AustLII Home | Databases | WorldLII | Search | Feedback

Melbourne University Law Review

Melbourne Law School
You are here:  AustLII >> Databases >> Melbourne University Law Review >> 2022 >> [2022] MelbULawRw 25

Database Search | Name Search | Recent Articles | Noteup | LawCite | Author Info | Download | Help

Psycharis, Matthew --- "Locating Plebiscites in the Australian Constitution" [2022] MelbULawRw 25; (2022) 46(1) Melbourne University Law Review 218



The plebiscite is well understood by political actors as a means of breaking parliamentary deadlocks or as a tool for government to establish a mandate. While it is recognised that a plebiscite may be lawfully procured by different means (by legislation or under the executive power), the law in Australia has so far struggled to characterise the substance of these activities. What is a vote may be a poll, a plebiscite or the gathering of statistics. Few detailed treatments of the topic exist. After identifying the different competences which support these activities, and after drawing together current understandings, this article offers a survey of the constitutional work performed by plebiscites in Australia and seeks to evaluate their place, as a species of political participation, within the normative system of representative government described by the Constitution.



The shape of government and society has changed since the time of Australia’s Federation. The law has recognised this change, with decisions of the High Court addressing the expansion of the franchise[1] (and efforts to limit the franchise),[2] the nature and effect of implied constitutional rights,[3] and the meaning of the constitutional ‘people’.[4] In so doing, successive decisions of the Court have affirmed that the Australian Constitution allows a large margin for experiments, and sometimes an ‘evolution’, in how the people of Australia engage with their government.[5] However, the Court has also held that there are constitutional constraints on this competence[6] — both in respect of the exercise of Parliament’s powers and of the executive power.[7] One product of this democratic experimentation, which finds no direct textual provision in the Constitution,[8] is the ‘plebiscite’ — understood, in Australia, as a legally non-binding poll of electors which does not cause a constitutional amendment.[9] Four plebiscites have been held, at the federal level, since Federation.[10] These have been authorised pursuant to ordinary legislation, or funded and executed under the executive power alone.[11] While legally non-binding, the result of these exercises may still be understood to impose strong political pressure on government or Parliament.[12]

At a time when the public may expect greater opportunity for participation in the decision-making that affects them,[13] and in a global era of populist politics,[14] there is a potential case for the future use of the plebiscite in Australia.[15] As a simple matter, unless the provision is repealed, legislation now provides that the Australian flag may only be changed by a plebiscite.[16] Among recent proposals for plebiscites on a range of other matters of more substantive policy,[17] it has also been suggested that a series of plebiscites might precede any later referendum on an Australian republic.[18] The question is how the constitutional function and effect of these democratic exercises is to

be understood.

The Constitution, which establishes a system of representative government and responsible government,[19] says little of exercises which go directly to ‘the people’.[20] While the states enjoyed a tradition of referendums (or plebiscites) pre-Federation,[21] a choice was made by the framers of the Constitution to restrict the use of legally binding, direct democratic exercises to proposals for constitutional change under s 128.[22] A decision was further made by the framers to reject the use of a referendum as a means of resolving deadlocks between the House of Representatives and the Senate[23] on ordinary legislation[24] — in part because it was thought that a referendum for deadlocks would ‘weaken the deliberative processes of Parliament and diminish ministerial and parliamentary responsibility’.[25] Instead, a double dissolution mechanism was provided in s 57.[26] Proposals for the constitutional provision of citizens-initiated referendums were also rejected at the time of Federation and on later occasions in the 20th century.[27]

Despite the use of legally non-binding plebiscites in subsequent practice and broad proposals for their expanded use from various political actors,[28] no complete account of the plebiscite’s function within the Constitution has been given, nor has there been, to date, any detailed judicial treatment of the topic. It has been posited by some commentators that these polls can be democratically useful in resolving moral controversies that divide parties or society,[29] or in building a case for constitutional change.[30] Politically, these observations may be apposite. However, existing legal analyses have revolved around the narrower concern of how executive government might fund a particular poll or the head of power under which any enabling legislation may be enacted.[31] While these narrower questions of legal competence are important, the anterior question of these polls’ function within a codified constitution, and their normative desirability within that architecture, remains doctrinally under-explored.[32] Four years after a ‘plebiscite’ on the matter of equal recognition of same-sex marriages transmuted to a ‘survey’ on the same question,[33] the question of characterising the substance of these exercises still begs a fuller description. This article is directed to that problem.

First, it will be observed that despite the different means by which a plebiscite may be procured (by legislation or under the executive power alone), plebiscites exhibit the common function of generating powerful political commitments, if political actors choose to opt into the process. Importantly, those commitments may have the practical effect of constraining the exercise of the legal powers of government or Parliament. In this way, while a plebiscite result is not legally binding, these robust political commitments generate outcomes which can impose a burden or benefit on the system of representative government and responsible government.

Second, a closer survey of past plebiscites is given, to better understand the contours of these outcomes. Here it is observed that, while plebiscites can be used cynically as pieces of political theatre, they have also tended to serve the constitutional function of resolving political deadlocks between the Houses (by supplying a politically decisive veto or approbation from the people), or otherwise of procuring by the executive power a political mandate for a policy in a manner which, by design or accident, imposes political barriers to effective parliamentary scrutiny or a Senate veto. This is where the machinery of government and the logic of the Constitution have tended to lead these votes.

Third, the posited danger of this function, which this article explores, is that by resolving certain institutional deadlocks or offering an efficient route around modes of parliamentary accountability, the plebiscite might, in some instances, pose a threat to established constitutional norms of responsible government and representative government. In trying to understand the consequences of this function, and the outer limits of government’s and Parliament’s capacity to engage in plebiscitary politics, it is suggested that the High Court’s decisions on measures which cause a burden (or benefit) on the system of representative government offer guidance. There, perhaps, might be situated a more complete doctrinal explanation of the plebiscite in its past and future usages.


There are at least two problems in writing on the plebiscite in Australia. The first is empirical. Very few plebiscites have been held at the federal level, with only four since Federation[34] (though many others have been proposed).[35] There have been few opportunities for the law to develop a response to what these exercises are and how they ought to function. Moreover, plebiscites have often been held at times of national emergency or under conditions of political urgency, when either by coincidence or by design the opportunity for debate about their merits, or for a detailed judicial examination, has been limited.[36] However, similar exercises have been held in the United Kingdom (‘UK’) and New Zealand;[37] and while these do not offer perfect comparisons, they do supply helpful points of contrast.

The second problem is doctrinal. In 2017, the High Court held that what might be ‘statistical information’ might equally be understood as a ‘vote’ or a ‘plebiscite’.[38] While that characterisation is likely confined to a construction of the legislation that the Court had before it (the Census and Statistics Act 1905 (Cth)), the thinness of counsel’s reasoning on that point[39] speaks to the general difficulty in characterising the substance of what is a relatively novel, and potentially flexible, political activity. There also remains an unresolved tension in the normative desirability of this activity. Few would say that regular elections are a bad thing, but many might dispute the need for regular plebiscites.[40] As Anne Twomey observes, the current understanding decomposes to the view that ‘[w]hile it is important that governments take responsibility for governing, plebiscites remain a useful tool for resolving contentious ... issues’.[41] Yet, as Twomey implicitly acknowledges, there are conflicting assumptions in that position.[42] If it is to be at all possible for those assumptions to be evaluated, then some understanding of the plebiscite, in substance, must be approached: not only to describe how they are procured, but also what work they do, and where limits on that work apply.

A How Are Plebiscites Procured?

In ordinary language, the terms ‘referendum’ and ‘plebiscite’ might be used interchangeably. In Australia, however, they have acquired different meanings. The Constitution provides for a referendum procedure under s 128, where this procedure is used to make binding changes to the text of the Constitution.[43] It is a ‘double’ referendum, such that for the result to carry, a majority of electors within the Commonwealth must vote for the change, as well as a majority of voters in a majority of the states.[44] The conduct of the referendum is regulated by the Referendum (Machinery Provisions) Act 1984 (Cth), which specifies a range of procedural matters[45] which are seen, among other things, to enhance the legitimacy of the process.[46]

Plebiscites are, by their definition, not a binding vote to amend the Constitution.[47] Unlike referendums, votes of this kind are not regulated by the provisions of the Referendum (Machinery Provisions) Act 1984 (Cth) (unless Parliament so provides). Unlike referendums, direct provision for which is found in s 128 of the Constitution, these votes require some other basis to be conducted lawfully. These indirect bases for procuring a vote are:[48]

a) by Parliament passing legislation[49] — either under one of the subject matter or purposive powers enumerated in s 51 of the Constitution,[50] under the ‘incidental power’ (with ss 51(xxxix) and 61 read together) as an incident of the executive power to undertake inquiries, or under the nationhood power.[51] In respect of the power to make inquiries, that power has previously been understood to encompass any ‘inquiry on behalf of the executive government for a purpose of government’.[52] However, regarding this, Paul Kildea observes that ‘the conduct of a nationwide poll that asks a question of 15 million voters is very different from a commission of inquiry’.[53] In respect of the nationhood power, that power vests the Commonwealth with ‘a capacity to engage in enterprises and activities peculiarly adapted to the government of a nation and which cannot otherwise be carried on’.[54] However, as Kildea observes, ‘the task of resolving public disagreement on contentious social issues ... might be said to rest “peculiarly” with national Parliaments or other democratic processes’,[55] not the executive government, and so the plebiscite finds only an imperfect home under this power; and

b) by the executive power alone, as a commercial ballot,[56] where, absent any fresh legislation, the Commonwealth contracts with the Australian Electoral Commission (‘AEC’) or some other agency to hold a poll.[57] This might be done pursuant to s 7A(1) of the Commonwealth Electoral Act 1918 (Cth), which provides that the AEC may ‘make arrangements

for the supply of goods or services to any person or body’. While the

1977 song poll was procured by something like this route,[58] as a consequence of the High Court’s later decisions in Brown v West,[59]

Pape v Federal Commissioner of Taxation (‘Pape’)[60] and both

Williams v Commonwealth[61] and Williams v Commonwealth [No 2][62] (‘Williams cases’), expenditure absent some authorising legislation must now fall into a set of exceptional categories,[63] and it has been the opinion of a series of practitioners that a national plebiscite would not easily fit into one of those categories.[64] In the case of the 2017 marriage equality survey, it appears that initial government plans to hold a fee-for-service ballot through the AEC were rejected for these reasons.[65]

After the events of 2017, it might also be said that a plebiscite, or something that achieves the same political effect as one, could be authorised through using the Australian Bureau of Statistics’ (‘ABS’) existing power to gather statistics under the Census and Statistics Act 1905 (Cth)[66] (or indeed some other agency with existing information-gathering powers).

B Plebiscites as a Democratic Device

If there is a preliminary observation about the plebiscite, it is that its legal basis is not fixed, and is indeed far from certain, depending on the case. If there is to be any value in talking of ‘the plebiscite’ as a distinct category of democratic exercise, then it cannot be found by looking at how the vote is authorised. Here, its legal forms may be various.

However, aside from its legal description, there is another characterisation of the plebiscite that may be posited through the lens of political constitutionalism.[67] It is well rehearsed that some aspects of Australia’s system of government were only elliptically identified in the text of the Constitution, or were otherwise left ‘unwritten’ as an assumed background of norms and political practices that may change over time.[68] Particularly in respect of those powers or functions under the Constitution which are not legally reviewable, the political process offers a useful further means of resolving disputes or ensuring government accountability.[69] In serving this function, Geoffrey Sawer observed that, under the Constitution, political activity may, from time to time, ‘ha[ve] the effect of reducing the ambit of a legal power or transferring its effective exercise to a person or institution other than the one designated by the legal power’.[70] It is suggested here that the plebiscite must be understood in similar terms — not only in terms of how it is funded or procured, but also as a political feature of the Constitution that might, in some cases, interact with or constrain legal powers (like the ability of Parliament to pass legislation on a certain topic). The means by which this is done is to be developed below.

In describing how plebiscites have typically been understood, it has been suggested that they are viewed, in law, as ‘no more than a formalised, national opinion poll’.[71] On this basis, the plebiscite might be used flexibly, and even cynically, by political actors as a piece of ‘political theatre’, designed to add weight to a policy proposal or to develop a consensus, but it might not carry any greater constitutional significance than that. For example, in 2011, the Leader of the Opposition, Tony Abbott, introduced the Carbon Tax Plebiscite Bill 2011 (Cth) which, given the composition of Parliament, stood little chance of being enacted, but which nonetheless offered an opportunity for colourful debate, at second reading, on the wider topic of the government’s climate policy.[72] This theatre is, of course, only carried on when participants in the political process choose to participate in it and to be bound by the result. In the case of the proposed carbon tax plebiscite, government MPs disavowed the need for a carbon tax vote[73] and civil society expressed little interest in a plebiscite on the topic,[74] and so the proposal melted away at second reading.[75]

Yet an account of the substance of the plebiscite which goes no further than characterising it as a cynical buttress to parliamentary argument or, at most, as a large-scale opinion poll would appear incomplete against the backdrop of the political constitution. The history and circumstances of Australia’s national plebiscites will be considered in the next part of this article. However, it is sufficient to say here that, on each occasion that a national plebiscite has in fact been held, government has never adopted a position contrary to the people’s veto (or failed to give effect to a positive plebiscite result); neither has Parliament passed legislation directly in conflict with the opinion produced by the poll.[76] This, of course, is not the case in respect of other measures of public opinion, which might often be discounted by government.[77] And while isolated political actors might still choose not to follow a plebiscite result,[78] consistent with the Australian example there is strong precedent elsewhere that representative bodies, when voting as a whole under majority rules, will not legislate in conflict with a plebiscite outcome.[79] Empirically at least, it appears that once actors agree to participate in the process, they appear ‘bound’, politically, by the result.

1 The Status of a Plebiscite Result

To ‘bind politically’ may seem a contradiction. It is true that, otherwise than as constrained by law, political actors are notionally free to adopt whichever position they choose. However, in practice, political conventions, the party whipping system, popular opinion, the popular mandate, party manifestos, and the relative fragility of party leadership offer tangible structural constraints.[80] The unsurprising result is that there may be strong normative reasons for an actor to choose not to adopt a position. In this analysis, the fact of a plebiscite result is another basis which supplies a strong normative constraint on

political activity.

In places other than Australia, the normative effect of a plebiscite result has been recognised as an important feature of constitutional arrangements, particularly where greater emphasis is placed on a political account of the constitution. In the UK, it is now understood that certain legislation concerning devolution, the method of voting, the currency, or the continuation of the Crown might lack legitimacy if not first submitted to an ‘advisory referendum’ (the equivalent of a plebiscite). It has thus been recognised by some that there has arisen a constitutional convention to hold popular votes on certain matters[81] (or otherwise the House of Lords might reject the Bill). In the language of Geoffrey Sawer, the ambit of Parliament’s legal power in these cases might, in practice, be circumscribed by political habits or a sense of political obligation that attaches to the advisory referendum.[82]

In New Zealand, in a recent dispute concerning an Act establishing a scheme for assisted suicide (where that scheme would only enter into force after an advisory referendum), counsel for the New Zealand Attorney-General submitted that, in the circumstances, ‘Parliament ha[d] divested or assigned the final step in the legislative process back to the electorate’.[83] The Court did not displace this appraisal.[84] The Attorney-General further submitted that, in seeking the opinion of the electors before the law entered into force, the ‘final step in the lawmaking process ha[d] been given back to the electorate directly with no discretion [reserved by Parliament]’.[85] While, strictly, the New Zealand Parliament remained free to later enact that the scheme might still enter into force despite the result,[86] it was the position of the government that an essential step in the legislative process would have been missed had the people’s voice been ignored in that way. This account approaches the result that there might, in some instances, be an entrenched legal requirement to put matters to a popular vote.[87]

In Australia, understandings of the effect of a plebiscite have not risen so high. It would be very hard to argue that there is a plebiscite ‘convention’, or that there is an entrenched legal requirement to hold an advisory vote in certain instances. Yet the capacity of a plebiscite result to generate normatively significant political commitments has still been recognised. Such was contemplated by the Senate Legal and Constitutional Affairs References Committee in 2015 in its report on the proposed marriage equality plebiscite.[88] In evidence given before that Committee, Kristen Walker, George Williams, Geoffrey Lindell, and the Gilbert + Tobin Centre for Public Law separately observed that even a legally non-binding reference to the electors (however procured) was still politically authoritative, and that this practice of committing government and its majority in Parliament to a position might upset the balance of what the High Court has understood to be the constitutionally prescribed system of representative government — if used without care.[89] Others have made similar observations.[90]

There appear to be two reasons for this concern, which, again, may be expressed in the language of Geoffrey Sawer. First, because the plebiscite may, in practice, constrain the policymaking latitude or legislative competence otherwise afforded to government and Parliament under law.[91] And second, because this practice of politically committing Parliament to a position might, in practice, upset other established mechanisms by which political disagreements are resolved, transferring the practical exercise of a power of review or veto out of the hands of one organ and into another politically authoritative body.[92]

2 A Constraint on Government and Parliament’s Legal Powers

In respect of the first matter, it has long been recognised by Australian MPs that it would be ill-advised to vote against a policy approved by a plebiscite.[93] Quite simply, the public censure may be severe. In the UK, the practical effect of political censure of this kind, on Parliament’s powers, has been considered by the courts. In a recent dispute, the Supreme Court of the United Kingdom examined an enactment passed unanimously by the devolved Scottish Parliament, which purported to establish a scheme allowing Scottish actors to declare certain policies of the Westminster government to be inconsistent with international human rights.[94] Relevantly, the scheme would not legally bind Westminster: its main mechanism was to create an institutionalised avenue for political censure of the British government.[95] Lord Reed, in giving the unanimous judgment of the Court, held that such political censure, despite being legally non-binding, would nonetheless ‘plainly affect Parliament’s power to make laws for Scotland, since it would impose pressure on Parliament to avoid the opprobrium [from the devolved nation]’.[96] The Court found that this purely political censure, by constraining Parliament’s options, would thereby offend parliamentary sovereignty, and the scheme was on that (and other) bases unlawful.[97] This scenario is, of course, different to a plebiscite in that it involves an institutionalised censure of alleged breaches of international law; but it nonetheless offers an example of a judicial recognition of the type of dynamic that Geoffrey Sawer identified, whereby purely political modes of accountability can rise so high as to impermissibly constrain otherwise sovereign legal powers.[98]

In a related example, the Scotland Act 1998 (UK) (‘Scotland Act’) provides that the Scottish government may only be abolished by referendum.[99] It remains that the Parliament of the United Kingdom could simply repeal the Scotland Act, and thereby repeal the requirement for popular consultation contained in that Act. However, the dictum of Baroness Hale in R

(Jackson) v Attorney-General (UK) might suggest that such a two-stepped disenfranchisement of the Scottish people may be unlawful.[100] David Feldman has posited that this is, in part, because such referendum provisions (though not legally entrenched) are nonetheless expositive of important political agreements or constitutional arrangements.[101] Again, there exist strong normative reasons for governments not to depart from those agreements — in these examples, to preserve the comity of the devolution settlement in an uncodified context.

Australia is not the UK, and different constitutional work is done by politics and convention. Yet the UK examples are not without their Australian parallels. Section 3 of the Flags Act 1953 (Cth) (‘Flags Act’) was amended in 1998 to require any change in the national flag to be preceded by a plebiscite.[102] Similar to the Scotland Act, it remains open to the federal Parliament to lawfully repeal that requirement — though, in similar fashion, it is likely to face political censure for doing so. In this way, plebiscites might partially entrench a policy position by imposing a strong cost for their repeal. On this basis, Simon Evans has suggested that the Flags Act might be a candidate for a ‘constitutional statute’ (if such a doctrine applied in Australia) and therefore would be immune from implied repeal.[103] Similarly, after the events of 2017, it is hard to imagine that Australia’s provision for marriage equality could be easily changed, politically, without some further form of popular consultation. This process of fixing political commitments is not necessarily an unhealthy result: in accounts of the Constitution which emphasise political (versus legal) controls on government, such a function might be seen to bolster public accountability.

Early proponents of the advisory referendum (or plebiscite, as it would now be called) saw this function too. AV Dicey understood that a legally non-binding poll could still act as a politically decisive ‘people’s veto’ on a government with a disciplined majority in Parliament, and could serve to commit that government to a position.[104] Dicey understood the constitutional import of this practice, and opined that ‘[s]ome theorist will object that any Act introducing [a plebiscite] will have little validity, since Parliament might by a subsequent statute undo its own handiwork’; however, he observed that

[t]his objection, whatever be its speculative force, is ... of no practical moment. ... Let a popular veto be established, and the popular veto will
command respect.[105]

3 A Political Circuit-Breaker for Constitutional Rigidities

At the same time as potentially constraining Parliament, the plebiscite, differently used, might also be deployed to soften legal rigidities in the Constitution — including the otherwise decisive veto of a hostile Senate. Dicey saw that legally non-binding referendums can offer a useful political release valve where a deadlock exists between parliamentary houses or within a party, or where the views of a majority of the electorate do not reflect the views of a majority of their representatives.[106] In the British constitutional crisis of 1911,[107] as one means of broaching the impasse between the House of Commons and the House of Lords, it was suggested that the result of a legally non-binding referendum initiated by the Commons, on a contested policy, would be hard for a hostile House of Lords to ignore (even if the Lords reserved a legal power of veto over that policy).[108] Lilian Tomn, writing at the turn of the 20th century in the Australian context, saw a similar potential use.[109] She further observed that Parliament could even provide for a plebiscite to be made legally binding, ‘not [in] an attempt to constitute the people sovereign, but to substitute their assent for that of the Upper House should the Upper House continue to reject a Bill passed by the Lower House’.[110] Such a potential use, in the case of deadlocks, was recognised by some of the framers of Australia’s Constitution too.[111] As with the British crisis of 1911, there is, here, the political capacity of a plebiscite to soften legal impediments to the passage of government policy.

4 The Substance of Plebiscites

The sum of these observations is that the substance of a plebiscite, in addition to its role as bare theatrics, is to create normatively robust political commitments. These commitments might partially entrench or veto a policy (in a way that imposes stronger than usual normative barriers to Parliament changing that policy), or resolve disagreements between different organs of government in a politically decisive way. This may serve a useful constitutional function, particularly under accounts of the Constitution that emphasise political, over legal, accountability.[112] However, as above, these results remain contingent on political actors opting into the process.

There are different accounts of why actors regularly choose to opt in, and why plebiscites, in such scenarios, are nearly always politically decisive (if not made to be legally binding).[113] One is that they appeal to ‘the people’, whose existence and consent to be governed are considered to form the basis of the legitimacy of the constitutional state.[114] Whether the reason for the normative force of the vote can be so easily reduced is a complex sociological question beyond the ambit of this article. However, it is sufficient to say here that contemporary judicial understandings of ‘the people’, in Australia, see ‘the people’ as the foundation of political and constitutional authority.[115] There is, then, some evidence that (politically at least, if not legally) the people’s voice sits prior to the authority of Parliament.[116] From such is derived not only the plebiscite’s power to resolve impasses created by the parliamentary system but also its capacity to diminish the status of Parliament as a deliberative chamber. What marks the plebiscite as unique is that, unlike other modes of gauging public opinion (eg opinion polling by private organisations), this process usually seeks to recapitulate the logic and form of the electoral process.

If there is credence to the view that plebiscites may act as a political check on Parliament — or as a political ‘release valve’ to get around either the supervisory role of the Senate or recalcitrant backbenchers — then there is a question as to whether this form of democratic participation (alongside other forms of participation, like political communication or voting in an election) would be, either in a given case or if used more expansively, compatible with the maintenance of the system of representative and responsible government otherwise prescribed by the Constitution. In the British context, Vernon Bogdanor has commented that ‘arguments against the referendum [or plebiscite] are also arguments against democracy’.[117] This may be true in Australia, but the Constitution does not provide for democracy simpliciter, and the question is whether, and to what degree, the practice of plebiscitary politics, as a means of democratic accountability and participation, sits within the principles of the particular form of democracy provided in the Constitution (as interpreted in the context of contemporary values).[118] This question may become particularly acute where a plebiscite is procured by the executive power alone, in circumstances where the government wields ‘the sword of the community’[119] as a political check on other methods of participation and accountability, including parliamentary deliberation.

C Plebiscites as a Form of Participation under the Constitution

Ordinarily, political practices, even those which might rise to be characterised as ‘conventions’, are not justiciable.[120] However, some aspects of political life are still regulated by law. Questions of this type — concerning the boundaries placed on measures which affect modes of purely political participation — are not alien to the law of Australian constitutional review.[121] However, the language of that review is yet to be applied to the plebiscite. The essential problem is to understand how the law can (or should) respond to a purely political activity (like political speech, participation in a plebiscite, or casting a vote) which is nonetheless capable of causing constitutional outcomes.

As observed above, not all aspects of civic life were codified at Federation — yet implied rights of political communication are a necessary incident of the textual and structural provision for free elections.[122] Limitations on the franchise (where the franchise is only elliptically defined in the text of the Constitution)[123] are also constrained by what are understood to be the requirements of representative democracy.[124] These interpretations of the Constitution derive their legitimacy, in part, from their alignment with Australia’s ‘unwritten’ constitutional norms.[125] While the law of implied rights, and the law of the franchise cases, have developed in partial separation, there is a common logic. As Joo-Cheong Tham observes, each is a species of a genus of disputes concerning efforts to reshape assumptions regarding how Australians participate in politics.[126] Or, more narrowly stated, they are about the free ‘choice’ of the people over their government, where a need for that choice is constitutionally prescribed.[127] These disputes are qualitatively different from other subjects of constitutional review, as they concern attempts to evaluate a posited threat to the system of political participation described by the Constitution, and may, for this reason, require a higher standard.[128] Cases of this type are answered by a form of proportionality analysis (or, earlier, a proportionality-like analysis,[129] the contours of which are subject to debate).[130] That standard requires not only an identification of bare legal competence, but also an analysis that looks to means and effect.[131] There is, in this way, a common language between the cases.[132] As a matter of constitutional interpretation, if a threat to the system of representative government is supposed, then the identification of a valid law must be with reference to how that law would function within the context of a constitutionally prescribed system of representative and responsible government.[133]

It is at least possible that this language, which brings in the political or democratic effect of a law, might be apposite to the plebiscite, which, given its democratic effects, might too be considered a sister species in the genus of exercises in political participation. Simply stated, exercises in political communication, the expression of the ‘people’s veto’ in a plebiscite, or the casting of a vote in an election, are all political activities which cut to the general object of ensuring political accountability under the Constitution by informing the people’s choice over their government. As with the other cases, in evaluating the plebiscite’s role within the normative system of representative government, it may not be sufficient to only identify some bare legal competence which might procure the vote. If the grammar of the communication cases could be applied to the franchise cases, then it may apply to other political activities too — at the very least, as a source of normative guidance if not as a basis of review.

However, these observations offer only vague boundaries to the problem of how plebiscitary politics, in all of its different possible manifestations, fits within the system of government prescribed by the Constitution. And it still remains that courts should not prescribe to government or Parliament what is in the public interest in using plebiscites[134] while the corresponding legislative competence in this respect is broad.[135] Analyses of this type necessarily involve a degree of factual and normative judgment which cannot be produced in the abstract.[136] Here, as Condillac would observe, it becomes an effort to build ‘the unknowable on the unknown’.[137] To seek out those unknowns, what, then, of the plebiscite in practice, and how might its potential burden on, or benefit to, the system be understood?

The importance of evaluating how political practice interacts with the legal constitution was understood by Walter Bagehot. In his description of the Westminster constitution as comprising two halves — the ‘dignified’ half, being that which is visible, ceremonial and provided for ‘on paper’, and the ‘efficient’ half, being how, politically, the machinery is put to use — he appreciated that identifying the politically ‘efficient’ use of a given constitutional mechanism was important.[138] Of the ‘efficient’ use, he observed that ‘[t]here are ... practical [people] who ... say, we want only to attain results’ and who ‘admit that ... [some] part[s] of a constitution [do] no business, or that a simpler machine would do equally well what it does’.[139] It is in this effort to find the ‘efficient’ working of the Constitution, particularly in respect of activity that is essentially political, that some answer may be given to any new measure’s effect on the settlement. This necessarily requires a closer survey of the plebiscite’s

historical use.


A Conscription Plebiscites: 1916 and 1917

Numerous historical and social accounts of the 1916 and 1917 conscription plebiscites have been given.[140] However, no detailed legal account exists,[141] and an effort is made here to highlight how the device of the plebiscite fits into the machinery of government.

1 The Need for a Vote

In the early 20th century, compulsory military service was strongly opposed within civil society[142] and within the governing Labor Party.[143] It was understood that the Commonwealth had the power to compel persons to serve in the armed forces,[144] but only within the territorial boundaries of the Commonwealth.[145] On the commencement of war in 1914, a volunteer force was sent overseas.[146] In that same year, the government succeeded in passing the War Precautions Act 1914 (Cth). That Act gave the Governor-General (on advice) a broad power to make regulations and orders in council in respect of the conflict.[147] That regulation-making power was limited in two ways. First, it was required to be used ‘for the defence of the Commonwealth’[148] — which, on one view, could impose a territorial limitation. Second, if any doubt lingered, the Defence Act 1903 (Cth) provided that no person compulsorily drafted shall be compelled to serve ‘beyond the limits of the Commonwealth’.[149] The result was that the Minister for Defence did not consider that he had the power to make regulations for the deployment of persons overseas where they were compulsorily recruited.[150]

The heavy loss of ANZAC life at the Battle of the Somme in 1916 caused a need for replacement personnel.[151] In New Zealand, in the month following the Somme, Parliament passed conscription legislation.[152] What precipitated the need for a plebiscite in Australia, instead of a simple parliamentary vote on the matter, was a state of intra-party and inter-house deadlock.[153] The Labor Prime Minister, William Hughes, who remained a staunch advocate of overseas conscription,[154] could not command a majority of his Cabinet on the policy[155] (and was threatened with the resignation of at least one of his Ministers should a conscription Bill be introduced).[156] Further, any such overseas service Bill would also likely face a veto by Labor votes in the Senate.[157] As the policy required no change to the Constitution, a constitutional referendum under s 128 would also be inappropriate.[158] It would also likely not be possible to hold the vote by means of regulations made under the broad powers in the War Precautions Act 1914 (Cth), because the Senate (which was hostile to conscription) had the power to disallow regulations, and there was a risk that power might be exercised.[159]

To break the intra-party and inter-house deadlock, senior members of government determined to hold an advisory plebiscite (at the time called a ‘referendum’),[160] though Inglis records that the proposal, while a compromise, ‘evoked no enthusiasm’ within the Prime Minister’s ranks.[161] It appears that even those opposition MPs who were advocates of conscription were nonetheless sceptical of resolving the matter by plebiscite, as it was ‘not ... a proper [issue] to submit to the people’.[162] The eventual introduction of the Military Service Referendum Bill 1916 (Cth) caused one government Minister to resign.[163] In that Bill’s passage through Parliament, and after ‘lengthy debate’,[164] it was opposed by 10 Labor Members in the House of Representatives and 9 Labor Senators.[165]

In the form as passed, the Military Service Referendum Act 1916 (Cth) provided that the Governor-General issue a writ to the Chief Electoral Officer to take a ballot on the question.[166] The question was wordy by contemporary standards and, among other things, asked whether, in a state of emergency, the elector was in favour of compulsory overseas service.[167] While legally non-binding, the vote was understood to ‘bind’ government and

Parliament politically.[168]

2 Features of the Vote

Several matters distinguished the conduct of this plebiscite from a referendum which might have been held under s 128 of the Constitution. The Act disqualified from voting those electors who were born in territories with which ‘Great Britain’ was now at war.[169] The Act also empowered electoral officers to question persons at polling booths regarding their place of birth, and prescribed tribunals to determine whether particular persons were of a disloyal national allegiance.[170] These additional qualifications would not normally apply to a constitutional referendum held under s 128 or regulated by the Referendum (Constitution Alteration) Act 1906 (Cth),[171] where the usual declaration required of the voter at the polling place, at the time, was that they simply be registered on the electoral roll of the relevant division (though restrictions on the franchise did apply).[172] Additionally, four weeks before the vote, Hughes exercised powers under the Defence Act 1903 (Cth) to call up eligible single men for domestic service, and further issued regulations under the War Precautions Act 1914 (Cth) such that returning officers on polling day were to ask single men whether they had responded to the call-up — and, if they had not, to set aside their vote.[173] In a number of ways, there was, here, a sui generis capacity to redefine the contours of the constitutional ‘people’ in the context of plebiscitary politics.[174]

Despite this modification of the franchise, the Military Service Referendum Act 1916 (Cth) sought to import to the plebiscite those other standards which ordinarily applied to parliamentary elections,[175] including those in relation to the appointment of electoral officers, the selection of polling places and methods of scrutineering.[176] In the degree to which the logic of parliamentary elections[177] was applied to the conduct of a legally non-binding vote, there appears something qualitatively different about this plebiscitary exercise to an opinion poll; normatively, at least, it appears nearer to a process of candidacy, where the electors confer a mandate on elected members. In the result, the plebiscite was defeated, with 52% of votes tallied being against conscription.[178] The defeat of the plebiscite meant that the Prime Minister lost the confidence of the party room; however, he succeeded in forming a new ‘Nationalist’ government,[179] but only on the mandate that he would not press

for conscription.[180]

3 The Need for an Additional Vote

Further loss of life in 1917 caused a further need for replacements,[181] and a fresh mandate was sought by the Prime Minister through a second plebiscite.[182] This second plebiscite was conducted by a different power to the first: not through passing primary legislation providing for a vote (as had been done the year before), but through the executive’s capacity to make regulations under the existing wartime legislation.[183]

It appears that this power was used to avoid further parliamentary scrutiny, which had been harsh the year before.[184] In September 1917, the Prime Minister had successfully moved that Parliament be adjourned for several months (until January 1918) saying that

[t]he Government does not propose to do anything during the adjournment which it would not do if the House were sitting. If [it is feared] that we intend to bring in conscription by regulation, I give ... the positive assurance that we do not.[185]

Speaking of the reason for that adjournment (some time later), the

Labor MP, JH Catts, said:

[We] adjourn[ed] in September, because we were given to understand Ministers desired a spell. Some of our friends said that the Government were tired of criticism. ... As a matter of fact, that adjournment took place in order that the Government might secretly, and behind the backs of the people, arrange another conscription campaign ...[186]

Catts may not have been wrong, and the Prime Minister may have been liberal with the truth, because in October 1917, while Parliament was adjourned, members of government began holding pro-conscription rallies.[187] On

7 November 1917, Cabinet met and resolved to hold a second conscription plebiscite.[188] Finally, on 10 November 1917, the Minister for Defence issued the War Precautions (Military Service Referendum) Regulations 1917 (Cth),[189] pursuant to his broad regulation-making powers under the War Precautions Act 1914 (Cth).[190] The Regulations provided for a plebiscite to be held on the question of overseas conscription.[191] On the same day, pursuant to the Regulations, the Governor-General issued a writ to the Chief Electoral Officer, directing the Officer to conduct a poll of the electors on 20 December 1917

(ie in about six weeks’ time) on the question of conscription.[192] That poll was held on the appointed date in December,[193] with Parliament only resuming its session in January 1918, after the plebiscite had been concluded[194] and with little opportunity to debate the plebiscite’s merits, other than the result.

The result of that second vote was again against the government, with 54% of those polled being against conscription.[195] The Nationalist government (despite its pro-conscription position) considered itself politically committed to honour the consecutive ‘no’ votes,[196] and legislation for overseas conscription was not introduced.[197] Indeed, such was the political import of the votes that the Prime Minister felt it fitting to promise to resign if the result did not carry in his favour (which, in the event, he did not).[198] This might have been a cynical stunt to tip the votes in favour of ‘yes’ (particularly if some of Hughes’ supporters were conflicted on the policy of conscription), yet Hughes’ precise motivation is irrelevant to the analysis. The critical point is that, for whatever reasons, the relevant political actors did not see it as fitting to depart from a plebiscite result.

B The National Song Poll: 1977

The next national plebiscite in Australia’s history requires less explanation. That poll occurred 60 years later and was in response to the unsettled question of what Australia’s national song ought to be (if not ‘God Save the Queen’).

1 The Need for a Vote

In 1974, the ABS had been directed to sample 60,000 people on their views on alternative national songs (which were ‘Advance Australia Fair’, ‘Waltzing Matilda’ and the ‘Song of Australia’).[199] 51.4% of respondents favoured ‘Advance Australia Fair’, and the Labor Prime Minister, Gough Whitlam, directed that this song would be the new national song.[200]

When the Labor government ended in 1975, the new Coalition government reinstated ‘God Save the Queen’.[201] However, the issue remained live. There was a concern that, at the upcoming 1976 Montreal Olympic Games, Australia could suffer embarrassment if the organisers were unable to play a distinct national song should its athletes place on the podium.[202] While Australia did not win gold (and thus the embarrassment never materialised), the concern to identify a modern national song remained.[203]

The Prime Minister, Malcolm Fraser, said he hoped the matter would be resolved ‘through a process of natural selection and choice’.[204] However, it was the stated policy of the government that its preferred alternative song was ‘Waltzing Matilda’ (as well as being the Prime Minister’s personal favourite).[205] Cabinet archives suggest that, in an effort to gain a mandate for ‘Waltzing Matilda’, the federal Cabinet had polled the six state premiers on their preferred song — where, unfortunately for the Prime Minister, a majority of the state premiers who gave a response indicated that ‘Advance Australia Fair’, and not ‘Waltzing Matilda’, was their preferred national song.[206] The Prime Minister had also requested the government whip to ‘secure reactions from Party Members concerning their views on Waltzing Matilda being ... the national song’.[207] However, no resolution was reached.

After the Montreal Olympics, at a meeting on 27 January 1977, Cabinet ultimately resolved to hold a national plebiscite on the question.[208] This was to be a plebiscite of all registered electors (unlike the limited survey of 60,000 people conducted by the ABS three years before).[209] As an expedient, this plebiscite was to be held alongside four constitutional referendums also to be held in 1977, with the votes to take place on the same polling day and using the same polling facilities.[210] The result returned was that the electors favoured ‘Advance Australia Fair’ more than any other proposed song, and this was later directed to be used as the national song.[211]

2 Features of the Vote

Like the earlier 1917 plebiscite, the direction to hold a vote was done through the executive power alone,[212] not through fresh primary legislation providing for a vote. Here, the relevant Minister directed the Chief Electoral Officer to prepare, distribute and count the ballots pursuant to something akin to a memorandum of understanding between government agencies.[213] Cabinet documents now available from the National Archives help explain the reason for this extra-parliamentary process. In a 1977 advice from the Attorney-General’s Department to the Prime Minister, it was said that

[w]e [the Attorney-General’s Department] have discussed the matter with the Australian Electoral Office ... who agree that it would be preferable not to introduce legislation for this purpose. The subject does not lend itself to a ‘yes/no’ argument and while an Act may serve as an extra argument against anyone still dissatisfied with the outcome, the result in any case would not be legally binding. Also, debate on a Bill in the Parliament could be protracted.[214]

Despite the extra-parliamentary method of procuring and funding the vote, there was still an effort to apply the usual standards that would pertain in a constitutional referendum or parliamentary election. A short Act was later passed[215] to the effect that the same polling stations to be used for the constitutional referendums to be held on the same day might also be used for the plebiscite.[216] However, in respect of executive government’s capacity to reshape the plebiscitary franchise, it is worth noting that while the vote had been extended to residents of the territories for the purposes of the song poll, they remained unable to vote in the referendums held on the same day.[217]

It need only be stated in summary that this plebiscite operated to resolve a disagreement within federal politics, and between the state governments and federal government, on the choice of the song.[218] Relevantly, unlike a constitutional referendum,[219] the result in the plebiscite carried on the basis of a simple national majority. In the result, a majority of states and territories did vote for ‘Advance Australia Fair’; however, had that not been the case, there may have arisen a question of whether federal interests had been sacrificed, depending on the Prime Minister’s choice.[220]

C Marriage Equality: 2017

The political and social factors which led to calls for a plebiscite on marriage equality, and the eventual survey, are complex and have been recounted elsewhere in some detail.[221] However, the legal context may be summarised

as follows.

1 The Need for a Vote

In 2013, the High Court had held that the federal Parliament was competent to legislate for marriage equality.[222] Despite this statement as to Parliament’s competence, the incumbent Coalition government did not choose to hold a free parliamentary vote to change the law (despite calls for such a vote).[223] Instead, in 2015, it was reported that senior members of government, including the Prime Minister, were considering either a referendum in order to expressly restrict the recognised constitutional power to legislate on same-sex marriage, or otherwise to hold a non-binding plebiscite on the issue.[224] This can be explained, in part, because despite a plurality of views within the party, some senior members of government were firmly opposed to the reform.[225] Speaking of the need for a plebiscite or referendum, the Prime Minister said that ‘the only way to successfully and satisfactorily settle this matter ... is to ask the people to make a choice ... we accept in our country that the people’s vote settles things’.[226]

While, at the time, the government could command a majority in the House of Representatives, it did not have control over the Senate.[227] And, in August 2015, the Senate referred the matter of the government’s proposed referendum or plebiscite to the Senate Legal and Constitutional Affairs References Committee, for the Committee’s consideration.[228] The view of the majority of the Committee was that the proposed referendum or plebiscite was constitutionally unnecessary (given Parliament’s competence on the subject matter) and potentially corrosive of representative and responsible government.[229] Evidence directed to the latter point was given by George Williams,[230] Geoffrey Lindell,[231] Kristin Walker,[232] and the Gilbert + Tobin Centre for Public Law,[233] which the majority of the Committee appeared to accept.[234] Some time after the Committee hearings, Anne Twomey observed that plebiscites might be democratically useful, but that the proposed plebiscite in this instance, and in these circumstances, had led some to ‘demonise’ the mechanism.[235] However, if these concerns as to the health of representative democracy were the contours of the evidence before the Committee in 2015, those more fundamental concerns did not resurface when the issue was ultimately litigated in 2017.

Despite the findings of the majority of the Committee, it remained the case that no free parliamentary vote could be held, in part because such a vote threatened a schism within the governing Liberal Party that risked the Prime Minister’s leadership.[236] After legislation providing for a plebiscite was rejected by the Senate in 2016,[237] senior members of government in early 2017 sought to consider alternative means of holding a plebiscite under the executive power, without the need for parliamentary support. The Attorney-General’s Department was asked to cost certain proposals.[238] In March 2017, the Minister for Immigration stated that

[o]bviously the Senate has knocked back the legislation and there is no prospect of the plebiscite proper passing through the Senate and there is an option ... in relation to the postal plebiscite ... others can comment on the legality or otherwise of it ...[239]

The 2017–18 Budget Paper, as tabled in May 2017, still contained a line item for the conduct of a plebiscite on the question.[240]

Around July or August 2017, proposals for a ‘postal plebiscite’, to be conducted by the AEC, were considered by Cabinet.[241] However, there was a view, publicly held, that the effect of the Pape and Williams cases was such that no expenditure on this proposal by the AEC could be lawful, absent some fresh parliamentary approval.[242] In the result, in August 2017, Cabinet resolved to hold a plebiscite,[243] conducted not by the AEC as initially planned, but by the ABS, which would rely on an existing statutory power under the Census and Statistics Act 1905 (Cth) to gather ‘statistical information’ from all electors.[244] That power to gather statistics would provide the required parliamentary approval for the expenditure.[245] The question to be asked by the ABS under that statutory power was the same as under the originally proposed AEC plebiscite: namely, ‘[s]hould the law be changed to allow same-sex couples to marry?’.[246] Participation was not compulsory, but approximately 12,700,000 forms were returned (ie about 80% of all registered electors returned their form).[247] When the results were released by the ABS on 26 November 2017, the response was 61.6% in favour of marriage equality and 38.4% opposed.[248] Following the result, government leadership permitted a free vote in Parliament,[249] which resulted in an Act legalising same-sex marriage being passed.[250]

In order to fund this activity (absent any new Act making available the public funds), the Finance Minister had made a determination under s 10(2) of the Appropriation Act (No 1) 2017–2018 (Cth) (‘Appropriation Act’).[251] This emergency provision, a Henry VIII power, enabled the Minister to amend the Appropriation Act to make added provision for any expenditures up to a limit of $295 million that were ‘urgent’ and which were ‘unforeseen’[252] at the time the Budget was handed down earlier that year in May 2017. In the event, the Minister’s exercise of the amendment power allocated approximately $122 million of public funds to the ABS to conduct the poll.[253] Importantly, s 10(4) of the Appropriation Act provided that the Minister’s determination under the Henry VIII power was not disallowable by the Senate (as might be the case for other legislative instruments),[254] and so the Senate’s capacity to scrutinise the expenditure was here diminished.

2 A Challenge to the Vote’s Legality

In responding to two challenges to the actions of the Finance Minister and the Australian Statistician,[255] the High Court upheld the legality of the exercise.[256] If it was the case that evidence before the Senate Legal and Constitutional Affairs References Committee in 2015 stressed the importance of an evaluation of the impact that these exercises would have on the maintenance of a constitutionally prescribed system of representative government and responsible government, then arguments of that type were almost entirely absent from the plaintiffs’ submissions.[257] The approach taken was to impugn the validity of the appropriation and expenditure.[258] Some effort was made by the plaintiffs to distinguish between ‘statistics’ on the one hand, and a ‘vote’ on the other, where the Census and Statistics Act 1905 (Cth) was argued only to empower the ABS to gather data of the former category.[259] However, the Court was almost entirely unpersuaded on this point,[260] holding that (for the purpose of the Census and Statistics Act 1905 (Cth)) there was no relevant legal distinction between the two, and what might be a ‘vote’ might also be

useful ‘statistics’.[261]

The fact that the Minister’s powers, in this case, had been used to effect a reallocation, or mere ‘earmarking’,[262] of funds already appropriated by Parliament for the purpose of emergency use (rather than an expenditure of appropriated moneys) also largely placed the conduct outside the confines set by the Pape and Williams cases.[263] Here, provided the appropriation was effective, the existing power to gather ‘statistics’ under the Census and Statistics Act 1905 (Cth) meant that the subsequent expenditure by the ABS found its statutory authority.[264]

Despite this reasoning, the result has been criticised on the basis that it left few avenues open for accountability in the context of what was a large expenditure on a topic that had vexed parliamentary politics for a number of years.[265] Indeed, the added feature of this vote, compared to the three plebiscites that preceded it, was that Parliament had spoken twice (both in the Senate and through the committee system) and, on both occasions, had rejected the proposal for a popular vote on this topic.[266] While it remains an open question of nomenclature whether this was truly a ‘plebiscite’, the activity at least bears the political hallmarks of the earlier exercises[267] and, at least in the context of the Census and Statistics Act 1905 (Cth), the Court held that a statistical exercise is not mutually exclusive with a ‘plebiscite’.[268]


If Dicey,[269] Tomn[270] and others[271] had hypothesised that the plebiscite’s efficient function in a Westminster context would be to offer a decisive political check on parliamentary power — in particular, as a means of overcoming deadlock within Parliament or within parties — then that hypothesis appears borne out in practice. While the plebiscite has assumed different legal forms and has taken place in different political contexts, there are functional similarities between these exercises.

In 1917, the vote was procured during a parliamentary adjournment, using delegated legislative powers, and the result was presented as a fait accompli upon the resumption of Parliament in 1918. In 1977, the vote was procured by the executive power alone (save for a short Act, later passed, which provided for the use of certain polling booths)[272] and was, on the advice of the Attorney-General’s Department, done so for the stated reason of avoiding parliamentary debate.[273] In 2017, the vote was again procured without fresh legislation[274] and done in a manner which served to avoid the potential for a veto from the Senate. Despite the relative lack of parliamentary scrutiny, the balance of political actors, in these cases, accepted the result. Of course, there were other political reasons why these votes were held, but the common thread is that, in these cases, the plebiscite appears effective at creating politically decisive outcomes, which, by delegating the role of the final arbiter of political debate to ‘the people’, bypasses Parliament’s (in particular, the Senate’s) usual role in that respect. Whatever the motivation for Mr Abbott in saying so, the empirics support his claim, made in the course of debate on the failed Carbon Tax Plebiscite Bill 2011 (Cth), that the people’s vote can ‘settle’ things.[275] Like the proposed solution to the House of Lords’ obstructive veto during the 1911 British constitutional crisis,[276] a recalcitrant government backbencher or a hostile Senate face strong political obstacles in continuing to oppose a course of action approved by a plebiscite. This is what Bagehot might call the ‘efficient’ function of the plebiscite.[277] As with the suggestion that the strength of the House of Lords’ legal veto might be softened by putting contested policies to an advisory referendum,[278] the result, in the Australian case, is also overtly constitutional: as Sawer opined, these practices might modify or ameliorate strict legal powers of oversight or veto.[279] However, unlike a constitutional referendum under s 128, which ordinarily requires parliamentary agreement prior to being held,[280] the examples above show that a plebiscite may be instigated on the motion of executive government alone.

These ‘efficient’ features of the plebiscite have not gone unrecognised in Australian discourse. Concerns over the diminution of the role of Parliament, and a concern over the undermining of responsible government, had been expressed by Members of Parliament in 1916 and 1917;[281] again, in passing, in the final report of the Constitutional Commission in 1988;[282] and in the evidence presented in the Senate Legal and Constitutional Affairs References Committee’s 2015 report.[283]

Having now surveyed the plebiscite, there appear to be four supposed vices. First, resorting to politically decisive popular votes ‘significantly weaken[s] the confidence of the voters in the ability of their elected representatives to reach their own decisions in the lawmaking system’[284] (undermining the logic of representative politics and diminishing the value of an elector’s choice). Second, a use of the plebiscite by means of the executive power alone could impose political barriers to (or a heavy political cost on) the exercise of a Senate veto over government policy (and thus diminish the Senate’s supervisory role), or might displace other, constitutionally prescribed oversight procedures. Third, a capacity to reshape the franchise might diminish certain electors’ political relevance or their choice in the policies of their government. And fourth, the vote may be structured in a way that upsets federal interests, particularly where results are counted on a bare national basis. The central mechanism by which these vices arise appears to be the capacity to generate politically decisive results that practically constrain government and Parliament, for reasons opined on in Part II.

Yet these enumerated concerns remain disjointed and do not present a clear legal schema for thinking about the problem. In the final analysis, bearing the substance of the plebiscite in mind, some firm grounding in doctrine must be sought if normative boundaries to the plebiscite can be understood in law as well as in politics. It is suggested that, on the basis of the description so far given in this article, doctrinal constraints might permeate along at least three axes. First, in the plebiscite’s capacity to offer a political check on dissent in the lower house, or on a hostile upper house, there is a potential for aspects of responsible government or accountability to be undermined. Second, as was hypothesised in Part II, the language and logic of constitutional review in the ‘participation cases’ might be another apposite source of norms. This is because if the supposed threat is to the meaning and value of the system of representative government and electoral choice envisaged by the Constitution, then a bare identification of legal competence may not be sufficient to answer the case.[285] Here, legal norms of representative government are engaged. Third, as also raised in Part II, the examples suggest that, depending on how the vote is counted, or if the vote is used to stymie a Senate veto, concerns relating to the federal balance may be enlivened. Each head is considered below.

A Norms of Responsible Government

The system of responsible government, as understood within the Constitution,[286] offers some basis to resist an expanded use of the plebiscite. In broad compass, the doctrine comprises conventions which provide that government ministers are drawn from Parliament and are accountable to that institution, and that the Governor-General acts on the advice of the ministers.[287] An important link in this chain of accountability is the Senate.[288] The executive power, which is held to account under this system, derives its basis from the text of the Constitution.[289] The boundaries of this executive power have been the subject of extensive debate.[290] Under this schema, the first limitation on the capacity of novel, plebiscitary politics to reshape democratic practice is political, in that strong conventions operate against conduct which undermines the ‘chain of accountability’ between government action and parliamentary scrutiny.[291] This may, however, have some legal import.

Dicey, in his vision for the advisory referendum’s use in the UK, had assumed that it would be procured with explicit parliamentary consent and that both Houses would agree to the process — and so too for Tomn in Australia.[292] Indeed, such has been the subsequent practice in the UK, where fresh primary legislation has always been passed authorising such votes.[293] However, in the Australian cases (other than in 1916), this legislative approval has only been indirect or applied after the fact.[294] In the analysis, a problem arises as to whether this process of partly entrenching a position (or checking another parliamentary actor) is normatively desirable when instigated without direct parliamentary approval.

In this respect, to the observer in the UK, or to the Diceyan observer,[295] the vice with the 2017 Australian exercise might be that Parliament twice spoke against the holding of a popular vote and twice vetoed the expenditure of public funds on the exercise.[296] In a system which observes a strong form of parliamentary sovereignty, the use of an executive power, or non-disallowable subordinate instrument, to act against an inferred parliamentary intention against such expenditure might raise concerns of the type in R v Secretary of State for the Home Department, Ex parte Fire Brigades Union (where the prerogative was used to fund a workers’ compensation scheme substantially different to that which Parliament had envisaged).[297] Or, difficult questions of intention aside, these circumstances might raise more direct concerns about an undermining of government accountability — particularly when it comes to the expenditure of large sums of public funds beyond further scrutiny and in a manner contrary to Parliament’s twice-stated position. In the UK, a rupturing of this ‘chain of accountability’ in the 2019 case of R (Miller) v Prime Minister (‘Miller II’) resulted in the Prime Minister’s actions (in proroguing Parliament to avoid oversight of Brexit arrangements) being ruled unlawful by the Supreme Court of the United Kingdom.[298] A litigant in the UK might also validly argue, as a matter of simple construction, that the Henry VIII power in the Appropriation Act that the Finance Minister relied on in 2017[299] could not be intended to authorise a reallocation of appropriated funds, the purpose for which had been the subject of several parliamentary votes, and which had been rejected at those votes.[300]

It is questionable whether Dicey, a staunch advocate of parliamentary sovereignty, would have envisaged that a plebiscite could be used by government in this way. In a more contemporary Australian critique, speaking at the National Press Club shortly before the trial of Wilkie v Commonwealth (‘Wilkie’),[301] George Williams similarly observed that ‘it would seem unlikely that the Court will permit the government to use this backdoor means of avoiding explicit parliamentary authorisation of its expenditure’.[302] These concerns can tend towards broad normative statements, but they do highlight the problem that arises when the plebiscite is untethered from a direct parliamentary footing.

Whether a use of plebiscitary measures to circumvent these oversight mechanisms is fair politics, or is unconstitutional, must be a matter of degree.[303] Moreover, whether something that is unconstitutional is then justiciable is another question.[304] Whereas the Miller II decision was seen by some UK commentators as an example of judicial overreach and a possible enforcement of political convention,[305] the position in Australia remains that these systems of accountability, as expressed in convention, are beyond the ambit of the courts.[306] This is not to say that these norms, in Australia, are entirely incapable of giving rise to legal claims: the High Court has informed its reasoning in other matters by reference to the existence of conventions of responsible government,[307] and George Winterton, Geoffrey Lindell and Leslie Zines have separately opined that, in the right case (like Miller II), these norms might themselves form the substantive basis of a claim.[308] In other contexts, Gabrielle Appleby and Joanna Howe have observed that the High Court is not afraid to ‘prod’ Parliament towards meeting its supervisory obligations, particularly where accountability over executive discretion or delegated lawmaking powers are concerned (as they were in the case of Wilkie).[309] A related aspect, which also remains unsettled after Wilkie, is whether the constitutional role of MPs and Senators, in holding government to account, might confer standing on them to challenge executive action which undermines that role through plebiscitary means or otherwise.[310]

Relevant to the conduct of plebiscites, there is, however, one area where these norms have risen, more unambiguously, to an enforceable legal standard, and that is in respect of the executive’s power to expend public money on novel measures. It was around this issue that the Wilkie litigation revolved,[311] and to which Paul Kildea’s earlier work was directed.[312] It was said above that, while in 1977 the government was able to expend money on the song poll without legislation authorising it to do so — indeed, such was the advice of the Attorney-General’s Department at the time[313] — by 2017 the view was that it was constitutionally necessary to find some statutory (or other) basis for the expenditure.[314] Such was the need for reliance on the ABS’s statistics-gathering power.[315] The contemporary view remains that only certain bases can lawfully ground executive expenditure, absent some express enactment authorising the expenditure.[316] Over the last quarter-century, the High Court has tended towards greater oversight of executive expenditure, with Wilkie perhaps representing a retreat in that trajectory.[317] This places substantive limits on the government’s capacity to fund plebiscites or like activities.[318] In respect of the exceptions to that rule, it would seem difficult to argue that there would be a common law prerogative to hold a plebiscite (since, historically, such activities were largely alien to British politics and, in modern usage, have been procured by legislation).[319] Moreover, it might be difficult to argue that holding such a vote (with only four having been conducted in Australia’s federal history) is part of the ordinary administration of the government’s functions[320] or otherwise falls within the nationhood power as an exercise peculiar to a national government (versus other deliberative institutions, like the Parliament).[321]

The result is that, absent some actionable claim in protection of government ‘accountability’, legal constraints along this axis appear mostly confined to attacks on the plebiscite’s mode of funding. It is only worth noting, however, that a standard similar to that deployed in Miller II might have led to the 1917 plebiscite (which was held during a parliamentary adjournment) to be invalid,[322] and possibly too, in the Wilkie case (where an inferred parliamentary intention against the expenditure could have been argued).[323] Australian doctrine, however, has not yet developed so far.[324] A solution to this paucity of oversight has, in other places, been proposed in the form of a ‘referendums Act’, which might set out not only those formal aspects of the vote (otherwise regulated by the Referendum (Machinery Provisions) Act 1984 (Cth), should Parliament choose), but also, importantly, the circumstances in which the need for a vote is enlivened, and the mode of its passage through Parliament (including the need for scrutiny at the committee stage).[325] While a standing Act of this type might, to some degree, ameliorate the deficiencies of a system which polices votes on a mostly political basis, it might also undercut the otherwise flexible use of the plebiscite, including by raising concerns about the possible entrenchment of plebiscite procedures.[326]

B Norms of Representative Government

While those political practices which ensure government accountability are mostly non-justiciable, a richer stream of legal doctrine has developed along the contours of how Australians participate in politics and exercise choice over their government. This may now be evaluated against the practice of plebiscites.

The broad contours of what representative government means, in Australia, are well rehearsed.[327] Beginning at the highest level of generality, in the 1926 case of Federal Commissioner of Taxation v Munro, Issacs J held that ‘the Constitution is for the advancement of representative government, and contains no word to alter the fundamental features of that institution’.[328] In

Lange v Australian Broadcasting Corporation (‘Lange’) (which has not been meaningfully displaced),[329] the High Court affirmed that the existence and nature of representative government arises from the text and structure of the Constitution and its provision for elections[330] — and that such representative government is a central feature of the Constitution.[331] The Court’s observations as to the sum of these considerations was that, drawing on AH Birch’s work[332]

[a]t federation, representative government was understood to mean a system of government where the people in free elections elected their representatives to the legislative chamber which occupies the most powerful position in the
political system.[333]

The result today is that ‘representatives of the people share ... in the making of political decisions’ (not ‘the people’ simpliciter).[334] In its identification of the sources of the Constitution’s representative norms, the Court appeared to acknowledge that the only other obvious means of direct popular participation in government (other than voting in elections), which is expressly provided by the Constitution, is given in s 128.[335] In combination, the Court described

(i) this referendum procedure, together with (ii) a system of voting for representatives in elections, which was (iii) underpinned by a doctrine of responsible government, to be ‘the system of government prescribed by the Constitution’.[336] In this description, other avenues of direct participation[337] were not seen to arise, at least not out of necessity.

This observation is certainly not to conclude that purely advisory, federal plebiscites are necessarily forbidden by the Constitution. Such would be an extreme position. The High Court in Lange,[338] and like cases,[339] was not concerned with opining on exercises in ‘direct democracy’. The observation is only to say that the framers of the Constitution decided not to provide for the conduct of these advisory votes in the text of the Constitution.[340] As above, the competence to hold these votes may find an indirect home elsewhere in the text.[341] The question must then become the degree to which the scope of these other, indirect powers is constitutionally circumscribed by the need to maintain a system of representative government.

