Harry Evans: Selected Writings

Papers on Parliament No. 52
December 2009

Prev | Contents | Next

An Elected President for an Australian Republic: Problems and Solutions

Harry Evans

It is self-evident to some people that it is anomalous for Australia to continue to share a head of state with another country on the other side of the globe, a country which is now not an independent nation but a member of a foreign quasi-federation. Following the referendum of 6 November 1999, however, it is also obvious that the Australian electorate is not willing to accept just any proposal to correct the anomaly.

A large part of the difficulty of making the change to a republic appears to be the resistance of the political elite to the strong public preference for a popularly elected head of state if there is to be any replacement for the Queen.[1] Overcoming this difficulty would be a substantial step towards reversing the result of November 1999. The proposal then put to referendum not only ignored this public preference, but proposed a replacement as far as possible removed from an elected president: an appointed head of state dismissible at any time by the prime minister. This proposal reflected the complete and dogmatic rejection by the official republican movement of the notion of an elected president. That rejection was based, in effect, on a claim that there would be insurmountable difficulties in combining such an office with the current system of government.[2] It is therefore desirable to analyse those alleged difficulties to see whether they are indeed insurmountable or whether there are solutions to them.

The underlying assumption of the official republican position is that we must keep the existing cabinet system of government, that is, the system whereby the executive government is carried on by a ministry who are members of the Parliament and who depend for their collective holding of office on the support of a majority of the House of Representatives. It is assumed that we wish to rule out of contention an executive presidency like that of the United States, or a hybrid system in which the head of state participates in the exercise of the executive power as in France. Pure democrats may contend that the electorate should be consulted on those options and given the opportunity to choose either of those systems, but for the purpose of analysing the alleged difficulties of combining an elected president with a cabinet system those possibilities may be put aside and the assumption of continuance of cabinet government accepted.

The alleged difficulties relate to the powers, the method of nominating and electing, and the method of removal, of the replacement head of state.


The current Constitution confers on the Governor-General the executive power, including the power to appoint ministers to administer the departments of state. Under the conventions of the monarchy and cabinet government, the Governor-General does not personally exercise the executive power, but normally acts on the advice of the prime minister and the ministry of the day, which is supported by a majority of the House of Representatives. Those conventions envisage, however, that in certain circumstances the Governor-General may exercise the powers of the office independently to preserve and facilitate the operation of the system of government. The powers exercised in those circumstances are referred to as the reserve powers.

The fear is that an elected president, replacing the Governor-General with powers unchanged, might use the ‘mandate’ of the office provided by popular election to exercise the reserve powers contrary to the conventions, or to take over the executive power and change the system of government. In short, we could find ourselves with an American-type executive presidency or a French-type hybrid system without intending to make that change.

This problem is related to the lack of specification in the Constitution of the rules of cabinet government as it actually operates.

The powers of the Governor-General hinge on the power to appoint and dismiss the prime minister and the ministry and the power to dissolve the House of Representatives (or both houses in the situation set out in section 57 of the Constitution) and to refuse a recommendation for a dissolution. It is these powers which are exercised normally on advice, or as reserve powers independently in certain circumstances. Other powers, for example, the power to assent to proposed laws or to withhold assent, are subsidiary to these main powers.

It is often said that, to avoid an elected president taking over the executive government, it is necessary to codify the powers of the head of state and the conventions of cabinet government. The use of the word ‘codification’ suggests that this is a lengthy and difficult task, full of opportunity for disputation about the conventions.

In reality, it is not a lengthy or difficult task. It is necessary only to make the following provisions in the Constitution:

  • to provide for the office of prime minister as the head of the ministry and the person who specifies the ministers and the departments of state they are to administer

  • to provide that the person appointed as prime minister is the person who has the support of a majority of the House of Representatives as indicated by a resolution of the House to that effect

  • to provide that the House of Representatives is to be dissolved only on the advice of the prime minister, unless it passes a resolution expressing lack of support for the prime minister and does not, within a specified time, pass another resolution expressing support for another person as prime minister

  • to provide that, where the conditions specified in section 57 of the Constitution exist, a dissolution is not to occur except in accordance with a resolution of the House of Representatives (the head of state would still have to be satisfied that the conditions exist).

