Sky News reveals tomorrow’s Newspoll will show Labor leading 54-46; primary vote Labor 46 per cent (down two points), Coalition 41 per cent (up one). Details to follow.
UPDATE: Preliminary article at The Australian.
UPDATE 2: The Australian’s graphic here.
Copyright © 2008 Private Media Pty Ltd, Publishers of Crikey.com.au. All Rights Reserved.
Independent news, blogs and commentary on politics, media and business.
Editor: Jonathan Green.
Publishers: Eric Beecher, Diana Gribble.
Level 7, 22 William St, Melbourne, 3000. Ph: 1800 985 502 Fax: (03) 8623 9975
697 Comments
Pages: « 1 … 5 6 [7] 8 9 … 14 » Show All
Andrew, getten sie real, kameraden: the chance of the LNP winning with 48% 2PP are virtually nil.
Won’t get interesting ‘98 style unless the LNP gets 48.5 at least; and in most realistic sceanarios they’d need 49+.
297
I couldn’t help it when i pictured the image of them trying to “cut through” in door knocking back packer hostels
BBD – The Power at Work stuff is dangerous propaganda designed to further the careers of some ACTU apparatchiks and hangers on. If you are at 16% in WA your finished sad to say!
Just look at the annual returns on the AIRC website. Rudd aint going to die in a ditch defending unions afterwards – he will be looking to grow his majority in QLD in 2010. Get out whilst you still can!
” Federal elections do not result in 56+ outcomes. It just doesn’t happen.”
That is a description of history not an observation of the present. A historically rare occurance yes – but with all polls agreeing [as they didn't last election] saying this is 54% – a self destructing government with something like proof they had plans to extend WorkChoices as came out todays – makes 54-55 a genuine possibility.
And Possum’s model using historical data suggests 55% and Jackman’s model suggests 54%. THUS 54.5% is certainly a real possibility especially as undecideds break 2/1 with the trend – which I gather is 67% of 10% undecideds will vote Rudd. [In 2001 it was about 59/41 break to Labor]
So regardless of history, the world keeps changing, this is a real possibility, not a certainty, a possibility and shouldn’t be ruled out because it hasn’t happened for Labor before.
No 301
Never fear, we’ll romp home with 60% of the vote, condemning the ALP to another decade in the wilderness.
Lefty E
Because the polls are taken over a time period where the thing being measured changes (or at least we assume some people change their votes) you can’t say that.
It’s not like flipping a coin where the more trials you do the more you can be sure of the number you’re getting.
Nevertheless the consistency of the polls all year would SUGGEST the true value lies around 54-55%, unless there is a sampling/weighting bias. Ok now the true psephs can jump in.
Mad Cow: I was bored at work, and since I’m a programmer I though WTF and I should do it. Slowly building it up over the night. i want to break it down per state swings. Also I’m making the bold assumption that the swing is spread via a normal distribution with the STD as 3%.
It gives results what we see by Antony’s calculator but can give us a better idea of what seats are in play and what’s the probability of them falling. Of course it doesn’t take in to affect that seats like Bennelong and Wenworth won’t fall as much, but we can see that they’re not 100% likely to fall anyway.
Michael @ 39
Re how preferences are allocated to produce 2PP estimates
As BBD says, Morgan does both methods of allocating preferences – asking people who their preference goes for, and calculating two series using that and 2004 election behaviour. More often thannot, they give the same result rounded, as Morgan does, to half a percentage point.
On average, though, the “what voters say” method produces a result a bit under 0.4 percentage points higher for ALP than does the “2004 election” method.
A priori, it’s not clear wich would be more accurate. Some people may be clear on who they’ll give their first preference to, but unsure after that. Others may be clear on whether they want to vote for or against Howard, but unsure at the time of polling whether to give their first preference to, say, Labor or the Greens.
I think the more reliable method is probably the one that produces the less variation in the 2PP vote from week to week. So I looked at the standard deviation of these two estimates over the past 3 years. The results were:
2PP based on what voters say – std dev = 3.98
2PP based on 2004 election behaviour – std dev = 4.09
On that basis, I’d say using what voters say is slightly more reliable, but not by much. We probably won’t be ably to test it on election day because the MOE of the polls is quite a lot more than the effect of choice of preference allocation.
Note that Newspoll normally allocates preferences using a simple version of the “2004 election behaviour” method (I’m not sure of they’ve switched during the campaign) whereas Nielsen uses “what voters say”, asking:
“[IF MINOR PARTY IN 1a OR 1b] At a Federal Election you will be required to vote
for all candidates in your electorate in order of preference. Given this, will you
give a higher preference to the Labor Party candidate or the Liberal/National
Party candidate?”.
Left E,
I’ve wondered that too. The MOE discussed here tend to be worked out assuming a random sample from a normal or binomial distribution. But the Polling organisations would ask lots of other questions such as age, education, income, etc etc which they could use to tighten up the MOE.
Anyone here know how the pollers used background info to adjust the raw numbers?
Mad cow – have been hunting around to see if I can find accurate numbers; I think there are between 750K-900K of them, and the voting turnout has been very poor in the past, because a lot of people thought that you paid double tax if you remained registered here. AEC has cleared up that misapprehension. Will keep hunting and get back to you. Found one story at http://www.brisbanetimes.com.au/news/national/expats-hide-so-labor-seeks/2007/09/30/1191090943392.html but it’s a bit old and vague on numbers.
Thats what Im getting at Matt – applying standard MOE to a consistent poll average figure across time would surely be un- or less necessary (as opposed to any one poll)?
What! Why only 60%. Doesn’t Howard just have to fart in the general direction of voters and they all rush over to vote for him?
ESJ
I told you three times now ALP will NOT save the Unions and you keep putting it back to me like an arguement- I AGREE!
Unions have the power to save themselves- do you think they came into existence by some parliamentary act or something?
Learn the history. If you dont like unions dont join them. If you really despise everything they stand for- dont take annual leave, superanuation, dont use medicare, dont wear safety equptment, whatever makes you feel less stained by unions.
In 1998 it was 53/47 to Labor three days before the election. I’m waiting until the last few days before I get an idea whether we’ve got a.) A close election (marginals) or b.) landslide.
Left e
I think what you’re talking about is the poll average as in Geoff Lamberts post. Probably why no ones talking about a 51 whenever Labor poll 54. The poll average is clearly approx 55. In my humble and half arsed opinion you’re right – the moe range can be narrowed down..
Bring back values (and morals) – honesty, decenty, truth, respect, hard work, fair deal, hand-in-hand, two way street, win-win, be firm but fair, give the boss some reassurance that you are looking after the interests of all concerned. Bring happiness and joy to the people you care about, be happy with yourself, stick it out, through tough times, be strong and stick to your beliefs. Have conviction.
The numbers wouldn’t tell the full story, because it isn’t compulsory. So you’d have to figure out how many are overseas, then how many are registered, then how many are actually going to vote.
307 will, did you manage to solve the waterbed problem? (push down on one spot and the rest goes up a bit)
“Australians unlikely to vote for change, expatriates say…”
The Gulf News today carries this cross-section of two(!) ex pats giving their views on how Australians will vote next Saturday.
One, who works for a security company, is “not a traditional Liberal” but strongly supports Howard this time.
The other, who is a director of a marketing company, describes himself as “traditionally a Labor voter” but strongly supports Howard this time.
The cross-section of two is therefore universal!
There you have it directly from GG’s counterpart in Dubai. These two swinging or Labor leaning voters are determined that Howard should be returned for another three years. And here I was all year thinking that swinging voters were moving to Rudd. Guess I’ve been in some parallel universe.
http://www.gulf-news.com/yoursay/main_story/10168710.html
I love the people who predict it will go down to the wire. I have listened to this negativity all year… to people saying how silly it is to think that Rudd can keep it going for so long. Now they are out of time and this great equalisation has to occur on the actual voting day, because the average joe will on the day just freak out and cast one for the rodent out of nothing else but sheer trepidation.
I don’t buy it, and guess what? That can go two ways… Where is it written that the Coalition will claw anything back? How about we look at the flip side, Labor could claw back even more seats. When the swing is on it is on.
Sure, people may scoff at that, but none of us can predict the future, the pessimists and realist would have never predicted we are were we are today a year ago, so who is to say they are right about a close result?
Grooski,
I respect your point of view, but would you please stop bagging out all the other posters, nearly all of whom contribute something worthwhile in terms of understanding various qualitative issues. As I don’t have access to Textores or Hugh McKay’s focus groups, this is the best I’m going to get during the campaign.
Obviously, Antony Green, Possum, Jackman are pseph people who don’t share your views about historical analysis. They apparently think that such a large number of polls over many months is predictive, and this Newspoll fits in with the 53-55 range over the past few weeks. In fact, Hugh McKay is adamant that people have already made up their minds by this time with a negligible percentage of undecideds or soft voters still out there.
Sorry to disappoint you, Grooski, but if you’re so miffed by all the rest of our posts, please just ignore us. I certainly don’t intend to read another of your posts if I can help it.
I like Adam’s analysis. He’s gone all Greek with his hubri and nemesi:
http://psephos.adam-carr.net/countries/a/australia/commentary4.shtml
Btw, Sportingbet has the Greens favourite to take the 2nd senate seat in ACT now.
Simon
You quote one poll from 98. Unlike now there was much more volatility in the polls back then.
I’m getting together my music for next Saturday night:
Bye, bye baby, goodbye (Col Joye)
The king has lost his crown (ABBA)
See you later, alligator (Bill Haley version)
Wipeout! (The Surfaris)
Goodbye-ee (WWI ?)
Chorus:
Goodbye-ee, goodbye-ee,
Wipe the tear, baby dear, from your eye-ee,
Tho’ it’s hard to part I know,
I’ll be tickled to death to go.
Don’t cry-ee, dont sigh-ee,
there’s a silver lining in the sky-ee,
Bonsoir, old thing, cheer-i-o, chin, chin,
Nah-poo, toodle-oo, Goodbye-eee
……
Sorry, I meant Portlandbet in post 323.
Your wrong on the history BBD
Unions couldnt organise in the 1890’s so they formed the ALP to get compulsory arbitration.
That has now been removed, you wont get it back with the ALP. Without that environment you cant survive.
People join unions out of belief, like churches people arent joiners anymore.
I am not saying that in a mean or partisan way but as fact.
Mad Cow: Poss has done the figures and there are sleepers there and uneven swinging averages of all poles that are missing local variability which will trend towards Labor e.g. youth vote. It is definitely the other way News Poll is at the lower end of MOE. Howard just put the final nail in his coffin with his uncompromising approach to Work Choices. Turnbull’s dropped his bundle, Abott’s crapping on himself, it is so close to a bun fight it doesn’t matter. Read the body language they know they are gone. Rock solid with increased margin. Get with the program fellas there ain’t no hope for Little Johnny. Oh well! cognitive dissonance is a wonderful thing.
Donation gratefully received from ESJ. Many thanks to all concerned.
Left E @ 301 – if my simulator is right, at 52/48TPP that would give the following chances: ALP: 89.1%, Hung parliament: 8.5%, LNP: 2.4%.
I’m loving this simulator. I think I better get back to it and create a UI over it some all you tragics can use it too.
My God, you Krudd-lickers are really setting yourselves up for a huge, shattering disappointment.
Did any of you see the interview with the PM and the Treasurer on Today Tonight? They were in great from – relaxed, jovial and confident. Beware…
” Federal elections do not result in 56+ outcomes. It just doesn’t happen.”
Actually, Curtin’s 1943 win is estimated as 55.8 2PP (I believe they didnt actually have pref system – but thats the estimate of what it represents in modern terms)
Special circumstances of course.
Bit of trivia: the ALP won *EVERY* federal seat in Curtin’s home state of WA in *BOTH* houses in 1943.
Thats right – every senate spot as well.
Geoff Lambert Says:
November 19th, 2007 at 11:02 pm
For the 2 Morgans,Galaxy, Newspoll and the AGB, the TPP after corrections for House Effects …
….
GL, I wonder what a “House Effect” is..?
Bob Green-Brown, Lateline. Will get there, this time.
Lefty E
Curtin won every seat in WA Senate because it was first past the post.
Hey guys,
The blog troopers are in full shrill!
We’ve got them on the outer over http://blogs.news.com.au/news/blogocracy/index.php/news/comments/weekend_talkback24/P180/
Give it a go folks.
[They were in great from - relaxed, jovial and confident.]………Costello was slumped in his chair thinking “what am I doing here with chucklenuts?”
Albert F – you cant “tighten” the sampling error by weighting for demographics. In fact, and weighting that is done actually increases the margin of error you end up with.
Because the pollsters use a mix of random sampling, quota sampling and apply some stratification weights to those results – the MoE’s given in the polls by the pollsters are actually , theoretically, less than the true statistical MoE.
However, so saying – polling is unusual and the MoE in practice appears to be roughly to what is given, and what is given is the MoE that would have occurred if only pure random sampling was undertaken. I keep finding political polling data to behave in ways that are quite unusual like that.
To @286 Lefty E :
Not being a stats head i think thats a interesting question.
The consistency of the polls over such a long time, with different polling companies, and their lack of sustained response to short term events, tells me that most people made up their minds months ago, (then switched off in sheer self defense) and are largely sticking to their decision. Even though the polls are independent events they are measuring the same thing at different times so changes (or lack of change) over time are significant. Rattus is in for a hiding of biblical proportions.
Thats my qualitative assessment. If there is any way of expressing it quantitatively (index of poll significance???) it will take brighter heads than mine.
I’d really like to see the ALP win this one, not just for my own satisfaction, but so that Gough lives long enough to see Howard lose his seat and make his concession speech. That seems appropriate to me somehow.
Wrong bloody link!!!
http://blogs.news.com.au/news/blogocracy/index.php/news/comments/this_election_is_so_last_week/
331
Steven Kaye Says:
November 19th, 2007 at 11:17 pm
“My God, you Krudd-lickers are really setting yourselves up for a huge, shattering disappointment. Did any of you see the interview with the PM and the Treasurer on Today Tonight? They were in great from – relaxed, jovial and confident. Beware…”
……
I saw them SK…. never seen anything so stilted, clearly staged and weirdly-comically- pointless… but I will beware. I’d hate these guys to be in charge of anything important…..
Soundtrack for Saturday’s Liberal Party win?
We are all bourgeois now – McCarthy
“There’s something wrong somewhere here,
so through unclean streets –
I made my way
With holes in my shoes,
and my children asleep at my feet
I paid my way
In every town on the way,
the people looked grey,
the buildings looked healthy
Then one day I met a man
with money to spare,
he said he would tell me
How it is:
“The state-” he began
“-has been propping up people too long.
for far too long.
“We all got lazy and couldn’t be bothered
to raise ourselves in the world.
But we are all bourgeois now –
once there was class-war,
but not any longer
because, baby,
we are all bourgeois now,
so go out and make
your way in the world –
you’re free to choose!
We’re all free to choose!”
Etc…
Steven K – beware of what? Two losers, one past it, the other unelectable, who hate each other?
Now, get back to working on that CV.
331 – You’re a sick puppy if you think that was confident and relaxed. They looked like they were confidently and relaxedly contemplating mutual seppuku. Which is pretty much the story, really.
Mad Cow: Nah, simple normal distribution of swing based around an average, and apply that to the 150 seats, then I repeat the whole process for a large number of times. If I get it to work via States, then that would probably the closest I would get to doing that, since Adam and Antony have mentioned that within states the swings are less spread.
Big Blind Dave,
“Accountants of the world unite!
We have nothing to lose but our….”
Suggestions please.
Jackie Kelly said in the Sun-Herald yesterday.
“A friend in parliament gave me his best line for dealing with more troublesome constituents – Sir I am paid $100,000 to look after 100,000 people, here is $3 come and see me after the next election”
There will be many going to see a different MP after this election.
The Australian people need to make a very clear decision with the election result. The country has been greatly divided by John Howard and if the result is very close either way, the country will be even further divided. I see a very sad future for Australia both environmentally and socially. The Religious Right and their infiltration of both major parties is the BIGGEST concern , but the electorate are not aware with what is happening. Individual freedoms and rights are greatly at risk and with The House of Unrepresentatives (not elected by proportional representation ) , there are thousands of citizens with no one to represent their views in the parliament.
Will, a N=100000 Monte Carlo? Wow, how long did that take?
cardigans
Pages: « 1 … 5 6 [7] 8 9 … 14 » Show All