1 An Irreducible Minimum Content?

As to legislative scope, it is understood that, beyond the rather sparse means of democratic participation provided for in the text of the Constitution, there must be room for democratic experimentalism.[342] Writing extra-curially,

Chief Justice Murray Gleeson observed that ‘[t]he framers of the Constitution did not approach their task with a rigid view of what constituted representative democracy ... Democracy is always in a state of evolution’.[343] He observed further that ‘[t]he Constitution does not seek to entrench more than the bare minimum of conditions for democratic government’, which, he surmises, leaves it ‘to the Parliament to fill in the details’.[344]

Until the 1970s, the federal Parliament’s power to reshape federal representative politics went largely unchecked.[345] However, since then, a series of decisions of the High Court (concerning voters’ rights,[346] the franchise[347] and restrictions on political communication)[348] have placed necessary constitutional limits on Parliament’s capacity to enact laws that reshape representative government.[349] As the High Court has observed, the tension within these cases of democratic experimentalism ‘concern[s] the relationship between the constitutionally mandated fundamentals and the scope for legislative evolution’.[350] However, within this ‘relationship’ there remains, still, some ‘irreducible minimum content’ to representative government.[351]

In cases of this class, the identification of the boundaries of that ‘irreducible minimum content’ appears determined by two related considerations. The first is an application of proportionality reasoning (or proportionality-like reasoning),[352] which seeks to determine whether the impugned activity impermissibly burdens the Constitution’s representative norms.[353] (As introduced in Part II of this article, proportionality reasoning is a standard of review which has applied to disputes over the franchise and over limits on political communication.) The second consideration is a more general juridical view of history and the Constitution, which informs the first question: that is, whether the Constitution should be seen to evolve, irreversibly, towards greater enfranchisement and democratic participation, or whether representative government is a nonlinear ‘living tree’, such that what is required to deliver it may ebb and flow.[354]

In respect of the first matter — ie the means of analysing the burden on representative norms — the High Court’s decision in the 2007 case of

Roach v Electoral Commissioner (‘Roach’) remains authoritative.[355] There, a law had sought to disqualify all persons serving a prison sentence from voting in federal elections.[356] The test for invalidity was whether that new democratic measure was ‘beyond what is reasonably appropriate and adapted (or “proportionate”) to the maintenance of representative government’.[357] In that case, this was answered in the affirmative.[358] The reasoning imports the idea that there is value to an elector’s choice of government, and that measures which limit or modify the existence or value of that choice must be justified.[359] A similar reasoning was adopted by the High Court in the later case of

Rowe v Electoral Commissioner (‘Rowe’).[360] Later, the Court was invited to apply a modified version of this standard in Murphy v Electoral Commissioner (‘Murphy’).[361] The form of review in these cases is a reflection of the language of the earlier political communication cases.[362] In respect of those communication cases, formulations of a structured proportionality test have since been given, but they are, for present purposes, largely similar in substance.[363] The cases of Roach, Rowe, Lange and later decisions in that line have been criticised by some as examples of judicial overreach,[364] but their position in Australian law seems firm.[365] Whether this analysis might also be applicable to other forms of representative experimentalism, beyond those limited matters pertaining to the electoral roll or political communication, remains an open question. There is at least the argument that political communication, participation in a plebiscite, or the casting of a ballot in an election are all species of a genus of political activity that allows Australians to exercise an informed choice over the shape of their government — or, as Isaacs J says in Judd v McKeon, activities which ensure that the government ultimately chosen is ‘expressive of the will of the community’.[366]

What can be said is that, in Australia, this standard of review is certainly not of general application.[367] However, within the rubric of those cases that burden representative government, the High Court has suggested that the standard is potentially broad and not confined to predetermined fact patterns — as in Murphy, where (despite rejecting the claim) French CJ and Bell J held that the reasoning in Roach and Rowe ‘did not set up a sui generis test of validity’[368] but instead that the reasoning represented a standard ‘[w]ithin the normative framework of a representative democracy’.[369] Murphy has shown, however, that different formulations of the standard (structured or otherwise) might be appropriate depending on the form of the challenge to the normative framework of representative democracy.[370]

Within this standard, the boundaries of experimentalism are also determined by the normative view of democratic history the Court adopts.[371] There is a question whether successive liberalising developments become entrenched over time,[372] or whether steps backwards and forwards in the relative empowerment of the people, Parliament and the executive may be taken — a ‘living tree’ model.[373] Without describing in too great detail what is a rich source of discussion,[374] the view to emerge from some of the majority judgments in Rowe[375] is that a linear, ‘evolutionary’ approach is the preferred understanding of Australia’s democratic development.[376] As Twomey describes, under this judicial view, ‘[e]ach liberalising step sets the new benchmark from which there can be no retreat, at least without a substantial reason’.[377] However, that view remains contested, with Twomey arguing that the reasoning effectively entrenches systems of democratic participation in a way that limits parliamentary sovereignty and the capacity of the people to redefine how they participate in politics.[378] Justice Kiefel’s dissent in Rowe also criticised this approach to interpretation.[379] In the more recent case of Murphy, the balance of the Court was against this evolutionary model, where Nettle J, Keane J and Gordon J each queried whether a retreat from an evolved standard was a basis for invalidity.[380]

The result of Rowe and Murphy is that, as Tham observes, there are at least two opposed normative camps, where

the perspective emphasizing breadth based on the benign view of [the power to remodel representative politics] ... sees the exercise of such power as enhancing Australia’s democracy ... [B]y contrast, the perspective emphasizing constraint based on a sceptical view of such power sees the risk of ... subverting the democratic process, therefore rupturing the link between [the people] and parliamentary supremacy.[381]

On the more sceptical view, new practices cannot go so far as to allow Parliament to be written out of the normatively supreme position it held at Federation. However, it remains to be considered, below, what forms that

threat might take.

2 A Constitutional Limit on Plebiscitary Politics?

As observed in Part II, the power to hold plebiscites finds an indirect basis in the Constitution (either pursuant to powers under s 51, or under the executive power). In the few instances where federal plebiscites have been held, the question of whether these powers have been used incompatibly with representative norms has never been posed directly in court (despite occasionally fierce opposition from within Parliament),[382] and isolated exercises in popular consultation might reasonably be seen to sit comfortably within a liberal, evolutionary model of representation. In submitting some matters to a popular vote, a useful function is served in enhancing the people’s choice over their government and its policies.[383]

On the other hand, it is also possible to say that there is a constitutional requirement of government by representation. This requirement may be adaptable, but its ‘irreducible’ central features — an elected body of representatives who are tasked with passing legislation and holding the government to account, and who occupy ‘the most powerful position in the political system’[384] — are firmly entrenched.[385] At some point, as Tham observes, the link between the people’s choice, and the value of Parliament as an institution, cannot be ‘ruptur[ed]’;[386] there must be some purpose and meaning in choosing our representatives under ss 7 and 24 of the Constitution. Neither can the function of Parliament as the supreme normative arbiter be displaced or reduced to a nominal function[387] (as the role of the Crown has been, in some respects, over the past centuries).[388] As Keane J observes in Murphy, the Constitution does not contemplate a future government by means of a ‘“sans-culottes” frenzy’ or by ‘the spontaneous manifestation of the popular will’.[389] If Keane J’s reasoning in Murphy is taken to express an ‘unvarying’[390] originalist standard for representative government[391] — especially when read alongside Gageler J’s more moderate view that there ought to be a ‘relatively stable and enduring understanding’ of what government by representation entails[392] — then this would work more strongly against political exercises which transfer political power away from Parliament, or which otherwise practically constrain its legal powers, such that the practical value of an elector’s choice over the shape of that institution is diminished.

The leap in the argument is to accept that, as Geoffrey Sawer posits, some political practices (such as being committed to follow the result of a plebiscite) may, depending on the circumstances, ‘ha[ve] the effect of reducing the ambit of a legal power’.[393] Or, in the words of Lord Reed, such practices might ‘plainly affect Parliament’s power to make laws ... since it would impose pressure on Parliament’.[394] If this is so, then this constraining effect on the relevant legal power (eg the Senate’s power of veto or the scope of Parliament’s lawmaking power) might be balanced when evaluating the constitutionality of the anterior steps which authorised the plebiscite vote. For example, if an Act (otherwise duly passed) authorised a plebiscite in a way that upset established norms or standards of representative government, or undermined the value of an elector’s vote under ss 7 or 24, then it might be susceptible to review.

Some threats of this type might be direct, such as where clear efforts are made to diminish the political value of a Senate veto (eg by establishing a practice whereby any Bills twice rejected are submitted to a plebiscite, similar to what was proposed in the case of the House of Lords in 1911).[395] This practice might reduce the importance or value of an elector’s Senate vote under s 7 of the Constitution.

Other direct threats to the normative position of Parliament (and the value of an elector’s choice over its composition) might arise in respect of a vote’s subject matter, particularly if it touches upon matters of revenue or of expenditure on government programmes (ie matters usually reserved for decision by Parliament and subject to close review by the Senate).[396] Examples of this type are not unheard of in the common law world. In New Zealand, in 2011, a government Bill proposed that annual plebiscites should be held in order to authorise increases in annual government spending (and changes to taxation) in the climate of austerity following the global financial crisis.[397] The New Zealand Treasury even opined on how amendments to the Referenda (Postal Vote) Act 2000 (NZ) might be effected to give operation to the proposal,[398] and flagged how a plebiscite requirement might fit into the ordinary budget cycle as a matter for further discussion.[399] In the UK, too, advisory referendums have been proposed to resolve disputes over the capacity to raise finance for major infrastructure projects.[400] Of course, in an uncodified constitutional context, there is greater freedom for such experiments.[401] In contrast, a practical constraint on the Australian House of Representatives’ competence to initiate laws raising revenue, and to fund government programmes, may severely undermine the logic of that chamber’s constitutional role and the ‘irreducible minimum content’[402] of what it means to elect a representative under s 24 of the Constitution. These may be extreme examples, but they do highlight the posited threat to the logic of representative government, and the potential impact that such practices may have on the value and meaning of an elector’s constitutional choice over the composition of both Houses of Parliament. While plebiscitary politics might pose other, less obvious threats to the system of representative government (eg a gradual loss of trust in Parliament as a decision-maker),[403] it is hard to imagine these latter cases being as easily embraced by legal constraints.

While future developments cannot be predicted with ease, progress in technology could be expected to make mass, rapid polling easier.[404] Alongside this, as Paul Kildea’s recent work illustrates, the capacity of governments to conduct mass polls or ‘plebisurveys’ — using Henry VIII powers or delegated legislation, in a manner that may avoid conventional means of scrutiny (including at Senate estimates) — offers one potential avenue by which these votes could be conducted in the future.[405] Such functions might even be partially privatised or automated.[406] However, the trend need not be in the direction of using executive powers alone; as the New Zealand example shows, such votes might also be procured by Parliament’s own hand in a time of perceived emergency (as in the wake of the global financial crisis when the Spending Cap (People’s Veto) Bill 2011 (NZ) was proposed).[407]

The sum of this behaviour is to impose a practical limitation on government or Parliament’s capacity to do business — or, in the language of Rowe, to impose a potential ‘burden’ on the normative system of representative government that must be weighed.[408] Of course, that weighing will not always yield to the result that steps which impose an impediment on government or Parliament’s legal powers will be disproportionate within a system of electoral politics, or necessarily impinge upon the meaning of an elector’s choice under ss 7 or 24 in an unconstitutional way. Writing extra-curially, Chief Justice Murray Gleeson gave what might be Delphic guidance, in saying that there is a need to distinguish ‘between [threats to] the essence of representative democracy, and [changes to] the forms of representative democracy resulting from accidents of history, or ideas fashionable at particular times’.[409]

In this respect, the examples given in the paragraphs above are relatively extreme cases. In a less remote example, if the question under this rubric is one of safeguarding the electors’ choice over the shape of Parliament, then, at the least, there appears an argument that a plebiscitary measure which confers a new mandate on government should be done in a way where the franchise matches as closely as possible the electoral franchise — particularly where Parliament commits to adopting the policy. Otherwise, it remains arguable that the electoral choice of certain electors, who are shut out of these other forms of political activity, is impaired.

The use of the plebiscite as a tool of executive government further highlights the problem. It was seen in the 1916 and 1917 plebiscites that the franchise was substantively limited by the Prime Minister under wartime regulations[410] — achieved while Parliament was adjourned — yet the result carried in a manner that politically checked government and Parliament. The historical cases show that it may be possible, in a future case, to attempt to alter the franchise in at least two ways: either under enabling legislation (including through any delegated legislation), or, in the event of something like a ‘plebisurvey’,[411] by administrative actors being directed to limit the class of ‘respondents’.[412] The potential disjunction between the parliamentary franchise and the plebiscitary franchise (and the capacity of executive government to shape the latter) raises a problem in respect of an elector’s constrained choice, or capacity to participate in a dialogue, over the form of their government. In respect of a ‘choice’ over that government, or a participation in a political dialogue, this may, as in the more extreme cases outlined above, enliven the implied limits articulated in the Roach, Rowe and Lange class of cases.[413] However, where this evaluation lands must depend on the severity of the burden and the vision of democracy that the court chooses to adopt. With a potential retreat from the evolutionary model heralded by Murphy,[414] and against the backdrop of a contemporary ‘common understanding’ which appreciates that some issues are helpfully resolved by popular votes,[415] it seems unlikely that all but the most far-reaching or egregious examples would be found wanting.

C Representative Norms and Federal Concerns

There is one final, related axis along which normative constraints might operate. As a sharp knife that cuts through intra-party or inter-house political deadlocks,[416] there exists the potential for a national plebiscite to upset the federal balance if votes are counted on the basis of a bare, national majority.[417] Unlike a referendum under s 128, which requires both state and federal majorities, a national plebiscite might under-weight the votes of the less populous states.[418] The tensions implicit in this result were raised at the Convention Debates and, to understand this concern, some recourse to the Constitution’s drafting history must be had.

In broad compass, by 1898, it had been agreed by the Convention delegates that the Constitution would be amended by means of a double referendum (ie s 128).[419] However, in the final weeks of debate, there was disagreement over whether a referendum mechanism should also be used to resolve parliamentary deadlocks.[420] Among other delegates,[421] Isaac Isaacs was in favour of using a ‘national referendum’ (ie a bare majority national vote, versus a double referendum) to resolve parliamentary deadlocks on matters of ordinary legislation.[422] In so advocating, Isaacs said that he was ‘fighting against ... the destruction of the powers of the House of Representatives by means of the decisive veto of the Senate’.[423] Speaking of the referendum as a means of resolving deadlocks, Issacs observed that ‘the objections to it are mainly that it will destroy ... representative government’.[424] However, Isaacs’ position was that a failure to resolve inter-house disagreements expeditiously, or by a more democratic means, might otherwise feed into ‘a strong and a growing feeling that representative government is becoming inadequate’.[425] Other delegates disagreed,[426] including Bernhard Wise, who argued that

[t]o make use of the referendum ... for the purpose of determining whether the Constitution shall be ... amended, is justifiable ... There is no other way of ascertaining what the people wish shall be the limits of the representative system. But when once a Constitution has been framed, based on parliamentary government ... to introduce into this Constitution, in any form, the principle of the referendum, is to introduce a subtle poison, which before long will altogether destroy the vitality of Parliament.[427]

Sentiments of the type of Mr Wise’s carried. The deadlock procedure agreed upon was a double dissolution of the House of Representatives and the Senate, now given in s 57 of the Constitution.[428] Had that not been the case, controversy over the conscription legislation, or possibly the passage of the marriage plebiscite legislation, might have been determined differently. More generally, it is apparent that the exclusion of an expanded use of referendum procedures from the Constitution, on issues of ordinary legislation, was deliberate.[429]

This result does not, however, preclude the hypothetical case of the major political parties agreeing that the best way to resolve disagreement between the Houses, on a discrete matter, is to submit a vexed policy to a plebiscite instead. This, it might be argued, would give the disputing Houses an indication of what the people’s true wishes were, on a given issue, in a politically authoritative way. If, for whatever reason, the plebiscite result was calculated on a bare national basis (rather than as a ‘double’ vote), then there is a risk that the value of an elector’s vote from a less populous state would be diminished. If the plebiscite result nonetheless politically constrained lawmakers in the Senate, practically limiting the exercise of their legal powers, then there is an argument that the value of an elector’s choice (from a less populous state) over the form of the Senate, exercised under s 7 of the Constitution, is also diminished. In practical terms, they remain capable of voting for a Senator at an election but, as a consequence of the political effects of the plebiscite, their interests may not be adequately represented (at least relative to other electors).

This reasoning should not seem foreign. Such a result may, for example, be analogous to a federal measure, the effect of which was to censor political communications in one state, but not another. In that case, too, the information available to a person from one state, and their capacity to shape electoral outcomes, may be diminished relative to another. However, this argument must be weighed against the High Court’s rejection of the view, in McGinty v Western Australia, that all votes must necessarily carry equal weight.[430]

The degree to which the Convention Debates should inform contemporary judicial exegesis of the Constitution remains in part unsettled.[431] However, if even limited recourse may be had to the Debates on the question of the resolution of disagreement between the Houses, then those records might, in a representation or communications case, militate against an expansive use of powers towards popular measures that undercut either the logic of a strong Senate veto or the relative value of electors’ choice over the composition of

that chamber.


The plebiscite has been described as ‘no more than a formalised, national opinion poll’,[432] but its position within the Commonwealth Constitution appears more complex. In locating the plebiscite in the Constitution, the first observation is that, as raised in Paul Kildea’s work,[433] there is no direct competence to hold plebiscites (unlike referendums under s 128), and so indirect bases must be found. These may be various.

However, the important additional observation is that, while plebiscites are not legally binding in their outcomes, accounts of political constitutionalism illustrate that they do appear capable of ‘binding’ government and Parliament politically.[434] So framed, the authorities might be capable of accommodating a view of the plebiscite as a singular form of participation that can reshape the meaning (or value) of an elector’s choice over their representative government in a constitutionally relevant way. It is as a species of political participation, and as an activity relevant to the value of electoral choice, that the plebiscite might find a clearer articulation in legal doctrine. The mode by which this choice is exercised might also have the practical effect of constraining the legal powers of Parliament or diminishing the political value of a Senate veto — where these results, too, engage legal norms of responsible government and concerns regarding the federal balance.

Within this understanding, what of the plebiscite’s future use? The authorities certainly do not rise high enough to suggest that plebiscites ought to have little or no role in the future of Australian democracy, or that any previous exercises have been unlawful. As Anne Twomey observes,[435] plebiscites can be democratically useful when, as Dicey and Tomn also saw,[436] a discrete policy issue divides the party, the people or Parliament. There is little which is democratically objectionable about seeking a mandate, or useful information, from the electorate.[437] But the observations made here do suggest a more principled approach to the plebiscite’s use is desirable — where the safeguards of parliamentary deliberation are not avoided, but embraced.

The various means by which this might be achieved ought properly be developed in a separate work.[438] However, in brief, and as suggested in Part IV, this might be done through a general Act (rather than ad hoc legislation) setting out certain formal aspects of votes (including the scope of the franchise) and prescribing a practice of putting proposed votes to the consideration of relevant parliamentary committees, or providing for other means of scrutiny.[439] Of course, such legislation would not be entrenched and would be open to amendment or repeal — but it would nonetheless create a strong presumption that government will follow certain procedures when conducting a popular consultation of this type. Even if such an ideal is not achieved (or is deemed unnecessary), the historical cases illustrate that efforts should be made, at least, to authorise votes directly through fresh legislation (even if on an ad hoc basis), rather than relying on executive authorisation alone.[440] The contemporary powers available to executive government (in respect of its power to pay for, and conduct, mass surveys) do not obviate the historical concern.

The concern arises because, in its more vexed, ‘efficient’ usages as a tool of executive government,[441] the plebiscite has served either as a substitute for other constitutionally prescribed means of resolving deadlock, or as a mechanism by which executive government may create commitments that politically constrain Parliament, without an opportunity for effective prior parliamentary scrutiny. Because of the narrow bases on which the Wilkie case was argued, it must remain an unsettled normative question whether powers put to procure this use of the plebiscite are constitutionally desirable, if so used. But, as Gageler J said in McCloy v New South Wales (speaking of the power to control matters of communication), the historical illustrations may show that there is at least a ‘systemic risk’ of their abuse.[442]

[*] Director of Studies of Law and College Teaching Officer in Law, St Edmund’s College, University of Cambridge; BA, JD (Melb), LLM (Cantab), PhD Candidate (Cantab); formerly a solicitor at Allens. Thank you to Alison Young and Jessie Smith for their invaluable assistance, and to the anonymous referees for their helpful comments.

