Aug 7, 2010

Nielsen Saturday

This brisk Saturday morning brings us a Nielsen poll via the Fairfax press running from a sample of 1369, giving us an MoE that maxes out around the 2.6% mark - the vote estimates, thei

Possum Comitatus — Editor of Pollytics

Possum Comitatus

Editor of Pollytics

This brisk Saturday morning brings us a Nielsen poll via the Fairfax press running from a sample of 1369, giving us an MoE that maxes out around the 2.6% mark – the vote estimates, their change since last week’s Nielsen and the pref flows –  broken down into gender and geography to the usual decimal place – come in like this:



The Preferred PM figures came in 49/41 to Gillard, with women having it 55/36 while Males were running with Abbott ahead 47/43. The full demographic tables of this Nielsen are here, while the three poll weighted average tables (the more robust estimates of the campaign so far) are here.

On the approval ratings we get:


Nielsen also ran additional questions on leaders attributes, asking respondents if they thought a number of statements described each leader:


You might notice that Gillard universally performs better among women than men. If we look at the the lead Gillard has over Abbott among both men and women, then compare how many more points favorable women are Gillard compared to Abbott (the gender gap) we get:


“Is easily influenced by minority groups” is in red as it’s the opposite to the rest of the questions since a smaller result is’ better’ than a larger result.

This is a two way street here – where Abbott is performing better among men as well as Gillard performing better among women. At the moment however, Abbott’s lead among men is larger than Gillard’s lead among women, giving Abbott the electoral advantage – a 10.2% lead among men on the two party preferred compared to Gillards TPP lead among women of only 7.2%.

Meanwhile, our Pollytrend now shows the all pollster trend sitting on a Labor two party preferred of 51.6% while our phone pollster trend ducks below 50 to come in on 49.6%.


Our usual Nielsen time series charts come in like this:

pmapprovalaug7 opapprovalaug7

netapprovalaug7 ppmaug7


Leave a comment

50 thoughts on “Nielsen Saturday

  1. dylan3

    I agree with Tom Jones- The last time Latham monstered a PM- ( Johnny Howard)- the Pm won – even though he wasn’t dearly loved by most people.

    Most ordinary folk dislike men who act like racecourse drunks and try to bully people- and Julia handled the incident so well and kept so calm, smiling through it all- that I am confident it did her a lot of good. ‘Add to that her bright, honest performance on Q&A on Monday night and one might say she’s had a great boost.


  2. biasdetector

    As if on cue, opinion polls come out showing no momentum for the coalition rofl lol

  3. steve

    A Galaxy in the Curious Snail.

    [TONY Abbott has his work cut out as the election enters the final two weeks, with an exclusive Galaxy poll finding one in two voters don’t believe he’s up to the top job.

    The Galaxy poll for The Courier-Mail shows the election remains neck-and-neck. If it were held today and preferences directed in the same way as 2007, the Government would win 51 to 49 per cent two-party-preferred.

    The poll also shows only 43 per cent of voters believe Mr Abbott’s Coalition is ready to govern.]

  4. David Richards

    Gusface – let’s hope this trend continues and carries on to election day. 2 points or so a week going the ALPs way will make a lot of people breathe a sigh of relief

    Gillard may not be what we really want… but the alternative doesn’t bear thinking about.

  5. Gusface


    big week ahead

    newspolls 52*48 has made me smile

    onward and upward


  6. Possum Comitatus

    Sprocket – yep, phone polls are landline based.

    The key to landline accuracy comes down to a single question:

    “Are the demographics that are mobile phone only, significantly different in their voting behaviour than the exact same age,gender, geographical and income demographics that are accessible through landlines?”

    If a random selection of 1000 19 year olds living in mobile phone only households vote approximately the same way as, say, a random sample of 1000 19 year olds that are accessible by landline – then the telephone polling is accurate.

    If there is a difference between the behaviour of those two groups, then the phone polling starts to get skewed.

    At the moment, there isnt much in the way of evidence that the two groups behave in a way that is significantly different – but there is evidence that it will likely occur in the not too distant future.

  7. Possum Comitatus

    BD went:

    [I don’t mean to sound disrepectful, as I think you have the best analysis on the web bar none.]

    Disrespectful? Hell no!

    *More* argument I say!! 😀

    It’s one of the most interesting things in public opinion – what actually defines “momentum”?

    One of the really interesting bits in the 2007 election campaign was how Howard pulled back 2 or 3 points of two party preferred in the last week.

    Before that, I thought that momentum was simply public opinion movement – in that, if a party is getting points, does it really matter how they’re getting them.

    Yet, even though Howard was gaining late points, did it actually generate him achieving “momentum” or was it just some sort of generic or inherent or general movement that had no real consequence in terms of a point gained leading to another point gained?

    I suppose it comes down to how we actually define what “momentum” is – and we could probably do a million of those.

  8. steve

    Via Grog’s day 22 report comes the costings so far under the Charter of Budget Honesty.

  9. sprocket_

    hi poss, just to confirm for me – phone polls are to landlines only?

    we are in a fast changing demographic where not only young people (eg my 3 kids who are first time voters) and plenty of transients and/or welfare types don’t have, and will probably never have, a traditional landline. What is the % weighting for this? Don’t kow, but it is growing.

    I figure landline only phone polls would be overweighted to Coalition demographics.

  10. biasdetector


    I don’t mean to sound disrepectful, as I think you have the best analysis on the web bar none.

  11. biasdetector

    Dear possum thankyou for your post.
    All those cases you suggest I would have said involved momentum.
    The Grech affair I thought did infact give labor a lot of momentum and they used it to hit the coalition over the head with threats of double dis election.
    Momentum doesnot mean anything will come from it. It is a temporary advantage and may be fleeting, like a football team that is behind and has kicked the last couple of goals has momentum but may still lose the game.
    If in a two horse race, the reasons for a move is less important.

  12. Ray Polglaze


    It looks to me like your feeling that “until the dumping of Rudd votes were moving to the greens, after the dumping the liberal vote started to increase” is basically correct.

    If you look closely at the Nielsen tables, it looks like a critical few percentage points of Labor voters had moved from Rudd to the Greens. When Rudd was replaced by Gillard, those voters plus a lot of Green voters went to Gillard Labor. But since then, the Green voters have gone back to the Greens and a few percentage points of Labor voters have again left Labor but this time they have gone to the Coalition.

    Is this how it looks to you Possum or have I missed something?

    I think there is a relatively clear explanation for this, but I am waiting to see if the Nielsen polls have now stablised after the Rudd-Gillard roller coaster.

  13. Possum Comitatus


    Remember back when the Grech affair blew up in Turnbull’s face – polls moved very strongly to Labor but they didnt really have any momentum from it.

    In 2001 when the polls moved back to Labor strongly in the campaign from the post Tampa, post Sep 11 Coalition highs – Labor didnt really have any “momentum” behind it.

    Who is generating the movement in public opinion? Is Abbott gaining it it or is Labor losing it?

    If one party is gaining from the silliness of the other – like Labor gained in 2001 – does that generate more movement in the public?

  14. Alfred E. Neuman

    Hi possum,

    Can you show a chart comparing pollytrend this time to last election and maybe 04?
    I’d love to see how it compares.

  15. biasdetector

    I am by no means saying that the coalition are going to win by any means. BUT how can anyone say that the coalition does not have momentum ???????
    Just look at the polly trend!!! I don’t think it could be any steeper.

  16. David Richards

    further to [email protected]

    not only misogyny and racism – but good old xtian bigotry towards nonxtians.. or for that matter the anti-intellectualism and sports mania – all the great bogan “values”

  17. Ozymandias

    Agnes @ 15
    [ latent misogyny like racism lurks just below surface in parts of Australia ]

    This would appear to be borne out in the 8 percentage point gap between the genders on PPM.

    Tony’s “no means no” comment last week was dog-whistling the misogynists.

  18. bakunin

    Possum, I notice in the Neilsen fine print that 6% undecided are “redistributed”. Is the method used to redistribute these votes known?

  19. fredn

    Possum what wold be really interesting is the poly trend with significant events marked.

    -The Oceanic viking
    -Dumping of the ets
    -Dumping of rudd.

    What would be also interesting would be the trend line of the green, liberal and labor vote with the same events marked. My feeling is until the dumping of Rudd votes were moving to the greens, after the dumping the liberal vote started to increase.

    If this is the case the graph would be political dynamite within the Labor party.

  20. deblonay

    Latham showed his thuggish side again,,remember his infamous attack on that taxi driver in Sydney some years ago!
    Julia was lucky he didn’t thump her !
    She handled the incident bravely,but looked alarmed…Good work Julia

    The official apology by Gyngell from Ch, 9 would suggest that they won’t show the incident or use Latham’s piece
    That may be the end of his short(thankfully) TV career
    Good riddance Latham!!!!!!!!!!!!!!

  21. Tom Jones

    The last time Mark Latham bullied a Prime Minister it helped that Prime Minister win.

  22. Rod Hagen

    Whoops! Scrub that! I clearly misread the numbers, inverting 49.4 as 49.6 and 50.6 as 50.4! (wish there was an “edit” button here! )

    Not sure how I managed that! How embarrassing!

    Nevertheless , the actual improvement is 1.6% from the 47.8% base in the previous poll , which, if rounded , would normally be seen as a 2% change rather than 1%. Further, if they achieved the same result this week coming they would be ahead of the coalition. A not insignificant change.

  23. Rod Hagen

    Interesting to see that many in the media are “truncating” rather than “rounding” in terms of the overall figures for the Neilsen Poll. The “swing back ” (yes, I know we all know that changes of this size are far from reliable) of 1.8% would normally be rounded by the media as “2%”, and the gross 2PP numbers treated as “50-50″, but on this occasion they have chosen to truncate instead, using 49 – 51 and speaking of a “1%” swing.

    I say “of 1.8%” because last time, looking at Poss’s figures in “Neilsen-THump” , last time they “rounded” rather than “truncating”. This exaggerated Labor’s result last time, but depresses it in this one. So we have a 2PP “swing” for labour from 47.8% to 49.6% over the pertiod.

    I guess any “comeback” by Labor might be seen as “momentum” , and make stories about “failing to make up ground” harder to justify.

    In fact the poll results , if “correct” would mean that Labor picked up more than twice the current “gap” between the parties in the last 7 days, and, if this continued to polling day, would end up with a 2PP of 53.2%!

  24. DemocracyATwork

    Missing from the polls is the level of undecided. This begs the question how skewed is the polling

  25. Yaz

    You have to have seen ‘The Princess Bride’ to understand…

  26. Chokyi Nyingpo

    “I don’t really think any side has any any real momentum – and hasn’t for two weeks, but rather Labor has just figured out new and interesting ways to lose a few points on the vote from their own actions.”

    Here’s another…

    “SYDNEY train commuters will get to travel for free Monday – but only as a goodwill down-payment by unions before they go on strike over a pay dispute, the Daily Telegraph reports.

    The planned industrial action over the next 12 days will cripple Australia’s largest city, with the first shutdown next Thursday when drivers and guards will abandon trains between 10:00am and 2:00pm (AEST)…”

    Sheesh, talk about shooting yourself while you’re down…

  27. agnesmack

    TV images Of Latham finger wagging & looming over Gillard visual portayal of my earlier comment on mysogyny , but hardly latent here.

    [email protected] Crikey has a seat tipping comp for subscribers

  28. David Richards

    which word?

  29. streetcred

    Thanks Possum. As Eliot once said, it may all end with a whimper rather than a bang. Such is life in the world of the hollow men.

  30. MDMConnell

    @11 Thanks Possum, I think Pollbludger reported that as 50-50 so that’s where the confusion came from.

    I was going to ask when the last time a Morgan poll gave a 2PP lead to the LNP but looking through your poll list there have actually been quite a few Morgan polls where the LNP led this year.

  31. MDMConnell

    “It is inconceivable that a government could be thrown out with the economic indicators the way they are, and after having asuccessfully avoided a recession.”

    You keep using that word. I do not think it means what you think it means….

    (someone had to say it)

  32. David Richards

    It is inconceivable that a government could be thrown out with the economic indicators the way they are, and after having asuccessfully avoided a recession.

    These poll results defy logic. The ALP SHOULD be miles in front. Not that I like them much, but when you consider the alternative and the fact that this government has been (in comparison to Keating and Howard) a good one, they don’t deserve to lose.

    I would prefer a third alternative – something closer to the centre in both economics and social libertarian terms, but there isn’t one.

    As a pointer – I’m even putting LDP and FF ahead of the ALP, and the Libs do get lucky last after all due to the disgusting way the media and the Libs have conducted this campaign.

    I didn’t like the Rudd Government, and I was even less happy with the lurch to the right under Gillard, but Phony Tony, the pugilistic pontif makes even Gillard look good.

  33. BeaglieBoy

    Possum u are a might more optomistic foe Labor that the Pollbludger

  34. Bobby

    Possum you kept me sane in 2007! I am relying on you again… Do you think the trend will turn around in the next two weeks or are Labor toast?

    Also is there a competition for picking the seat results this time around?

  35. shepherdmarilyn

    It would be poetic justice for David Bradbury to lose Lindsay after getting it on the back of racism only to lose it on the back of pandering to racism.

    Kelly thought all her mob were racist shits, now Bradbury does too.

  36. agnesmack

    Hard to avoid conclusion that latent misogyny like racism lurks just below surface in parts of Australia. I say latent because it remain unnoticed in some, possibly older males until they are confronted by the prospect of a woman at the top of the power pole. “I don’t want any sheila telling me what to do” is rarely voiced in these seemingly more enlightened times, but I fear it is a sentiment still nurtured in the darker corners of a few psyches.

  37. jenauthor

    [The methodology of phone polling is a known quantity with known accuracy and has an excellent historical track record.]

    If that is the case — then the current phone figs are very depressing!

  38. Possum Comitatus

    Streetcred – I don’t really think any side has any any real momentum – and hasn’t for two weeks, but rather Labor has just figured out new and interesting ways to lose a few points on the vote from their own actions.

    I think at the moment were in a period of random drift where neither side is taking the election by the scruff of the neck – where the Coalition can’t because they have nothing to do it with, and Labor aren’t because they’re too busy making problems for themselves.

  39. streetcred

    Possum – What do you take from these figures about campaign momentum at this stage? Seems to look fairly static (?)

  40. Possum Comitatus

    MDM – the last Morgan phone poll, 49.5 for the ALP

    They’re all listed here going back to January 2008:

  41. MDMConnell

    What is the other pro-Coalition poll you have in the Pollytrend for August?

    There’s the two Nielsen polls 52-48 and 51-49 but you’ve got one around 50.5-49.5 to Coalition. I must have missed this poll, or is it the Galaxy marginal poll translated to a national 2PP?

  42. Possum Comitatus

    [Poss, at the risk of sounding like a dope]

    Better than looking like one which I often do around these parts! 😛

    The methodology of phone polling is a known quantity with known accuracy and has an excellent historical track record.

    Other methodologies have uncertainty. Face to face polling in Australia (that Morgan does for instance) shows a lean towards Labor compared to phone polling, and online panel polling such as Essential does is a relatively unknown quantity (as this is Essential’s first public go at an election) and where Essential sometimes leans towards Labor compared to phone polling, and sometimes does not.

    So I run two trend lines – the all pollster trend that contains everyone, and the phone pollster trend that contains only the methodology with the best track record.

  43. Traveller

    Poss, at the risk of sounding like a dope, why’s it important to separate phone polling from the rest?

  44. Tri$tan

    Darn, the comment must be the same. It was looking fwds from that point in time.

    You can’t change it with the benefit of hindsight. That’s hindsight bias.

    “Hindsight bias is the inclination to see events that have occurred as being more predictable than they were before they took place. Hindsight bias has been demonstrated experimentally in a variety of settings, including politics, games and medicine.[1]”

    Fact is there were multiple outcomes at the time. And just because something happened doesn’t mean that it HAD to happen and was predictable.

    It also does show the difficulty in predicting results rather than looking at the current state of play and the past. This blog is traditionally really great at revealing a current state of play – but prediction is a really different kettle of fish.

  45. Possum Comitatus

    Hasn’t affected it at all Darn.

  46. Darn


    How does this affect your comment of about a week ago that you would still prefer to be in the position of Labor than the Libs

  47. rationalist

    I think even though the TPP is slightly better than last week, this poll will count as an ALP ouchpoll due to the sharp decline in preferred PM.

  48. Possum Comitatus

    Dogma, there was -there usually always is. But it’s not a predictive one – the two move together.

  49. dogma

    Hi Poss,

    Bit crisp this morning up in qld. In 2007 wasn’t there a correlation between preferred PM and the primary vote?

  50. Tweets that mention Nielsen Saturday – Pollytics --

    […] This post was mentioned on Twitter by Possum Comitatus, STVFTA. STVFTA said: RT @Pollytics: On Pollytics: Nielsen Saturday The usual demographic breakdowns, results to a decimal point etc etc […]

Leave a comment

Share this article with a friend

Just fill out the fields below and we'll send your friend a link to this article along with a message from you.

Your details

Your friend's details