Possums Pollytics

Politics, elections and piffle plinking

For Whom the Inertia Tolls

Posted by Possum Comitatus on August 12, 2007

In-er-tia : the property of polling by which a political party retains its voting level at a state of rest, or its voting level along a straight line so long as it is not acted upon by an external force….

…… such as the actions of its political opponents.

Take a squiz at this:

The ALP and Coalition TPP swings, by month, since the last election using Newspoll and where the bottom axis is the month/year:

c2pps11.jpg a2pps11.jpg

There is no doubt that Rudd has made an impact upon the ALP vote, but that impact does not appear to be the orthodox narrative that gets peddled around the pages of some of the daily political commentariat.

Back in the pre-Rudd days, those long forgotten days when the Coalition occasionally won a poll (and strangely enough, when I used to be able to get a decent Caesar Salad) – the seeds of the Howard governments destruction appeared to have already been sown.

Bit by bit, polling point by polling point had Puntersville turned on Howard in the years since the 2004 election. The trend away from the government was clouded by the small recoveries, but each recovery continued to claw back less support than they had lost.

Not only can it be seen in the swing graphs above, but it can also clearly be seen in the TPP vote itself. If we regress the Coalition TPP vote on a constant and a time variable that starts at 1 for November 2004, 2 for December 2004 etc right through to 34 for August 2007, and graph the results we get the following:


R-Squared = 0.838 and the time variable has a coefficient of 0.306 with t-stat of -12.9 and a p-value of 0.0000. This tells us that a linear decline in the Coalition two party preferred vote of 0.3% per month explains approximately 84% of the movement in the Coalition TPP vote since the last election, at an extremely high level of statistical significance.


Likewise if we do the same for the ALP TPP vote:


As expected, we get the same result. R-square= 0.838, with the time variable having a coefficient of +0.306.

But if we throw in a Rudd dummy variable that has a value of 0 for the months he wasn’t leader and a value of 1 for the periods Rudd was the leader of the ALP and run the regression again we get:


Here we get an R-squared= 0.87, Adjusted R-Squared=0.86, with the time trend variable having a coefficient of 0.24 and the Rudd Dummy variable coefficient having a value of 2.02.

This tells us that the ALP were gaining 0.24 percentage points a month (and the Coalition losing the same) without Rudd, and that once Rudd was elected leader he brought to the table an extra 2% of the TPP vote for the ALP.

If the ALP TPP was already growing, how can one explain the massive surge in the ALP primary under Rudd?

Simple – the ALP TPP was getting boosted off the back of the preference flows from the growing minor party vote until Rudd came along. After that, the minor party vote declined and many of those votes that ended up with ALP through preferences came across to become ALP primary votes instead. All the while the ALP TPP vote continued to increase as it had before.

Longer term readers here will already know this to be the case. Voters started deserting the Howard government a long time ago, but parked their primary vote with the minors and preferenced the ALP.

If we take the difference between the ALP TPP vote and the ALP primary vote, and graph that difference against the minor party vote – they will follow a similar path and shape if what we are saying here is true.

So let’s do it and see:


Lo and behold.

Which gets us into this nonsense about Rudds vote being somehow “soft”.

The ALP TPP vote before Rudd was soft, as it relied on a large amount of preference flows from a large amount of minor party voters. But now those voters have come across and put their primary vote with the ALP instead of their preferences. That isn’t a soft vote – it’s as hard as it gets. It’s a primary vote. So the next time you read some dullard waxing lyrical about the soft Rudd vote – remember this:

What that person is really saying is that there are a number of voters which could easily switch back, not to the minor parties from whence they came and where they used to preference the ALP, but back to the government that they deserted over the last 3 years, mostly starting after the 2005 budget and whom have gone out of their way to vote for anybody BUT the government.

The voting behaviour running against the government has been inertial. The TPP trends show that clearly, and what the Rudd leadership has done is merely change the nature of that inertia.

Where before Rudd it was just a general retreat away from the government, it is now a general charge to the ALP. It has changed from people not voting FOR the government and letting preferences flow to the ALP via the minors, to people voting for the ALP directly.

For the government to win the election, not only do they have to turn the vote around, but they have to turn around 3 years of momentum running against them in 3 months.

BTW – I’m back now, so I’ll get stuck into answering the comments from tomorrow.

AddThis Social Bookmark Button


36 Responses to “For Whom the Inertia Tolls”

  1. Hoots said

    Great to have you back. I’ve been reading your analysis for a while now and the last couple of weeks left a hole.
    The trends you point out in this post are fascinating. Bring on the election.

  2. Hoots said

    Actually, something just struck me about the trend lines being applied to the poll figures. I know bugger all about statistics, but I was interested in the debate that followed the “Great Climate Swindle” doco when some of the commentators pointed out how those trend lines can be altered so dramatically depending on where you choose to start tracking them. It makes sense to track from the last election, but much has been made of a trend since March or April this year back to the Government. It looks like even that trend seems to be turning around (not that it had built enough steam to get the Coalition over the line). Looking at your graphs, that 2007 “trend” may just have been a speed-bump.