[1] See generally Anne Twomey, ‘The Federal Constitutional Right To Vote in Australia’ [2000] FedLawRw 6; (2000) 28(1) Federal Law Review 125; Joo-Cheong Tham, ‘Political Participation’ in Cheryl Saunders and Adrienne Stone (eds), The Oxford Handbook of the Australian Constitution (Oxford University Press, 2018) 979. See also McGinty v Western Australia [1996] HCA 48; (1996) 186 CLR 140, 201 (Toohey J), 221–2 (Gaudron J), 286–7 (Gummow J) (‘McGinty’); Rowe v Electoral Commissioner [2016] FCA 1081; (2010) 243 CLR 1, 18 [18]–[22] (French CJ) (‘Rowe’).

[2] See generally A-G (Cth) ex rel McKinlay v Commonwealth [1975] HCA 53; (1975) 135 CLR 1 (‘McKinlay’); Roach v Electoral Commissioner (2007) 233 CLR 162 (‘Roach’); Rowe (n 1); Murphy v Electoral Commissioner [2016] HCA 36; (2016) 261 CLR 28 (‘Murphy’). See Day v Australian Electoral Officer (SA) (2016) 261 CLR 1, 23–4 [51]–[54] (French CJ, Kiefel, Bell, Gageler, Keane, Nettle and

Gordon JJ).

[3] See generally Nationwide News Pty Ltd v Wills [1992] HCA 46; (1992) 177 CLR 1 (‘Nationwide’); Australian Capital Television Pty Ltd v Commonwealth (1992) 177 CLR 106 (‘ACTV’); Theophanous v The Herald & Weekly Times Ltd [1994] HCA 46; (1994) 182 CLR 104 (‘Theophanous’); Lange v Australian Broadcasting Corporation [1997] HCA 25; (1997) 189 CLR 520 (‘Lange’); Coleman v Power [2004] HCA 39; (2004) 220 CLR 1 (‘Coleman’); McCloy v New South Wales [2015] HCA 34; (2015) 257 CLR 178 (‘McCloy’); Murphy (n 2);

Brown v Tasmania (2017) 261 CLR 328 (‘Brown v Tasmania’).

[4] Love v Commonwealth (2020) 270 CLR 152, 261–2 [295], 278 [354]–[355] (Gordon J), 286–7 [391], 314–16 [452]–[454] (Edelman J). See generally Elisa Arcioni, ‘The Core of the Australian Constitutional People: “The People” as “the Electors”’ [2016] UNSWLawJl 16; (2016) 39(1) University of New South Wales Law Journal 421. See also Justice RS French, ‘The Constitution and the People’ in Robert French, Geoffrey Lindell and Cheryl Saunders (eds), Reflections on the Australian Constitution (Federation Press, 2003) 60, 75–6.

[5] Mulholland v Australian Electoral Commission [2004] HCA 41; (2004) 220 CLR 181, 189–90 [9]–[10], [14] (Gleeson CJ), 206 [63], 213–14 [78] (McHugh J) (‘Mulholland’). The power of Parliament to shape aspects of representative democracy has long been recognised: Smith v Oldham [1912] HCA 61; (1912) 15 CLR 355, 358 (Griffith CJ), 361 (Barton J), 362 (Isaacs J); Judd v McKeon [1926] HCA 33; (1926) 38 CLR 380, 383 (Knox CJ, Gavan Duffy and Starke JJ), 385 (Isaacs J) (‘Judd’). See also Chief Justice Murray Gleeson, ‘The Shape of Representative Democracy’ [2001] MonashULawRw 1; (2001) 27(1) Monash University Law Review 1, 7; Anne Twomey, ‘Rowe v Electoral Commissioner: Evolution or Creationism?’ [2012] UQLawJl 9; (2012) 31(2) University of Queensland Law Journal 181, 185, 191; Gabrielle Appleby and Brendan Lim, ‘Democratic Experimentalism’ in Rosalind Dixon (ed), Australian Constitutional Values (Hart Publishing, 2018) 221, 226.

[6] See, eg, Roach (n 2) 174 [7], 182 [24] (Gleeson CJ), 199 [85], 202 [95] (Gummow, Kirby and Crennan JJ); Rowe (n 1) 19 [22], 20–1 [25] (French CJ), 56 [150] (Gummow and Bell JJ), 107 [328] (Crennan J); McCloy (n 3) 226–8 [112]–[118] (Gageler J).

[7] McCloy (n 3) 226–8 [112]–[118] (Gageler J).

[8] See Appleby and Lim (n 5) 224–6.

[9] See Anne Twomey, ‘Plebiscites and Referenda’ (2015) 89(12) Australian Law Journal 832, 832.

[10] These plebiscites have concerned conscription in 1916 and 1917, the national song in 1977, and marriage equality in 2017: Parliamentary Library, Department of Parliamentary Services (Cth), Parliamentary Handbook of the Commonwealth of Australia 2020: 46th Parliament (35th ed, 2020) 414 (‘Parliamentary Handbook’). The status of the 2017 exercise as a ‘plebiscite’ remains unsettled, though some of its proponents have since labelled it as a plebiscite: Tony Abbott, ‘Problems with a Plebiscite for a Republic’ (2018) 30 Upholding the Australian Constitution: Proceedings of the Samuel Griffith Society 154, 154.

[11] See Twomey, ‘Plebiscites and Referenda’ (n 9) 834; Paul Kildea, ‘The Constitutional and Regulatory Dimensions of Plebiscites in Australia’ (2016) 27(4) Public Law Review 290, 292 (‘Constitutional and Regulatory Dimensions’).

[12] Geoffrey Lindell, Submission No 4 to Senate Legal and Constitutional Affairs References Committee, Parliament of Australia, Matter of a Popular Vote, in the Form of a Plebiscite or Referendum, on the Matter of Marriage in Australia (30 August 2015) 5 [13] (‘Submission

No 4’); Michael Kirby, ‘250 Years of the Crown in Australia: From James Cook to the Palace Papers (1770–2020)’ (2021) 95(7) Australian Law Journal 520, 528 (‘250 Years of the Crown’).

[13] See, eg, Department of Prime Minister and Cabinet (Cth), Australia’s Second Open Government National Action Plan 2018–20 (Plan, September 2018) 5; Department of Industry, Innovation and Science (Cth), Hidden in Plain Sight: Building an Understanding of How the Australian Public Service Can Unlock Community Expertise To Improve Policy, Programmes and Service Delivery (Discover Phase Report, 2017) 4–6.

[14] The political phenomenon is broad, but in the Australian context, see Benjamin Moffitt, ‘Populism in Australia and New Zealand’ in Cristóbal Rovira Kaltwasser et al (eds), The Oxford Handbook of Populism (Oxford University Press, 2017) 121, 121–2, 123–7, 130–4. See also Simon Jackman, ‘Populism and Discontent: Comparing the United States and Australia’ (Senate Occasional Lecture, Parliament House, 17 February 2017).

[15] On ‘direct democracy in an age of instability’, see Stephen Tierney, Constitutional Referendums: The Theory and Practice of Republican Deliberation (Oxford University Press, 2012) 1, 6–11.

[16] Flags Act 1953 (Cth) s 3(2) (‘Flags Act’), as amended by the Howard government in Flags Amendment Act 1998 (Cth) sch 1 item 1. See below Part II(B)(2).

[17] See, eg, Carbon Tax Plebiscite Bill 2011 (Cth) cl 5, which proposed a plebiscite on Labor’s proposed carbon tax; Plebiscite (Future Migration Level) Bill 2018 (Cth) cl 5(2), which proposed to poll electors on the question: ‘Do you think the current rate of immigration to Australia is too high?’; Plebiscite (Restricting Non-European Migration) Bill 2018 (Cth) cls 5(2)–(4), which proposed that three plebiscite questions be asked, concerning first, the current rate of immigration; second, the need to ‘preserve the current ethnic composition of society’; and third, the desirability of Muslim immigrants. Plebiscites have been proposed on various further topics, including assisted suicide, though they have not made it to the Bill stage: Garry Linnell, ‘The Plebiscite We Should Have: Is There a Right To Die?’, The Sydney Morning Herald (online, 28 July 2016) <> , archived at <>; ‘Peter Reith Suggests More Plebiscites Needed in Australian Politics’, OUTinPerth, (online, 26 July 2016) <> , archived at <>.

[18] George Williams and David Hume, People Power: The History and Future of the Referendum in Australia (UNSW Press, 2010) 6; Alister Henskens, ‘Can There Be a Plebiscite on the Road to an Australian Republic?’ (2018) 30 Upholding the Australian Constitution: Proceedings of the Samuel Griffith Society 147, 149–50; Kirby, ‘250 Years of the Crown’ (n 12) 528.

[19] See generally Amelia Simpson, ‘Parliaments’ in Cheryl Saunders and Adrienne Stone (eds), The Oxford Handbook of the Australian Constitution (Oxford University Press, 2018) 563, 563; Lange (n 3) 557–9 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ).

[20] McKinlay (n 2) 56 (Stephen J).

[21] See Lilian Tomn, ‘The Referendum in Australia and New Zealand’ (1897) 72 (July–December) Contemporary Review 242, 242–6. Since Federation, 33 plebiscites have been held in the states and territories: Antony Green, ‘Plebiscites in Australian History’, ABC News (online,

30 September 2016) <>, archived at <>; Anne Twomey, Submission No 6 to Senate Legal and Constitutional Affairs References Committee, Parliament of Australia, Matter of a Popular Vote, in the Form of a Plebiscite or Referendum, on the Matter of Marriage in Australia (31 August 2015) 5 (‘Submission No 6’).

[22] See Nicholas Aroney, The Constitution of a Federal Commonwealth: The Making and Meaning of the Australian Constitution (Cambridge University Press, 2008) 299–301, 308–15 (‘The Constitution of a Federal Commonwealth’).

[23] Ibid 240–2.

[24] However, deadlock over a Bill proposed under s 128 of the Constitution may be resolved by putting the proposed law directly to the electors in a referendum.

[25] Aroney, The Constitution of a Federal Commonwealth (n 22) 242.

[26] Harry Hobbs and Andrew Trotter, ‘The Constitutional Conventions and Constitutional Change: Making Sense of Multiple Intentions’ [2017] AdelLawRw 3; (2017) 38(1) Adelaide Law Review 49, 78–81.

[27] Ibid 73–6. See also George Williams and Geraldine Chin, ‘The Failure of Citizens’ Initiated Referenda Proposals in Australia: New Directions for Popular Participation?’ (2000) 35(1) Australian Journal of Political Science 27, 36–8 (‘The Failure of CIR Proposals’); George Williams and Geraldine Chin, ‘Australian Experiments with Community Initiated Referendum: CIR for the ACT?’ [1998] GriffLawRw 16; (1998) 7(2) Griffith Law Review 274, 277–80 (‘CIR for

the ACT?’).

[28] See above nn 17–18.

[29] Anne Twomey, ‘Let’s Not Demonise Plebiscite in the Heat of the Debate over Our Marriage Laws’, The Australian (Sydney, 7 October 2016) 28 (‘Let’s Not Demonise Plebiscite’); Twomey, Submission No 6 (n 21) 3; Abbott (n 10) 154; James Allan, Submission No 19 to Senate Legal and Constitutional Affairs References Committee, Parliament of Australia, Matter of a Popular Vote, in the Form of a Plebiscite or Referendum, on the Matter of Marriage in Australia

(4 September 2015) 1 (‘Submission No 19’).

[30] Williams and Hume (n 18) 6–8, 249–51.

[31] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 295–9; Paul Kildea, ‘Setting the Ground Rules for the Same-Sex Marriage Plebiscite’, Australian Public Law (Blog Post,

27 April 2016) <>, archived at <> (‘Setting the Ground Rules’); Anne Twomey, ‘A Tale of Two Cases: Wilkie v Commonwealth and Re Canavan(2018) 92(1) Australian Law Journal 17,

17–19 (‘A Tale of Two Cases’); Anne Twomey, ‘Wilkie v Commonwealth: A Retreat to Combet over the Bones of Pape, Williams, and Responsible Government’, Australian Public Law (Blog Post, 27 November 2017) <>, archived at <> (‘Wilkie v Commonwealth: A Retreat to Combet’); Angus Brown, ‘Responsible Government and Parliamentary Intention: The Impact of Wilkie v Commonwealth[2019] SydLawRw 23; (2019) 41(4) Sydney Law Review 517, 518–19, 527–32; Ryan Goss, ‘Explainer: With No Free Vote for Now, Where Next for Marriage Equality?’, The Conversation (online, 7 August 2017) <>, archived at <>; Paul Kildea, ‘Using the ABS To Conduct a Same-Sex Marriage Poll Is Legally Shaky and Lacks Legitimacy’, The Conversation (online, 10 August 2017) <>, archived at <> (‘Using the ABS’).

[32] The existing literature has tended to focus on the plebiscite as a purely political device, on the formal aspects of how a vote might be counted or scrutinised, or on the narrower legal question of the means by which funding is authorised, yet few challenges to its normative constitutional standing are articulated in detail: see Williams and Hume (n 18) 6–8, 249–51; Twomey, ‘Plebiscites and Referenda’ (n 9); Paul Kildea, ‘Constitutional and Regulatory Dimensions’

(n 11) 295–310; Brown (n 31) 517–19, 527–32; Matt Qvortrup, Caroline Morris and Masahiro Kobori, ‘Australasia’ in Matt Qvortrup (ed), Referendums around the World (Palgrave Macmillan, 2018) 237, 250–1; Graeme Orr, ‘The Conduct of Referenda and Plebiscites in Australia: A Legal Perspective’ (2000) 11(2) Public Law Review 117, 117–25, 131–2; Graeme Orr, ‘Electoral Reform as a Tonic for Referenda and Federalism: A Response to Professor Craven’ (2005) 20(2) Australasian Parliamentary Review 83, 84–90; Graeme Orr, ‘Voluntary Voting for Referendums in Australia: Old Wine, New Bottle’ in Ron Levy et al (eds), New Directions for Law in Australia: Essays in Contemporary Law Reform (Australian National University Press, 2017) 359, 359–61 (‘Old Wine, New Bottle’); Kirby, ‘250 Years of the Crown’ (n 12) 528; Michael Kirby, ‘Marriage Equality Law and the Tale of Three Cities: How the Unimaginable Became Inevitable and Even Desirable’ [2016] AukULawRw 2; (2016) 22 Auckland University Law Review 11, 33 (‘The Tale of Three Cities’); Michael Kirby, ‘Beyond Marriage Equality & Skin Curling’ (2018) 6(2) Griffith Journal of Law and Human Dignity 1, 6–7 (‘Skin Curling’); George Williams, ‘Distrust of Representative Government: Australian Experiments with Direct Democracy’ in Marian Sawer and Gianni Zappalà (eds), Speaking for the People: Representation in Australian Politics (Melbourne University Press, 2001) 80, 94–8; Paul Kildea, ‘Australia’s Same-Sex Marriage Survey: Evaluating a Unique Popular Vote Process’ [2020] MonashULawRw 13; (2020) 46(2) Monash University Law Review 107, 109–11 (‘Australia’s Same-Sex Marriage Survey’). Cf Henskens

(n 18) 155–62.

[33] Wilkie v Commonwealth [2017] HCA 40; (2017) 263 CLR 487, 509 [6], 514–15 [25]–[28] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ) (‘Wilkie’).

[34] A total of 33 plebiscites have been held at the state or territory level. For an account of the plebiscites held at the state or territory level, see Green (n 21); Twomey, Submission No 6

(n 21) 5. For plebiscites in colonial Australia, see generally Tomn (n 21).

[35] See above nn 17–18.

[36] See, eg, Wilkie (n 33) 521–2 [56] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and

Edelman JJ), where in giving reasons on the legality of the 2017 marriage equality survey, the Court observed that aspects of the argument were far from fully developed, as the trial, of necessity, had been brought on within four weeks of the survey being announced. In 1917, Parliament was adjourned when the conscription plebiscite was proposed, and only resumed after the plebiscite had been effected: Rob Lundie and Joy McCann, ‘Commonwealth Parliament from 1901 to World War I’ (Research Paper, Parliamentary Library, Parliament of Australia, 4 May 2015) 28. See also JH Catts, Labor and the War: Stumbling Blocks to Reform (Government Printer, 1918) 23–4.

[37] See Adam Webster, ‘Referendums in the UK and Plebiscites in Australia: When To Seek the Views of the People?’ [2019] (October) Public Law 746, 746–9; Matt Qvortrup, ‘Citizen Initiated Referendums (CIRS) in New Zealand: A Comparative Appraisal’ (2008) 44(1) Representation 69, 71–5.

[38] Wilkie (n 33) 545 [142] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ).

[39] Ibid 545 [143].

[40] See generally Lindell, Submission No 4 (n 12).

[41] Twomey, ‘Plebiscites and Referenda’ (n 9) 835.

[42] Ibid. See also Kirby, ‘250 Years of the Crown’ (n 12) 528; Webster (n 37) 750–1.

[43] Since Australia federated, there have been 44 referendum questions put, of which only eight carried a result leading to a change in the Constitution: Parliamentary Handbook (n 10) 413.

[44] Constitution s 128. See also Twomey, ‘Plebiscites and Referenda’ (n 9) 833.

[45] Procedural matters in the Referendum (Machinery Provisions) Act 1984 (Cth) include how writs for the vote will be issued: pt II; where and how the polling will take place: pt III; the appointment of scrutineers: s 27; counting of the ballots: pt VI; and mechanisms for disputed returns: pt VIII. The Act also contains provisions to ensure relative government neutrality on the vote: see, eg, s 11 regarding the ‘in favour’/’against’ pamphlet. It also provides for neutrality in polling places: ss 126, 131–2; and that voting, as in Parliamentary elections, shall be compulsory: s 45.

[46] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 300–1; Paul Kildea, ‘Setting the Ground Rules’ (n 31).

[47] The term has been mobile. The 1916 and 1917 conscription plebiscites were called ‘referendums’ and the 1977 national song plebiscite has been variously called a ‘referendum’ or ‘poll’: see Constitutional Commission, Report of the Constitutional Commission (Final Report, 1988) vol 2, 867 [13.72].

[48] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 295.

[49] Ibid 295–9.

[50] See, eg, Constitution s 51(xxi). See also ibid 295–7.

[51] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 297–9.

[52] Victoria v The Australian Building Construction Employees’ and Builders Labourers’ Federation [1982] HCA 31; (1982) 152 CLR 25, 156 (Brennan J).

[53] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 298.

[54] Victoria v Commonwealth [1975] HCA 52; (1975) 134 CLR 338, 397 (Mason J) (‘AAP Case’), applied in

Pape v Federal Commissioner of Taxation [2009] HCA 23; (2009) 238 CLR 1, 91–2 [241]–[242] (Gummow, Crennan and Bell JJ) (‘Pape’); Davis v Commonwealth [1988] HCA 63; (1988) 166 CLR 79, 94 (Mason CJ, Deane and Gaudron JJ).

[55] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 298 (emphasis added).

[56] Ibid 299–301.

[57] Ibid 299.

[58] A direction was made from a government minister to the Australian Electoral Office to issue the ballots: see Department of Parliamentary Services (Cth), Bills Digest (Digest No 22 of

2016–17, 11 October 2016) 4 n 19.

[59] (1990) 169 CLR 195, 205, 208 (Mason CJ, Brennan, Deane, Dawson and Toohey JJ), quoting Auckland Harbour Board v The King [1924] AC 318, 326 (Viscount Haldane for Viscount Haldane, Lords Sumner, Parmoor and Wrenbury and Sir Charles Darling) (‘Auckland

Harbour Board’).

[60] Pape (n 54) 55 [111] (French CJ), 74 [183] (Gummow, Crennan and Bell JJ), 104 [292], 105 [296] (Hayne and Kiefel JJ), 210–12 [601]–[604] (Heydon J).

[61] [2012] HCA 23; (2012) 248 CLR 156, 187 [27], 192–3 [36]–[37], 216–17 [83] (French CJ), 236–9 [150]–[159] (Gummow and Bell JJ), 353 [524] (Crennan J) (‘Williams [No 1]’).

[62] [2014] HCA 23; (2014) 252 CLR 416, 463–5 [59]–[65] (French CJ, Hayne, Kiefel, Bell and Keane JJ), 471 [99] (Crennan J) (‘Williams [No 2]’).

[63] Anne Twomey, ‘Post-Williams Expenditure: When Can the Commonwealth and States Spend Public Money without Parliamentary Authorisation?’ [2014] UQLawJl 2; (2014) 33(1) University of Queensland Law Journal 9, 9 (‘Post-Williams Expenditure’).

