The 2019 federal election

29 June 2020

PDF version [1496KB]

Dr Damon Muller
Politics and Public Administration

Executive summary

  • This paper provides an overview of the 2019 federal election, including the results, the political environment, and the major features of the campaign.
  • The 2019 federal election was held on Saturday 18 May 2019.
  • The Liberal-National Coalition, led by Prime Minister Scott Morrison, won a majority in the House of Representatives with 77 of the 151 seats and 51.5 per cent of the national two-party preferred vote, a net increase of one seat on the 2016 election result.
  • The Australian Labor Party won 68 seats, with the Australian Greens retaining their single seat. Three independents were elected to the House of Representatives (including Helen Haines, who replaced retiring independent Cathy McGowan in the division of Indi), along with one Centre Alliance MP and one Katter’s Australian Party MP.
  • The Coalition won 19 seats in the Senate, with the ALP winning 13, the Greens winning six, and the Jacqui Lambie Network and Pauline Hanson’s One Nation winning one seat each. When combined with the ongoing senators, the Coalition has 35 votes in the Senate, Labor 26, the Greens nine and there are six other cross-benchers.
  • Mr Morrison was elected as Liberal Party leader and Prime Minister on 24 August 2018 following the ousting of Malcolm Turnbull. The Nationals also contested the election with new leader and Deputy Prime Minister Michael McCormack, who replaced former leader Barnaby Joyce in February 2018.
  • The election followed the ‘citizenship crisis’ of the 45th Parliament in which 15 senators and members of the House of Representatives were disqualified or resigned from Parliament. Election candidates were required to complete a qualification checklist with their nomination, and only one citizenship case was referred to the High Court following the election, which was dismissed by the Court.
  • The Coalition won the election having trailed Labor in the polls for essentially the entire term of the 45th Parliament, representing an unprecedented failure of polling for Australian federal elections. The polling failure has resulted in at least one inquiry and may lead to changes in the performance and transparency of Australian polling.
  • Clive Palmer, who contested the 2013 federal election with the Palmer United Party, returned to electoral politics with the renamed United Australia Party. While the party failed to win any seats, the estimated $60 million advertising spend by the party eclipsed the spending of all other major parties, and has led to some calling for expenditure caps in federal elections.


Executive summary
Timing of the election
The election campaign
The opinion polling failure
Cost of the election and campaign
Early voting
Electoral participation
Appendix A: Parties contesting the 2019 federal election
Appendix B: Section 44 disqualifications
Appendix C: Final opinion poll results prior to the 2019 federal election
Appendix D: Public election funding paid


This paper provides an overview of the 2019 federal election. This includes the political and electoral context in which the election occurred, the results in the House of Representatives and the Senate, the timing of the election and the election campaign. The paper also discusses key democratic indicators such as early voting, campaign spending, and electoral participation.

Held on Saturday 18 May 2019, the 2019 federal election returned Prime Minister Scott Morrison (Cook, NSW) and the Liberal-National Coalition to government with a majority of 77 of the 151 seats in the House of Representatives. The result meant that Mr Morrison was the third Liberal Prime Minister since the 2013 federal election when Tony Abbott won government from a minority Labor Government. The Coalition was widely expected to lose the 2019 election, having trailed the Labor Opposition, led by Bill Shorten (Maribyrnong, Vic.), in the polls for almost the entire parliamentary term.

The electoral context

While every election is unique in some respects, the 2019 federal election came after a particularly turbulent term that followed the 2016 double dissolution election. After emerging from the 2016 federal election with a slim majority of 76 seats, the Coalition lost a number of its MPs and senators due to ineligibility (as did Labor), a prime minister due to internal instability, and its majority in the House before the 2019 federal election.

Parliament’s citizenship crisis

On 14 July 2017 Greens Senator Scott Ludlam (WA) announced by tweet that he was resigning from the Senate after he was revealed to have been a New Zealand citizen.[1] Section 44(i) of the Australian Constitution holds that anyone holding citizenship of another country is incapable of being chosen or sitting as a senator or member of the House of Representatives.

Senator Ludlam’s resignation was particularly notable as it led to widespread citizenship examinations that eventually resulted in a total of 15 disqualifications due to section 44(i)—seven from the House of Representatives and eight from the Senate. A full list of the disqualified members and senators is in Appendix B.

While Senator Ludlam was the first of the section 44(i) citizenship casualties, he was the third senator to have been found ineligible in the 45th Parliament under section 44. The High Court finding former Pauline Hanson’s One Nation (PHON) senator Rodney Culleton (WA) ineligible in February 2017 due to having been convicted and under sentence for a crime (section 44(ii)),[2] and in April 2017 the Court finding Family First Senator Bob Day (SA) ineligible under section 44(v) due to a direct or indirect pecuniary interest with the public service of the Commonwealth.[3]

While most of the members of the House of Representatives who were disqualified or resigned prior to being disqualified were returned at subsequent by-elections, the recounts required by the High Court in the Senate led to considerable change in the composition of the Senate. This resulted in changes in the party composition as replacements were either elected from a different party (Nationals Senator Fiona Nash was replaced by Liberal Jim Molan in NSW) or left the party for which they ran at the 2016 election.[4] As such the Senate, which elected all of its 76 senators at the 2016 double dissolution election, looked quite different immediately before the 2019 federal election.

As a response to the citizenship crisis, the Parliament passed the Electoral Legislation Amendment (Modernisation and Other Measures) Act 2018 in February 2019. One effect of this legislation was to require candidates nominating for election to the House of Representatives or the Senate to complete a candidate qualification checklist as part of their nomination process. The checklist requires candidates to list information about their citizenship and family history, and for the 2019 election completed checklists were made available on the Australian Electoral Commission (AEC) website. No successful citizenship challenges following the election suggest that the initiative was successful in obliging candidates to consider their citizenship prior to nominating.

Leadership change and the minority government

Scott Morrison led the Liberal Party (LP) to the election, having taken over as Liberal leader and Prime Minister in August 2018 following the ousting of Malcolm Turnbull (Wentworth, NSW) less than nine months before the election. Prior to the leadership change, Kevin Hogan MP (NP, Page, NSW) had indicated that if there was a leadership spill he would move to the crossbench in protest—a move he duly made.[5]

Mr Turnbull subsequently resigned from the Parliament, triggering a by-election in his seat of Wentworth on 20 October 2018. The seat was narrowly won by independent Dr Kerryn Phelps. Following the loss of Wentworth, the Coalition was reduced to 74 seats of the 150 seat House of Representatives, two seats short of an absolute majority of 76 votes. While an absolute majority is only required in a small number of situations, the Speaker of the House (one of the Coalition’s 74 members) does not vote except to have a casting vote in the case of a tie.[6]

Coalition numbers were reduced further when, on 27 November 2018, Liberal Julia Banks (Chisholm, Vic.) announced that she was leaving the LP due to disillusionment with the party and would sit as an Independent.[7] As a result of these various developments, despite winning 76 seats at the 2016 federal election, the Coalition went into 2019 as a minority government (even before some of the electoral redistributions, which are discussed below).

The election was also the first for Michael McCormack (Riverina, NSW) as leader of the Nationals (NP); Mr McCormack had taken over as leader in February 2018 from Barnaby Joyce (New England, NSW). Mr Joyce had resigned as NP leader and deputy Prime Minister following weeks of controversy over his personal life.[8]


House of Representatives

The Liberal-National Coalition was returned to government at the May 2019 election with 51.53 per cent of the two-party preferred vote. The Coalition had a net gain of one seat compared to the 2016 federal election, winning a total of 77 seats in the 151 seat House of Representatives (a majority of one seat following the appointment of a Liberal Speaker). Labor won 68 seats (a net loss of one), with three seats going to independents, one to the Greens, one to the Centre Alliance (CA) (formerly the Nick Xenophon Team) and one to Katter’s Australian Party (KAP) (Table 1 below).

Table 1: House of Representatives seats won by party by state, 2019 federal election

NSW 24 15 - 7 - 1 - - 47
VIC 21 12 - 3 1 1 - - 38
QLD 6 - 23 - - - 1 - 30
WA 5 11 - - - - - - 16
SA 5 4 - - - - - 1 10
TAS 2 2 - - - 1 - - 5
ACT 3 - - - - - - - 3
NT 2 - - - - - - - 2
Total 68 44 23 10 1 3 1 1 151

Source: Compiled by the Parliamentary Library from AEC election results data.

The 2019 federal election continued the trend of decreasing first preference votes for both of the major parties, with both parties returning the lowest primary vote for some years (Figure 1 below). This election saw the ALP receive the lowest primary vote it had received since 1931 (when it received 27.1 per cent) and the lowest primary vote for the LP since it first contested federal elections in 1946.[9] In two-party preferred (TPP) terms the election saw a slight swing of 1.17 percentage points towards the Coalition compared to 2016, leaving the Coalition with a 1.53 per cent winning margin.

Figure 1: First preference and two-party preferred vote, federal elections 2004–

Figure 1: line graph showing first preference and two-party preferred vote, federal elections 2004–

Source: Compiled by the Parliamentary Library from AEC election results data.

Table 2: House of Representatives national results (top 15 parties), 2019 federal election

Party Votes Percent Seats Swing
        Votes Seats
Australian Labor Party 4 752 160 33.34 68 -1.39 -1
Liberal 3 989 404 27.99 44 -0.69 -1
The Greens 1 482 923 10.40 1 0.17 0
Liberal National Party of Queensland 1 236 401 8.67 23 0.15 +2
The Nationals 642 233 4.51 10 -0.11 0
United Australia Party 488 817 3.43 0 3.43 0
Independent 479 836 3.37 3 0.55 +1
Pauline Hanson's One Nation 438 587 3.08 0 1.78 0
Animal Justice Party 116 675 0.82 0 0.12 0
Christian Democratic Party (Fred Nile Group) 97 513 0.68 0 -0.63 0
Katter's Australian Party 69 736 0.49 1 -0.05 0
Centre Alliance 46 931 0.33 1 -1.52 0
Shooters, Fishers and Farmers 41 479 0.29 0 0.18 0
Country Liberals (NT) 38 837 0.27 0 0.03 0
Formal 14 253 393 94.46   -0.48  
Informal 835 223 5.54   0.48  
Total 15 088 616 91.89   0.89  

Source: Compiled by the Parliamentary Library from AEC election results data.

PHON polled better than it has in recent elections, with its national first preference vote of 3.1 per cent an increase on the 1.29 per cent it received in 2016. The result was well short of the 8.4 per cent national primary vote it received in 1998, however, or even its next-best general election result of 4.3 per cent in 2001. The party ran candidates in 59 seats (primarily in Queensland (Qld) and Western Australia (WA)) and did not win any lower house seats. In Qld ON received 8.86 per cent of the first preference vote, a positive swing of 3.34 percentage points since 2016.

The Greens received a very slight positive national swing (0.17 per cent), returning incumbent Adam Bandt in the division of Melbourne (Vic.), and KAP received a very slight negative national swing (-0.05 per cent), also returning the party’s only incumbent candidate, Bob Katter, in Kennedy (Qld).

The Centre Alliance, which only ran candidates in three South Australian seats, experienced a significant negative national swing (-1.52 per cent) compared to the previous general election. CA incumbent Rebekha Sharkie was returned in the division of Mayo (SA).

Three independents were elected. Andrew Wilkie was returned in the division of Clark (Tas.) for his fourth term.[10] Helen Haines, the designated successor of two-term independent Cathy McGowan, won the division of Indi (Vic.), the first time an independent has been replaced by another independent.[11] Independent Zali Steggall won the division of Warringah (NSW) from eight-term incumbent Liberal member and former Prime Minister Tony Abbott.[12]

Independent Kerry Phelps, who won the division of Wentworth (NSW) at a by-election on 20 October 2018 following the resignation from Parliament of former Prime Minister Malcolm Turnbull, was not successful at the 2019 general election. The seat of Wentworth was returned to the Liberals, and represented no net change for the division compared to the 2016 general election.

The United Australia Party (UAP) ran candidates in all 151 electorates and received a primary vote of 3.4 per cent, failing to elect any of its candidates. While not strictly a like-to-like comparison, the Palmer United Party received 5.5 per cent of the primary vote in 2013, electing Clive Palmer to the division of Fairfax (Qld).

Table 3: Two-party preferred by state (percentage), 2019 federal election

State Coalition Australian Labor Party Swing to Coalition
New South Wales 51.78 48.22 1.25
Victoria 46.86 53.14 -1.31
Queensland 58.44 41.56 4.34
Western Australia 55.55 44.45 0.89
South Australia 49.29 50.71 1.56
Tasmania 44.04 55.96 1.40
Australian Capital Territory 38.39 61.61 -0.48
Northern Territory 45.80 54.20 2.86
National 51.53 48.47 1.17

Source: Compiled by the Parliamentary Library from AEC election results data.

While the Coalition won the majority of the TPP vote in only three of the eight Australian states and territories, these were three of the four largest by number of voters (see Table 3 above and Figure 2 below). The ALP won a majority of the TPP vote in a majority of the states and territories, however as these included the smallest states and territories it meant the Coalition led the ALP in TPP terms Australia-wide by about 436,000 votes. If Qld were excluded, the ALP would have won a majority of the seats (62 for the ALP compared to 54 for the Coalition).

Figure 2: Two-party preferred margin by state, 2019 federal election

Figure 2: map of Australia showing two-party preferred margin by state, 2019 federal election 

Source: Compiled by the Parliamentary Library from AEC election results data.

Figure 3: Two-party preferred swing to Coalition by state, 2019 federal election

Figure 3: map of Australia showing two-party preferred swing to Coalition by state, 2019 federal election 

Source: Compiled by the Parliamentary Library from AEC election results data.

All states and territories except Victoria and the ACT recorded a TPP swing towards the Coalition at the election (Table 3 and Figure 3 above). At the electorate level, most states and territories had electorates that swung both to and against the Coalition. However in WA and Qld there was a much stronger swing towards the Coalition across electorates. In Victoria electorates tended to swing more to Labor, although to a lesser extent than the swing to the Coalition in electorates in Qld (Figure 4 below).

When examined by vote type, the Coalition performed more strongly in early voting, leading the ALP by 9.2 percentage points in TPP terms. Election-day votes were much more even, however, with Labor having a slight two-party preferred lead in all types of election-day votes (Table 4 below).

Figure 4: Two-party preferred swings to the Coalition by electorate for each state, 2019 federal election

Figure 4: graph showing two-party preferred swings to the Coalition by electorate for each state, 2019 federal election

Note: This chart shows what proportion of divisions in each state swung to each party and by how much. The division of Whitlam was excluded due to the AEC not reporting a TPP swing, and the NT has too few electorates for a distribution to be calculated.

Source: Compiled by the Parliamentary Library from AEC election results data.

Table 4: Two-party preferred vote share by vote type, 2019 federal election

Vote type Coalition (votes) ALP (votes) Coalition (%) ALP (%)
Postal 699 547 514 117 57.64 42.36
Pre-Poll Declaration 300 964 292 914 50.68 49.32
Pre-Poll Ordinary 2 206 882 1 859 965 54.27 45.73
Total Early 3 207 393 2 666 996 54.60 45.40
Absent 267 677 309 819 46.35 53.65
Ordinary 3 849 803 3 905 494 49.64 50.36
Provisional 19 940 26 271 43.15 56.85
Total Election Day 4 137 420 4 241 584 49.38 50.62

Source: Compiled by the Parliamentary Library from AEC election results data.


Over two-thirds (almost 70 per cent) of the divisions were decided by second or later preferences, meaning that the winning candidate received less than 50 per cent of the first preference vote. Seats decided by first or later preferences are given by winning party in Table 5 below.

Table 5: Seats decided by first or later preferences by party, 2019 federal election

Decided By LP ALP LNP (Qld) NP GRN IND KAP CA Total
First preference 22 14 4 5 0 1 0 0 46
Later preferences 22 54 19 5 1 2 1 1 105

Source: Compiled by the Parliamentary Library from AEC election results data.

The number of seats decided by later preferences at this election follows the trend in recent elections whereby the number of electorates decided by later preferences has steadily increased (Figure 5 below).

Figure 5: Electorates decided by first or later preferences, federal elections 2004–

Figure 5: line graph showing electorates decided by first or later preferences, federal elections 2004–

Source: Compiled by the Parliamentary Library from AEC election results data.

Of the 151 divisions at the election, 15 were what the AEC refers to as ‘non-classic divisions’, where the final two candidates (the two-candidate preferred—or TCP—candidates) were not Labor or Coalition candidates. The high (by historical standards) number of ‘non-classic’ divisions makes examining preferences more complex—preferences in ‘classic’ divisions always go to either Labor or the Coalition, whereas in ‘non-classic’ divisions preferences will be counted towards another party, or an independent candidate.[13] When considering ‘classic’ divisions only, it is apparent that Labor won 54 seats overall on preferences compared to the Coalition’s 46 seats, and that, across all vote types, Labor gained more than half of preferences (and in some cases considerably more than half). However, early voting in ‘classic’ seats, particularly postal and pre-poll ordinary votes, was more favourable to the Coalition in terms of preference flows than election-day votes (Figure 6 below).

Figure 6: Share of TCP votes by vote type in ‘classic’ divisions, 2019 federal election

Figure 6: bar chart showing share of TCP votes by vote type in ‘classic’ divisions, 2019 federal election

Source: Compiled by the Parliamentary Library from AEC election results data.

The observation that Labor did better than the Coalition in terms of preferences in ‘classic’ seats may appear counter-intuitive given that the Coalition won the election (in terms of both TPP vote and seats); however, while Labor did receive a greater proportion of the preferences, it also received considerably fewer first preference (primary) votes. As can be seen in Figure 7 below, the Coalition received more primary votes in ‘classic’ seats than the combined Labor primary votes and Greens preferences flowing to Labor. Labor’s preferences flows, while strong, were insufficient compared to the Coalition’s much stronger lead on primary votes.

Figure 7: Source of votes and preferences in ‘classic’ seats by TCP party, 2019 federal election

Figure 7: bar chart showing source of votes and preferences in ‘classic’ seats by TCP party, 2019 federal election

Source: Compiled by the Parliamentary Library from AEC election results data.

The strong preference flows to Labor appear to have been mostly driven by the Greens (Figure 7, above). Around 1.35 million Greens votes flowed as preferences overall to TCP parties, with over one million of those going to Labor and around 250,000 going to the Coalition. The number of Greens preferences flowing to Labor therefore tends to overwhelm the preferences of every other party, most of which tended to favour the Coalition (Figure 8, below). For example, PHON and UAP preferences were about twice as likely to flow to the Coalition than to Labor.

Figure 8: Preference flows to the ALP and Coalition in ‘classic’ divisions, 2019 federal election

Figure 8: bar chart showing reference flows to the ALP and Coalition in ‘classic’ divisions, 2019 federal election

Note: ‘classic’ divisions are divisions where one of the two candidate preferred candidates is from Labor and the other is from the Coalition, and includes ‘three cornered contests’ where there is more than one Coalition candidate in the division.

Source: Compiled by the Parliamentary Library from AEC election results data.


The 2019 Senate election saw the predicted effect of the 2016 Senate voting system changes on minor party representation in the Senate realised.[14] Subject to a normal half-Senate quota, the only small parties other than the Greens that won seats were PHON in Qld and the Jacqui Lambie Network (JLN) in Tasmania (both of which incidentally returned senators who had been disqualified during the section 44 crisis of the 45th Parliament). Both PHON and the JLN were successful due to receiving a high proportion of the vote in their respective states. In Qld PHON received the third highest vote by group, with 0.72 quotas, and in Tasmania the JLN received 0.62 quotas, the fourth highest vote in the state. The new Senate voting system has had the effect of allowing small parties to gain Senate seats on the basis of votes rather than preference deals.

The results of the Senate election by party and state are in Table 6 below.

Table 6: Seats won by state in the Senate, 2019 federal election

State ALP LNP (Qld) LP NP Coalition
NSW 2 - 2 1 3 1 - - 6
VIC 2 - 3 - 3 1 - - 6
QLD 1 3 - - 3 1 - 1 6
WA 2 - 3 - 3 1 - - 6
SA 2 - 3 - 3 1 - - 6
TAS 2 - 2 - 2 1 1 - 6
ACT 1 - 1 - 1 0 - - 2
NT 1 - 1 - 1 0 - - 2
Total 13 3 15 1 19 6 1 1 40

Source: Compiled by the Parliamentary Library from AEC election results data.

Figure 9: National Senate primary votes (dots) and swings (arrows) by party group, 2019 federal election

Figure 9: chart showing national Senate primary votes (dots) and swings (arrows) by party group, 2019 federal election

Note: only includes parties that received at least 0.5 per cent of the national first preference vote. Some parties occur multiple times due to running under different group names in different states—the AEC does not provide the data required to calculate swings for combined groups.

Source: Compiled by the Parliamentary Library from AEC election results data.

Smaller parties tended to do well in the Senate election, with many recording a positive swing compared to the 2016 election, with notable exceptions in Derryn Hinch’s Justice Party and the Liberal Democrats—two parties that were represented in the Senate prior to the 2019 election (Figure 9 above). In general, however, these positive swings were not sufficient to make the parties competitive with the more established parties in terms of winning seats. The largest negative national swing was recorded by the CA, which saw a 3.11 percentage point drop in its vote from 2016.

Figure 10: National Senate votes (blue) and seats (red) by party group, 2019 federal election

Figure 10: chart showing national Senate votes (blue) and seats (red) by party group, 2019 federal election

Source: Compiled by the Parliamentary Library from AEC election results data.

Nationally, the Senate seat share of the LP, the Greens, the JLN and the ALP tended to out-perform the parties’ vote share. The Liberal National Party (LNP) of QLD received an unusually proportional 7.5 per cent of the seats for 7.7 per cent of the vote (Figure 10 above). The high seat performance relative to votes for the LP is likely due to having done relatively well in small states (WA, SA and Tasmania), where a large state-wide vote share is relatively small on a national level, however still results in winning seats.

As discussed in the introduction to this paper, due to the section 44 citizenship crisis the composition of the Senate immediately prior to the 2019 election looked quite different from that of the Senate elected in 2016. The net change as a result of the 2019 election was generally a loss of most of the non-Greens minor party senators, with the seats being picked up by the Coalition (Table 7 below).[15] As elected, the resulting Senate was much more favourable to the Government for passing legislation. If the ALP and the Greens chose to use their 35 combined votes against the Government in the Senate, the Coalition Government only needed four more votes from the six cross-benchers, in addition to its 35 votes, to pass legislation. With the resignation of independent Senator Cory Bernardi (SA) in January 2020 and his replacement by Liberal Senator Andrew McLachlan on 6 February, the Government gained one more Senate vote, meaning that it needs just three votes from five crossbench senators to pass legislation.[16]

Table 7: Senate composition following the 2019 federal election

Party Terms end 30/06/2022 Terms end 30/06/2025 Terms end next election Total Change
LP 14 15 2 31 5
ALP 13 11 2 26 0
GRN 3 6 0 9 0
NP 2 2 0 4 -1
CA 2 0 0 2 0
PHON 1 1 0 2 0
IND 1 0 0 1 -1
JLN 0 1 0 1 1
Others 0 0 0 0 -4

Note: The following parties were no longer represented in the Senate following the 2019 election: Derryn Hinch’s Justice Party (DHJP), the Liberal Democratic Party (LDP), the United Australia Party, and, after 25 June 2019, the Australian Conservatives (AC). Senator Cori Bernardi disbanded the AC party to sit as an independent prior to the Senate first sitting following the 2019 election. Territory senators do not have fixed terms; their terms end at the next general election.

Source: Compiled by the Parliamentary Library

Timing of the election

The timing of the 2019 federal election was the subject of considerable speculation from late 2018. The 2016 double dissolution election meant that a half-Senate election was required to be held between July 2018 and May 2019 for the senators elected in 2016 for a short term.[17] Further complicating the choice of election date were fixed-term elections due in NSW (23 March 2019) and Victoria (24 November 2018), and the Federal Budget, traditionally held in early May. The Budget was moved to 2 April 2019 to accommodate the election timing.

Media reports stated that Prime Minister Malcolm Turnbull had intended to call an election on 27 January 2019 for a 2 March election day, in order to go to the polls before the NSW state election.[18] By April 2019, however, new Prime Minister Scott Morrison was leaving open the possibility of a May election.[19]

There was some debate over election timing in respect of how late an election could be held.[20] Prior to the announcement of the election there was some speculation that it could be held as late as 25 May 2019,[21] and while the Electoral Commissioner did not dismiss the possibility, he noted that ‘the later we go, the more complex it becomes for the AEC but that does not mean it is impossible.’[22]

On the morning of Thursday 11 April 2019 the Prime Minister visited the Governor-General and advised the dissolution of the 45th Parliament for a general election on Saturday 18 May 2019.[23] The key dates the Prime Minister nominated (and consequential dates that follow from those) are listed in Table 8 below.

Table 8: Key dates for the 2019 federal election

Milestone Date
Issue of the writs Thursday 11 April 2019
Close of the rolls Thursday 18 April 2019
Close of nominations Tuesday 23 April 2019
Declaration of nominations Wednesday 24 April 2019
Early voting commences Monday 29 April 2019
Election advertising blackout commences Wednesday 15 May 2019
Polling day Saturday 18 May 2019
Writs returned Friday 21 June 2019
Last day for return of the writs Friday 28 June 2019

Source: Compiled by the Parliamentary Library

The Parliament was prorogued at 8.29am on 11 April 2019, and as such the Senate Estimates hearings scheduled for 11 and 12 April 2019 did not proceed.

While recent elections have generally been announced on a weekend, with the writs issued early the following week, the 2019 election timing provided complications with both Easter and Anzac Day occurring within the election period.[24]

The timing of the election is dictated by the Commonwealth Electoral Act (1918) (CEA), which nominates a minimum 33 day period between the issue of the writs and polling day.[25] The 2019 election period was 37 days, slightly above the average for elections since 1984 of 35.6 days. The writs were due to be returned 41 days after polling day, in time for the new Senators to commence their terms on Monday 1 July 2019, and were returned with a week to spare on Friday 21 June 2019.[26]

The CEA requires that close of the rolls must happen seven days after the issue of the writs (section 155) and the close of nominations between 10 and 27 days after the issue of the writs (section 156). Issuing the writ on Monday 15 April would have meant that the close of rolls would have been on Easter Monday (and for an 18 May election the close of nominations would have had to have been on Anzac Day).

While the campaign period included public holidays, early voting did not commence until Monday 29 April (five days after the declaration of nominations, which is 24 hours after the close of nominations, under subsection 200D(4) of the CEA), and so was not interrupted by any public holidays. Prior to the election being called the Electoral Commissioner had stated at Senate Estimates that if Anzac Day had fallen within the pre-poll period then pre-poll voting would not be available on that day.[27] The Prime Minister and Leader of the Opposition agreed not to campaign or run political advertising on Good Friday, Easter Sunday or Anzac Day (however this truce was not observed by all campaigners).[28]


One year after a new Parliament first meets, the CEA requires that the AEC use the latest population statistics to determine how many electorates each state and territory is entitled to using a formula based on section 24 of the Australian Constitution. The entitlement determination that occurred in the 45th Parliament required that SA lost one seat and Victoria and the ACT gained one seat each due to their relative population changes.[29] For the first time 151 Members of the House of Representatives were elected at the 2019 federal election.