  3. Lomandra said

    Possum is back!

    And there was much rejoicing.

  4. codger said

    The wicked ppm theory, not wiggling nor wriggling; & Bryan’s gallaxy ghoul monsta trend inta da futcha…welcome back Possum.

  5. Baz said

    Welcome back Possum – it’s been difficult with everything that’s been happening and you not here to tell us your angle.

  6. Backa said

    With the minor party votes going directly, rather than via preferences, to ALP does this not mean that the remaining minor party ALP preferences are therefore diluted and an adjustment needs to be made to predicted preference flows?

  7. Graeme said

    Is your final graph – particularly the way the minor vote tracks the ALP prefs – in any way conflated by the fact that Newspoll doesn’t actually measure second preferences? (My understanding is they just rely on 2004 pref flows, and hence are using a set allocation throughout the term).

  8. Jarusa said

    As a stats nut and a politics junky I think your website is a marvellous read.

    Missed your analysis for a few weeks, great to have you back.

  9. Evan said

    Welcome back Possum. Been hanging out for some decent analysis.

  10. David Walsh said

    I can’t agree with your casual definition of ‘hard’ and ‘soft’.

    The “soft” vote is really just another term for the swing vote. It needn’t have anything to do first or second preferences.

    I would assert that most swing voters vote primarily for a major party. To my mind, these are largely unengaged voters and its them that the major parties fight so hard for. These voters want nothing to do with the minor parties who they see either as too extreme or just a non-event.

    On the other hand, many/most minor party voters are rusted-on partisans in two party terms. Take Greens voters for instance. For most of them, they vote Green because the ALP isn’t left-wing enough. But not in a mad fit would they consider preferencing John Howard ahead of Labor.

    In short, primary vote isn’t necessarily hard and preference vote isn’t necessarily soft.

  11. Bushfire Bill said

    After a discussion last week on “soft” versus “hard” votes on another blog, Surfdom, particularly the Morgan version (“Is the country heading in the right direction?”), I got interested enough to phone Gary Morgan personally (after all, he posts his phone number on the Morgan web site).

    I asked him straight-out: what’s with this “heading in the right direction?” concept for determining softness of vote?

    He said it was solely his intuition that this was the way to measure “softness” of vote. In other words, he made it up.

    No correlations, no focus group backup, no other stats to measure it against, no correlations with other questions… it was just his experience and intuition that this was The Question to ask.

    Given the generality of the question, and the possibility of ambiguous interpretations of it, I thought it was possible that some voters could think the country was moving in the right direction because Howard was about to get turfed out (Morgan has 61.5% at last count thinking Labor will win), or maybe that things were so good even Labor couldn’t f**k it up, or that the economy had nothing to do with Howard… did Gary’s crew ask those kinds of follow-up questions?

    Nope. Just the one question. Based purely on his experience about how people think.

    So there you have it. Mystery solved.

    BTW, he believes Labor is only marginally likely to win. He says the real odds are 50/50 with rabbits out of hats to come from Howard.

    He also reckons that the public will warm to Work Choices seeing as (he believes) anyone on an AWA is paid more than a worker in an equivalent job on a standard agfreement. In his opinion, WCs is a PLUS for Howard and we’ll soon be seeing that, starting from Western Australia and spreading out from there. He is absolutely convinced this is true.

    I guess that puts paid to another theory – that Morgan is a Labor stooge.

    From what he said to me, that ain’t the case at all.

  12. […] For Whom the Inertia Tolls In-er-tia : the property of polling by which a political party retains its voting level at a state of rest, or its […] […]

  13. steve everist said

    The state by state difference in TPP voters intention from the Nielson poll published Monday 13 August makes an interesting comparison to the Table of state by state swings in the table in Pollycide Part 3 (based on a summary of the quarter’s polls).

    Very Close

    National 10 vs 9.8
    NSW 12 vs 12.2
    QLD 13 vs 11.1

    Way Apart

    VIC 2 vs 9
    SA 24 vs 10.4
    WA -12 vs 5.4

    The Nielson sate analysis for WA is in marked contrast to the Saturday 11 August WestPoll that shows a TPP advantage to the ALP.

  14. Possum Comitatus said

    trends do depend on where you start.Maybe this movement back to the government is something real, maybe its just noise (and there is a hell of a lot of noise in any poll trend simply because of its nature).The government should get a bit of movement back to it if history is a guide as undecided voters tend to (but not always) move to the incumbent in greater proportions than they move to the challenger.But you’re only talking a few points there anyway.

    Oh Lomandra – if you keep that up I’ll have to add you to the christmas list 😉

    I would have thought that to be the case, but both ACNielson and Morgan polling which asks for preferences, suggest that there is no evident dillution of preference flows to the ALP at all.On the contrary, it suggests that preferences are flowing more strongly to the ALP than they did at the 2004 election which really is bizarre.It’s something I might give a bit more thought to over the next few weeks and see if we can pull something out of the data that might explain what’s going on there.

