May 3 miscellany: RedBridge poll, teal seat polls and the path ahead (open thread)

As the starter’s gun is fired, one pollster finds Labor with its nose in front, while another suggests a tough battle for the two Melbourne teals.

The ball is now officially rolling on a campaign for a May 3 election, with nominations to be declared in two weeks and a not-quite-two-week early voting period starting on Tuesday, April 22. The latter reflects the interruption of Easter, extending from April 18 to 21, with Anzac Day presenting a further interruption on April 25. Some attractions to make a visit to Poll Bludger part of your daily routine during the campaign period:

• I may regret saying this, but I will at least aspire to publish posts on a daily basis, which will either break news of major polls as soon as they report, or appear overnight to summarise developments of the previous day and the contents of the morning newspapers.

• The Poll Bludger election guide, which I’ve spent much of the past week bringing up to speed, offers an overview page reviewing the electoral terrain and summarising the seats to watch, and detailed interactive guides to all 150 lower house seats and eight state and territory Senate contests,

• The BludgerTrack poll aggregate tells you as much in one glance about the state of federal polling as any sensible person needs to know, and uniquely offers regularly updated trend measures at state level.

• Come the big day and the weeks to follow, the site will offer live results reporting that will wipe the floor with all comers, as an examination of its Western Australian state equivalent should readily attest. Innovative features include a model for calculating win probabilities three ways in complex contests and a new colour-coded results map feature along the lines of that recently added for the Western Australian results.

Show-don’t-tell time:

• A new poll from RedBridge Group, conducted March 13 to 24 from a sample of 2039, has Labor leading 51-49 on two-party preferred, unchanged from the previous poll of March 3 to 11. The primary votes are Labor 34% (up two), Coalition 38% (up one), Greens 11% (down one) and others 17% (down two). The Coalition continues to do better among those who profess themselves “solid” in their choice, leading 54-46 among that group. Twenty-nine per cent felt themselves able to name something the Albanese government had done that had made their lives better (electricity rebates being most cited), compared with 54% who couldn’t; 23% reported themselves more or less in favour of tariffs, with 35% more or less opposed; and 68% registered concern about Chinese naval vessels off the coast, with 24% less or not at all concerned. The addition of the voting intention numbers to BludgerTrack has lifted its reading of the Labor primary vote by a grand total of 0.1%, and left all other indicators unchanged.

• Thursday’s Herald Sun had JWS Research polling from the two teal-held seats in Melbourne, conducted “a fortnight ago” from samples of 800 each for Australian Energy Producers. The results have Zoe Daniel trailing Liberal challenger Tim Wilson 54-46 in Goldstein, compared with her 52.9-47.1 win in 2022, with Wilson on 44% of the primary vote (40.4% in 2022) to Daniel’s 24% (34.5%), Labor on 21% (a somewhat counter-intuitive improvement on their 11.0% in 2022) and the Greens on 5% (7.8% in 2022). Monique Ryan is credited with a 51-49 lead over Liberal candidate Amelia Hamer in Kooyong, likewise compared with a 52.9-47.1 win for Ryan in 2022, from primary votes of 40% for Hamer (42.7% for Josh Frydenberg at the 2022 election) and 32% for Ryan (40.3% in 2022), with Labor on 11% (6.9%) and the Greens on 9% (6.3%). Peter Dutton was credited was leads over Anthony Albanese as preferred prime minister of 51-35 in Goldstein and 40-36 in Kooyong.

• The Australian National University’s Centre for Social Policy Research has results from the second wave of a series tracking changes in attitudes ahead of the election, the first having been conducted from October 14 to 25 from a sample of 3622, this latest having run to January 29 to February 12 with a sample of 3514, 2380 of whom also participated in the first wave. On this occasion there is no straightforward reading of voting intention, but it finds confidence in the federal government fell from 52.9% in its honeymoon period to 33.7% in the latest survey. Satisfaction with democracy “remains relatively stable at 66.2%”, and there has been a general decrease in sentiments associated with populism from August 2018 to January 2025, such as a drop from 67.9% to 61.3% in the share of respondents who felt “government is pretty much run by a few big interests looking out for themselves”. Nonetheless, “life satisfaction” has been on a downward trend since pre-pandemic, which has resumed since 2023 after recovering from a slump while it was on, and the percentage reporting financial difficulty stabilised in the low thirties in 2024 after a steady ascent from late 2020.

• The Tasmanian government announced earlier this week that the periodic Legislative Council elections, which this year encompass the seats of Nelson and Pembroke, would be postponed from their usual date in the first week of May to May 24, to exclude the possibility of a clash with the federal election.

Federal polls: Morgan, DemosAU Victorian results, and more (open thread)

Another strong result for Labor from Roy Morgan, another finding of 51-49 to federal Labor in Victoria, plus seat polling for Brisbane and various teal targets.

Notwithstanding yesterday’s suggestion of a pre-budget lull, poll news continues to come in at an exhausting clip:

• The weekly Roy Morgan poll maintains the series’ recently acquired form of remarkably favourable polling for Labor, who lead 53-47 on respondent-allocated preferences (in from last week’s 54.5-45.5 blowout) and 54-46 on previous election preference flows (in from 54.5-45.5). The primary votes are Labor 33.5% (up one), Coalition 35.5% (up one-and-a-half), Greens 12.5% (down one) and One Nation 4% (down one). The poll was conducted Monday to Sunday from a sample of 1683.

• DemosAU has a poll on both federal and state voting election from Victoria, the latter of which I’ll cover when I do a post on a Resolve Strategic poll result I’m expecting next week. In common with both the Newspoll and Freshwater Strategy breakdowns, the federal result shows Labor leading 51-49, in from 54.8-45.2 at the 2022 election. The primary votes are Labor 29%, Coalition 34%, Greens 15% and One Nation 8%, compared with respective results of 32.9%, 33.1%, 13.7% and 3.8% in 2022. The poll was conducted last Monday to Friday from a sample of 1006.

• The Courier-Mail reports a uComms poll of the seat of Brisbane by Liberals Against Nuclear has Trevor Evans of the LNP on 31.2%, Greens incumbent Stephen Bates on 24.2% and Labor candidate Madonna Jarrett on 23.2%, with some of the remainder likely allocated as undecided. The poll was conducted last Thursday from a sample of 1184. Another privately conducted poll of the seat published last week had the LNP at 36.4%, Labor on 29.5% and the Greens on 18.1%. Neither is encouraging for the LNP, with the Greens competitive on the first set of numbers but not the second.

• An article by Mike Seccombe in The Saturday Paper sketchily relates encouraging polling for teal independents by Climate 200, putting Nicolette Boele ahead of the Liberals 52-48 in the Sydney seat of Bradfield; Caz Heise 53-47 ahead over Nationals incumbent Pat Conaghan in the Mid North Coast New South Wales seat of Cowper, where Conaghan won 52.3-47.7 on Heise’s first attempt in 2022; and the Liberals leading only by 51-49 in both the South West Western Australian seat of Forrest and the Mornington Peninsula seat of Flinders, which are respectively being contested by Sue Chapman and Ben Smith. We are further told that the polls have the Nationals on 39% of the primary vote in Lyne, and the LNP on 43% and 42% in the seats of Fisher on the Sunshine Coast and McPherson on the Gold Coast. Respective sample sizes of 980, 867 and 1047 are provided for the Cowper, Lyne and Bradfield polls, with the latter conducted on February 3.

Canadian election called for April 28

Mark Carney calls the Canadian federal election just before parliament was due to resume, with the Liberals narrowly ahead of the Conservatives in the polls.

Guest post by Adrian Beaumont, who joins us from time to time to provide commentary on elections internationally. Adrian is a paid election analyst for The Conversation. His work for The Conversation can be found here.

