Possums Pollytics

Politics, elections and piffle plinking

And Thus Spake Newspoll

Posted by Possum Comitatus on September 19, 2007

My humblest apologies for the delay on the Newspoll low-down; I’ve been organising a few things for the election and will have an announcement about that soon (insert mysterious music)

When the Newspoll rolled out on Monday night, I was expecting to wake up to headlines reading “Howard poised to take Denison” or some equally silly interpretation of polling noise. But no, journos and commentators across the land generally kept it in their pants.

As we’ve been saying for months, the polls are moving about 2 points around a long run mean of 56-57/43-44 on TPP and 47-48/38-39 on primaries. Adding in the final September poll to our monthly aggregate of Newspolls, we end up with:



Again – it’s just more of the same. If you remember back to the last Newspoll when the headline 59 figure came out, we said:

The headline numbers will undoubtedly have the politicians, commentators and pollyjunkies running around like headless chooks in varying states of euphoria or complete panic, depending on ones political bent.

But underneath these headline figures it is nothing more than business as usual.

Nothing has actually changed since March. The nature of polling series is such that they wander a few points around their true level simply because of the probabilistic nature of sampling that occurs with opinion polls.

But so saying, break out the popcorn because the fallout is going to be worth watching

And it really was all rather amusing over the last 2 weeks. They’re fickle, predictable little creatures those pollies.

There were a couple of other interesting bits in the Newspoll. The Coalition leadership question which suggested that Petulant Pete was about as popular as nappy rash and Malcolm Turnbull was about as popular as, well…. Malcolm Turnbull.

But the other bit was on the question of the strength of voter support:


What I haven’t seen picked up anywhere much around the place was talk on these figures (but so saying, I could well just have missed it).

The ALP is ahead by a large margin on primaries, yet these figures clearly suggest that the ALP vote is actually stronger than the Coalition vote.61% of the ALPs primary vote is rusted on while only 58% of the Coalitions vote is the same. The “probably will” voters have the Coalition one point in front 29 to 28.But the Coalition has soft support at 11 compared to the ALPs 10.

But to fully absorb the power of these figures, you have to place them in the context of the true primary vote being close to 10 points higher for the ALP.

So not only is the ALP vote higher, but it is at least as strong as the Coalition vote.

To highlight this, if the “maybe” voters for each party changed between now and the election, representing the best and worse case scenario for each party, the ALP primary would be reduced by 10% (about 5 points) from 47-48 to 42-43.However, if the soft voters for the Coalition all shifted, the Lib/Nat primary vote would fall 11% (or about 4 points)to around 34/35 *.

A scary prospect indeed for the conservatives if we take these numbers at face value, and assume that he who wins the election campaign will make the greatest inroads into their opponents soft vote.The best the Coalition can do if the “soft vote” deserts the ALP is to come roughly equal to the ALP on primaries – only to lose on preferences.


Which gets us back to a question I asked earlier – what happens if Rudd wins the campaign?

These numbers put some meat on the bones of that particular question.

The conclusion from the Newspoll is pretty simple – its business as usual on the primaries, business as usual on the TPP, Howard is the most popular Coalition leader by a powerful lack of alternatives and the Coalition vote is not only dismally small, but considering its size it’s also dismally weak.When your primary is hanging around 39 and you get the rough end of the pineapple on preference flows, 4 points is the difference between a post election result where there’s more party room members than shadow portfolios to divvy out, or close to the the reverse.

* Thanks to BB for pointing out an earlier mistake.

AddThis Social Bookmark Button


17 Responses to “And Thus Spake Newspoll”

  1. RobertBe said

    Well done once again Possum.

    You were dead on about the need for popcorn. It was indeed most entertaining. I had the good fortune to be on holiday in Noosa so I had the added benefit of watching (and eating my popcorn) under the shade of a palm tree.

    I loved the comment from one of the pundits in the Oz who said that watching Downer and Co front up to Howard was like watching a bunch of 7 year old girls poking a brownsnake. Lots of squealing and running away.

  2. RobertBe said

    My mistake. Credit for the 7 year old girls analogy belongs to Anabelle Crabb at the SMH.

  3. Martin B said

    Who the heck are the one percent who deliberately give an incorrect answer in regards to their voting intention, and then *admit it*? 🙂

  4. Bushfire Bill said

    Don’t quite get it Possum, re. the soft vote swap.

    If all 10% of Labor’s soft votes disappeared off their scorecard that would be 4.7% wouldn’t it? Leaving Labor at 42.3% primary.

    And if all 11% of the Coalition’s soft vote went west, that would be 4.5%… leaving their primary at 36.5%. No “2” in front of that number.

    A more interesting exercise might be if 3/11 of the Coalition’s soft vote (they get to keep 8/11) went to Labor and 7/10 of Labor’s soft vote (they only keep 3/10) went to the Coalition.

    This would leave the respective primaries at Labor 43.7 and Coalition at 39.7. My 2PP calculations (based on the 2004 election distribution) deliver a 2PP result of around Labor 53.7% to Coalition 46.2.

    Still a Labor win, although a closerish one than some anticipate. The Coalition has to keep most of it’s waverers, while Labor can shed quite a lot.

  5. Possum Comitatus said

    You’re right Bill, I’ve had a brain fart and forget how to count to 100.I better go and fix that up!

  6. disenfranchised Gippslander said

    I’ve been thinking along the same lines as Bushfire Bill, but my analysis of the “soft”vote for either party is a little different. I propose that the ~10% soft vote is made up of 2 components.. 1) voters whose vote depends on extraneous factors eg, is it raining, did they like the HTV personnel, did they get lucky last night,etc.
    2) those who are going to be affected by the campaign.
    Now group 1) are the true random portion. it is only possible to assume that the net swing will be zero for this group.
    group 2) are the ones whose change if any will be predominantly one directional.. ie they will tend to be to the coalition, or to the ALP or there’ll be litlle change. We have no way of knowing which of the 3 broad possibilties ..Pro coal, pro ALP or no change … is going to occur.. and any talk of probabilities is absurd.
    As to the relative size of the two groups, let’s call them the “dolts”, and the “thinkers”, we can say a little more. We know that the % of the combined dolts plus thinkers is about 10% of the population in each of the 3 elections surveyed. I have the impression that the movement in the campaign period has been about 2%. This would imply that the dolts are about 8%, and the thinkers about 2%.
    I realise that this is little different than saying the movement in this election campaign is likely to be the same as in previous campaigns if one side or the other decisively wins the campaign otherwise about nil. However, the analysis does afford the possibility of more detailed thinking, in paricular about third party preferences vs primary votes. But as we get more detailed, we tend to be “ässumption driven” rather than “data driven”.
    I haven’t mentioned the ~30% unlikely but possible to change their vote. In my view the self declared small possibility makes it reasonable to ignore. IMO these changes are likely to be caused by some major event, such as the Sept 11th 2001 outrage plus Tampa, and this group are the likely source of the cataclysmic fall in Beasley’s figures. To a lesser extent the Latham implosion would have knocked some of these people off the ALP bandwagon in 2004.. by their nature these things are impossible to predict (although I was less than surprised by the Latham event).
    These are all very nice theoretical musings. Putting my mouth in charge, i’d guess that the Newspolls are indicating a 57/43 split, with an individual poll MOE of about 3%, but with no sign of any secular change. Because of this stability, I’d guess the “true” ALP vote in the range 55.5%-58%. with 2% up for grabs in the campaign, I’d predict the final tesult in the range:
    59%-53% ALP, 47%-41% Coalition, a pro ALP swing of 6%-12%

  7. disenfranchised Gippslander said

    On re reading my last post, one futher point occurs to me.. The likely changers % (L) is a fair indication of possible campaign swings. Assuming the Dolts are about 8% of the population, then possible campaign swing % is L-8

    OH, wouldn’t it be nice to have these figures for each poll!

  8. stevet said

    Possum, my Dad keeps on saying, don’t worry they’ve [the electorate] made up their minds, and when I see extrapolations such as yours today, I think he’s dead on the money.

    And Bushfire Bill, you have suggested that Labor 53.7% to Coalition 46.2% would produce a close result. To give you an idea, in 1983 Labor 53.2% to Coalition 46.8% and in 1996 Coalition 53.6% to Labor 46.4%.

    The result both times? A landslide. I would be more than happy with a result like this.

  9. PaulC said

    Possum, are you taking into account the margin of error on those figures for likelihood of changing votes? Am I right in thinking the MoE is higher than for Newspoll as a whole as the sample sizes are effectively halved?

    Supplementary question – how do those figures stack up for the previous elections? Looking at Bryan Palmers site the way I read it Labor was on 54 TPP 8 weeks out from the 04 election.

    I wonder how accurate peoples answers are on the likelihood of their vote changing. Still, I think it’s a better method than Morgan’s “heading in the right direction” question.

  10. Heretic said


    Wont the next Newspoll be another September data point? I assume they will sample on 28-30 Sep and publish on October 2. Thus we will probably see the Jul-Sep consolidated numbers and the state and age breakdowns in the first week of October.

  11. Bushfire Bill said

    Stevet, my exact word was “closerish. But I take your point. I’d be happy with 53%+.

    Got into a discussion at the barber’s this morning with a Lib so rusted on she think it’s just “normal” to vote for Howard. She was in her sixties and not at all aggressive or obnoxious.

    I was in the chair and heard a discussion on morning TV about the “big ask” that 16 seats are for Labor to win. Some discussion of marginals needed to be won in NSW ensued on the tele.

    Paraphrasing Possum I said (to the gilded youths, youthettes and one sixty-ish lady sitting waiting their turn) that the swings were smaller in the marginals than in the “safe” (10%) seats. Therefore if the Libs were relying on the 16-seat hurdle to save them they were (politely put, of course) up ’emselves with a blunt pineapple.

    The lady, ever so demurely said, “Oh just you wait. Kevin Rudd will make a blunder somewhere along the track. You’ll see.”

    To which I replied, “It’d have to be a big blunder love, with the poll figures as they are.”

    She said, “But the poll figures are good! John Howard has pulled of a coup over the weekend and caught up all the ground he lost!”

    Just goes to prove the power of positive suggestion, eh?

    Oh for the barber shops of old, where disagreements were settled first with drunken shouting, then with fists, then with knives and (if all else failed) with pistols (‘”Murder! BLOODY murder!”, cried the Man From Ironbark’)

    Instead of continuing the dispute outside in the street by engaging in some gutter brawling with broken beer bottles, I dipped me lid and said, “Well, gotta go love. Mind you, Beecroft is bolshy. We scored the biggest swing against Ruddock last election.”

    I left her staring at my newly shaven scone agape, asking, “What does ‘bolshy’ mean?”

    Howard should be careful that ladies like her don’t believe the propaganda too much, or they might get a nasty surprise.

  12. Greensborough Growler said

    Bushfire Bill I have a policy of never arguing with some one who is using sharp pointed objects that could be used as weapons. This specifically applies to wives, ex-wives and barbers.

    He who laughs and runs away, lives to fight another day.

  13. Bushfire Bill said

    GG – agreed.

  14. stevet said

    Bushfire Bill, you are in Beecroft, but my story comes from my homw town of Mt Druitt, solid Labor territory.

    My wife told me that she was speaking to an old woman in the area the other day who remarked that I probably wouldn’t like her because she loved John Howard. She then told my wife that all the oldies loved him.

    I told my wife she should have cited my father as an oldie who definitely doesn’t like him (hate & detest don’t even begin to cover it), just for starters. I had to wait for five minutes whilst she stopped roaring with laughter.

  15. […] And Thus Spake Newspoll My humblest apologies for the delay on the Newspoll low-down; I’ve been organising a few things for the election and […] […]

  16. Stunkrat said

    But alas, as one read through the piece it quickly became apparent that the article had much to be humble about.

    To the contrary, the article contains precisely nothing requiring humility.

  17. Stunkrat said

    Damn, wrong topic.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: