Via The Oz and Newspoll comes some specific Qld marginal seat polling undertaken in the six electorates of Dawson, Flynn, Bowman, Dickson, Longman and Herbert. The average two party preferred in these marginals was given as 54/46 to the ALP.
That equates to an average swing of 2.7% to Labor in these six electorates (The Oz has it at 2.9 but that’s not what I get), which should put the Qld wide swing at the moment something in the order of 3-4% or more. If a party receives a given swing at an election, that swing is usually smallest in their own safe seats, largest in the safe seats of their opposition, with the size of the swing in the marginal seats falling somewhere in the middle.
If we use the last election as an example and measure the different swings in certain categories of seats, it tells the story.
So a swing of 2.7% in the marginals in Qld should equate to a slightly larger Qld wide swing.
Morgan released a tidbit of info in their usual face to face polling release about a telephone poll they took over November 11/12 that had the two party preferred slashed to 52/48 from a small sample of 573. If we feed that into our Pollytrend charts, we get:
However, without having the primary vote details I can’t calculate the poll averages – so we’ll use the result for our trend charts but not the averages.
Over the weekend, the Greens also released a small bit of polling they commissioned by Galaxy on public perceptions of emissions trading. Rather than simply repeat what Andrew Norton has said about this – since I agree completely with him – I’ll simply point you in his direction.
16 Comments
The dip in the ALP TPP that started around mid-April (according to Pollytrend) was, I recall, not really precipitated by anything in particular, which is somewhat more ominous than a knee-jerk reaction to a topical issue (since it would indicate simply a Rudd fatigue setting it).
The even quicker reversal in June was of course all about the Ozwincaremailgrechgate affair. Could it be that this served to mask the original natural death of the Rudd honeymoon, which is instead being seen now thanks to some good old racial wedge politics?
Rightly or wrongly (hint: wrongly), Labor has no real counter to the Liberals on asylum seekers, so will have to regain support on some other issue. Climate Change? Not likely; the Labor position has little support from the left or the right, and the almost certain failure of Copenhagen will play into the Coalition’s hands (even if it will make Malcolm have to make some hard choices). Most of Labor campaign promises on health and education are not progressing and there will be no extra cash with which to pork-barrel once the recovery kicks in and the automatic austerity measures limiting spending set in.
Sure Rudd and Swan will say they saved us from financial oblivion, but that’s done and dusted now, yesterday’s news. What have you done for me lately? Is there anything of much substance behind the Rudd Government’s impressive public front? I’d like to think so, but I fear not. I think more and more Australians are coming to this view, although of course for some disappointing reasons.
Indeed, Rudd has a disillusionment problem. Just about everyone of the left persuasion that I know is disillusioned with Rudd’s social conservatism and general spinelessness on matters of principle in his bid to keep the centre. But none of those people would think of voting for the Libs in a million years, nor preferencing them over Labor.
The only real risk for Rudd is if the Libs tap into the general Rudd disillusionment by offering something attractive to the disillusioned. Not holding my breath on that one.
For all their faults, Rudd and Co are not yet anywhere near the problems of Labor in NSW, where diehard lefties are itching to vote for the Coalition just to rid their own party of its cancer.
Possum, I don’t agree with your first statement about where the swing occurs at all. I can point to many elections where the largest swing was in marginal seats. In 1996 the largest swings were in marginal seats, in 1998 the exact opposite, which was why the Howard government was re-elected with less than a majority of the 2PP vote.
Your 2007 figure for safe Coalition seats includes swags of seats like Dickson and Longman which were classed as safe simply because of Labor’s withdrawal from the contest in 2001 and 2004. I’d predict we will see much bigger swings at the 2010 election in marginal seats like McEwen, La Trobe and Macarthur than we will see in safe seats like Higgins and Wentworth.
You’d normally resort to regression of 2PP swing against 2PP margin on this subject, which I think is a more reliable measure than the categorisation of swing.
This is just the standard swing vs. margin regression divided into 3 chunks. Of course you can find examples of where it doesnt happen – both elections and specific seats particularly – but all things being equal, a slightly proportional manifestation of the swing is what we expect to see over any arbitrary number of elections.It’s also why we can predict fairly accurately the ultimate election result in terms of total number of seats changing hands from marginal seat exit polls (a small subset) – it’s effectively the same thing just done in a slightly more complicated manner.
Wouldn’t any discussion of size of swing in marginals HAVE to take into account the previous result in these seats? Flipped from one side to the other (from previous election) would contain some “sophomore surge” as opposed to those marginals which didn’t change hands. 1996 to 1998 comes to mind here as an example of many seats in this category – which happens only after a previous landslide in their favour.
Nipper, it would certainly help if you were trying to figure out what’s happening in any given, particular, seat.
Over the last 20 years I’ve produced enough of those table of swing by margin to disagree with you but can’t be bothered digging around for all the data. Overall swings are almost always under 5%, so swings in seats outside of the marginals don’t come into play even if there is a difference.
Possum, if you believe that’s how come you don’t include it in the monte carlo?
Extremly interesting poll results I think the Liberals would win NSW if an election were held now as they did in 1974
Antony, I agree they rarely make a difference. But next time you have the data available, don’t run a linear regression – look at something like a nearest neighbour style regression and look at the difference in the tails of each.
Tri$tan – it doesnt make any real difference in terms of total seats that would change hands. An extra 1% swing against a seat that’s sitting on a 15-20% margin is really neither here nor there in terms of result.
I love it when Antony and Possum fight, its so cute
Two point, Poss:
“two party preferred slashed to 52/48 from a small sample of 573″.
That number of respondents give plus or minus 4% with 95% certainty does it not?
So even assuming a low outlier, that maxes Labor support at 56% – consistent with the earlier bigger poll.
And speaking of outliers, doesn’t it add credence to my previous comments that Newspoll was not an outlier?
Rather, Labor got hammered over Asylum seekers until Rudd took to the airwaves – 14 interviews in 72 hours.
What Nielsen picked up was an element of the post-blitz recovery – but notably was still down on where Labor had been.
Still the point remains – all of the major polls are now reporting statistically significant losses from Labor’s peaks.
And the situation is more interesting given the usual over-estimation of the ALP vote in Morgan.
polling polling polling.. keep them pollsters polling.. raw data!
Peter, or the Newspoll triggered a bandwagon effect among the lumpen.
Really, anyone who entertains the merest thought of giving their primary vote to the Libs over Asylum Seekers is despicable. Giving the Libs your primary vote over anything is crazy enough, but over Asylum Seekers it goes beyond crazy.
test
As a Greens member, I’m disappointed with the quality of the Galaxy poll. Push polls like this one only hurts Greens’ credibility. If the Greens are going to start polling the public, I would’ve thought it’d be far more useful to probe where the public is willing to bend to meet emissions targets, then start prodding the government with the results.