Philosophical Conversations Public Opinion Junk for code
parliament house.gif
Think Tanks
Oz Blogs
Economic Blogs
Foreign Policy Blogs
International Blogs
Media Blogs
South Australian Weblogs
Economic Resources
Environment Links
Political Resources
South Australian Links
"...public opinion deserves to be respected as well as despised" G.W.F. Hegel, 'Philosophy of Right'

a cautionary tale « Previous | |Next »
November 17, 2007

It's going to go down to the wire in the local campaigns in the marginal seats and be determined by the swinging voter. It's old news, I know. But it does look as if there is a chance that Labor may get over the line, as the Liberals are not getting traction.

I do not think that Labor is going to romp home, as the polls imply, since about 25 per cent of people make up their minds in the last week and 10 per cent on the day as they go into the polling booth. I know, that too is old news.

Geoff Pryor

I've switched off listening to Labor. As John Hewson pointed out on Lateline:

Rudd is really from central casting. They've done all the focus group research, got the lines right and been very disciplined - I agree - in delivering the lines. They've got five dot points and it doesn't matter what you ask, you get the five dot points. And they keep doing it and they've been very disciplined.

What they have achieved is that Labor will get a dozen of the 16 seats it needs. It's the last 4 that will be difficult.

As Rod Cameron pointed out on Lateline:

Labor has to get a record preferred vote and has to get swings of 6, 7 or 8 per cent in either Queensland or Victoria or NSW to win this election and it's only just sitting there at the moment at that level....The polls are showing that Labor is sitting on the swing that it needs to win election, a bit above it. Labor's going to need 52 per cent of the preferred vote to win. It's a smidgin above that at the moment and, you know, one week can knock a per cent off easily.And there's a certain gravitational settling here. I mean, Labor to be polling at such stratospheric levels is just not going to happen. There will be just normal settling, even if Labor has a good week. There will be just people who will come back to the fold.

The numbers are very hard, particularly if Labor doesn't win the bottom 16 seats and has to go to higher margins to get them, or loses a seat or two as seems likely in WA. So how much is the drift back to the Liberals in the key margginal seats in percentage terms?

| Posted by Gary Sauer-Thompson at 2:47 PM | | Comments (19)


The editorial in the Weekend Australian says:

THE whistle has not yet blown but, given the results of today's Newspoll of 18 marginal seats, only an unprecedented rush of support back to the Coalition in the final week will save John Howard from defeat. The Coalition strategy of defending marginal seats in the hope of restricting their losses to fewer than 16 seats appears to have failed. After four terms, John Howard may ultimately be judged to have stayed too long.

They say that the prospect of defeat allows us to reflect on Howard's more strategic mistakes and after naming them add that during periods of economic bounty such as these, elections are for governments to lose rather than for Oppositions to win and Howard's strategic failings are clearly coming home to roost.
Sounds like The Australian has been mugged by reality.

I differ in one area, I suspect they'll gain 14 seats for a total of 74 with the rest needing to be worked for.

Its a fair analogy Gary.

The possibility of picking up a couple in W.A has thrown a spanner in the works a bit.

With so much campaigning in Queensland there may be an air out there from the lower states that they will be ruled by Queenslanders. That may turn some off.

As I said early there is not much cheering out there for Rudd as he campaigns. Its as if people don't really like him but the poll numbers reflect they like Howard less and I wonder if some of your mood feelings are perhaps a little Ho Humness coming from the electorate. Ho Hum we are sick of it. Get it out of the way of Xmas.

I heard a report on the radio of the possibility of the coalition picking up a seat in S.A. But I cant see one. Do you?

16 - 18 seats is unattainable I feel.


Even if the figures quoted are right about the proportion of undecided voters, a proposition that might not be right on this occasion, the same literature also suggests that in most cases it breaks in the same proportion as the existing proportions of decided voters - that is it is unlikely to significantly affect the relative standing of the parties.

The Coalition will lose 3 seats--Kingston, Wakefield and Makin-- and gain none in SA. What is in doubt is whether the 3 becomes 4 --Boothby. Sturt still appears to be defended.

Okay, the crunch is not 12 or 14. It's the last 2 or 4 that is going to be hard. The Coalition has only to hold onto 1 of the 16 to retain power. It's that last seat where the battle will be decided. I wonder which seat it will be?

There has been a massive shift in the market today with centerbet.
Earlier Labor was showing @1.38 and Coalition @ 3.50

Now it is Labor @ 1.17 and Coalition @5.10

Either there was a plunge today in the range of 2-10 million on Rudd or there is some insider knowledge involving Health issues/Scandal.
The polls have some information that Rudd wont have enough seats in Qld.
What to make of it?
Take the 5.10 with an equal bet on Natalie to win Idol @ $3 and hope that neither one has had a heart attack!

Possum Pollytics agrees with you:

the “soft vote”, that bottom draw cliché that’s been dragged one too many times around the commentariat’s back paddock of late, does exist even if it’s a magnitude of order less than some would have us believe. The polling on this has been telling us all year, from the Crosby Textor Oztrack 33 through to Newspoll and ACNielsen qualitivate polling that the “soft vote” is relatively small, but it’s also evenly split - perhaps with even a slight tilt toward the Coalition.

I'm in no position to judge. Possum says that the message is clear - the game is over--its the ALP for sure-- then adds that is what makes it so dangerous for the Coalition:
And here lies the danger. If Coalition members lose the plot, if recrimination starts creeping into the last week of the campaign as government members fail to come to terms with the electoral reality, if members become more concerned with saving their own skin now that the fight to save the government is lost - it could all get very ugly very, very quickly.

For the Coalition the story is stark - the difference between 45 and 43 is a functioning party. Gaining less than 30% of the seats in the House, the parliamentary consequence of losing their own soft vote, will repudiate the Howard years, the Howard agenda and what the Liberal Party has become under his stewardship.

this is not 'the decline and fall of the roman empire'.

relax, one side of pollies may be turfed out, another side gets in the saddle. some horses will be better off, others worse.

but horses still, and no amount of chatter can disguise that. and horses are ridden for the benefit of riders, not horses. national policy will continue to be set for electoral advantage.

The Coalition's last bastion of defence is WA. It's where the political miracle might happen as the ALP is nervous about Cowan and the Liberals reckon they will hang onto Hasluck and Stirling.

The Liberals reckon good local members will save them from them being overwhelmed by the Rudd onslaught. Will that be enough?

My guess in Qld.Liberals lose 4---Bonner, Moreton, Blair, Herbert and maybe 6--Leichardt and Forde. Flynn is a hopeful.

What's your reading?

my guess at what Gary and Vee are calling ALP 'in the bag' or likely.

Tasmania 1
SA 3
NSW 4-5
Qld 4
WA none; lose one (Cowan). It's a liberal stronghold.
NT maybe 1, but I don't know enough

That's 12 or 13. I'm not sure where the next 3-4 are.Will it be Victoria? Corangamite, Deakin , McEwan? I guess we have to watch Howard and Rudd's movements.

I wouldn't give up any seats in Qld at this stage. Some are too close to call. The undecided still has a too high percentage and I am expecting most of them to go to the Liberals. I don't see the public has been turned off the Nationals.
My earlier prediction of 6-10 falling across Australia I will stick to for now.
Howard will hold Bennelong.

Peter S,

I'm calling 74-87 ALP seats, I could stretch as far as 90 but its no sure thing the ALP will win. Ideally I would like to see a hung parliament, there is scope for that but it is unlikely.

I see that you hold the 'surge' theory.

What's your breakdown state by state? Are you in a position to do that? Where does it differ from my more cautious breakdown? Victoria?


The Newspoll marginals poll has Labor walking it in, but the Galaxy marginals poll has it much closer. Confusing.

They've both polled thousands of people, but across so many seats the margin of error gets too big. Phone polls only call landlines and miss the youth vote.

Possum's warning was based on the Newspoll which, if it's anywhere near right, will see the coalition implode with even Costello in trouble.

The shift in the betting market Les points out is also shifting in the marginals in the same direction, and so is Sol Lebovic.

Then you have the added confusion when pollsters and commentators have their own reasons for doing things the way they do. And they're modelling the TPP on 2004 preference flows which won't necessarily hold this time around. And much of the received wisdom is based on 2004 demographic distributions which have also changed.

Peter S,

I am uncertain to what the 'surge' theory is. No I am not in a position to do a State by State breakdown.

My guess is that the 'surge theory' is that the ALP surges past the Coalition's marginal set defences, which are based around the strengths of the local candidates. All politics is local etc. I guess that it is a bit like the incoming tide and a line of sandcastles.

It's what the national polls show, and many in the ALP hope for. I'm not willing to grant it one week out.

More problems with the polls:

The surge theory assumes uniform swing. We already know it's not uniform across states, seats or even booths within seats, however it does hold for some demographics.

The two marginal seat polls both have fat margins of error on a seat by seat basis and were conducted at weird times which account for neither launch, the Howard launch but not the Rudd launch, or the Howard launch and either during or immediately after the Rudd launch. TV coverage is important, as Gary's other post today points out, so what people made of the Rudd launch will have some impact on late deciders.

Rudd and Howard are not the only ones having an effect here. What if some campaigner in the last crucial seat has halitosis and a flatulence problem, and knocks on every door in the electorate over the coming week?

And Howard has just discovered the last remaining minority he hasn't picked on yet. How do the people in marginal electorates feel about junkies? How many voting junkies live in marginal electorates?