[64] Katherine Richardson, James Emmett and Surya Palaniappan, ‘Potential Postal Plebiscite on Same Sex Marriage’ (Joint Opinion, 4 August 2017) 1 [2], 6 [28]; Michael Koziol, ‘Postal Vote on Same-Sex Marriage Would Be Invalid without Legislation: Lawyers’, The Sydney Morning Herald (online, 5 August 2017) <>, archived at <>; George Williams, ‘Dual Citizens and the Postal Survey: What Might the High Court Say?’ (Speech, National Press Club, 30 August 2017) (‘Dual Citizens and the Postal Survey’). See also Goss (n 31); Paul Kildea, ‘Using the ABS’

(n 31).

[65] See Goss (n 31).

[66] Wilkie (n 33) 544–6 [140]–[143], [146]–[148] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ). See also Paul Kildea, ‘Australia’s Same-Sex Marriage Survey’ (n 32) 115.

[67] See Gabrielle Appleby, ‘Unwritten Rules’ in Cheryl Saunders and Adrienne Stone (eds), The Oxford Handbook of the Australian Constitution (Oxford University Press, 2018) 209, 209–10, 214–16.

[68] Ibid 209–10.

[69] See ibid 217–18.

[70] Geoffrey Sawer, Federation under Strain: Australia 1972–1975 (Melbourne University Press, 1977) 179 (emphasis added).

[71] Senate Legal and Constitutional Affairs References Committee, Parliament of Australia, Matter of a Popular Vote, in the Form of a Plebiscite or Referendum, on the Matter of Marriage in Australia (Report, September 2015) 10 [2.16], citing George Williams, Submission No 32 to Senate Legal and Constitutional Affairs References Committee, Parliament of Australia, Matter of a Popular Vote, in the Form of a Plebiscite or Referendum, on the Matter of Marriage in Australia (3 September 2015) 2 (emphasis added) (‘Submission No 32’).

[72] Commonwealth, Parliamentary Debates, House of Representatives, 22 August 2011, 8727–8 (Tony Abbott, Leader of the Opposition) (‘Parliamentary Debates (22 August 2011)’).

[73] See, eg, ibid 8844–5 (Gai Brodtman).

[74] See, eg, George Williams, ‘National Plebiscite on Carbon Tax an Expensive, Bad

Idea’, The Sydney Morning Herald (online, 21 June 2011) <>, archived at <>.

[75] Commonwealth, Parliamentary Debates, House of Representatives, 15 September 2011, 10300–1.

[76] See below Part III.

[77] In the case of marriage equality, polls had indicated a majority of Australians supported reform; however, this was not translated into manifesto items or firm political commitments until much later: see, eg, ‘Senior Australian MP Calls for Gay Marriage Referendum’, BBC News (online, 13 August 2015) <>, archived at <>.

[78] In 2017, after the return of the survey results, some MPs still did not vote in favour

of the Marriage Amendment (Definition and Religious Freedoms) Bill 2017 (Cth):

Paul Karp, ‘Marriage Equality Law Passes Australia's Parliament in Landslide Vote’, The Guardian (online, 7 December 2017) <>, archived at <>.

[79] See Michael Gallagher, ‘Conclusion’ in Michael Gallagher and Pier Vincenzo Uleri (eds), The Referendum Experience in Europe (Macmillan Press, 1996) 226, 246; David Altman, Direct Democracy Worldwide (Cambridge University Press, 2011) 48–9.

[80] See Alan Fenna and Rob Manwaring, ‘Political Parties and the Australian Party System’ in Alan Fenna and Rob Manwaring (eds), Australian Government and Politics (Pearson, 2021) 112, 113–16.

[81] Vernon Bogdanor, Beyond Brexit: Britain’s Unprotected Constitution (IB Tauris, 2019) 111.

[82] Sawer (n 70) 179.

[83] Transcript of Proceedings, Hospice New Zealand v A-G (NZ) (High Court of New Zealand, CIV-2020-485-000176, Mallon J, 25 May 2020) 13, 166 (DR La Hood) (‘Transcript of Proceedings, Hospice New Zealand v A-G (NZ)’).

[84] Hospice New Zealand v A-G (NZ) [2020] NZHC 1356; [2021] 3 NZLR 71, 77 [6], 82–3 [38] (Mallon J).

[85] Transcript of Proceedings, Hospice New Zealand v A-G (NZ) (n 83) 166 (DR La Hood).

[86] See Philip A Joseph, Joseph on Constitutional and Administrative Law (Thomson Reuters, 5th ed, 2021) 628–30.

[87] See Timothy Shiels and Andrew Geddis, ‘Tracking the Pendulum Swing on Legislative Entrenchment in New Zealand’ (2020) 41(2) Statute Law Review 207, 212, 215–16, which discusses the possibility of the entrenchment of manner and form requirements (including in respect of a referendum).

[88] Senate Legal and Constitutional Affairs References Committee (n 71) 10 [2.13]–[2.14].

[89] Kristen Walker, Submission No 36 to Senate Legal and Constitutional Affairs References Committee, Parliament of Australia, Matter of a Popular Vote, in the Form of a Plebiscite or Referendum, on the Matter of Marriage in Australia (16 September 2015) 2–4; Williams, Submission No 32 (n 71) 3; Lindell, Submission No 4 (n 12) 2–5 [5]–[13]; Rosalind Dixon, Paul Kildea and Andrew Lynch, Submission No 11 to Senate Legal and Constitutional Affairs References Committee, Parliament of Australia, Matter of a Popular Vote, in the Form of a Plebiscite or Referendum, on the Matter of Marriage in Australia (3 September 2015) 5–6 (‘Submission No 11’).

[90] See, eg, Kirby, ‘Skin Curling’ (n 32) 5–7; Kirby, ‘The Tale of Three Cities’ (n 32) 33; Kirby, ‘250 Years of the Crown’ (n 12) 528; Henskens (n 18) 150, 159–62; Twomey, ‘Plebiscites and Referenda’ (n 9) 835; Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 298.

[91] Sawer (n 70) 179.

[92] Ibid.

[93] See Maurice Blackburn, The Conscription Referendum of 1916 (Anti-Conscription Celebration League, 1936) 6.

[94] Re United Nations Convention on the Rights of the Child (Incorporation) (Scotland) Bill [2021] UKSC 42; [2021] 1 WLR 5106 (‘UNCRC Case’).

[95] Ibid 5116 [31], 5122 [52] (Lord Reed PSC, Lord Hodge DPSC and Lords Lloyd-Jones, Sales and Stephens JJSC agreeing).

[96] Ibid 5122 [52].

[97] Ibid 5122–3 [51]–[55].

[98] Sawer (n 70) 179.

[99] Scotland Act 1998 (UK) s 63A(3).

[100] [2005] UKHL 56; [2006] 1 AC 262, 319 [163]–[164] (‘Jackson’).

[101] David Feldman, ‘Legislation Which Bears No Law’ (2016) 37(3) Statute Law Review 212, 216; David Feldman, ‘Legislation as Aspiration: Statutory Expression of Policy Goals’ (Speech, Statute Law Society, 16 March 2015) 6.

[102] Flags Amendment Act 1998 (Cth) sch 1 item 1.

[103] Simon Evans, ‘Why Is the Constitution Binding? Authority, Obligation and the Role of the People’ [2004] AdelLawRw 6; (2004) 25(1) Adelaide Law Review 103, 120.

[104] AV Dicey, ‘Ought the Referendum To Be Introduced into England?’ (1890) 57 (January–June) Contemporary Review 489, 499, 510.

[105] Ibid 499.

[106] Ibid 510–11, where it is observed that ‘[t]he principle of the [plebiscite] ... is to place, at any rate as regards important legislation, parties, factions, and sections under the control of the national majority’.

[107] The crisis had resulted in a number of government Bills being rejected by the Lords, including money bills: see Jackson (n 100) 274–9 [9]–[20] (Lord Bingham).

[108] See Lucy Atkinson, Andrew Blick and Matt Qvortrup, The Referendum in Britain: A History (Oxford University Press, 2020) 46–8.

[109] Tomn (n 21) 242.

[110] Ibid 243.

[111] Aroney, The Constitution of a Federal Commonwealth (n 22) 240–4.

[112] An early exposition of the importance of political modes of accountability is given in Robert Randolph Garran, The Coming Commonwealth: An Australian Handbook of Federal Government (Angus and Robertson, 1897) 149.

[113] Regarding the legally non-binding referendum in the UK, on entry into the European Communities in 1975, Vernon Bogdanor observes that

Edward Short, then Leader of the House of Commons, insisted to the House that ‘This referendum is wholly consistent with parliamentary sovereignty. The Government will be bound by its result, but Parliament, of course, cannot be bound’. He then added, ‘Although one would not expect honourable members to go against the wishes of the people, they will remain free to do so’.

Vernon Bogdanor, The New British Constitution (Hart Publishing, 2009) 95.

[114] Tierney (n 15) 12–15. See also Martin Loughlin, Foundations of Public Law (Oxford University Press, 2010) 221–7. For a summary of the Australian understanding of ‘the people’ as a tenet of constitutional authority, see generally Brendan Lim, ‘Legitimacy’ in Cheryl Saunders and Adrienne Stone (eds), The Oxford Handbook of the Australian Constitution (Oxford University Press, 2018) 315.

[115] Kirmani v Captain Cook Cruises Pty Ltd [No 1] (1985) 159 CLR 351, 383 (Murphy J);

R v Smithers; Ex parte Benson [1912] HCA 96; (1912) 16 CLR 99, 108–9 (Griffith CJ), 109–10 (Barton J);

Leeth v Commonwealth (1992) 174 CLR 455, 484 (Deane and Toohey JJ); Nationwide (n 3) 74 (Deane and Toohey JJ); ACTV (n 3) 138 (Mason CJ); Theophanous (n 3) 172–3 (Deane J); McGinty (n 1) 230 (McHugh J). See also French, ‘The Constitution and the People’ (n 4) 71, 76–80; Patrick Keyser, Open Constitutional Courts (Federation Press, 2010) 140–4; Leighton McDonald, ‘The Denizens of Democracy: The High Court and the “Free Speech” Cases’ (1994) 5(3) Public Law Review 160, 178; George Winterton, ‘Popular Sovereignty and Constitutional Continuity’ [1998] FedLawRw 1; (1998) 26(1) Federal Law Review 1, 1, 3–4.

[116] See, eg, McGinty (n 1) 230 (McHugh J).

[117] Vernon Bogdanor, The People and the Party System: The Referendum and Electoral Reform in British Politics (Cambridge University Press, 1981) 93.

[118] McGinty (n 1) 223 (Gaudron J), 230 (McHugh J); Dietrich v The Queen [1992] HCA 57; (1992) 177 CLR 292, 319 (Brennan J); French, ‘The Constitution and the People’ (n 4) 76–7; Twomey,

Rowe v Electoral Commissioner: Evolution or Creationism?’ (n 5) 183, 192.

[119] Alexander Hamilton, ‘Federalist No 78’ in Alexander Hamilton, James Madison and John Jay, The Federalist Papers, ed Michael A Genovese (Palgrave Macmillan, 2009) 235, 236, quoted in Murphy (n 2) 90 [188] (Keane J).

[120] Joseph Jaconelli, ‘Do Constitutional Conventions Bind?’ (2005) 64(1) Cambridge Law Journal 149, 153.

[121] See, eg, Tham (n 1) 979, 1005; Susan Kiefel, ‘Standards of Review in Constitutional Review of Legislation’ in Cheryl Saunders and Adrienne Stone (eds), The Oxford Handbook of the Australian Constitution (Oxford University Press, 2018) 488, 499–500; Lange (n 3) 561, 567, 575 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ).

[122] See above n 4.

[123] See Arcioni (n 4) 426–7; French, ‘The Constitution and the People’ (n 4) 75–6.

[124] See above nn 2–3.

[125] Appleby (n 67) 233–4.

[126] Tham (n 1) 980, 1009–10.

[127] Ibid 991–3.

[128] See ibid 979, 1005; Kiefel (n 121) 488, 499–500.

[129] On the longstanding test of ‘reasonably appropriate and adapted’, see Adrienne Stone, ‘Proportionality and Its Alternatives’ (2020) 48(1) Federal Law Review 123, 124, 126–8.

[130] McCloy (n 3) 193–6 [2]–[4] (French CJ, Kiefel, Bell and Keane JJ). But see LibertyWorks

Inc v Commonwealth [2021] HCA 18; (2021) 391 ALR 188, 255–6 [249] (Steward J) (‘LibertyWorks Inc’), which doubted whether the freedom of political communication existed at all. See also Kiefel (n 121) 501–7; Shipra Chordia, ‘The Trajectory of Structured Proportionality in Implied Freedom of

Political Communication Cases: Brown v Tasmania’, Australian Public Law (Blog

Post, 2 November 2017) <>, archived at <>.

[131] Kiefel (n 121) 500. Such has also been the approach in the UK courts in cases of this type: R (Miller) v Prime Minister [2019] UKSC 41; [2020] AC 373, 406–7 [49] (Baroness Hale PSC and Lord Reed DPSC for the Court) (‘Miller II’).

[132] Murphy (n 2) 51 [34] (French CJ and Bell J), where it is said that Roach (n 2) and Rowe (n 1) ‘did not set up a sui generis test of validity’. See also Murphy (n 2) 50 [33], 52 [36], where French CJ and Bell J spoke of the ‘general rubric of proportionality long applied in Australian public law’. However, the precise form of that rubric (structured or otherwise) might differ depending on the case: at 53 [38].

[133] See, eg, ACTV (n 3) 143–4 (Mason CJ), 186–7 (Dawson J), 234–5 (McHugh J). See also Kiefel (n 121) 501–5; Stone (n 129) 135.

[134] See ACTV (n 3) 159 (Brennan J); Theophanous (n 3) 156 (Brennan J). Occasionally the term ‘deference’ is used. But it has been held that the proper inquiry is not as to appropriate ‘deference’, but as to the limits of the competence of the courts, as against that of Parliament and the executive: McCloy (n 3) 220 [91] (French CJ, Kiefel, Bell and Keane JJ).

[135] Tham (n 1) 981, 984.

[136] Chordia (n 130).

[137] Peter Gay, The Enlightenment: An Interpretation (Alfred A Knopf, 1975) 139.

[138] Walter Bagehot, The English Constitution, ed Miles Taylor (Oxford University Press, 2001) 5–11 (emphasis omitted).

[139] Ibid.

[140] See generally Robin Archer et al (eds), The Conscription Conflict and the Great War (Monash University Publishing, 2016); KS Inglis, ‘Conscription in Peace and War, 1911–1945’ in Roy Forward and Bob Reece (eds), Conscription in Australia (University of Queensland Press, 1968) 22; Jeff Kildea, ‘Australian Catholics and Conscription in the Great War’ (2002) 26(3) Journal of Religious History 298.

[141] For a brief discussion of the two plebiscites, see Williams and Hume (n 18) 6–7. See also Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 293.

[142] On the moral basis of this objection, see DH Monro, ‘Civil Rights and Conscription’ in Roy Forward and Bob Reece (eds), Conscription in Australia (University of Queensland Press, 1968) 1, 20–1. See generally Robin Archer, ‘Labour and Liberty: The Origins of the Conscription Referendum’ in Robin Archer et al (eds), The Conscription Conflict and the Great War (Monash University Publishing, 2016) 37 (‘Labour and Liberty’). Opposition was particularly strong amongst members of the working class and urban Catholic communities. The Catholic Archbishop Daniel Mannix, in Melbourne, was a prominent leader of the anti-conscription movement: Inglis (n 140) 36–9, 42.

[143] Inglis (n 140) 24–5; Archer, ‘Labour and Liberty’ (n 142) 38.

[144] Inglis (n 140) 23, discussing Defence Act 1909 (Cth) s 125.

[145] Defence Act 1903 (Cth) s 49, as enacted.

[146] Lundie and McCann (n 36) 10 n 44, 20.

[147] Section 4(1) of the War Precautions Act 1914 (Cth) provided:

The Governor-General may make regulations for securing the public safety and the defence of [Australia], and for conferring such powers and imposing such duties as he thinks fit ... upon the ... members of the ... Military Forces of [Australia].

Section 5 made similar provision, in respect of orders in council.

[148] Ibid ss 4(1), 5.

[149] Defence Act 1903 (Cth) s 49.

[150] Leslie C Jauncey, The Story of Conscription in Australia (George Allen & Unwin, 1935) 109–10.

[151] Inglis (n 140) 33.

[152] Ibid. This was the Military Service Act 1916 (NZ).

[153] Inglis (n 140) 34.

[154] See Archer, ‘Labour and Liberty’ (n 142) 39.

[155] Ibid 37.

[156] Inglis (n 140) 34.

[157] Where Labor held 31 of 36 seats: Archer, ‘Labour and Liberty’ (n 142) 37.

[158] See also Twomey, Submission No 6 (n 21) 2–3 on the use of a referendum on

same-sex marriage.

[159] Robin Archer, ‘Stopping War and Stopping Conscription: Australian Labour’s Response to World War I in Comparative Perspective’ (2014) 106 (May) Labour History 43, 58 (‘Stopping War and Stopping Conscription’).

[160] Inglis (n 140) 34.

[161] Ibid 33–4.

[162] See, eg, Commonwealth, Parliamentary Debates, House of Representatives,

20 September 1916, 8663 (Hugh Sinclair). Sinclair was a former Anti-Socialist Party MP and then Nationalist MP in Queensland: see, eg, ‘Mr Sinclair’s Campaign’, The Brisbane Courier (Brisbane, 26 November 1906) 5; ‘Reversal at Moreton’, The Daily Telegraph (Sydney, 2 June 1917) 12.

[163] Inglis (n 140) 34.

[164] Archer, ‘Labour and Liberty’ (n 142) 40.

[165] Ernest Scott, Australia during the War (Angus and Robertson, 3rd ed, 1938) 340. Strict party voting discipline has historically been observed in the Labor Party: see Simpson (n 19) 577–8.

[166] Military Service Referendum Act 1916 (Cth) s 4(1).

[167] Ibid s 5. In its full length, the question was:

Are you in favour of the Government having, in this grave emergency, the same compulsory powers over citizens in regard to requiring their military service, for the term of this War, outside the Commonwealth, as it now has in regard to military service within the Commonwealth?

[168] Maurice Blackburn, a member of the Labor Party and staunch opponent of conscription, said that a ‘yes’ plebiscite result would have constituted a ‘mandate which only a handful of Parliamentarians would have dared to disobey’: Blackburn (n 93) 6.

[169] Military Service Referendum Act 1916 (Cth) s 7(a).

[170] Ibid s 9.

[171] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 305.

[172] See Referendum (Constitution Alteration) Act 1906 (Cth) s 4(1). The franchise excluded members of ‘unlawful associations’: Jennifer Norberry and George Williams, ‘Voters and the Franchise: The Federal Story’ (Research Paper No 17 2001–02, Parliamentary Library, Parliament of Australia, 28 May 2002) 25. The Commonwealth Franchise Act 1902 (Cth) also excluded ‘those of unsound mind, persons convicted and under sentence for any offence punishable by imprisonment of one year or longer, and ... “aboriginal natives of Australia[,] Asia[,] Africa or the Islands of the Pacific (except New Zealand)”’: s 4, as quoted in Twomey, ‘The Federal Constitutional Right To Vote in Australia’ (n 1) 131.

[173] Inglis (n 140) 35.

[174] On the contours of ‘the people’, as electors or otherwise, under the Australian Constitution, see generally Arcioni (n 4).

[175] Military Service Referendum Act 1916 (Cth) s 7.

[176] Had the vote taken place the year before, it is also possible that it might have been subject to the requirement of compulsory voting: Orr, ‘Old Wine, New Bottle’ (n 32) 359–60; Compulsory Voting Act 1915 (Cth) ss 3–4.

[177] See generally Anne Twomey, ‘Compulsory Voting in a Representative Democracy: Choice, Compulsion and the Maximisation of Participation in Australian Elections’ (2013) 13(2) Oxford University Commonwealth Law Journal 283; Lisa Hill, ‘Compulsory Voting: Residual Problems and Potential Solutions’ (2002) 37(3) Australian Journal of Political Science 437. More recent decisions of the High Court have understood the compulsory requirement to vote to be an important aspect of the maintenance of Australia’s system of representative government: see, eg, Roach (n 2) 173 [5] (Gleeson CJ).