The 45th Parliament also saw electoral redistributions in Qld, the NT and Tasmania due to seven years having passed since those jurisdictions last had redistributions.

The redistribution of Qld was completed on 27 March 2019, and made relatively minor changes to the boundaries. The redistribution retained the names of all 30 seats and adjusted the boundaries of 18 seats, with 33,604 electors (1.09 per cent of all enrolled Queensland electors) changing seats.[30]

The redistribution of Victoria was completed on 13 July 2018 and resulted in the creation of a new division of Fraser located in the western suburbs of Melbourne (in honour of the former Prime Minister Malcolm Fraser), resulting in a total of 38 electoral divisions. The redistribution also renamed the division of McMillan to Monash, in honour of Sir John Monash; the division of Melbourne Ports was renamed to Macnamara in honour of Dame Annie Jean Macnamara; the division of Murray was renamed to Nicholls in honour of Sir Douglas Ralph Nicholls and Lady Gladys Nicholls; and the division of Batman was renamed to Cooper in recognition of William Cooper. The names of the remaining 33 divisions were retained. A total of 769,523 electors (18.93 per cent of enrolled Victorian electors) changed their division as a result of the redistribution, and an additional 435,141 electors (10.71 per cent of enrolled Victorian electors) were in a division that was renamed.[31]

The redistribution of the ACT was completed on 13 July 2018 and resulted in the creation of a third electoral division for the ACT, which was named Bean in honour of Charles Bean. The naming of the electorate resulted in a number of objections, with the augmented Electoral Commission voting four to two in favour of Bean. The redistribution resulted in 151,218 electors in the ACT (52.44 per cent of all enrolled ACT electors) changing their division.[32]

The redistribution of SA was completed on 20 July 2018 and resulted in the abolition of the division of Port Adelaide. In addition, the division of Wakefield was renamed to Spence in honour of Catherine Helen Spence. The boundaries of all South Australian divisions were adjusted and 255,784 electors (21.40 per cent of all enrolled South Australian electors) changed division as a result.[33]

The redistribution of the NT was completed on 7 February 2017 and resulted in an adjustment of the boundaries of the two divisions and the movement of 2,680 electors (2.07 per cent of enrolled NT electors) who changed divisions.[34]

The redistribution of Tasmania was completed on 14 November 2017 and resulted in adjustments of the boundaries of Bass and Lyons and renaming of the division of Denison to Clark in honour of Andrew Inglis Clark. The redistribution resulted in 22,800 electors (6.08 per cent of enrolled Tasmanian electors) changing divisions.[35]

The abolished seat of Port Adelaide in SA was held by Labor, but the new seats of both Bean (ACT) and Fraser (Vic.) were notional Labor, giving Labor a national net gain as the result of the redistributions. The divisions of Corangamite and Dunkley in Victoria, both won in 2016 by the Liberals, were notionally ALP following the redistribution (by 0.03 percentage points and 1.25 percentage points respectively).[36]

The election campaign

The 2019 federal election campaign was generally unremarkable. The Coalition went into the election behind in the polls, however while the Coalition was out-polled by Labor throughout the campaign, the gap between the two parties closed across the election period.

The Coalition had its third Prime Minister since Tony Abbott led the Coalition to government in 2013, while the ALP’s leadership team was largely unchanged since Bill Shorten took the leadership of the party in October 2013 following the defeat of the Rudd government.

Most commentators expected an ALP victory, if possibly a narrow one.[37] In particular, it was predicted that Labor would pick up two to three seats from the Liberals in Victoria and win seats from the Liberals in NSW, but would probably not win additional seats in Qld.[38]

The campaign was shaken by the death, two days before the election, of former Labor Prime Minister Robert ‘Bob’ Hawke. Media reports suggested that both major parties held the view that the death of Mr Hawke would be generally beneficial for Labor’s election efforts.[39]

The ALP was the first of the major parties to hold a campaign launch, doing so in Brisbane on Sunday 5 May. The launch featured former prime ministers Kevin Rudd and Julia Gillard, and Bill Shorten articulated Labor’s campaign promises on health, tax cuts, education, climate change and fairness.[40] The Coalition campaign launch was held in Melbourne on Mothers’ Day, Sunday 12 May. Commentators noted that the launch was focused almost exclusively on Scott Morrison as leader, in part to shift focus from the leadership instability that the party had experienced in the past six years.[41]

While it had been the case that the parties had an agreement not to claim travel allowances after the campaign launch, a media report prior to the election stated that at the 2019 election ‘both sides of politics can travel on chartered and domestic flights and claim travel allowance all the way to election day’.[42] Post-election media reports stated that in April, May and June cabinet ministers’ staff accumulated $5 million in travel expenses, and shadow ministers’ staff around $1.6 million.[43]

The policy contest

The federal Budget, usually delivered in May, was delivered on 2 April 2019, and was viewed by commentators as an ‘election budget’. The Budget featured tax cuts, both immediate and into the future, and a commitment to return the Budget to a small surplus, though was otherwise restrained in terms of new spending.[44] The Coalition largely relied on the Budget, delivered two weeks before the election period began, as the foundation of its campaign, with a focus on maintaining a strong economy.[45]

A statement by the Prime Minister proposed that voters ‘have the choice between the LNP that is delivering a strong economy and will continue to do so, or Bill Shorten and the Labor Party, who are proposing massive new taxes that will only weaken our economy’.[46] Bill Shorten, in a statement to voters, asserted that the Government ‘has no agenda for the next three years apart from a plan to give a $77 billion tax handout to the top end of town’.[47]

Labor went to the election with largely the same suite of policies that had delivered it an almost three per cent swing and 49.6 per cent of the two-party preferred vote at the 2016 federal election. Labor’s promises amounted to an extensive list of expenditure measures along with a number of savings measures to fund them. The highest profile savings measures were the abolition of negative gearing on all but newly-built investment properties (grandfathering existing negative gearing), and the cessation of tax rebates for franking credits.[48] These two policies were calculated to save the budget $14.2 billion over the following four years and $32 billion in a decade, respectively.[49] The Opposition Leader conceded that Labor’s campaign promises, particularly in relation to the savings issues, ‘have given our opponents plenty of ammunition for their endless scare campaigns’.[50]

One aspect of the campaign that was considered significant (mostly in retrospect) was the position of the ALP on the proposed Adani coal mine in Queensland. Labor’s climate change position, which was viewed as being anti-coal, was reported prior to the election as damaging Labor’s chances of picking up rural and regional Queensland seats, although it was expected that Labor would do better in Brisbane.[51] While Labor generally attempted to avoid taking a position on the Adani mine, noting that environmental approvals had already been granted by the Government, one significant union threatened to campaign against candidates if it looked like Labor would not support the mine proceeding.[52]

The Greens argued that the Adani mine ‘can’t be built’, and former Greens leader Bob Brown led a convoy from Hobart to the mine site in the Qld Galilee Basin to protest against the mine.[53] After stopping at Clermont in central Qld, the convoy finished at Parliament House in Canberra prior to the election.[54] In the wake of the election, Greens leader Richard Di Natale concluded that the convoy had harmed Labor in the election ‘because the Labor party refused to take a clear position’.[55] Similarly, a Labor MP who lost a central Qld seat agreed that the way Labor’s climate policy was conveyed alienated blue-collar workers.[56] According to Labor’s official review of its campaign:

Labor’s ambiguous language on Adani, combined with some anti-coal rhetoric and the Coalition’s campaign associating Labor with the Greens in voters’ minds, devastated its support in the coal mining communities of regional Queensland and the Hunter Valley.[57]

Candidate disendorsements

The election saw an unprecedented number of candidates either resign or lose their party endorsement. There is no provision in the CEA for the removal from the ballot paper of candidates who resign or who are disendorsed by their party after the close of nominations. For the 2019 federal election ten candidates resigned or were disendorsed by their party after the close of nominations:

  • Gurpal Singh (Scullin, Vic.), Liberal
  • Luke Creasey (Melbourne, Vic.), Labor
  • Peter Killin (Wills, Vic.), Liberal
  • Jeremy Hearn (Isaacs, Vic.), Liberal
  • Jessica Whelan (Lyons, Tas.), Liberal
  • Steve Dickson (Qld Senate), One Nation
  • Wayne Kurnorth (NT Senate), Labor
  • Jay Dessi (Lalor, Vic.), Greens
  • Tony Pecora (Melbourne, Vic.), United Australia and
  • David Paull (Parkes, NSW), Greens.

While none of the disendorsed candidates were elected, some polled well despite their disendorsement (Jeremy Hearn received 35 per cent of the vote in Isaacs, coming second behind the winning Labor candidate, for example).

As a response to the string of disendorsements the Victorian Liberal Party brought forward its preselection process for the next federal election, closing nominations in January 2020.[58] However the candidates’ votes, due to be conducted in March 2020, have at the time of publication been delayed indefinitely due to the COVID-19 pandemic.[59]

Candidates who are disendorsed by their party after the close of the nominations continue to have their original party listed against their name on the ballot paper, and the party will receive per-vote public funding for any votes received by such candidates. The public funding scheme was changed prior to the 2019 federal election, with parties and candidates being required to submit a claim to the AEC for reimbursement of campaign expenditure up to the public funding amount. On the basis of the public funding rate and the votes the candidates received, the Liberals were eligible for around $246,000 in public funding from the votes from their disendorsed candidates, Labor around $51,000, and the Greens around $31,000. None of the disendorsed Senate candidates received sufficient votes to reach the four per cent threshold to qualify for public funding.

Electoral advertising on social media

Authorisation requirements

Following changes to the CEA by the Electoral and Other Legislation Amendment Act 2017 (Cth), the requirements to authorise electoral advertising were expanded and in some cases clarified to explicitly cover electoral advertising by social media.[60]

The AEC is responsible for administering and enforcing the authorisation provisions. A pre-election report in The Guardian noted the AEC’s strategy for policing social media advertising:

The AEC said in a statement that generally unauthorised paid electoral ads infringe on the Commonwealth Electorate Act. But given the volume of material on Facebook and other channels, the AEC said it “benefits” from public reports to identify where the law has been breached.

Once alerted, it contacts the Facebook group and asks for the ad to be authorised or removed.

“If the infringement is not addressed, the AEC can contact the relevant social media provider to advise them to remove the unauthorised electoral advertisement,” a spokesman said. “The AEC may also consider legal action, such as an injunction to prevent the further display of the unlawful advertising, and, in serious cases, civil penalties for breaching the Electoral Act.”[61]

More recently the AEC has indicated that it had positive interactions with social media providers in respect of the 2019 election. In December 2019, in evidence to the parliamentary Joint Standing Committee on Electoral Matters (JSCEM) inquiry into the 2019 federal election, the Electoral Commissioner stated:

At the last event, we had more interaction and, I would say, better cooperation with the social media companies than we've ever had previously. It was the first election where our interaction was comprehensive before and during, and we're pretty happy with what we achieved with them as a one-off... It wasn't just with Twitter, Google and Facebook; we also reached out to WeChat. As you know, in Australia there are a very large number of users of WeChat as well. [62]

In its post-election submission to the JSCEM inquiry, the AEC stated that it had received 1,072 enquiries and complaints relating to electoral communication (an increase on what was received in relation to the 2016 election). It provided advice in relation to 544 of these and investigated 528, of which 109 (some 20 per cent) related to social media communications. The outcomes of the 528 investigations are detailed in Table 9 below.[63]

Table 9: Electoral communication investigated by the AEC at the 2019 federal election

  Communications investigated No breach Breach (s 321D) Breach (s 329)
Social media 109 81 25 3
Signs and print 223 183 47 3
Other communications 186 175 10 1
Total 528 439 82 7

Source: Australian Electoral Commission[64]

Note: Section 321D of the CEA relates to improper authorisation of electoral communication and section 329 relates to electoral communications that mislead or deceive an elector in relation to the casting of a vote.

In the report the AEC stated that it issued 78 warnings in relation to breaches and took further action on 11 matters where the infringement was not addressed following the warning by asking the social media company to remove the material.[65]

Transparency of electoral advertising on Facebook

In light of recent criticisms of the lack of transparency of political advertising on Facebook, Facebook has introduced an ‘Ad Library’, which aims to make public otherwise ‘dark’ (that is, not publicly viewable) targeted political advertising. An ABC report shortly after the 2019 election noted that Facebook’s Ad Library did not include information such as the demographic targeting of ads, as is available in some regions. It was also reported that Facebook would make the data available globally in June 2019 (Ad Library now contains Facebook ads from Australian political parties and candidates).[66]

In evidence to the JSCEM the AEC noted that, according to Facebook, Ad Library was not available in relation to political advertising for the 2019 federal election due to a number of elections internationally.[67] In April 2019 Facebook had published a blog post in which it discussed Ad Library in the context of the election, stating that it would ‘shin[e] a brighter light on advertising and Pages [which] makes both Facebook and advertisers more accountable, which is good for people and good for democracy’. Facebook also stated that it would block foreign electoral advertising:

Combating foreign interference is a key pillar of our approach to safeguarding elections on our platform. As part of this commitment, we’re temporarily not allowing electoral ads purchased from outside Australia ahead of the election in May.

The restriction will take effect the day after the election is called and will apply to ads we determine to be coming from foreign entities that are of an electoral nature, meaning they contain references to politicians, parties or election suppression. We also won’t allow foreign ads that include political slogans and party logos.[68]

Due to the limited transparency of political advertising on Facebook at the time of the 2019 election, both the Guardian and the ABC launched initiatives encouraging people to take screenshots of political advertising and to share these with the media organisations for archiving. The Guardian’s archive is still available, although the ABC archive appears no longer to be accessible.

The Guardian published some analysis of the Facebook ads that were submitted to it and data provided by Facebook’s Ad Library, finding that ads sponsored by political campaigns were targeted in a range of ways, for example towards particular demographics or specific interests that might be affected by significant election policies. The Guardian’s analysis also found ‘many’ ads without proper authorisation information, including ads from anonymous pages.[69]

One particular stream of social media election advertising that has been investigated in some depth related to material suggesting that Labor had plans to introduce a death tax if it was elected. According to an investigation by The Guardian, the death tax material was widely shared on Facebook, and the belief that Labor would introduce a death tax was encountered by Labor MPs campaigning for the election.[70] The Guardian report found that claims Labor was in favour of a death tax could be traced back to a Daily Telegraph article from July 2018 regarding Australian Council of Trade Unions support for an inheritance tax;[71] a follow-up discussion on the ‘Sunrise’ television program the next day; and a subsequent media release by the Treasurer in January 2019.[72]

The report states that at first the spread of the material was relatively organic:

The Sunrise clip was shared by One Nation leader Pauline Hanson’s official Facebook account the same day it aired. Hanson’s post alone was viewed 106,000 times. LNP members Ian MacDonald and George Christensen began to push the death tax messaging from their Facebook pages in late January and early February, but their posts achieved relatively little engagement.[73]

The Guardian noted that the material spread rapidly in the lead-up to the election and was forwarded to people via direct Facebook message. The same message was also being spread by minor parties and others:

The death tax messaging on social media was being amplified by minor parties, including Clive Palmer’s United Australia party, which was engaged in saturation advertising, by fringe right-wing groups, and by individual Coalition senators. The LNP backbencher George Christensen published three Facebook posts on the topic on one day, 22 April. Christensen’s first post was titled “Labor’s secret plans for a DEATH TAX”. A day later, Rite-On, a right-wing group with a significant online following, published a post headlined “DEATH TAX – INHERITANCE TAX – DEATH DUTY, call it what you will, it will hit us all”.[74]

Exactly who was responsible for the rapid spread of the death tax message remains unclear; however the information in the report might seem to indicate that it was more than just organic spread of a message that caught the attention of a sub-set of the electorate.

Even if Facebook’s Ad Library had been in full operation in the 2019 federal election, it is difficult to know how effective it would have been in revealing any organisation behind the death taxes campaign. The Guardian reports that the campaign did involve some paid Facebook advertising by candidates and campaigns, which should have been caught by Ad Library; but sharing by groups and individuals would likely not have been exposed by Ad Library. It is therefore questionable whether Facebook provides the tools to render transparency to these sorts of campaigns. 

In a policy shift that happened after the federal election, Facebook has recently stated that posts shared by politicians are not subject to Facebook’s community standards or fact-checking.[75] Facebook has stated that it exempts politicians from its fact-checking program, but that if a politician shares previously debunked content it will be ‘demoted’ (however paid ads must be consistent with Facebook’s ‘Community Standards’, and debunked content cannot be displayed in paid ads).[76] Google has stopped allowing political ads to be targeted to search users, and Twitter has stopped accepting paid political advertising.[77]

Expenditure on social media electoral advertising

How much of the electoral advertising spend in Australia goes on social media is impossible to accurately determine. Australian federal political finance laws do not require parties to disclose their election campaign spending—parties are only required to report annual total income, donations, expenditure and loans, and are not required to state what proportion of the expenditure related to a campaign.

Media monitoring companies also tend not to provide social media campaign spending estimates as they do with traditional broadcast media, at least partly because much of the advertising is micro-targeted and so only appears to certain users.[78] Television, radio and print advertising, on the other hand, is by definition public, and more closely monitored, allowing spending estimates to be made. For example, for the 2016 election an estimated $16m was spent on television, radio and print advertising, 90 per cent of it by the Coalition and the ALP; for the 2019 federal election, by mid-March 2019 Clive Palmer alone had already spent an estimated $21m.[79] In contrast, it has been estimated that the major party spends on online and social media advertising would constitute around 20 to 30 per cent of the total ad spend.[80]

Election appeals: citizenship and purple signs

In its inquiry report on the section 44 crisis, the JSCEM recommended that, in the absence of a referendum to change the Constitution, the Government should consider strategies to mitigate the impact of section 44, such as ‘that all candidates for election be required to make a public disclosure of family citizenship history at time of nomination.’[81] The Government responded to this recommendation by passing the Electoral Legislation Amendment (Modernisation and Other Measures) Act 2019 (Cth), which requires that, as part of the nomination process, all election candidates are required to disclose certain information in relation to the eligibility criteria in section 44.[82] In relation to the 2019 federal election, this measure appears to have been successful, as only one substantive section 44 challenge—which was subsequently dismissed by the court—was filed in the wake of the election (however former senator Rodney Culleton was referred to police by the AEC for submitting a nomination form stating he was eligible while he was listed as an undischarged bankrupt[83]).

The section 44 challenge that resulted from the election was to the eligibility of the member for Kooyong (Vic.), federal Treasurer Josh Frydenberg. Mr Frydenberg’s eligibility had come under question at the height of the section 44 crisis, but the question of his eligibility was not referred to the High Court.[84]

The High Court’s 2018 decision in Alley v Gillespie confirmed that the only way for the eligibility of a member or candidate to be challenged on the basis of section 44 was by a referral from the relevant Chamber of Parliament, or in the 40 days after the return of the writ for the election under section 355 of the CEA.[85] The petition lodged by Michael Staindl against Mr Frydenberg claimed that Mr Frydenberg was not eligible to sit in the Parliament as he was a citizen of the Republic of Hungary.[86] The challenge against Mr Frydenberg’s eligibility was labelled by critics as being anti-Semitic and ‘disgraceful’ (Mr Frydenberg’s mother was a Holocaust survivor).[87]

The Federal Court, sitting as the Court of Disputed Returns, dismissed the challenge on 17 March 2020, concluding that Mr Frydenberg ‘was not, and never has been, a citizen of Hungary’. Mr Staindl was ordered to pay costs.[88]

On the subject of election signs, two other petitions related to the same election signage in both Mr Frydenberg’s electorate of Kooyong and in the division of Chisholm, which elected Liberal candidate Gladys Liu.[89] The High Court, sitting as the Court of Disputed Returns, referred the petitions for trial to the Federal Court of Australia.[90] The petitions related to signs at polling places in the two electorates that were authorised by the Liberal Party. The signs were in Chinese characters and were in the distinctive purple colour (or a very similar colour) that has been a feature of the AEC’s branding since 2007. The petitions argued that the signs were likely to mislead voters as they translated to instructing voters that the ‘correct’ way to vote was to put a ‘1’ next to the LP candidate.[91] 

On 24 December 2019 the Federal Court, sitting as the Court of Disputed Returns, dismissed the petitions. The Court concluded that the signs were misleading and deceptive in that they purported to be from the AEC and they stated that the ‘correct’ way to vote was to vote for the Liberal Party. The Court found that, even if voters were influenced by the signs, it would not have been in sufficient numbers to affect the result of the election.

The Court ordered that the LP official who authorised the signs demonstrate to the court why their authorisation of the signs was not an illegal act under section 329(1) of the CEA.[92] On 20 February 2020 the Court found that it had insufficient evidence to determine that the official had the requisite state of mind and knowledge to have breached the Act and did not proceed any further with the case.[93]

Clive Palmer’s High Court case on the early release of election results

Under Australia’s preferential voting system, each election in each division comes down to a contest between two candidates, with the preferences on the votes for all of the other candidates eventually counting towards one of those two candidates. To expedite the count, the AEC chooses the candidates most likely to be the final two candidates in each division, and conducts and publishes a preference count against those candidates. The count is referred to as the ‘two-candidate preferred’ (TCP) count, and these candidates are the TCP candidates. The TCP candidates are not made public by the AEC until after the polls have closed in that division.

In April 2019 Clive Palmer took the AEC to the High Court, arguing that the AEC’s practice of choosing two candidates in each division might reduce the incentive for voters in parts of the country where the polls are still open due to time zone differences to vote for minor parties or independents. He sought orders from the Court to prevent the AEC publicly revealing the predicted final two candidates in any electorate until all of the polls in the country had closed.[94]

The High Court unanimously dismissed the case on 7 May 2019. In reasons published in August 2019, some months after the election, the Court stated that Mr Palmer’s contention that releasing the predicted final candidates was not authorised under the CEA, and that releasing the information would distort the voting system ‘lacked a factual foundation’.[95]

The opinion polling failure

In the time period between the 2016 federal election and the 2019 federal election the Coalition only received a TPP opinion (voting intentions) poll result above 50 per cent twice, both in 2017 in the short-lived YouGov poll series.[96] For the majority of the term of the 45th Parliament, and for the entirety of the term in the case of the most prolific pollsters, Labor led the polls in TPP terms (Figure 11 below). While the polls in the lead-up to the election showed a narrowing of the gap, the final polls of all of the major polling companies predicted a narrow win for Labor (see Appendix C).

Figure 11: Opinion polls over the 45th Parliament

Figure 11: line graph showing opinion polls over the 45th Parliament

Note: curved lines represent polling averages for each party and the ‘+’ indicates the 2019 election result for the party.

Source: Compiled by the Parliamentary Library from various sources.

What is particularly notable about the failure of polling in the 2019 federal election was not so much that the polls consistently failed to predict the outcome (although this is significant), but that the polls were so poor in 2019 after having been so accurate in previous elections.

The Library’s analysis of the polling for the 2016 federal election noted:

The accuracy of the pollsters has been analysed by a number of sources, including by polling company ReachTel, who were pleased to report that it was ‘accurate to less than a single percentage point’ for its TPP results. In fact, all of the final polls from the major polling companies came to within one percentage point of the final TPP result, with Essential and Newspoll coming within 0.1 percentage point of the final result. Newspoll had each of the major party primary votes to within 0.3 percentage points of their actual result.[97]

That is, the polls were remarkably accurate in 2016 in Australia, despite significant failures in other western democracies. The 2016 federal election paper went on to note:

Given the significant failures of polling in recent elections in other, similar, western democracies (such as the United Kingdom (UK) in 2015 and the United States in 2016), there is a question as to why the national polling in Australia continues to be so accurate. The increasing number of households without landlines was thought to undermine the representativeness of political polling, however, the one Australian polling company that still uses live phone calls, Ipsos, was substantially less accurate than those who use robopolling (automated phone calls) and online panels, such as ReachTel and Newspoll.

It may be that Australia’s compulsory voting is one of the reasons that polling still works for Australian elections. A review of the failure of the polls at the 2015 UK general election found that three groups were underrepresented in the polling: older voters, who predominately voted Tory; young non-voters, who were polled less frequently than young people who did intend to vote; and busy voters, who were more likely to vote Tory.

Compulsory voting means that Australian polling has much firmer grounds for extrapolating from demographic sub-samples. As long as some older people respond to an online poll, and those people are reasonably representative of the views of older people, it is not difficult to extrapolate to the wider voting population with a degree of accuracy. Sophisticated turnout models to determine which demographics will vote are not necessary. Polling experts also note that the Telemarketing and Research Industry Standard, which allows polling companies to contact numbers on the Do Not Call Register if the polling is for research purposes, also adds to the accuracy of Australian polling. [98]

The major factors that existed in 2016 essentially continued relatively unchanged in 2019: a trend away from fixed phone lines to mobile phones; lack of a phone directory mapping mobile phone numbers to geographical addresses; and compulsory voting and high levels of turnout. Any analysis of the 2019 polling failure should be viewed within this wider context.

It has been noted that while seat-level polling continues to be highly unreliable, and similar failures in state election polls have occurred recently, the amount of polling at the federal level in the lead-up to an election tends to smooth over any inconsistencies between single polls when the entirety of the polling is aggregated.[99] As such, the failure in 2019 was essentially unprecedented.

One political scientist has argued that the polling suggesting that Labor would win likely affected the style of campaigns run by Labor and the Coalition, which as a result may have affected the outcome of the election.[100]

A number of theories have been proposed as to why the polls may have failed so notably in 2019, but as yet there is little evidence to promote one theory over another. The theories that have been advanced for the failure include:

  • herding, where polling companies deliberately or inadvertently adjust their results to be more consistent with other polling results (either their own previous polls or the polls of other polling companies)[101]
  • reluctance of polled voters to tell pollsters that they planned to vote for the Coalition (referred to as the ‘shy Tory’ effect in the UK)[102]
  • a late swing to the Coalition, not captured by any of the polls (however at least one polling expert’s view is that all the evidence suggests that there was no such late swing)[103]
  • voters deciding late, who had told pollsters they ‘don’t know’ who they were voting for, favouring the Coalition[104] and
  • poor sampling procedure that systematically missed some voters who prefer the Coalition.[105]

The Australian Market and Social Research Organisation (AMSRO), a peak body for social research companies, announced an inquiry into the polling failure.[106] However it has been noted that of the pollsters who published national polls for the 2019 federal election, only Ipsos appears to be a member of AMSRO,[107] and the media organisations that had commissioned Ipsos’ polling have ended the relationship after the 2019 federal election.[108]

In a statement YouGov, which owns Galaxy and conducts Newspoll, stated that it would submit a written statement to the inquiry, but that as ‘the polling conducted under our brand YouGov Galaxy during the federal election was commissioned by paid clients it is not appropriate to provide raw data from these polls retrospectively to third parties’.[109]

AMSRO released an interim discussion paper on 20 May 2020 for consultation,[110] focusing on methodological transparency, rather than offering a view as to what went wrong in 2019.[111]

A complicating factor for examining the cause of the polling failure is the lack of transparency in the ways in which polling is conducted in Australia. One long-time polling expert has noted that many Australian polls do publish their sample size and the dates on which the polling is conducted, but do not publish the sampling variance of the poll, and some do not publish the questions that were asked.[112]

In conducting a poll a polling company will typically identify a number of parties and ask the respondent which of those, or ‘other’, would receive their first preference. If the answer is ‘other’ then a list of minor parties may be read out. Analysts have noted that including a minor party in the first list tends to elevate the numbers of respondents who chose that party. Changes in the parties that are part of that list over time leads to difficulties in comparing results over time, particularly if a polling company is not transparent about the changes. Pollsters can also be inconsistent and non-transparent about how they treat the respondents who report they do not know who they are voting for.[113] Even the questions of who owns a polling company and the interests of the owners, and who has paid for a particular poll, are often unclear.[114]

Another polling analyst has reported that while Galaxy, which produces Newspoll, used online methods and robopolls, it was not known what proportion of each sampling approach was used to produce the outcome.[115] Other Australian pollsters use a variety of measures for contacting respondents:

The four active pollsters at this election were YouGov Galaxy, which conducts Newspoll, Ipsos, Essential and Morgan. Galaxy uses online methods and robopolling, Essential uses online methods, Ipsos uses live phone polling and Morgan uses face-to-face interviews. No pollster does only landline polling – Ipsos calls mobiles.[116]

In September 2019 YouGov announced that it would be adjusting its polling procedure in the wake of the 2019 federal election opinion polling failure and that it supported the creation of an Australian Polling Council, along similar lines to the British Polling Council. YouGov stated that it was ‘supportive of a common goal of greater transparency and ongoing public confidence’.[117] In April 2020 YouGov announced the creation of the Australian Polling Council, with Essential and UComms as members. The three members constitute the majority of political opinion polling in Australia. The statement said:

The objectives of the Council are to adhere to a strict set of guidelines intended to:

◦ Ensure standards of disclosure that provide an adequate basis for judging the reliability and validity of the results of survey results that are published in media

◦ Encourage the highest professional standards in public opinion polling and advance the understanding of how polls are conducted and how to interpret poll results

◦ Inform media and the public about best practice in the conduct and reporting of polls

Inspired by similar organisations in the US and Great Britain; YouGov, Essential and UComms believe that the future of Australian polling should operate with guidelines informed by a deep understanding of the market and the complexities of the local landscape.

These guidelines aim to provide an appropriate balance between responsible disclosure to the public, the needs of media outlets, and the intellectual property rights of the polling companies.  Active polling companies in the industry are invited to join in to develop these guidelines.

The Australian Polling Council believes that though published political polling needs to have appropriate standards of transparency, existing standards of disclosure are not fit for purpose. The guidelines ought to be determined by active pollsters as well as accepted by the Australian media. [118]

The organisations stated that they would release a new set of standards and disclosure for polling before July 2020.

In the absence of complete transparency it is difficult to have an informed view of the accuracy of polling, or to diagnose what went wrong in 2019, and whether it is likely to continue to go wrong in the future. It is likely that there is not any one cause of the polling failure, but a combination of several factors, possibly exacerbating each other.

Cost of the election and campaign

Accurate expenditure details are not available for Australian federal elections because parties are not required to report their electoral expenditure. Annual returns to the AEC listing donors over the disclosure threshold ($13,800 for the 2019 federal election) and total party income and expenditure were released by the AEC in February 2020 for the 2018–19 financial year, which includes the May 2019 election.

Amendments to the CEA contained in the Electoral Legislation Amendment (Electoral Funding and Disclosure Reform) Act 2018, which came into effect before the election, changed the election funding provisions to require parties and candidates to submit a claim for any public funding over $10,000. The AEC has not published the details of the claims made by the parties, only whether the claim was accepted in whole, or accepted in part and refused in part, and the amount paid.[119] As such, the changes have not provided substantial additional public information about campaign expenditures. The total amount of public funding paid was $69,647,101.79, which includes $584,640 in automatic payments and $69,062,461.79 in accepted claims.[120] A detailed breakdown of the per candidate and party amounts of public funding is available in Appendix D.

In the absence of any actual figures on expenditure, a number of commentators have provided figures which appear to be either estimates or leaks. While it is not clear where the figures come from, one media article has claimed that Clive Palmer spent $53 million on ads for the campaign, the Greens spent $320,000 and Jacqui Lambie spent $50,000.[121] Analysis published in The Guardian estimated that in the final three weeks of the election Clive Palmer spent $15 million, the Coalition $6.3 million, Labor $7 million and the ACTU $3 million.[122] GetUp reported that it spent $3.5 million on its election campaign.[123] The Liberal Party reportedly spent $1.5 million on its unsuccessful campaign to retain the division of Warringah.[124]

Another media article claims that Labor spent $31.9m on its campaign.[125] The article notes that Labor’s spending was based on its expected election funding at the level it was polling prior to the election, and the resulting lower than anticipated primary vote had resulted in a $1 million deficit for the campaign. Similar estimates of the total Coalition spending do not appear to have been published.

Other reports claim that Clive Palmer’s UAP spent $60m on election advertising, including $8 million on advertising in the final week of the campaign.[126] Mr Palmer himself estimated that the spend was $65 to $70 million.[127] It was estimated that Mr Palmer was the sixth-largest advertising spender in Australia between June 2018 and May 2019, outspending Coles, Toyota and McDonalds, with no other political party in the top 50 spenders.[128] Mr Palmer’s spending was cited in submissions to the JSCEM inquiry into the 2019 federal election as a reason to introduce expenditure caps for federal elections.[129]

The AEC reported that the total cost of the election was $372.5 million, which included $69.6 million in public election funding. The break-down of the costs is reproduced in Table 10 below.[130]

Table 10: Costs of the 2019 federal election

Item Cost ($)
Employee expenses 103 330 537
Labour hire 38 334 000
Property, office supplies and services (inc. venue/equip hire, security, stationery) 38 104 760
Election cardboard and supplies (inc. electoral forms, envelopes, ballot papers, screens) 21 672 260
Contracted services (excluding labour hire) 36 102 260
Travel 2 491 455
Advertising, printing and media services 22 509 500
ICT costs 1 167 533
Mailing and freight services 31 970 699
Other expenses 7 142 707
Sub total 302 825 920
Public funding (a) 69 647 102
Total expenses 372 473 022

Source: AEC Electoral Pocketbook 2019, p 38.

Early voting

The continuing rise in early voting is no longer surprising, having followed a relatively predictable trend over the last decade from the introduction of pre-poll ordinary votes for the 2010 federal election.[131] While pre-poll declaration votes and postal votes have remained relatively stable over recent elections, the proportion of votes cast as pre-poll ordinary votes has continued to climb at an essentially linear rate, single-handedly driving the increase in early voting. For the 2019 federal election the total number of early votes (pre-poll ordinary, pre-poll declaration, and postal) was 6.16 million votes, or 40.8 per cent of total votes.

Figure 12: Type of vote, federal elections 2004–

Figure 12: line graph showing type of vote, federal elections 2004– 

Note: Postal votes, pre-poll ordinary and pre-poll declaration votes are early votes.

Source: Compiled by the Parliamentary Library from AEC election results data.

In its report on the 2016 federal election, the JSCEM recommended that pre-poll voting be restricted to no more than two weeks prior to the election.[132] The CEA allows pre-poll voting to commence five days after the declaration of nominations, which, for the 2019 federal election, resulted in a three-week pre-poll voting period (in 2013 and 2016 early voting commenced on the Tuesday three weeks before the election, resulting in a pre-poll period that was one day shorter). Similar calls for restricting the pre-poll voting period (specifically for pre-poll ordinary votes, but not for postal votes) have been made in relation to the JSCEM’s inquiry into the 2019 federal election.[133]

It is difficult to determine what effect restricting the amount of time available for pre-poll voting would have on the number of early votes cast in future elections. While pre-poll voting runs for between 18 and 25 days,[134] fewer people tend to cast their pre-poll vote early in the period. Since 2010, around half of all pre-poll voters have cast their vote in the five days before the election, with around ten per cent of pre-poll voters voting in the first week of the pre-poll period.[135] If pre-poll voting had been restricted to two weeks for the 2019 election, and none of the roughly 660,000 people who voted in that first week of the 2019 pre-poll period voted later in the period, the number of pre-poll votes in 2019 would still have exceeded the total number of pre-polls in 2016 by more than 1.1 million votes.

Figure 13: Pre-poll ordinary votes received by day, federal elections 2010–

Figure 13: line graph showing pre-poll ordinary votes received by day, federal elections 2010–

Source: Compiled by the Parliamentary Library from AEC data.

The significance of pre-poll voting, however, is not just in the number of early votes, but also relates to the impact of the three-week period on campaigning. The continued rise in early voting would at this point not be a surprise for political parties. Parties and candidates increasingly plan their campaigns around the expectation that many voters will only see the campaigning they do and policies they release in the first few weeks of election periods, rather than the traditional model of a build-up over the election period to an intensive phase and launch in the last week before the election.

Parties realise that any policies they intend to announce to convince voters to vote for them must be released early in the campaign in order to affect the largest number of voters, and that there is ‘no point making a campaign announcement in the final week of the campaign when perhaps 30 per cent of the electorate has already voted’.[136] However, it is also true that the majority of early votes are cast in the last week of the campaign, so in recent elections the traditional party campaign launch a week or two out from election day has still had the potential to swing a substantial proportion of the early voting population (the campaign launches are discussed on page 21, above).

One practical issue that the early voting period raises for candidates and parties is their ability to attend early voting centres in electorates and hand out campaign material such as how-to-vote cards. Smaller parties and independents may find attending early voting centres for three weeks to be particularly challenging. Fewer people handing out how-to-vote material may also affect the formality of votes cast in early voting centres, particularly in areas where there are higher proportions of voters who do not speak English fluently or large numbers of candidates on the ballot paper (both are factors associated with higher levels of informality according to AEC research).[137] For the 2019 election at least, according to Parliamentary Library analysis pre-poll ordinary voting tended to have a slightly lower level of informality compared to election-day voting, suggesting that any effect on informality of access to how-to-vote material is not yet apparent in respect of the 2019 election.[138]

While the continued increase in the early voting rate has been the subject of considerable commentary, less noticed is the apparent peak of postal voting, with a slight decrease in the proportion of votes cast as postal votes in 2019 (albeit with a small rise in the absolute number of postal votes).

In addition, the source of postal vote applications (PVAs) has changed considerably over recent elections.[139] Political parties regularly send PVA forms to supporters and other electors, which are then generally returned to the party and forwarded on to the AEC. This practice has drawn criticism over the years, but it has been a longstanding feature of Australian federal elections.[140] Postal votes may also be applied for by filling in a form and returning it to the AEC, or applying online directly with the AEC.[141]

At the 2010 federal election roughly one-third of postal vote applications came via the Liberal Party, one-third via the ALP, and one-third directly to the AEC. By 2019, the vast majority of applications (almost three quarters of all applications) came directly to the AEC. While the Liberal Party remains the largest party source for PVAs, it now constitutes a declining proportion of total applications, down from around 30 per cent of all applications in 2010 to around 15 per cent in 2019 (Figure 14 below).

Political parties have resisted any diminution of their involvement in the postal voting process,[142] however the additional time involved in PVAs being sent to parties and then being forwarded to the AEC has meant that a number of PVAs have not been received in time for postal votes to be sent to voters.[143] Postal vote campaigns are also expensive for parties to run compared to other forms of campaigning, with parties reporting that the postal costs went up 40 per cent between 2013 and 2017.[144] As fewer voters apply for postal votes through parties, it is possible that parties may in time decide that the benefit of running these campaigns is no longer justifiable from a cost perspective.

Figure 14: Source of postal vote applications, federal elections 2010–

Figure 14: bar chart showing source of postal vote applications, federal elections 2010–

Note: AEC includes online and paper postal vote applications plus general postal voters; not all postal vote applications result in a postal vote—according to the AEC around 10 per cent of the people who applied for a postal vote ended up voting in person.

Source: Compiled by the Parliamentary Library from AEC data

Electoral participation

The story of electoral participation in recent federal elections is a distinctly positive one; however it is concealed to an extent by the complexity of the data. Turnout—how many people voted—and formality—how many people voted correctly—are the main indicators of participation.


Turnout in Australian elections is traditionally calculated as the total number of votes (formal and informal) as a proportion of the enrolled population. The turnout for the 2019 federal election was 91.9 per cent, which is a slight improvement on the 2016 rate but below the rate for 2010 and 2013.

The introduction of direct enrolment and online enrolment in 2012, however, has significantly increased the proportion of the population enrolled going into the 2019 election. The enrolment rate of 96.8 per cent for the 2019 federal election was hailed by the AEC as the ‘best electoral roll in history’.[145] So, although the turnout rate for 2019 was lower than that of 2013, the underlying enrolment rate was significantly higher.

Table 11: Enrolment and turnout rates, federal elections 2010–

Election Votes Eligible population Enrolled population Enrolment rate Turnout VEP turnout(a)
2010 13 131 667 15 499 743 14 088 260 90.90 93.21 84.72
2013 13 726 070 15 925 415 14 712 799 92.40 93.29 86.19
2016 14 262 016 16 493 096 15 676 659 95.00 90.98 86.47
2019 15 088 616 16 960 337 16 424 248 96.80 91.87 88.96

(a) Turnout as a proportion of the voting eligible population (VEP)

Source: Compiled by the Parliamentary Library from AEC election results data.

The two factors of enrolment and turnout can be combined by considering turnout as a proportion of the eligible population (those who have enrolled and those who are eligible to enrol but have not yet done so) to get a more complete view of participation. This is referred to as ‘VEP (Voting Eligible Population) turnout’.

When examined in terms of VEP turnout (Table 11, above), electoral participation has been steadily increasing since 2010 (the earliest election for which the AEC publishes the estimated eligible population). Even the 2016 federal election, which had the lowest turnout since the introduction of compulsory voting, showed a slight increase in VEP turnout compared to the previous elections due to enrolment rate increases.

The increase in VEP turnout shows that an increasing number of Australians are participating in federal elections over time (Figure 15 below). A similar observation about increasing enrolment and decreasing enrolled turnout has been made by the Victorian Electoral Commission in its submission to the Victorian Parliament’s Electoral Matters Committee’s inquiry into the 2018 Victorian state election.[146]

Figure 15: Enrolment and turnout rates, federal elections 2010–

Figure 15: line chart showing enrolment and turnout rates, federal elections 2010–

Source: Compiled by the Parliamentary Library from AEC election results data.

Despite what is generally a good news story for participation, enrolment rates differ markedly between jurisdictions, with the NT having historically low enrolment rates. For the 2019 federal election, the NT VEP turnout of 65.3 per cent means that roughly one-third of NT residents who potentially could have voted in the election did not do so (Table 12 below). When the informality rate is factored in, only 62.2 per cent of the NT’s eligible population cast a vote that could be counted. Low levels of participation has also been observed in recent NT territory-level elections, with the electorate of Arnhem recording a turnout of less than 50 per cent at the 2016 NT election.[147]|

Table 12: Informality, enrolment and turnout rates by state, 2019 federal election

State Informal rate Enrolment rate Turnout VEP turnout
NSW 7.01 98.30 92.09 90.48
VIC 4.66 96.70 92.60 89.56
QLD 4.95 95.50 91.22 87.12
WA 5.44 96.00 90.08 86.47
SA 4.81 97.00 93.06 90.31
TAS 4.39 97.30 94.27 91.69
ACT 3.49 99.20 93.13 92.35
NT 4.69 83.80 77.95 65.30

Source: Compiled by the Parliamentary Library from AEC election results data.


Of the states and territories, NSW had the highest informality rate (the proportion of votes that could not be counted because they were filled in incorrectly or not at all) for House of Representative votes, with eight of the ten top divisions for informality located in NSW. More generally, however, the divisions that had informality rates of 10 per cent or higher were outliers (Figure 16 below).

Figure 16: Distribution of divisional informality rate, 2019 federal election

Figure 16: chart showing distribution of divisional informality rate, 2019 federal election

Note: This chart shows what how many of the divisions in each state swung had what informality rate. NT has too few electorates for a distribution to be calculated.

Source: Compiled by the Parliamentary Library from AEC election results data.

The majority of the high informality divisions in NSW were in Western Sydney, centred around Fowler, Blaxland and Watson. Mallee in Victoria and Lyne in NSW, both regional divisions that returned Nationals members, were notable in having the largest increase in informality since the 2016 federal election.

In its analysis of informal voting at the 2016 federal election the AEC discussed possible causes of informal voting:

– Higher levels of informality are likely to be associated with higher levels of social exclusion or relative disadvantage.

– A change in the number of candidates between elections is a significant predictor of changes in informal voting.

– Voter confusion about the differences between state and federal electoral systems may be contributing to some categories of informal ballots (particularly for HoR ballots with incomplete numbering or where ticks and crosses have been used as the first preference).

– As some informal votes are cast intentionally rather than representing an error on the part of the voter, voters’ attitudes to and opinions of the electoral system or politics in general will also likely influence informality.[148]

In respect of these possible causes, the 2019 NSW state election (which uses optional preferential voting) was held only two months before the federal election—timing which may have contributed to the informal voting level in NSW at the federal election. Additionally, the number of nominations in WA also increased substantially, from 86 in 2016 to 132 in 2019 for the same number of seats.

There was also considerable overlap between the seats across the country with the highest levels of informality in 2016 and those in 2019, suggesting that the underlying causes of informality did not change substantially from one election to the other.

Figure 17: Informality rate in Greater Sydney divisions (NSW), 2019 federal election

Figure 17: map showing informality rate in Greater Sydney divisions (NSW), 2019 federal election

Source: Compiled by the Parliamentary Library from AEC election results data.


Of the 458 candidates who nominated for the Senate, 38 per cent were women (compared to 36 per cent in 2016). In the House of Representatives 32 per cent of the 1,056 candidates were women (compared to 31 per cent in 2016). The candidate gender split of each of the larger parties in the House of Representatives and the Senate is set out in Table 13 below (the Liberal, Nationals, and Country Liberal parties are represented together as the Coalition).

Table 13: Candidate gender by party, 2019 federal election

House of Representatives Male Female Unspecified Male
Coalition 116 46 0 71.6 28.4 0.0
Australian Labor Party 85 66 0 56.3 43.7 0.0
The Greens 88 62 1 58.3 41.1 0.7
United Australia Party 123 28 0 81.5 18.5 0.0
Pauline Hanson’s One Nation 40 19 0 67.8 32.2 0.0
FRASER ANNING’S CNP 40 8 0 83.3 16.7 0.0
Derryn Hinch’s Justice Party 4 4 0 50.0 50.0 0.0
Katter's Australian Party (KAP) 6 1 0 85.7 14.3 0.0
Centre Alliance 0 3 0 0.0 100.0 0.0
Others 212 104 0 67.1 32.9 0.0
Total 714 341 1 67.6 32.3 0.1
Australian Labor Party 17 21 0 44.7 55.3 0.0
Coalition 21 17 0 55.3 44.7 0.0
The Greens 6 29 0 17.1 82.9 0.0
FRASER ANNING’S CNP 18 4 0 81.8 18.2 0.0
Pauline Hanson’s One Nation 9 3 0 75.0 25.0 0.0
Jacqui Lambie Network 2 1 0 66.7 33.3 0.0
Katter's Australian Party (KAP) 2 1 0 66.7 33.3 0.0
Centre Alliance 1 1 0 50.0 50.0 0.0
Derryn Hinch’s Justice Party 1 1 0 50.0 50.0 0.0
Others 203 100 0 67.0 33.0 0.0
Total 280 178 0 61.1 38.9 0.0

Source: Compiled by the Parliamentary Library from AEC election results data.


The main outcome of the 2019 federal election was to essentially retain the status quo. The Government was (unexpectedly) returned with a slightly improved position in the House of Representatives and without control of the Senate, although with an easier path to the passage of legislation through the upper house. The section 44 crisis that permeated the previous parliamentary term appears to have had limited impact on the election itself, with the requirement to complete candidate nomination forms presumably having led to more thorough vetting of candidates.

The election is likely to be remembered for Labor’s failure to win despite expectations arising from the opinion polls; the unprecedented failure of these polls; and the campaign spend of Clive Palmer’s UAP. While the true cause of the opinion polling failure will likely remain a mystery, it may lead to a greater transparency in polling processes which could bring Australia closer to the standards set by similar democracies such as the UK. The formation of the Australian Polling Council is a positive step, however it remains to be seen exactly what changes will occur in the Australian political polling environment (and how quickly they will occur).

The reported $60 million campaign spend of the UAP has led to calls for campaign expenditure caps in federal elections. The political finance scheme at the federal level is much more lightly regulated than the current schemes in almost all Australian states and territories, three of which (NSW, SA, and the ACT) currently have political party campaign expenditure caps for state and territory elections .[149] The Government has not proposed any legislation to introduce caps; Leader of the ALP Opposition Anthony Albanese has indicated that he would introduce expenditure caps in the light of UAP spending in 2019. [150]

Appendix A: Parties contesting the 2019 federal election

Abbreviation Name
ABFA Australian Better Families
AFN Australia First Party
AJP Animal Justice Party
ALP Australian Labor Party
ASP Shooters, Fishers and Farmers
AUC Australian Christians
AUD Australian Democrats
AUP Australian Progressives
AWP Australian Workers Party
CDP Christian Democratic Party (Fred Nile Group)
CEC Citizens Electoral Council
CLP Country Liberals (NT)
CPP Child Protection Party
DHJP Derryn Hinch's Justice Party
DLP Labour DLP
FLUX VOTEFLUX.ORG | Upgrade Democracy!
FUT Science Party
GAP The Great Australian Party
GRN The Greens
IMO Involuntary Medication Objectors (Vaccination/Fluoride) Party
IND Independent
KAP Katter's Australian Party (KAP)
LAOL Love Australia or Leave
LDP Liberal Democrats
LNP Liberal National Party of Queensland
LP Liberal
NCP Non-Custodial Parents Party (Equal Parenting)
NP The Nationals
NP National Party
ON Pauline Hanson's One Nation
REAS Reason Australia
RUA Rise Up Australia Party
SAL Socialist Alliance
SEP Socialist Equality Party
SPP Sustainable Australia
UAP United Australia Party
VNS Victorian Socialists
XEN Centre Alliance

Source: AEC

Appendix B: Section 44 disqualifications

Disqualified name Party Date resigned Date disqualified Reason Electorate Replacement Replacement party Method of replacement Replacement date
Bob Day FF 1/11/2016 5/04/2017 44 (v) SA Lucy Gichuhi LIB Recount 19/04/2017
Rodney Culleton PHON NA 3/02/2017 44 (ii) WA Panagiotis Georgiou PHON Recount 10/03/2017
Scott Ludlam GRN 14/07/2017 27/10/2017 44 (i) WA Jordon Steele-John GRN Recount 10/11/2017
Larissa Waters GRN 18/07/2017 27/10/2017 44 (i) Qld Andrew Bartlett GRN Recount 10/11/2017
Barnaby Joyce NAT NA 27/10/2017 44 (i) New England - - By-election 2/12/2017
Malcolm Roberts PHON NA 27/10/2017 44 (i) Qld Fraser Anning KAP Recount 10/11/2017
Fiona Nash NAT NA 27/10/2017 44 (i) NSW Jim Molan LIB Recount 22/12/2017
Stephen Parry LIB 2/11/2017 8/12/2017 44 (i) Tas. Richard Colbeck LIB Recount 9/02/2018
John Alexander LIB 11/11/2017 11/11/2017 44 (i) Bennelong - - By-election 16/12/2017
Jacqui Lambie JLN 14/11/2017 8/12/2017 44 (i) Tas. Steve Martin NAT Recount 9/02/2018
Skye Kakoschke-Moore XEN 22/11/2017 13/02/2018 44 (i) SA Tim Storer IND Recount 16/02/2018
David Feeney ALP 1/02/2018 1/02/2018 44 (i) Batman Ged Kearney ALP By-election 17/03/2017
Katy Gallagher ALP NA 9/05/2018 44 (i) ACT David Smith ALP Recount 23/05/2018
Justine Keay ALP 10/05/2018 NA 44 (i) Braddon - By-election 28/07/2018
Susan Lamb ALP 10/05/2018 NA 44 (i) Longman - By-election 28/07/2018
Josh Wilson ALP 10/05/2018 NA 44 (i) Fremantle - By-election 28/07/2018
Rebecca Sharkie XEN 11/05/2018 NA 44 (i) Mayo - By-election 28/07/2018

Source: Compiled by the Parliamentary Library.

Appendix C: Final opinion poll results prior to the 2019 federal election

Source Date n Primary TPP
      Coalition Greens Labor Liberal Nationals One Nation Others Coalition Labor
Essential 14/05/2019 1 201 38.5 9.1 36.2 36.2 3.6 6.6 9.6 48.5 51.5
Morgan Face 14/05/2019 1 265 38.5 10 35.5 - - 4 8.5 48 52
Ipsos 15/05/2019 1 842 39 13 33 - - 4 8 49 51
YouGov-Galaxy 15/05/2019 1 004 39 9 37 - - 3 9 49 51
Newspoll 17/05/2019 3 038 38 9 37 - - 3 8 48.5 51.5
Election Results 18/05/2019 41.44 10.4 33.34 36.93 4.51 3.08 8.31 51.53 48.47

Source: Compiled by the Parliamentary Library from various sources.

Appendix D: Public election funding paid

Public election funding was paid for the 2019 federal election at the rate of $2.75642 per first preference vote for parties and candidates who received at least four per cent of the first preference vote. Parties and candidates who qualified for public funding received an initial payment of $10,080 and were permitted to submit an interim and/or final claim, based on election spending, up to a maximum set by their per-vote amount. With a small number of exceptions, parties and candidates successfully claimed around the full amount to which they were entitled.

Party / Candidate Name Votes Initial Payment ($) Interim ($) Final ($) Total ($)
Adam Blakester 13 804 10 080.00 0.00 27 963.82 38 043.82
Alex Dyson 10 797 10 080.00 10 080.00
Alice Thompson 11 975 10 080.00 0.00 22 923.10 33 003.10
Andrew Bock 4 581 10 080.00 0.00 2 545.24 12 625.24
Andrew Wilkie 33 761 10 080.00 0.00 82 965.32 93 045.32
Animal Justice Party 4 393 10 080.00 0.00 2 027.11 12 107.11
Arthur Chesterfield-Evans 4 295 10 080.00 0.00 1 757.02 11 837.02
Australia First Party (NSW) 4 094 10 080.00 10 080.00
Australian Greens 415 894 10 080.00 1 079 317.67 56 806.19 1 146 203.86
Australian Labor Party (ALP) 8 954 603 10 080.00 23 440 130.69 1 233 829.89 24 684 040.58
Centre Alliance 43 849 10 080.00 0.00 110 767.84 120 847.84
Christian Democratic Party (Fred Nile Group) 28 252 10 080.00 0.00 67 782.51 77 862.51
Country Liberals (Northern Territory) 77 350 10 080.00 0.00 203 096.60 213 176.60
Craig Brakey 7 619 10 080.00 0.00 10 917.96 20 997.96
Damien Cole 5 131 10 080.00 0.00 4 061.00 14 141.00
Derryn Hinch's Justice Party 8 905 10 080.00 0.00 14 462.18 24 542.18
Fiona Leviny 5 240 10 080.00 0.00 4 361.44 14 441.44
Grant Schultz 7 585 10 080.00 0.00 10 824.26 20 904.26
Hamish MacFarlane 2 123 10 080.00 10 080.00
Helen Haines 32 664 10 080.00 0.00 79 941.98 90 021.98
Huw Kingston 6 068 10 080.00 0.00 6 643.41 16 723.41
Innes Larkin 5 165 10 080.00 0.00 4 154.74 14 234.74
Jacqui Lambie Network 31 383 10 080.00 41 909.65 2 885.77 54 875.42
Jamie Christie 7 683 10 080.00 0.00 11 094.35 21 174.35
Jarrod Bingham 8 363 10 080.00 0.00 7 764.47 17 844.47
Jason Modica 8 795 10 080.00 0.00 14 159.02 24 239.02
Jeremy Miller 5 169 10 080.00 10 080.00
Julia Banks 13 367 10 080.00 0.00 26 759.45 36 839.45
Katter's Australian Party 63 854 10 080.00 157 606.54 8 295.08 175 981.62
Kerryn Phelps 29 109 10 080.00 0.00 70 144.40 80 224.40
Kevin Mack 19 926 10 080.00 0.00 44 836.06 54 916.06
Liberal Democratic Party 22 598 10 080.00 0.00 52 200.00 62 280.00
Liberal Party of Australia 10 705 090 10 080.00 26 599 996.30 959 533.79 27 569 610.09
Louise Stewart 6 902 10 080.00 8 494.81 447.10 19 021.91
Nathan Herbert 5 473 10 080.00 0.00 5 003.59 15 083.59
National Party of Australia - N.S.W. 472 855 10 080.00 2 432 199.87 128 010.52 2 570 290.39
National Party of Australia - Victoria 138 715 10 080.00 610 663.54 700 674.39 1 321 417.93
Oliver Yates 8 890 10 080.00 0.00 14 420.84 24 500.84
Pauline Hanson's One Nation 1 077 825 10 080.00 1 883 207.10 947 479.35 2 840 766.45
Queensland Greens 571 036 10 080.00 1 485 510.46 78 184.76 1 573 775.22
Ray Kingston 8 621 10 080.00 0.00 13 679.48 23 759.48
Reason Australia 8 895 10 080.00 0.00 3 027.54 13 107.54
Robert Oakeshott 25 847 10 080.00 0.00 61 154.33 71 234.33
Senate Group - Anthony Pesec and Gary Kent 12 604 10 080.00 0.00 24 656.62 34 736.62
Shooters, Fishers and Farmers Party 30 513 10 080.00 70 313.13 3 700.69 84 093.82
Simone Karandrews 3 849 10 080.00 10 080.00
Sue Fraser-Adams 2 684 10 080.00 10 080.00
The Australian Greens – Victoria 833 005 10 080.00 2 171 397.69 114 284.09 2 295 761.78
The Greens (WA) Inc 333 747 10 080.00 864 240.40 45 486.34 919 806.74
The Greens NSW 805 028 10 080.00 0.00 2 118 372.89 2 128 452.89
The National Party of Australia (WA) Inc 18 673 10 080.00 0.00 41 382.79 51 462.79
Tim Bohm 4 062 10 080.00 0.00 1 114.87 11 194.87
Timothy Jerome 4 220 10 080.00 0.00 1 550.32 11 630.32
United Australia Party 235 988 10 080.00 0.00 640 302.92 650 382.92
Van Tran 5 306 10 080.00 0.00 4 543.34 14 623.34
Victorian Socialists 12 453 10 080.00 23 028.45 1 212.02 34 320.47
Will Landers 6 730 10 080.00 10 080.00
Zali Steggall 40 034 10 080.00 0.00 100 253.70 110 333.70
Total   584 640.00 60 868 016.30 8 194 446.49 69 647 102.79

Note: Votes in the above table include the total votes for each individual party’s candidates where the party was a member of a group on the Senate ballot paper, however maximum public funding amounts for Senate groups are calculated on the basis of all of the votes for each of the candidates in the group, and the respective parties make their own decision as to how much of the total group funding each party receives. As such, in some cases the received public funding equals more than the vote-based entitlement would indicate.

Source: Parliamentary Library calculations based on AEC election results.

[1].          C Knaus, ‘Scott Ludlam resigns from Australian Senate after finding out he has New Zealand citizenship’, The Guardian, 14 July 2017.

[2].          Prior to this, in January 2017, the President of the Senate had notified the WA Governor that Mr Culleton’s Senate seat was vacant as a result of Federal Court bankruptcy findings against Mr Culleton in December 2016: S Parry (President of the Senate), ‘Qualifications of Senators’, Senate, Debates, 7 February 2017, p. 2.

[3].          Mr Day had resigned from the Senate in November 2016: B Day, Senator Bob Day resignation, media release, 1 November 2016.

[4].          Bob Day’s replacement Lucy Gichuhi left Family First to join the Liberals; Malcolm Roberts’ replacement Fraser Anning left One Nation to sit first as an independent, then as a Katter’s Australian Party senator, and then once again as an independent; Jacqui Lambie’s replacement Steve Martin left the Jacqui Lambie Network to sit as a Nationals senator; and Skye Kakoschke-Moore’s replacement, Tim Storer, left the Centre Alliance to sit as an independent.

[5].          K Hogan, ‘Statement by Kevin Hogan’, Kevin Hogan MP website, accessed 25 June 2020.

[6].          Australian Constitution, section 40. See also N Horne, ‘Hung Parliament 2018 – continued’, Flagpost, Parliament Library blog, 6 November 2018.

[7].          J Banks, ‘Statements on Indulgence: Member for Chisholm’, House of Representatives, Debates, 27 November 2018, pp. 11571–2.

[8].          P Karp and G Hutchens, ‘Barnaby Joyce quits as Australia's deputy prime minister and Nationals leader’, The Guardian, 23 February 2018.

[9].          The Liberal primary vote situation is complicated by the fact that it has run as a combined party with the National Party in Queensland since 2010; S Barber, Federal election results 1901-2016-Reissue #2, Research paper series, 2019–19, Parliamentary Library, Canberra, 5 July 2018.

[10].        The division of Denison was renamed to Clark in 2017.

[11].        M Koslowski, ‘”School fete on steroids” creates history’, The Age, 21 May 2019, p. 12.

[12].        Tony Abbott was first elected to Warringah at a by-election on 26 May 1994, was re-elected at the 1996 general election and went on to serve eight full terms. He was Prime Minister from 18 August 2013 to 15 August 2015.

[13].        The AEC does also calculate two-party preferred totals, which are counted to Labor or the Coalition regardless of the TCP candidates in the contest, however unlike TCP preferences does not indicate which candidates those preferences came from.

[14].        D Muller, The new Senate voting system and the 2016 election, Research paper series, 2017–18, Parliamentary Library, Canberra, 25 January 2018.

[15].        Senator Bernardi, the sole senator for the Australian Conservatives, was not up for election however dissolved his party prior to the election to sit in the Senate as an independent.

[16].        Australia, Senate, Journals, 39, 10 February 2020.

[17].        Section 13 of the Australian Constitution requires the election for Senate vacancies within one year before the position becomes vacant. The terms for the senators elected for short terms in 2016 expired on 1 July 2019.

[18].        L Tingle, ‘Election date preferences are quickly exhausted’, Australian Financial Review, 17 November 2018, p. 47.

[19].        R Ferguson and P Riordan, ‘Morrison vows to avoid Howard slips’, The Australian, 3 April 2019, p. 18.

[20].        R Lundie, ‘So when is the next election?’: Australian elections timetable as at 1 September 2016, Research Paper Series, 2016–17, Parliamentary Library, Canberra, 2016.

[21].        A Remeikis, ‘Morrison dismisses concerns over taxpayer-funded ads as election date delayed’, The Guardian, 7 April 2019.

[22].        T Rogers (Electoral Commissioner), Evidence to Finance and Public Administration Legislation Committee Estimates hearing, 4 April 2019.

[23].        A Hawke, ‘Issue of writs for election of members of the House of Representatives and of senators to represent the Australian Capital Territory and the Northern Territory’, 11 April 2019.

[24].        The writ is a direction to the Electoral Commissioner to hold an election and nominates the dates when key events for the election must happen.

[25].        Commonwealth Electoral Act (1918)

[26].        Australian Electoral Commission (AEC), ‘AEC returns writs for 2019 federal election’, AEC media release, 21 June 2019.

[27].        T Rogers (Electoral Commissioner), Evidence to Finance and Public Administration Legislation Committee Estimates hearing, 9 April 2019.

[28].        A Workman, ‘Easter Sunday election blackout’, The Australian, 10 April 2019, p. 2.

[29].        T Rogers, Notification of determination, 31 August 2017.

[30].        Redistribution of Queensland in electoral divisions, March 2018.

[31].        Redistribution of Victoria into electoral divisions, July 2018; T Rogers, Determination of names and boundaries of federal electoral divisions in Victoria, 13 July 2018.

[32].        Redistribution of the Australian Capital Territory into electoral divisions, July 2018; T Rogers, Determination of names and boundaries of federal electoral divisions in the Australian Capital Territory, 13 July 2018.

[33].        Redistribution of South Australia into electoral divisions, July 2018; T Rogers, Determination of names and boundaries of federal electoral divisions in South Australia, 20 July 2018.

[34].        Redistribution of the Northern Territory into electoral divisions, February 2017; D Cowdroy, Determination of names and boundaries of federal electoral divisions in the Northern Territory, 7 February 2017.

[35].        Redistribution of Tasmania into electoral divisions, November 2017; D Cowdroy, Determination of names and boundaries of federal electoral divisions in Tasmania, 14 November 2017.

[36].        AEC, National seat status.

[37].        See, for example, R Viellaris, D Atkins, F Whiting, D Houghton, S Wardill and L Hunder-Nolan, ‘The verdict’, Courier Mail, 18 May 2019, p. 56.

[38].        P Durkin, ‘Vic Libs brace for backlash’, AFR Weekend, 18 May 2019, p. 9; M Ludlow, ‘Labor hopes for swag of Qld seats fade’, AFR Weekend, 18 May 2019, p. 13.

[39].        P Coorey, ‘Larger than life: Hawke’s final gift to the Labor party’, AFR Weekend, 18 May 2019, p. 6.

[40].        M Ludlow, ‘Unity the core message of Labor’s launch’, Australian Financial Review, 6 May 2019, p. 5.

[41].        T Bramston, ‘Forget the party: it’s all about me’, The Australian, 14 May 2019, p. 4.

[42].        S Maiden, ‘Have promise, will travel: the true cost of election jetsetting’, The New Daily, 5 May 2019; it is not clear from the report whether the agreement applied to all parties or only the major parties

[43].        M Koziol and E Bagshaw, ‘Politicians charged taxpayers millions for election travel’, Sun Herald, 1 August 2019, p. 1.

[44].        M Grattan, ‘View from the hill: budget tax-upmanship as we head towards polling day’, The Conversation, 2 April 2019.

[45].        The Budget was not able to be passed through Parliament prior to the election and a supply bill was passed as an interim measure.

[46].        S Morrison, ‘It’s a choice between jobs and strong economy or higher taxes’, Courier Mail, 12 April 2019, p. 29.

[47].        B Shorten, ‘ALP is the “fair go” team’, Courier Mail, 18 May 2019, p. 54.

[48].        The negative gearing policy had been taken to the 2016 election and the franking credits policy was introduced in March 2018. C Emerson and J Weatherill, Review of Labor’s 2019 federal election campaign, ALP. November 2019

[49].        G Hutchins, L Scarr, A Hennessy and P Wearne, ‘Election issues by the dozen’, West Australian, 18 May 2019, p. 86.

[50].        B Shorten, ‘ALP is the “fair go” team’, Courier Mail, 18 May 2019, p. 54.

[51].        M McKenna and J Owens, ‘Adani is hurting us in rural Queensland, Labor concedes’, The Australian, 17 April 2019, p. 1.

[52].        G Brown, ‘CFMEU waning on Adani approval’, The Australian, 22 April 2019, p. 1.

[53].        P Riordan, ‘Di Natale dials up climate heat on ALP’, The Australian, 15 April 2019, p. 4; R Wuth, ‘Adani protest tour sets off’, Hobart Mercury, 17 April 2019, p. 12.

[54].        ‘Adani protest rally signs off with island serenade’, Courier Mail, 6 May 2019, p. 12.

[55].        E Bagshaw, ‘Brown’s Green convoy hurt Labor: Di Natale’, The Age, 7 October 2019, p. 5.

