Executive Summary

In November 1999 Australian voters participated in two referenda on the questions of whether Australia should become a republic and whether a new preamble should be added to the Commonwealth Constitution. Voters received a government-sponsored booklet spelling out the proposed changes to the words of the Constitution, and including two thousandword arguments on the YES and NO cases. These were the 43rd and 44th constitutional referenda since the first in 1906 and, like most attempts, these two failed. Over the years, only eight amendments have been made: single changes in 1906, 1910, 1928, 1946 and 1967; and three changes in 1977. On five occasions a national majority has been gained, though not a majority of States, causing the proposed alteration to fail.

Section 128, the constitutional provision which provides for referenda emerged from the Convention negotiations of the 1890s as a classical political compromise. In drawing up mechanisms for initiating referenda proposals and ratifying the outcomes, s. 128 gave something to the States as well as to the Commonwealth, and to the people as well as to the politicians.

There have been a number of significant amendments of the Constitution. The 1910 and 1928 cases provided a means for the Commonwealth to ease the burden of State debts, with the latter also providing constitutional backing for the Financial Agreement and the Loan Council. Many Australians have benefited from the social services amendment of 1946, while Australia's indigenous populations regard the Aboriginals referendum of 1967 as a major landmark in their relationship with the Australian political system. The issue of the replacement of Senators was one of the major points of political contention during 19745 and two years later an amendment attempted to remedy this problem. In the same year, residents of Australia's Territories gained the vote in constitutional referenda.

Although constitutional referenda are important occasions of political activity, remarkably little research has been undertaken on Australians' voting behaviour at such times. However, referenda have been much written about and commented upon and this paper discusses some of the views that have been propounded about Australia's voting record in them particularly since the 1999 referenda which produced an unusually large number of studies.

Many critics of our constitutional referenda have been concerned by what they see as the relatively poor record of constitutional change in this country. Some believe that achievement of the 'double majority' is too difficult, while others criticise official support for the handing out of NO cases. There is a common claim concerning voter ignorance and conservatism which is said to account for so many negative votes.

By contrast, others suggest that voters are well aware of what is involved, citing different patterns to the returns as evidence. Amendments perceived to weaken the federal system have generally been given short shrift, whereas attempts to repair the perennially-flawed finance chapter have been received sympathetically. Proposed changes to the Senate are likely to be repelled but matters dealing with Australian society are usually considered closely. It has been claimed that such varying outcomes are a reminder that voters go to a referendum polling place possessing some consciousness of Australian history and politics which influences the way they vote. Other evidence for this has been the fact that particular results have often indicated a voter awareness of what was involvedthe vastly different results in the two instances in 1967 for instance. Australian voters have been cautious, and conservative at times, but they have shown their willingness to accept change when persuaded that such change is sensible or right.

Whatever one's view of constitutional referenda, there seems general agreement among commentators that the major parties have treated the electorate with disdain in the years since the first referendum in 1906. Not the least of the criticisms of the parties has been the inconsistency in their stance on constitutional amendment. Most notably this has involved supporting issues while in government, which are then opposed when submitted by their opponents.

But the parties have erred in more fundamental ways. They have often tried to do too much at one time, they have sometimes been less than earnest in putting YES cases, and they have sometimes ignored the political times which have made some questions unlikely to be accepted. They have also been unable to avoid opportunistic, short-term responses that have put the embarrassment of their opponents ahead of a careful consideration of the issues under discussion. The parties have also overlooked strong community views on particular questions, and they have often antagonised the Premiers, who have participated in quite a number of NO campaigns. Overall, then, the history of Australian constitutional referenda has been one of controversy, with many observers criticising the voters, but others directing their criticism at the parties.

Introduction

The process by which the Australian Constitution is amended is complex and probably little understood by the general community. This paper is an attempt to explain the process of constitutional amendment that is found in s. 128 of the Constitution and supporting legislation.

The paper follows an earlier Department of the Parliamentary Library publication, by Scott Bennett and Sean Brennan, which was partly a discussion of the politics of constitutional amendment (by Bennett) and partly a discussion of the constitutional aspects of s. 128, and ways in which the rigidities of the section might be circumvented (by Brennan).⁽¹⁾ It was noted in that earlier paper that 'although a great deal has been written and spoken about Australian constitutional referenda, remarkably little research has been conducted into Australians' voting behaviour on these occasions'.⁽²⁾ Interestingly, this cannot now be said about the two referenda in November 1999, and ideas and conclusions from the findings of scholars published since then have been incorporated into this text by Bennett. In the interests of space, Brennan's legal discussion has been removed, though interested readers are referred to the 1999 paper.

The paper is divided into five sections:

• the amendment process
statistical and other details of the 44 referenda
problems posed by s. 128
the politics of constitutional amendment, and
the voters.

Amending the Constitution

The Constitutional Provisions

The Commonwealth Parliament initiates constitutional amendments. Section 128 of the Australian Constitution requires that a proposal to amend the Constitution must first take the form of a Bill submitted to the Commonwealth Parliament. Between two and six months after it leaves the Parliament, the proposal 'shall be submitted' in a referendum to the voters in the various States and Territories.

For an amendment to be ratified, the so-called 'double majority' is required. There must be a majority of voters saying YES in a majority of the States (i.e. at least four of the six), but there must also be a nation-wide affirmative vote. Territory votes are included in the national total, but not in any State figure (for the wording of s. 128 see Appendix).

The YES and NO Cases

Parliament prescribes the manner in which referendum votes are taken. In most referenda since the Referendum (Constitution Alteration) Act 1912 (No. 2), each elector has received a pamphlet containing arguments in favour and arguments against any proposal upon which the elector is voting. Normally, these arguments must be no more than two thousand words in length, and must be authorised by a majority of those parliamentary members who voted for or against the proposed law. When more than one referendum is held on the same day, a particular argument may be more than two thousand words, provided the average of YES or NO arguments is not greater than two thousand words.⁽⁴⁾

In some cases, NO arguments are not produced. In 1967, for instance, arguments for and against the Parliament alteration were produced, but only a YES argument in relation to the Aboriginals amendment. The Referendum (Machinery Provisions) Act 1984 (s. 11) states that such arguments as are produced must be submitted to each voter, 'not later than 14 days before the voting day for the referendum'.

Typically, the YES cases are usually led by the sponsoring government, while the NO cases are argued by informal groups of individuals often combining their efforts with the Opposition of the day. A marked departure from this norm was seen in the Establishment of Republic referendum in 1999. Unusually, the cases were managed by two advertising campaign committees appointed by the Government, with the YES committee chaired by Malcolm Turnbull, of the Australian Republican Movement, and the NO committee by Kerry Jones, of Australians for Constitutional Monarchy. The Howard Government's role was limited to ensuring that each committee's proposals met the 'basic standards' set for 'the activities to be covered by the public funding', as well as accounting for the use of the funds. The two publicly-funded committees were, therefore, 'responsible for the conduct of the campaign'.⁽⁵⁾ Governments and Oppositions have not previously been prepared to hand over so much control to others.

The Ballot Paper

From the first ballot in 1906, the style of question was uniform for many years, voters being asked 'Do you approve of a proposed law for the alteration of the Constitution entitled [here the title of the proposed law was inserted]'. For example:

In recent years, however, there has been occasional tinkering with the style and wording of the question. For the 1974 referenda, for example, the ballot paper read:

For the four referenda in 1988, the instructions read:

In 1999 the format was again slightly different:

Referenda Facts and Figures

There have been 44 attempts to amend the Constitution, of which eight have been successful (Table 1):⁽⁶⁾

Table 1: Constitutional Referenda 19061999

Note: Successful amendments to Constitution in bold.

Some amendment proposals have dealt with a single issue. The Senate Elections (1906) question, for example, was whether the commencement date for Senators' terms should be altered, while the Marketing (1937) question sought to remove the restrictions imposed on the Commonwealth Parliament by s. 92 of the Constitution.

Others, by contrast, have involved a number of issues. Trade and commerce (1911) sought to extend the Commonwealth Parliaments powers over trade and commerce, the control of corporations, labour and employment and combinations and monopolies. Rents and Prices (1948) sought to give the Commonwealth Parliament ongoing power to make laws with regard to both rents and prices. In both the 1911 and 1948 cases a single YES or NO for or against the entire package was all that the voter could registerthe voter could not support or reject parts of each question (see below, pp. 1415).

As can be seen in Table 1, there have been eight amendments to the Constitution: single changes in 1906, 1910, 1928, 1946, and 1967, and three changes in 1977 (for details, see pp. 813).

Six of the successful referenda have been carried comfortably. The State Debts (1910) and Social Services (1946) amendments had the narrowest margins, at 54.9 per cent and 54.4 per cent, respectively. Five of the questions that received 50 per cent of the total vote or better, failed to get a majority of States in favour. In nine other cases the national YES vote was in the range of 49.0 to 49.8 per cent (Table 2):

Table 2: YES Votesrange of votes

Table 1 shows that referenda have been held on 19 separate occasions. On five of these, voters have had to vote in just a single referendum. On the other 14 occasions, multiple referenda have been held, with the six questions in 1913 being the largest number considered on the same day (Table 3):

Table 3: Number of Referenda Per Polling Day