Queensland election: late counting

Click here for full display of Queensland state election results.

Friday evening

Another seat has come on to the radar (I am consistently indebted here to comments thread contributors, as I really only have about half-an-hour or so a day to devote to analysing the results at present) in the shape of Aspley, which has joined neighbouring Pine Rivers in trending in Labor’s favour in late counting. This is in fact part of a broader trend, evident to those who have been following the statewide two-party preferred estimate on my results entry page, where the LNP’s two-party lead exceeded 54-46 after the pre-poll voting centres had finished reporting on Sunday, but has since wound back to 53.5-46.5. This is mostly down to absent election day votes, which have recorded fairly modest swings for no obvious reason I can think of.

The situation is Aspley is that the late reporting votes have collectively been only slightly harmful to Labor, whereas projections based on election day and early voting results on Saturday and Sunday presumed they would be more substantially so. Based on how preferences seemed to be flowing when the ECQ was conducting a two-candidate preferred count, I only get to an LNP winning margin of 62 votes. However, I presume only late-arriving postals remain, and this will assuredly favour the LNP. Labor’s hope is that the preferences of late reporting votes were more favourable than election day and early votes.

That said, my expectation that only postal votes remain keeps getting confounded in South Brisbane, the one count I’m monitoring closely: today’s counting saw the addition of 210 absent early votes, 203 absent election day votes, 292 in-person declaration votes and only 109 postals. These were collectively unhelpful for the Greens, who need the LNP to finish ahead of Labor: 270 of the votes counted today were for Labor and 231 LNP, together with 251 for the Greens and 37 for One Nation.

Thursday evening

Today’s counting from South Brisbane saw the LNP catch up 95 votes on Labor through absent early votes, and fall back 19 on in person declaration votes and one on postals. I’m not sure if there’s any more to come in the way of absents, but I believe there will be around 900 postals which should narrow a gap that currently sits at 696 to around 600. The LNP would then need their share of One Nation preferences to be around 50% higher than Labor’s out of a three-way split inclusive of the Greens. As noted below, the difference in 2020 in this seat was 44%. For those of you who have just joined us, the issue here is that Labor wins the seat if they make the final count ahead of the LNP, and the Greens do if they don’t. We presumably won’t know the answer until the full distribution of preferences, which will presumably be late next week.

Wednesday evening

What I had previously rated the slim prospect of Labor getting over the line in Pine Rivers might yet come to pass – indeed, my own results system is calling it for them, but this is based on very rosy assumptions for Labor about how late postals will behave. The ABC also projects a Labor lead, in their case of 50.4-49.6. The ECQ’s way of doing things is not at all conducive to projecting results late in the count – it stops conducting its notional two-preference count and, still more confoundingly, records its initial and check count results for the primary vote separately. I have just switched over from the former to the latter in my results display, which is important in the case of Pine Rivers since a number of corrections in the check count were substantially to Labor’s advantage. However, it also means a lot of results, particularly of absent votes, that were in my system before are not there now, and will not return to it until the check count catches up with them.

Another late turn in counting is that Mirani now looks like going to the LNP rather than One Nation-turned-Katter’s Australian Party member Stephen Andrew, who has done distinctly badly on absent and to a lesser extent postal votes.

The Greens’ chances in South Brisbane took a hit yesterday with the addition of further absent election day results, which my assessment from yesterday had not accounted for. These were evidently from a Labor-friendly area, with 321 (35.9%) of the newly added batch being cast for for Labor and 241 (27.0%) for the LNP, which accordingly widens the gap the LNP will need to close on postals if Labor is to drop out and the Greens are to win the seat.

Tuesday evening

Belated recognition of what’s been evident to comments thread denizens, that the Greens’ defeat in South Brisbane is not indeed an accomplished fact owing to the outside chance that Labor will not make the final count. With four candidates in the count, One Nation are a very distant last on 984 votes, and Labor leads the LNP 9730 to 9043. The two variables in the equation are the extent to which that gap will narrow on last postals and in person declaration votes, and whether One Nation preferences will flow to the LNP strongly enough to close it altogether.

My highball estimate of the number of outstanding postals is 1300, which will put the lead at about 10120 to 9580 if they behave the way postals have to this point, with One Nation on about 1030. When One Nation was excluded in the seat in 2020, 62% of their votes went to the LNP, 18% to Labor and the balance to the Greens. Applying that to the above will close the gap between Labor and the LNP to barely more than 100.

In person declaration votes are less likely to be helpful to the Greens – of 688 formal votes in 2020, 215 were for Labor (31.25%) and only 107 for the LNP (15.55%), compared with respective overall vote shares of 34.42% and 22.84%. The situation is nonetheless worth monitoring, and will be the focus of whatever further updates I provide over the coming week or so.

Sunday evening

Counting today caught up with the incomplete early voting centres, leaving three consequential categories of vote type unaccounted for: absent early votes, absent election day votes, and around a quarter to a third of the postals which will trickle in between now and the cut-off point next Wednesday. My results system isn’t giving 18 seats away, but it rather errs on the conservative side in late counting. Only where the projected margin is inside 1% do I reckon the late counting to be worth following in detail, though a surprise may well emerge somewhere or other.

That means Maryborough, Pine Rivers and Pumicestone, all with the LNP ahead by respective margins of 1.0%, 0.4% and 0.7%. As will shortly be explained, even here the odds are fairly substantially against a late reversal. Leaving them out of the equation gives the LNP 50 seats, Labor 34, Katter’s Australian Party four, the Greens one, independent one and three in doubt. To deal with the latter in turn:

Maryborough. If 2020 is any guide, there should be a swag of over 5000 absent early votes here, giving Labor at least some hope of closing a gap that currently stands at 14903 to 14088, a deficit of 815. The surprise would be considerable though, as they only broke 54-46 Labor’s way in 2020, compared with nearly 62-38 overall. Absent election day votes were more favourable to Labor, though below par overall, and there should be less than 1000 of them. Then there’s a further 2000 postals that will surely favour the LNP, so this probably won’t stay on the watch list for long.

Pine Rivers. The LNP leads 14488 to 14894, or by 406, to which upwards of 2000 outstanding postals should add around 200. There should be around 2000 absent early votes and 1300 absent election day votes, which collectively scored similarly to the overall result in 2020.

Pumicestone. Here the LNP lead is a formidable 962, or 14916 to 13954. My system is not giving this away because absent early and absent election day votes, of which there should respectively be around 5000 and 1000, broke around 59-41 in their favour in 2020, compared with 55-45 overall. However, that suggests a Labor gain of only around 300, and even that should be partly offset by around 2500 late-arriving postals favouring the LNP.

Saturday evening

After what initially looked like a remarkably weak result for the Liberal National Party early in the evening, the Queensland election in many ways played to script, once late-reporting pre-poll votes were shown to have swung harder than votes cast on election day. The LNP appears headed for a modest majority in its own right, built largely on a long-anticipated regional backlash against Labor that encompassed all three Townsville seats, the Cairns and Cape York Peninsula seats of Barron River, Mulgrave and Cook, and further northern Queensland losses in Keppel, Mackay and likely Rockhampton.

Labor perhaps did better than expected in regional seats further south, potentially pulling off an upset win in Bundaberg. However, they appear likely to lose Hervey Bay, Caloundra, Nicklin and Maryborough, whose retiree-heavy population delivered the party rare victories in 2020. However, my results system is not giving any Labor seats away in Brisbane, although they are behind the eight-ball in Aspley, Pine Rivers, Pumicestone, Redlands and, somewhat surprisingly, Capalaba. Labor has its nose in front in its only Gold Coast seat, Gaven, and will more than likely recover its by-election loss of Ipswich West.

It was a disappointing night for the minor players, particularly for the Greens, who far from expanding their inner urban footprint have been defeated in South Brisbane, where they were not granted a repeat of the LNP’s decision to preference them ahead of Labor in 2020, and fell well short in their other target seats, leaving them only with Maiwar. Talk of Katter’s Australian Party sweeping all before it proved off the mark, although they retain their firm grip on Traeger, Hill and Hinchinbrook. Stephen Andrew, who defected to the party from One Nation, is fighting off a challenge from the LNP in Mirani. One Nation emerged empty-handed, and Sandy Bolton in Noosa remains the parliament’s only independent.

This post will be progressively updated over the coming days on late counting in seats in doubt, of which there are a good many. Quite a few pre-polling centres had yet to report as of the close last night, so we should see large numbers added to the count today.

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics. View all posts by William Bowe