Alternative Law Journal
A constitution is fundamentally about providing sufficient political power to government to perform the will of the people and, at the same time, limiting that power to ensure that the people’s liberty is not compromised. If a constitution is insufficiently entrenched, the government may attempt to change the mechanisms that limit its power and the abuse of the people’s liberty could follow. For this reason, it is important that in any democracy, the following condition is met: the politics of constitutional change should be separated from the realm of normal politics.
In considering the way that Australia’s constitution is formally altered, this condition has only been partially met. The way that constitutional referendums are initiated has meant they are firmly embedded in the realm of normal politics. Indeed, the formal constitutional amendment procedure is caught in a ‘cycle of inertia’ with three key elements:
• the amendment procedure,
• the referendum results, and
• criticism of the amendment procedure.
Each element of this cycle drives the next. The amendment procedure has produced proposals for change of a certain type and, consequently, Australia’s constitutional referendums have rarely succeeded. This in turn has led to criticism of the amendment procedure. This criticism has resulted in changes in, or attempts to change, certain aspects of the procedure, thus completing the cycle. The elements of this cycle will now be explored.
Australia’s constitution can only be formally altered through the initiation and ratification of a constitutional referendum (these conditions are laid out in s.128 of the Constitution). Constitutional referendums must be initiated in the Commonwealth parliament. However, the development of the disciplined mass party has resulted in the governing party at the federal level controlling the initiation of constitutional referendums. A rider to this is that while a constitutional amendment bill may pass through both houses of parliament in the normal fashion, there is also provision in s.128 for a referendum bill to pass through the same house twice (in theory, a hostile Senate could propose change against the wishes of the government but in practice, the Governor-General will only sign the writ for a constitutional referendum on the recommendation of the Prime Minister).
The control that the government of the day has over the initiation procedure means that the politics of constitutional change has been tied to the struggle between the major political parties at the federal level. The key role of political parties in the constitutional amendment process is reflected in a number of ways.
The quest for partisan advantage has meant that successive Labor and non-Labor governments have produced proposals for formal constitutional change that have characteristically been about concentrating, not dispersing political power. Most proposals have been about changing the ambit of Commonwealth influence (32), especially in relation to the economy (23, including the Powers referendum of 1944). Others have been designed to change the machinery of government (9, including the recent republic proposal), especially the timing of Senate elections (4). There have also been proposals designed to change the mode by which the Constitution is formally altered (2) and to add a preamble to the Constitution.
The partisan dimension is also reflected in the frequency and timing of referendum proposals. In considering the period from 1910 (around the time when the existing Australian party system emerged) to the present, the partisan colour of the federal government has changed 11 times (from Fisher Australian Labor Party through to the Howard Liberal–National coalition). Labor has been more eager to change the Constitution than non-Labor. Labor governments have initiated more amendment proposals than non-Labor governments (25 compared with 19) and have held, on average, considerably more constitutional referendums for each year in office (0.8 compared to 0.4 for non-Labor). Also, on average, Labor governments have attempted to change the Constitution by referendum much sooner after coming to power (an average of about 20 months) than non-Labor governments (an average of about 37 months).
The central role of party could be seen in the recent referendum on the republic issue, though the dynamics contrasted with past experience. Usually, the Prime Minister as the leader of the government (and therefore the major party grouping) plays a leading role in advocating change during a referendum campaign. The republic referendum was the first time that a Prime Minister actively campaigned against a referendum question. Some argued in the course of the referendum campaign that the Prime Minister used his position to spoil the republic referendum’s chances of success. This was particularly evident in the wording of the referendum question that explicitly set out that the President would be elected by Parliament (but did not mention more electorally popular elements like the public nomination process). In response, Senator Natasha Stott Despoja, on the night the referendum question passed through Parliament, said, ‘I think the Prime Minister should hang his head in shame. He is not a statesman. We have a biased question, we have a Prime Minister who said he would opt out of this debate and take a back step, and yet he interfered’. Others have focused their criticism on the Constitutional Convention as an event where the result was stage-managed. This was perhaps not unexpected since it was a Convention that was organised by the government and managed by elites from the major parties. The partisan dimension could also be seen in the opinion polls leading up to the referendum (Labor voters were more likely to vote Yes).
While the power of party modified the republic debate, referendums also have a powerful effect on party. In non-referendum mode, the normal behaviour of a political party is to present a consistent public face on an issue. Referendums cause political parties to split and the Liberal Party split on the republic issue during the campaign (for example Howard, Reith and Costello each publicly held a different view). Indeed, the impact of constitutional referendums in untangling and recasting different interest group coalitions has meant that they have played a key role in the development of Australian politics. Referendums contributed to the formation of the early Ministerial Party, the modern Liberal Party and the separation of the Democratic Labor Party from the Australian Labor Party.
While the initiation of proposals to formally change the constitution is embedded in the realm of normal politics, their ratification is not. Proposals for constitutional change are ratified through a referendum. A special majority of the Australian people (an overall majority plus majorities in a majority of States) must vote Yes in order for a proposal to pass. The referendum effectively provides Australia voters with a veto role similar to that of a very strong Upper House. The people have used this role to great effect. With the recent defeat of the proposals to change Australia to a Republic and to add a preamble, 36 of 44 proposals have been vetoed. Referendum defeats have tended to be massive (31 proposals received less than 50% of the total vote and 11 failed to receive majorities in any of the six States). Successful proposals have tended to be of low significance in that they have not changed the status quo. When it comes to formal constitutional change in Australia, ‘inert’ therefore seems to be a fitting description (informal constitutional change continues of course in the areas of intergovernmental relations and judicial review).
The failure of constitutional referendums to pass has resulted in criticism leveled at every aspect of the formal constitutional amendment procedure: the way referendum campaigns are conducted, the way referendums are ratified and the way referendums are initiated.
While commentators have often considered the need for a long gestation period to allow a consensus to emerge, referendum campaigns tend to be short, sharp and highly partisan affairs. Over time, the characteristic failure of referendums has meant greater expense and more professional attempts to influence the result (but not more referendums passed).
Most criticism has directed at the role of the voter in ratifying constitutional referendums. Indeed, the image of the Australian as an ‘unthinking No voter’ or ‘simple partisan’ are the ones most frequently presented in the literature. However, there is little evidence to support either image. Analysis of constitutional referendum voting indicates that, historically, there has been high turnout, low informal voting and a demonstrated ability to distinguish between proposals when two or more are put to the vote. Arguments about the importance of partisanship are tempered by analysis showing that the States as regions play a key role in structuring referendum voting.
The large number of referendum proposals that have failed to pass is less a sign of the incapacity of the Australian voter than it is a reflection of the kind of proposals that have been put to the vote. As outlined above, formal constitutional change in Australia has been subordinate to the partisan considerations of Labor and non-Labor governments. The kind of proposals put to the vote, and consequently Australia’s record of change, would be qualitatively and quantitatively different if the power to initiate Constitutional referendums was extended beyond the Commonwealth government, to the State governments or to the Senate for example. These institutions would arguably be more likely to produce proposals for change designed to disperse, rather than concentrate political power. Further, referendum proposals of this kind would be more likely to pass (the cycle would no longer be inert and would be allowed to move forward).
In the meantime, inertia in formal constitutional change is not a bad thing if the alternative means that the responsiveness of the government is compromised. It is this point that many critics of Australia’s record of formal constitutional change fail to pick up. Critics have suggested that the referendum record, in terms of the number that have passed, is evidence that the referendum has failed as a means of constitutional amendment. Indeed, Australia does not fare well in bald comparisons with the amount of constitutional change that has occurred in other countries. But the number of changes made is a poor, even irrelevant, measure of a satisfactory system of constitutional amendment. In terms of the types of changes made, Australia compares well.
For example, in the United States, the system of constitutional amendment involves either initiation by two-thirds of the State legislatures and ratification by three-quarters of the State legislatures or simply a two-thirds majority of both houses of Congress. This institutional structure has meant that few changes have been made other than in the areas of expanding certain democratic practices and asserting governmental protection of certain basic rights.
In New Zealand, seven of the nine nationwide referendums held between 1949 and 1993 received a majority. However, the two New Zealand referendums which failed to receive a majority, in 1967 and 1990, were unique in that they were the only ones which, if they passed, could be said to have reduced the power of the people over their elected representatives (by allowing for less frequent elections). Certainly, referendums which resulted in reducing the majoritarian aspects of the New Zealand political system have been considerably more successful (in 1992 and in 1993). While Australia has never had constitutional referendums which would have worked similarly in the direction of reinforcing the consensual elements of its democratic system, the results of nationwide referendums in New Zealand, especially those which have failed, are arguably consistent with the pattern of results of constitutional referendums in Australia.
Finally, with similar political institutions and processes, Canada’s experiences with constitutional amendment are perhaps the most appropriate to compare with Australia’s. Before 1982, certain key components of Canada’s Constitution were amended by the British Parliament at the request of Canada’s federal government. Systematic analysis of Australia’s and Canada’s record of change shows that while Canada has had much more formal amendment than Australia (24 amendments up to 1982), they have been mostly minor. The experience of the two countries differs much more in the change which has been attempted (with the Australian governments much less content with existing constitutional arrangements), than the change which has resulted. The constitutional amendment procedure in Canada changed with the passing of the Constitution Act 1982. Now, the Canadian Constitution is amended in Canada, by the passing of resolutions within the House of Commons, the Senate and a specified proportion of provincial legislatures. Despite the considerable efforts of the Canadian federal government, only minor changes have been made to the Constitution since 1982. The failure of the Meech Lake agreement meant that a different approach was taken with the Charlottetown Accord in the early 1990s. Part of this approach was the adoption of a referendum with strict ratification requirements: the referendum had to pass in every province (the referendum was non-binding but the government adopted these strict ratification requirements). The referendum held in 1992 was unsuccessful. It has been argued that its form may serve as a precedent for all but the least significant of constitutional changes in the future.
It seems then that, from the perspective of the type of constitutional change that has occurred, the system of constitutional amendment in Australia has produced results comparable to that of similar liberal democracies.
There has been speculation in the media that while the republic referendum of 1999 was lost, the opportunity to vote in another referendum on the issue will avail itself fairly soon. Indeed, immediately after the 1999 republic referendum failed, the Opposition leader, Mr Beazley outlined a four-step plan which is, of course, tied to the political fortunes of the Labor Party.
• winning the next federal election in 2001;
• holding a plebiscite in 2004 to determine whether Australians want a republic;
• holding another plebiscite in 2005 asking voters to decide between direct election and appointed president models; and
• federal parliament constructing a referendum question in 2007 on the basis of the result of the second plebiscite.
If Beazley’s plan is put in place, we may be voting again on the republic issue by 2007. However, a constitutional referendum will again fail unless three key lessons from Australia’s experience with the referendum are heeded. What are these lessons?
The first lesson, and perhaps the most counter-intuitive, is that the constitutional referendum is not there to change the constitution. Rather, the referendum is there to protect the Australian people from proposals for constitutional change with implications that are unsound or unclear. It could also be said that the constitutional referendum is not there to make the constitution more palatable, by adding a preamble for example. The preamble that was voted on in November contained words produced through a highly non-consultative process. By design, the preamble was to have had ‘no legal force’ if the referendum passed (see proposed s.125A). The preamble had symbolism thrust on it by the Prime Minister.
The second lesson is that it is essential to generate a truly federal consensus before and during the referendum campaign. The aim of those campaigning for a change to the Constitution must not simply be a result that just slips over the 50% mark nationally. The double majority requirement of the ratification phase makes this vital. Australia is composed of a federation of States and the different populations of the different States means that, in a sense, a Tasmanian’s referendum vote is worth more than that of someone in Western Australia and a Western Australian’s referendum vote is worth more than that of someone in New South Wales. Analysis of referendum voting suggests that the way the issues of the referendum are shaped by actors and institutions at the State level is frequently an important factor in the way Australians vote at referendums. These facts make it important that forums designed to debate and produce potential referendum proposals are not dominated by people from the south-east corner of Australia (as was the case with the Constitutional Commission of 1988). Such forums ought to contain, for example, equal numbers of delegates from each of the six States and, given the reasoning outlined above, are best not dominated by the major parties at the federal level. This was clearly not the dynamic that operated in the Constitutional Convention of 1998. Further, if the goal is symbolism that is going to generate a federal consensus, then holding conventions in Old Parliament House in Canberra is clearly not the way to go. There must have been quiet laughter on the part of No campaigners who had seen other referendums fail on the back of slogans like ‘Say No to Canberra’ when the Constitutional Convention was held there.
The phased approach of arriving at a final republic model that Beazley has outlined seems a sensible one but the pivotal step is missing. While the Australian people would choose in a plebiscite between a model where the president is directly elected or one where the president is appointed by parliament, the forum for determining what these models will look like is yet to be determined. Mr Beazley has already indicated that the Labor Party, because of the costs involved, would probably not organise another major convention on the republic issue. One thing is certain: the history of referendums in Australia shows that the best place to decide on a republic model is not on the floor of parliament, behind party room doors, or in a joint parliamentary committee room. An alternative forum needs to be hit upon, an alternative that recognises that Australia is a federation. The fact that the majority of people from rural areas and lower income-earners voted against the recent republic proposal was instructive. So too is the fact that the ACT is the only Territory/State that has consistently voted a majority Yes in referendums (that is, since 1977 when the Constitution was altered to incorporate the vote of Australia’s Territories in the national referendum vote). The key to whether a constitutional referendum on the republic issue finally passes will be whether the ‘ordinary Australian’ feels included in the process by which a republic model emerges. It must be a ground-up rather than top-down approach. In order to generate a federal consensus, creative solutions must be found.
The third lesson is that those who want constitutional change must understand and anticipate the advantage of those who want to maintain the status quo in a constitutional referendum. Those who propose a constitutional change are necessarily tied to explaining the relative benefits of the proposed change. This includes arguing that the change will not or can not go further than the provision of these benefits. Those who argue against such a change are not tied in the same way. For example, in the republic referendum, advocates of the Yes case strained very hard to keep us focused on one issue. Quoting Neville Wran: ‘From the republican point of view it is imperative to keep it simple, because the substance of the question is: Do we Australians want an Australian as our Head of State?’ This is further evidenced by the Yes case in the official referendum pamphlet that stated ‘Becoming a Republic simply means having an Australian Head of State instead of the Queen’ (coupled with the issues of national identity that were discussed during the campaign in relation to the preamble, it is clear that for many the November referendum proposals were about matters of the heart).
In taking this approach, advocates of the republic model failed to put at centre stage arguments about the balance of power between the Australian people and their government, the constraints on Executive power and the implications for Australian democracy. In other words, the Yes campaigners chose to make their initial and primary focus a given (and apologies to the monarchists out there): the majority of Australians do want an Australian Head of State.
The failure to put at centre stage arguments about the implications for Australian democracy was a poor campaign strategy because it provided plenty of ammunition to those who didn’t want the referendum to pass. As noted above, people who advocate the status quo for whatever reason (in this instance the ‘direct electionists’ and the monarchists) are not constrained in the same way as are people who want change. Opponents of a proposed change can portray it as an unknown quantity with effects that may go much further than those argued by advocates of change. In the last week of the referendum campaign, Jason Yat-Sen Li complained that whatever the result, we would witness one of the quickest divorces in Australian history when the ‘direct electionists’ and the Monarchists woke up on the Sunday after the referendum. Indeed, many pro-republicans criticised the work of the monarchists in partnership with the ‘direct electionists’ as an exercise in cynicism, but it was effective. Rather than simply muddying the waters, the truth is that they very quickly redefined the focus of the campaign. It moved from being a debate about the nationality of our Head of State to one about the proposed procedures for the appointment and dismissal of the President. This translated into the more general question of how much power rests with politicians, especially the Prime Minister (the contrast in campaign approach may be noted by reading the official cases for and against). This shift has become a very familiar theme in constitutional referendum campaigns in Australia and may be considered a central issue for those campaigning in constitutional referendums to consider and address.
In this article, I have briefly described the three key elements of what may be described as a ‘cycle of inertia’:
• the amendment procedure,
• the referendum record, and
• criticism of the amendment procedure.
It is clear that the underlying force in the ‘cycle of inertia’ is the central role of the Commonwealth government in initiating constitutional referendums. The major party groupings that control the Commonwealth government have been guided by the possibility of gaining some short-term partisan advantage or manipulating the referendum process to achieve certain ends.
The politics of constitutional change should transcend the realm of normal politics. The fact that constitutional change is always subordinate to partisan considerations is therefore symptomatic of a design fault in the machinery for constitutional amendment prescribed in s.128. This fault might be fixed if the power to initiate constitutional referendums were extended to other institutions. In these circumstances, there would be less scope for manipulation of the amendment process by the Prime Minister and his/her party and a greater likelihood of proposals designed to disperse, rather than concentrate political power.
When it comes to the ‘cycle of inertia’ in which the procedure for formal constitutional change is caught, the control over the initiation procedure by political parties is the heart of the matter. The major political parties have been blind to many lessons that can be drawn from Australia’s referendum experience. I have outlined three. The first lesson is that the constitutional referendum is there to protect the Australian people from proposals with implications that are unsound or unclear. The second lesson is that creative solutions are required to generate a truly federal consensus. The third lesson revolves around the need to understand the advantage of those who are campaigning for the status quo in a referendum campaign. These and other lessons must be wrestled with if Australia is to become a republic.
[*] Richard Miles is a Policy Research Analyst at the Western Australian Department of Training and Employment. His MA thesis, entitled 'Partisanship and Explanations of Voting in Australia's Constitutional Referendum' was completed at the University of Western Australia in 1997.
 See the Australian, 13 August 1999, p.4 for Senator Stott Despoja’s statement.
 While some critics have pointed to the severity of the special majority requirement, only five proposals have failed because they did not receive the necessary majorities in the States. In 1974, the desire to make ratification of constitutional referendums easier led the Whitlam government to an unsuccessful attempt to water down the required majorities for a referendum to pass.
 Note that while the final cost of the recent referendums on the Republic and the preamble will not be known until March/April 2000, it was in the order of $91m, excluding the cost of holding the Constitutional Convention of 1998 which cost around $27.5m. Thanks to Ian Stringall from the Australian Electoral Commission for information on the cost of the 1999 referendums.
 The images of the voter in the referendum process are explored in Miles, Richard, ‘Australia’s Constitutional Referendum: A Shield, Not a Sword’, (1998) 35(4) Representation: Journal of Representative Democracy 237-46.
 This evidence is detailed in Miles, Richard, Partisanship and Explanations of Voting in Australia’s Constitutional Referendums, unpublished Master of Arts thesis, University of Western Australia, 1997.
 Note that it was proposals of this kind, on the issue of proportional representation, that succeeded in New Zealand in the early 1990s.
 For an index estimating the relative difficulty of constitutional amendment processes, see Lutz, Donald S., ‘Toward a Theory of Constitutional Amendment, in Sanford Levinson (ed.), Responding to Imperfection: The Theory and Practice of Constitutional Amendment, Princeton University Press, 1995, pp.237-74. Note though that Lutz incorrectly calculated the index of difficulty for Australia to a value of 4.7. It should be 4.25. For comparison of the Australian record of formal constitutional amendment with the United States, see Altman, Dennis, ‘Obstacles to Constitutional Change’, (1979) Australian Quarterly 103-112 at 105; Galligan, Brian, ‘The referendum process’, in A Federal Republic: Australia’s Constitutional System of Government, University of Cambridge Press, 1995, p.121; and Hunt, Erling M., ‘The Australian and American Provisions for Amendment’, in American Precedents in Australian Federation, Columbia University Press, 1930, pp.210-21. For comparison with New Zealand, see Hughes, C.A., ‘Australia and New Zealand’, in David Butler and Austin Ranney (eds), Referendums around the World: The Growing Use of Direct Democracy, Macmillan, 1994, pp.154-7. For comparison with Canada, see Russell, Peter H., ‘The Politics of Frustration: The Pursuit of Formal Constitutional Change in Australia and Canada’, in Bruce W. Hodgins and others (eds), Federalism in Canada and Australia: Historical Perspectives, The Frost Centre for Canadian Heritage and Development Studies, Trent University, 1988, pp.59-85; and Gibbins, Roger, ‘Constitutional Turmoil and Frustration: From Trudeau to Mulroney’ in Michael S. Whittington and Glen Williams (eds) Canadian Politics in the 1990s, Nelson, Canada, 1995.
 See Russell, above, ref 7, p.3.
 Note that Mr Beazley’s plan for the republic is drawn from the Australian, 8 November 1999, p.3.
 See Wran, Neville, ‘The merits of minimalism’, pp.20-6 in John Uhr (ed.), The Australian Republic: The Case for Yes, The Federation Press, 1999, p.20.
 Jason Yat-Sen Li was an elected, independent delegate to the 1988 Constitutional Convention, co-chair of the NSW Yes coalition and federal government appointee to the national Yes committee for the 1999 referendum. The complaint was made in a radio debate broadcast on the ABC, Thursday, 4 November 1999.