South Australian election minus two days

Nick Xenophon wins debate, but loses favouritism in Hartley; Liberals told off by Electoral Commissioner; Labor cuts deal with Australian Conservatives.

Nothing new on the polling front, but other developments worth noting:

• A leaders’ debate last night, which was deemed to have been won by Nick Xenophon by 30% out of the audience of 106 undecided voters, compared with 22% for Steven Marshall and 19% for Jay Weatherill.

• Electoral Commissioner Mick Sherry has ruled in favour of a Labor complaint against Liberal Party electoral material promising a $302 cut in household power bills, and a Liberal complaint against Labor Senator Alex Gallacher’s claim that the Liberals had a “secret plan” that would cut the state’s GST revenue by $577 million. The rulings arise from the state’s unique provision against “inaccurate and misleading” election advertising, which places the Electoral Commissioner in the awkward position of having to adjudicate on matters of political controversy.

• Labor has cut a preference deal with Australian Conservatives in which the latter will direct preferences to Labor ahead of the Liberals in the crucial seat of Lee, and run split tickets in two other tight Labor marginals, Light and Newland. In return, Labor will recommend the party be given third preference after the Greens in the upper house, although the experience of the 2016 election suggests this will have very little impact. The Conservatives’ state leader, Dennis Hood, told the Advertiser the party was not directing preferences to the Liberals in Lee because the Liberals hadn’t asked.

• Betting markets have been leaning towards Labor as party to form government, but Ladbrokes has recently shortened the Liberals in a number of seats where they are under challenge from SA Best. Most notably, Vincent Tarzia is now a $1.72 favourite in Hartley, with Nick Xenophon out to $2. Liberal odds have also been shortened against Labor in Mawson, Elder and Black.

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.

75 comments on “South Australian election minus two days”

Comments Page 1 of 2
1 2
  1. Yeah we could do with a debate about potential pensioner changes.

    Still no one knows how the cards will fall.

    I just want to see things improve. My faith in either side is not strong.

  2. Some simple modelling by me based on 2016 federal election data shows, not unexpectedly, that as soon as the ALP or LIB primary vote drops below 40 (when there is a significant third party vote) the party in front on primaries is in danger of dropping behind on the 2PP vote after preferences.

    Based on the 2016 SA federal data for various seats, and then recalculating the 2PP for the various opinion polls published by the Advertiser recently using the published primary votes gives the following model predictions:

    For LIB v ALP contests (model error +/- 1.7%):
    Lee: (Poll said 50-50, mine says ALP leading 50.2-49.8)

    Mawson: (Poll said 50-50, mine says ALP leading 52.2-47.8)

    Dunstan (Poll said 53-47 to LIB, mine says LIB leading 54.7-45.3)

    For SAB v LIB contests (with lower confidence):
    Morialta (Poll said 52-48 to LIB, mine says SAB leading 52.1-47.9)
    Heysen (Poll said 51-49 to LIB, mine says SAB leading 53.2-46.8)
    Hartley (Poll said 51-49 to LIB, mine says SAB leading 54.3-45.7)

    Not enough data to calculate models for ALP v SAB contests sadly.
    The flow of preferences in the Galaxy polls appear to be a little more
    generous to the Liberals (on average) than the 2016 federal data.

  3. Sykesie
    That approach works very well in predicting TripleJ’s Hottest 100.
    It might have a few flaws in politics though. Bernard Finnigan would have been a big winner for eg.

  4. Actually Sykesie, Google (the company) didn’t make that prediction. If you can see Google search data as having a life of its own, I suppose you could say it made the prediction – but it was really an outfit called Kwasi Studios that analysed the data and used it to tentatively make a prediction. I guess it’s as good as any for this totally unpredictable election.

  5. Yeah I must say some of your modelling results seem more plausible, Syksie. But I do like the outcome of Kwasi’s guesses – 23-15!!!

  6. It is all very confusing.

    However it seems to be a view that the margin between SA Best having a very good election and totally missing out is very fine. Northern suburbs Labor seats Enfield, Taylor and Elizabeth are being seen a genuine chances for SA Best along with Giles, Hartley, Heysen and maybe some safe Liberal seats.

  7. The South Australian Hoteliers Association has been very active in the Sa election, spending heaps of money on the SA election. Xenophon says $2m. I know some hotels are currently full of “anti Nick” posters. The publicity is all about the impact on jobs, music venues and suppliers. They suggest the SA-BEST policy of reducing the number of pokies by 50% over 5 years will damage all of these things – and have spent heaps of $ saying so. Really?

    Perhaps ….. not.

    Will the pubs and hotels remove the posters they have in numerous bars in the festival state – 2 days before the election.

    I doubt it. Even so, have opinions been formed already? So … if the election result ends up being questionable, and this is a possible reason, could it be challenged in court?

    Xenophon is a lawyer after all!!

  8. why would shorten make a major and controversial it seems policy announcements days before a major state election and federal bi-election? is the judgment to win votes? does shorten really have a clue about PR?

  9. “why would shorten make a major and controversial it seems policy announcements days before a major state election and federal bi-election? is the judgment to win votes? does shorten really have a clue about PR?”

    Well it is great policy and shows a decisiveness and courage most of us wouldn’t have expected from Shorten. The he’d take on the disgusting greedy filth that have come out violently angry at having their unneeded and undeserved Govt handout, shows he is fit to be a good PM, whereas before that was in doubt.

  10. @jack I put in numbers based on the last Newspoll but corrected for seats being contested. SAB got 21 seats in that simulation. Either we’re all missing something or everyone, but especially the media who’ve been running the Xenophon has lost support line are going to get a massive shock on election night.

    It was interesting to hear on the ABC this morning that Antony Green thinks the result will be known on the night, and he thinks he knows who it will be!

  11. Liberal Party have finally hit the front in Sportsbet if only marginally. Current standing is Libs $1.87, Lab $1.90 and SA Best $21.00.

    Anyone care to make a prediction? I’ll go Lib- 21, Lab- 21, Ind- 2 (Bedford and Brock), SA Best 2 (Heysen and Hartley)

  12. Ok Max – with about 2% confidence in its accuracy, I’ll say Lab 20, Lib 15, SAB 9, others 3. Ie the real liberals to steal a number of seats from the misleadingly-named “Liberals” but hardly any from Labor.

  13. I’m not sure why the late money for the Liberals; based on nothing but the vive I suspect. We haven’t seen any published polling that would breed that sort of confidence-I guess that could all change later today though. Just the vive and the media telling us it’s time I guess.

  14. Matt
    At this point I still have no idea. Anything from Lab-minority through Lab- SA coalition to Lib – SA minority still seems possible. The horse trading after Saturday night will probably determine who is premier. Shades of October 2017 in New Zealand.

    If that is the case personalities will play a big part. Does anyone have a feel for how both X and his candidates would line up in choosing to back Weatherall or Marshall?

  15. Agreed Socrates, I don’t know how strong the ‘it’s time’ factor is supposed to be, but the fact that it’s this close speaks volumes about Weatherill; showing strength on an issue that the media has tried to portray as a weakness.

    If Labor loses and Jay steps down, does Koutsantonis become opposition leader? That might result in 16 years of Liberals!

  16. Bonza

    Depending on who X alliances with, that is a distinct possibility. It would be funny if both X and Marshall lost, and a Liberal – SA Best coalition then had to be stitched together.

  17. If Marshall loses his seat, there will be almost certainly be some sort of Labor government. If Dunstan goes, Adelaide will already have been lost.

  18. I think we can say with reasonable certainty that Peter Malinauskus will be the next leader of the SA Labor party.

    Depends on tomorrow’s result. If Labor win again and Weatherill decides to quit mid-term, I doubt he’d want to captain the Titanic to defeat.

    However, I definitely agree it’s a near certainty he will be the next Labor opposition leader, whether it’s next week or in 2022 (And yes, if Labor get back in tomorrow, it will definitely be the last time.)

  19. Has anyone given a thought to the Legislative council result.
    My thoughts being ALP 3, Lib 3 or 4, SAB 3. If the Libs do get 4 (or Labor for that matter) then that leaves the Greens, Aust Conservatives and Kelly Vincent fighting for the remaining seat. My premise for giving the Libs 4 seats is that (and I may be wrong) is that I have a feeling that SA Best will get more votes from disaffected Labor voters than from disaffected Libs.

  20. Liberal Party have finally hit the front in Sportsbet if only marginally. Current standing is Libs $1.87, Lab $1.90 and SA Best $21.00.

    FWIW, the others still have Labor as narrow favourites but gaps have all closed. The odds are (I’ll just put Lab v Lib, as they’ve all written-off Xenophon):


    Labor: $1.90
    Liberal: $1.87


    Labor: $1.80
    Liberal: $2.00


    Labor: $1.80
    Liberal: $2.10


    Labor: $1.70
    Liberal: $2.10

    Please note: These odds aren’t final and are still subject to chance. Also, let me know if there are any other legit agencies I’ve overlooked (I’m not going to list every bookie in the world that has a stake here.)

    The only take-away I have from those numbers is nobody really knows. The last minute movement to the Libs is probably just a movement of those who have left it to the last minute and are following the conventional favourite – and they could be right for all I know. But, yeah, there’s nothing really significant to read into this stuff IMO.

  21. Has anyone given a thought to the Legislative council result.

    Yeah, my tentative (albeit more confident than any HoA prediction) guess is: 3 Labor, 3 Liberal, 3 SA Best, 1 Green, 1 Conservative.

  22. I’m calculating ALP 22 / Lib 19 / SAB 3 / IND 3.

    First draft I had 20/18/6/3 but I’m trusting Sportsbet’s pessimism on SAB…

    Looking at your numbers Max, I’d say we have pretty similar assumptions although I’ve calculated SAB with a chance in either Taylor or King, and Bell as likely in Mount Gambier. I’d say you’re then guessing Adelaide and Mawson vote Liberal?

  23. I do hope that, if Labor lose, they don’t blame their big vision ideas and “go back to basics” in future elections in order to try and win the grey/NIMBY vote. Also, I hope the Catholic Right don’t try and blame the progressive nature of the government and successfully push for a more socially conservative Labor on the premise that it will appeal to the “middle” more.

    Be like (non-SA) Liberals. Treat it as a mere setback. Certainly learn from genuine mistakes but keep pushing forward. No retreating.

  24. Jack- 2% confidence is about where I am at as well.

    MM- had Labor keeping all their current seats under the revised redistribution, and then gaining Adelaide and Newland.
    I’m not convinced at all that Troy Bell will hold Mt Gambier so have given that to the Libs. I would have given Lee away as well, but I think Stephen Mullighan might just have a strong enough personal vote to keep it by the barest of margins.

    Re: labor seats going to SAB- definitely wouldn’t surprise to see 2-3 go. If I had to choose one I would probably say Enfield. I am basing that purely on John Rau’s behaviour over the last week and the fact that he will be friendless when it comes to preferences without getting near a 50% primary.

  25. And, if the Liberals lose, I hope they finally have that long-overdue introspection and stop blaming everyone else or think they just need a charismatic leader, and actually start changing their policies and approach to make themselves a more effective and electable opposition that suits the needs of the state, not just Liberal Party ideology. It’s up to them to adapt to the electorate – not the electorate to adapt to them.

  26. I’m holding out hope for one more poll tonight, then I will attempt some form of House of Assembly prediction.

  27. I would be pretty confident we will get a Newspoll tonight. I also think I heard that Reachtel were in the field. So I’d say definitely one, possibly two polls tonight. How much they actually help us though is very questionable, unless there’s a sharp move one way or the other, probably not much.

  28. Most people don’t know that the way we vote in State elections has now changed. Yyou can now vote for more than one Party “above the line” on the large white Legislative Council (upper house) voting paper. In the past, you could only mark 1 in one Party’s box above the line, and you lost control of your preferences. Now, you can choose where to send your preferences by numbering as many or as few Parties as you want.

  29. Liberal now 1.80 favourite with Sportsbet, and IF you believe the seat markets, they are now favourite in 23 seats. So IF they win all these, add in Mt Gambier and they are across the line with 24.
    I’ve always found there to be lots of inefficiencies in politics betting markets, so take that with however many grains of salt you like.

  30. I was lacerated by some forum-ites when I forecast a result of 27 Lib, 16 ALP, 3 Ind, 1 SA Best.

    Apparently was optimistic for the ALP, which is likely to lose seats to SA Best.

  31. Been playing with the brilliant SA votes calculator and also 2016 3PP Senate figures broken down by electorate, linked to by Ben Raue on the Tallyroom in December

    Based on the most recent Newspoll it would seem to me that SAB will do a lot better than most people expect. Unless the that Newspoll is dramatically wrong, I think they will displace the Liberals as the second largest party.
    Since they are getting preferences from both the other major parties, they could conceivably emerge as the largest party, even though they will have the lowest vote, and the Liberals the highest! That’s the way preferential voting works. Its amazing that this preferencing situation hasn’t received more attention as its a huge change from the situation in the federal election.

    I predict SAB will win the following seats: Finniss, Heysen, Kavel, Narungga & Davenport – all Liberal seats with no sitting member. I’d add to these Giles (ALP) and Chaffey in the country and Gibson, Morialta, Hartley, Hurtle Vale (ALP), Reynell (ALP) and Waite in the metro area. The northern suburbs ALP seats look harder because of the high primary vote but I’d flag Taylor and Playford as the two most vulnerable. Enfield has been mentioned, but the Liberal vote there looks too high. This is also likely to stymie them narrowly in marginal north-eastern seats like Wright and King, while in Elder a sitting ALP member should ensure SAB are third in an otherwise eminently winnable Liberal seat.

    By my calculations there are about 10 seats that SAB should win if the Newspolls and the 2016 federal senate results hold. In addition there would be around 8 more which should be basically line ball.

    Just like Qld in 2015, the Trump victory and the recent UK result, the writing is actually on the wall, people are just failing to see it.

  32. Nick Xenophon has been ordered to delete online electoral ads for SA Best following a complaint by the Greens to SA’s Electoral Commission, which found the materials to be “inaccurate and misleading”.

    According to the commission, a Facebook post by Mr Xenophon from March 5 and an associated web page make false claims about Greens preferencing.

    In the post targeting Greens voters, Mr Xenophon claimed the Greens were part of a “dirty deal” to put both the Liberals and Labor ahead of SA Best in seats including Hartley, which the former senator is contesting.

    Mr Xenophon further claimed the Greens had “defied principle and logic … they are siding with the Liberals, many of whom are still captured by the coal lobby”.

    The Greens — who admit to a preference deal with Labor, but not the Liberals — referred the material to Electoral Commissioner Mick Sherry, who was satisfied no such deal with the Liberals had been done.

Comments Page 1 of 2
1 2

Leave a Reply

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