Monday’s daily figures from YouGov are out. The topline figures with changes from Friday’s figures are CON 39%(-1), LAB 29%(+5), LDEM 20%(-1). The poll was conducted yesterday and today.

That’s a big jump in Labour support for the first day of the conference. It’s probably a combination of normal sample variance, the fading of the Lib Dem publicity bounce and the beginning of publicity from Labour’s own conference. That said, the figures from last Sunday had a high Labour score too, so there’s a possibility that it could be a timing thing: daily polling is a new development and there’s much to learn.

Looking at the other questions, Sky are obviously going to concentrate on the Leaders’ Debate question, which showed 60% of people would like to see a debate during the next election campaign. 77% of people also said they wanted it between all three main party leaders, as opposed to just Brown v Cameron.

Turning to other questions, in the discussion here we’ve occassionally pondered the difference between Brown’s image in other countries as having made a huge contribution to organising the global response to the credit crunch and his atrocious approval ratings in this country. An interesting question is whether this was because people in the UK didn’t think he had made a positive contribution to the global response, or because it is just outweighted by their opinion on his domestic performance.

Today’s YouGov figures suggest that its at least partially the latter. 49% of people agreed that “Gordon Brown has played a significant role in working with other world leaders to tackle the global recession”, including 30% of Conservative voters. This doesn’t stop 65% of people continuing to think that he’s doing a bad job as Prime Minister.

On other questions, on Brown’s future 50% of people said he should be replaced as Labour leader, 36% wanted him to stay on (including the vast majority, 70%, of remaining Labour voters). If he was replaced, YouGov asked whether there should be an immediate election. 46% said “yes, even if this means holding the election in November or December”, 47% said “No […] we are nearing the end of the Parliament and the election must be held by next spring anyway”. It suggests we are getting close enough to the election that Labour could change their leader without a huge clamouring for an immediate dissolution.

Finally the poll also contained a series of “How would you vote if X were leader” questions. All my caveats I made yesterday about this sort of question apply, but these were at least asked on an even footing when it came to likelihood to vote. The figures were:

Jack Straw – Tory lead of 6
Alan Johnson – Tory lead of 8
David Miliband – Tory lead of 9
Normal unprompted question – Tory lead of 10
Harriet Harman – Tory lead of 13
Ed Miliband – Tory lead of 14

In terms of comparative performance, it’s not too different from ComRes’s findings. I suspect it is not co-incidence though that Jack Straw is probably the best known of these politicians, and Ed Miliband probably the least known (though it’s pretty damning for Harriet Harman, who doesn’t have the excuse of being low profile).

21 Responses to “YouGov Daily Poll – 39/29/20”

  1. Do we think they are weighting the results differently, or not accounting for voter turnout likelihood? It seems odd that they had Labour at 30% last week.

    Having said that, regular poll had them on 26% on 24/09 and this daily poll came up with roughly the same figure.

    I just find it hard to see a real shift in voting inentions by 6 points to and from Labour over the course of one week. People are not that indecissive surely?

  2. Owen. Weights are exactly the same, but YouGov never take any account of likelihood to vote.

    The figures in the YouGov/Telegraph poll were much the same as in the daily ones taken at the same time
    YG/Telegraph was CON 39%, LAB 26%, LDEM 20%
    The average of the two daily polls taken at the same time was CON 38%, LAB 26.5%, LDEM 21%

    We should expect slightly more volatility than normal YouGov polls because the samples are half the size (but that does not mean twice the volatility – the margin of error goes from 2% to 3%). A normal ICM poll has a sample of 1000, and they certainly aren’t volatile.

    My own guess is that public opinion is simply very frothy during conference season, with each party getting a solid week of publicity, one after the other. Had this poll happened in August, I think we’d have probably seen pretty flat lines.

  3. Hard to imagine that Jack Straw as leader would cut the Tory lead to 6. I’ve yet to meet anyone who likes the man.

  4. I’ve been reading this site regularly for several months now, and really enjoy the commentary.

    I have to say I think these daily tracker polls are a bit pointless during party conference season; it would be better to do a full analysis just before the lib dem conference, then repeat it just after the tory conference to see whether the conference season has had any overall net effect on public opinion.

