Research Brief Index

Research Brief no. 13 2004–2005

Commonwealth Election 2004

Scott Bennett
Politics and Public Administration Section

Gerard Newman and Andrew Kopras
Statistics Section
14 March 2005

Contents

Part One:

Executive Summary
Introduction
Background to the election

House of Representatives electorates
Enrolments
Nominations
The leaders

The battle for the House of Representatives

The state of the parties
A likely Coalition victory?
The campaign battle
Campaigning innovations
The House of Representatives result
The states and territories
Local contests

Adelaide and Hindmarsh (both SA)
Bass (Tas)
Bennelong (NSW) and Mayo (SA)
Braddon (Tas)
Cunningham (NSW)
Greenway (NSW)
Parramatta (NSW)
Richmond (NSW)
Wentworth (NSW)
The ‘Scoresby’ seats
The benefits of incumbency?
The personal impact of electoral redistributions

The battle for the Senate

The state of the parties
The Greens under attack
The Coalition gains control
The Greens, Labor’s preferences and Family First
‘Vote for Me’

Over the longer term

Coalition versus Labor
Liberal versus National
What can Labor do?
Beyond the city centre
Labor and the Senate
Preferential voting

Electoral system problems

Postal votes
Informal votes

The aftermath
Appendix: The passing parade

House of Representatives

Table1: Retiring members
Table 2: Defeated members
Table 3: New members

Senate

Table 4: Retiring Senators
Table 5: Defeated Senators
Table 6: New Senators

Women

Table 7: Women elected
Endnotes

Part Two:

Symbols and abbreviations

Table 8: House of Representatives: National summary
Table 9: House of Representatives: State summary
Table 10: House of Representatives: Regional summary
Table 11: House of Representatives: Party Status Summary
Table 12 House of Representatives: Socio-Economic Status Summary
Table 13a: House of Representatives: Electoral division summary
Table 13b: House of Representatives: Electoral division summary
Table 14: House of Representatives: Electoral division detail
Table 15: House of Representatives: Two-party preferred vote, state summary
Table 16: House of Representatives: Two-party preferred vote, regional summary
Table 17: House of Representatives: Two-party preferred vote, party status summary
Table 18: House of Representatives: Two-party preferred vote, Socio-economic status summary
Table 19: House of Representatives: Two-party preferred vote, Electoral division summary
Table 20: House of Representatives: Electoral pendulum
Table 21: House of Representatives Election: Electoral divisions ranked by two-party preferred swing to LP/NP
Table 22: Senate: National summary
Table 23: Senate: State summary
Table 24: Senate: Composition after 1 July 2005
Table 25: Senate: Candidate details
Table 26: Comparison of Senate and House of Representatives votes
Appendix 1: Electoral division classification
Appendix 2a House of Representatives elections 1949–2004
Appendix 2b Senate elections 1949–2004

Executive Summary

The main question in an election is whether or not enough voters are prepared to support the party that is challenging the government. If that government is in a position of strength, there may be little the challenger can do, other than go through the campaigning motions. The 2004 Commonwealth election appeared to be a similar case.

Although at the time of the announcement of the election, Labor appeared to have a chance of winning, the position soon altered, and the Government remained comfortably ahead of its rival in first preference terms for much of the time until polling day. The impression of there being solid primary vote support for the Coalition was reinforced by opinion poll figures indicating the consistent support for Prime Minister John Howard of about half of those polled. Opposition leader Mark Latham also polled consistently well on polls exploring satisfaction levels for performance in office, but Howard’s popularity remained ominously high from Labor’s perspective.

Another relevant measure seemed to be the question of who was seen as the preferred Prime Minister. Approximately one-third of voters preferred Latham, but Howard’s ‘preferred Prime Minister’ figure remained close to the 50 per cent mark. It seems clear that many of those polled saw merit in Latham’s performance as leader, but it is also clear that the Prime Minister had retained a great deal of support, despite his length of time in office. Howard was not showing the drop in popularity that many leaders experience. Here also was evidence suggesting that Labor was probably not going to win in 2004.

Despite this, the general impression conveyed through the media and polling bodies was that the election race was virtually too close to call. So close did the contest seem, that some observers speculated about the possibility of a hung parliament. Various journalists greeted the campaign with the assertion that the campaign would be ‘crucial’. But with Australia possessing a healthy economy bolstered by low interest rates, and with living standards continuing to rise, the Opposition would probably have needed some catastrophic collapse by the Government for it to win.

The campaign was the usual babble wherein government and opposition attempted to attract votes by promises, by warnings and by threats. Despite this, it is quite likely that few votes shifted. The Government attempted to ensure such a result by a preparedness to spend on projects that it saw as necessary to match or better promises made by Labor—what one journalist described as ‘repeated tactical vote-buying from the surplus’. The Government’s hand remained steady throughout, while there were sufficient uncertainties in Labor’s performance and policies to make it unlikely that it would shake enough voters loose from the Government’s grip.

In the event, in the House of Representatives the Coalition has been left in a stronger position than before the election. Apart from increasing its House majority from 14 to 24 seats, its marginal seats have been reduced, with many now a great deal harder for Labor to win at the next election. By contrast, seats that are marginal for Labor have increased, suggesting that if Labor’s vote does not increase at the next election, it could lose more seats. Labor’s first preference vote of 37.6 per cent was its lowest vote since the elections of 1931 and 1934.

The election confirmed a steadily strengthening Liberal position, which suggests that future non-Labor victories are quite likely to see the Liberals gaining a parliamentary majority in their own right. This was actually achieved in 1996; in 2004 the Liberals were just one seat short. If the Liberals start doing this regularly, there may be increased internal pressure to cut the Nationals adrift. From the Nationals’ perspective, this alteration in the relationship between the Coalition partners has been exacerbated by the apparent permanent loss of seats by the rural party to the Liberal Party.

The result made clear the trough that the Labor Party had been in since before the 1996 Commonwealth election. With Labor failing to achieve even 40 per cent in five of the six elections since 1990, its poor electoral health can be more clearly appreciated. The 2004 election should therefore be seen as part of Labor’s long-term troubles rather than a failure of its campaign. The need for the party from now to the next election is to increase its vote to a winning first preference figure; the dilemma is establishing how this can be achieved.