    Graeme – it might, but the difference between 2004 preference flows and what Morgan and ACNielson are picking up suggests that the difference is only a percent or so.Nothing of a magnitude that would throw that relationship out, especially since it’s probably less of a difference than the error margins involved in the actual polling itself.

    I’m not sure how to define hard and soft without having some polling organisation actually asking questions about the strength of voting intention.I’ve given this hard/soft issue a fair bit of thought over the last few weeks and tried all sorts of data manipulation to try and pull something meaningful measure of it out – but without some actual baseline to judge an artificial measure against – the whole exercise became kind of pointless 😉

    What I do believe though, is that being a party with a high TPP based on primaries is preferable to being a party with a high TPP based on preference flows.

    From the swinging voter analysis on the site we can see that voters went from the Coalition to the minors, and from the Morgan and ACNielson polls over the last 3 years (which measure preferences) we can also see that those voters started preferencing the ALP rather than the Coalition from whence they came.From the swinging voter analysis we can then see those new minor party voters probably moving across to Rudd (I say probably as I’m of the opinion its more likely those new minor party voters rather than the rusted on minor party voters were the ones that moved, and it had to be one of those groups that moved to explain the polling levels).

    Thanks for that.I’m as bewildered as you are about that measurement of ‘soft’ for all the reasons you mention.

    Gary has a lot of interesting things to say about how people think, and he’s always worth listening to – hell, he makes his living out of it and has for eternity.But I cant agree with him on this one.Especially considering the way Workchices is playing out in that whole Interest Payments to Disposable Income/Reduced Discretionary Spending Household Budgets/Self Perceived Living Standards nexus as it relates to where the vote swings seem to be happening.

    The ACNielson State breakdown just doesnt have the sample size to be confident about what it says for every State.The Newspoll quarterly breakdowns based on 10 thousand people or so is probably the best State based indicator we have at the moment.As you say, Vic, WA and SA are out there on the fringes.A bit like every Westpoll seems to be – they need a larger sample.I have no idea why Westpoll doesnt just half their number of polls and double their resources for the polls they do.Who knows, maybe they have high fixed costs or something.

  15. EconoMan said

    Bushfire Bill, on the theory that Morgan is a Labor stooge: Morgan were the preferred pollsters of the Liberal party before Textor. ‘Nuf said.

    By the way BB, did you ask Gary whether they ask ‘who do you think will win’ before they ask ‘heading in the right direction’?

    Steve, as possum says, the state-by-state numbers are on samples that are just too small. See ozpolitics which says take them with ‘a grain of salt’. There is no way VIC is 51/49. It’s almost never been that good for Howard.

    Possum, I also think Morgan is out of his mind if he thinks WorkChoices is a winner. And workers on AWAs get paid less than workers in the same jobs. All evidence supports that. Comparisons saying AWAs get paid more conflate higher paid workers (managers, mining) with the rest to distort the result.

  16. Leopold said

    The presumption that those minor party preferences in 2006 were soft votes is dubious.

    An equally plausible scenario is that much of the primary boost under Rudd has come from progressive left voters who couldn’t stand Beazley – but were still 100% rock-solid ALP preferences, and some of whom would have reluctantly come back to the primary fold on election day. The Greens have plunged 3%+ across all polls, for example.

    It may sound logical to assert that primary votes are ‘solid’ – but Bomber was polling in the low-to-mid 40s through most of the 1998-2001 term and got 37.8% on the day. Latham polled 41%+ right through 2004 than dropped around 3% on BOTH primaries and 2PP once the election was called. I don’t immediately see any great evidence that having a higher primary vote makes your 2PP voting intention any more stable (it may be higher, but we’re talking about whether votes are ‘soft’ or not).

    I don’t disagree with much of what you say above. The government has been in trouble in the polls for a long while now, and Rudd has given Labor a 2PP bounce of 3% or so (you say 2 – something between 2 and 4 I’d say). But the presumption that a 2PP of 55 off a primary of 46 is significantly more reliable than a 2PP of 54 off a primary of 42 is just that – presumption. How reliable it is depends on a bunch of things that can’t be measured in quantitative polls.

  17. The Doctor said

    that first and largest spike in the ALP swing is due to WorkChoices isn’t it?
    I think you could just as easily made a WC Dummy variable of value 2, and gotten results not dissimilar to the Rudd effect with it, either that or use two dummies of 1 + 1.

  18. David Flude said

    All this discussion of swings leads me to ask –

    What is the new Senate likely to look like/

    Secondly if Labour wins and the Senate blocks legislation what would be the impact on the senate of a double dissolution election.