  5. Personally I would like to see a cessation to all polling during the conference season as we are getting such ridiculously fluctuating percentages nothing solid can possibly be learnt by them. To suggest one day that labour is on 23% the next 29% is beyond MOE and belies any true voting intentions. Did you gov by any chance question labour activists in this recent poll? Even amongst them 29% is probably the best they can do :-)

  6. obviously a truimph of Browns liason in the UN kitchens

  7. Labour on 30% do me a favour, did they poll Brighton outside or inside the conference.

  8. Whilst this is a useful exercise , the danger is that changes from one poll to the next may be down to simple M Of E changes or conversely the polls may stay static or move in one direction simply caused by a more Conservative/Labour or Libdem bias in the samples .
    I would have preferred the polling sample to have been exactly the same people in each poll and changes in party support would then be clear .

  9. Anthony,

    What is the point of this daily poll experiment in the conference season between YouGov and Sky news, when everyone who follows these events knows that before Christmas it will all settle down again to a 15- point lead for the Tories?

    David J

  10. David E. Jones makes a valid point, although lots of polls are always good fun for those of us who follow them of course.

  11. Mark – there’s several problems with that.

    Firstly, it isn’t completely possible. You can invite the same people every day, but you can’t force them at gunpoint to actually complete the survey, so in reality the sample still changes each day.

    Secondly, you can only do it for so long – people get bored, stop doing it and the panel dwindles away (and renewing it rather undermines the point).

    Finally there’s panel effect. Asking people a question changes their views (if nothing else, people like to be consistent, so it would be artificially unvolatile!). In some contexts panel effect isn’t necessaily a bad thing, but for something like voting intention where you do want the opinion of people who don’t pay much interest to politics too, it would be a bad thing.

    David – describing it as an experiment as good as answers your question! If conference isn’t a very interesting period to poll I’m sure you can think of more interesting periods. Next year perhaps.

  12. Even though, like most people on here I believe we need to get the conference season (and pbr) out of way before judging if the 14-16pt gap has widened, narrowed or stayed put I like this experiment.
    I don’t share DEJs view that the15pt lead is fixed and what the daily polls may well demonstrate is how much could change if things go well or badly for one party or another.
    So we have a high watermark of 23% for the LDs and so far 30 for Labour (low 23%) the Tories could high score anywhere between 40 and 46. (low 38%)
    This suggest potential votes so we can say it is not fanciful for Labour to score 30 (plus 1-2% others returning). Also potentially the Tories may go as low as 38%. This means a hung parliament is a possibility if the extremes of potential are reached on the upside for Labour and downside for the Tories.
    Of course,should the gap be 10% or lower for a few weeks in early 2010 we may well see a further ant-Tory coalescing around Labour . Conversley should the lead become around 18% we could see a games up mentality amongst anti-tories and a move to minority parties. (possibly the LDs but chances are not).
    I am aware that most posters here can not conceive of Labour improving but this is a polling site and the evidence suggest it is at least possible, if not likley.


    Reports an Ipos Mori Poll showing the Lib Dems ahead of labour for the first time since 1982/.

    Con -36%
    Lib – 25%
    Lab – 24%

  14. Anthony , I accept your points except that doing this for a 2 week period should not lead to a very great boredom effect .
    The problem remains though with a new sample every poll – how do you identify a real movement for M of E changes .

  15. So this poll is saying that the more popular and left-wing the candidate, the more the voters won’t vote for him/her. Doesn’t look good for the future of the Labour Party.

  16. ‘WOLF
    So this poll is saying that the more popular and left-wing the candidate, the more the voters won’t vote for him/her. Doesn’t look good for the future of the Labour Party.’

    Why? Labour is not popular and it certainly isn’t leftwing…

    Hard to imagine that Jack Straw as leader would cut the Tory lead to 6. I’ve yet to meet anyone who likes the man.’

    Conversely they dont dislike him either; a bonus I think for all politicians…

  18. Jack,

    “.. Labour ……certainly isn’t leftwing…”

    Perhaps Blairite NuLab was not, but Brown is trying hard to give that impression at present. Or is there another reason for the sickeningly bright scarlet platform set at Brighton ?

  19. Pls, how do i participate in the opinion polls? I mean, how can i vote on issues raised by the YouGov Daily Poll?

    You guys are doing a fantastic service though!!!

  20. Tee Pee –

    You can join the YouGov panel at then wait. People are invited to surveys randomly, and only 1000 people a day, so in theory you might be waiting for 250 days…

  21. Anthony

    Please – not a 250 day conference season !

    On teh other hand, maybe You-Gov will keep up the dail polls right up to teh GE. After all, it is now fewer than 250 dyas to teh last possible date in June 2010, still less to the most likely date of 6th May