Senate and Page polls

Morgan has published a survey of Senate voting intention aggregated from its polling over October. As usual the minor party figures look a little inflated, while major party support reflects the slight improvement the Coalition seems to have managed during the campaign. We also have a poll of 300 voters in Page conducted by Grafton’s Daily Examiner and Lismore’s Northern Star, which they stress is “not intended to be scientifically accurate”. It shows Labor’s Janelle Saffin with a decisive primary vote lead over Nationals candidate Chris Gulaptis, 44 per cent to 41 per cent. A poor level of recorded support for the Greens is not of interest in itself, but it elicits an admission from candidate Theo Jongen that the party’s vote is “running at six per cent”, compared with 10.8 per cent in 2004.

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.

986 comments on “Senate and Page polls”

Comments Page 1 of 20
1 2 … 20

  1. In an earlier thread (before the unfortunate disruption) someone quoted the GG as reporting comments from SportingBet that the betting over the last week was behind the Coalition. I recommend the following summary of betting intentions on electorates in the gripping hours after the Coalition launch (and of course the new Newspoll)

    http://jackman.stanford.edu/oz/Aggregate2007/bettingmarkets/sportingbet/currentFluctuationsDotPlot.pdf

    which suggests that the only people who were not put to sleep by the Coalition launch rushed gasping for breath to plunge money on the ALP

  2. I wouldn’t be surprised that we see a drop in the Green primary vote (almost all heading to Labor) from Page right through to the Gold Coast (that’s my turf – I’m not going to speculate on other areas).

    It will be interesting (in Page at least and possibly McPherson) which way the remaining prefs split. The decision to pref Labor in Queensland may wash out the loss of disaffected Labor voters a bit. Not sure which way the prefs are going in Richmond.

  3. I heard Antony on PM tonight expressing what I thought was a fair degree of scepticism re the accuracy of Senate polling.

    Is ther a reason why that is so and if there is, can I guess it’s the method of voting?

  4. My old University supervisor Dr Ernie Chaples did a lot of work in the early 1980s that looked at identifying ‘softly committed’ voters. Out of that came the conclusion that while people could tell you who they were going to vote for in the House, they were much vaguer in their Senate vote. McNair Anderson used to do exit polls in the 1980s, and asked people who they voted for in both houses, and there was always a ‘can’t remember’ vote for the Senate. I suspect that people just make up their minds later, and haven’t concentrated at all on their Senate vote. And before someone berates me that lots of people take the Senate vote more seriously than their House vote, the numbers in surveys don’t back that up.

    The same applies to asking people who they are directing preferences to in the House. You’ll note that both Newspoll and AC Nielsen stopped asking that after 2004. Both companies got the primary votes pretty right, but the 2PP from voters responses was wrong. Everyone’s gone back to estimating 2PP based on past preference trends.

    Another factor I think is important in both preferences and the Senate is that no survey can catch the how-to-vote effort that has a big impact on preference flows, and also the ballot paper effects that distort the Senate vote from surveys.

  5. Richmond – Greens prefs going to Labor, Dems prefs going to Labor. Labor to get donkey as Justine Elliott (sitting member) got 2nd place on ballot with Nationals candidate @ No 4 spot on the paper.

  6. Antony Green Says:
    November 13th, 2007 at 8:41 pm

    My old University supervisor Dr Ernie Chaples did a lot of work in the early 1980

    Wouldn’t the above the line option have changed that? I would have thought that work would need to be done again.

    It is easy to remember an above line selection and only people who are seriously into politics vote below the line, the serious should be able to recall their preferences a little.

  7. A bit off target – I hav been watching simon Jackman’s blog over the past couple of weeks – the individual seat betting is gradually, painstakingly closing the gap between the individual seat betting and the overall result betting. see qute below from his latest blog.

    “ALP favoured to win in 77 of 150 seats (# seats with average ALP win prob > .5)
    ALP Expected Seat Count: 79.7 out of 150 seats. Yesterday: 78.93 out of 150. Computed as the sum of the 3-agency-average seat-by-seat ALP win probabilities.”

  8. I’m sorry to keep repeating myself, but all Senate polls are bunk. 90% of voters decide which party they are going to vote for, then they take their party’s HTV and they follow it, for both houses. (That’s why they can’t remenber who they preferenced – they let their party make that decision for them.) The only exception to this used to be that the Democrats got a significantly higher vote in the Senate than in the House, because a lot of Liberals voted for them in the Senate. The Greens don’t draw that vote, so now there is hardly any difference between the two houses.

  9. Antony, given a lack of forethought about preferences, how closely do you or your research think people follow how to vote cards on the whole?

  10. No, the research was after above the line voting was introduced. I would add one caveat. Like British Liberal supporters, the Australian Democrats voters were always hard to find in academic research. In the 1979 Australian Election Study, only around 4% of people interviewed admitted voting Democrat in 1977 or intended voting that way in 1980, nothing like what occurred at both elections. There was always a fair slice of Democrat support which was simply from people not choosing to vote for a major party with their primary vote. Transient voters were a significant part of Democrat support, especially in the Senate where their vote was often twice their House vote. But before somebody claims that point as relevant today, I’d point out that for the last 5 years, the Democrats have been on 2% in opinion polls and ended up with 2% at the general election.

    I’d stick with the first point I made. Later decisions on Senate voting, the fact that the Senate vote is not about who forms government, and the random factors introduced by the ballot paper, make Senate polls less reliable than House polls.

  11. Haha. Barnaby Joyce is on Agenda an appealing for people to ‘try to maintain a bit of balance’ by voting Coalition to avoid an annihilation.

  12. Thanks Anthony, still looks like the end for the Democrats and fortunately the end of the Libs holding outright majority in the Senate.

    Can anyone in SA tell me if Nick X is running on any issues or is it just his profile that counts?

    I know Stuart Henry here in Hasluck is running on grafitti and car hoons, so I suppose Nick might take “no pokies” to the federal level????

    I just took him to be more intelligent than Stuart Henry.

  13. Antony, I just want to know, how hard do you find it not to try and be the first one on the TV to call the election for either side?

    And also are we going to see a lot of graphs on the 24th of seats with “95 votes counted” and swings of 55%??!

  14. On the question about preferences…

    “Preference deals” are nothing more than decisions about the order in which votes are distributed on how to vote cards right?

    Meaning, if I number 1 box above the line on my senate paper and I have no subsequent preferences to be redirected and that party doesn’t enough votes to get a senator, then my votes is goes nowhere else?

  15. 12: That irritates me, darn statistics, get 3 votes and they’ve already extrapolated the conclusion!

    Which reminds me of a science fiction story about a future where the election is decided by a single voter, and all the science goes into a super computer who job it is to find the right voter to cast the single vote necessary.

  17. 13-s

    Senate preference deals are a bit more binding than that if you vote ‘above the line’ you will be in reality filling out all the ‘below the line’ boxes in accordance with how that ‘above the line’ party have registered thier senate preferences.

  20. S your vote could end up with the fishing party and ellect one of them if you’re not careful. I am partly responsible for getting FF as I voted ATL in the last election (never again)

  21. I love how Antony has to say “obviously that too early to tell”.

    Can I say though I hate the “experts” who always rabbit on about “my scrutineers tells me that one’s safe”.

    Antony, does the ABC computer take into account the booths when it is working out those early swings? Surely that would give a bit more of a picture – given all the info is on the AEC web site, surely you could get come boffins on the case? (bit too late though now I guess)

  22. Antony #9
    Interesting about the 4% Dem HR polling which resulted in much higher numbers. This is easily understood as meaning, when polled, ‘I really do support one or the other major party’ but at election time like the idea of giving some support to the Dems while knowing full well that they have no hope (but being fully aware of their preferencing).
    At the moment the Greens are polling considerably less than at the 2004 election.
    Do you think the phenomenon in above is at play again, now that Labor is similarly placed (in an apparently commanding electoral position) to what the Libs were in 79?