  19. Possum Comitatus said

    Analysing the Senate possibilities….. that way lies madness! 😉

    I have absolutely no idea how the Senate will play out.Will people that have changed their votes since the last election hedge their bets by voting for minor parties in the Senate? Will Some marginal Coalition supporters do the same? Will ex-Coalition supporters that will vote for Rudd take out insurance by voting for the Coalition in the Senate? Will the swing to Rudd in the House be copied in the Senate? How will the preference deals play out since important groups like Family First and the Greens dont seem to pay much attention to how to vote cards?

    Because of the lack of regular polling on the Senate vote estimations – we cant really tell.And even if we had that type of regular polling, would you trust it?

    I’m pretty much bamboozled by the Senate possibilities.

    One thing a double dissolution does for Senate numbers is increases the possibility of minor parties ending up in the Senate as it halves the size of the quota needed to get in.Especially in those States that tend to elect 3 Labor and 3 Coalition Senators.

    The coefficient on those dummy variables are actually calculated by regression software rather then me giving them a value.What the software does is account for the structural change in the linear trend when those dummy variables have a value of 1 (which represents the periods when Rudd was leader) compared to when those dummy variables had a value of zero(which represents the months when Rudd wasnt the leader).It then calculates what effect that dummy variable (the Rudd leadership) had in changing the linear trend by telling us the size of the structural change in the vote trend that occured when Rudd became leader.Regression uses a process that minimises the sum of the squared errors, and that process suggests that the Rudd leadership brought to the ALP table an extra 2% of the TPP vote.

    That first ALP spike is actually in November 2005.I have no idea what caused it – but it must have been something pretty juicy for the polls to spike like that over 2 months (which would have been 4 to 6 Newspolls).

    Hi Leo,
    The problem with what your saying is that it isnt reflected in the data.We know that after the 2005 budget, the minor party vote went up at the expense of the coalition vote – about 5% worth of voters over the ensuing few months (all the while, the ALP primary vote didnt really move from its 2004 result).. We also know via ACnielson and Morgan (that ask for preference flows in the polling) that those voters that deserted the Coalition and went to the minors started preferencing the ALP.What we also know is that when Rudd became leader, the minor party vote dropped below what it was before the 2004 election, the Coalition vote dropped by around 5% as well and all of that from both those groups went straight across to the ALP as primaries.

    So it wasnt just progressive left voters (although some of them were certainly in there), as progressive left voters wouldnt have been voting for Howard before the 2005 budget anyway.It’s a different group altogether than made that shift.You can follow the patterns and a rough approximation of the numbers of how this played out here:


    The problem with 98 is that the One Nation vote affected both parties primaries in a way that hadnt happened before.I’m pissed off at One Nation – they’ve made modeling the longer term vote trends a hell of a lot more difficult than it would have been 😉

    2001 had the Tampa and September 11 which was another unique thing.The government primary vote moved 13 points in 6 months.That has only happened once in the 21 years of Newspoll data I use.Then it droped back 6 points in the next month.

    So the few months leading up to the 98 election and the Tampa period through to the 2001 election add alot of unique exogenous interference to the relationship between the primary vote and the TPP vote that isnt repeated outside of those two periods.

    The problem with softness is how you define it.If we take a page out of John Howards little red book of bizarre cliches and use the “common sense pub test” to define it literally as how likely it is that a person will change their vote between now and the election – we have that group of voters, about 5%, that moved from the coalition to the minors, that started preferencing the ALP, and then moved across to become ALP primary voters.

    How soft would the vote be of that group?

    Would people that deserted Howard years ago and have voted for anyone but Howard since be likely to move back to him?

    I cant see it.

    Some of that other group that moved directly from the Coalition to the ALP with Rudd (about another 5% or so) may well be soft by the common sense pub test, but that first group?

    I just cant see it.

    So the ALP TPP is already boosted simply as a result of that 5% or so of 2005 Coalition deserters (that I cant see going back due to their behaviour) not being able to deliver a preference vote elsewhere because their primary is given to the ALP.It’s effectively reduced the number of swinging voters by 5% and boosted the ALP primary by 5%, which in turn has boosted the baseline ALP TPP by 5%.

    I think the ALP TPP of 55 is more stable with a high primary vote, simply because it’s added 5% to its baseline with those 2005 Coalition deserters.As a result, its effectively minimised the size of the downside risk of the ALP TPP vote. vote.

    However, all of that said – I really wish the delightful Martin O’Shannessy would tack a couple of questions onto his Newspoll that sought to measure the strength of current voting intentions.ALthough if he did that – we’d all have one less thing to argue about! 😉

  20. Leopold said

    So how do you explain the substantial decline in the Green vote with Rudd in the job if not a return of Beazley-hating lefties to the fold? Are the Greens moving to Howard?

    And an intention to vote minor party and preference Labor can be as rock-solid as an intention to vote Labor first. To say that people who say they’re voting 1 for a minor party are ‘swinging’ while people who are saying they’ll vote one for Labor are ‘solid’ is still just assertion. Just to give an example of election results where people swung to minor parties and preferenced the opposition: in 1996, John Howard got a swing of just 3% on primaries – but that went to 5% after preferences. Ted Baillieu in 2006 got a swing of around 1% on primaries – which went to 3.6% after preferences. I’m sure there would be many others if I looked around. I don’t question Rudd has found 2% or so of middle ground voters Beazley hadn’t, though I question whether it will last into a campaign – but I do question whether you can make safe assertions regarding the level of a primary vote and the stability of a given 2PP position.

    On your last comment -I seem to remember either Graham Young or Mark Bahnisch (forget which, no doubt that would offend them both 😉 ) saying that research has found that ‘how likely are you to change your vote’ is an ineffectual question. Don’t know how true that is, but I’m inclined to think it would be.

  21. The Doctor said

    I checked Oz politics for Nov. 2005, as I thought it was WorkChoices passing thru Parlt. at the time. Interestingly, ACN did not record a similar spike.

  22. Possum Comitatus said

    The Green vote didnt move after the 2005 budget – it just stayed where it was, around the 6-9 percent mark.The Green vote didnt drop until Rudd came along, where some of those those Green voters shifted directly across to the ALP.So that post 2005 budget shift from the Coalition vote to the “others” vote had very little, if any interactivity with the Green vote.The Green vote moving to Rudd is a completely different group from those that deserted Howard at after the 2005 budget.

    Its not the Green shift that is rock solid – that shift doesnt really matter becuase those preferences will flow back to the ALP anyway.Its the post 2005 budget shift away from the Coalition to the “others” camp that started preferencing the ALP, and then shifted directly across to the ALP primary after Rudd that is the important 5% that has bolstered the hardness of the ALP baseline TPP. That group, that post 2005 budget “5%” group has moved from a Coalition primary, to a “others” primary AND started preferening the ALP, to shifting across to the ALP primary.

    For that group to be soft would mean that those voters would have to not only move back to the “others” vote from where they preferenced the ALP, but right back to giving their vote to the Coalition primary.They have to take 2 steps backward in their political voting history to move a TPP vote from the ALP to the Coalition.

    I dont know how accurate the results would be from polling questions on the solidity of voting intention – a lot of it would depend on the way the questions were framed.Ideally you would just ask the questions to those people whose current voting intention is different from from their vote at the 2004 election.That would slash the sample size compared to the rest of the poll, but if it was run over a couple of months worth of polls – it would provide enough data to give a decent estimation of the solidity.If you ran it as a moving 4 or 5 poll compilation you would get interesting trend data out of it as well.At least something that could then be data matched to various satisfaction ratings and derivatives thereof for modeling purposes.

    Thanks for that.I’ve been wondering about that spike for a while.Workchoices going through the parliament gives some sense to the spike.

    ACN doesnt seem to pick up the volatile movements that Morgan and Newspoll ofen capture.It’s a pretty sedate kind of series.

  23. Leopold said

    I understand you worked hard on that analysis of swinging voters, but I draw your attention to Bryan Palmer’s graphs of minor party support in the second half of 2006 and the decline since.

    In Newspoll, the Greens have slipped from around 7% to around 4%. In ACN, from 11% to 7-8%. In Morgan, from almost 9% to about 6%. So about 3% or slightly more. The ‘other’ vote has fallen almost 3% in Newspoll, about the same in Morgan and slightly above 3% in ACN. There’s been no 5% decline in the non-Green minor party vote, in any published poll.

    Labor has (as of now) gained 6-7% on the primary vote compared with Beazley’s last 3-4 months as leader. In raw numbers (which are necessarily a bit simplistic) around half of that came from the Greens, and half from ‘Other’. The Coalition is essentially back (now) where it was in late 2006 in terms of primary vote.

    And you still haven’t (IMO) shown the slightest evidence that someone who says ‘Labor’ when asked who they’ll vote for is NECESSARILY a more solid voter than someone who says ‘Green, Democrat, Labor’ or ‘Family First, Labor’. Labor’s primary vote has slipped at least 3% in the last 3 months; there ain’t nothing to stop them dropping another 3 points in the next 3.

    Beyond these points, I think we should agree to disagree. Cheers.

  24. Stig said

    Hi Possum,

    Glad you’re back, hope you enjoyed the break. On your observation that ACN is more sedate – I’ve noticed that it tends to have large sample sizes compared to many of the other polls. This woould tend to get rid of some of the statistical noise, I would think. It’s polling methods may also reduce comparative volatility somewhat?

    I’m with Economan on state-wise sample sizes. Workchoices is a lemon too, despite what the ACC says…

  25. michael said

    Workchoices is once again fingered as the culprit that engineered the Coalition’s demise.

    People don’t want it, they did not vote for it and they want it expunged. One of my theories about the high ALP vote is that people are saying we will give you the majority in both houses to get rid of this insidious legislation.

