Morgan: 60-40

Roy Morgan has caused a disturbance in the force by releasing its weekly poll a day early, perhaps hoping to strike while the Strippergate iron is lukewarm. Conducted over the last two evenings, the phone poll of 633 voters shows an intriguingly strong result for Labor, whose lead has widened to 60-40 from 58.5-41.5 at the last comparable poll a fortnight ago. Separately published leadership approval figures are remarkable enough to raise suspicions about the sample: the Prime Minister’s approval rating is at its lowest level since he was elected, his disapproval rating (up nine to 51 per cent) shooting past his approval rating (down nine to 44 per cent), while Kevin Rudd’s approval rating has reached a Bob Hawkeian 74 per cent (up 2 per cent). Rudd’s lead as preferred prime minister has widened from 51-44 to 52-38, and 87 per cent of respondents profess them unconcerned about the incident in New York.

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.

336 comments on “Morgan: 60-40”

Comments Page 7 of 7
1 6 7
  1. Paul, it is a shame that the only pro-Howard contributors also tend to be wind-up merchants who spout a whole heap of lies and nonsense either for attention or to derail debate. This includes Glen, Steven Kaye, Nostrodamus, and Cerdic. The only solution is to ignore them or at least not address them directly.

    Where are the rational, reasonable-minded Howard supporters? Or is it that rational, reasonable-minded people no longer support Howard?

  2. Paul, i dont think it needs to be.

    Really nobody on here should need to barack for a side. We should all be electoral nerds 😛 and fuss over if movements are significant and if question wording technique is influencing the polls. There are plently of places to support a political party on the net, here we can barack for the various pollsters.

    On a completely random thought, id like to see a great polling debate. 4 teams of regular posters on here have a structured debate on which polling company is the best. It think it would be really interesting, although i got a feeling possum would win by producing graphs and crazy formulas that the rest of us have to assume are correct 😛 (i’m sure they are)

  3. # 277 – Victoria 1999 – in particular the election-morning Newspoll was very close (50-50 or 51-49 if I remember rightly). The biggest hint to me that something serious was on was finding out (thanks to a friend who lived in the area) that there had been a last-minute Liberal campaign blitz in a regional seat on a 9% margin, something which you’d only expect if someone’s internal polling was telling them something. (Didn’t help them – they lost that seat anyway).

  4. According to newspoll on September 18, the day of the election Primary votes ran 46 – 46. actual result was ALP 45.6 – Coal 47.0

  5. Pollfight! Pollfight!

    Seriously though I don’t have a favourite pollster

    I think Morgan is the best “first impression” pollster.I have a theory (work in practice 😉 ) that Morgan picks up how issues play out in the electorate as a first impression.

    I think Newspoll is the best at picking up how much of that first impression gets translated into moderately serious voter movement.

    ACN seems to be the best at producing a longer term conservative estimation of voter change.

    Galaxy doesnt produce enough polls.

  6. Timbo,

    That’s interesting as I remember we all ignored the polls becasue all the media was saying that it would be Kennett easily.

  7. History shows us that we are now entering a phase of a polling fog. Polling seems to become more volatile in the lead up to the election, particularly for the Coalition’s vote. We can less certain that changes in polling numbers from now on, represent real swings or just election static.

    For polling junkies, it will be like waiting for the astronauts to re-enter the atmosphere from orbit; there’s a period of radio silence.  We’ve got a pretty good idea that they are OK; but we can’t be sure.  That’s what happens with campaign polling.

    Interested parties can have a look here.

  8. Possum (or anyone), is there anywhere on the web I can find any detail about the methodologies of Australian polling organisations?
    How do they choose from within an electorate (when booths can vary so much in party support).
    How closely do they stick to proportions of the demographic. If 18-24 year olds are low on numbers do they weight with what they’ve already got?
    Are phone numbers called always landlines?
    Why is the Nats number always so low?
    Can anyone point me to any info?

    I guess they’d all keep things close to their chest, but there must be some conventional wisdom from which to build their own particular polling methods.
    Their confidence (MoE) levels match according to sample size, but results vary more widely than that too often. This suggests to me some fundamental variations in their methods.
    I am talking about fair dinkum polling. Not the likes of this:
    I was polled by Galaxy about the NSW state election in March. A question BEFORE being asked who I would vote for included the option “Is this the worst government NSW has ever had”. Pretty dodgy!