[56].        G Brown, ‘Labor climate policy “put the fear of God in people and alienated blue-collar base”’, The Australian, 15 October 2019, p. 6.

[57].        Emerson and Weatherill, op cit.

[58].        B Preiss and R Harris, ‘Battle for Liberal hearts and minds in Menzies’, Sunday Age, 15 March 2020, p. 25.

[59].        S Hutchinson and K Loussikian, ‘Hold that thought on preselection’, The Age, 19 March 2020, p. 4.

[60].        D Muller, ‘Online political communication—does this post need to be authorised?’, FlagPost, Parliamentary Library Blog, 23 April 2019.

[61].        N Evershed and C Knaus, ‘Lies, Trump enthusiasts and car parks: the Australian election campaign waged by Facebook ads’, The Guardian, 1 May 2019.

[62].        T Rogers (Electoral Commissioner, AEC), Evidence to Joint Standing Committee on Electoral Matters Inquiry Conduct of the 2019 federal election and matters related thereto, 6 December 2019.

[63].        AEC, Submission to Joint Standing Committee on Electoral Matters Inquiry Conduct of the 2019 federal election and matters related thereto, 2019.

[64].        Ibid, p. 34.

[65].        Ibid.

[66].        A Bogle, ‘Clive Palmer’s ad deluge shows Google and Facebook need to step up transparency, experts say’, ABC News, 22 May 2019.

[67].        A Johnson (Acting Chief Legal Officer, AEC), Evidence to Joint Standing Committee on Electoral Matters Inquiry Conduct of the 2019 federal election and matters related thereto, 6 December 2019.

[68].        M Garlick, ‘Working to safeguard elections in Australia’, Facebook Newsroom, blog, 4 April 2019.

[69].        N Evershed and C Knaus, ‘Lies, Trump enthusiasts and car parks: the Australian election campaign waged by Facebook ads’, The Guardian, 1 May 2019.

[70].        K Murphy, ’”It felt like a big tide”: how the death tax lie infected Australia’s election campaign’, The Guardian, 8 June 2019.

[71].        S Holderhead, ‘Grave fears on death tax’, Daily Telegraph, 21 July 2018, p. 17.

[72].        J Frydenberg (Treasurer), Death taxes - you don't say, Bill!, media release, 24 January 2019.

[73].        K Murphy, ’”It felt like a big tide”: how the death tax lie infected Australia’s election campaign’, The Guardian, 8 June 2019.

[74].        Ibid.

[75].        K Cox, ‘Political ads can lie if they want, Facebook confirms’, Ars Technica, 11 October 2019.

[76].        N Clegg, ‘Facebook, elections and political speech’, Facebook Newsroom, 24 September 2019.

[77].        E Glazer, ‘Facebook won’t limit political ads’, Weekend Australian, 11 January 2020, p. 28.

[78].        A Bogle, ‘Clive Palmer’s ad deluge shows Google and Facebook need to step up transparency, experts say’, ABC News, 22 May 2019.

[79].        C Pash, ‘Cash splash: A record-breaking federal election ad spend approaches’, AdNews, 19 March 2019.

[80].        F Hunter & N Gladstone, ‘As voters tune out, ‘old fashioned’ ad campaigns are still in favour’, Sydney Morning Herald, 30 April 2019.

[81].        JSCEM, Excluded: the impact of section 44 on Australian democracy, Canberra, May 2018, p. 60.

[82].        Electoral Legislation Amendment (Modernisation and Other Measures) Act 2019

[83].        P Karp, ‘Electoral commission accused of failing to investigate Palmer candidates’, The Guardian, 2 May 2019.

[84].        M Dreyfus, ‘Not “going after” Josh’, Australian Jewish News, 15 December 2017, p. 15.

[85].        Alley v Gillespie [2018] HCA 11.

[86].        Australia, Parliament, Court of Disputed Returns —Election petitions—Michael Robert Staindl v. Joshua Anthony Frydenberg in respect of the election of a member for the House of Representatives for the Division of Kooyong, Parl. Paper, 1 August 2019.

[87].        ‘Frydenberg defended’, Australian Jewish News, 20 September 2019, p. 7; R Baxendale, ‘Bid to save Treasurer all about seats: challenger’, The Australian, 20 August 2019, p. 7.

[88].        Staindl v Frydenberg [2020] FCAFC 41.

[89].        Australia, Parliament, Court of Disputed Returns —Election petitions—Oliver Tennant Yates v. Joshua Anthony Frydenberg in respect of the election of a member for the House of Representatives for the Division of Kooyong, Parl. Paper, 1 August 2019; Australia, Parliament, Court of Disputed Returns —Election petitions—Naomi Leslie Hall v. Gladys Liu in respect of the election of a member for the House of Representatives for the Division of Chisholm, Parl. Paper, 1 August 2019.

[90].        Australia, Parliament, Court of Disputed Returns —Election petitions—Copy of the order of the High Court of Australia remitting the election petitions: Oliver Tennant Yates v Joshua Anthony Frydenberg & Anor, and Vanessa Claire Garbett (formerly Naomi Leslie Hall) v Gladys Liu & Anor, Parl. Paper, 19 September 2019.

[91].        P Karp, ‘Josh Frydenberg and Gladys Liu election challenges should be thrown out, AEC says’, The Guardian, 8 October 2019.

[92].        Garbett v Liu [2019] FCAFC 241.

[93].        Garbett v Liu (No 2) [2020] FCAFC 14.

[94].        Palmer v Australian Electoral Commission [2019] HCA 24

[95].        Ibid.

[96].        YouGov acquired Galaxy Research in late 2017 and discontinued its own political polling series.

[97].        D Muller, Double, double toil and trouble: the 2016 federal election, Research Paper Series, 2016–17, Parliamentary Library, Canberra, 2017, p. 11.

[98].        Ibid.

[99].        K Bonham, ‘The miracle is over: the 2019 Australia federal election poll fail’, Dr Kevin Bonham’s blog, 21 May 2019.

[100].      R Tiffen, ‘How the polls mapped a road to Liberal Victory’, Canberra Times, 14 September 2019, p. 29.

[101].      B Schmidt, ‘The mathematics does not lie: why polling got the Australian election wrong’, The Guardian, 20 May 2019; M Koziol, ‘Pollster trashed out-of-step research’, Sunday Age, 9 June 2019, p. 6.

[102].      N Bonyhady, ‘Calls for transparency, national standards after polls miss result’, Sydney Morning Herald, 24 May 2019, p. 4.

[103].      P Lewis, ‘As pollsters, we are rightly in the firing line after the Australian election. What happened?’, The Guardian, 21 May 2019; M Goot, ‘Did late deciders confound the polls?Inside Story, 19 September 2019.

[104].      Lewis, op. cit.

[105].      RP Scully, ‘How did pollsters get the Australian election results so wrong?’, New Scientist, 20 May 2019.

[106].      AMSRO, ‘Inquiry into the performance of the opinion polls at the 2019 federal election’, AMSRO website.

[107].      W Bowe, ‘Essential Research leadership polling’, The Poll Bludger, 2 July 2019.

[108].      T Maguire, ‘We’re pressing pause on our political polling’, Sydney Morning Herald, 24 May 2019, p. 22.

[109].      YouGov, YouGov supports establishing Australian Polling Council, YouGov website, 19 September 2019.

[110].      AMSRO, op cit.

[111].      AMSRO, Discussion paper: Disclosure standards for election and political polling in Australia, Inquiry into the performance of the opinion polls at the 2019 Australian federal election, May 2020.

[112].      M Goot, ‘Who controls opinion polling in Australia, what else do we need to know about the polls, and why it matters’, Inside Story, 15 May 2019.

[113].      K Bonham, ‘Field guide to opinion pollsters: 46th Parliament edition’, Kevin Bonham’s Blog, 25 December 2019.

[114].      Ibid.

[115].      A Beaumont, ‘Newspoll probably wrong since Morrison became PM; polling has been less accurate at recent elections’, The Conversation, 23 May 2019.

[116].      A Beaumont, ‘Newspoll probably wrong since Morrison became PM; polling has been less accurate at recent elections’, The Conversation, 23 May 2019.

[117].      YouGov, op cit.

[118].      YouGov, YouGov, Essential and UComms to form Australian Polling Council, media release, 21 April 2020.

[119].      AEC, ‘Transparency Register’.

[120].      AEC, 2019 federal election: election funding payments finalised, media release, 12 December 2019.

[121].      R Denniss, ‘The maw of averages’, The Saturday Paper, 6 July 2019, p. 7.

[122].      K Murphy, ‘Clive Palmer outspends McDonald’s, Toyota and Coles to advertise his political party’, The Guardian, 16 August 2019.

[123].      A Tillett, ‘GetUp to go easy on Labor's pro-coal MPs’, Australian Financial Review, 17 October 2019, p. 7.

[124].      K Loussikian and S Hutchinson, ‘Abbott falls on his sword for Liberals’, Sydney Morning Herald, 29 November 2019, p. 2.

[125].      S Maiden, ‘A $31.8 million election bill leaves Labor in the red’, The New Daily, 6 October 2019.

[126].      P Karp, ‘Clive Palmer $60m election spend shows need for cap, advocates say’, The Guardian, 23 September 2019.

[127].      S Maiden, ‘$65 million and counting, Clive Palmer labels Labor a sore loser on election spend’, The New Daily, 20 October 2019.

[128].      ‘Clive Palmer outspends McDonald’s, Toyota and Coles to advertise his political party’, op cit.

[129].      Ibid.

[130].      AEC, 2019 Electoral Pocketbook, 2020.

[131].      Pre-poll ordinary votes are early votes cast in person by a voter at a pre-poll voting centre in their own electorate. Pre-poll declaration votes are early votes that are cast at a polling place outside the voters’ home electorate.

[132].      Joint Standing Committee on Electoral Matters, Report on the conduct of the 2016 federal election and matters related thereto, Canberra, November 2018.

[133].      P Karp, ‘Calls to cut pre-poll voting amid claims it is eroding “integrity” of elections’, The Guardian, 3 October 2019.

[134].      Section 200D of the Commonwealth Electoral Act 1918 specifies that early voting commences five days after the close of nominations, so the length of the early voting period is dictated by the date of polling day relative to the date of the close of nominations.

[135].      Almost 14 per cent of the pre-poll voters in 2019 voted in the first week, which was higher than the average for 2010 to 2019 of 10.5 per cent in the first week, however the first week of early voting was one day longer than previous elections, which may account for this difference.

[136]. R Smith, I Brightwell, R Buckland, M Drum, J Harbord, A McIver, S Mills, C Morgan, M Radcliffe and R Wen, Implications of changes to voting channels in Australia, Electoral Regulation Research Network, December 2018, p. 68.

[137].      AEC, Informal voting: 2016 House of Representatives elections, 2016.

[138].      On average the informal voting level for pre-poll votes was around 1 percentage point lower than for election-day votes and in only seven divisions, most of which had a fairly low level of informality, was the informal voting rate for pre-poll votes higher than for election-day votes.

[139].      Not all postal vote applications will result in a postal vote being cast as voters may apply for a postal vote but may actually cast their vote by another method.

[140].      H Davidson, ‘Political parties’ postal vote mailouts spark concerns voters could be misled’, The Guardian, 17 April 2019.

[141].      In each of these cases the actual postal vote is sent directly to the elector by the AEC—it is only the application form that may be mediated by political parties.

[142].      N Kelly, Directions in Australian electoral reform, ANU E Press, Canberra, 2012.

[143].      AEC, Submission to the Joint Standing Committee on Electoral Matters, op cit., p. 17.

[144].      R Smith et al , Implications of changes to voting channels in Australia,  op. cit.

[145].      AEC, ‘The best electoral roll in history’, AEC media release, 23 April 2019.

[146].      Victorian Electoral Commission, ‘Submission to the Electoral Matters Committee’, EMC Submission No 92, 30 August 2019.

[147].      ‘Voter apathy rife in the NT’, Northern Territory News, 22 August 2019, p. 7.

[148].      AEC, Informal voting: 2016 House of Representatives elections, 2016, p. 4; at the time of publication the AEC had not released similar analysis for the 2019 federal election.

[149]. Queensland is also considering implementing expenditure caps for state elections. D Muller, ‘Election funding and disclosure in Australian states and territories: a quick guide’, Research paper series, 2018–19, Parliamentary Library, Canberra, 28 November 2018; S Parnell, ‘Labor’s $1m election limit for unions, GetUp’, The Australian, 30 October 2019, p. 8.

[150].M Wray, ‘Palmer cash not welcome’, Courier Mail, 24 February 2020, p. 10.


For copyright reasons some linked items are only available to members of Parliament.

© Commonwealth of Australia

Creative commons logo

Creative Commons

With the exception of the Commonwealth Coat of Arms, and to the extent that copyright subsists in a third party, this publication, its logo and front page design are licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Australia licence.

In essence, you are free to copy and communicate this work in its current form for all non-commercial purposes, as long as you attribute the work to the author and abide by the other licence terms. The work cannot be adapted or modified in any way. Content from this publication should be attributed in the following way: Author(s), Title of publication, Series Name and No, Publisher, Date.

To the extent that copyright subsists in third party quotes it remains with the original owner and permission may be required to reuse the material.

Inquiries regarding the licence and any use of the publication are welcome to

This work has been prepared to support the work of the Australian Parliament using information available at the time of production. The views expressed do not reflect an official position of the Parliamentary Library, nor do they constitute professional legal opinion.

Any concerns or complaints should be directed to the Parliamentary Librarian. Parliamentary Library staff are available to discuss the contents of publications with Senators and Members and their staff. To access this service, clients may contact the author or the Library‘s Central Entry Point for referral.