Queensland election live: day two

Thursday morning

I’ve now taken Gaven, Cook and Burdekin off my watch list, the first two credited to Labor, the latter to the LNP (see below for further detail on Burdekin). That gets Labor to 45, which then becomes 46 if Margaret Strelow is right to have conceded defeat in Rockhampton, which she presumably is (more on that below also). To get to 47, Labor would need one out of the following: to retain Macalister, which will happen if independent Hetty Johnston can’t close a 3% gap against the LNP on preferences (which I would rate somewhat likely); Townsville, which is going down to the wire with Labor very slightly behind; and to be the beneficiary of Scott Emerson’s defeat in Maiwar, which seems somewhat more likely to go to the Greens. The ABC computer is predicting 48 for Labor, but I’m not sure why, because it only projects them with leads in 47.

The latest iteration of my results table looks as follows, with explanatory notes to follow:

Rockhampton

The big news of the day was independent Margaret Strelow’s concession that she is not going to win, contrary to most back-of-envelope projections to this point. The kicker is apparently a very tight 60% flow of preferences from the LNP to the One Nation, which will cause the latter to overtake Strelow at the second last exclusion, by a fairly comfortable margin of around 400 votes on my reckoning. One Nation would need about 55% of Strelow, LNP and Greens preferences to overtake Labor, and evidently Strelow’s are favouring Labor enough that this is not going to happen. It seems a full preference count will be conducted today.

Macalister

The count here seems unusually advanced, so there will presumably not be much change to the current results – which is good for independent Hetty Johnston, who has been getting smashed on postals. The key to the situation is the LNP’s 26.66% to 23.33% lead over Johnston, which she needs to close to poach the seat from Labor. The sources of the preferences will be the Greens on 6.54% and three minnows on 6.82% between them. Out of a three-way split of preferences, Johnston’s share will need to be about 25% higher than the LNP’s. Buried deep in a typically eyeroll-inducing report from the Courier-Mail is the news that Labor is very confident that this won’t happen.

Townsville

Not much progress in the count yesterday, with 90 postal votes breaking about evenly, and Labor clawing back about 30 on rechecking of booth votes. I still have the LNP a few dozen votes ahead, but there are perhaps 2000 absent votes that are yet to be counted, which might turn up something for Labor – though they were in fact slightly favourable to the LNP in relative terms in 2015. The same goes for maybe 700 outstanding out-of-district pre-poll votes. Also to come are around 600 declared institution, polling day declaration and uncertain identity votes, whose idiosyncrasies cancelled each other out last time.

Maiwar

All that was added yesterday were 130 postals, which increased the Greens’ primary vote lead from 37 to 43. Still to come: about 1500 absent votes, which are historically strong for the Greens; about 400 out-of-district pre-polls; a trickle of postals and 200 or so odds and sods. With scrutineer talk of a strong flow of Greens preferences out of the 737 votes for independent Anita Diamond, Labor will need to do extremely well on the outstanding count to get their nose in front.

Burdekin

My projection that the LNP would pull away here is looking pretty good after 652 postals were added to the count yesterday, breaking 430-222 to the LNP if preferences behaved as before. I’m projecting a 637 vote LNP win, and while this is probably inflated by an overestimate of the number of outstanding postals, I’m no longer regarding it as in doubt.

Hinchinbrook

Not really anything to follow here, as we won’t know the real situation until the preference distribution. However, it looks very much to me like One Nation’s narrow lead over Katter’s will be eliminated by Labor preferences, and that Katter’s will then ride home over LNP member Andrew Cripps on One Nation preferences. That’s unless Labor gets a strong flow of preferences from independent Peter Raffles and the Greens (3.04%), in which case Labor will close a 21.02% to 18.83% deficit against Katter’s, causing the latter to be eliminated in fourth place. In this case, there would need to be a Labor preference share around 30% higher than that to the KAP – plausible in the Greens’ case, but there doesn’t seem any reason to think preferences from Raffles, who wants statehood for north Queensland, will not go strongly to Katter. As top candidate on the ballot paper, some of Raffles’ vote would be of the donkey variety, and that vote won’t harm Katter’s.

Tuesday evening

The ECQ have unhelpfully pulled the notional two-party counts from their site. That makes it particularly difficult to track Burdekin, where Labor today picked up a bonus I hadn’t been factoring in: a strong pre-poll booth at Collinsville, which would have narrowed Labor’s two-party deficit from 366 to about 150. However, I’m still projecting the LNP to gain a couple of hundred votes on remaining postals.

Labor had a much better day today in Townsville, getting 35.3% of the primary vote from a batch of 635 postals, compared with 28.5% from the earlier batch of 885. The LNP’s 37.7% vote in the first batch fell to 33.2% in the second. Based on earlier reported preference flows, I’ve got Labor paring their deficit back from 78 to 31, and the projected losing margin down from 312 to 154 – and with perhaps 3000 voters yet to come, there’s a fairly substantial margin for error on that.

In Maiwar, the Greens are now 37 votes ahead of Labor on the primary vote, pending the unknown quantity of the preferences of independent Anita Diamond, who is on 734 votes. Kevin Bonham hears scrutineer talk that the Greens are getting a strong flow of preferences from those votes, to the extent that they should boost them by about 200. The two main outstanding categories of vote are absent votes, both of the pre-poll and polling day variety. If these favour the Greens like they did in 2015, I’m projecting the margin to increase by 135 votes. That does not factor in what will presumably be a few hundred outstanding postals, which have so far been fractionally more favourable for Labor.

Monday evening

A better day for Labor today, with three indicative two-party counts in seats where the ECQ had picked the wrong top two all bringing good news. In short:

Cook. A Labor-versus-One Nation throw records One Nation receiving 64% of preferences, when they need more like three quarters. The only remaining question is whether it will indeed be One Nation facing Labor in the final count, the other possibility being Katter’s Australian Party, who might get a better flow of preferences. However, there are likely to be only about 2500 votes left to be added to the count – in which case KAP would need to outpoll One Nation by nearly 10% of the outstanding vote, when they are closely matched at present.

Maiwar. Labor will clearly defeat the LNP’s Scott Emerson if it make it to the final count ahead of the Greens (I’m not actually clear in this case why the ECQ wasn’t conducted an LNP-Labor count all along). Presumably Labor preferences would go even more strongly to the Greens, to the outstanding question is who gets over the hump. The Greens currently have a lead of 19 votes, subject to the slight impact independent candidate Anita Diamond’s preferences will have.

Burdekin. Previously identified as a technical LNP gain, meaning a retain in a seat the redistribution had made notionally Labor. With a Labor-LNP throw now conducted, it has emerged that Labor has enjoyed a strong flow of One Nation preferences, and trails by only 34 votes. The seat is prompted me to add it to the summary chart below, where it constitutes a potential Labor to gain to compensate for the fact that I’ve now called Pumicestone for the LNP along with a detailed projection. The latter suggests today’s development is a false alarm for the LNP, who have a huge advantage on postals that is yet to flow through to the published two-party count.

Not featured in today’s two-party throws: LNP versus KAP in Hinchinbrook; Strelow versus Labor in Rockhampton; Johnston versus Labor in Macalister. Next to nothing happened today in Gaven and Townsville.

Sunday evening

Today’s counting has yielded two notable developments, both of them unfavourable to Labor. The LNP has roared back into the race in Townsville, performing very strongly at the city’s pre-poll centre and in the first half of postal votes. Postals have swung to the LNP by 8.9%, pre-polls by 6.8%, with the latter doubling in number since 2015. Some activity of the Defence Force that I’m not aware of may have had a bearing here.

Labor’s lead in Aspley has also withered from 2.2% to 0.6%, with postals swinging to the LNP here as well. However, that seems to most of the postals accounted for – most of the outstanding votes now are absents, which are likely to favour Labor.

I now have detailed projections for the three seats I am reading as straightforward Labor-versus-LNP contests, which are Gaven, Pumicestone and Townsville. These suggest Labor is in real trouble in Townsville and has little chance in Pumicestone, but will most likely win Gaven.

Not much has happened in the count today in Gaven, so what it says below is much the same as yesterday. In Pumicestone, Labor had a raw vote lead of 309 last night, but I was calculating this would become a 53-vote deficit when primary votes in the count were added on two-part. I then projected a 228 winning margin for the LNP on the final count, with the LNP to gain 341 on postals and 217 on absents. Once again though, postals have been bad for Labor, swinging against them 4.5%, such that I am now projecting the LNP to win by 535.

Including Gaven and Aspley, I can see a clear 44 seats for Labor; losses in Cook or Macalister I would still rate as unlikely, but they simply cannot be ruled out given the lack of hard information about preferences. That leaves them still needing an extra seat to reach the magic 47, for which their best chances are squeezing out the Greens in Maiwar or hanging on in Townsville.

Saturday evening

As I see it, in the race for 47 seats, Labor is on 43 and the LNP is on 38; there are at least two for Katter’s Australian Party, one for One Nation and one independent; and then there are eight seats that I’m treating as up in the air in one way or another. First up, there are eight seats that I’m treating as having changed hands. No doubt I’ll be proved wrong about some of them, but I figure you’ve got to start somewhere.

Aspley. Labor has held a stubborn lead of a bit over 2%, which doesn’t look like being overturned.

Redlands. Surprisingly, Labor’s only entirely clear gain from the LNP, off a swing of 6.3%.

Noosa. Independent Sandra Bolton seems to have surprised everybody by topping the primary vote in Noosa. Bolton appears to be exquisitely inoffensive, so there is no chance of the LNP chasing her down on preferences.

Nicklin. With the retirement of independent Peter Wellington, Nicklin returned home to the LNP.

Bundaberg. Gained by the LNP from Labor on a 1.2% swing, putting them 0.7% ahead, which will surely increase on late counting.

Mirani. This looks very much like a case of LNP dropping out and deciding it for One Nation over Labor on preferences. It may be within the realms of possibility that One Nation would tank so badly on late counting they finished third, in which case they might push the LNP ahead of Labor. But I’m putting that in the long shot column for now. For one thing, I’d think veteran Labor MP Jim Pearce would do okay on preferences.

Burdekin. In a seat held by the LNP, but made notionally Labor by the redistribution, this is a near three-way tie on the primary vote. If Labor drops out, the LNP wins. If One Nation drops out, I guess Labor has a chance (its preferences were directed to them). If the LNP drops out, One Nation wins. But the LNP does in fact have a slight lead, which will presumably increase on late counting. So for now I’m calling it an LNP gain from Labor.

Maiwar. Lost by the LNP, but not known whether to Labor or the Greens.

Then there are a further seven seats that I really don’t care to call, for one reason or another. I will be adding summaries of the situation in these electorates as I complete them. To start with though, here’s what I see as a summary of the situation:

UPDATE: For now, I have completed my analysis/projection of Gaven – the others I plan to do will have to wait until later today. The table below shows actual results in the first four columns, and my best attempt at projections in the last two columns. This requires estimates both of the number of outstanding votes, which involves at least as much art as science, and the two-party split. In the case of postals, for which about half the anticipated total have been counted, I have projected the results from the counted votes on to the uncounted. This is bad for Labor, as postal votes were weak for them to begin with, and appear to be recording no swing.

For other types of vote, it is assumed they will observe the same idiosyncrasies as in 2015. On this basis, Labor is projected to do well enough on absent votes to hold back the tide on postals, which largely reflects a strong Greens vote on absents in 2015.

For the other seats I’m listing as doubtful, just the briefest of rundowns for now:

Maiwar. The Greens have a raw 0.7% lead ahead of Labor in the race to finish second and, presumably, win the seat from the LNP on the preferences of the other. No absents or postals have been counted; the former should be good for the Greens, the latter bad, and there should be roughly equal numbers of each. So the Greens would seem favoured, but it’s certainly not done and dusted.

Pumicestone. Labor has a raw lead of 309 votes (0.9%) on the two-party count, but there won’t be much of it left when votes that have presently been counted only on the primary are added to two-party preferred. Postals should as usual favour the LNP, but Labor’s big hope is that the LNP tanked on postals in 2015. None of either have been counted yet.

Cook. With Labor on 39.3%, and a crush of others just shy of 20% (One Nation 18.9%, LNP 17.9%, Katter’s 17.6%), one of the latter will need a strong flow of preferences from the other two to make it home. I would expect that a Katter candidate in the final count would be most threatening to Labor, followed by One Nation, followed by the LNP.

Macalister. Labor faces a threat here from independent Hetty Johnston, but it’s a long shot — she trails the LNP 26.4% to 24.2% on the primary, which she needs to chase down with either preferences or an unusually strong late count performance for an independent.

Rockhampton. With Labor’s vote on only 31.8%, independent Margaret Strelow would seem assured of taking this if she finishes second. However, the LNP looks like bowing out before One Nation, who it had second on its how-to-vote card. So it would seem possible that Strelow will actually run third, in which case I imagine her preferences would decide the result for Labor. For all I know though, there may be a One Nation surprise lurking in wait here. Labor could wear a defeat at the hands of Strelow, a Palaszczuk-backed Labor preselection candidate who could potentially be lured back to the party, or perhaps made Speaker.

Thuringowa. The order here clearly runs Labor, LNP and One Nation about even on second, and Katter’s fourth, with the latter’s preferences presumably set to secure second place for One Nation. The question then arises as to whether LNP preferences go cleanly enough to One Nation to finish the job for them. UPDATE: They don’t – what I had thought was an ABC estimate is actually a real preference count that makes clear One Nation can’t win. So the only conceivable threat to Labor is the LNP, and that’s a long shot.

Hinchinbrook. The LNP incumbent here is on 30%, and then there’s a crush of One Nation, Katter’s and Labor around 20%. Provided Katter’s can stay in the count when the field is reduced to three, they would seem set to take the seat. Otherwise, the final count looks like being LNP versus One Nation, with Labor preferences saving the day for the LNP.

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