  9. Ha! Beedle that pre-poll question is priceless. I’m sure we could all come up with a heap of questions to bias a poll against the Howard Government.

  10. Just because Nostro Steven Cedric and myself hold differing views from you doesnt mean what we say are lies or spin…by your logic i could claim Pi and STROP are just as bad if not worse than myself and others…

    You’re all so used to everybody agreeing with your left wing views that you cannot tolerate any opposing views…how are we suppose to have an honest debate on policy when whenever you make a post you rail about how myself and others are liars or off with the faries ect…

    My comment to make is dont count the chickens just yet because you could get a bad surprise come November if things dont go your way…regardless of how well Labor are doing in the polls atm 16 seats would be a tough assignment for anybody let alone a bloke whose been the leader of the opposition for less than 12months…

    If you want an honest debate try acting civil towards those with differing opinions try it and you may get some good debates happening here.

  11. ” Polling seems to become more volatile in the lead up to the election ”

    I would agree with this statement in regards to TPP but Labor’s Primary seems fairly solid to me. I get teh feeling talking to people who have switched from voting Howard in 2004 to Rudd in 2007 is that they’ve made up their minds and they’re not listening to the government anymore.

  12. ” If you want an honest debate try acting civil ”

    This from the guy who labels anyone who disagrees with him a leftard, leftie, Communist, etc, etc. And tell him you voted Howard last time and want to vote Rudd this time and he’ll call you an outright liar who always voted for labor. Sounds like the pot calling the kettle black to me.

  13. Glen good to have you back, don’t let those lefty bullies upset you.
    Lets debate the Wickapediagate scandle.
    What do you know and when did you know it?
    I heard about it today and am out raged that Dept. 0f Defence is involved in Politics and rewriting history. Sounds like something out of George Orwells “1984”.
    Wasn’t that a novel not a manual?

  14. Glen I see where you’re coming from and its true that the Coalition are still more likely to be returned than not.

    However, it would do you much good if you at least tried to be a bit more objective and a little less delusional sometimes. Any feasible debate on policy relies on people being less tied to a particular party. I tend to ignore posts on here that are those of obvious party members who will never disagree with the policy of a party.

    I don’t see a single post from a Howard supporter on here where I even think for a second they aren’t a Liberal party member. Even then I’ve spoken with Liberal Party members who are critical of moves made of Howard. You never see it on here.

    Finally, if you throw lines in like this: “regardless of how well Labor are doing in the polls atm 16 seats would be a tough assignment for anybody let alone a bloke whose been the leader of the opposition for less than 12months…” you have to expect a bad response.

    How long had Bob Hawke been Opposition Leader for before he won the election?

    Acceptance that there’s a problem is the first step to recovering from it.

  15. Beedle, most polling organisations are the same in one respect, in that they use a random telephone poll of households to get their raw data.(Morgan face to face polls are different of course)

    Where the telephone pollsters differ is in what they do after that.Once they get a random sample of households made up of various respondents of various demographics, they then use their own weighting mechanisms to make the numbers of respondents in each particular demographic reflect what they believe is the actual population they are attempting to measure.

    It’s in those weighting mechanisms that they differ and the way they treat their data.I think all pollsters keep their weighting magic secret, but some like Galaxy I think keep it super duper secret.The weightings are the key to accuracy… if you have a better weighting system than other pollsters, you’ll be more accurate than other pollsters.

  16. I think we should be ready to call this election for the Coalition. The turning point has come.

    Tony Abbot has just announced the regulation of solariums.

    Nice try KRudd… it’s all over now.

  17. ” the regulation of solariums ”

    How dare Abbott try to tell honest business people how to run their businesseses. The low down socialist. He obviously a Commie at heart.

  18. Thanks, Possum. You’re a trooper.
    You have filled in some of the many blanks in my understanding.

    So if it’s likely that phone polls are random and not targeted to electorates, then I can see large errors occurring with weightings if a certain demographic or electorate type is under or over represented. This might be a chronic problem with 18/24 year olds if landlines are used exclusively for phone polling.
    But there does seem to be a general similarity across many polls and polling companies. No one is saying Lib 53/Lab 47 for example (Maybe Galaxy next Monday!!)

    Given this, is there any way of knowing who is more “accurate”? Surely this can’t ever be known for sure. The accuracy of voting intention polling seems to be measured by the closeness of their polls to actual voting on election day. Just because Newspoll has it 55/45 now and the elections turns out to be 50/50 doesn’t make Newspoll wrong because 55/45 could have been right at the time but there was no way of knowing it back then.
    Morgan’s high Labor numbers might all be true!!

