ACNielsen: 55-45

The Fairfax papers today carry their monthly ACNielsen poll, which shows a narrowing of Labor’s two-party lead from 58-42 to 55-45. Labor’s primary vote is down from 49 per cent to 46 per cent, while the Coalition is up from 39 per cent to 41 per cent. The movement most likely represents a correction from a somewhat excessive result last time. Now please, for the love of Christ, no more polls until next Tuesday …

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.

558 comments on “ACNielsen: 55-45”

Comments Page 1 of 12
1 2 12
  1. Michelle Grattan does the thing I dislike most about opinion poll reporters. She puts far too much emphasis on the movement. Polls themselves are error-prone enough on their own, and to subtract the result of one error-prone poll from another error-prone poll really is the height of silliness.

    That said, 55-45 seems about right. Since it more or less matches what the other pollsters are saying.

    Looking back, I tend to agree that last months 58-42 result was “somewhat excessive”. Which is kind of ironic since it accorded well with the previous Nielsen results; but most other pollings outfits were reporting a closer contest at the time.

  2. Considering most pollsters ring around 6pm when people are having dinner etc, I wonder what the rejection rate is and whether they only have a set number of people to ring per electorate, or do they just keep ringing people until they get their quota for the area ?

  3. Hmm, I wonder if Friday’s Sharemarket collapse had anything to do with the small bounce back to the Govt?

    “THE Australian stock market posted its biggest one-day fall in six years yesterday following falls on Wall Street, which is still spooked by the rout in the US sub-prime mortgage sector.

    The largest fall since the September 11, 2001 US terrorist attacks wiped almost $53 billion from the value of the Australian market, rattling investor confidence and eroding superannuation-linked gains.

    The benchmark S&P/ASX200 index was 229.6 points lower at 5936, while the All Ordinaries retreated 222.5 points to 5965.2.

    At the close of day trading on the Sydney Futures Exchange, the September share price index contract was 217 points lower at 5934, on a volume of 36,791 contracts. ”,25197,22224173-20142,00.html

  4. 55-45.

    In Six months the ALP have dropped from 61 to 55 and with about three months to go I think we can knock of another 3 making the Election result

    ALP 52 – Liberals 48. I find the state voting intentions interesting with the ALP dominating in South Australia, NSW Queensland and WA continues to be strong for Liberals, but Victoria appears to have moved toward the Liberals.

    I can’t think of any reason for the Victorian numbers except and maybe this is a long bow to draw but were the polled responding to Brumby becoming Premier.

    I know that may sound silly, but the news in Victoria has generally been poor for the Federal Government.

    What happened to polling at Federal level when Carr departed.

    While this poll was done after the rate announcement I wouldn’t expect to see an impact until the next Newspoll.

    I agree with the point made by Jackman on his Blog, the swing will be bigger in the 4-9% seats than the Marginals, but I think this as all ways been the way.

    As I’ve been thinking for quite sometime while tipping an easy ALP win vote wise, but when looking at the seats I’ve not been able to see the ALP winning more than 80 seats without having seats marked as ALP gains that they don’t normally win.

    I note that the previous six months have seen little to no movement in the TPP until this poll, Howard looks gone but and maybe this is something while Rudd has played Howard very nicely.

    He appears to have eased off the gas a little and there is a danger that the ALP will let Howard back by not maintaing the front foot which got them the lead in the first place.

  5. bmw , Rudd is out there announcing a big housing policy ($600M) today. I think he needs to get into the ‘front row’ and lay a few tacklesin NRL speak now.

    His timing has been spot on to date, but now I think now he needs to convince the electorate that he has a plan for Australia that is DIFFERENT to JWH to enough extent that people can recognise an alternative approach in key policy areas like health, housing, education, environmental issues (climate change) and WATER to back up his trump card (industrial relations:Workchoices). The negative press for JWH will abate, eventually, even in the polls as this poll may be forewarning.

  6. State by state breakdowns from AC Nielsen Poll:

    N.S.W: ALP: 56, COAL: 44
    VIC: ALP: 51, COAL: 49(WTF)
    QLD: ALP: 57, COAL: 43(WTF)
    SA: ALP: 62, COAL: 38
    WA: COAL: 56, ALP: 44

    Victoria I don’t believe(is Howard really that popular down there?), wasn’t the council amalgamation thing said to hurting Rudd in QLD, and why contradictory WA results?

    I think I’d rather trust Newspoll, far more reliable.

  7. The WA sample size would be about 140 out of the 1400 polled, would it not? Built-in dodginess. If we add these voters to the Westpoll lot from last week I think we get about a 52/48 ALP/Lib split, which is probably closer to the ‘truth’.

  8. As I pointed out in today’s post on the rental subject –

    “Labor seems to think that it can keep the price of petrol, groceries, rent and child care down while introducing a climate change policy that calls for a CO2 reduction of something like 100% by next Wednesday and at the same time control inflation in a wages-growth market. Good luck with that. If Labor wins the next election then we’ll be lucky not to have double digit interest rates within their first term.

    Labor leader Kevin Rudd has an understanding of economics akin to Gough Whitlam and the temperament of Mark Latham. It’s not a good combination.”

    When the ALP starts costing some of this stuff then the Coalition will close the gap quickly.

  9. Out of curiousity, would it be possible for the Govt to win with 48% of the vote?

    It’s possible for a government to win with around 25% of the vote (50% + 1 of the vote in 50% +1 of the seats; 0 elsewhere) but it’s not probable.

    Howard in 1998 won with the lowest share of the 2PP vote since proper records have been kept at a tick over 49%. So a victory with 48% of the 2PP vote would have to not just break the current record, but smash it.

    It’s exceedingly unlikely.

  10. It’s been mentioned before, but the fury that surrounds the release of one poll’s findings is just rampant nonsense. Today’s A C Nielsen release is no exception. From interpreting state swings from samples of 150, to concluding that last week’s interest rate rise was of political benefit to the Coalition. The media’s full of it – in more ways than one.

    It reminded me of the example given in Carl Sagan’s Cosmos, where he was explaining how, in the absence of absolute facts, the mind can run wild with speculation. It related to scientists in previous centuries, trying to come to terms with the planet Venus and the difficulty they had in not being able to see the surface of the planet with their telescopes, because of the thick clouds in the atmosphere.

    “The absence of anything to see on Venus led some scientists to the curious conclusion that the surface was a swamp, like the Earth in the Carboniferous Period. The argument – if we can dignify it by such a word – went something like this:

    ‘ I can’t see a thing on Venus.’
    ‘Why not?’
    ‘Because it’s totally covered with clouds.’
    ‘What are clouds made of?’
    ‘Water, of course.’
    ‘Then why are the clouds of Venus thicker than the clouds on Earth?’
    ‘Because there’s more water there.’
    ‘But if there’s more water in the clouds, there must be more water on the surface. What kind of surfaces are very wet?’

    And if there are swamps, why not cyacads and dragonflies and perhaps even dinosaurs on Venus? Observation: There was absolutely nothing to see on Venus. Conclusion: Dinosaurs.

    The featureless clouds of Venus reflected our own pre-dispositions….Venus turns out not to oblige our pre-dispositions.”

    It seems the last 8 months polling and the opinions of over 70,000 respondents, does not oblige the pre-dispositions of the Canberra media either.

  11. Herald front page had a Pythonesque Black Knight quality about it. JWH is still a limbless political torso but we now discover there’s a couple of teeth left.

    Interesting conversation at weekend with a rusted on Lib with close connections to the Howard inner circle. Essentially thinks he’s lost it. Inside the circle Howard was thinking out loud a while back about having a crack at Menzies’ record. Problem is he now actually believes he can do it. And with Arthur gone there’s no-one left that can tell him otherwise. Sad and a bit of a worry because this guy is still PM.

  12. And to complete the King George picture there’s no-one capable of moving against him. The Right would rather wait and salvage what they can from the post-election wreckage than triggering something that might end up delivering the prize to the Left. And the Costello forces just don’t have anywhere near the numbers so they’re just waiting for the election to destroy the Right.

    Sad and ugly.

  13. Indeed, that polls indicating a victory that would be the largest ever for the ALP, and the largest for any party in 30 years are being hailed as good news for the government really shows what kind of a mess they are in.

    (and yes, yes, not getting carried away, long time to go, things could change…)

  14. I note the bookmakers odds on the Govt winning moved out dramitcally after the rate rise and a series of poor regional polls. The AC Neilson poll have moved the odds back in. They are implying a 58% chance of an ALP win. I think this is about right.

    The Govt is still pounding away with a tax payer funded advertising campain. I noticed two or three govt ads per break on Sunday night. They still have a pile of cash to throw at marginals. So if the can gain 3 or 4 percent across the board and defend a few keys marginals they are definitely in the game.

    Still this is a big ask. I get the impression the ALP is holding back a lot of fire power at the moment. Just avoiding the wedges. I wonder when they will switch to offense?

  15. “Howard in 1998 won with the lowest share of the 2PP vote since proper records have been kept at a tick over 49%.”

    He won in 1998 with just a tick under 49%: 48.9% and that still gave him a 13 seat majority. Assuming a ‘uniform national swing’ in 2007, the Government will hold on with 48.6%.

    It’s unlikely that they’ll hold on with 48.0% of the TPP, but I’m not sure ‘exceedingly’ is the correct adverb to qualify the unlikelihood with.


  16. The SMH and Age are in the garbage business this morning. What’s going on? News Ltd is dumping Howard and now Grattan and Hartcher are doing Dennis Shanahan’s job?

    The headline figure is an imaginary one–the primary figures are the ones that count–and this has been spectacularly consistent–Labor with 46-49% for a long, long time.

    I think we are getting tunnel vision here. Let’s all go and get another hobby, like stamp collecting or animal cruelty.

  17. He won in 1998 with just a tick under 49%: 48.9% and that still gave him a 13 seat majority. says 49.02% but I won’t quibble.

    About that anyway: if I read the pendulum correctly Bennelong, and government, would fall with a uniform national swing of 4.0% from last election which would be 51.3/48.7 ALP/Coalition, so on that basis (and to the nearest 0.1%) the Coalition could ‘expect’ to win with a uniform 2PP vote of 48.8% but not less.

    I still think that beating this mark by 0.8%, and the previous record by 1.0% is very, very unlikely but I concede that excess is in the eye of the beholder.

  18. In a lot of these polls the 2pp is worked out assuming preference go the same way as the last election. How reasonable is that?

    My impression is that a lot of ALP voters went for minor parties in the last election and preferenced the ALP. It would seem that most of these voters have now returned to the ALP primary vote. But does this mean the remaining minor party vote will be less favourable to the ALP than the last election and hence the 2pp vote for the ALP is overstated?

  19. 3 things to take out of this poll:

    – the PM’s approval rating of 50% is nothing short of phenomenal, when you consider he’s been in the top job for 11 years. If anyone needs reminding, after 9 years the supposedly mega-popular Hawke was languishing in the mid-20s. So despite the squealings from the media, John Howard is definitely not on the nose in the electorate and still comands a great deal of support. This will come into play during the election campaign.

    – 83% of those interviewed said that the experience of the candidates would be critical in influencing how they vote. It’s no coincidence, then, that Ministers like Downer and Abbott say in interview after interview that the electorate will ultimately re-elect the PM and his team because of their experience and their amazing record – concerns about Rudd’s inexperience are obviously coming through in the Liberals’ own research. Watch for the Government to keep hammering this point over the next couple of months.

    – the Government’s primary vote is now above the 40% mark; indeed, from now on any poll that has the Coalition below 40% should be automatically discounted. And, as I mentioned in another thread, when the naturally pro-Labor Nielsen has the parties at 41/46, then you know the reality is much, MUCH closer. And once again, all astute observers ignore the 2PP figure because it’s just a figure plucked out of thin air by the pollster.

    Everything here points to a Government victory in November. You have been warned.

    PS: Yet more evidence of how hard our media is campaigning for Labor: at ninemsn’s news site, the story about this poll has already been pushed off the front page, whereas yesterday the story about the poll in the Bennelong was highlighted there for the whole day. And when you do jump to this story, you find that the hack responsible couldn’t resist reminding us about the PM’s supposed troubles in his seat. Lovely. I hope there’ll be a lot of payback all around when the Government is re-elected.

  20. Nice to see renters getting a sniff in the policy wars today. Thats the sort of thing that will keep green prefs tight, and convert more than a few to ALP first pref – nice shoring work on the left flank by Rudd.

    Back in the main game: if Rudd sorts out the kinder gap for 3-4 yo (ie the current model is pretty cap where both parents work, seems to be based on giving home Mum a break, there’s very few long day care options in the kinder mix) he’ll solve one major cross-spectrum headache for the punteriat. I predict some moves there.

    This latest poll is ace: will keep Lord Wentworth at Bay, and Rodent araldited to the losing seat.

    Seriously though – I reckon thats the boomer retirees swinging back some. Check the age group breakdown. Remember they just landed the mother of all bribes last month – the tax free super free kick holiday special.

    On that thesis: can someone tell me what the 55+ 2PP vote was last month? August has it 48-52 pro Rodent.

  21. “If you torture the data long enough, it will confess to anything,” as they say in medical research. The same is true of polls. Polling is an inexact science. Movements of 1% here and 2% there don’t mean very much, so everyone should calm down. All we can usefully say about the polls is that most of them have shown Labor’s 2PV in the 54-56% band for the last three months, after having mostly shown it in the 58-60% band in March-April. So Labor is well ahead, but not as far ahead as they were. *IF* that trend continues (a big if), things will be about evens by November, and the election will then depend on the campaign, as most elections do. The debates will be particularly important.

  22. I’ve found some of the reporting of the Nielsen poll mystifying. One writer this morning suggested the swing back to the government would be a relief to the government after the Galaxy poll in Bennelong. Which is a very odd thing to write, given both polls showed a 7% swing and they were both done in the same few days. The same result from two different polls can’t mean different things.

  23. Martin B,

    Correct me if I am wrong but I think Bennelong is the 15th most marginal seat so if Labor won all the seats up to Bennelong that would make it 75-73-2. I presume you are saying that the Government would fall on the basis that at least one of the two independents would support Labor to avoid a hung parliament.

    To have a majority Labor would need to win the 16th seat which is Dobell on 4.8%.

    All this is academic of course because swings aren’t uniform.

  24. Wouldn’t it be funny if the Man of Steel pushed the election back to January 2008? That would leave Labor looking very silly over all their ‘Kevin 07’ hype !

    They’d have to throw away all those ridiculous T-shirts, and change the website. Their rhyming slogan (what genius thought that one up, you’d have to wonder !) would be totally meaningless, and they’d be left looking the bunch of fools that they are.

    So says Cerdic Conan.

  25. Here’s an interesting poll on interest rates – funny this one doesn’t seem to get much publicity …..

    “Mr Howard seized on a finding in the ACNielsen poll that only seven per cent of Australians believe interest rates would be lower under a Labor government.

    “I tell you what, the 93 per cent who don’t think they would be lower under Labor are dead right,” he said.

    “Labor has a bad record on interest rates. They hit 17 per cent when they were last in office and on average housing rates under us have been four and a half per cent than what they were under Labor.”

    Mr Howard urged voters not to put the economy at risk by electing a Labor government.”

  26. AC Nielson actually asks which party people will preference, unlike Newspoll, which assumes last election’s preference flows hold. This poll is in agreement with Newspoll in 2PP, but not on PV, where News has had Labor at 47-48 and Coalition at 39-40 for the last 3 polls. On PV, this poll is close to Galaxy, so I think it can be considered an outlier in terms of the ACN series.

    There is some evidence of skew in this poll, with the Dem vote going from 1% to 0% (not

  27. Antony Green Says: The same result from two different polls can’t mean different things.

    Actually, that’s exactly what it could mean. Trying to second-guess the rationalle between why someone chooses to say X or Y in a poll is not possible with polling.

    The only appropriate information you can get out of polling, is a trend over time, and even then, you’re not going to be privy to the reasons behind the people who make the decisions that go into the polls. That is what is so good these days about having so many polls, and so much analysis. The ‘rogue’ polls are pulled back into the average, and that gives you a clear indication of what poeple are going to vote.

    But not WHY they are going to vote a certain way.

  28. Which is a very odd thing to write, given both polls showed a 7% swing and they were both done in the same few days. The same result from two different polls can’t mean different things.

    You appear to be stuck in Enlightenment rationality 😉

  29. Martin J:

    If it came to it, then I strongly suspect that Tony Windsor would support an ALP government, but in any case the point was about the 2PP% that the govt could win with, so “government lose its majority” was really the appropriate point rather than “ALP win a majority” regardless of what I wrote.

    But as you said…

  30. My message has been chopped in half (Aaargh!)

    I was saying the Democrat decline from 1% to 0%(not less than 0.5%) implies 0 out of 1400+ were going Democrat. This is very significant as std err is prop. to sqrt(p*(1-p)); this could be evidence of skew.

    About the state 2PP breakdowns, 2PP for NSW, Vic and Qld were given in July. Here are July and August Labor 2PP for these states and averages.

    State Jul Aug Avg
    NSW 63 56 59.5
    Vic 54 51 52.5
    Qld 48 57 52.5

    No evidence that amalgamations are hurting Labor in Qld.

  31. “…the PM’s approval rating of 50% is nothing short of phenomenal, when you consider he’s been in the top job for 11 years.”

    Steven, it is very possible that some people are approving of Howard, not for his good governing, but for his capacity this year for undermining the Liberal’s chances of a victory. In other words, a LOT of people want to see Howard gone, so if he is seen as doing the things which they think will reduce his chances of winning, then he gets their tick of approval.

    Personally, I haven’t been polled, but if I were, I can easily imagine saying YES to a question about whether I approve of Howard’s performance lately. I think he has done a fine job at digging himself into a hole!

  32. “I’ve found some of the reporting of the Nielsen poll mystifying.”

    So true A.G. What gets me is Auntys ‘good news for Howard spin’, when the Liberal so called ‘pickup’ is within Nielsen margin for error.

  33. And Antony Green, further to that, the same thing goes for the commentariat on it. Amazing as it may seem to you, there are some people that actually beleive things, that to you, are completely nonsensical. It’s not necessarily a reflection on their intelligence (although, with some it clearly is, e.g. CC) but the order in which a particular belief structure was created.

    There are a number of thought experiments for this. (and again, i’m paraphrasing N.N.Taleb) You can ask someone how many Esperanto speakers there are in Australia, how many kilometres it is from the earth to the moon, and then ask how many kilometres it is from saturn to the sun. The order in which you ask these questions will influence the average values that people answer with. It’s seems to be a biological process of our brains to see patterns where none exist, and to form constructs based upon incomplete material. We then, when we see further data, fit the data into already pre-conceived notions.

    Poll analysis (like market analysis) is tailor made for this false reasoning.

  34. Bennelong recorded a significant anti-Howard swing at the last election, which was very much against the national average. For Maxine to get a further large swing when people know it’s a real contest this time would be phenomenal, and could only be done in a real Labor landslide, or at least one in Sydney.

Comments Page 1 of 12
1 2 12

Leave a Reply

Your email address will not be published. Required fields are marked *