009] SydLawRw 25; (2009) 31(4) Sydney Law Review 607, 616–18.

[233] Rowe (2010) 243 CLR 1, 141 [463] (Kiefel J). See also Cunliffe [1994] HCA 44; (1994) 182 CLR 272, 327 (Brennan J).

[234] McCloy [2015] HCA 34; (2015) 257 CLR 178, 195 [2] (French CJ, Kiefel, Bell and Keane JJ) (emphasis added).