[178] Parliamentary Handbook (n 10) 445. See also Jeff Kildea, ‘Australian Catholics and Conscription in the Great War’ (n 140) 298 n 1.

[179] John Connor, ‘Why Was It Easier To Introduce and Implement Conscription in Some English-Language Speaking Countries than in Others?’ in Robin Archer et al (eds), The Conscription Conflict and the Great War (Monash University Publishing, 2016) 148, 158.

[180] In observing these matters, the New Zealand Defence Minister, James Allan, wrote to the New Zealand Prime Minister, William Massey (where conscription had been introduced by a parliamentary vote alone), and said of the plebiscite ‘[t]hank goodness we have avoided that blunder’: ibid 158.

[181] Inglis (n 140) 42.

[182] Archer, ‘Stopping War and Stopping Conscription’ (n 159) 59.

[183] The power was found in the War Precautions Act 1914 (Cth) s 4(1).

[184] See Scott (n 165) 340.

[185] Commonwealth, Parliamentary Debates, House of Representatives, 26 September 1917, 2852, 2856, 2859 (William Hughes, Prime Minister). The Senate was adjourned the following day on the motion of a Nationalist Senator: Commonwealth, Parliamentary Debates, Senate,

27 September 1917, 2871 (Edward Davis Millen).

[186] Catts (n 36) 7–8.

[187] Ibid 8.

[188] Ibid.

[189] The question was: ‘Are you in favour of the proposal of the Commonwealth Government for reinforcing the Australian Imperial Force overseas?’: War Precautions (Military Service Referendum) Regulations 1917 (Cth) reg 6.

[190] War Precautions Act 1914 (Cth) s 4(1).

[191] War Precautions (Military Service Referendum) Regulations 1917 (Cth) regs 5(1), 6.

[192] Governor-General, Writ for a Referendum (NAA: Series No A11803, Control Symbol 1918/89/94, Item Barcode 351376, 10 November 1917).

[193] Parliamentary Handbook (n 10) 446.

[194] See Catts (n 36) 23–4.

[195] Parliamentary Handbook (n 10) 446.

[196] See Inglis (n 140) 41, 43.

[197] Jauncey (n 150) 335; Scott (n 165) 446. See also Archer, ‘Stopping War and Stopping Conscription’ (n 159) 59.

[198] Inglis (n 140) 43.

[199] National Archives of Australia, Australia’s National Anthem (Fact Sheet No 251, May 2020) 1 <>, archived at <> (‘Australia’s National Anthem’).

[200] Ibid; Parliamentary Handbook (n 10) 447.

[201] Australia’s National Anthem (n 199) 2.

[202] See Cabinet Minute, 4 May 1976 (National Archives of Australia, NAA: Series No A10756, Control Symbol LC91 PART 1, Item Barcode 7828879) 98–9 (‘Decision No 546’).

[203] Australia’s National Anthem (n 199) 2.

[204] Commonwealth, Parliamentary Debates, House of Representatives, 2 June 1976, 2907 (Malcolm Fraser, Prime Minister).

[205] Commonwealth, Parliamentary Debates, House of Representatives, 4 May 1976, 1842–3, 1848 (Malcolm Fraser, Prime Minister).

[206] Decision No 546 (n 202) 98. Only the Victorian Premier supported the use of ‘Waltzing Matilda’.

[207] Ibid.

[208] Letter from Attorney-General’s Department to Malcolm Fraser, 11 February 1977 (National Archives of Australia, NAA: Series No A1209, Control Symbol 1978/1103 PART 4, Barcode 8843997) 77 (‘Letter from Attorney-General’s Department’).

[209] Ibid; Australia’s National Anthem (n 199) 2.

[210] Letter from Attorney-General’s Department (n 208) 77–8.

[211] Commonwealth, Gazette: Special, No S 142, 19 April 1984. ‘Advance Australia Fair’ received 43.3% of the vote; ‘Waltzing Matilda’, 28.3%; ‘God Save the Queen’, 18.8%; and ‘Song of Australia’, 9.7%: Parliamentary Handbook (n 10) 447.

[212] Letter from Attorney-General’s Department (n 208).

[213] See ibid; Department of Parliamentary Services (Cth), Bills Digest (n 58) 4 n 19.

[214] Letter from Attorney-General’s Department (n 208) 77 (emphasis added).

[215] Referendum (Constitution Alteration) Modification Act 1977 (Cth).

[216] Ibid s 3.

[217] Parliamentary Handbook (n 10) 435–7, 447.

[218] See Australia’s National Anthem (n 199) 1–2.

[219] ‘[I]f in a majority of the States a majority of the electors voting approve the proposed law, and if a majority of all the electors voting also approve the proposed law, it shall be presented to the Governor-General for the Queen’s assent’: Constitution s 128.

[220] Cf Flags Act (n 16) s 3(2), which provides that the flag may only be changed if supported by a plebiscite, but that the results are counted on a national basis (unlike a s 128 referendum). In respect of subsequent changes to the national song, see Justice François Kunc, ‘Options for the Voice to Parliament Released’ (2021) 95(3) Australian Law Journal 163, 164–5.

[221] See generally Alex Greenwich and Shirleene Robinson, Yes Yes Yes: Australia’s Journey to Marriage Equality (NewSouth, 2018); Kirby, ‘The Tale of Three Cities’ (n 32); Kirby, ‘Skin Curling’ (n 32) 2–18.

[222] Commonwealth v Australian Capital Territory [2013] HCA 55; (2013) 250 CLR 441, 452 [1]–[2] (French CJ, Hayne, Crennan, Kiefel, Bell and Keane JJ).

[223] Mary Anne Neilsen, ‘Same-Sex Marriage’ (Briefing Book, Parliamentary Library, Parliament of Australia, December 2013) 130–1.

[224] ‘Senior Australian MP Calls for Gay Marriage Referendum’ (n 77).

[225] Daniel Hurst, ‘Tony Abbott Says “It’s OK To Be Gay” but Insists Same-Sex Marriage Is Not OK’, The Guardian (online, 24 August 2015) <>, archived at <>.

[226] Joint Doorstop Interview with Tony Abbott (12 August 2015) <>, archived at <> (emphasis added) (‘Interview with Tony Abbott’).

[227] Martin Lumb, ‘Composition of the 44th Parliament’ (Briefing Book, Parliamentary Library, Parliament of Australia, December 2013) 4–5.

[228] Senate Legal and Constitutional Affairs References Committee (n 71) 1 [1.1].

[229] Ibid 31 [4.1].

[230] Williams, Submission No 32 (n 71) 3.

[231] Lindell, Submission No 4 (n 12) 2–4 [5]–[11].

[232] Walker (n 89) 3.

[233] Dixon, Kildea and Lynch (n 89) 5–6.

[234] Senate Legal and Constitutional Affairs References Committee (n 71) 31 [4.1].

[235] Twomey, ‘Let’s Not Demonise Plebiscite’ (n 29) 28. However, that view was partially reversed after the conduct of the later survey: Twomey, ‘Wilkie v Commonwealth: A Retreat to Combet’ (n 31).

[236] Paul Karp, ‘Peta Credlin Says Marriage Equality Plebiscite Bill May Fail’, The Guardian (online, 28 June 2016) <>, archived at <>.

[237] The government still did not have control over the Senate: Hannah Gobbett, ‘Composition of the 45th Parliament: A Quick Guide’ (Research Paper Series 2016–17, Parliamentary Library, Parliament of Australia, 29 August 2016) 2.

[238] Australian Marriage Equality Ltd, ‘Plaintiffs’ Submissions’, Submission in Australian Marriage Equality Ltd v Minister for Finance, M106/2017, 23 August 2017, 3–4 [11]–[22].

[239] Doorstop Interview with Peter Dutton (23 March 2017) <,-Parliament-House.aspx>, archived at <>.

[240] Commonwealth, Budget 2017–18: Budget Strategy and Outlook (Budget Paper No 1, 9 May 2017) 9-5, 9-11.

[241] See, eg, Wilkie (n 33) 514–15 [25]–[28] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ). See also discussion of the Finance Minister’s affidavit at 543–4 [135]–[137]

(Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ).

[242] Richardson, Emmett and Palaniappan (n 64) 1 [2], 6 [28]–[29]; Koziol (n 64); Paul Kildea, ‘Using the ABS’ (n 31); Williams, ‘Dual Citizens and the Postal Survey’ (n 64).

[243] Wilkie (n 33) 514 [25]–[28] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ).

[244] The direction was made by the Treasurer: see Treasurer (Cth), Census and Statistics (Statistical Information) Direction 2017 (9 August 2017).

[245] Ibid s 2.

[246] Australian Bureau of Statistics, ‘Australian Marriage Law Postal Survey, 2017’ (Catalogue No 1800.0, 15 November 2017) <>, archived at <>.

[247] Ibid.

[248] Ibid.

[249] ‘Same-Sex Marriage Bill Introduced to Parliament, with Senators Set To Debate

Religious Exemptions’, ABC News (online, 15 November 2017) <>, archived at <>.

[250] Marriage Amendment (Definition and Religious Freedoms) Act 2017 (Cth) sch 1 item 3.

[251] Minister for Finance (Cth), Advance to the Finance Minister Determination (No 1 of 2017–2018) (F2017L01005, 9 August 2017) (‘Advance to the Finance Minister Determination’).

[252] Appropriation Act (No 1) 2017–2018 (Cth) s 10(1)(b).

[253] Advance to the Finance Minister Determination (n 251).

[254] See, eg, Legislation Act 2003 (Cth) s 42.

[255] For a background to the case, and description of the litigants, see Greenwich and Robinson

(n 221) 199–203.

[256] Wilkie (n 33) 547 [151] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ).

[257] See ibid 493–501.

[258] Ibid.

[259] Ibid 497.

[260] Transcript of Proceedings, Wilkie v Commonwealth [2017] HCATrans 174, 1660–2245

(Kiefel CJ, Edelman, Gordon, Keane, Bell, Gageler and Nettle JJ and Ms Foley).

[261] Wilkie (n 33) 545 [142]–[143] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and

Edelman JJ).

[262] Transcript of Proceedings, Wilkie v Commonwealth [2017] HCATrans 175, 4425, 4717

(Mr Donaghue QC).

[263] Wilkie (n 33) 524–32 [64]–[95] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and

Edelman JJ). See especially at 530–1 [88]–[89].

[264] See ibid 544–6 [139]–[148].

[265] See, eg, Twomey, ‘Wilkie v Commonwealth: A Retreat to Combet’ (n 31); Twomey, ‘A Tale of Two Cases’ (n 31) 17–19; Brown (n 31) 522–7.

[266] In 1917, Parliament did not vote on the proposal, as it had been adjourned; and in 1977, a Bill providing for the poll was, quite deliberately, never introduced into Parliament: Letter from Attorney-General’s Department (n 208) 77.

[267] The survey has been understood as a ‘plebiscite’ or ‘vote’ by various actors: see, eg, ‘Same-Sex Marriage Bill Introduced to Parliament, with Senators Set To Debate Religious Exemptions’

(n 249). Justice Hamill of the Supreme Court of New South Wales, in dealing with proposals for celebrations on ‘yes’ day, used similar language: Commissioner of Police (NSW) v Marshall (Reclaim the Streets) [2017] NSWSC 1589, [1]. In Explanatory Statement, Minister of Finance (Cth), Advance to the Finance Minister Determination (No 1 of 2017–2018) (F2017L01005,

9 August 2017) 1, the Minister for Finance explained that the funds were being made available to fund a ‘postal plebiscite’.

[268] Wilkie (n 33) 545 [142] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ).

[269] Dicey (n 104) 510–11.

[270] Tomn (n 21) 242.

[271] See, eg, Aroney, The Constitution of a Federal Commonwealth (n 22) 240–4.

[272] Referendum (Constitution Alteration) Modification Act 1977 (Cth) s 3.

[273] Letter from Attorney-General’s Department (n 208) 77.

[274] Though the government later introduced legislation formalising aspects of the survey’s conduct, but only after the High Court’s decision on the legality of the survey’s funding: see Marriage Law Survey (Additional Safeguards) Act 2017 (Cth) s 3.

[275] Parliamentary Debates (22 August 2011) (n 72) 8729. See also Interview with Tony Abbott

(n 226).

[276] Atkinson, Blick and Qvortrup (n 108) 46–8.

[277] Bagehot, The English Constitution (n 138) 7 (emphasis omitted).

[278] Atkinson, Blick and Qvortrup (n 108) 46–8.

[279] Sawer (n 70) 179.

[280] A deadlock-breaking referendum procedure does apply to laws proposed under s 128. This procedure has only been used on four occasions (all in 1974), in respect of government Bills for the amendment of the Constitution: Jack Richardson, ‘Resolving Deadlocks in the Australian Parliament’ (Research Paper No 9 2000–01, Parliamentary Library, Parliament of Australia, 31 October 2000) 29 n 74.

[281] See, eg, Catts (n 36) 5–6.

[282] Constitutional Commission (n 47) 868–9 [13.75]–[13.80].

[283] Senate Legal and Constitutional Affairs References Committee (n 71) 31 [4.1]. The use, and potential abuse, of the Finance Minister’s powers under s 10 of the Appropriation Act (No 1) 2017–2018 (Cth) has also been the subject of a recent parliamentary research paper: Glenn Ryall, ‘Wilkie v Commonwealth and Parliamentary Control of Appropriations’ (Papers on Parliament No 70, Procedure and Research Section, Department of the Senate, Parliament of Australia, December 2019) 78, 86–8, 91, 95–6.

[284] Lindell, Submission No 4 (n 12) 3 [9].

[285] Kiefel (n 121) 500. See generally Tham (n 1).

[286] Simpson (n 19) 574.

[287] Ibid.

[288] See Harry Evans and Rosemary Laing (eds), Odgers’ Australian Senate Practice (Department of the Senate, 14th rev ed, 2016) 28–9.

[289] Constitution s 61 and the other specific powers of the Governor-General: see, eg, s 5. See generally Ruddock v Vadarlis [2001] FCA 1329; (2001) 110 FCR 491, 537–9 [176]–[179] (French J), where it is described how s 61 not only gives expression to those powers granted to the executive by statute, but also contains that residual power (inherited at the time of Federation) that might be called the ‘prerogative’ power.

[290] For a discussion of recent controversies, and an overview of the contours of the debate as to the content and scope of the executive power, see generally Robert French, ‘Executive Power in Australia: Nurtured and Bound in Anxiety’ [2018] UWALawRw 17; (2018) 43(2) University of Western Australia Law Review 16; Amanda Sapienza, ‘Chief Justice French on Non-Statutory Executive Power: A Timely Reflection’, Australian Public Law (Blog Post, 20 December 2016) <>, archived at <>; Catherine Dale Greentree, ‘The Commonwealth Executive Power: Historical Constitutional Origins and the Future of the Prerogative’ [2020] UNSWLawJl 32; (2020) 43(3) University of New South Wales Law Journal 893.

[291] Simpson (n 19) 574. See also Appleby (n 67) 214–15.

[292] Dicey (n 104) 498–500; Tomn (n 21) 242–3. See also discussion of Dicey in Anne Twomey, ‘Dicey on Brexit and the Conservative Nature of Referendums’ (2019) 93(12) Australian Law Journal 981, 982.

[293] See, eg, European Union Referendum Act 2015 (UK) s 1 (in respect of the ‘Brexit’ vote); Scottish Independence Referendum Act 2013 (Scot) s 1 (concerning independence from the UK); Referendums (Scotland and Wales) Act 1997 (UK) ss 1–2 (concerning referendums in Scotland and Wales on devolution); Scotland Act 1978 (UK) ss 83, 85 (which provided for a referendum to approve proposals for devolution); Wales Act 1978 (UK) s 80 (which provided for a referendum to approve proposals for devolution); Referendum Act 1975 (UK) s 1 (in respect of ‘remain[ing in] ... the European Economic Community’). In addition, the conduct of referendums is now regulated by the Political Parties, Elections and Referendums

Act 2000 (UK).

[294] See Referendum (Constitution Alteration) Modification Act 1977 (Cth) s 3; Marriage Law Survey (Additional Safeguards) Act 2017 (Cth).

[295] ‘Diceyan’ refers to the customary understanding of Dicey as favouring parliamentary sovereignty. His own views on the referendum were not always consistent: see Rivka Weill, ‘Dicey Was Not Diceyan’ (2003) 62(2) Cambridge Law Journal 474, 474, 493.

[296] Twomey, ‘Wilkie v Commonwealth: A Retreat to Combet’ (n 31). As in Australia, parliamentary oversight of public expenditure is an old and important aspect of the UK settlement: Auckland Harbour Board (n 59) 326–7 (Viscount Haldane for Viscount Haldane, Lords Sumner, Parmoor and Wrenbury and Sir Charles Darling).

[297] [1995] UKHL 3; [1995] 2 AC 513, 513.

[298] Miller II (n 131) 406 [46], 411–12 [69] (Baroness Hale PSC and Lord Reed DPSC for

the Court).

[299] Appropriation Act (No 1) 2017–2018 (Cth) s 10.

[300] But that argument, again, relies on identifying some intention from the act of negativing a legislative proposal.

[301] Wilkie (n 33).

[302] Williams, ‘Dual Citizens and the Postal Survey’ (n 64).

[303] See Gleeson (n 5) 10–11.

[304] The High Court has long held that even those textual aspects of the Constitution which regulate the intramural mechanisms of the Parliament are not to be the subject of judicial intervention: see, eg, Wilkie (n 33) 523–4 [63] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and

Edelman JJ).

[305] See, eg, Mark Elliott, ‘Constitutional Adjudication and Constitutional Politics in the United Kingdom: The Miller II Case in Legal and Political Context’ (2020) 16(4) European Constitutional Law Review 625, 628–30.

[306] Appleby (n 67) 217–19.

[307] Egan v Willis [1998] HCA 71; (1998) 195 CLR 424, 449 [36], 451–2 [42], 453 [45] (Gaudron, Gummow and Hayne JJ), 473–4 [96]–[97] (McHugh J), 500 [150] (Kirby J) (‘Egan’); R v Toohey; Ex parte Northern Land Council [1981] HCA 74; (1981) 151 CLR 170, 184, 192 (Gibbs CJ), 236 (Aickin J).

[308] George Winterton, Parliament, the Executive and the Governor-General: A Constitutional Analysis (Melbourne University Press, 1983) 124–39; Geoffrey Lindell, Responsible Government and the Australian Constitution: Conventions Transformed into Law? (Federation Press, 2004) 3; Leslie Zines, The High Court and the Constitution (Federation

Press, 5th ed, 2008) 340.

[309] Gabrielle Appleby and Joanna Howe, ‘Scrutinising Parliament’s Scrutiny of Delegated Legislative Power’ (2015) 15(1) Oxford University Commonwealth Law Journal 3, 5, 10–11.

[310] Wilkie (n 33) 521–2 [56] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ). No finding on standing was made because the plaintiffs lost on the substantive case: at 547 [151].

[311] Brown (n 31) 518–22.

[312] See Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 301.

[313] Letter from Attorney-General’s Department (n 208) 77.

[314] Richardson, Emmett and Palaniappan (n 64) 1 [2], 6 [28]; Koziol (n 64).

[315] Census and Statistics Act 1905 (Cth) s 9(1). It was a ‘long held’ view that s 81 of the Constitution gave executive government the power to spend money appropriated from the consolidated revenue fund: James Stellios, Zines’s the High Court and the Constitution (Federation Press, 6th ed, 2015) 384. In the 2009 case of Pape (n 54), the High Court held that no such general power derived from s 81, and a further lawful basis was required: at 55 [111] (French CJ), 78 [197] (Gummow, Crennan and Bell JJ), 105 [296] (Hayne and Kiefel JJ), 211 [602] (Heydon J). This was further qualified in Williams [No 1] (n 61) 224 [114] (Gummow and Bell JJ), 245–6 [187]–[188], 248 [191] (Hayne J), 341 [478] (Crennan J); Williams [No 2] (n 62) 453 [20], 455 [25] (French CJ, Hayne, Kiefel, Bell and Keane JJ).

[316] Twomey, ‘Post-Williams Expenditure’ (n 63) 9–10.

[317] Kirby, ‘Skin Curling’ (n 32) 6–7; Twomey, ‘Wilkie v Commonwealth: A Retreat to Combet’

(n 31); Twomey, ‘A Tale of Two Cases’ (n 31) 17–19; Brown (n 31) 530.

[318] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 295.

[319] Bogdanor, Beyond Brexit: Britain’s Unprotected Constitution (n 81) ix, where Clement Attlee is quoted as saying that the referendum was ‘alien to all our traditions’.

[320] Cf Wilkie (n 33) 540–2 [123]–[128] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and Edelman JJ), where arguments of a related type, in respect of appropriations, were rejected on justiciability grounds.

[321] See Lange (n 3) 559–60 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ).

[322] See Miller II (n 131) 407 [50], 408–9 [55]–[56], 410 [61] (Baroness Hale PSC and Lord Reed DPSC for the Court).

[323] See Wilkie (n 33) 523 [60]–[61] (Kiefel CJ, Bell, Gageler, Keane, Nettle, Gordon and

Edelman JJ).

[324] In respect of Parliament’s relationship with, and power over, its Members, see generally Egan (n 307), which may be a departure point.

[325] See, eg, Referendum Bill 1898 (NZ); Referendum Bill 1900 (NZ); Referendum Bill 1901 (NZ); Referendum Bill 1902 (NZ); Referendum Bill 1903 (NZ); Referendum Bill 1904 (NZ); Referendum Bill 1905 (NZ); House of Lords Reform Bill 1922 (UK) s 10(5); Referendums Criteria Bill 2019 (UK); Referendums Criteria Bill 2020 (UK). See also United Kingdom, Parliamentary Debates, House of Lords, 29 March 1911, vol 7, cols 721–2 (Earl Beauchamp), col 755 (Earl of Ancaster), discussing Reference to the People Bill 1911 (UK); United Kingdom, Parliamentary Debates, House of Commons, 5 July 2010, vol 513, col 80; Atkinson, Blick and Qvortrup (n 108) 47–8.

[326] For a criticism of entrenching democratic processes, see Twomey, ‘Rowe v Electoral Commissioner: Evolution or Creationism?’ (n 5) 185, 189.

[327] See generally Simpson (n 19).

[328] [1926] HCA 58; (1926) 38 CLR 153, 178.

[329] See Chordia (n 130).

[330] Lange (n 3) 557–9 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ).

[331] Ibid 577–8. The Court observed that s 7 of the Constitution provides for a Senate, chosen by the people of the various states, and s 24 provides for a House of Representatives chosen by the people of the Commonwealth, where those chambers are imbued with legislative competence. Other provisions relating to elections were also considered.

[332] AH Birch, Representative and Responsible Government: An Essay on the British Constitution (George Allen & Unwin, 1964) 17.

[333] Lange (n 3) 559 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ) (citations omitted).

[334] Birch (n 332) 13 (emphasis added), quoted in Simpson (n 19) 566.

[335] Lange (n 3) 559, 561, 567, 571, 575 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ).

[336] Ibid 567.

[337] Ibid 559–60. Political discussion, protest, and ‘free political communication’ are important, indirect forms of popular participation in government. There remains an open question whether truly non-binding plebiscites (ie those which bind neither legally nor politically) are also exercises in ‘indirect’ participation.

[338] Lange (n 3) 557 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ).

[339] On the nature of representative government, see, eg, Roach (n 2) 173–4 [5], [7] (Gleeson CJ), 186 [44]–[45], 198–9 [79]–[80], [83], 202 [95] (Gummow, Kirby and Crennan JJ); Rowe (n 1) 18–21 [18]–[22], [25], 38–9 [77] (French CJ). On the interaction of representative government with political communication, see, eg, Nationwide (n 3) 74 (Deane and Toohey JJ); ACTV

(n 3) 143–4 (Mason CJ), 186–7 (Dawson J), 234–5 (McHugh J); Theophanous (n 4) 120–3, 128, 133–4 (Mason CJ, Toohey and Gaudron JJ), 149 (Brennan J), 163, 168 (Deane J), 190

(Dawson J); Coleman (n 3) 48–53 [88]–[100] (McHugh J); McCloy (n 3) 193–4 [2] (French CJ, Kiefel, Bell and Keane JJ); Murphy (n 2) 50 [33] (French CJ and Bell J), 61 [63] (Kiefel J), 67 [85] (Gageler J), 86–7 [178] (Keane J); Brown v Tasmania (n 3) 359 [88] (Kiefel CJ, Bell and Keane JJ), 375–6 [156], 377–8 [162] (Gageler J).

[340] See Gleeson (n 5) 7, 9.

[341] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 295.

[342] See Appleby and Lim (n 5) 223–4, 241; McKinlay (n 2) 56 (Stephen J).

[343] Gleeson (n 5) 7.

[344] Ibid. See also Mulholland (n 5) 189–91 [9]–[10], [14] (Gleeson CJ), 206 [63] (McHugh J), citing McKinlay (n 2) 56–7 (Stephen J).

[345] Simpson (n 19) 571.

[346] See generally Rowe (n 1).

[347] See generally Roach (n 2); Murphy (n 2).

[348] See generally Nationwide (n 3); ACTV (n 3); Theophanous (n 3); Coleman (n 3); McCloy (n 4); Brown v Tasmania (n 3).

[349] Simpson (n 19) 571–2.

[350] Roach (n 2) 187 [45] (Gummow, Kirby and Crennan JJ).

[351] Mulholland (n 5) 189 [10] (Gleeson CJ), 206 [63] (McHugh J).

[352] On differences between this and the ‘reasonably appropriate and adapted’ test, see Stone

(n 129) 146–8.

[353] See Kiefel (n 121) 501–7.

[354] See, eg, Twomey, ‘Rowe v Electoral Commissioner: Evolution or Creationism?’ (n 5) 183–4.

[355] Roach (n 2).

[356] Electoral and Referendum Amendment (Electoral Integrity and Other Measures) Act 2006 (Cth) sch 1 item 15.

[357] Roach (n 2) 202 [95] (Gummow, Kirby and Crennan JJ).

[358] Ibid.

[359] Ibid 174 [7], 179 [19] (Gleeson CJ).

[360] Rowe (n 1) 59–60 [161]–[164] (Gummow and Bell JJ), 118–21 [374]–[384] (Crennan J), which concerned government legislation imposing restrictive cut-off dates to enrol to vote (where such dates might prejudice certain classes of voters). See also Tham (n 1) 1006–9.

[361] Murphy (n 2) 46 [22] (French CJ and Bell J). Cf McCloy (n 3) 281 [308] (Gordon J).

[362] For the leading judgment, see Lange (n 3) 567–8 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ). See also McCloy (n 3) 193–5 [2] (French CJ, Kiefel, Bell and Keane JJ); Brown v Tasmania (n 3) 363–4 [104] (Kiefel CJ, Bell and Keane JJ, Nettle J agreeing at 416 [277]), 375 [156] (Gageler J); LibertyWorks Inc (n 130) 207–9 [76]–[85] (Kiefel CJ, Keane and Gleeson JJ, Edelman J agreeing at 240 [194]), 255 [247] (Steward J).

[363] See McCloy (n 3) 193–5 [2] (French CJ, Kiefel, Bell and Keane JJ), where it is asked whether a law burdening a particular guarantee, immunity or freedom is nonetheless valid as being a measure suitable, necessary and adequate in its balance. See also Murphy (n 2) 53 [38]

(French CJ and Bell J); Brown v Tasmania (n 3) 363–4 [104] (Kiefel CJ, Bell and Keane JJ,

Nettle J agreeing at 416 [277]), 375–6 [156] (Gageler J); LibertyWorks Inc (n 130) 207–9 [76]–[85] (Kiefel CJ, Keane and Gleeson JJ, Edelman J agreeing at 240 [194]), 255 [247] (Steward J). For a discussion of developments in this area, see generally Chordia (n 130).

[364] See LibertyWorks Inc (n 130) 255–6 [249] (Steward J); Twomey, ‘Rowe v Electoral Commissioner: Evolution or Creationism?’ (n 5) 202; 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, 777–82 (‘The Three “Rs”’); Nicholas Aroney, ‘Towards the “Best Explanation” of the Constitution: Text, Structure, History and Principle in Roach v Electoral Commissioner[2011] UQLawJl 10; (2011) 30(1) University of Queensland Law Journal 145, 147, 149; Simpson

(n 19) 573.

[365] See Stone (n 129) 126; Eric Chan, ‘A Proportionate Burden: Revisiting the Constitutionality of Optional Preferential Voting’ [2017] UWALawRw 13; (2017) 42(1) University of Western Australia Law Review 57,


[366] Judd (n 5) 385.

[367] See, eg, Murphy (n 2) 53 [38] (French CJ and Bell J); Stone (n 129) 135–6. The older ‘reasonably appropriate and adapted’ test is also applied to the construction of the purposive powers under s 51: Sir Anthony Mason, ‘The Use of Proportionality in Australian Constitutional Law’ (2016) 27(2) Public Law Review 109, 114–15.

[368] Murphy (n 2) 51 [34].

[369] Rowe (n 1) 21 [25] (French CJ). This reasoning was adopted by French CJ and Bell J: ibid.

[370] Murphy (n 2) 52–3 [36]–[38] (French CJ and Bell J).

[371] See Gleeson (n 5) 7.

[372] This is what Anne Twomey identifies as a more linear, ‘evolutionary’ approach to democracy: Twomey, ‘Rowe v Electoral Commissioner: Evolution or Creationism?’ (n 5) 184.

[373] Ibid.

[374] See, eg, ibid 183–6; Ruth Greenwood, ‘A Progressive Court and a Balancing Test:

Rowe v Electoral Commissioner [2010] HCA 46(2010) 14 University of Western Sydney Law Review 119, 128–9; Chan (n 365) 70. See also Hayne J’s and Heydon J’s dissenting judgments in Roach (n 2), which expressed more conservative views of democratic evolution and practice: at 206 [110]–[113] (Hayne J), 224 [179]–[180] (Heydon J).

[375] Rowe (n 1) 18–19 [18]–[22] (French CJ), 107 [328], 115 [356] (Crennan J).

[376] See, eg, Roach (n 2) 186–7 [45] (Gummow, Kirby and Crennan JJ). See also Twomey,

Rowe v Electoral Commissioner: Evolution or Creationism?’ (n 5) 187; Chan (n 365) 71–2.

[377] Twomey, ‘Rowe v Electoral Commissioner: Evolution or Creationism?’ (n 5) 184. See also Allan, ‘The Three “Rs”’ (n 364) 778; Graeme Orr, ‘The Voting Rights Ratchet: Rowe v Electoral Commissioner(2011) 22(2) Public Law Review 83, 88.

[378] See Twomey, ‘Rowe v Electoral Commissioner: Evolution or Creationism?’ (n 5) 181.

[379] Rowe (n 1) 129–31 [417]–[423].

[380] Murphy (n 2) 99 [222] (Keane J), 106 [243] (Nettle J), 125 [309] (Gordon J). Justice Gageler was more equivocal, but still appeared to express doubt over the need for an ‘evolutionary’ model, emphasising the more ‘stable and enduring’ nature of representative government: see especially at 69 [91]–[92].

[381] Tham (n 1) 1011 (emphasis added) (citations omitted).

[382] See, eg, Catts (n 36) 5–8, 20.

[383] See Twomey, ‘Let’s Not Demonise Plebiscite’ (n 29) 28.

[384] Lange (n 3) 559 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ).

[385] See Constitution ss 7, 24.

[386] Tham (n 1) 1011. See also Murphy (n 2) 87 [179] (Keane J) on the value and meaning of an elector’s choice.

[387] Something like this concern was expressed in Lindell, Submission No 4 (n 12) 3 [8]. See also Dixon, Kildea and Lynch (n 89) 5–6.

[388] Garran (n 112) 147.

[389] Murphy (n 2) 89 [186] (Keane J).

[390] Roach (n 2) 219–20 [161]–[162] (Hayne J), cited in ibid 93 [200].

[391] For a discussion of this general point, see Brendan Hord, ‘Murphy v Electoral Commissioner: Between Severance and a Hard Place’ (2017) 39(3) Sydney Law Review 399, 413.

[392] Murphy (n 2) 69 [91].

[393] Sawer (n 70) 179.

[394] See UNCRC Case (n 94) 5122 [52] (Lord Reed PSC, Lord Hodge DPSC and Lords Lloyd-Jones, Sales and Stephens JJSC agreeing).

[395] United Kingdom, Parliamentary Debates, House of Lords, 2 March 1911, vol 7, col 262 (Lord Balfour).

[396] Twomey, ‘Post-Williams Expenditure’ (n 63) 9–10, 21, 25.

[397] The Spending Cap (People’s Veto) Bill 2011 (NZ) proposed to

limit the annual increase in core Crown expenses by linking it to the rate of inflation and the rate of population change. The Bill also provide[d] for eligible voters to approve, by a majority of valid votes cast in a binding referendum, an increase in the [fixed] spending cap.

Explanatory Note, Spending Cap (People’s Veto) Bill 2011 (NZ) 1.

[398] New Zealand Treasury, Spending Cap (People’s Veto) Bill (Regulatory Impact Statement,

27 April 2011) 15.

[399] Ibid 19.

[400] See, eg, HS2 Funding Referendum Bill 2015 (UK).

[401] See Bogdanor, The New British Constitution (n 113) 18–19.

[402] Mulholland (n 5) 189 [10] (Gleeson CJ).

[403] Lindell, Submission No 4 (n 12) 3 [9].

[404] ‘iVote’, which was used in a New South Wales state election, is suggested as a way of holding plebiscites: Twomey, Submission No 6 (n 21) 3.

[405] Paul Kildea, ‘Australia’s Same-Sex Marriage Survey’ (n 32) 108–9, 115.

[406] On the trend towards privatisation and automation, see Yee-Fui Ng, ‘Institutional Adaptation and the Administrative State’ [2021] MelbULawRw 4; (2021) 44(3) Melbourne University Law Review 889, 907–16. On government automation in general, see Jack Maxwell and Joe Tomlinson, Experiments in Automating Immigration Systems (Bristol University Press, 2022).

[407] See generally Spending Cap (People’s Veto) Bill 2011 (NZ).

[408] See Rowe (n 1) 21 [26] (French CJ), 96 [289] (Heydon J).

[409] Gleeson (n 5) 9, discussing McGinty (n 1) 272 (Gummow J), quoting John Stuart Mill, Utilitarianism, Liberty, and Representative Government (JM Dent & Sons, 1910) 228.

[410] These were made under the War Precautions Act 1914 (Cth); Inglis (n 140) 35.

[411] Paul Kildea, ‘Australia’s Same-Sex Marriage Survey’ (n 32) 108–9, 115.

[412] In the context of the marriage plebiscite, it was argued by some that the franchise should be extended to 16- and 17-year-olds. It was proposed that this might better be done on a legislative footing (rather than by administrative action alone): George Williams, ‘Voting in the Same-Sex Marriage Plebiscite Should Be Extended to 16 and 17 Year Olds’, Sydney Morning Herald (online, 21 February 2016) <>, archived at <>.

[413] Roach (n 2) 174 [7] (Gleeson CJ), 199 [85], 202 [95] (Gummow, Kirby and Crennan JJ); Rowe (n 1) 18–21 [17]–[25] (French CJ), 46–50 [118]–[126] (Gummow and Bell JJ), 105–17 [324]–[368] (Crennan J); Lange (n 3) 559, 561, 567, 575 (Brennan CJ, Dawson, Toohey, Gaudron, McHugh, Gummow and Kirby JJ).

[414] Murphy (n 2) 69 [91] (Gageler J), 99 [222] (Keane J), 106 [243] (Nettle J), 125 [309]

(Gordon J). See also Hord (n 391) 413, 417.

[415] See Senate Legal and Constitutional Affairs References Committee (n 71) 2 [1.8], 10


[416] On ordinary mechanisms, see generally Richardson (n 280).

[417] It has been argued that a plebiscite which preceded a vote on the same topic under s 128, if it failed to take account of federal concerns, would in some way undercut the manner and form requirements of that section, and therefore also be invalid: see Kirby, ‘250 Years of the Crown’ (n 12) 528; Henskens (n 18) 159–62.

[418] For general discussion of the double referendum, see Aroney, The Constitution of a Federal Commonwealth (n 22) 299–334.

[419] Ibid 324–5. The Debates were subject to a division across at least two axes: one, as between liberals and conservatives; and another, as between those in favour of states’ rights, and those who were ‘nationalists’ (in the sense of preferring the majority national interest): at 188.

[420] Ibid 240–2. See also Christopher Puplick and Larry Galbraith, Submission No 10 to Senate Legal and Constitutional Affairs References Committee, Parliament of Australia, Matter of a Popular Vote, in the Form of a Plebiscite or Referendum, on the Matter of Marriage in Australia (2 September 2015) app A.

[421] Hobbs and Trotter (n 26) 79.

[422] Official Record of the Debates of the Australasian Federal Convention, Melbourne,

10 March 1898, 2174 (Isaac Isaacs). See also Aroney, The Constitution of a Federal Commonwealth (n 22) 188.

[423] Official Record of the Debates of the Australasian Federal Convention (n 422) 2180 (Isaac Isaacs).

[424] Ibid 2182.

[425] Ibid 2183.

[426] Aroney, The Constitution of a Federal Commonwealth (n 22) 242.

[427] Official Record of the Debates of the Australasian Federal Convention (n 422) 2188

(Bernhard Wise).

[428] By convention, the dissolution is effected by the Governor-General on the advice of the Prime Minister: Procedure Office, House of Representatives, ‘Double Dissolution’ (Infosheet No 18, July 2020) 1. Seven such dissolutions have occurred. These took place in 1914, 1951, 1974, 1975, 1983, 1987 and 2016: at 2.

[429] A citizen-initiated referendum mechanism was also rejected: Hobbs and Trotter (n 26) 73.

[430] McGinty (n 1) 175, 178 (Brennan CJ, Dawson J agreeing at 189), 229–30 (McHugh J), 275 (Gummow J).

[431] See, eg, Cole v Whitfield [1988] HCA 18; (1988) 165 CLR 360, where the High Court held that it may not have recourse to the subjective intention of the framers, but may (among other things) only use the Convention Debates to determine ‘the nature and objectives of the movement towards federation from which the compact of the Constitution finally emerged’, or to resolve ‘the contemporary meaning of [the] language used’: at 385 (Mason CJ, Wilson, Brennan, Deane, Dawson, Toohey and Gaudron JJ). Cf Stephen Donaghue and Thomas Wood, ‘The Evolving Role of History in Constitutional Interpretation’ (2020) 31(3) Public Law Review 248, 250, where an expanded use of the framers’ words is advocated.

[432] Senate Legal and Constitutional Affairs References Committee (n 71) 10 [2.16], citing Williams, Submission No 32 (n 71) 2.

[433] Paul Kildea, ‘Constitutional and Regulatory Dimensions’ (n 11) 295; Paul Kildea, ‘Australia’s Same-Sex Marriage Survey’ (n 32) 115.

[434] Williams, Submission No 32 (n 71) 3. See UNCRC Case (n 94) 5122–3 [51]–[55] (Lord Reed PSC, Lord Hodge DPSC and Lords Lloyd-Jones, Sales and Stephens JJSC agreeing).

[435] Twomey, Submission No 6 (n 21) 3; Twomey, ‘Let’s Not Demonise Plebiscite’ (n 29) 28.

[436] Dicey (n 104) 510–11; Tomn (n 21) 243.

[437] See Dicey (n 104) 500, 510–11.

[438] In the UK context, see generally Leah Trueblood, ‘Legislating for Referendums in the United Kingdom’ [2020] (January) Public Law 49. In the Australian context, see generally Paul Kildea, ‘Australia’s Same-Sex Marriage Survey’ (n 32).

[439] See above n 325.

[440] See Pape (n 54) 55 [111] (French CJ), 74 [182]–[183] (Gummow, Crennan and Bell JJ), 104–5 [292], [296] (Hayne and Kiefel JJ), 210–12 [600]–[604] (Heydon J); Williams [No 1] (n 61) 179–80 [4], 184–5 [22], 190 [31] (French CJ), 233 [139] (Gummow and Bell JJ), 249 [193] (Hayne J), 342 [484]–[485], 348 [507] (Crennan J), 361 [558], 370 [582]–[583] (Kiefel J); Richardson, Emmett and Palaniappan (n 64) 1 [2], 6 [28]; Koziol (n 64).

[441] See Bagehot (n 138) 7.

[442] McCloy (n 3) 227–30 [115]–[125].

AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback