Morgan face-to-face: 53.5-46.5 to Coalition

The final Morgan face-to-face poll for the year is a surprisingly good one for Labor, producing a set of primary vote figures that scans out to 50-50 on two-party preferred if minor party and independent preferences are allocated in the same proportions as per the 2010 election result (the method used by all pollsters other than Morgan, with good reason). Labor’s primary vote is unchanged on the previous poll at 36.5 per cent, but the Coalition is well down from 46.5 per cent to 43 per cent, with the Greens up three points to 13 per cent. The poll was conducted over the past two weekends, from a combined sample of 1950.

Morgan’s headline two-party figure of 53.5-46.5, which is obtained by asking minor party and independent supporters to whom they would direct their preference, is quite a bit less impressive for Labor, and underscores a very odd feature of Morgan’s face-to-face polling this year: their respondent-allocated preferences have been splitting about 50-50 between Labor and the Coalition. Not only is this entirely unlike any election result in recent history, it also flatly contradicts the other pollster which publishes a respondent-allocated two-party result: Nielsen, which has given Labor a mid-60s share of minor party and independent preferences in each of its last five monthly polls, consistent with what election results would lead us to expect.

For this reason, Morgan’s 50-50 result is obviously a more meaningful extrapolation from its primary vote figure. However, this brings us to Morgan’s other problem: that those primary vote figures appear to be consistently biased to Labor. Labor’s 36.5 per cent in both the past two polls compares with 29 per cent from the most recent Nielsen, 31 per cent from Newspoll and 35 per cent from Essential Research (which tends to have Labor higher and the Greens lower than the phone pollsters – the limited evidence available suggests it is Essential which is more accurate).

Finally, the Poll Bludger will be going into hibernation from Christmas Day until January 8 inclusive, which is to say the site will still be accessible but comments will be closed. Thanks to all (well, almost all) who have contributed throughout the year, and I look forward to hearing from you again in a fortnight or so.

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.

2,200 comments on “Morgan face-to-face: 53.5-46.5 to Coalition”

Comments Page 42 of 44
1 … 41 42 43 44

  6. Happy Christmas all, and HNY.

    Thanks again William, and all the residents here.

    I’ve been coming here since 2007. I did not have much spare time during 2011, but I read-up on PB when I can. I continue to learn from the residents here, the exchange of ideas and thoughts (plus some uncensored abuse) all help me along my own path.

    See you on the flipside!!

  13. Let’s hope that Castle is right and

    That’ll be the end of Mr. Rabbit.
    Now he’s developed that bad habit.
    He’ll finish up all petered out
    And go off with a whimper, not a shout.

    Happy Christmas everyone! With special thanks to William for providing such a great forum!

  14. [This little black duck

    Posted Saturday, December 24, 2011 at 9:02 pm | Permalink

    Ch Nein

    Neigh sayers?
    ]

    Very old aus.tv usenet refereference – Ch Nine was always referred to to as Nein cos of their atitude to programming.

  17. William said:

    I think if you’d used previous-election preferences it would look more or less as I anticipated – with the Labor bias in face-to-face polls disappearing when Labor’s vote collapsed after the carbon tax, but reasserting itself more recently.

    Well, if you use the previous election prefs it changes a data point or so but it doesn’t change the trend. I just happened to take a look at both.;-)

    See http://www.rodhagen.id.au/Aussie_Pollitics/AusPol_bits%26pieces.html for the comparison.

    If nothing else it suggests that the long standing shibboleth that Morgans face to face always substantially favours Labor more than their phone polls hasn’t held up for most of this year. It may be that Morgan have changed their processes (different sampling, perhaps) or it may be that voters are now behaving differently in polls. I have no objective reason for pointing one way or the other, but my gut feeling is that the later is more likely than the former. I’m sure you are in a far better position than me to ask Morgan whether they’ve changed anything.

    The last available comparison using the “last election prefs” that you prefer, rather than suggesting that the old shibboleth is ‘reasserting” itself, suggests that Morgan face to face now favours the Coalition, not Labor , when compared to their phone polls.In fact, from something like a whopping 6% plus bias towards Labor in Feb 2011 we are now looking at about a 1% bias towards the Coalition when we compare Morgan Face to Face with Morgan phone.

    Unless you can fill me in on some substantial changes in Morgan methodology, I reckon this points to a real change in voter behaviour when dealing with pollsters face to face.

  18. Many thanks to William and all posters for making this blog such an enjoyable and stimulating place to be over the past year.

    Merry Christmas to all PBrs and their families.

  19. Interesting that I’m to the right of most people on here in regards to the economy.

    That’s probably because I believe in a market economy (with a social safety net grafted onto it to protect the needy and the vulnerable) and that people should take responsibility for themselves rather than whinge when the government cuts services. That is also consistent with supporting the ALP as they’ve been responsible for the largest reductions in taxes and government spending in this nations history.

    I suspect many Abbott supporters (including Abbott) would find themselves to the left on the economy as they’re big government conservatives. Read Andrew Norton for more on the subject.

  20. [The dogs eat the shells and heads.]

    Being inland, not a lot of fresh prawns around here. But a question for dog keepers: It’s said that one should not feed cooked bones to dogs. The bones should be raw.

    The local IGA sells roast chooks, which are discounted to around $3 after a day or two.

    I religiously strip the meat: dogs love ’em. Is this an old wife’s tale? Will cooked chook bones hurt them? Just to get back on topic: Newspoll. There.

  21. I used to live at Gawler SA, which had in its district the market gardening town of Virginia.

    I had a secret (and unfulfilled) ambition one Christmas season to add a poster to their town sign posts proclaiming,
    [Yes Santa, there is a Virginia!]

  23. Rod, my next problem is that the Morgan phone polls you’re using have small samples (500-600) and are therefore going to be erratic in any case. I would be more interested to see how Morgan face-to-face polls compare with a combined poll trend, or failing that with a more regular and less flaky series like Essential or Newspoll.

  24. janice2

    Thank you for your kind wishes. I got my chickens as 1 week old chicks, they are now 8 weeks old and such characters I love sitting out in the garden watching their interactions. I have one that seems to have a bit of a poor memory and keeps forgetting that she can get in and out through the open door and doesn’t need to pace up and down the coop squeeking at her friends in or outside the coop.

  25. I’m not for a moment suggesting that Morgan Polls are as good as they should be, William.

    I do think, though, that we can’t simply trot out the line that Morgan face to face “primary vote figures appear to be consistently biased to Labor.” any more.

    If anything the evidence this year suggests that their face to face polls now treat Labor more harshly than their phone polls.

  28. [ I would be more interested to see how Morgan face-to-face polls compare with a combined poll trend]

    I thought Possum did just that?

    Besides, what’s far more interesting than the current polls is where they’re going to next year.

    Now one of the reasons I come to Pollbludger is in the hope some people have got something insightful to say about the underlying mechanics of why people poll the way they do and why they vote the way they do.

    In other words, some good working theories that might have predictive ability..

    Oh well. .one can dream.

  30. Rod, I’ve gone to the effort of plotting the difference between ” rel=”nofollow”>Morgan face-to-face polls against the admirably solid Essential Research series, and while it’s up and down like the Assyrian Empire, it’s broadly U-shaped in the fashion I’ve been suggesting – in the 10 observations after the carbon tax announcement Morgan had Labor higher than Essential in five but lower in the other five, but in all the 15 observations since Morgan has been higher. Averaging results for all polls this year gives you the following, with ALP primary vote first and two-party vote second:

    Morgan F2F: 35.9%, 47.8%
    Essential: 34.1%, 45.9%
    Newspoll: 30.7%, 45.3%
    Nielsen: 29.5%, 43.2%

    Note that Essential is a lot kinder to Labor than Newspoll and Nielsen on primary but not on two-party – this is because it has the Greens a fair bit lower. As I’ve noted frequently, it’s very likely Essential which is getting this right. Phone polls have consistently been overrating the Greens at recent elections. As a result, I’ve never believed that Labor’s primary has been in the high 20s, as a number of polls have suggested.

  31. Scringler,
    Let’s put it this way, you could feed the cooked chicken bones to your dog for ages with nothing happening or never. But one day an bone is not digested because it is cooked and it gets stuck inside somewhere or pierces the bowel. Much pain, an emergency operation and then dog doesn’t make it anyway.

    Keep getting the bones out. Make sure you get all the little rib ones.

  34. [ I’ve gone to the effort of plotting the difference between Morgan face-to-face polls against the admirably solid Essential Research series, and while it’s up and down like the Assyrian Empire, it’s broadly U-shaped in the fashion I’ve been suggesting]

    Yep, definitely something in there..

    Again, would love a working theory 🙂

  35. suggest people catch Debt Defying Acts on ABC iView when it goes up. Very funny. Highlights are:

    Rupert Murdoch as King Lear
    Julia and Kevin duet
    Bob Brown and SHY selling out
    Barry O’Farrell medley

    thankfully, very little of :mrgreen:

  40. Thank you William and all who make this blog such an addictive read. See you all, after the break, for the next instalment in power games, Australia.

  46. Two different questions here, William. One involves a possible systematic pro Labor difference between Morgan generally and other polls.

    The one I’ve been focused on, though, is more confined, and looks at the matter of whether Morgan face to face favours Labor when compared to their phone polls. SSpencer & Chris Kenny have been making great play on the latter presumed difference in various places over the last few days (occasionally citing you as their authority along the way).

    I’ve simply been trying to point out that this Morgan FacetoFace/ Morgan Phone Poll bias just ain’t true on the evidence from this year.

  47. Wow William, there’s a good pattern there.
    Honestly, I wonder what a proper sampling of the entire electorate would show.
    And William, thanks again for the wonderful forum you provide. I hope all goes well for you in your PhD quest.

Comments are closed.

Comments Page 42 of 44
1 … 41 42 43 44