On March 9, Mark Carney, former governor of the Bank of Canada and Bank of England, was elected Liberal leader with 86% of points and 87% of the popular vote. On March 14, Carney replaced Justin Trudeau as Canadian PM. On Sunday (Canadian time), Carney called the Canadian federal election for April 28, about six months early.

Parliament had been due to resume on Monday after it was prorogued for the Liberal leadership election. The governing centre-left Liberals won 160 of the then 338 seats at the September 2021 election, ten short of a majority, and have been reduced to 152 through by-election losses and defections. Carney is not an MP, so he could not address parliament (he will contest Nepean at the election). Perhaps owing to these difficulties, Carney called the election early.

There will be 343 seats elected by first past the post at this election, up from 338 in 2021, so 172 seats will be needed for a majority. The CBC Poll Tracker was updated Sunday, and it gives the Liberals 37.5%, the Conservatives 37.1%, the left-wing New Democratic Party (NDP) 11.6%, the separatist left-wing Quebec Bloc (BQ) 6.4% (28.4% in Quebec), the Greens 3.8% and the far-right People’s 2.2%. Seat predictions are 174 Liberals, just over a majority, 134 Conservatives, 26 BQ, seven NDP and two Greens.

In early January, just before Trudeau announced he would resign once a new Liberal leader had been elected, vote shares in the Tracker were 44% Conservative, 20% Liberal and 19% NDP. At this point, the Conservatives looked headed for a massive landslide with well over 200 seats, while the Liberals could have fallen into third behind the BQ.

Donald Trump is probably most responsible for the Liberal revival, with his tariffs and his talk of making Canada the 51st US state pushing Canadians back to supporting the Liberals. Trump is expected to impose more tariffs on April 2, possibly assisting the Liberals further. I believe Trump’s tariffs and associated stock market falls have also helped Labor in Australia.

However, I don’t believe in momentum in elections: just because one party is gaining ground in the polls doesn’t mean that party will continue to gain ground. The massive surge for the Liberals could reverse during the election campaign, perhaps as voters refocus on stuff they don’t like about the Liberals after nearly ten years of Liberal government since Trudeau was first elected in October 2015.

US and Portugal

In Nate Silver’s aggregate of US national polls, Trump has a net approval of -2.1 (49.6% disapprove, 47.5% approve). His net approval turned negative on March 11. Trump has fallen from +12 net approval at the start of his term.

A Portuguese parliamentary election will be held on May 18, only 14 months after the March 2024 election. The early election came after the conservative AD, which governed in minority with support from the far-right Chega, lost a confidence vote. Polls indicate another AD-led minority government is likely. Portugal uses proportional representation by region to elect its 230 MPs.

Federal poll aggregates: Newspoll and Freshwater Strategy (open thread)

Two pollsters break down their recent federal polling by state, and in Newspoll’s case much else besides.

We can expect a lull in polling this week ahead of the budget, followed by an avalanche next week in its wake – which, according to Simon Benson of The Australian, is “expected” to follow upon an announcement by the Prime Minister next weekend of an election for either May 3 or May 10. In the absence of new federal voting intention results, two pollsters have produced detailed breakdowns of their recent polling aggregated from the start of the year, one being the especially extensive quarterly Newspoll aggregation in The Australian, the other encompassing results for the four largest states from Freshwater Strategy in the Financial Review.

Newspoll has two-party results of 50-50 from New South Wales, unchanged from the previous quarterly aggregate, suggesting a Coalition swing of around 1.5% compared with the 2022 election result; 51-49 to Labor in Victoria, compared with 50-50 last quarter, a swing of around 4%; 57-43 to the Coalition in Queensland, a substantial move from the 53-47 result last quarter, a Coalition swing of around 3%; 54-46 to Labor in Western Australia, unchanged on last quarter for a swing of around 1%; and 50-50 in South Australia, after Labor led 53-47 last quarter, suggesting a Coalition swing of 4%. This combines results of three polls from a national sample of 3757, ranging from 271 for South Australia to 1149 for New South Wales. Also featured is an array of demographic detail that you can access by sifting through the BludgerTrack poll data feature, together with reams of further polling detail since the 2022 election.

Only primary votes were provided from Freshwater Strategy, but the results imply a Coalition lead of around 52-48 in New South Wales, a Labor lead of around 51-49 in Victoria, a Coalition lead of 54-46 in Queensland and a Labor lead of 56-44 in Western Australia. This combines three monthly poll results from a national sample of 3152, ranging from 318 for Western Australia to 947 for New South Wales.

Federal polls: YouGov and RedBridge Group (open thread)

Two new federal polls offer further indications that Labor is in a stronger position than it was at the start of the year.

The now weekly YouGov poll records a 50-50 result after two weeks with Labor in their nose in front, from primary votes of Labor 31% (steady), Coalition 37% (up one), Greens 13% (down half) and One Nation 7% (down half). Anthony Albanese is down two on approval to 43% and up one on disapproval to 50%, while Peter Dutton is steady on 42% and down one to 47%, with Albanese’s lead as preferred prime minister is in from 45-39 to 45-40. The poll was conducted Friday to Wednesday from a sample of 1500.

There is also a new federal poll from RedBridge Group showing Labor leading 51-49, after the Coalition led 51.5-48.5 at the poll from early last month. Labor is up one on the primary vote to 32% and the Coalition is down three to 37%, with the Greens up one to 12%. The gap favouring the Coalition on firmness of voting intention has narrowed, 61% of Coalition voters professing themselves solid, down four, compared with a steady 51% for Labor voters.

The RedBridge poll also finds 51% holding that the country is “generally headed in the wrong direction” compared with 29% for the right direction. Fifty-three per cent agreed with the statement “the Albanese government’s renewable energy policies and timelines are pushing the costs of energy through the roof”, with only 23% disagreeing, and 38% agreed that “if Australia were to produce nuclear energy it would be cheaper for consumers like me than renewable energy”, with 28% disagreeing. A question on whether Australia should be “more assertive“ or “do more to build a positive relationship” with China produced an even result of 39% and 38% respectively. The poll was conducted March 3 to 11 from a sample of 2007.

DemosAU has published federal voting intention numbers from the poll it conducted during the last week of the state election campaign, which proved highly accurate with respect to the result (Labor 43% compared with a result of 41.5%, Liberal 30% compared with 28.2%, Nationals 5% compared 5.3%, Greens 11% compared with 10.8%). The federal numbers are Labor 36% (36.8% at the 2022 election), Coalition 38% (34.8%), Greens 11% (12.5%), One Nation 6% (4.0%), with Labor leading 52-48 on two-party preferred (55.0-45.0). The poll was conducted March 4 and 5 from a sample of 1126.

Federal polls: Essential, Morgan, Greens seat polling (open thread)

Encouraging new polls for Anthony Albanese and Labor, plus a more objective account of Greens seat polling than you’ll get from News Corp.

Federal polling continues to come thick and fast, with a general pattern of improving results for Labor and Anthony Albanese:

• The Guardian reports the fortnightly Essential Research poll has the major parties unchanged on the primary vote, with Labor on 29% and the Coalition on 35%, and the Greens down one to 12%, with the remainder including an undecided component of 6%. The pollster’s 2PP+ measure has a 47-47 tie (the balance being undecided), after the Coalition led 48-47 last time. The stability on voting intention is in contrast to much improved personal ratings for Anthony Albanese, who records his first net positive result in this series since October 2023, the month of the Indigenous Voice referendum. Albanese is up five on approval to 46% (The Guardian report says up four, but the previous published result was 41%) and down four on disapproval to 45%. Peter Dutton is steady on 41% approval and up two on disapproval to 46%. The poll also finds 31% supporting and 39% opposing Peter Dutton’s proposal to reduce work-from-home arrangements for public servants (with women particularly opposed), which he has since dialled back. Albanese’s suggestion that Australia might send peacekeeping troops to Ukraine was supported by 33% and opposed by 40%. The poll had a sample of 2256 – double the usual size – and was conducted from Wednesday to Sunday. The full report is available here.

• Roy Morgan, which is notably more volatile than other Australian pollsters, has turned up an eyebrow-raiser with its regular weekly federal voting intention result, putting Labor 54.5-45.5 ahead on respondent-allocated preferences, out from 51.5-48.5 last week. Labor is up two-and-a-half on the primary vote to 32.5%, the Coalition is down three to 34%, and the Greens and One Nation are steady on 13.% and 5%. The result is also 54.5-45.5 on previous election preference flows, which usually favours Labor more than the other method, out from 52-48 last time. The poll was conducted Monday to Sunday from a larger than usual sample of 2097.

• The ABC has polling from Talbot Mills on attitudes to Donald Trump, finding approval down from 41% since February (well on the high side of such polling in Australia) to 37% and disapproval up from 49% to 51%, including a six-point increase for “strongly disapprove”. Twenty-two per cent of Australians somehow approve of steel and aluminium tariffs on Australian exports, with 65% disapproving. The survey was conducted between March 6 and 12, so before the tariffs were actually confirmed, from a sample of 1051.

Yesterday’s News Corp papers reported on polling of six seats of interest to the Greens, conducted for right-wing activist group Advance by Insightfully, whose principal is Leanne White, formerly of Crosby Textor. Reported under headlines promising a “federal election wipeout” for the party, it in fact suggests the party will gain a second seat in Victoria and retain one or two of their three seats in Brisbane. To deal first with the latter, for which the greater detail is provided:

• In Griffith, the poll has the LNP on 38.6% (30.7% at the 2022 election), the Greens on 31.3% (34.6%), Labor on 22.6% (28.9%), independents on 1.5% and others on 5.9%. No two-candidate preferred is provided, but a conservative estimate based on 2022 election flows would give the Greens a winning margin of between 2% and 3%, compared with 10.5% at the 2022 election.

• In Ryan, the primary votes are LNP 39.6% (38.5% in 2022), Greens 27.4% (30.2%), Labor 21.9% (22.3%), independent 7.1% and others 3.9%. A two-candidate result of 51.6-48.3 in favour of the LNP is presumably based on respondent-allocated preferences, with flows from the 2022 election suggesting a lineball result. The Greens’ winning margin over the LNP in 2022 was 2.6%.

• The results suggests Labor would gain Brisbane with 29.5% of the primary vote (27.3% in 2022) to the LNP’s 36.4% (37.3%), since they would receive most of the third-placed Greens’ 18.1% (27.2%) as preferences, with independents on 9.7% and others on 6.3%. A two-party result of 53-47 result in favour of the LNP (the basis of the “Brisbane to turn blue” table headline) appears to presume the Greens would come second, when the primary votes clearly suggest otherwise, and seems excessive in its LNP preference flow besides. The Greens won by 3.7% over the LNP in 2022, and Labor won the two-party preferred count over the LNP by 4.4%.

Sketchier detail is provided for the three Victorian seats canvassed:

• In Melbourne, Adam Bandt is on 50.1% (49.6% at the 2022 election, which reduces to 44.9% on my own determination of the redistribution), Labor is on 19.2% (25.0% and 25.6%) and the Liberals are on (I think) 21.6% (15.2% and 19.4%).

• In Wills, the only detail provided for the primary vote is that the Greens are up 4.8% on what seems to be the 2022 result without adjustment for the redistribution, suggesting 33.1%. Labor is said to have a 53.7-46.3 lead after preferences (54.2-45.8 based on my own post-redistribution determination), which is hard to assess in the absence of the other primary votes.

• In Macnamara, the Liberal are on 37.6% (29.0% on the 2022 election result, which will do because of the modest impact of the redistribution on this seat), the Greens are on 27.9% (29.6%) and Labor is on 25.9% (31.8%). No two-candidate preferred result is provided, but the Greens would certainly close that gap on any normal accounting of Labor preferences, which may hold true even if Labor put the Liberals ahead of them on their how-to-vote cards, as counselled by News Corp’s James Campbell.

The polls targeted “about 600 voters” in each seat, suggesting a margin of error of around 4% (effectively higher if the results were heavily weighted, as was presumably the case), with field work dates not disclosed.

Freshwater Strategy: 51-49 to Coalition (open thread)

Another increment of evidence of an improving picture for federal Labor.

The Financial Review reports the latest Freshwater Strategy poll has the Coalition’s lead in from 52-48 to 51-49. Labor’s primary vote is unchanged at 31%, but the Coalition is down two to 39% and the Greens are up one to 14%. The exact figures are not yet provided, but we are told that Anthony Albanese’s net approval rating is up one to minus 10% while Dutton’s is down four to minus 12%. (UPDATE: Albanese is up two on approval to 37% and up one on disapproval to 47%; Dutton is down one to 35% and up two to 46%.) Albanese’s lead as preferred prime minister is out from 45-43 to 46-42.

Forty-two per cent now expect Labor to win the next election, up five (though only 7% expect a majority Labor government), with the Coalition down five to 45% (18% for a majority). Field work dates and sample size don’t appear to be featured in the online reports (UPDATE: 1051 and Thursday to Saturday, apparently), but the gaps will presumably be filled when the print edition is available. Polls in this series have typically been spaced four weeks apart, but this time it’s only been three.

Western Australian election: late counting, week two

A new post to cover late counting in Western Australia, including a general review of the situation.

Click here for full display of Western Australian state election results.

Wednesday, March 26

The Pilbara preference distribution has now been published, which is particularly of interest because no running two-candidate count was conducted, leaving us flying blind as to how the final result would look based on primary votes of Labor 36.1%, Liberal 24.0% and Nationals 16.4%. My preference estimate had Labor winning by about 0.8%, and the ABC’s slightly higher. With the conclusion of the count on Thursday, we were told only that Labor had won. Now we know how much by: 207 votes, or 0.6%.

Monday, March 24

The West Australian reports a recount following the preference distribution in Kalamunda confirmed a Liberal win by 82 votes, one vote less than the initial count (although the numbers in the media feed, and hence my results page, don’t reflect this). So the lower house numbers are now confirmed at Labor 46, Liberal seven and Nationals six. Geraldton is the first seat for which a preference distribution has been published, and it shows independent Shane van Styn dropped out at the last exclusion with 7126 votes to defeated Labor member Lara Dalton’s 7775 and Nationals victor Kirrilee Warr’s 11,339. The WAEC relates the upper house will not be concluded until the second week of April, pleading “a high number of voters marking more than one square above the line, which has resulted in a more complex count”.

Saturday, March 22

Labor’s win in Kalgoorlie by a 1.6% margin was confirmed today, the Nationals evidently having failed to have closed the gap with the Liberals on preferences and potentially winning the seat on a stronger flow of preferences than the Liberal candidate could manage. The preference count in Kalamunda will be conducted tomorrow, and will confirm a narrow Liberal victory unless the process turns up an error. Two useful posts from Antony Green, one on the upper house count, in which he raises doubts about Labor’s presumed sixteenth seat based on how the totals might look when below-the-line votes have added, and another comparing results from 2017 and 2025.

Thursday, March 20

The distribution of preferences was conducted yesterday in Albany, confirming the Nationals’ win the the Liberals’ failure to reach the final count. Albany and the other known known of Churchlands are among the seats identified by the WAEC as “Awaiting Declaration” rather than “Count in Progress” – if these means the preference distributions have been conducted (UPDATE: Confirmed), this presumably also means confirmation of a Nationals win in Warren-Blackwood and a Labor win in Pilbara. The other seats so identified are Bateman, Collie-Preston, Fremantle, Geraldton, Jandakot, Murray-Wellington and Riverton.

Elsewhere, the Liberal lead in Kalamunda fell from 83 to 64 as one presumably last batch of absents broke 50-31 to Labor. That increases by about 20 when factoring in the result at Wooroloo Primary School, which for some reason is lacking a TCP count. That leaves declaration votes, of which there are likely to be about 300. Comparison with other metropolitan seats that have counted their provisionals suggests a Labor gain of about 30 would be par for the course, but a 270-151 split in Riverton offers one hopeful precedent for Labor.

Wednesday, March 19

The West Australian reports the preference distribution was conducted today in Churchlands, though the result hasn’t been published yet, confirming Basil Zempilas the winner with a margin of 0.7%, or 376 votes. It also reports Albany and Warren-Blackwood will follow, both to be won by the Nationals from Labor if the Liberals remain in third place, which they will unless benefit from very unusual preference flows. Nothing new today from Kalamunda, where Labor can vaguely hope a few hundred outstanding provisional votes can overturn an 83-vote Liberal lead.

Tuesday, March 18

Follows of my live results have a new toy to play with in the shape of a colour-coded map, with light shadings for seats still in doubt and dark ones for those that have been called. This may not look all that impressive with most of the results on, but on future election nights it will present the spectacle of starting out white and then filling out as votes come in and seats are called.

We’re starting to see Provisional Votes entering the count, and clearly these include election day enrolment votes (which weren’t a thing at previous elections), because they’ve gone from being a handful of votes per seat to several hundred. They have strongly favoured Labor over Liberal in the three metropolitan seats where they have been added, though not in Geraldton and the Labor-versus-independent count in Fremantle. They have not been counted yet in Kalamunda, where a tiny handful of votes today reduced the Liberal lead from 88 to 83. My system now says Liberal ahead because I’ve turned off the projection, which was insisting on putting Labor’s nose in front for reasons that could probably use looking into.

Provisionals contributed to a further reduction in Basil Zempilas’s lead in Churchlands, breaking 244-188 to Labor on top of a 369-305 break to Labor on absents. His lead has gone from 852 to 375 over two days, but as the count is now at 90.3% of the enrolment, slightly exceeding the final turnout in 2021, that may be the end of the matter. Updates in Pilbara continue to not tell us what we need to know, which is the Labor-versus-Liberal two-party preferred.

The Liberals have been fading over the past few days in the upper house count, suggesting they may not get to 11 seats. My revised assessment of the situation is Labor 16, Liberal ten, Greens four, Nationals two and One Nation, Legalise Cannabis and Australian Christians one each, with the last two seats likely a game of musical chairs between the eleventh Liberal, the second One Nation and the first Animal Justice.

Monday, March 17

Results would be getting close to final in a lot of cases now, with the deadline for postal votes having been on Thursday, although a lack of provisional votes suggests we’re not there yet (UPDATE: On reflection, the fact that votes have only been counted for 75% of enrolled voters suggest that very substantial numbers of votes are yet to be counted). The Liberal lead in Kalamunda went from 98 to 88 today, with postals favouring Labor 436-418 and absents favouring Liberal 397-389. My projected Labor lead naturally assumes there are still some votes left out there, but as noted that may not be the case. There were only five provisional votes in 2021, but it may be different this time if the category encompasses those who took advantage of the introduction of election day enrolment.

Labor did well on yesterday’s counting in Churchlands, winning the absent votes 1721-1497 and a presumably final batch of postals 463-328, but this only gets as far as reducing Basil Zempilas’s 852-vote lead to 493. A batch of postals in Fremantle broke 313-295 to independent Kate Hulett, reducing Labor member Simone McGurk’s lead to from 491 to 473. Even without the correction of an apparent error that I believe to be penalising them by 140 votes in South Perth, Labor now looks home and hosed there after splits of 801-611 on absents and 313-255 on postals pushed the lead from 315 to 563 – certainly the ABC is now calling it for Labor.

Postals in Kalgoorlie broke 105-83 to Labor, increasing their lead over the Liberals from 394 to 416, although the outside chance of the Nationals doing better on preferences than I’m presuming and sneaking past first the Liberals and then Labor is something we won’t know about until the full preference distribution. The distribution continues to hold the secret of the result in Pilbara, where the counting of 1208 absent votes did not fundamentally change the situation.

Saturday, March 15

A new thread for what remains of the Western Australian state election count, the earlier one being in danger of falling off the bottom of the landing page. For those who have just joined us, what I estimate to be a 12.4% swing off the superlative 2021 result (the ABC only gets to 11.9%, based on what are probably more careful preference estimates than my own) has yielded the Liberals remarkably little in the way of seats: only Carine, Nedlands and Murray-Wellington are being called as Liberal gains by my highly conservative system, though the ABC is undoubtedly on safe ground in adding Churchlands.

In the regions, Labor has lost Geraldton to the Nationals and has conceded defeat in Albany (my own system would seem to be crediting Labor with too many preferences in the latter case). It’s unclear whether the winner will be the Nationals, as my system considers more likely, or the Liberals, whom the ABC favours. It’s a similar story in Warren-Blackwood by the ABC’s reckoning, but I’m now calling it for the Nationals after adjusting the parameters that were allowing for the possibility of the Liberals making the final count. The Liberals won’t win Kalgoorlie unless something unusual happens on absent votes, but it’s mathematically possible that the Nationals will make the final count in their stead and sneak home on a stronger flow of preferences than the Liberals are receiving.

There are a number of seats the ABC’s system is calling for Labor which mine isn’t, where I see no reason to doubt the ABC. Labor has survived an early scare at the hands of an independent in Fremantle, although there are suggestions of a legal challenge. The ABC isn’t calling South Perth for Labor, but I believe there’s an error in the WAEC’s numbers that will tip it over when it’s corrected. All told, I make out three serious points at issue: Kalamunda, which is going down to the wire between Liberal and Labor; Pilbara, where we must await the final preference count to see if Nationals preferences flow tightly enough to the Liberals to get them over the line; and the aforementioned Liberals-versus-Nationals race in Albany. If the Liberals fail in all three, they will suffer the ignominy of failing to recover official opposition status from the Nationals.

The Greens will hold the balance of power in their own right in the reformed 37-seat upper house, adding four seats to what looks like 16 for Labor. The remainder will go Liberal 11, Nationals two and one each for One Nation, Legalise Cannabis and Australian Christians, with one still up in the air. Animal Justice are on 0.43 quotas and One Nation’s second candidate will be on 0.35 after the election of the first, with the latter presumably to do well out of preferences from the Libertarians, Stop Pedophiles! and Shooters Fishers and Farmers (the latter of whom flopped after committing the rookie error of identifying themselves on the ballot paper as “SFFPWA”). An unknown quantity is “Independent” on 0.48 quotas – if most of this went to the ticket headed by former Legalise Cannabis member Sophie Moermond, rather than the ungrouped candidates, she too could be in contention. (UPDATE: It’s been pointed out to me that the ungrouped candidates haven’t been counted yet because these are below-the-line votes, so she would seem a real chance).

For those of you who haven’t just joined us, an update on Saturday’s counting. There was, at last, substantial progress, though most of it was in seats that aren’t in doubt. One exception was Murray-Wellington, which both the ABC and myself are calling for the Liberals after 5498 absents broke 2776-2719 in their favour, putting their candidate 837 ahead. My projection of Labor’s lead in Pilbara was cut from 1.4% to 0.8% due to what I think must have been a correction in an error from the first batch of postals, though it’s hard to disentangle because new postals were added to the count as well. In any case, Labor had an implausible 54.6% of the postals as of Friday, but has a far more plausible 32.3% now. Clearly we’re going to have to wait for the preference distribution here — my estimate has Labor ahead by 0.8% and the ABC’s has it at 0.5%, but estimates is all they are. The WAEC has decided against conducting a Labor-versus-Liberal preference throw that would clarify the matter, seemingly due to the potential for a Nationals win that my recalibrated model is now ruling out entirely. Labor continues to claw back on absents in Warren-Blackwood, today’s batch breaking 470-398, but my narrower error margins mean my system is now calling it for the Nationals.

Page 2 of 576
1 2 3 … 576