2022 Post-Federal Election Pendulum

With 16 members elected to the crossbench in the new House of Representatives, drawing up a new electoral pendulum based on the 2022 Federal election result strains the traditionally used two-sided format.

However, I’ve gone with the traditional format with the non-major party seats separated bottom right on the opposition side of the pendulum. However, the expanded size of the crossbench means this group of seats deserves more attention than its bottom of the table position suggests.

Inside this post I provide a post-election pendulum for the House of Representatives, along with some general comments on the overall result.

General Comments on the Results

Before moving on to the new electoral pendulum, I thought it worthwhile making some general comments.

  • The nearly complete two-party preferred count appears to be Labor 52.1% Coalition 47.9%, a swing of 3.6%. This is roughly the swing that Labor needed to win based on the pre-election pendulum. The two-party result hides the importance of both the Greens and Independents to the election outcome, the success of both being responsible for the Coalition doing much worse in seat terms (58) than you would expect from a 47.9% two-party preferred vote.
  • There were 84 seats that recorded Labor two-party preferred majorities against 67 with Coalition two-party preferred majorities.
  • On first preferences the Coalition polled 35.7%, Labor 32.6%, at 68.3% the lowest combined vote for the major parties since two-party politics became the norm in 1910.
  • There were 27 seats that did not finish as traditional two-party contests, 20 that recorded Coalition majorities on the alternative two-party preferred count and seven for Labor.
  • Only 15 of the 151 electorates were won by candidates recording a majority on first preference votes, eight Labor, four Liberal, two National and one LNP. 136 electorates required some or all preferences to be distributed before the winning candidate achieved a majority of the vote. It is by far the highest number of seats where preference distributions were required.
  • Labor won 77 seats, the Coalition 58, Independents 10, Greens four, one Katter’s Australian Party and one Centre Alliance.
  • There were 16 seats won by trailing candidates. Seven were won by Labor (Bennelong, Boothby, Gilmore, Higgins, Lyons, Robertson, Tangney), seven by Independents (Curtin, Fowler, Goldstein, Kooyong, Mackellar, North Sydney, Wentworth), and two by the Greens (Brisbane from 3rd place and Ryan).
  • If winners were declared based on first preference simple majority leaders, the Coalition would have won 74 seats and Labor 70 with two Greens, three Independents and one each for Katter’s Australian Party and the Centre Alliance. Of course, had the election been fought under simple majority rules, it is likely that many voters for third parties might have voted differently knowing they did not have the option of preferences.
  • The results after preferences again highlight how much the accrued political advantage of preferential voting has shifted from the Coalition to Labor. Preferential voting was introduced in 1917 to allow electoral co-operation between competing non-Labor parties. Until 1990 it worked consistently in favour of the non-Labor parties. But the 2022 result is the latest to highlight that since 1990 preferential voting has worked overwhelmingly in favour of Labor.
  • Queensland was the only state to record a Coalition two-party preferred majority with 54.0%. The state also saw the largest swing to Labor, 4.4%, but Labor lost one seat to the Greens (Griffith) along with two LNP seats that recorded Labor two-party majorities in 2022 (Brisbane, Ryan). The swing to Labor was 4.9% in South-east Queensland and 3.3% in the rest of the state.
  • Tasmania was the only state to record a swing to the Coalition, a 1.6% swing though no seats changed hands and Labor’s recorded a two-party majority with 54.3%.
  • The swings in other states, noting that counting continues in some seats, were 3.2% to Labor in NSW, 1.5% in Victoria, a huge 10.5% in Western Australia, 3.3% in South Australia, 5.3% in the ACT and 1.3% in the Northern Territory.
  • In Victoria, the state swing hides a wide variety of results. Labor recorded large swings in its favour in key seats it needed to retain or gain, but suffered significant swings against it in a string of its own safe seats in northern and western Melbourne.
  • The Liberal party suffered a massive rejection in Western Australia, losing Swan, Hasluck, Pearce and Tangney to Labor on double digit swings, as well as Curtin to Independent Kate Chaney. Of the 15 seats that recorded the largest swing to Labor, 12 were in Western Australia. The result in Western Australia was the difference between minority and majority government for Labor Party.
  • Support for the Greens rose substantially from 10.5% to 12.3%, the party gaining an extra 300,000 votes in the lower house. This is a vastly greater number than reported by Peter Hartcher in the Sydney Morning Herald on 28 May. In the same article Hartcher stated that Labor had lost 600,000 votes, but the final results show Labor had roughly an unchanged number of votes, though in percentage terms Labor’s vote fell 0.8%.
  • Support for Pauline Hanson’s One Nation rose from 3.1% to 5.0%, but largely due to a near trebling of the number of seats the party contested. In the seats One Nation did not contest in 2019 the party polled 4.0%. In the seats it contested at both elections, the party’s support fell from 8.0% to 6.4%. After Labor, the Coalition and the Greens, One Nation remains the nation’s fourth largest party.

The seats changing party (with margin) were –

  • Labor gains from Coalition (10) – Chisholm VIC (0.5), Boothby SA (1.4), Higgins VIC (2.6), Reid NSW (3.2), Swan WA (3.2), Robertson NSW (4.2), Pearce WA (5.2), Hasluck WA (5.9), Bennelong NSW (6.9), Tangney WA (9.5)
  • Green gains from Coalition (2) – Brisbane QLD (4.9), Ryan QLD (6.0)
  • Green gain from Labor (1) – Griffith QLD (2.9)
  • Independent gain from Labor (1) – Fowler NSW (14.0)
  • Independent gain from Coalition (6) – Wentworth NSW (1.3), Kooyong VIC (6.4), Goldstein VIC (7.8), North Sydney NSW (9.3), Mackellar NSW (13.2), Curtin WA (13.9) (WA)
  • The Coalition lost 18 seats, by state six in NSW, four in Victoria, two in Queensland, five in Western Australia and one in South Australia.

New Electoral Pendulum

The pendulum below sets out the new seats in margin order, the most marginal seats at the top of each list, the safest at the bottom. Labor (Government) seats are shown in the left hand column, Coalition seats at the top of the right-hand column. Independents, Greens and other parties are listed at the bottom of the right hand column, again in ascending margin order.

In 124 electorates the margins are shown in traditional two-party format, either Labor versus Coalition or Coalition versus Labor. The other 27 electorates are shown with a margin for the winning party and the opposing party shown in brackets. (e.g Grayndler is shown with (v GRN) after the name and state indicating a Labor versus Greens margin.) Seats in bold changed party status at the election.

Based on the margins, a swing of around 1% would deprive the Albanese government of its majority.

However, it is almost impossible to calculate a swing needed for the Coalition to return to office. The Coalition needs 18 seats, which assuming all crossbench members are re-elected, means a swing of 6.3%. A gain of 10 seats from Labor on a uniform swing of 3.3% would give the Coalition more seats than Labor though still well short of a majority.

But there are nine crossbench seats on margins under 5%. This includes the six seats the Coaliyion lost to Independents, as well as two seats the Coalition lost to the Greens. Every seat the Coalition wins back from the crossbench is one fewer seats the Coalition needs to win from Labor.

If you follow this link, I’ve also prepared a simpler pdf version that you can print on two-sides of an A4 page.

2022 Post-Federal Election Pendulum
Labor Seats (77) Coalition Seats (58)
Margin Electorate Margin Electorate
ALP 0.2 Gilmore NSW LIB 0.2 Deakin VIC
ALP 0.9 Lyons TAS LIB 0.5 Sturt SA
ALP 0.9 Lingiari NT LIB 0.7 Moore WA
ALP 1.0 Bennelong NSW LIB 0.7 Menzies VIC
ALP 2.1 Higgins VIC LIB 1.4 Bass TAS
ALP 2.3 Robertson NSW LIB 1.5 Casey VIC
ALP 2.4 Tangney WA LIB 1.7 Dickson QLD
ALP 3.3 Boothby SA NAT 2.3 Cowper NSW (v IND)
ALP 3.3 McEwen VIC LIB 2.8 Aston VIC
ALP 3.3 Paterson NSW LIB 2.9 Monash VIC
ALP 4.0 Hunter NSW LIB 3.1 Longman QLD
ALP 4.6 Parramatta NSW LIB 3.2 Banks NSW
ALP 5.2 Reid NSW LIB 3.4 Bonner QLD
ALP 5.2 Blair QLD LIB 3.4 Leichhardt QLD
ALP 5.8 Shortland NSW LIB 3.6 Canning WA
ALP 5.8 Werriwa NSW NAT 3.8 Nicholls VIC (v IND)
ALP 6.0 Hasluck WA NAT 3.8 Flynn QLD
ALP 6.3 Dunkley VIC LIB 3.9 Wannon VIC (v IND)
ALP 6.4 Chisholm VIC LIB 4.2 Forde QLD
ALP 6.5 Dobell NSW LIB 4.2 Bradfield NSW (v IND)
ALP 6.6 Bruce VIC LIB 4.3 Durack WA
ALP 6.9 Isaacs VIC LIB 4.3 Forrest WA
ALP 7.1 Holt VIC LIB 4.4 Petrie QLD
ALP 7.6 Corangamite VIC LIB 5.5 Bowman QLD
ALP 7.6 Hawke VIC LIB 6.3 Lindsay NSW
ALP 7.8 Macquarie NSW NAT 6.6 Capricornia QLD
ALP 7.8 Richmond NSW LIB 6.7 Flinders VIC
ALP 8.2 Eden-Monaro NSW LIB 6.9 Groom QLD (v IND)
ALP 8.5 Macarthur NSW LIB 7.0 O’Connor WA
ALP 8.6 Wills VIC (v GRN) LIB 7.0 Hughes NSW
ALP 8.7 Cooper VIC (v GRN) LIB 7.7 Hume NSW
ALP 8.8 Swan WA LIB 8.0 Braddon TAS
ALP 8.9 Hindmarsh SA LIB 8.7 Fisher QLD
ALP 9.0 Pearce WA LIB 8.7 La Trobe VIC
ALP 9.1 Rankin QLD LIB 9.0 Fairfax QLD
ALP 9.1 Moreton QLD LIB 9.3 McPherson QLD
ALP 9.4 Solomon NT NAT 9.7 Calare NSW (v IND)
ALP 9.5 McMahon NSW LIB 9.8 Berowra NSW
ALP 10.0 Gorton VIC LIB 10.1 Grey SA
ALP 10.1 Whitlam NSW NAT 10.1 Hinkler QLD
ALP 10.5 Lilley QLD NAT 10.4 Dawson QLD
ALP 10.8 Makin SA LIB 10.6 Fadden QLD
ALP 10.8 Cowan WA LIB 10.7 Mitchell NSW
ALP 11.5 Greenway NSW NAT 10.7 Page NSW
ALP 11.5 Gellibrand VIC LIB 10.9 Wright QLD
ALP 11.6 Oxley QLD LIB 11.2 Moncrieff QLD
ALP 11.9 Adelaide SA NAT 11.3 Wide Bay QLD
ALP 12.1 Bendigo VIC LIB 11.8 Herbert QLD
ALP 12.2 Canberra ACT (v GRN) LIB 12.4 Cook NSW
ALP 12.2 Macnamara VIC NAT 13.8 Lyne NSW
ALP 12.3 Jagajaga VIC NAT 14.8 Riverina NSW
ALP 12.4 Calwell VIC LIB 16.4 Farrer NSW
ALP 12.4 Maribyrnong VIC NAT 16.5 New England NSW
ALP 12.8 Lalor VIC LIB 16.6 Barker SA
ALP 12.8 Corio VIC NAT 17.8 Parkes NSW
ALP 12.9 Spence SA NAT 19.0 Mallee VIC
ALP 12.9 Bean ACT NAT 20.6 Gippsland VIC
ALP 13.0 Ballarat VIC NAT 22.1 Maranoa QLD
ALP 13.5 Chifley NSW
ALP 13.7 Franklin TAS Others (16)
ALP 14.3 Hotham VIC Margin 0.0 Electorate
ALP 14.5 Kingsford Smith NSW IND 1.3 Curtin WA (v LIB)
ALP 14.7 Cunningham NSW IND 1.6 Fowler NSW (v ALP)
ALP 14.8 Perth WA IND 2.5 Mackellar NSW (v LIB)
ALP 14.9 Blaxland NSW GRN 2.6 Ryan QLD (v LIB)
ALP 15.1 Watson NSW IND 2.9 Goldstein VIC (v LIB)
ALP 15.2 Burt WA IND 2.9 North Sydney NSW (v LIB)
ALP 15.5 Barton NSW IND 2.9 Kooyong VIC (v LIB)
ALP 15.6 Scullin VIC GRN 3.7 Brisbane QLD (v LIB)
ALP 15.7 Fenner ACT IND 4.2 Wentworth NSW (v LIB)
ALP 16.4 Kingston SA IND 9.1 Indi VIC (v LIB)
ALP 16.5 Fraser VIC GRN 10.2 Melbourne VIC (v ALP)
ALP 16.7 Brand WA GRN 10.5 Griffith QLD (v LIB)
ALP 16.7 Sydney NSW (v GRN) IND 11.0 Warringah NSW (v LIB)
ALP 16.9 Fremantle WA CA 12.3 Mayo SA (v LIB)
ALP 17.1 Grayndler NSW (v GRN) KAP 13.1 Kennedy QLD (v NAT)
ALP 18.0 Newcastle NSW IND 20.8 Clark TAS (v ALP)

