by frog
So, a TVNZ poll out last Sunday had National 8 points ahead. A TV3 poll out today has Labour 9 points ahead – 45% plays 36%. That’s a seventeen-point swing in a few days. Now, the TVNZ and TV3 polls are reputed to favour National and Labour respectively – but that doesn’t begin to explain this seventeen-point discrepancy. There’s volatility here that appears quite unprecedented. What on earth is happening? I half expect the political pundits in the Press Gallery to throw their hands in the air, say “God, I don’t know”, give up, and go home.
The best theory I can come up with (and, I stress, it’s just a theory – probably with little statistical merit) is that National got a couple of very large spikes in the polls off the back of its tax cut and race announcements, but that these advantages quickly evaporated. You can imagine voters hearing the National message, saying “yeah, that sounds great”, then thinking about it for a while and concluding “nah, I prefer the Labour lot”. If this theory is right, one would conclude that the political environment is fundamentally conducive to a third Labour term, but that National has been successful at scratching a couple of itches which give it short-term poll boosts that soon disappear.
By the same token, I wouldn’t be at all surprised if this polling madness continued, and we had hugely vacillating leads for both Labour and National for the remaning days of the campaign. I’m genuinely bamboozled. I guess we’re probably going to go into election night not quite knowing what to expect.
As for the third parties: the Greens are up 1 to 7%, a heartening result, while NZ First is down 1 to 5%. I remain convinced that the Greens are the only party sucessfully avoiding the third party squeeze.
TV3 also asked some other questions which were revealing about the mood of the electorate. Asked whether they consider Clark and Brash capable leaders, 85% said yes for the Labour Leader and 49% said yes for the National Leader. Asked whether they consider the two “rather inexperienced”, 9% said yes for Clark and 69% said yes for Brash. These are Brash’s two problems: next to Clark, he seems neither competent nor experienced enough. Asked whether tax cuts would be the issue that decided their vote, 85% said no and 14% yes. Taken together, these results make pretty sober reading for National. However, I have to say, all this volatility makes me very gun-shy about making bold pronouncements about this data
Now that National has declared a favoured coalition partner in United, we have two quite distinct teams to compare: Labour/Progs/Green vs National/United. In the TV3 poll, LPG leads NU by 52% to 37.5%; on frogblog’s averager LPG leads by about 8 points, 49.86% plays 41.90%. Now that we know the configuration of the two prospective governments, I’ve added two more graphs to the info I upload with each poll update: showing how much of the party vote and how many seats each “side” has according to the polls. At the moment, LPG have 60 seats and NU have 50. Enjoy.
![]()
Published in Campaign by frog on Wed, September 7th, 2005
Tags: environment






on the trolls and those who are unable to keep on topic
Given, that polls could influence the election. Especially polls so close to it. Are there any objective standards that polls have to be judged by. Imaging if the last poll before the election showed the greens below the 5% mark?
Like or Dislike:
0
0 (0)
If these polls are correct, (and with the Hearld and Fairfax out in the next few days to confirm it), Dunne will be absolutely kicking himself for that stupid love in he had with Brash. Hes really going to be first on Clarks phone list now. I hope the polls are confirmed and she rebuffs a chance to have a love in with him. What a dork.
Also I think Labours poll lead will continue to harden as the full extent of the Brethern scandal comes out.
Like or Dislike:
0
0 (0)
bloody hell frog, how high is the resolution on your monitor set? all the graphs are huge and go way way off the screen (apart from Left vs Right – party vote % – can’t you make them all like that?) When it comes to Excel graphs you truely are the antithesis of Edward R. Tufte!!!
Also, you could edit the label on the tabs so that instead of sheet 2, sheet 3, etc it actually says what the graph is that that sheet contains.
Like or Dislike:
0
0 (0)
your mrs sure got experience fwwog we credit that, just she overexperienced so we lets her go stockholm soon,
now yous blithely exaggerate again, we confident Epsom plus 3%, put that in yous graph fwwog, because even you got to admit you cant read these reports, theres plus 4 seat ACT,
Like or Dislike:
0
0 (0)
I think its unlikely that Act will get in this time around PQ. Hide trails the Nat guy in Epsom apparently by quite a margin, along with sub-5% poll rating.
Like or Dislike:
0
0 (0)
One positive thing about the ‘volatile’ poll numbers, is you’re seeing a lot more transparency and self-critical analysis from media organisations of their own polling. Not quite up to the standard of the Washington Post and NY Times – who put on line their methodology, questions and detailed data – but things are getting better.
Like or Dislike:
0
0 (0)
And who knows… if the pollsters get it terribly wrong, we might see some media outlets might start redirect their polling budgets (which I’ve heard run to millions of dollars a year) on recruting, resourcing and retaining experienced journalists. I don’t believe the revolving door between newsrooms, lobbyists and corporate/government spin machines is healthy for the media or democracy. As I’ve said elsewhere, what does it say when two of the most experienced journalists in in Parliament aren’t in the Press Gallery because they’re chief spin doctors for Helen Clark and Don Brash?
Like or Dislike:
0
0 (0)
Heh, yeah i think all this does is highlight the mound of salt these polls need to be taken with.
You know what I reeeeally hate – when Campbell Live and the Herald use dodgy Stuff internet polls to run stories and give statistics without giving it any kind of intelligent background.
Like or Dislike:
0
0 (0)
My reply at
http://gonzofreakpower.blogspot.com/2005/09/fuck-polls.html
Like or Dislike:
0
0 (0)
I liked this remark, zippy:
The electorate may be volatile but it is not bipolar.
But I think you’re wrong… I think “bipolar” might be quite a good description for the wild mood-swing voters. Who say “Yeah! Bloody right!” when Brash bashes the bros (not the brethren!) but feel slightly ashamed the next day. Similarly on tax. A surge of self-interest followed by an egalitarian reflux.
Looks like the excitement is with National in this campaign, but they can’t keep it up long. I predict they will turn out relatively limp and shrivelled on election day.
Like or Dislike:
0
0 (0)
Stymied:
Indeed. I find ‘em interesting, but you do have to be very careful about the inherit bias of self-selecting samples. I’m not saying they’re complete worthless (it’s interesting to see what kind of people feel strongly connitted on an issue to participate), but you must have the qualification in there.
Like or Dislike:
0
0 (0)
BTW, Frog, good to see the Greens managed to inspire National Radio to censor its own presenter – see
http://www.nzpundit.com/archives/2005/09/08/05.46.55
Look forward to the post condemning this outright censorship from our state-owned broadcaster.
Like or Dislike:
0
0 (0)
peterquixote:
Rodney will not get Epsom. The poll that showed Rodney leading the Nats was done by Act party members and apparently overseen by an Aussies polling company that is no longer operating in NZ. In other words it is a load of bollocks. A last ditch effort to prevent oblivion.
Alan
Like or Dislike:
0
0 (0)
ah now if you’re looking for a interesting self-selecting sample internet poll, may I recommend:
http://www.politics.org.nz/
they need some more people to take the various surveys – seems to be only 400 so far – you need to register and vote to be able to see the results.
Craig, well you got the post you were looking forward to:
http://blog.greens.org.nz/index.php/2005/09/08/an-odd-overreaction/
Lik e or Dislike:
0
0 (0)