  25. Has anyone done any work on what would the impact of optional preference voting?
    When I lived in QLD, I loved it in the state elections. I like the preferential system above first past the post, but I do think optional preference is errr.. more prefered!

  26. I’ve only numbered every box on one occasion – when I was doing a postal vote.

    To be honest, I had no idea that preference redirection like that occurred.

    How dumb am I? (Don’t answer that)

  27. How accurate to people think those Roy Morgan Senate figures are?
    Based on the numbers and Antony’s Senate Calculator, it seems that the Greens will win/retain a seat in NSW/Vic/WA/Tas/ACT.
    In SA other is only 6.5% I thought that Mr X would poll >10% himself.
    Also could Tas go 2/2/2 ALP/Lib/Green? Could any other state elect 2 non ALP/Libs?

  28. Grog, we use the booths. The computer gives an exact matched booth swing. To be honest, I don’t usually hear what the ‘experts’ on the panel say. The acoustics of the tallyroom mean that on the set it is hard to hear anyone more than a few feet from you, so I only hear Nick Minchin and the other guest on the panel through my earpiece. That is only in one ear, so if I’m diving through figures in the computer, I tend to tune out what they are reporting their ‘scrutineers’ as saying. Usually their ‘scrutineer’ is an assistant sitting at the back of our set using the same ABC computer system as me. Guests will not call an election as over, especially if defeated, before the party leader appears. From years of doing this, the losing party guest always goes on ‘my scrutineers tell me we still have a chance’ or ‘there are x outstanding postals in that seat’. From years of tracking back through the numbers, the computer predictions are always more accurate than what the political guests say.

    And I must say, I can’t ever remember getting a 55% swing on early figures. You can’t get a figure like that using matched polling places.

  29. How dumb am I?

    Been approached by Four Corners lately?

    🙂

    FWIW, I always always always fill out every box. I know who I want to put last, and I’m going to do it if even it takes me all day. Which it sometimes has in the NSW elections – everyone else remember the “tablecloth” of a few elections back?

  30. I always like the part where Antony’s computer breaks down at the crucial moment and he announces that the Save the Speckled Drongo Party is going to win Gellibrand.

  31. Big Blind Dave

    Here are Mr X issues / positions (excerpts from http://www.nickx.net.au/html/faq.html)

    I know you are against Pokies but what other main issues you will fight for in Canberra?
    Water is the big issue that is critical to South Australia’s future. We deserve special consideration from Canberra because we are at the tail-end of a once great river system that has been abused and neglected. Our Riverland irrigators are facing economic ruin – this is a challenge for SA to be a world leader in water management, and we deserve Commonwealth help to achieve it.

    What are your views on the war in Iraq – should the troops be withdrawn?
    We should never have gone to Iraq in the first place. The government needs to acknowledge that it was a mistake. The big dilemma is if you are to leave now do you leave things worse than they are. There’s a role for reconstruction for Australia but it can’t be open-ended commitment. It’s a question of trying to extricate ourselves in a way that minimizes bloodshed to the local community. My view is that we should be spending money on aid and building up the political infrastructure. At the same time we need to be starving the terrorists of oxygen.

    What about WorkChoices – is it delivering a fair deal to Australian workers?
    The WorkChoices legislation went too far. There definitely needs to be some revisiting of it – you need to consult with small businesses that are nervous. There’s scope to have a good look at it away from the heat of the federal election campaign and scope to improve it. A big issue for me in Canberra would be industrial health and safety. In terms of the guarantee of minimum conditions, if someone doesn’t want a certain shift or expanded shifts there should be some safeguards.

    What’s your view of the Murray Darling Basin rescue package?
    South Australia needs to be given special consideration as we are at the tail end of failed water policy in the Eastern States. There’s an interesting constitutional argument under section 100 of the Constitution for the Commonwealth to override Victoria if they continue to hold it up. I think we should push the envelope and ensure that federal legislation guarantees SA gets a fair share. The other thing is that headline grabber – Howard’s $10 million water plan. That’s over 10 years. A lot of that money is back ended. Not much is happening in the next three years. We need that money to be spent sooner rather than later to help irrigators.

    Should Australia sell uranium to India?
    We have to be careful about where uranium is sold and the safeguards so it doesn’t end up being used other than for peaceful purpose. I’m a bit concerned India has not signed the nuclear non-proliferation treaty and that worries me. If they are dinkum about it why won’t they sign?

    Should the pulp mill be built in Tasmania?
    I wouldn’t have voted for it because the Tasmanian Government fast-tracked it. They didn’t have appropriate scientific evaluation and didn’t go through the right process. You can’t say it’s world best practice because dioxins will be released into the water system. I think it’s an act of folly.

    Should Australia ratify the Kyoto Protocol?

    Yes. It isn’t that radical a target. If you know what the rules are, businesses can plan accordingly. But if it’s kept vague you are not going to get that impetus for change. Australia needs to be leading the world on environmental change.

  32. Bryce, the Greens tend to pull in the same votes as the opinion polls. I haven’t done this systematically, but my impression is they do slightly better than Newspoll, but rarely as well as Morgan. My supposition is that the difference between the Greens and Democrat support is that the Green vote tends to be a committed vote, which is why it is the same in both houses, while the Democrat vote included a fare few ‘bet each way’ votes. Green voters are much more positive about their intended vote, while a Democrat vote was often an anti-politics vote.

  36. Now now Adam, we haven’t had a computer problem in a long long while. We had one problem in Melbourne at the Victorian election, where due to the way Linux drew a distinction between zero and null that our previous operating system didn’t, I was unable to over-rule early predictions of the Greens winning Melbourne. But in live television, you sometime have to live with these bugs.

  37. Antony

    Thanks for adding your input here – I always enjoy your contributions. Any chance of a hello to the bludger community on the night? In code? Something like ‘the computer statistics may be wrong, but our projections are showing the LNP losing 80 seats’? 🙂

  38. Diogenes – I am sick to death about questions on Wentworth. It is an irrelevant sideshow to this election. It might matter to people who live there, but it is a sideshow to the election result. It is a campaign being driven by the dinner table gossip of rich people played out on the front pages of local papers.

  42. Antony,

    Which are the best and worst State Electoral computer systems you have dealt with ?

    Are they of a uniform standard etc ??

  45. So is the Greens candidate Theo or Hank Jongen? Or did Theo run last time and Hank is running this time? Or is Theo is real name and Hank his nickname??? Confused!

  46. Sorry to tease you Antony. As you know I never actually watch election night TV because I am in a polling booth scrutineering until about 10pm, and then at a party which is either a celebration or a wake, depending, but in any case too noisy to hear the TV.

  47. Thank you Antony –
    you do a fantastic job. Although I am waiting on election night for you to say that there is an ever -increasing chance that the No.2 Greens candidate is about to cross the line on preferences, having seen RDN romp it in on primaries. Given what Steven Fielding pulled off last time…
    (relax Adam – I’m joking).

    And what happened to PB?
    I kept getting some scary message telling me there had been a fatal thingy. I was starting to think that Helen Coonan had realised wht was happening on her broadband and had put a stop to such subversion.
    J.

  49. Bludgertariat,

    Be nice to Antony, he is a psephelogical maestro and we are lucky to have him visit this little site (0% irony intended).

    Adam, a good craftsman never blames his tools (be they predictive psephelogical computer programs or otherwise). I think we should respect Antony’s right to spend the whole election-cycle looking for/working on new software but still use his own judgement on the night.

    Antony’s analysis of the major party’s experts spinning on the night is excellent and will be well worth remembering on Saturday night. (That is, whatever’s happening, most of it is BS until after the concession speech – sorry for summarising so bluntly and succintly).

Comments are closed.

Comments Page 1 of 20
1 2 … 20