In the Senate contest, the Coalition won half the Senate seats in five states and the territories, but won four seats in Queensland. This unique event was due to at least two circumstances—the increased Liberal vote in the state, plus the ability of the Nationals to gain enough votes to win a seat. The Coalition thus won 19 state Senate seats in 2004. When combined with its two territory seats, and the 18 state senators due to retire in 2008, the Coalition will hold 39 of 76 Senate seats from 1 July 2005. This will be the first time a government has controlled the Senate since 30 June 1981.

Despite earlier claims by Senator Bob Brown that the Greens would gain the balance of power in the Senate, the Coalition performance across the nation was too even to make that a realistic outcome. The Greens won Senate seats in Western Australia and Tasmania, and came relatively close in Queensland and Victoria. The new Family First party won a Victorian seat and came close to a seat in Tasmania.

The Australian Democrats lost all of its seats being contested, and will have only four seats in the new Senate—the same number as the Greens. The single One Nation Senator was defeated.

Among the trends emerging from the election were:

• the figures over the past two decades suggest that the problems of the Labor Party have been as much long-term as a number of particular electoral failures
• a measure of Labor’s decline is that it is now the major beneficiary of the compulsory allocation of all preferences on House of Representatives ballot papers
• the Liberal Party now has a strong grip on the seats surrounding Sydney, and if it maintains this, it will continue to be difficult for the Labor Party to regain office
• Labor has seemingly lost touch with many of its traditional supporters, a touch that it needs to regain
• the shifting electoral strength of the Coalition partners suggests future non-Labor victories are quite likely to see the Liberals gaining a House majority in their own right, and
• Labor not only needs to improve its House performance, for its performance in Senate elections since the last increase in the size of the Senate has been weaker than its performance in House of Representatives elections.

Two aspects of election management revealed problems: postal vote arrangements in Queensland were found wanting, and an increasing informal vote, especially in Sydney, is a matter of concern.

Introduction

This paper follows a similar format to the Parliamentary Library studies of the 1998 and 2001 Commonwealth elections.⁽¹⁾ Once again the paper is divided into two parts.

Part One is written by Scott Bennett of the Politics and Public Administration Section. It is written partly as a journal of record, partly as a discussion of the election campaign, and partly as a discussion of the election outcome. An appendix lists departing Members of the House of Representatives and Senators and their replacements.

Part Two comprises a comprehensive set of statistics compiled by Gerard Newman and Andrew Kopras of the Statistics Section. Tables contain national, state and regional vote summaries, details concerning electoral divisions, two-party preferred figures, and the party strengths in the two houses of the Commonwealth Parliament. Two appendices complete this section of the paper. The first shows the classification for each electoral division for the various classifications used in the brief. The second gives figures for Senate and House of Representatives elections held between 1949 and 2004.

Background to the election

The lead-up to the 2004 Commonwealth election was notable for the intense speculation that built up over the possible date. The first question was whether or not there would be a double dissolution election over the issue of Telstra, a debate that died as the year wore on. Many in the press then became certain that an election would be held on 7 August, the first available date for a joint House of Representatives and half-Senate election. This certainty then shifted to 4 September and then moved on to 18 September. The Australian was amused, noting that the speculation did have one positive political outcome for the Prime Minister: ‘it is driving journalists short of a real story absolutely nuts’.⁽²⁾

Eventually, 9 October 2004 was announced as polling day. The official process began with the Prime Minister’s announcement on 29 August 2004 and was programmed to conclude no later than 8 December 2004:

2004 election timetable

House of Representatives electorates

Queensland had gained a House of Representatives seat (Bonner) and South Australia had lost a seat (Bonython) as the result of redistributions in those states.⁽³⁾ The Northern Territory had been projected to lose one of its two seats, but legislation passed by the Parliament ensured that the Territory would retain both for at least the 2004 election.⁽⁴⁾ The number of House of Representatives seats therefore remained at 150.

Enrolments

The Australian Electoral Commission defines the polling day enrolment figure as:

… enrolment at the close of rolls with subsequent adjustments such as the removal of the names of electors who have died after the close of rolls, and the reinstatement of eligible electors previously removed from the roll.⁽⁵⁾

As is usual, therefore, there were two sets of enrolment figures—the figure when the rolls closed on 7 September and the figure on polling day, which was substantially larger. When rolls closed on 7 September 2004 there were 13 021 230 electors enrolled, an increase of 384 599 (3.0 per cent) on the previous election. By polling day, this figure had increased by more than 70 000 voters. The largest electorate was Fraser (ACT) and the smallest was Solomon (NT).

Enrolment figures

Source: Australian Electoral Commission

Nominations

A total of 1421 nominations was received, 97 more than for the 2001 election—but 14 fewer than for the 1998 election. There were 1091 House of Representatives nominations (+52) and 330 Senate nominations (+45). There were 52 parties represented, three more than in 2001 and 16 more than in 1998.

Women made up 28.6 per cent of all nominations, a fall of 0.2 percentage points. Of these, 299 nominated for the House of Representatives (27.4 per cent) and 107 nominated for the Senate (32.4 per cent).

• Graeme Campbell, independent candidate, Kalgoorlie (WA)—former ALP, later Independent, Member for Kalgoorlie (1980–98)
• Susan Davies, ALP candidate, La Trobe (Vic)—Independent Member of the Legislative Assembly for Gippsland West, 1997–2002
• Brian Deegan, independent candidate, Mayo (SA)—magistrate and father of a Bali bombing victim
• David Ettridge, independent candidate, Senate (NSW)—former One Nation organiser
• Peter Garrett, ALP candidate, Kingsford Smith (NSW)—environmental activist, singer
• Pauline Hanson, independent candidate, Senate (Qld)—founder One Nation Party, former One Nation Member for Oxley (1996–98)
• Rev. Fred Nile, Christian Democratic Party candidate, Senate (NSW)—Christian Democratic Party Member of the Legislative Council (NSW) since 1981
• Michael Ronaldson, Liberal candidate, Senate (Vic)—former Liberal Member for Ballarat (1990–2001)
• Malcolm Turnbull, former chair of the Australian Republican Movement
• Andrew Wilkie, Green candidate, Bennelong (NSW)—former Office of National Assessments employee, and
• six Senate candidates, one per state, who were associated with a Channel Seven ‘reality’ television show (see below pp. 28–9).

The leaders

John Howard (Bennelong, NSW) led the Liberal-National coalition into an election for the fifth time, second only to Robert Menzies’ nine campaigns (1940, 1946–1963). The Nationals’ John Anderson (Gwydir, NSW) was contesting his second election as party leader, while for Mark Latham (Werriwa, NSW) it was his first election as Labor leader.

The battle for the House of Representatives

The state of the parties

As a result of the Cunningham by-election (2002) and redistributions in Victoria, Queensland and South Australia, the notional position of the parties had altered slightly since the 2001 election, with the Coalition in a slightly stronger position, and the Opposition slightly weaker:

House of Representatives—party status

Source: Australian Electoral Commission

The Government therefore held 82 of the 150 seats in the House of Representatives at the end of the 40^th Parliament. Although a Coalition loss of just seven seats would produce a hung parliament, it required a nett increase of 13 seats for Labor to gain control of the House of Representatives.

The Australian Greens, united by the decision of the Greens of Western Australia (GWA) to join the national party, were expected to increase their vote well above the five per cent  they and the GWA had gained in 2001. Their leader, Senator Bob Brown (Tas), predicted one million votes for the party, well in excess of the 569 075 House of Representatives votes in 2001.⁽⁶⁾ Such a return would probably see their House of Representatives preferences playing a major part in the outcome of the struggle for government. They hoped to retain Cunningham (NSW).

A likely Coalition victory?

The main question in an election is whether or not enough voters are prepared to support the party that is challenging the government. If that government is in a position of strength, there may be little the challenger can do, other than go through the campaigning motions, as was the case in 1977 and 1984. The 2004 election appeared to be a similar case. At the time of the announcement of the election, it seemed there were enough voters prepared to turn their backs upon the Howard Government to bring about its defeat. Newspoll had the ALP’s primary vote (42 per cent) three points ahead of the Coalition. Its two-party preferred margin was a remarkably healthy 8 percentage points:

Voting intention and two-party preferred  figures July–October 2004 (Newspoll) (per cent)

Source: http://www.newspoll.com.au/home.html

It was soon clear, however, that the 13–15 August poll was an aberration. In the following poll, taken 27–29 August, the position was reversed, with the Government now apparently three percentage points in the lead, and it remained comfortably ahead of its rival in first preference terms for much of the time until polling day. Between the 27–29 August Newspoll and the final Newspoll before the election, the Coalition averaged 44.4 per cent to Labor’s 39.9 per cent of first preferences, indicating that the Government was always likely to be victorious.⁽⁷⁾ The Coalition was, in fact, quite likely to gain more votes than that, due to the fact that National Party opinion poll figures are likely to be lower than in a general election.

Although it thus seemed unlikely that there would be enough voter support to remove the Coalition from office, the impression conveyed through the media and polling bodies throughout the election campaign was that the election race was virtually too close to call. The polls seemed to confirm what one journalist described as ‘a distinct air of a fading government’.⁽⁸⁾ So close did the contest seem, that some observers, including the pollster, Gary Morgan, were prepared to speculate about the possibility of a hung parliament.⁽⁹⁾ This is despite the fact that the first preference figures generally seemed to suggest a comfortable margin for the Government.

Why should this be so? The answer seemed to lie with the media preference to focus on the two-party preferred calculation rather than the first preference figure. The table above indicates that on six occasions in the three months prior to election day, the ALP two-party preferred figure was ahead of that for the Government, sometimes by a wide margin. On four other occasions there was a two-party preferred tie. In only one poll (1–3 October) was the Government ahead, and then by the narrowest of margins. If one chose to highlight the two-party preferred poll figures, it would therefore have seemed quite possible that the ALP would win the election—or that the result would be extremely close.⁽¹⁰⁾

The constant media emphasis on the two-party preferred figure meant that many observers overlooked how much the Labor first preference vote remained in the trough it had been in since before the 1996 Commonwealth election. In the four elections of the 1980s, Labor’s first preference vote averaged 47 per cent, contrasting with the 39.8 per cent the party has gained in the six elections since. Elections can be won with a first preference vote of less than 40 per cent (1990 and 1998), but such victories are rare. To speak of Labor ‘leading’ at any stage in the last three months of the campaign therefore seemed anomalous when it was clear that the party’s first preference figure was very low during that time, whereas the Coalition seemed likely to receive a first preference vote of about 45 per cent.

How many votes are needed to win a House election? If we look at all elections since 1958, we find that the average first preference vote of the winning party or coalition has been 46.3 per cent. ⁽¹¹⁾Only in 1961, 1990, 1998 and 2001 did the winners secure less than 43 per cent of first preferences; only in 1990 did a party winning 43 per cent of the vote fail to win office.⁽¹²⁾ Even though recent elections have seen a fall in the average first preference vote of the winning party—the average has been 43.5 per cent since 1990—a party has much more chance of success if its vote reaches at least 43 per cent than if it falls below. With Labor failing to achieve even 40 per cent in five of the six elections since 1990, its electoral trough can be more clearly appreciated. A 43 per cent first preference return was not guaranteed to hand government to either side, but unless Labor could lift its 2004 first preference figure at least to that level, its chances of regaining office seemed slight. According to Professor Dean Jaensch, Labor’s first preference vote needed to be much higher ‘to have any real hope of winning government’.⁽¹³⁾ As former Labor Senator, Peter Walsh, put it, a party that cannot get a vote above 40 per cent is unlikely to win a Commonwealth election.⁽¹⁴⁾

Perhaps election analysts have become used to focusing on the two-party preferred vote, and now tend to put insufficient weight on the first preference figure. The two-party preferred figure, after all, with its over-simplified reduction of the contest to just government versus opposition, is less complicated for the analyst to deal with. On polling day 2004, the Weekend Australian gave the final Newspoll figures with a story headed: ‘Latham within striking distance’.⁽¹⁵⁾ This was despite the fact that Newspoll gave the Coalition’s first preference lead as six points (39–45 per cent). Labor’s Newspoll figure in the final six polls had fluctuated between 39 and 41 per cent, while the Coalition readings were between 43 and 46 per cent (see table p. 11). The Weekend Australian indicated why the election was being called in this fashion. Having noted that the parties were apparently equal on the two-party preferred vote, the journalist explained that although the Coalition lead on first preferences would ‘normally’ see it returned to office, it was the fact that people’s likely preference allocation would strongly favour Labor that ‘could neutralise the primary vote’.⁽¹⁶⁾ Clearly the two-party preferred vote was being focussed on rather than the first preference vote.

