Menu lock

Polling

Jan 7, 2010

The Liberal Party Disaster in the Capital Cities

Think back over the last two years of the Rudd government and keep in your thought orbit the never ending commentaries of Labor facing some “political crisis” or  “

Possum Comitatus — Editor of Pollytics

Possum Comitatus

Editor of Pollytics

Think back over the last two years of the Rudd government and keep in your thought orbit the never ending commentaries of Labor facing some “political crisis” or  “increased pressure” or dangerous “political test” which was inevitably going to end in failure. Think back on the multitude of dour proclamations by the usual suspects – for some reason routinely published in The Australian – that the Labor government is/was/always will be on the precipice of some public revolt.

Now compare that….. quality insight …. to the reality – using Newspoll quarterly data (click the charts to expand)

primstpp

Yeah, I know – why do these people bother?

Anyway, while that chart says a lot – such as the absence of movement telling an important story all of its own – it also hides a lot of seriously major things happening underneath the headline figures, things which we’ll have a bit of a squiz at over the next few posts.

First up, we’ll look at the capital cities, as it brings up a question worth pondering: Who do you think would be more politically popular in the capital cities – Malcolm Turnbull or Tony Abbott?

While we don’t have the answer for that yet – we’ll find out in due course –  if Turnbull is more popular, to the point where the capital city vote for the Liberal Party is higher under a Turnbull leadership then it ordinarily would be, here’s a chart to send shivers down the spine of Liberal supporters everywhere. It’s the latest Newspoll quarterly results of the Coalition primary vote, by state and geography.

lnpprimsq4

Around 1 in 3 voters in the nation’s capitals would have given the Coalition their first preference were an election held in the last 3 months of 2009 – which in this case would be the Liberal Party since the Nats don’t really run in the cities.

If Turnbull was holding that up, it will be nightmare territory if it drops further.

However, for those that answered Abbott to the question – there’s some evidence to support the view.

The capital cities have been a big problem for the Libs generally, but Turnbull particularly, since 2007. If we track the Coalition vote in the capitals since the last election, we see Brendan Nelson started increasing the Coalition voteshare after the election washout – although some of that was inevitable as the “honeymoon” started to unwind. Yet under Turnbull, the Coalition capital city vote started to erode after his first three months as leader (2008Q4 figure) and never really stopped.

libcapcit

I suppose the big question for that chart and the Liberal Party capital city vote generally, is how much if any of the erosion was due to Turnbull and how much if any was due to the circus of the broader party and the Nats. I’m sure Minchin and Turnbull would have two completely separate answers on that.

While the Libs are having trouble in the cities, the Labor party on the other hand are pretty stable.

alpcaps

The Labor primary vote appears to have stabilised in the capitals between 45 to 47, giving a two party preferred of 58-60.

Something the Coalition might want to pay attention to federally is the way the Queensland conservatives managed to play themselves out of electoral success for 20 years. The Coalition (and later the LNP) in Qld allowed Labor to dominate the Brisbane city vote in much the same way that Rudd is doing now with capital cities across Australia.

Labor cannot lose while it has a strong city vote. But worse, the stronger Labor becomes in the cities, the fewer metropolitan representatives the Coalition ends up with in a given Parliament – forcing the policy and leadership choices the Coalition takes to any later election being mostly designed and supported by non-metropolitan interests.

That generally alienates any metro voters that have even a slight interest in political modernity – which usually happens to be most of them.

Lawrence Springborg as LNP leader getting flicked the bird by Brisbane voters for 3 elections on the trot now is a good case study. If the Coalition loses too many metro seats this year, their political problems will have only just begun.

The other interesting thing about the capital city vote is the way the increase in the Greens vote since the last election has been a part of a broader voter shuffle, with voters moving from the ALP to the Greens and from the Libs to the Greens in roughly the same proportions as Greens preferences were distributed in 2007 – which is around 75% of new Green voters coming from the ALP and 25% coming from the Libs, give or take a few percent.

Both Nielsen and Morgan in their polling ask voters how they would distribute preferences in order to get a two party preferred result, as well as distribute just the primary vote estimates on the basis of the 2007 election result. Ultimately, the difference between the two distributions for both pollsters is not statistically different from zero, even though the Green vote has increased from around 8% to as high as 13% and 14% depending on the poll. That’s very handy, because knowing that the Greens growth is coming from both parties in the same proportion as it historically flows as preferences, allows us to use the Newspoll results (which allocate prefs on the basis of 2007 rather than ask respondents) in the following way.

If we look at the ALP leads in the primary and two party preferred vote and then take the difference between the two, this is what we get:

alpcitygap

That “gap” between the two leads is very stable at 7 to 8 points regardless of the size of the ALP two party preferred. This suggests that in net terms, there is close to a 1 for 1 movement in swinging voters between Labor and the Coalition.

If a substantial number of Coalition voters were ‘parking’ their vote with the minor parties, we would see that come out in the Morgan and Nielsen figures with a lower ALP two party preferred vote when respondents allocate preferences themselves compared to the two party vote we get from allocating prefs on the basis of the 2007 election results – since more minor party respondents than usual would be preferencing the Coalition over Labor. But since we see no difference between respondent allocated and 07 election allocated preferences, it suggests that there isn’t any “Parked Vote” phenomenon at play in net terms.

That means that votes gained or lost by the majors are generally coming from the other major party in net terms – which will have some pretty important consequences in the election campaign when it comes to voter targetting.

We recommend

From around the web

Powered by Taboola

28 comments

Leave a comment

28 thoughts on “The Liberal Party Disaster in the Capital Cities

  1. Predicting the Election « A Senex View

    […] 2007, Kevin Rudd’s Australian Labor won the election with 52.7% of the Two Party Preferred (2PP) vote.  The exact same figure John Howard’s […]

  2. Tad Tietze

    Thanks for the clarification, Poss. I suspected that you may be inferring from the preference flows. Of course that carries an assumption that voters of both the ALP and the Libs tend to send preferences back to their “old” parties when they switch to the Greens, but that seems reasonable enough.

    It underlines that a serious, targeted strategy by the Greens to hunt for Liberal votes would probably be more harmful than helpful as ALP voters continue to represent the biggest pool of potentials.

    It’s sad that the pollsters don’t ask questions like in the Australian Election Survey that ask about who you support now and then who did you vote for last time. It might stop pundits making silly assertions about how changes in the breakdown reflect underlying shifts (which may be more complex). But then again, the MSM pundits seem too addicted to making silly statements that fly in the face of obvious facts!

  3. peach1

    shark 1975

    Re the Gold Coast, it is god’s greatest waiting room in Australia. The impact is described in my preceding post.

    PS. I am not joking. I pity the young people who will have to suffer trough this, being affected by people whose vision of the future is the next meal break.

  4. peach1

    The problem with this country is that pea brained women and the lot in god’s waiting room are allowed to vote. As was pointed out this lot kept Johnny the mean spirited, mendacious and obnoxious in power for so long.

    Heaven help us as this lot increases in numbers through the dumbing down of one and the aging of the other.

  5. Albert Ross

    >>And I should say of course my effective preference will still go to the ALP

    A lot of Green voters would prefer not to number past 1 or at least exhaust before the big parties at the next Federal election which is why, I imagine, the Another Liberal Party seems reluctant to bring back OPV for the lower house and eliminate ATL voting in the Senate with its egregious back room preference deals which allowed the likes of Fielding to emerge from under a stone.

  6. Chasing the Norm » Blog Archive » Can Nerds deliver wedgies?

    […] alienate labor from some rural electorates, but could help divest a lot of economic liberal voters/few remaining city voters from the Coalition to Labors waiting hands. (I’d be keen to hear any readers suggestions of […]

  7. Barking

    Re the ‘actual’ green vote. We do have two fantastic real polls coming up. Tassie and South Aus will give us some real stuff to look at, in particular the rural/regional/capital city stuff. i have this feeling that the Coalition types are in complete denial, they keep saying things like, ‘Oh we don’t think the polls reflect whats going on out there, and in the bush we are going to do better than people expect.”
    The SA and the Tas polls will give us lots of info to cross check some of the issues raised here. Cracker of a post Possum, but then what else would we expect.

  8. Jack A Randa

    And I should say of course my effective preference will still go to the ALP

  9. Jack A Randa

    Poss and Doug (way back at 2) – speaking as a sample of one I can say that yes, definitely, in this sample 100% of the shift from ALP to Greens has been the result of dissatisfaction with the response to global heating-up

  10. Possum Comitatus

    Lacoon went:

    [Was the”Parked vote” phenomenon evident pre the 2007 election? ]

    It was for the ALP between about 2005 and November 2006 – where the Greens and ‘Others’ vote inflated, giving the ALP a rising TPP vote, but from relatively low ALP primaries. When Rudd came along, that transformed into a strong ALP primary feeding a stronger ALP TPP, ultimately leading to the election result.

    As for the question about the Dems vs. Greens in terms of “parked vote”, that’s an interesting question – but alas occurred before I was paying attention to this sort of stuff in detail, so I can’t really give an answer that means a great deal!

  11. Possum Comitatus

    My Say and Don – yes, Tassie – bless its heart (and home to my favourite city in the world – Launceston) – is left out.

    The major pollsters don’t break down their numbers into State based figures that include Tasmania because the number of Tasmanian respondents in any national poll with a sample size of 1100 or so is usually 30 or less. The margin or error on a sub-sample of 30 maxes out at around 18%!

    That means any figure we get is a bit pointless. Imagine a poll result which says “Labor has the lead in Tasmania on a TPP basis of 54/46” with an MoE of 18%! We’d all laugh!

    If you run it through the Poll Cruncher:

    http://blogs.crikey.com.au/pollytics/thepollcruncher/

    … you can see it for yourself.

  12. Possum Comitatus

    Tad Tietze (cracker nic BTw!)

    The big piece of data comes from analysis of the two party preferred numbers from Morgan and Nielsen. Both pollsters produce two sets of TPP numbers – the first from asking poll respondents how they would allocate preferences, the second comes from allocating preferences for their primary votes based on the way preferences flowed at the 2007 election.

    The difference between both sets of allocations from both pollsters are statistically not different from zero – which, when combined with the Greens currently getting a higher vote than they achieved at the 2007 election, suggests that new Greens voters are coming to them from the ALP and the Libs in roughly the same proportions as their preferences flowed in 2007.

    If those new Green voters were mostly ex-ALP or mostly ex-Lib, we’d expect to see their preference flows reflect that with a point or 2 higher vote for either the Libs of the ALP on the respondent allocated TPP compared to the 2007 election allocation – but we aren’t.

  13. Laocoon

    Possum

    Was the”Parked vote” phenomenon evident pre the 2007 election?

    And also, if one were to have a vew that Greens are more left than Labor, versus the Democrats being in between Labor and Liberal, would one have seen more “Parked votes” in the “old days” when Democrats were the 3rd choice compared to now?

  14. Cuppa

    No complaints here. It’s all good.

  15. David Richards

    It’s not good for them, is it George?
    A spiral reminiscent of the Ouzelum bird.

  16. george

    “…the stronger Labor becomes in the cities, the fewer metropolitan representatives the Coalition ends up with in a given Parliament – forcing the policy and leadership choices the Coalition takes to any later election being mostly designed and supported by non-metropolitan interests.”

    Ouch! Catch twenty-two anyone?

  17. don

    [email protected]:

    [poss is there one captital city you have forgotten.]

    You’re right, Tassie has been left off again! And what about voters in the NT and ACT?

  18. my say

    poss is there one captital city you have forgotten.

  19. Tad Tietze

    Poss,

    You write: “The other interesting thing about the capital city vote is the way the increase in the Greens vote since the last election has been a part of a broader voter shuffle, with voters moving from the ALP to the Greens and from the Libs to the Greens in roughly the same proportions as Greens preferences were distributed in 2007 – which is around 75% of new Green voters coming from the ALP and 25% coming from the Libs, give or take a few percent.”

    Intuitively this seems right to me (it’s historically true also), but what data have you used to come to this conclusion? Is there some aspect of the polling data that would tell you this? (Maybe I’m missing something obvious?)

  20. David Richards

    Combined with the age demographic problem, this capital city problem should have
    Liberal Party stratagists slashing their wrists.

    Allied with the decline of the National Party, the coalition no longer has the easy road to the Lodge that it once enjoyed, and now has to earn the treasury benches. Are they up to it? Or is the Liberal Party ultimately doomed? The two problems facing them that Possum has identified exacerbate their other problem – lack of funds. If the perception among its erstwhile backers that they are no longer worthy of receiving patronage is increased by such problems, will they have the resources to fight elections effectively?

  21. zoomster

    Always happy to learn, poss.

  22. Possum Comitatus

    Zoomster,

    What’s interesting with the Greens vote is that regardless of the pollster and their methodology, they all have the Greens running at a long term average of around 10%.

    Some have it a point down over an arbitrary number of polls, some have it a point up – but it all averages out pretty consistently – suggesting that the Greens have grown their vote by a few percent with most of the variability being slightly on the upside rather than the downside.

    Shark,

    Unfortunately we don’t get that level of geographic breakdown published by the pollsters. All we have to work with in most cases is Capital vs Non-Capital City vote in the demographic cross tabs – and then only Nielsen publishes that with every poll, Newspoll publishes it every three months and Galaxy will sometimes publish it.

    Essential Report doesnt, which is a pity because they have some serious sample sizes happening. I can’t remember Morgan every publishing breakdowns that way.

    That said, even if the pollsters did publish that level of regional breakdown, the sample sizes involved in.. say … just the Gold Coast would be so small that the margin of error involved would make the result for the Gold Coast pretty much meaningless.

  23. shark1975

    Very, very interesting!

    Are you planning to undertake the same analysis for (say) the top five significant but non metropolitan provincial locations? I’m thinking Gold Coast etc, non capital but significant population.

  24. zoomster

    I thought whether or not the Greens vote has increased was still largely a matter of conjecture, as newspoll changed its polling to deliberately favour the Greens since the last election.

    Until we have another election, we don’t know whether Newspoll’s new methodology is accurate.

    From memory – I think I had a look at them before Christmas, and have lost a few braincells since – Morgan and other pollsters don’t show the same strength for the Greens that Newspoll does (Morgan’s figures swing widely for the Greens, but the basic pattern seems very similar to before the last election, with perhaps a slight increase, 1% or so).

  25. Possum Comitatus

    Doug,

    What ever the issue(s) may be for the increased Green vote, we unfortunately don’t have enough of the type of polling data required that would let us know. Asking why respondents are voting for a particular party isnt something that Oz pollsters focus a lot on – and being a rather difficult exercise in and of itself doesnt help!

    Jimmy – sorry about that. It’s on my list of things to do before Monday. I havent got the Vic or SA figures updated yet either (though I now can as Newspoll has published the Vic numbers on their site).

  26. JimmyD

    Poss, I know this is completely unrelated to this post, but you haven’t updated the NSW state govt polling sidebar for the Nov-Dec Newspoll.

  27. doug hynd

    Does your comment on the lack of evidence for minor party vote parking imply that the increased Green vote is being driven by some specific factors – like discontent over response to global warming?

  28. Tweets that mention The Liberal Party Disaster in the Capital Cities – Pollytics -- Topsy.com

    […] This post was mentioned on Twitter by Jeff Waugh, Pollytics. Pollytics said: The Liberal Party Disaster in the Capital Cities http://is.gd/5Q1sv A closer look at political dynamics in the nation's capitals […]