    They are not walking into a change of Government, they are venting their displeasure.

    The MSM and other hystrical supporters can rabbit on as much as they like, but the message is clear.

    Abolish Workchoices and restore the balance between work and family.

  26. Possum Comitatus said

    From May to Oct 2005, the non-major party primary went from 15.5 to to 21.5, the Coalition went from 46 to 40 and the ALP went from 38.5 to 38.5.Yet the ALP TPP moved over the same time from 47.5 to 50.5.The TPP not only moved according to 2004 preference distribution assumptions for the election, it moved according to Morgans respondents allocating preferences as well.

    That was the group of voters that moved from from being Coalition primary voters to ‘others’ voters and preferencing the ALP after the 2005 budget.

    Between then and November 2006 you had some noise, but at November 2006 the non-majors were on 21, the Coalition on 41 and the ALP on 38.Similarly the ALP 2pp was at 51.5, the Coalition on 48.5.Again, the “others” preferences were driving that ALP TPP vote.The Greens vote hadnt moved beyond the MoE since the 2005 budget.

    Then Rudd came along where the instant effect was to reduce the COalitions primary by 2 points, from 41 to 39 and to reduce the non-majors from 21 to 15 for the ALP primary to have a total gain of 8.But that non-major vote went right back to the level it was before the 2005 post budget Coalition desertion.

    Since then, the ALP primary has gained a couple of points – further at the expense of the others – but the Coalition primary has been stuck on 39-39.5 by monthly newspoll aggregates.So you’ve had a further voter movement from the ‘others’ to the ALP since then.

    The non-green others over the period of May to Oct 2005 grew from 9/10% to 13/14%.At November 2006 they were still at 13%.That others is now a consistent 9%.That difference between their 13s and 14s and their current 9s has moved to the ALP primary since Rudd was elected.

    The Greens since Rudd have dropped about 2-3 points which has moved across to the ALP.

    So we have 3 clear groups that have moved.The 2005 post-budget Coalition deserters that went from the Coalition to non-green “others” to Rudd.About 4 points, maybe 5.

    The 2 points that moved directly from the Coalition primary to the ALP primary under Rudd.

    And the 2-3 points that moved from the Green vote to the ALP primary under Rudd.

    All up that makes 8-10 points gain to the ALP primary since Rudd.From 38s to 40s pre-Rudd to 48 as the average over the last 6 months.

    As for the ALP primary slipping over the last 3 months – here’s the monthly Newspoll averages for the last 5 months:
    April 49, May 49.5, June 46 (single poll in June), July 47.5, so far in August 48.
    There’s no 3% movement according to Newspoll.

    As for soft voting – I’m of the view that the ALP vote TPP has been bolstered by the group that moved after the 2005 budget from the coalition to the others.For that group (and only that group) to move back to the Coalition primary requires them to take 2 steps back and nearly 18 months of behaviour to vote for the Coalition.The only time I can find in the last 21 years of newspoll history where an identifiable group of voters of around 5% or more moved from a long term voting behaviour rapidly to an oppositie position was Tampa/S11.

  27. Possum Comitatus said

    You could be right Stig.ACN have large samples and probably a weighting methodology that reduces the volatility.Whatever they doi it works – they’re the mogadon of Australian polling results.

  28. Paul C said

    A question I have wondered for a while now, and as someone who has only just started reading, I apologise if it has been answered before.

    Do the pollsters ask how people voted last election? Assuming that people answered correctly, I would have thought that would be a good way to reduce sampling risk – if your sample voted roughly 52-48 in favour of the coalition last time, then it’s more likely a representative sample. I’d love to see percentages of people who change their votes, and their reasons.

  29. Leopold said

    Okay, from the top.

    You are using Newspoll numbers from early-to-mid-2006 and November 2006, comparing them with Rudd, and saying presto, this is what happened. The polls shifted definably during the period from June 2006. Labor stopped polling 38 or so and started polling 41 or so in both News and ACN. The non-Green others dipped from the 13% in Newspoll you’re quoting to a sustained run at around 11%, and the Coalition ALSO dropped off 1-2% of their vote during this period. Labor and the Greens were up, the Others and Coalition were down.

    There was a spike in ‘Other’ (to 13%) and a dip in Labor support in November’s Newspolls – but that wasn’t shown across the other polls. ACN showed 42, 40, 41 between October and December for Labor, and an easing in ‘Other’. The Morgan poll could be argued to have shown a drop of about 1% in Labor’s primary but no more than that. The ‘Others’ had been easing down in ACN and Morgan before Rudd took the leadership. If there was a 5% ‘Others’ group to be discussed, a significant proportion of them must have already made the shift to Beazley Labor. Kevvie is currently tracking 6% above where Beazley was on primaries; the Greens and Others are both down about 3% on where they were before Rudd took over. I repeat: there is good reason to think much of the increase in Labor’s primary vote has come from people who were rusted-on Labor preferences anyway.