  19. Beedle, the weighting mechanisms used can take into account under representation and more.

    Let’s say for instance that a polling company knows that 15% of 18-25 year olds dont have a landline.And let us also say that through qualitative research they know those that people that dont have a landline are 10% less likely to vote for the coalition, 5% more likely to vote for the greens and 5% more likely to vote for the ALP.

    They can then use the original sample they get from their landline polling and weight it to reflect not only their estimate of the true proportion of 18-25s in the country that are enrolled to vote, but also account for the 15% of people that dont have landlines.

    The more a pollster understands his samples and understands their relationship top the true population, the more the weighting mechanism can reflect that to produce more accurate polls.

    The fact that all of the pollsters are generally in the same ballpark (Westpoll excluded) suggests that they all have a fairly accurate weighting system (or alternatively, are all completely wrong in exactly the same way;-) ).

    We dont have enough elections to keep a close track on who is the best (for that we can probably all be thankful 🙂 ) but we do know that on primary votes, all of the pollsters nearly all of the time get their last poll before the election within their margin of error to the actual primary vote election result.

    Extrapolating up from the primary vote to the TPP though does a lot of pollsters in some times.

  20. The Morgan phone poll was accurate on primary vote for the last Vic, Qld and Fed elections. If this one’s anywhere near the ballpark, it’s bye, bye Howard and company, with a HUGE Labor landslide looming.

  21. im sorry but i cant understand why it has to be such a big dama.
    if 10 people are polled or 10.000 they give there answers and then the pollsters calculate the percentages.after that why have such a big hoha about bias ,moe etc. this is what they polled this is the result full stop. or is it a case of any figuers can be made to say what you want them to say ?

  22. I am suprised at the lack of polls coming from Queensland considering the importance of all those semi marginal seats. Any one knows what is going on up there?

  23. 328
    nath Says:
    August 24th, 2007 at 5:11 pm

    I am suprised at the lack of polls coming from Queensland considering the importance of all those semi marginal seats. Any one knows what is going on up there?

    At the moment , Nath, not much happening up here. There have been a few polls in seats affected by the ‘amalgamations’ issue published in the Courier Mail and articles claiming it is ‘hurting’ in those seats.

    It was an error of judgment on the part of the Premier to think the issue would ‘hurt me’ (the State Labor Party at the next State Election) and not ‘hurt’ Rudd and Labor at the Federal election.

    As things stand, I wouldn’t be counting on more that 5 seats going to Labor up here at the present time, but the election is some way off yet and I THINK the jury is still out on how the most State parochial place in Australia will react to to Howard sticking his nose in.

    Some say the amalgamations issue will keep some voters with the Coalition because of it and there is an equally strong argument that others will not let this issue decide for them, they will be annoyed at Howard for his big show about it, but will let Federal ‘big picture’ issues decide who they vote for.

    I think it will cost Labor seats it might otherwise have ‘easily’ won and for no good reason – Mind you, I think there is merit in the argument that the State Government should be free to do as they see fit and be accountable to the STATE at a STATE election if the electorate think it sux.

  24. Labor now need a minimum of 17 seats gain to get a bare majority – remember that they expelled Harry Quick in Franklin, Tasmania and that must now be counted as a non-Labor seat. They are no guarantee by any means to retrieve it.

  25. Off topic, sure – but Nicola Roxon is actually holding up well to an interview for the first ever time! Lateline… She’s whooping Christopher Pyne for six. (no probs there – my cat would would him for six too, but…) Roxon 1, Pyne 0

  26. Nostro,

    If you’re going to use that sort of logic why don’t you just double the 16 and say that Labor needs 32 seats to win or even triple it to 48 seats. Makes as much sense as the dripple you spout.

  27. For John Rocket cmt 320

    The reason News Ltd has two polls is that the papers can differentiate between The Australian (with its long association with Newspoll) and the rest of the stable which decided to go with Galaxy at the 2004 so that it could market its “own” poll against that of The Australian.

    The rest of the News Ltd stable pays for Galaxy, the Oz looks after Newspoll.

    Sort of why Coca-Cola Amatil sells Coke and every other variety of soft drink and water that you can think of – market differentiation.

Comments are closed.

Comments Page 7 of 7
1 6 7