The 2019 House of Representatives election saw a record vote for minor parties, and as a consequence, a record number of seats where preferences needed to be distributed.
As has been the case for three decades, the Labor Party benefited most from preferences, both in flows of preferences and in seats won from second place. But matching the decline in Labor’s first preference support in 2019, preference flows to Labor were weaker than at any election since 2001.
As in the past, Green preferences overwhelmingly favoured Labor, though it was a different story with other parties. The United Australia Party and Pauline Hanson’s One Nation polled 6.5% between them, but where previously preferences flows from both parties had only weakly favoured the Coalition, in 2019 voters for both parties had a clearer preference for the Coalition over Labor.
The Minor Party Vote
The graph below plots the combined vote for minor parties and Independents at House of Representatives and Senate elections since 1949. Support in the House rose from 23.2% in 2016 to a record 25.9% in 2019. In the Senate, third party support declined from 35.0% to 33.1% as the number of minor parties contesting election fell due to the new electoral system.
A consequence of a higher minor party vote was an increase in the number of seats going to preferences. Of the 151 House divisions, 46 were decided on first preferences. There were 105 contests where preferences needed to be distributed, at 69.5%, the highest proportion ever recorded. Twelve of the 102 seats were won on preferences by second placed candidates, down from 14 in 2010, 15 in 2013 and 16 in 2016.
The Coalition won 77 seats, 31 on first preferences and 46 after leading on first preferences. (Figures corrected 4/1/22) Of Labor’s 68 seats, it won 14 on first preferences, 44 after leading on first preferences and another 10 from second place on preferences. Of the six crossbench members, Independent Andrew Wilkie won a majority of the first preference vote, Rebekha Sharkie in Mayo and Helen Haines in Indi won from second place after preferences, the three other crossbenchers being elected after leading on the first count. (There is more on wins from behind further down this post.
Preference Flows by Party
The table below lists the preference flows of the three largest minor parties, and summarises all other parties into a single total. (A sortable table separately listing all parties can be found at the bottom of this post.)
2019 House of Representatives Election – Summary of Preference Flows
| to Coalition | to Labor | First Preferences | ||||
|---|---|---|---|---|---|---|
| Party | Votes | Pct | Votes | Pct | Votes | Pct |
| The Greens | 263,830 | 17.8 | 1,219,093 | 82.2 | 1,482,923 | 10.4 |
| United Australia Party | 318,413 | 65.1 | 170,404 | 34.9 | 488,817 | 3.4 |
| Pauline Hanson’s One Nation | 286,049 | 65.2 | 152,538 | 34.8 | 438,587 | 3.1 |
| Others/Independents | 583,710 | 49.3 | 600,321 | 50.7 | 1,184,031 | 8.3 |
| Total Minors and Independents | 1,452,002 | 40.4 | 2,142,356 | 59.6 | 3,594,358 | 25.2 |
Over the years I have many times been misquoted as to the meaning of the above data. Voters, not parties, control preferences. The preferences of every ballot paper flow 100% where the voter directs them. The above data is simply an aggregation of each voter’s individual decision. Anyone who interprets this as the Greens giving 17.8% of preferences to the Coalition doesn’t understand preferential voting.
One Nation’s preference flows to the Coalition were considerably higher than at any previous election. When it first burst on to the scene in 1998, One Nation polled 8.4% and only 53.7% of preferences flowed to the Coalition. At its second election in 2001, the party polled 4.3% with 55.9% flowing to the Coalition. From a small number of contests in 2016, One Nation preferences flowed 50.4% to the Coalition, and actually favoured Labor in Herbert and Longman, the two Queensland seats gained by Labor.
As shown in the table, One Nation preferences flowed 65.3% to the Coalition in 2019, in line with preference flows at the 2017 Western Australian and Queensland elections. One Nation preferences favoured the Coalition in all 59 seats that it contested.
Clive Palmer’s United Australia Party party recommended preferences for the Coalition in every seat in 2019, as had the Palmer United Party in 2013. In 2013 the Palmer United Party polled 5.5% with 54.0% of preferences flowing to the Coalition. In 2019 United Australia polled only 3.4% but preferences to the Coalition were 65.1%.
Green preference flows to Labor rose slightly from 81.9% to 82.2% in 2019.
But Green support has slipped as a proportion of minor party support, sliding from a high point of 64% in 2010 to only 41% at the 2019 election. Support for other minor parties and Independents has more than doubled since 2010, rising from 6.7% to 14.8% in 2019.
Between 2016 and 2019, other party support rose but the flow of preferences to Labor declined from 49.2% to 43.7%. With non-Green parties making up a greater proportion of the other party vote, the overall flow of preferences to Labor declined from 63.6% in 2016 to 59.5% in 2019.
This was the lowest overall flow of preferences to Labor since 2001, the last election before the Australian Democrats were supplanted by the Greens as Australia’s largest minor party. The graph below shows the overall minor party preference flows at elections since full preference data was first published in 1996. The Greens have been separated from other parties for elections since 2004.
The increase in preferences to the Coalition from One Nation and the United Australia Party may be a function of the party’s lower support. Rather than being a vehicle for an anti-politics vote as in the past, in 2019 both parties drew support from a smaller cohort of voters sending a clearer anti-Labor message.
At One Nation’s first election in 1998, the Coalition suffered a 14.4% loss of first preference support in Queensland, Labor’s vote rose 2.9%, and One Nation polled 14.4%. But One Nation preferences flowed only 57.7% to the Coalition, producing a net 7.2% swing to Labor after preferences. Did One Nation’s campaign against the Howard government in 1998 both take votes of the Coalition, and encourage those voters to preference Labor?
In 2019, One Nation polled 8.9% in Queensland and abandoned its strategy of recommending preferences against sitting MPs, a policy that usually hurt the Coalition. In Queensland One Nation preferences flowed 67.2% to the Coalition. In non-metropolitan seats that One Nation contested in both 2016 and 2019, the changes in first preference support looked very much like One Nation taking votes from Labor.
Contesting only 59 seats at the 2019 election, including 29 in Queensland, One Nation’s vote and its preferences were more important than the United Australia Party. There were 14 seats where One Nation polled more than 10% , 11 of those in Queensland, and two others in the NSW coal mining seats of Hunter and Paterson. One Nation preferences flows favoured the Coalition in every seat.
If you treat the Palmer United Party from 2013 and United Australia Party in 2019 as the same party, you see a similar though less politically important shift. In Queensland in 2013, the defeat of the Rudd Labor government saw Labor and the Greens suffer an 8.5% fall in first preference support, and the Coalition lose 1.7%. The Palmer United Party polled 11.0%, but only 54.9% of preferences flowed to the Coalition, so in the end Labor suffered only a 1.9% two-party preferred swing in Queensland.
In 2019 the United Australia Party polled only 3.5% in Queensland, but its preferences flowed 68.9% to the Coalition. Across Australia, the United Australia Party contested all 151 divisions, but its vote passed 5% in only 19. United Australia’s preference flows didn’t significantly impact any results, and were probably less important than the anti-Labor message and broadcast weight of Clive Palmer’s political advertising.
Across the country the Coalition’s first preference vote fell 0.6%, the combined first preference support for Labor plus the Greens fell 1.2%, and the two-party preferred swing against Labor was 1.1%. All that the extra preferences to the Coalition did was fill the gap left by the drop in its first preference support, the overall two-party preferred swing reflecting the overall loss of first preference support for left parties.
The picture is a little more complex in Queensland, where Coalition first preference support rose 0.5%, Labor’s fell 4.2% and the Greens rose 1.5%. The seat by seat picture suggests Labor lost support to the Greens and the Coalition in the south-east, where in the rest of the state the shift was more from Labor to One Nation, especially in seats One Nation also contested in 2016. Many others have already pointed to the impact on Labor’s vote of the Adani protest convoy, and the pattern of vote loss specific to Queensland would back those suggestions.
My observation on One Nation and United Australia preference flows is that for the first time at a Federal election, both parties delivered preference flows that fitted more closely to their position on the political spectrum.
With Labor attacking both parties, Labor being attacked in return, and the Coalitiojn concentrating on attacking Labor, anyone voting for One Nation or United Australia had lots of messaging on which major party to preference.
Despite the rise in minor party vote, in most seats Australian politics still comes down to a two-party contest. The two-party swing at the election was 1.1%, the flows of preferences doing little to alter the shift in first preference vote shares of the major parties. The stronger flows of preferences from One Nation and United Australia suggest that the underlying two-party contest was framing how third party voters made their decision on preferences.
Green voters have always been ideological rather than strategic in directing preferences. For the first time at a Federal election, it appears that One Nation and United Australia Party voters may have behaved the same way, the two-party contest between Labor and their Coalition framing their preference choices, at the same time as they sent a general protest message with their first preference.
Non-Major Party Contests
There were fifteen seats that did not finish as two-party preferred contests, down from 17 in 2016. Several were safe rural seats where the contest was not close (Cowper, Farrer, New England, Kennedy, Maranoa). Hobart based Clark was easily won by Independent Andrew Wilkie. The remaining nine seats had more intriguing contests.
Tony Abbott’s efforts to retain his seat of Warringah against Independent Zali Staggall attracted more attention than any contest at the election. In the end Abbott was defeated by a 12% decline in his first preference vote, polling only 39.0% to Steggall’s 43.5%. First preference support for both Labor and the Greens fell below 7%, a lot of strategic voting for Steggall going on to ensure Abbott’s defeat. Overall 78.6% of preferences flowed to Steggall, including around 86% of Labor and Green preferences.
Across Sydney Harbour in Wentworth, Independent Kerryn Phelps was trying to win re-election in the seat she had won after Malcolm Turnbull’s resignation. On first preferences Liberal Dave Sharma led 47.4% to 32.4% for Phelps. Receiving 80.8% of preferences, including around 85% of Labor and Green preferences, was not enough for Phelps to win given how close to 50% Sharma had started the count.
There were three Labor versus Green contests where Liberal preferences were distributed. In line with preference policy since 2013, the Liberal Party recommended preferences for Labor ahead of the Greens. Liberal preferences flowed 66.1% to Labor in Cooper (formerly Batman) and 60.7% in Grayndler. Liberal preferences flowed 51.5% to Labor in Wills, where the Liberal candidate had been disendorsed and presumably there were fewer Liberal how-to-votes.
Kooyong and Melbourne finished as Liberal versus Green contests. In both seats Labor recommended preferences to the Greens, and the flow was 83.4% in Kooyong and 80.4% in Melbourne. Labor had disendorsed its candidate for Melbourne during the campaign.
The two remaining non-traditional contests, Indi and Mayo, were won by second placed candidates as discussed below.
Wins from Behind on Preferences
Set out below are the basic details for the 12 seats won on preferences by second placed candidates. (10 Labor, one Independent, one Centre Alliance)
| First Preferences | Two-Candidate | Preference Flows | ||||||
|---|---|---|---|---|---|---|---|---|
| Electorate (State) | Gap | ALP | L/NP | Oth | ALP | L/NP | ALP | L/NP |
| Corangamite (VIC) | -6.9 | 35.5 | 42.3 | 22.2 | 51.1 | 48.9 | 70.3 | 29.7 |
| Cowan (WA) | -1.3 | 38.1 | 39.4 | 22.5 | 50.8 | 49.2 | 56.6 | 43.4 |
| Dunkley (VIC) | -1.4 | 38.5 | 39.9 | 21.6 | 52.7 | 47.3 | 65.8 | 34.2 |
| Griffith (QLD) | -10.0 | 31.0 | 41.0 | 28.1 | 52.9 | 47.1 | 78.0 | 22.0 |
| Lilley (QLD) | -5.1 | 35.6 | 40.8 | 23.6 | 50.6 | 49.4 | 63.6 | 36.4 |
| Macnamara (VIC) | -5.6 | 31.8 | 37.4 | 30.9 | 56.2 | 43.8 | 79.3 | 20.7 |
| Macquarie (NSW) | -6.6 | 38.3 | 44.9 | 16.9 | 50.2 | 49.8 | 70.6 | 29.4 |
| Moreton (QLD) | -5.7 | 35.2 | 40.8 | 24.0 | 51.9 | 48.1 | 69.7 | 30.3 |
| Perth (WA) | -3.0 | 34.4 | 37.4 | 28.2 | 54.9 | 45.1 | 72.8 | 27.2 |
| Richmond (NSW) | -5.2 | 31.7 | 36.9 | 31.4 | 54.1 | 45.9 | 71.2 | 28.8 |
| Gap | IND | LIB | OTH | IND | LIB | IND | LIB | |
| Indi (VIC) | -2.7 | 32.4 | 35.1 | 32.6 | 51.4 | 48.6 | 58.5 | 41.5 |
| CA | LIB | OTH | CA | LIB | CA | LIB | ||
| Mayo (SA) | -3.5 | 34.2 | 37.7 | 28.2 | 55.1 | 44.9 | 74.4 | 25.6 |
Had the 2019 election been conducted under first past the post voting, the 12 seats listed above would have been won by the Coalition, delivering 89 seats the the Morrison government. However, assuming voters would not have changed their vote under first past the post voting is an unrealistic assumption. The Morrison government would have won more seats under first past the post voting, but Labor and cross bench members would still have won some of the above seats.
One other seat could be added to the above list. In the NSW seat of Gilmore, Labor’s candidate led on first preferences, but with 36.2%, trailed by 5.5% the combined vote of the Liberal candidate (29.2%) and National candidate (12.5%). A disendorsed Liberal candidate contested as an Independent and polled 7%, and his preferences slightly favoured Labor 54% to 46%. Preferences from lower placed candidates delivered Labor victory without needing National preferences, which in the end flowed 83.7% to the Liberal party.
In Mayo, the Centre Alliance’s Rebekha Sharkie received 74.4% of other party preferences to defeat the Liberal Party’s Georgina Downer for the second time. This included 80% of Labor preferences and 77% of Green preferences.
The victory of Independent Helen Haines in Indi is even more remarkable given her seat was a three-cornered contest with both Liberal and National candidates. The Liberal candidate polled 35.1% and National 9.4%, meaning Haines trailed the combined Coalition vote by 12%. Haines received 74.6% of third party preferences, including 81.6% of Labor preferences and 83.9% of Green preferences. Haines needed only a tiny leakage of National preferences to win, but received a higher share at 19.0%.
One other preferential contest of note was the usually ultra-safe National seat of Mallee in Victoria. The retirement of the sitting National MP over an unedifying personal matter saw 13 candidates, including a Liberal and a number of Independents, contest the seat. One of the Independents, Jason Modica, finished fourth and did well enough on preferences to almost pass the third placed Labor candidate. After nine candidates had been excluded, Modica still trailed the Labor candidate by 248 votes. If Modica had passed Labor, there was an outside chance he could have won the seat. Instead, Modica’s preferences pushed the Labor candidate into second place 338 votes ahead of the Liberal, resulting in Liberal preferences electing new National candidate Anne Webster. Had Labor stayed in third place, Labor preferences could have helped deliver the seat to the Liberal Party.
Links to related posts on previous elections
Preference Flows at the 2016 Federal Election
Preference Flows at the 2013 Federal Election
Preference Flows at the 2010 Federal Election
All Party Preference Flows 2019
The table below includes totals of preferences for all minor parties towards Labor and the Coalition. The table can be sorted by any column.
| Party Name | Cands | Votes | % Vote | to L/NP | to ALP |
|---|---|---|---|---|---|
| The Greens | 151 | 1,482,923 | 10.40 | 17.8 | 82.2 |
| United Australia Party | 151 | 488,817 | 3.43 | 65.1 | 34.9 |
| Independent | 95 | 479,836 | 3.37 | 40.6 | 59.4 |
| Pauline Hanson's One Nation | 59 | 438,587 | 3.08 | 65.2 | 34.8 |
| Animal Justice Party | 46 | 116,675 | 0.82 | 38.4 | 61.6 |
| Christian Democratic Party | 42 | 97,513 | 0.68 | 74.4 | 25.6 |
| Fraser Anning's Conservative National Party | 48 | 77,203 | 0.54 | 71.8 | 28.2 |
| Katter's Australian Party | 7 | 69,736 | 0.49 | 67.0 | 33.0 |
| Centre Alliance | 3 | 46,931 | 0.33 | 32.9 | 67.1 |
| Shooters, Fishers and Farmers | 8 | 41,479 | 0.29 | 59.1 | 40.9 |
| Sustainable Australia | 19 | 35,618 | 0.25 | 46.0 | 54.0 |
| Liberal Democrats | 10 | 34,666 | 0.24 | 77.2 | 22.8 |
| Derryn Hinch's Justice Party | 8 | 26,803 | 0.19 | 46.2 | 53.8 |
| Western Australian Party | 15 | 25,298 | 0.18 | 49.0 | 51.0 |
| Australian Christians | 13 | 23,802 | 0.17 | 80.8 | 19.2 |
| DLP - Democratic Labour Party | 8 | 18,287 | 0.13 | 39.8 | 60.2 |
| Rise Up Australia Party | 14 | 14,032 | 0.10 | 60.4 | 39.6 |
| Science Party | 7 | 12,617 | 0.09 | 32.5 | 67.5 |
| Victorian Socialists | 3 | 12,453 | 0.09 | 12.4 | 87.6 |
| Reason Australia | 3 | 8,895 | 0.06 | 31.2 | 68.8 |
| Australian Progressives | 5 | 7,759 | 0.05 | 32.8 | 67.2 |
| Australia First Party | 4 | 6,786 | 0.05 | 56.4 | 43.6 |
| The Great Australian Party | 5 | 5,355 | 0.04 | 53.1 | 46.9 |
| Citizens Electoral Council | 2 | 3,267 | 0.02 | 26.4 | 73.6 |
| Socialist Equality Party | 4 | 2,866 | 0.02 | 36.9 | 63.1 |
| Socialist Alliance | 3 | 2,447 | 0.02 | 20.2 | 79.8 |
| Non Affiliated | 2 | 2,143 | 0.02 | 32.4 | 67.6 |
| Australian Better Families | 1 | 2,072 | 0.01 | 64.1 | 35.9 |
| Australian Democrats | 1 | 2,039 | 0.01 | 30.9 | 69.1 |
| Australian Workers Party | 1 | 1,676 | 0.01 | 58.7 | 41.3 |
| Love Australia or Leave | 1 | 1,564 | 0.01 | 54.5 | 45.5 |
| Child Protection Party | 1 | 1,219 | 0.01 | 45.4 | 54.6 |
| Non-Custodial Parents Party | 1 | 1,213 | 0.01 | 51.3 | 48.7 |
| Involuntary Medication Objectors (Vaccination/Fluoride) | 1 | 1,179 | 0.01 | 36.4 | 63.6 |
| VOTEFLUX.ORG | Upgrade Democracy! | 1 | 602 | 0.00 | 46.2 | 53.8 |
| All minor parties and independents | 743 | 3,594,358 | 25.22 | 40.4 | 59.6 |