This is all that is necessary for a ‘codification’ of the powers of the head of state and the conventions of cabinet government. These provisions are not novel. Most of them were contained in a bill to amend the Constitution which was passed by the Senate in 1982 but which was refused passage by the then government in the House of Representatives, the Constitution Alteration (Fixed Term Parliaments) Bill 1982. The only novelty in the above list is the suggestion that a section 57 dissolution occur only in accordance with a resolution of the House, to prevent the head of state exercising that dissolution power independently. That bill should have been put to a referendum by the incoming Labor government in 1983, and, having the support of the Labor party, the minor parties and a considerable number of the non-Labor senators, it would have had a good chance of success. For reasons not explained, it was dropped by the incoming government.

As the title of the bill indicates, its purpose was not to make the change to a republic but to provide for fixed term parliaments. The provisions listed above could be inserted in the Constitution with or without provision for a fixed term parliament, but the latter has the advantage of depriving the government of the day of the power to go to an election at a time of its choosing, and is seen by many as an additional reform which could be made in the change to a republic.

Other adjustments to the executive power should be made. One such necessary change, which is seldom mentioned, is the abolition of the power of the executive to prorogue the Parliament (section 5 of the Constitution). This monarchical power, much used by Stuart monarchs to dispense with troublesome parliaments, and more recently by the New South Wales Government to stop the Legislative Council inquiring into its activities, has no place in a republic.[3] It could be used to disrupt the system of government; for example, a prime minister about to be overthrown by the House of Representatives could resort to prorogation to cling to office.

Under the proposed provisions, in conjunction with current constitutional provisions which would remain unchanged, the head of state would still have ample scope to be the ‘umpire’ of the system of government, as well as to ‘represent the nation to itself’ as only an elected person could.

Moreover, the provisions should be thought of not as regulating or limiting the powers of the head of state, but as clarifying the position of the executive government, of which the head of state is the apex, in its relationship with the legislature. In normal circumstances it is prime ministerial power which is exalted by the lack of specification of the executive power.

The suggested provisions ignore a large red herring which has been dragged across the trail. This is the claim that, in the change to a republic, it will be necessary to change the powers of the Senate so that the Senate cannot ‘refuse supply’. Those making this claim, if pressed, justify it by reference to the events of 1975 and make the further claim that it is necessary to avoid a repetition of those events. If pressed further, they claim that an elected president might be in a stronger position to dismiss a government and decree a dissolution, of the House of Representatives, or of both houses in a section 57 situation, contrary to the wishes of the government of the day, where there is a disagreement between the houses over appropriation bills. This formulation, however, makes it clear that the problem is not the legislative powers of the Senate but the ability of the head of state to dismiss the government and call a dissolution independently. That situation pertains under the current arrangements and regardless of any change to a republic. The provisions listed above, by ‘codifying’ the powers of the head of state, remove the problem. Under those provisions, it would not be open to the head of state, Governor-General or president, to dismiss the government and call a dissolution independently, whether or not there is a disagreement between the houses over appropriations.

The suggestion that the change to a republic should be combined with ‘stopping the Senate blocking supply’ raises numerous difficulties. Such terms as ‘supply’ and ‘budget measures’ have no fixed meaning. Any bill can be turned into an appropriation bill. It is therefore difficult to formulate a constitutional change whereby the government could pass its appropriations in the absence of the consent of the Senate, without creating loopholes whereby the government could pass any laws without the consent of the Senate. This would create de facto unicameralism and a prime ministerial dictatorship; with control of the House of Representatives a government could legislate by decree. Even if the power to bypass the Senate could be confined strictly to annual appropriations, this would be an unacceptable elevation of executive power: governments can commit enormities through annual appropriation bills.[4] Those who consider that the executive government is entitled to automatic endorsement of its appropriations, or any legislation, should explain why we bother to elect houses of parliaments at all. Lack of government control over the Senate is a residual safeguard, in an otherwise deteriorated parliamentary system, against abuse of the legislative power.

There is and would continue to be a residual risk of a deadlock between the two houses over financial legislation where the double dissolution provisions of section 57 of the Constitution would be too slow to provide a resolution. Such a deadlock could have arisen, for example, over the States Grants (Primary and Secondary Education Assistance) Bill 2000: the majority of the Senate could have refused to pass the bill while it contained grants to wealthy private schools, and the government could have refused to pass the bill without those grants, and schools could have been forced to close. This would have been a genuine deadlock. The absence of such events hitherto indicates that preserving the legislative safeguard is worth the slight risk. Voting patterns suggest that many electors see it that way.

Reference to the electors’ perception leads to an equally serious difficulty with allowing legislation to bypass the Senate. By combining such a power shift with the change to a republic, the latter would be likely to be poisoned by the association with another agenda, and a republic would again be rejected because of the defective model offered. The proposal for a republic, in this combination, would be represented as a ‘grab for power’ and as ‘Gough’s revenge’. Apart from the enormous enhancement of prime ministerial power, it would raise objections associated with federalism and the equal representation of states in the Senate. Those who wish to have a republic and incidentally solve the problem of the unspecified powers of the head of state would wish to avoid poisoning the well by this combination with another agenda.

Nomination and election

The opponents of an elected president have raised difficulties with nomination and election procedures. These difficulties boil down to the alleged propensity of any popular nomination and election process to produce a partisan president. It is said that however you devise the nomination and election procedures, a partisan contest and therefore a partisan president is inevitable.

A preliminary point needs to be made about this thesis. In any system of popular election, the electors would choose the president, and it is necessarily open to the electors to choose a partisan nominee. If that is the choice of the electors, true democrats and republicans are constrained to accept the people’s choice.

A partisan president is much less to be feared if the provisions listed above are made in relation to the office. If the powers of the office are provided as suggested, a partisan head of state could be merely an embarrassment but not a danger to the system of government.

It is possible, however, to devise nomination and election procedures to discourage partisan contests for the office and partisan choices by the electorate.

There is one set of proposed measures which needs to be put aside. Provisions intended to prohibit direct participation by political parties in nominations and elections in an attempt to achieve a non-partisan process are superficially attractive. Examples of such measures are prohibitions on candidates being, or having been, members of political parties, prohibitions on political parties sponsoring nominations, prohibitions on campaigning or campaign expenditure by parties, and so on. All such measures are likely to be full of loopholes and eventually circumvented. If political parties are determined to field partisan candidates and to run partisan campaigns, they will soon find a way around prohibitions of this sort. Intended candidates could resign from parties while entering into secret agreements to represent those parties, parties could organise nominations under another guise, campaign expenditure could be channelled through other organisations, and so forth. Constitutional prescriptions of such prohibitions are likely to be lengthy (how, for example, is ‘political party’ to be defined?), but still full of loopholes. If the details of the prohibitions are left to ordinary legislation rather than constitutional prescription, the danger of partisan manipulation of the rules arises, and there could also be constitutional difficulties, such as conflict with the implied freedom of political communication which the High Court has found in the Constitution. In short, attempts to exclude political parties are likely to be counter-productive.

Under any nomination and election system for a president, political parties would probably realise that it would be in their interests to support non-partisan candidates generally sympathetic to their view of the world but more attractive to the electors than conventional politicians. The relevant provisions should be framed to encourage them, rather than to require them, to do so.

In relation to nominations, this formulation would involve avoiding any arrangements which give political parties effective control of nominations, or even a leading role. In any event, such provisions would revive objections to a ‘politicians’ republic’ which were a feature of the defeat of the 1999 proposals. On that basis, we should avoid confining nominations to parliaments or their members or other bodies or persons in the political structure.

This leads to nomination by the electors. Nomination of candidates for president could be by the signature of a particular number or a particular percentage of persons enrolled as electors. The number or percentage would be set fairly high to ensure that only candidates with significant public support would be nominated. Similarly, signatures should be required from all states and territories, with the numbers or percentages varying by state and territory, to ensure that candidates have wide geographical support. The numbers or percentages of signatures required could be varied by legislation but with a formula set by constitutional prescription.

This would mean that the collection of signatures would have to be organised, which would mean that organisations would be involved in collecting signatures. Political parties could organise the collection of signatures, but so could other bodies. Electors nominating candidates would know in advance the candidates they were being called on to support; it would not be a matter of simply endorsing a party or organisation, so there would be an incentive to put forward popular candidates with widespread support before the process of collecting signatures began. Any candidate with the required number of signatures, duly verified, would appear on the ballot. The time of nomination would be fixed sufficiently in advance of polling day to allow time for verification of nominations.

In this connection the importance of having a fixed presidential term is evident. It is necessary that the election timetable be known in advance so that there is time for the nomination and voting processes to be completed in an unhurried manner, to produce a president-elect well before the end of the presidential term. A fixed presidential term entails elections for president separate from parliamentary elections, unless provision for fixed-term parliaments is also to be made. As has been pointed out elsewhere, separate elections for president, although more costly, would discourage partisan contests for the office.

In relation to election campaigning, there would be no restriction on organisations campaigning for particular candidates. Any attempt to restrict campaigning would, as has been indicated, probably be in vain, and, if done by ordinary legislation, possibly unconstitutional. This would mean that there would be nothing to prevent more money being spent in support of some candidates than others. No democracy has solved the problem of limiting campaign expenditures, particularly the expenditure of ‘soft money’ by organisations other than political parties, and it would be unrealistic to expect a solution in this context. It must be hoped that the electors would view massive spending on election campaigns with distaste because of the nature of the office, and that any attempts to buy elections would be counter-productive when exposed.

Direct election of course rules out any intermediate body, such as an electoral college.

Voting should be preferential, and the candidate with the most votes, after distribution of preferences if necessary, should be elected. Australians are accustomed to preferential voting, and there would seem to be no reason for adopting the more expensive alternative of run-off elections.

If the nominating procedure described above were adopted, there would seem to be no reason to require a special majority, such as a majority including majorities in three or four states, or some percentage of the votes in a number of states, as is sometimes suggested. The nomination procedure should ensure a sufficient geographical spread of support for a candidate.


For an elected president, removal from office should not be an easy process. Certainly removal by prime ministerial decree, as in the 1999 proposal, would not be acceptable. Nor should either or both houses of the Parliament be able to remove an elected president at their discretion.

The provisions in section 72 of the Constitution for the removal of federal judges provide the appropriate model: removal by resolution of both houses of the Parliament only on the grounds of proved misbehaviour or incapacity. These grounds have been the subject of much exposition, and are now fairly well understood. With judges there is the additional safeguard that the concurrence of the Governor-General is required, but it is not clear on what grounds, if any, the High Court would be able to review a removal. For a president an additional safeguard of a special majority of each house of the Parliament could be required.

There remains the question of the replacement of the president in case of removal, death or resignation. It is suggested that the houses of the Parliament be empowered to appoint, by a special majority, an acting president until an election can be held. Provision could be made for the houses to meet speedily in case of a vacancy suddenly occurring (the abolition of the prorogation power would assist here). If either or both houses were dissolved when a vacancy occurred, the government could make a temporary appointment until the houses met again.

The alternative, suggested by the Keating Government’s Republic Advisory Committee, that the most senior of the state governors act as president, unduly depends upon the states having the same constitutional arrangements as the Commonwealth. Some or all of the states may choose different arrangements, for example, having no separate head of state, or having an executive governor. This presumption that the states will have similar arrangements should be avoided. There would seem to be no other suitable offices to provide an ex officio acting president. Modern communications make it unnecessary to provide for the overseas absence of the president.

The essential ingredients

The following elements have been identified as governing the change to a republic:

  • replacing the Queen with a non-executive head of state

  • preserving the system of cabinet government

  • deferring to the public preference for an elected head of state

  • providing democratic procedures and avoiding the defects of a ‘politicians’ republic’ or an ‘elitists’ republic’

  • filling gaps in the existing Constitution, particularly the lack of specification of the powers of the head of state

  • avoiding any combination with other agendas, such as the ‘grab for power’ (executive domination) or ‘Gough’s revenge’ (removing or weakening the safeguard provided by the Senate).

A consideration of these elements strongly points to the kind of scheme here outlined.

The proposed provisions would provide a truly republican republic, combining the quintessential republican elements of popular sovereignty and safeguards against undue concentrations of power. The Australian electorate has shown a sound, instinctive understanding of these republican principles.

1.   J. Kelley et al., ‘Monarchy, republic, parliament and the people: 20 years of survey evidence’, Australian Social Monitor, Melbourne Institute of Applied Economic and Social Research, November 1999, pp. 104–113; C. Hull, ‘Get used to it, we want direct election’, Canberra Times, 5 February 2000, C3.

2.   ‘The proper balance of authority between president and prime minister would be gravely threatened, if not severely distorted, were a directly elected president to co-exist with a prime minister unable to claim such a direct popular mandate’., G. Winterton, Monarchy to Republic: Australian Republican Government. 2nd edn, Melbourne, Oxford University Press, 1994, pp. 112–13.

3.   H. Evans, ‘Power to prorogue a relic of imperial past’, Canberra Times, 1 February 1996.

4.   In 1995 the government attempted, by an annual appropriation bill, to pay the legal fees of a minister in a matter unconnected with the minister’s Commonwealth ministerial capacity. The Senate amended the bill to remove the offending sum.

Prev | Contents | Next

Back to top