    As to your use of monthly Newspoll aggregations to deny movement since about April (four rather than three months, I acknowledge) – for heaven’s sake. Open your eyes. The aggregated numbers, whether calculated by Bryan Palmer or Reuters, are in unanimous agreement that Labor has gone from a peak of around 50 to about 47. This obsession with the Newspoll primary figures as the Word of God is unhealthy, Possum.

  30. Possum Comitatus said

    Hi Paul, no problem – welcome aboard the polling season 😉

    If the pollsters ask how people voted at the last election, they dont publish that information.I can understand why they dont, because the results could give superficially dubious poll by poll results.If, for example, 6 in 100 people have changed their vote from the last election, it only takes a very few people in the polling sample to inflate or deflate that proportion because of the small number of people involved.

    However, I wish that informaton were published because over a long enough time frame, the volatility in that movement could be mathematically removed which would be a real handy little piece of information to have.

    It’s not simply a matter of choosing arbitrary periods of time, taking numbers and comparing them to come up with conclusions.Go here:

    Follow the methodology.It measures changes by sums of incremental differences for party polling results since the last election.It allows voter movements to be tracked across time in terms of who went where and when.That isnt made up, it’s just maths.

    The polling numbers and what they mean in terms of voter movement for any given month since the election, and the differences between polling numbers and what they mean for voter movement for any given two periods of time since the last election are accurately described by that process of determining the swinging voter.If you use just raw polling observations and compare them across time – there are many different possibilities that could explain what is happening.If you treat the polling numbers as a coherent time series, there is only one explanation.The one I’ve described.

    Because of MoE issues, some of the movements may not be 5%, they may only be 4.But one thing is crystal clear – in the months following the 2005 Budget, a group of Coalition voters worth about 5% moved from being Coalition primary voters to being non-major party voters and started preferencing the ALP.That group then underwent a bit of polling noise and wobbled around but that pattern of Coalition primary voter desertion held right through until Novermber 2006.Then Rudd became leader and they moved across to the ALP primary – dropping the others vote level to where it was at the 2004 election.Since then that others has slowly leaked a few points to the ALP.

    There was no numerical evidence of that “others” group moving to Beazley in any statistically significant way.There’s significant evidence the others moved away from the Coalition post 2005 Budget.There’s significant evidence that nearly all of that Coalition deserter group plus a couple of points of the Greens moved across to Rudd.But there is no statistically significant evidence that the COalition deserters moved to Beazley.None, zip, zilch, nadda.A couple of points of volatile movement in that others group around August 2006 suggested a movement to Beazley might be happening AT THE TIME- but the 4 months following August killed that possibility off.That possibility was just polling noise.

    When you say “there is good reason to think much of the increase in Labor’s primary vote has come from people who were rusted-on Labor preferences anyway.” – there is no evidence for that.Around 5% points of Labors primary vote came from Coalition primary voters that deserted after the budget in 2005.

    The Government primary vote at the 2004 election = 46.7%
    Government primary vote according to Newspoll at October 2005 (yes 2005) = 40.
    From that 40, the government primary vote has never recovered and the only improvements have turned out to be illusion fuelled by polling noise.

    Yet throughout that time the ALP TPP was growing, but the ALP primary was not – remaining at slightly below 40.Then Rudd came along and the ALP primary grew, folding that ex-coalition 6.7% primary vote into their own as well as pinching an additional couple of points here and there.It was a crystal clear movement from the Coalition primary to ‘others’, then on to the ALP with the Greens playing very little role.The growth in the ‘others’ didnt come from the ALP primary, it came from the Coalition primary voters which then preferenced the ALP until RUdd came along whereby it went to the ALP primary.

    I cant put it any clearer than that.That’s how the polling time series itself stacks up.

    I use a monthly aggregated newspoll series as my primary series for modeling for very good reasons.Firstly, it produces a time consistent series that spans back 21 years but secondly, and most importantly, monthly aggregation reduces, for any given month, the MoE and volatility of the polling series.

    Was the May Newspoll that had the ALP on 52 as a primary really 52?
    Probably not – the real figure statistically could have anywhere from 49 to 55.
    Was the newspoll that had the ALP primary on 46 really 46?
    Probably not – the real figure statistically could have been anywhere from 43 to 49.

    What we do know is that the newspoll figures over the last 13 polls were:
    47, 52, 49, 50, 48, 50, 49, 47, 52, 46, 48, 47, 48

    What we can say from that data, using that data alone is that the most likely possibility is that the ALP primary is stable around the 48-49 mark over the last 13 polls, and most likely has been for the period of those 13 polls, and we can say that with a solid amount of statistically significant evidence.

    What we CANNOT say with any statistical significance is that there has been a 3% drop in the ALP primary according to Newspoll.

    If the ALP primary is really 49 and one fortnight the Newspoll gives a figure of 52, the next fortnight it gives a figure of 46 – the healines would scream “Howards recovery!”.The reality on the other hand would have been no change.