The Newspoll figures during the three months prior to the election showed little alteration in the first preference margin between the parties, nor was there any evidence of any voter volatility. The figures suggested, then, that as the Coalition’s first preference vote was healthy, the most likely result was a Government victory.

The impression of there being little movement in voter intention was reinforced by opinion poll figures indicating the consistent support for the Prime Minister of about half of those polled—as journalist Paul Kelly noted later, ‘the people never really rejected him’.⁽¹⁷⁾ Opposition leader Latham also polled consistently well on polls exploring satisfaction levels for performance in office, but his performance made no apparent dent in Howard’s standing:

 Leader satisfaction (per cent)

Source: http://www.newspoll.com.au/home.html

It seems clear that many of those polled saw merit in Latham’s performance as leader, but it is also clear that the Prime Minister had retained a great deal of support, despite his length of time in office. Howard was not showing the drop in popularity that many leaders experience. Here also was evidence suggesting that Labor was probably not going to win in 2004.

Another relevant measure seemed to be the opinion poll question of who was seen as the preferred Prime Minister. Generally Leaders of the Opposition poll behind Prime Ministers in this type of poll, but there have been some exceptions. In 1998, for example, Kim Beazley out-rated Prime Minister Howard 41 to 40 per cent at the time of the election, after having kept relatively close in previous polls.⁽¹⁸⁾ In 2004, however, Mark Latham was seen as preferred Prime Minister by only one-third of voters, by contrast with John Howard, whose ‘preferred Prime Minister’ figure hovered steadily close to the 50 per cent mark. Here also, there was no evidence of voter volatility:

Preferred Prime Minister (per cent)

Source: http://www.newspoll.com.au/home.html

To sum up this poll evidence, throughout the campaign it seemed that there was enough general support for the Government to ensure its victory. Labor would have been hard-pressed to persuade enough voters to change their preference. Three weeks before polling day, Professor John Wanna put it thus:

The nationwide polls—and in particular Newspoll’s findings over the past three months—indicate that a substantial proportion of the electorate firmly made up their minds months ago about how they are going to vote. ⁽¹⁹⁾

The campaign battle

Despite this, as is now usual, various journalists greeted the campaign with the assertion that the campaign would be ‘crucial’—or as Michelle Grattan put it, the campaign would produce:

… an unpredictable and exciting election in which neither leader starts with a decisive break, either could take the prize, and absolutely anything could happen in between.⁽²⁰⁾

The Prime Minister had announced the election on 29 August with the words:

This election will be about trust. Who do you trust to keep the economy strong and protect family living standards? Who do you trust to keep interest rates low? Who do you trust to lead the fight on Australia’s behalf against international terrorism? Who do you trust to keep the Budget strong so that we can afford to spend more on health and education?(21)

This encapsulated much of the Government’s strength and the Opposition’s problems. With a healthy economy bolstered by low interest rates, and with living standards continuing to rise, the Opposition would probably need some catastrophic collapse by the Government to move to the Treasury benches.

As is now usual in the lead up to an election, a leaders’ television debate was held, on 12 September. Commentators were generally impressed by Latham’s performance.(22)

The perceived importance of Queensland to the result was clear in the fact that both parties held their formal policy speeches in Brisbane, the Coalition on 26 September, and Labor three days later, both much later than once was the case. This is a reflection of the fact that MPs can claim travelling allowances only until their parties’ formal campaign launch, after which time they or their parties must pick up the tab.(23) A measure of this lateness can be seen in the dates when policies were issued. By the time of the Coalition launch, for example, 27 of a total of 73 policies had already been announced, and only 12 were issued at the launch. Forty-three of the Labor Party’s 62 policies had been made public before their launch, at which only three saw the light of day.(24)

Like all campaigns, the battle was fought between two sides determined to attract voters with their policies while endeavouring to counter their opponents’ offerings, all the time attempting to avoid any serious mistakes. Essentially, the Government’s hand remained steady, while there were sufficient uncertainties in Labor’s performance to make it increasingly unlikely that they would shake enough voters loose from the Government’s grip:

• Labor released a number of complex policies very late in the campaign, the detail of which would probably have been too difficult for voters to digest.⁽²⁵⁾ The complexity of Labor’s proposal for the restructuring of family payments was held to be one example⁽²⁶⁾
• Labor’s Medicare Gold proposal for free hospital care for Australians over 75 years was criticised for being unrelated to need.⁽²⁷⁾
• according to its critics, Labor’s forest policy, particularly in relation to Tasmania, was a sign of the party being too close to the Australian Greens, and the opposition of people like the Tasmanian Labor Premier was an indication of a party divided. Professor Rodney Tiffen of the University of Sydney has suggested the late announcement of the policy was Labor’s major tactical error, which allowed the Prime Minister to claim to a large public meeting of Tasmanian timber workers that the Liberal Party was the only party concerned about the jobs of ‘ordinary’ Australians,⁽²⁸⁾ and
• many Labor advertisements focused on a perceived voter dislike of Treasurer Costello—elect Howard, they said, and in due course the nation would gain Costello as Prime Minister. Critics, including the Liberal Party’s federal director, suggested that this missed the point that the Treasurer probably had a much higher standing in the community than the Labor Party gave him credit for.⁽²⁹⁾

Labor was probably also hurt by specific Coalition criticism:

• Latham’s period as Mayor of Liverpool was described as a disaster, and hardly the background that Australia needed for its Prime Minister. Advertisements featured a learner driver’s L-plate as the first letter of the Labor leader’s name⁽³⁰⁾
• Latham’s promise to bring Australian troops home from Iraq by Christmas was held to be inappropriate to Australia’s diplomatic and defence needs⁽³¹⁾
• the way in which Labor’s policy for the funding for private schools was spelled out seemed designed to punish those parents who chose to send their children to particular private schools⁽³²⁾
• Labor’s preference arrangement with the Greens was described as being in the thrall of a party that was described as potentially dangerous for Australia’s future⁽³³⁾ (for more on views critical of the Greens, see below, p. 26), and
• the Coalition spoke continually of the higher interest rates that were experienced during the Hawke–Keating period. They warned that the financial management of a Labor government was certain to produce higher interest rates, and thus greater pressure on home-owners with a mortgage: ‘The reality is that Labor governments over the last 30 years have always delivered higher interest rates’.⁽³⁴⁾ This was an attack that the Labor Party did surprisingly little to rebut, which Tiffen sees as the party’s major strategic error.⁽³⁵⁾

Throughout the campaign the Government showed its preparedness to spend money on projects that it saw as necessary to match or better promises made by Labor—what a journalist described as ‘repeated tactical vote-buying from the surplus’.⁽³⁶⁾ The Coalition also made many extravagant promises in regard to projects that were designed to attract or retain voters.⁽³⁷⁾ The extent to which this was done was remarkable for a Coalition Government. The Liberal Party, in particular, has long described itself as the main protector of the Australian federal system of government from Labor ‘centralists’. But in 2004 it was quite prepared to promise work on projects such as $800 toolboxes for apprentices, the construction of new private technical colleges, or the improvement of some bridges and roads in marginal seats, matters that clearly were areas of state power and responsibility. ⁽³⁸⁾

So, the campaign was the usual cacophony wherein government and opposition attempted to attract votes by promises, by warnings and by threats. Both parties spoke throughout as if the vote was continually fluctuating. Labor ‘insiders’, for instance, blamed a Labor ‘loss of momentum’ on the Jakarta bombing on 9 September.⁽³⁹⁾ Despite such views, as mentioned earlier, it is probable that few votes shifted, due to most voters having already made up their minds about the parties. This was the view of social analyst, Hugh Mackay, whose qualitative research throughout 2004 suggested that enough voters were satisfied with the Howard Government’s performance to ensure its re-election. More specifically, Mackay’s research suggested that as economic and security issues were central to the election, and the Government occupied ‘the high ground’ on both, it was unlikely to be defeated.⁽⁴⁰⁾ As Labor MP, Bob McMullan, put it, his party simply ‘hadn’t made the case for change’.⁽⁴¹⁾ To such a view, Professor Ian McAllister of the ANU added the point that voters see little difference between the major parties:

If voters see little difference between the parties, it is hardly surprising that they don’t bother to examine policies released just weeks or even days before a poll. In such a situation, they will opt for the party that has brought them sound economic management. ⁽⁴²⁾

Campaigning innovations

The 2004 campaign saw a further impact of technology on campaigning practices.⁽⁴³⁾ Of most interest, and causing some controversy, was the use of pre-recorded telephone messages by the Prime Minister and Treasurer —‘advocacy calls’ as the Liberals called them. Although Labor’s Wayne Swan described this ‘bizarre’ tactic as ‘another dirty trick’ imported from the USA, some voters expressed interest in it.⁽⁴⁴⁾ Others were less happy, especially as some messages apparently were made to silent numbers. The Liberals’ Mark Textor claimed that advocacy calls actually aided the Liberal effort: ‘… people appreciated the fact that they got a direct and unfiltered message from a political leader in a new, effective way’. The Liberals also claimed that the calls helped them win six seats.⁽⁴⁵⁾

It had been expected that the parties would use the Internet far more than had been the case in 2001. This occurred, although it seemed to be truest of the minor parties which saw the Internet as a cheap method to contact voters and to publicise their views. The Australian Democrats were said also to have used a viral email facility to spread their message.⁽⁴⁶⁾ A few Liberal MPs, including the Prime Minister, used spam emails to send messages to voters in their electorates.⁽⁴⁷⁾

The House of Representatives result

This election was notable for the fact that there were significant seat gains and losses on both sides of the House of Representatives. The Liberal Party won nine seats (the new electorate of Bonner plus Greenway, McMillan, Bowman, Hasluck, Stirling, Kingston, Bass and Braddon), but also lost three seats (Parramatta, Adelaide and Hindmarsh). Although Labor lost eight seats (Greenway, McMillan, Bowman, Hasluck, Stirling, Kingston, Bass and Braddon), it picked up five (Cunningham, Parramatta, Adelaide, Hindmarsh and Richmond). The Nationals (Richmond) and the Greens (Cunningham) each lost a seat.

The Liberal first preference vote of 40.5 per cent was 3.4 percentage points higher than in 2001. This was the party’s highest first preference vote since the landslide of 1975 (41.8 per cent), and only the fourth time since its creation that the party had secured 40 per cent of the national total. The party’s first preference share rose in 112 of the 133 seats it contested. The 2001 and 2004 results more than restored the Liberal vote lost in 1998: the 2004 vote of 40.5 per cent exceeded the 1996 tally by 1.8 percentage points. By contrast, the Nationals’ vote of 5.9 per cent only moved slightly (+0.3 percentage points)—their third consecutive return under 6 six per cent. Their vote rose in 18 of the 24 seats they contested.

Overall, the Coalition was left in a stronger position than before the election. Apart from increasing its House majority from 14 to 24 seats, its marginal seats—that is, seats that would be lost on a five per cent swing—had been reduced to 17, with many now a great deal harder for Labor to win at the next election.

Labor’s first preference vote of 37.6 per cent was its lowest vote since the elections of 1931 and 1934, when the ALP and Lang Labor were in competition. It was the lowest vote by a united Labor Party since the 1906 election. Its first preference vote fell in 82 of the 150 seats it contested. By contrast with the Coalition, seats that are marginal for Labor have increased to 20, suggesting that if its vote does not increase at the next election, it could lose more.

The Australian Democrats almost disappeared in the House of Representatives contest. They did not contest all 150 seats as they had done in 2001, though they still nominated 125 candidates. The party’s first preference vote fell by more than 475 000 votes or 4.2 percentage points; their average votes per seat fell from 4134 to just 1158 votes.

As in 2001, the Australian Greens contested all electorates. Their first preference vote of 7.2 per cent was 2.2 percentage points higher than in 2001, though the party’s vote tally of 841 734 votes was well below Senator Brown’s predicted one million. This effort was therefore described by some writers as a failure, but the party actually increased its first preference vote in 146 of the 150 seats, and Green preferences helped elect Labor MPs in at least 21 seats.⁽⁴⁸⁾ Ironically, one of the five where their vote fell was in Cunningham, where their first preference return was 2.9 percentage points lower than when they won the seat in the 2002 by-election. Cunningham was lost.

The states and territories

The Liberal and Green votes increased in every state and territory; Labor’s rose in three states and both territories:

Party votes by state and territory (per cent)

Source: Australian Electoral Commission

As noted, four seats changed hands in New South Wales. The Liberal Party has not topped 40 per cent in this state since 1966, but its vote of 36.2 per cent was its best tally since 1975. Labor has now been below 40 per cent in three of the last four elections, its lowest returns since 1940. The Nationals’ vote did not move, while the Greens enjoyed their second-highest state return with a 3.3 percentage point lift in their vote. Since 1983, about one-third of Labor’s total House membership has come from New South Wales—yet its overall performance in the largest state has weakened considerably. Between 1983 and 1993 it won 58.5 per cent of all New South Wales seats, but this has declined to 41.5 per cent in the elections since. This has been the result of a marked decline in its vote—47.8 per cent during 1983–87 (Liberal 32.6 per cent) to 40.4 per cent in the elections after 1987 (Liberal 32.5 per cent). In 2001 and 2004, the Labor first preference vote was only 36.6 per cent, barely ahead of the Liberals’ 34.9 per cent, the closest the parties have been since 1966. Labor’s difficulties can be seen particularly in the outer Sydney region, once an important component of its New South Wales electoral strength (see below pp. 31–2).

Only one seat changed hands in Victoria. Labor gained its second-highest state return here, though its vote actually fell (-1.3 percentage points), and its total first preference vote was its poorest since 1990. Although the state is only the Liberals’ third-best in terms of votes won, their 43.2 per cent was the best return since 1954. The Nationals’ vote increased slightly (+0.4 percentage points). The Australian Democrat vote fell sharply (-5.1 percentage points), whereas the Green vote increased slightly to 7.5 per cent.

One seat changed hands in Queensland. The Labor vote was remarkably low when one considers the party’s strength in the 2001 and 2004 state elections. In national elections between 1980 and 1993, Labor managed an average Queensland vote of 43.4 per cent. The last four elections have seen this average drop to an average of 34.7 per cent, with these elections returning Labor’s four lowest Queensland votes since the 1901 election. The Liberals, on the other hand, have recorded their three highest Queensland votes in these same four elections, with their 39.4 per cent in 2004 being their highest vote in the state since the party’s founding in 1944. The Liberal average vote since 1996 has averaged 36.5 per cent, 1.8 percentage points higher than Labor, despite the party’s contesting significantly fewer electorates than its opponent. In fact, the Liberals have outstripped Labor in those seats they have both contested:

Liberal and Labor votes, Queensland 1996–2004 (per cent)

Source: Australian Electoral Commission

Despite the Nationals’ vote climbing in 2004 (+0.6 percentage points), their last six elections have produced their six lowest votes since 1949. Queensland is the poorest state for the Greens (5.1 percentage points). The One Nation vote fell by 5.1 percentage points but still the party secured 2 per cent of the vote.

Two seats changed hands in South Australia. The Liberal vote rose slightly, giving the party its second-best return since 1975. The Labor vote improved more than in any state (3.1 percentage points), but its last six electoral returns have been its worst six in South Australia since 1934. In 2001, the Australian Democrats gained 10.5 per cent of the vote; three years later their vote crashed to 1.9 per cent. The state is the second-poorest for the Greens (5.4 per cent).

Two seats changed hands in Western Australia, where the Liberal vote climbed by the greatest amount in any state (6.7 percentage points), giving the party its best return since 1993. Labor’s vote was its worst since 1977, said to have been a consequence of an unpopular state government and the absence of Kim Beazley as the national leader.⁽⁴⁹⁾ In the 1980s, its first preference vote averaged 47.4 per cent; its vote since has been 36.2 per cent. The Green vote increased to 7.7 per cent (+1.7 percentage points).

Two seats changed hands in Tasmania. This was Labor’s highest state vote (44.6 per cent), but it also lost two seats. The Liberal vote jumped 4.9 percentage points, but was still relatively weak at 42 per cent. As expected, Tasmania produced the highest state Green vote (9.9 per cent).

Labor’s vote jumped more in the Australian Capital Territory than in any other jurisdiction (3.3 per cent), with its vote topping 50 per cent for the third time in the past six elections. The Liberal vote also rose (2.8 percentage points) but it was still 15.1 percentage points behind Labor, which won both seats. The highest Green vote (10.8 per cent) and the greatest increase in the Green vote (3.7 percentage points) also occurred in the ACT.

The major parties and the Greens all saw their votes increase in the Northern Territory, with the CLP and the ALP sharing the two seats. The One Nation vote of 8.1 per cent in 1998 has disappeared.

Local contests

Adelaide and Hindmarsh (both SA)

In the 1993 election, Trish Worth’s victory in Adelaide, and Chris Gallus’s victory in Hindmarsh, were both narrowly achieved and unexpected. At the next three elections, Labor always seemed to have a good chance of beating both Members, yet they retained their seats on each occasion. In Adelaide, Parliamentary Secretary Worth’s first preference vote was never high, averaging just 45.2 per cent between 1993 and 2001. Gallus’s average vote at the same elections was 47.1 per cent. In 2004, Adelaide was on a notional margin of 0.6 per cent and the margin was just 1 per cent for Hindmarsh—they were the fourth and fifth most marginal Liberal seats across the nation. Labor therefore once again gave itself a good chance of winning these inner Adelaide electorates, particularly as Gallus was retiring from Parliament and its candidate was well known from the previous two elections. Although Labor trailed in both seats on first preferences, its candidates were successful, probably due to the party’s benefiting from the collapse in the Australian Democrat vote. In eight Adelaide electorates, the Democrat vote fell by 9.6 percentage points, with the Labor vote rising by 5 percentage points. With two-party preferred margins of 1.4 per cent in Adelaide and 1 per cent in Hindmarsh, these two South Australian electorates remain among the most marginal in Australia.

Bass (Tas)

Since 1975, Bass has been held by the Liberal Party more often than not. In recent elections it has become rather less safe than previously, with Labor’s Michelle O’Byrne winning the seat in 1998 by only 78 votes. In 2001 she increased her 1998 margin, but her first preference vote of 42.8 per cent was still only 1.3 percentage points ahead of her rival and five points behind Labor’s statewide vote. As a seat with an important timber sector, and as Labor’s eighth most marginal electorate, Bass was vulnerable.

Prime Minister Howard visited this electorate four times during 2004, and from July 2003 it received $12.2 million in Regional Partnership grants, the largest total for any Australian electorate.⁽⁵⁰⁾ Given the final margin, it is likely that Bass was lost before 7 October. But when, on that day, the Prime Minister told over 3000 timber workers in a packed Albert Hall that their jobs would not be sacrificed, he probably made Labor’s loss certain.⁽⁵¹⁾ To make sure, the Forest Industries Association of Tasmania ran full-page advertisements in the Examiner on 8 and 9 October, stating ‘Howard kept his word!’⁽⁵²⁾ Tasmanian Premier, Paul Lennon, also made O’Byrne’s task difficult when he stated that ‘nothing less than full support’ for the Regional Forests Agreement was acceptable to his government.⁽⁵³⁾ Liberal challenger, Michael Ferguson, increased his party’s first preference vote by 7.7 percentage points to 49.1 per cent, while O’Byrne’s vote fell to 39.2 per cent, her party’s poorest performance in the seat since 1990.

Bennelong (NSW) and Mayo (SA)

The seats held by the Prime Minister and the Minister for Foreign Affairs were unlikely to be lost, but each was of interest because of the presence of a well-publicised candidate who had been very critical of the Government. Both challengers did remarkably well, though their efforts had little impact upon the chances of the incumbents.

Andrew Wilkie, formerly of the Office of National Assessments, stood as an endorsed Green candidate in Bennelong. He won 16.4 per cent of the vote, a jump of over 12 percentage points in the Green vote. The Prime Minister’s vote fell to 49.9 per cent, a drop of 3.2 percentage points, and his second-lowest vote since winning the seat in 1974. The Labor vote also fell (-2.5 percentage points).

Brian Deegan, Adelaide magistrate and father of a Bali bombing victim, nominated as an independent in Mayo. Deegan managed to win 15.2 per cent of the vote, most of which presumably came from the Australian Democrats whose vote fell by 11.9 percentage points. Alexander Downer’s vote remained at 53.6 per cent.⁽⁵⁴⁾

Braddon (Tas)

If Bass was always in some danger, Braddon seemed to be a different case. Won by Labor’s Sid Sidebottom in 1998, and retained in 2001 on a first preference tally of 48.4 per cent, the seat had looked safe during much of 2004. The Liberals’ Mark Baker seemed to have a difficult task in winning a seat that needed a 6 per cent, two-party preferred swing to be lost. In the event Baker’s first preference gain (8.2 percentage points) was even higher than Ferguson’s in Bass, and he defeated Sidebottom comfortably. Braddon was another Tasmanian electorate in which the timber industry was significant, and it seems likely that it was this issue which brought the sitting member undone. Like The Examiner in Bass, The Advocate ran full page advertisements, such as that by the Timber Communities Australia, which stated that ‘Mark Latham will lock up our Forests & Destroy Thousands of Jobs’.⁽⁵⁵⁾ In Bass, Michelle O’Byrne had stated that Green preferences would be crucial, but in neither Bass nor Braddon did the Green vote come close to being a factor in the result.⁽⁵⁶⁾

Cunningham (NSW)

The victory by the Greens’s Michael Organ in the Cunningham by-election was the first minor party victory in a House of Representatives contest since 1946. It was due largely to the Liberal Party decision not to contest the by-election, as well as a significant drop in the Labor vote (-6.1 percentage points).⁽⁵⁷⁾ With a Liberal candidate running in 2004, it was unlikely that Organ would hold his seat due to his probable third place finish in the first preference count. This duly occurred, and Organ was too far behind the Liberal candidate to be able to make up the leeway on preferences before being dropped from the count. Interestingly, although the Green vote was 13.5 percentage points higher than in 2001, it had fallen 2.9 percentage points on the party’s tally in the by-election. Labor’s first preference vote (39.6 per cent) was 4.5 percentage points lower than in 2001 and only 1.5 percentage points higher than in the by-election, but Sharon Bird easily won the seat for Labor on preferences, turning around her 2002 by-election defeat.

Greenway (NSW)

The Liberal Party has gradually pushed the ALP out of urban-edge seats that ring the Sydney metropolitan area (see below pp. 35–6), with Greenway being the latest. Labor had held the seat since it was first contested in 1984, though with a marked slippage in its vote since the 1996 election. The 2001 election saw the margin between Labor and Liberal narrow to the point where the seat could be lost on just a 3.1 two-party preferred swing.

In 2004 the Liberal Party nominated Louise Markus, a social worker and member of the popular local Hillsong Church, something that caused commentators to speculate about the possible electoral impact of voters who were regular churchgoers.⁽⁵⁸⁾ Labor’s candidate was Ed Husic, a member of the NSW right and a union organiser, a member of a Muslim family, a factor that was described by one journalist as forming ‘a subtle backdrop’ to the Greenway contest.⁽⁵⁹⁾ Apart from these local matters, though, Greenway had seemed a likely Liberal gain because of the marked shift in voter preferences away from the ALP in this part of Sydney. A seven per cent first preference swing to the Liberal Party, and a fall in the Labor vote of 2.6 percentage points, were enough to see Markus take the seat narrowly with a two-party preferred vote of 50.6 per cent.

Parramatta (NSW)

Parramatta was the Liberal Party’s sixth most marginal electorate and was vulnerable to a two-party preferred swing of 1.2 per cent. Liberal Parliamentary Secretary to the Treasurer, Ross Cameron, who had taken the electorate from Labor in 1996 with 47.7 per cent of first preferences, had not been able to match this figure in either the 1998 or 2001 elections. Clearly, it was a possible Labor gain.

On 14 August 2004, the Sydney Morning Herald’s Good Weekend Magazine published a profile on Cameron in which he confessed to infidelity: ‘I have been an unfaithful husband’. He also acknowledged that, ‘If my constituents want to vote for a great family man, they should probably vote for the other guy’.⁽⁶⁰⁾ As the media seized the story, a number of Cameron’s party colleagues expressed their support for him: ‘I would say to the people of Parramatta, if you want an energetic, hard-working representative, vote for Ross Cameron’, was the way the Prime Minister put it.⁽⁶¹⁾

In the event, Cameron’s first preference vote dropped only 1.8 percentage points with the Labor candidate, Julie Owens, trailing him by 2.6 percentage points. Cameron’s confession may have affected just enough voters to move the seat to the ALP, for Owens eventually won by just 1.6 percentage points after preferences.

Richmond (NSW)

The north coast New South Wales seat of Richmond had been long dominated by the Country/National Party; for 58 years it had been held by a member of the Anthony family. However, during the 1980s and 1990s, the combined impact of an inflow of ‘sea-changers’ and ‘alternatives’ had altered the nature of the seat. In the 1980s, the National vote always topped 50 per cent, but that figure was last achieved in 1987. The first sign of major change came during 1990–96 when Labor held the seat. Larry Anthony recaptured Richmond for the Nationals in 1996, though with a primary vote of only 35.4 per cent.

Between 1996 and 2004, Anthony increased his primary vote at each election, but a vote of 45.8 per cent was not high enough to protect him from the impact of Green preferences in 2004. Labor’s Justine Elliot secured a first preference vote of only 35.6 per cent, Labor’s average vote since 1990, but was able to win the seat narrowly by 301 votes. Anthony criticised the how-to-vote cards put out by the Liberals for Forests and the Ex-Service, Service & Veterans parties for being deliberately misleading and affecting the result,⁽⁶²⁾ but it is likely that the demographic changes in the seat were of greater importance. These changes, which are continuing as people retire to the area, may make it difficult for the Nationals to regain the seat if the Liberal Party contests it at the next general election.

Wentworth (NSW)

In 2001, the sitting member for the inner-Sydney electorate of Wentworth, Andrew Thomson, was challenged successfully for Liberal pre-selection by Peter King, former State President of the New South Wales division of the Liberal Party. For a time it had appeared as if former republican movement head, Malcolm Turnbull, would also be a challenger. King duly won this seat that has always been held by parties of the right.

In 2004, it was Turnbull’s controversial challenge that stripped King of party pre-selection. After much intra-party controversy, the sitting member nominated as an independent, a move that opened up the possibility of there being a tight three-way struggle between King, Turnbull and the Labor candidate, David Patch. Some speculation suggested that the Labor Party might win the seat, but as that party’s vote in the previous election had been only 29.5 per cent, this seemed unlikely. A much more likely prospect was of King being ahead of Labor after the second-last count, and defeating Turnbull on preferences—as had occurred in Curtin and Moore (both WA) in 1996.⁽⁶³⁾ In the event, the Liberal vote fell 10.3 percentage points, but it still topped 41 per cent, a healthy total under the circumstances. Labor’s vote fell by 3.2 percentage points, but Patch’s 26.3 per cent was still well ahead of King (18.0 per cent). Turnbull consequently won on King’s preferences with the Liberals’ two-party preferred vote falling only 2.4 per cent.⁽⁶⁴⁾

The ‘Scoresby’ seats

Occasionally a major state issue can impact upon a Commonwealth election—as in the case of the Franklin Dam in 1983. In Victoria, the issue of whether or not tolls should be paid on the soon-to-be-built Scoresby Freeway in Melbourne’s south-east seems to have been such an issue. The Liberal Party joined with their state counterparts to criticise the Bracks Labor Government’s determination to impose tolls. A Liberal advertisement close to polling day stated:

A strong vote against Labor on Saturday will force Steve Bracks to back down on tolling the Scoresby. There’s only one way to stop Labor’s tolls on the Scoresby.⁽⁶⁵⁾

The election analyst, Malcolm Mackerras has pointed out that the Liberals’ best performance anywhere in Australia was in the seven seats most affected by the freeway—Aston, Casey, Deakin, Dunkley, Holt, Isaacs and La Trobe, where Liberal seats are now safer and Labor seats very marginal.⁽⁶⁶⁾ Aston, in particular, a seat which runs from Forest Hill to Rowville, has been made the safest Liberal seat in Melbourne:

 ‘Scoresby’ seats—two-party preferred margins (per cent)

Source: Australian Electoral Commission

The benefits of incumbency?

This election saw some discussion of the perceived financial advantages that sitting MPs have in their efforts to retain their seats. Former Victorian Liberal Party President, Michael Kroger, talked of this, and was quoted as calculating that the financial benefit of incumbency for members of parliament to be $1.5 million over a three-year term. In this total, Kroger was including staff, offices, telephones, printing and mail allowances.⁽⁶⁷⁾ This was a point taken up by Professor Dean Jaensch of Flinders University, who spoke of the ‘massive benefit’ of being a sitting member. According to Jaensch, the largesse mentioned by Kroger helped guarantee the high level of stability in the House of Representatives, and he concluded that ‘the mechanics of the election process are generally fair, but these biases to the incumbents certainly are not’.⁽⁶⁸⁾ If there is an advantage, it probably aids the winning side more than the losers. Labor’s vote rose in 72 seats, but it held only 30 of the seats, and it lost two of these.

The table below suggests that incumbency is a major factor in Australian elections primarily because governments are usually stable rather than because of such benefits:

Electoral performances by incumbents