You have seen how one, possibly (probably?) “rogue” Newspoll shifted the reporting landscape. Suddenly Kevin Rudd is beatable and actually quite a few commentators now know precisely what he is doing wrong politically and why.
This makes a welcome change for the media consumer, and will probably do him good as well.
ACNielsen in Fairfax papers tomorrow. All eyes on voting intentions, but leaders’ approval ratings will also be interesting. I reckon much of this measure these days is respondents’ judgements about how politicians are doing politically, so a significant drop in Rudd’s approval and/or rise in Turnbull’s might reflect the huge publicity given to last Tuesday’s Newspoll.
Or it might simply indicate dissatisfaction with his handling of asylum seekers – about which there will presumably also be questions, results perhaps not published until Tuesday.
I wonder why Newspoll is always regarded as THE significant poll while Morgan and Essential are Tagged Pro Labor? Isn’t Newspoll actively Pro LIberal? ANd published in a Murdoch paper?
So which is an impartial poll and is there such a thing?
71 of the 90 Essential polls since the 2007 election have had Labor on 58% 2PP or above. Similar proportion amongst Morgan polls. Nielsen and Newspoll have had Labor on or above 58% only a third on the time. Labor has never polled above 53.2% in its history. Labor polled in the high 50s all through 2007 and ended up polling 52.7%. You tell me why Essential’s consistently high Labor vote should be trusted.
So why does this happen?
Do Essential and Morgan poll different areas to Nielson and Newspoll?
Or to a differnt demograhic?
Specific suburbs?
Self selected or random selection?
I do not necessarily trust Essential research, but neither do I trust Newspoll.
And in the previous fortnight weren’t Newspoll and Essential really close?
And aren’t pre – elction polls ( Those close to the real election)
apt to go to a more even 50/50 approach.
Have now seen the results of Nielson 56-44 hardly a seismic shift.
So Morgan , Essential, and Nielson all tell the same basic story -so the question remains. How do you explain last weeks Newspoll?
Essential possibly have a skewed ‘community’ (mentioned here http://mumble.com.au/?p=302). Morgan’s face to face favour Labor, but their phone polls are more in line with Newspoll, Nielsen and Galaxy. So probably something about the type of people who are letting that nice man/woman with the cards into the door.
You have highlighted the fact that methodology leads to different results. That’s why the trend of each of these polls should be the focus and not the raw numbers. Following last week’s Newspoll, I would have expected the figures from the Morgan & Essential polls to show falls from their very high levels of support for Labor. Say from 58%/59% to 54%/55%. It didn’t happen. And the Neilsen poll [whose methodology is I understand similar to that of Newspoll] backs up the absence of a quantum shift in ALP support. I think Crikey’s Bernard Keane nailed it in yesterday’s comment about the OZ’s penchant for setting agendas.
And for the first time I heard someone [Barrie Cassidy on the Insiders program] state in very dismissive tones that some people believe the figures are manipulated. Assuming that there is no manipulation [and I take Peter's point on this, made under another heading, that there isn't any] why would anyone want to publish a poll that that appears internally inconsistent? You know, a big fall in the ALP primary vote, big increase in the Coalition vote, but no change in the Greens and “other” votes, and virtually static figures in PM preference and performance stakes. The messy Oceanic Viking issue would surely have affected the other figures. It doesn’t comprehend. Why didn’t Newspoll query this before supplying the figures to the OZ? Why not go back and do the poll again to avoid the damning term “rogue poll”? Because I think the OZ saw the poll result as one that suited a particular agenda. This is the point that Keane was making.
JA: the pollster has to publish, whatever the result. It’s not up to them to second guess their data. And there is no time to redo it (or do another one as well) in time for Tuesday publication; and there would be the added expense.
Every other pollster would do the same. Nielsen published similar survey results last year.
Yes I suppose it is a case of publish and be damned. Mind you, I don’t think cost is an issue given the number of times Newspoll has been published outside the fortnightly cycle. Timing is based on when it suits the OZ.
The scepticism about the poll however has I think made the OZ somewhat defensive about the way it covers matters concerning the Rudd govt. Note this morning’s article by Elliot and van Onselen defending the paper’s approach. Top of the pile in Breakfast Politics. It points out that Rudd swears [shock horror!] and is thin skinned, the Fairfax media are pussycats when it comes to reporting the activities of the Rudd govt, how tough it was reporting the AWB scandal, & for good measure, quotes editor Chris Mitchell who despises Rudd. It was pathetic. Sadly the OZ is damaged goods, losing credibility by the day. It needs a new editor and owner.
[...] electoral event, or unexpected opinion poll, come the Reasons. That blippish 52 48 federal Newspoll a few months ago delivered quite a few op-ed Reasons for Kevin Rudd’s [...]