28 thoughts on “2022 Post-Federal Election Pendulum”

  1. Thank you. You do an amazing job. Factual unbiased reporting. I feel There needs to be more focus on education about our voting system particularity at the high school year eleven and twelve level.

    1. Antony for parliament to be truly representative of the populace it should be like a random selection of the demos…but it isn’t?

      Parliament seems to be dominated by particular vocations and as such a particular mindset dominates, is this good?

      What I’m trying to get at is…would it be a positive if all candidates had to list beside their name their profession before putting themselves forward as representatives?

      COMMENT: Candidates provide an occupation on their nomination forms and it is available where candidate are listed on the ABC website. If you select and electorate and click the down arrow next the candidate name, occupation is one of the fields listed. It is available in some of the download files. However, someone describing themselves as a “businessman” or “analyst” may not be very helpful.

  2. Important work presented with great clarity. Thank you.
    At some stage, on completion of counting, it would be interesting to see an analysis of non-participation across the 151 seats. This may tell us a lot about the success of our democracy at this time in history.

    COMMENT: You can find the turnout figures by division at this link. The overall turnout is slightly down on 2019.

    The lowest turnouts are all in electorates with large indigenous populations. The government has already stated it intends to give the AEC more resources to deal with the low rates of enrolment and turnout amongst first Australians.

  3. Great work Antony,
    From the numbers of seats won by Labor (77) and in which Labor had a 2PP majority over the coalition (84) there are 7 seats which Labor held a 2PP majority but didn’t win. 4 are obviously those won by the Greens, and I’m guessing Mayo, Clark and Fowler are the other 3 – is that correct? (CORRECT)

  4. I thought that in labor-held seats like Macnamara and Richmond they were against the greens and the margin was a lot smaller than listed here. Is it something to do with the ordering of candidates? Or am I just dumb?

    COMMENT: The margin shown is for the final two candidates after the full distribution of preferences. There are several seats where the winner was determined by the much closer race between the final three candidates. The gap at that point will not be available until the full distribution of preferences is published. Seats in this category include Brisbane, Macnamara and Richmond.

  5. Thanks Antony. A follow up to your observations on preferential voting: How deep did the count have to go? Was the result mostly clear after second preferences were distributed? Or were there seats where third and fourth preferences actually mattered?

    COMMENT: The question is how many candidates had to be excluded before a continuing candidate reached 50%. I suspect a lot more seats than in the past but it is impossible to calculate until the preference distributions are published. Even then, it is a laborious task to add up the details.

  6. Many thanks Antony. This is a great analysis. I am particularly interested in your comments on preferential voting over time. Given this analysis and the ongoing difficulties it will give the Coalition in winning in the future, is their next gambit an argument as to why Australia needs a first past the post system?

    COMMENT: Voting using preferences is so widely entrenched in Australian that abandoning it all together is not an option, especially given preferences are still used in the Senate and in the states. If the Coalition wanted to change the system then optional preferential voting is more likely, used in NSW since 1980, in Queensland 1992-2015 and proposed in South Australia by the Liberal Party two years ago.

  7. Antony, Very interesting analysis.
    You state the coalition recorded 35.7% on first preferences. Much has been made of the fact that this is higher than the ALP’s first preference.
    What were the FP votes separately for the Liberal Party, National Party, LNP in Queensland and CLP in NT?
    Is the 35.7% simply the sum of all those votes vs the total number cast nationally?
    Is this even a reasonable and valid question to ask?

    COMMENT: The Coalition percentage is simply the sum of the four parties that make up the Coalition. In 147 of the 151 divisions there was only one Coalition candidate and there seems no good reason to complicate analysis of alternative government by artificially splitting up support for parties who when in majority have formed government together for the last century.

  8. Very interesting but of course it doesn’t show the true picture with 3 cornered contests. The Greens lost Mcnamarra by 300 off votes but because they came third the ALP hold it by 12%

  9. Hi Antony,
    Just wondering if you know around when the AEC normally publishes the distribution of preferences tables for house seats? Is it a couple of months after the election?

    COMMENT: They will be published but I am not sure when.

  10. I think I would prefer it if, instead of always refrrring to the ‘Coaliyion’ you recorded the results for the Libs and Nats as separate parties as? From my perspective? It was only the zlibs who suffered such staggering losses and we all know that was directly about Morrison. Even the rabid support of the Murdoch media, and to a lesser extent the Costello-led Nine mob couldn’t counter the toxicity of Morrison and many of his team, Teal wins notwithstanding.

    COMMENT: The seats held by the two parties are separately identified in this post. My one mention of the Coalition’s percentage vote was while highlighting the decline in major party vote and noting the record level of support for third parties and Independents. I think you’ve rather missed the point of the reference.

  11. Thanks for the great data driven analysis. It’s so refreshing compared to the agenda driven commentary from much of the modern day media.

    1. The Mackerras Pendulum worked when there were only two party groupings in the House. This election has deviated significantly from the assumptions on which the Mackerras version was based. If the proportion of independent/non-aligned members increase, it may become time to rethink the pendulum.

      COMMENT: I think he might have been referring to the old layout of pendulums rather than my inverted version designed for the web.

  12. Thanks Antony. Very professional and puts the election in perspective. I wish that I had the energy to do first preference votes excluding seats where ALP or LNP voters obviously tactically voted. In many of these seats ALP candidates obviously ran “low key” and were merely there to ensure extreme rusted on voters’ preferences arrived at the desired destination?

  13. Thanks Antony. So if the Coalition get a uniform swing of 4.0%, they get a majority. They win 11 seats from the ALP, 5 from the Independents and 2 from the Greens, giving them the 18 seats they need. So in terms of the pendulum, that is the required swing. Of course, ‘uniform swings’ never happen, but 4% is the magic number.

    COMMENT: No. You can’t mix and match between two-party and two-candidate preferred swings. For any given national two-party preferred swing, there is no guarantee that the swings in the non-2PP seats will be anything like the 2PP swing.

  14. “All politics is local” A phrase (with variations) used since 1932 in the USA. I live in Greenway and work in Lindsay. Greenway showed a 2PP 8.73% swing to the ALP, Lindsay a 2PP 1.30% swing to the LP. Admittedly Lindsay is becoming more gentrified, but in my mind the late nomination of candidates by the losing side in each case (due to the internecine squabbling that is such a part of NSW politics) played a large role in the outcome.

    COMMENT: Not just about politics being local. Also about candidates and intensity of effort. The Liberal Party was very late in announcing a candidate in Greenway and clearly never took the seat seriously. The Liberals held Lindsay and put mret effort into defending it than they did in trying to win Macquarie and Greenway. Labor caused some controversy in its choice of Andrew Charlton as candidate for Parramatta and had a much weaker result than in neighbouring Reid and Greenway. And Labor had a shocker in Fowler for obvious candidate reasons.

  15. The assumption of uniform swing is already quite shaky even only looking at 2PP (as this election shows), and I’m not sure extending it to ‘non-classic’ 2CP as Hugh did is so problematic.

  16. Great work Antony. Can you tell me where I can find the preference flow figures for Dickson? Thank you.

    COMMENT: They are yet to be published.

  17. CENSUS.
    I think the figures from the Census may mean some significant changes to federal electorates and their boundaries.

    Section 24 of the constitution says the number of members of the House of Representatives “shall be, as nearly as practicable, twice the number of the senators”. Currently 76 and 151.
    Tasmania is guaranteed 5 members of the House of Representatives under the constitution and Northern Territory is guaranteed 2 members under the Electoral Act.

    Aside from Tasmania and NT, the remaining number of members is determined by population (although once a the number for each state is determined, the size within the state is determined by voters, not population).

    COMMENT: I’ve removed your calculations as they are not correct. Seats are not allocated based on the Census. The numbers used are the ABS’s quarterly population statistics, the most recent being published on 28 June. You can find the report at this link. The report includes an explanation of how the Census has been used to re-base the population. The current estimates are that Victoria will lose one seat, WA will gain one and NSW could lose one.

  18. How soon Will the AEC publish a colour-coded State-by-State map showing the new Members by Party as it it gives some general indication of changed votes & demographic changes?

    COMMENT: I have no idea. The AEC might know though.

    1. Hey Antony, you usually publish something called the Rise of the Greens on certain seats with high greens votes, but is their any certain threshold of first preference green votes that is what makes it a “rise of the greens” seat? And another questing relating to the greens, in wa there is one seat with the rise of the greens and that is freo while they have never really got anywhere there but they have really boosted their vote in Perth, do you think Fremantle can still be won/boosted vote-wise by the greens?

      COMMENT: See my response to your previous comment. The problem for the Greens with Fremantle is the seat is more outer suburban than the state seat of Fremantle where the Greens finished second in 2021. The Greens might match Labor in the portside core of Fremantle, but not in the outer suburbs to the south east.

      At the moment the best the Greens can hope for in Fremantle is to pass the Liberal Party and reach second place, the party’s current position in Grayndler, Sydney, Cooper and Wills. In each of these seats, the Green candidate passing the Liberal to reach second isn’t enough to win given the current Liberal policy of recommending preferences to Labor ahead of the Greens.

  19. Hey Antony, in the seats where the greens achieve a high amount of the vote, you usually publish something and that seats Page called “The rise of the Green” What is the threshold of the vote for the greens that makes you I guess classify that seat as a greens target.

    COMMENT: The Greens usually define their own targets. At the 2022 election my opinion was that Brisbane, Griffith and Macnamara were the seats where the Greens had the best chance of winning by getting ahead of Labor. Ryan was also possible, but victory relied on more than just outpolling Labor as there also needed to be a significant first preference and two-party against the LNP for them to lose to one of Labor or the Greens. The Greens also named Higgins and Richmond, but I thought both were seats where it would harder for the Greens to get past Labor.

  20. Hi Antony. I think I recall you quoted a figure a couple of weeks ago that Monique Ryan’s vote share split 80% to Labor in the notional 2PP count for Kooyong. I was just wondering where that information can be found? I am just interested in finding out the notional 2PP preference flows to Labor in the QLD seats won by the Greens. Thanks a lot.

    COMMENT: I hope to publish the data next week after I get my NSW election guide published. The data is not generally available though not secret. In Brisbane the Green to Labor preference flow was 88.4%, in Griffith 87.3% and Ryan 88.2%. In each case the flow was 4-5% higher than the reverse Labor to Greens preference flows, which were 84.5%, 82.1% and 83.5% respectively.

Leave a Reply