The Monday after Super Saturday

An all-new forum for discussion of Super Saturday and its aftermath.

For those wanting a more psephologically focused forum for discussion of the by-elections than the main thread, which is the Newspoll post directly below this one, I offer the following.

As well as that, some scattered notes and observations:

• Hear Ben Raue and I trade thoughts on the results in a podcast at The Tally Room.

• For those of you still following the count, Braddon and Longman (and presumably others) yesterday saw counting of postals and special hospital booth votes, together with rechecking. In Braddon, 3967 postals and 1224 followed the overall pattern in not swinging at all, leaving the Labor lead at 52.5-47.5 from a favourable swing of 0.3%, which is unlikely to change much from here. In Longman, 7775 postals and 943 special hospital votes swung somewhat more heavily to Labor (4.9% and 7.1%) than the election day result (3.4%). Since the LNP nonetheless won the postals 52.3-47.7, the raw lead has come down from 5.4% to 4.5%, but I’m now projecting a final margin of 4.3% rather than 3.4%.

• It’s not news anymore, but I thought it worth noting that the Daily Telegraph had a report on insiders’ expectations for Braddon and Longman on July 21 that proved unusually prescient, but which escaped my notice at the time – other such commentary having generally been unduly bullish from the conservatives’ perspective. According to the report, “a senior Liberal strategist said the polling was much tighter in Braddon and that while the LNP was still competitive in Longman, Labor would have to be considered in the box seat”. Also quoted as a “senior Labor source” who said Labor had “made considerable ground in the past couple of weeks”, and that the party was now feeling “pretty confident”.

• If by-election booth results in a spreadsheet-style format are of any use to you, I am maintaining them online for my own purposes for Longman and Braddon.

The Sunday after Super Saturday

A good night for Bill Shorten as Labor lands a surprisingly emphatic win in Longman, and does enough to get home in Braddon.

While Labor’s by-election performances were nothing special in historical terms, it was undeniably a good night for the party, thanks largely to an unexpectedly clear win in Longman. Five campaign opinion polls had Labor slightly behind in the seat, before the election eve Newspoll found them edging to a 51-49 lead. Labor actually appears headed for a winning margin of around 4%, bolstering a fragile 0.8% margin with a swing of 3.4%. The big surprise was the near double-digit fall in the Liberal National Party primary vote, which leaves them struggling to crack 30%. This is well below the 34% attributed to them by Newspoll, to say nothing of a series of ReachTEL results that had them approaching 40%.

The LNP slump rendered redundant what everyone imagined would be the decisive factor, namely the flow of One Nation preferences. Despite this, One Nation were the other big winner in Longman, adding around 7% to their 9.4% vote from 2016. This indeed flowed a lot more strongly to the LNP than in 2016, reflecting the party’s how-to-vote card recommendation and the fact that they clearly picked up much of the LNP’s lost support. After receiving 56.5% of One Nation preferences in 2016, Labor looks to have scored only a third this time.

The Braddon result was less good for Labor, notwithstanding that they have clearly won, and that this looked in doubt throughout the campaign. The main change from the 2016 result is that independent Craig Garland scored a creditable 11.0% (although it may come down a little in late counting), chipping a few percent away from each of Labor, Liberal and the Greens. Rebekha Sharkie’s win in Mayo was of about the anticipated scale: her present lead over Georgina Downer after preferences is 8.6%, compared with her 5.0% margin in 2016. Sharkie’s primary vote performance was even more robust, up from 34.9% to around 45%. This bespeaks one poor aspect of the by-elections for Labor – after playing dead at two successive elections, its vote in Mayo has fallen all the way to 6.0%.

In the two WA seats, Josh Wilson did notably better in Fremantle than Patrick Gorman did in Perth, although neither was in the least bit troubled. Wilson gained 11.6% to gain a clear majority on the primary vote, with the Greens treading water at 17% and the Liberal Democrats garnering enough stray Liberals to land in the low teens. Despite the 42.3% Liberal vote from 2016 being up for grabs (compared with 36.9% in Fremantle), Labor only made a negligible gain on the primary vote in Perth, with the Greens also only up slightly. The rest spread among a large field of 15 candidates, with independent Paul Collins the strongest performer among claimants to the Liberal vote. Turnout was notably subdued in Perth and Fremantle, and looks likely to settle at around 70%.

If you click on the image below, you will find an accounting of the swings in Braddon and Longman and, in the former case, an projection of the final result. Since the swing on votes counted in Braddon thus far is exactly zero, it concludes Labor’s existing margin of 2.2% will be maintained. Also featured are regional breakdowns for Braddon and Longman, with the former broken into the larger towns (Burnie, Devonport and Ulverstone) and the remainder, and the latter into Bribie Island area and the remainder. This doesn’t turn up anything particularly interesting: especially in Longman, the swings were remarkably uniform. Craig Garland’s vote was a little lower in the larger towns, but there was otherwise little distinction to speak of in Braddon.

Super Saturday live

Live coverage of the counting for the Super Saturday by-elections (especially Braddon and Longman).

9.31pm. So much for the last entry. Patrick Gorman’s projected two-party vote is now 61.6%.

9.18pm. It’s only based on 1000 votes from Leederville and Northbridge, and suburbs further out might tell a different story, but the first two-party results in Perth suggest Patrick Gorman is headed for a fairly modest winning over the Greens. Conversely, Josh Wilson is outpolling the Greens two to one in Fremantle.

9.00pm. The Devonport PPVC result is also nothing special for the Liberals, so the door has closed on them in Braddon now.

8.55pm. First figures from Perth don’t bear out ReachTEL’s indication of a strong Liberal Democrats result, with more of the homeless Liberal vote going to independent Paul Collins.

8.53pm. Labor think they have Braddon in the bag, and the Burnie PPVC result that has just come through might explain why: the Liberals have copped an 8.7% hit on the primary vote there. Only Devonport and Queenstown PPVCs, plus postals, yet to come.

8.20pm. Labor still tracking for a winning margin approaching 5% in Longman.

8.13pm. The first PPVC in from Braddon is Ulvertstone, which has only 513 votes. It’s indeed a bit better for the Liberals than polling day votes, with their primary vote down by 2.1%, compared with 6.6% overall – although that’s still less than what they need.

8.10pm. So there has been no swing at all out of the polling day booths in Braddon, which have produced around 47,000 votes. That means the Liberals need to conjure a swing of around 7.5% out of probably 10,000 pre-polls and 7000 postals. Which doesn’t seem terribly likely.

8.02pm. Only the PPVCs to come from Braddon now. Polls have closed in Perth and Fremantle.

7.52pm. The projected Labor margin in Braddon is down to 2.0%. Two more polling day booths to report in Braddon, plus the four pre-poll voting centres, which could yet disgorge many thousands of votes. However, it seems to be agreed that it’s time to pull the plug on the LNP in Longman, and Georgina Downer is conceding defeat in Mayo.

7.40pm. A bit of dispute on the ABC as to whether the LNP is getting 70% or 60% of One Nation preferences in Longman. They will need for it to be very high indeed, and for something highly dramatic to happen in the pre-poll voting centres.

7.38pm. A very quick count in Braddon, where the momentum against Labor is levelling off. The AEC now projects a 2.6% Labor margin, but the potential for a different dynamic from the pre-poll voting centres means a measure of caution is still advised.

7.25pm. I’ve now got primary vote swings of 8.6% against Labor and 6.7% against Liberal in Braddon, with Craig Garland now barely into double figures. Labor’s projected two-party is continuing its slow decline, now at 2.2%. Given the trend as larger booths come in, and the unknown factor of the large pre-poll voting centres that will come in later in the evening, I wouldn’t be ready to call this.

7.24pm. We’ve now gone from around 7000 to around 14,000 primary votes counted in Longman, and the earlier trend is continuing: Labor holding steady, Liberal down around 14%, One Nation on around 15%. So looking like a surprisingly solid Labor win. No surprises in Mayo.

7.18pm. The situation is still a little elusive in Braddon: two-party projections point to a Labor win, but the primary count is well ahead of it, and the Labor primary vote there is beginning to sag. Bit of a lull in Longman counting, but the early indications are extremely strong for Labor.

7.12pm. Picture beginning to change in Braddon as larger centres do less well for Craig Garland. The ABC now has the Labor swing at 2.7%, which is down from over 4% earlier. The AEC computer is calling it for Labor, but given that trend, this should be treated with caution at the moment.

7.10pm. Count progressing a little quick than I’d figure in Longman, and here too Labor are doing quite a bit better than expected, holding their own on the primary vote while One Nation gouges double figures out of the LNP.

7.02pm. There are some reasonably serious primary vote numbers in now from Braddon, and that earlier picture is still holding: Craig Garland is on 16.3%, and Liberal are down more than Labor. The first two-party results suggest this is converting into a two-party swing to Labor.

6.50pm. Still only 1591 votes counted, but the early dynamic in Braddon is that Craig Garland is doing very strongly, coming in at the high teens, and he’s gouging the Liberals twice as heavily as Labor. If this kept up, Labor would win pretty handily on preferences from 35% of the primary vote.

6.40pm. The first two small booths in from Braddon are Moorleah (183 votes), where Craig Garland has a fairly spectacular 26.8%, and Waratah (139 votes), where he has a rather more modest 10.8%. For the time being though, he’s gouging double figures out of both major parties’ primary votes.

6pm. Polls have closed in Braddon and Longman, and will do so in half an hour in Mayo, and two hours in Perth and Adelaide. Results for small booths in Braddon should start coming in very shortly, but it will have to wait an hour or so for anything meaningful from Longman. For my own benefit more than anything, I have mocked up summarised booth results for Braddon and Longman which will, when there’s actual data to plug into them, will show booth-level primary vote totals, percentages and swings for the Labor, Liberal(-National), Greens, One Nation and Craig Garland (both the latter two are identified, wrongly, as “IND”). Other than that, the AEC publishes its own perfectly good booth-matched projected results (though not nearly enough besides), which can naturally be found at the official results page.

Newspoll: 51-49 to Labor in Longman and Braddon

The biggest and presumably best polls for today’s by-elections are better for Labor than some, but still suggest Braddon and Longman could go either way.

As Super Saturday dawns, The Australian brings us perhaps the two most heavy-duty seat polls yet published in Australia, with Newspoll surveying 1015 respondents in Longman and 1002 in Braddon between Tuesday and Thursday. In both cases Labor records a lead of 51-49, putting them at the high end for Labor’s polling record throughout the campaign. Labor is credited with a solid improvement on the primary vote compared with the 2016 election result, from 35.4% to 40%, while the Liberal National Party is down from 39.0% to 36%. This gives Labor the buffer it needs to deal with One Nation, up from 9.4% to 14%, and their stronger flow of what I presume are respondent-allocated preferences to the LNP. The Greens are on 5%, compared with 4.4% in 2016. In Braddon, the primary votes are Labor 40% (unchanged on 2016), Liberal 43% (up from 41.5%), the Greens 5% (down from 6.7%) and independent Craig Garland 8%.

As did the YouGov Galaxy polls for the News Corp tabloids last week, the poll asked respondents how they would vote if Anthony Albanese rather than Bill Shorten were leading the Labor Party, which in both cases would supposedly have lifted Labor to leads of 55-45. Candidate approval questions found 37% satisfied with Labor’s Susan Lamb in Longman and 47% dissatisfied, while Justine Keay in Braddon was on 42% and 41%. Trevor Ruthenberg’s military medal issues apparently haven’t registered, with 61% saying they would not influence their vote, 22% saying it made them less likely to vote LNP, and 16% of people who were presumably going to vote LNP anyway implausibly indicating they would be more likely to vote LNP.

Super Saturday minus one day

New polls show the Greens performing underwhelmingly in the Western Australian seats; and some historic context for potential Labor defeats in Longman and/or Braddon.

Legalise Vaping Australia has published results of ReachTEL polling from the neglected seats of Perth and Fremantle, which turn up no surprises so far as their finding that both seats will be easily retained by Labor is concerned. However, the results are notably weak for the Greens, who record 18.6% in Perth and 17.0% in Fremantle after exclusion of the undecided (8.7% and 6.1% respectively). In particular, the Greens are placed slightly behind the Liberal Democrats in Fremantle, who clearly stand to harvest votes from homeless Liberal supporters.

On to more important matters, namely the Longman and Braddon by-elections, and some historical context for them. Much is being made of the fact that the only time a government party has gained a seat from the opposition at a by-election was in the fairly unusual circumstance of Kalgoorlie in 1920, after Labor member Hugh Mahon was expelled from parliament for seditiously describing the British Empire as a “bloody and accursed despotism”, apropos its recent conduct in Ireland. However, by-elections in highly marginal opposition-held seats have been a far from common occurrence, particularly in recent times. The last time there was a by-election in an opposition-held seat with a margin of 5% or less was in November 1983, early in the life of the Hawke government, when the Liberals retained the Brisbane seat of Moreton with a slight swing in their favour.

The histogram below shows the spread of swings for or against the government at the 67 by-elections that produced a Labor-versus-Coalition two-party result since 1949 (sourced from this Parliamentary Library paper). Thirteen recorded swings to the government, of which eight occurred in opposition-held seats, but in no case were they sufficient to gain them the seat. Ten of the pro-government swings were bigger than Labor’s 0.8% margin in Longman, and five were bigger than the 2.2% margin in Braddon.

The mean result, indicated by the peak of the distribution curve, is a 4.0% swing against the government. However, the dispersal is great enough to suggest 0.8% pro-government swings occur 20.7% of the time, and 2.2% swings occur 14.5% of the time. On this basis, the chance of one or the other being lost by an opposition that faced both simultaneously would be 32.2% (assuming the two results were independent of each other). This is no doubt inflated by two outlier results: Australian Capital Territory in 1970 (13.8%) and McPherson in 1981 (16.2%). With these removed, the likelihood of a 0.8% swing to the government is 13.5%; the likelihood of a 2.2% swing is 8.2%; and one or the other is 20.6%.

Super Saturday minus three days

Liberals claim to be encouraged by internal polling in Braddon, as the travails of minor party and independent candidates dominate the headlines.

Developments, and further reading:

Matthew Denholm of The Australian reports a Liberal tracking poll sample of 500 voters has Brett Whiteley edging to a 51-49 lead over Labor’s Justine Keay in Braddon. Party sources credit this to their targeting of independent Craig Garland, who is said to be down from 9% last week to 5% this week. Garland is a crusty local fisherman who polled 3.1% at the state election on a campaign against fish farming, but Michael Koziol of Fairfax reports he is “building a brand beyond that niche, connecting with disaffected working class voters and even some Liberals”. With Garland directing preferences to Labor, the Liberals fear a repeat of the 2016 election, when the Recreational Fishers Party polled 5.7% in Braddon and 4.9% in Bass, respectively producing preference flows of 63.1% and 67.0% to Labor. The Liberals have been duly keen to portray Garland as an “extreme green”, and to exploit an assault conviction recorded against him in 1993 over an altercation involving a group of his friends and two off-duty police officers. This extended to the Prime Minister telling reporters that “violence against women, violence against police, can never be accepted, must always be condemned”.

• Pauline Hanson has taken an oddly timed break from politics to go on a cruise off Ireland and Scotland, citing exhaustion. Meanwhile, the party’s candidate in Longman, Matthew Stephen, has faced a series of claims of unpaid debts and wages from subcontactors and workers for his tiling business. However, Renee Viellaris of the Courier-Mail reports One Nation’s campaign has been more professional than recent form: its how-to-vote cards, thought previously to have been “too wordy and hard to understand”, have been made visually punchier and clearer in their recommended preference order (in this putting the LNP ahead of Labor), and the party promises an organised polling booth presence and tighter scrutineering.

• Acres of psephological goodness on offer in Kevin Bonham’s review of the Longman by-election.