What happened

How a dramatic change in the behaviour of minor party and independent preferences powered Labor to the cusp of victory in Queensland, plus results broken down by region.

I’m going to progressively add a region-by-region post mortem to this thread as I go, so be sure to hit “refresh” every now and then if you’re hovering around the site and you think that might of use to you (UPDATE: Scratch that – I’ll do it tomorrow). I’ll start by answering a question that will be on the lips of many: how did the polls (and by implication the poll aggregators – see the sidebar) get it so wrong? The answer to this is simple: by allocating preferences as they flowed at the 2012 election. In fact, the pollsters did well at predicting the primary vote. As noted two posts back, all three late-reporting pollsters essentially had the LNP on 42% and Labor on 37%. The present vote totals are 40.8% and 38.1%, and they will probably edge closer in the direction of the poll results during late counting. This tends to argue against the notion that there was a violent late surge to Labor – and also the notion that there is likely to be a particularly big move back to the LNP on late counting.

The real key to the surprise is that Labor’s share of minor party and independent preferences went from 27% to 45%, the LNP’s went from 22% to 15%, and the exhaustion rate fell from 51% to 39%. This is based on slightly incomplete data, but it should be probably be near enough. Applying those preference flows to the current primary vote totals, it’s Labor and not the LNP that comes out 52-48 ahead, contrary to what all three of the aforementioned polls said. If the pollsters had been operating on accurate assumptions concerning preferences, they would have come out at 51-49 to Labor. Which gives them a lot to think about going forward, particularly with another optional preferential election on the way in New South Wales in a little under two months.

Anyhow, stay tuned.

UPDATE: Or I could leave it for tomorrow. Yes, I think that’s the go actually. Here’s a table that should set you a long way towards working out what I’ll end up saying. Seats are rated in doubt if Labor is ahead by less than 1.8% or the LNP is ahead by less than 1.2%, according to a crude 0.3% estimate of how much late counting generally favours the conservatives. The six in question are Mansfield (Labor leads by 0.1%), Mount Ommaney (LNP by 0.9%) and Redlands (LNP by 0.9%), in Southern Brisbane; Ferny Grove (Labor by 1.6%), in Northern Brisbane; and Whitsunday (LNP by 0.2%) and Maryborough (a special case, in that it will come down to who finishes second out of Labor and independent Chris Foley) on the Central Coast. Since Maryborough is in doubt between Labor and an independent, the ceiling for the LNP is 43, which includes the long shot of Ferny Grove.

Inner Brisbane 41.6% 40.1% 15.7% 2 8 0 0
-8.2% +7.2% +1.5% -7 +7 0
Northern Brisbane 41.3% 44.3% 8.9% 3 8 0 1
-11.4% +14.0% +0.6% -9 +8 0
Southern Brisbane 38.7% 47.1% 8.7% 2 10 0 3
-11.1% +13.7% +1.2% -11 +8 0
Ipswich 29.0% 54.3% 7.0% 0 3 0 0
-9.4% +20.4% +1.3% -2 +2 0
Gold Coast 48.1% 30.2% 7.8% 10 0 0 0
-10.2% +6.6% +1.1% 0 0 0
Sunshine Coast 45.2% 24.2% 12.7% 6 0 1 0
-12.4% +7.8% +0.7% 0 0 0
Central Coast 35.2% 38.2% 4.5% 4 6 0 2
-6.8% +13.1% +0.3% -5 +4 -1
Northern Coast 36.8% 40.6% 6.1% 1 7 0 0
-3.4% +13.1% +0.4% -6 +6 0
Urban Hinterland 45.0% 30.5% 6.1% 5 0 0 0
-5.6% +11.3% +0.3% 0 0 0
Interior 46.6% 23.5% 4.0% 5 0 2 0
-5.5% +8.5% +0.7% 0 0 0
TOTAL 40.8% 38.1% 8.4% 38 42 3 6
-8.9% +11.5% +0.9% -40 +35 -1

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.

274 comments on “What happened”

Comments Page 6 of 6
1 5 6
  1. I’m certainly hoping for a Labor government and the news about KAP’s openness to deal just posted by ajm adds even more hope. But if it went the other way there’d be plenty of consolation for me. The result is still extraordinary for Labor – way beyond most people’s expectations. A minority LNP government is I suspect more likely to be fractious than a minority ALP government, and, fractious or not, is likely to be unpopular and so help the Federal ALP in Qld in 2016. I also wonder, though am glad to be contradicted, whether there is a risk of an ALP government running into political strife early with so many newbie MPs. Whatever happens it’s not too shabby!

    On a different subject, there was a great episode on This American Life tonight about Internet forums and trolling among other things ….. Some of the stories resonated deeply with life in PB world. Hilarious really.

  2. I doubt Palmer candidate in Maryborough had anyone handing out HTVs (even if he did allocate preferences) would expect his preferences to go everywhere & unlikely to put Foley ahead of Labor.

  3. Don’t think so, daretotread 249. see http://results.ecq.qld.gov.au/elections/state/State2015/results/district49.html only shows 612 prefs exhausted in the 2PP count of a possible 2,654 pref votes & about 1000 prefs not counted. 2PP also shows a 500 vote “narrowing” from the 1st pref count. I’d reckon there are still a 1000 prefs still to be counted in Mansfield, and they are flowing pretty solidly to Labor.

    All still before counting of declaration votes etc, of course.

    Don’t write it off yet.

  4. 253 Rod
    The 2PP votes are counted by booth, not by candidate. Its not possible that there are greens votes that haven’t been added to the 2PP count, unless there’s a mistake.

  5. For what its worth, daretotread, in Mansfield the Greens HTV card preffed ALP. The independent, Wirth, didn’t lodge one at all. I’d suspect many of those exhausted prefs actually come from a block count of Wirth’s prefs (or lack of ’em).

  6. Yes. You’re right too. I forgot about the informals, in which case it may well be too high an ant heap to climb for the ALP if the other uncounted categories flow close to even for the LNP.

  7. I don’t see any realistic path to government for the LNP. Wellington has already made his position clear, which will influence KAT’s position – and they were already leaning to the ALP. Plus the LNP has no leader at this critical juncture.

  8. Rod the votes uncounted in the notional count for Mansfield are from the out of electorate Pre poll booths. Most are from the Brisbane CBD Pre-Poll. Most of these votes are of course for the main parties and only just over 50 are Green votes so there will be hardly any help there.

  9. It seems to me most likely that the ALP will win Maryborough as well.

    With Wellington that’s 45. At that point, KAT can play spoiler of the speaker deficit, OR get a ministry / have a big influence on government. What they cant do is form government with the LNP.

    No brainer.

  10. Here’s something further to Wow you on a huge Wow weekend.

    Henry Palaszczuk ‏@HenryPalaszczuk · 5m5 minutes ago  Brisbane, Queensland
    In the electorate of Inala, @AnnastaciaMP, 7 out of 16 booths returned results above 80% 2pp, with Richlands booth 87.5%. amazing #qldvotes

  11. Yeah Rod, to think otherwise it to misunderstand what KAT represents. Labor is actually far closer the old-style NATs that KAT represents.

    At least since the ALP dropped privatisation, that is.

  12. Rod

    It is 625 votes and they are all the ones from the mixed electorate prepoll centres.

    Had a close look at them there seems to be about 268 Alp, 256 LNP, 36 Independent (his family I assume) plus 71 Green. Once included the ALP should go ahead by 40-50 votes.

  13. Kev Johnno. It seems as if all the votes are excluded not just minors so Labor gains 10 or so votes immediately before distributing the 108 preferences.

  14. thanks kevjohnno. Actually I was originally pointing out that the Greens prefs had really flowed quite strongly to ALP in Mansfield, but yes, I had over-estimated the number left uncounted by the looks of it.

  15. They seem to have included most in Ferny Grove but there are about 35 not included. These break heavily Labor’s way.

    Anyone notice the Mt Nebo booth. Small only 355 voters but Greens vote was 44%. LNP 18%.

  16. Rod

    It is going to be very hard for Labor to catch up 500 votes in Mansfield.

    Unless there was a fabulous job done on postals.

  17. http://andrewelder.blogspot.com.au/2015/02/bad-reporting-bad-decisions.html
    [01 February 2015
    Bad reporting, bad decisions

    If you believe that Anna Bligh led Labor to defeat with a huge swing in 2012 over asset sales, then you must also believe that Campbell Newman led the LNP to defeat last night for the same reason.

    Policy matters in political analysis. Only the analysis about what state government is actually for, the focus on schools and hospitals and law-and-order, makes any sense of why Queenslanders voted as they did.

    The people of Queensland have permanent interests, rather than permanent friends or enemies in politics. Once you understand that – if you can understand it – you can get past the idea that it is the political class that is fickle and obtuse, and not the electorate.

    Queensland’s optional preferential system, combined with the significance of parties outside the LNP and ALP, makes a mockery of polling in other jurisdictions. On the ABC election coverage Antony Green once again fought a brave but losing battle with his own software. I was half expecting it to rise up, like Frankenstein’s monster, and proclaim Aidan McLindon the next Premier. Green’s on-screen troubles pointed to a wider problem in the quasi-profession of political punditry.]

  18. Mansfield is definitely lost to Labor. The 2128 Postal, Special Postal, Unenrolled Votes that have been counted so far went 814 to ALP (39.25%) and 1,260 to LNP (60.75%)(54 exhausted). By the way the ECQ site seems to have a programming bug in that it always says there are 42 polling booths in the electorate regardless of how many there are. Most of the time its a lot less. In Mansfield there are 35 ‘booths’ but 19 of these are ‘pre-poll’ booths in other electorates or All-District Centres with almost no votes attached to them.

  19. looks like Maryborough goes to Labor, Mansfield goes to the LNP.
    So it all comes down to whitsunday…

    i would be very surprised if in coming weeks at least an LNP member of parliament did not resign in wake of their sleepwalking to defeat as they clearly expected to still be victorious.

    A by-election up for grabs for labor to win.

Comments Page 6 of 6
1 5 6

Leave a Reply

Your email address will not be published. Required fields are marked *