    The headline figures must be be accomodated by their MoE band.

    I dont particularly care what the Reuters weighted poll says – the weighting process, by its very nature (ostensibly a double moving average process), reduces the infomation contained within the series being modeled…. that’s the point of using it.It sacrifices information for smoothness to produce an aesthetically pleasing graph.

    So saying, the ALP primary may have dropped by a point or so, but I’ll continue to use statistically rigorous verification processes to inform me what the data is doing rather than pretty weighted 5x3MA graphs, as only one of those can tell me what is or is not statistical reality.

    Newspoll primary figures arent the word of God – but I use them as my primary modelling series because they are:
    a) time consistent
    b) regular

    Just because you dont like what the Newspoll data says is no reason to get shirty about it Leo 😉

    When the Newspoll data shows a statistically significant change in the ALP primary vote, I promise to put it up 😉

  31. Crikey Whitey said

    Looks like Glenn Milne is putting in a last ditch effort to influence the Newspoll responses. See Tin Tin goes to NY to satisfy his lascivious predelictions!

  32. steve said

    I think you will find that November 2005 was when workchoices was introduced into the Federal Parliament.


  33. Leopold said

    First of all, I know the Liberal vote was 40% in Newspoll in October 2005. It went back up again in early 2006 and stayed up until August, at around 42-43 or better. Then it quite suddenly dropped a couple of points and stayed down through to December, when another couple of points dropped off.

    Second, you’re assertion of a 5% fall in the Coalition primary vote after the 2005 budget only shows up in Newspoll. It’s not visible in either ACN or Morgan. If one pollster shows something and the others don’t, I generally tend to conclude the one pollster is wrong. Your entire analysis is dependent on a period in early 2005 when Newspoll tracked 3 points or more to the right of ACN in terms of Coalition primaries. I submit to you your analysis is therefore (at best) arguable.

    Third it has nothing to do with liking or disliking the Newspoll data. It has to do with applying my mind to all the available polling evidence and drawing conclusions based on logic. Galaxy was showing 49 in April and May, now 44-46. ACN was showing 48-50, now showing 46. Morgan, from having pro-Labor outliers of 52.5, is now showing pro-Labor outliers of 49.5. Newspoll has shown less movement than the others to date, but nonetheless I repeat: you’re in denial if you think the government is in as bad a position now as they were in April.

    And I am trained in econometric analysis thank you – well enough trained to understand the difference between a statistical model and empirical proof. 😉

  34. Possum Comitatus said

    Hi Leo,
    Over 2006, the Coalition primary held steady to a long run average of 41-42.As I said in the last reply about this movement over the 2006 period:”That group then underwent a bit of polling noise and wobbled around but that pattern of Coalition primary voter desertion held right through until Novermber 2006″.41-42 is not a recovery from 40 compared to the 46.7 they came off from the last election, nor from the 46-48 they were on before the 2005 budget.

    ACN and Morgan both reflect exactly what happened in Newspoll.The Morgan Monthly aggregate particularly so (with its -1 point systemic difference from Newspoll for the Government primary and its +1.5 to +2 point systemic difference from Newspoll for the Oppositions primary vote).Likewise, ACN over the period followed exactly that path within its MoE.

    That can be seen clearly at Bryans site:

    ….. without me having to post the numbers unless you’d really like them.

    We have 3 pollsters all showing the same thing.

    As for the May-AUgust differences.

    Galaxy in Apr/May period had the ALP primary (taking into account the MoE) at 46->52, 41->47, 43->49.
    Galaxy tracks the ALP primary lower than the other 3, and the Coal primary higher… the opposite of Morgan.Their latest July has them on 41->47.

    ACN had in May for the ALP 45->51.Currently the ACN has the ALP at 43->49.Suggesting no movement between 45 and 49.

    NP had in may for the ALP 4 polls headlining at 47->52 for an MoE band of 44->55.Taking out the rogue 27-27 May poll it was a band of 44->53.Currently NP has the ALP at 45->51.Suggesting no movement between 45 and 51.

    Morgan had 4 polls in May, with an overlap MoE band of 48->52.Currently Morgan has the ALP on 46.5->52.5.Again, no movement between 48 and 52.

    So we have Galaxy at the top between a stable 43->47.
    ACN stable between 45 and 49 with a primary of around 48 fitting all polls.
    NP stable between 45 and 51 with a primary of around 48 fitting all polls.
    Morgan stable between 48 and 52 with a primary of around 49 fitting all polls.

    Galaxy, like always, is the lowest for the ALP. ACN and Newspoll reflect each other like always and Morgan is best for the ALP as always.

    What you call movement is empirically nothing more than random walks within the MoE according to the available data.

    As such, I can only believe what the data tells me.

  35. wpolscemamymocneseo said

    You completed certain good points there. I did searching on the subject matter and found most persons will go together with your blog

  36. weight loss chart kids

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: