Mike Smithson has the latest Angus Reid poll for PoliticalBetting, conducted over the weekend just gone. The topline figures are CON 39%(+1) LAB 22%(-2) LDEM 21%(+1). I’m on the way to a meeting, and will post properly on my return – in the meantime feel free to discuss it here.

52 Responses to “And at the other extreme”

1 2
  1. Phuuee What a relief I was worried there for a minute

  2. Is it just me or do these Angus Reid polls produce some weird results compared to other polling companies (Labour 22%??)

  3. Pardon my French. These guys should go back to Canada ! Some garbage.

  4. Now that looks more likely and about right-lol
    Although I must admit these Angus Reid Polls are abit odd compared with most of the others!

  5. This poll MUST be accurate. Just joking!

    It does perhaps confirm what I said about Clegg turning the Mori poll to the Lib Dems advantage.

  6. Don’t agee that it is garbage although the Labour figue is probably understated.The Tory figure is at or about the level at which it has been for some time now. If anything these pollsters have always had lower Tory figures than other polling Orgs. The figure for others is as always with this Org high. Mori is the one that looks rogue

  7. AngusReid ask how you would vote in your constituency. In safe seats that might boost the Others as people may feel freer to vote other than the big 3.

  8. Another poll showing Others way up there? Main parties make up only 82%.

    I personally believe that Labour’s rock-bottom, unassailable core vote is around 22-23% so whilst this poll isn’t inconceivable, it seems very low. How on earth would Labour have fallen over so badly in the current circumstances? Rogue, rogue, rogue… with an overinflated Others score to blame I’d wager..

  9. Undoubtedly it is the minus 2 movement of Labour’s score that is the significant thing.

    Note that when Angus took their last poll on the 5th of November Yougov also took one. Yougov polled for Labour 27 on that day.

    This poll may therefore be taken to mean that Labour are on 25 – a figure I have often suggested reflects Labour’s true position.

  10. Interesting, even with the poll leads going haywire it looks like levels of Tory support are fairly consistant – it’s the Labour/Lib Dem axis that seems to be constantly shifting…

  11. Probably the most accurate poll since the beginning of time.

  12. The 3 Angus Reid polls so far published all seem to be rogues!

  13. I think if you take the average of this poll and the MORI poll you probably will be somewhere near the true picture.

  14. I have noticed that Angus constantly gives Labour a lower rating and the Lib-Dems a higher one than theother pollsters. Something funny going on here!
    Could it be their on-line polling source?

  15. @ GRAHAM

    “The 3 Angus Reid polls so far published all seem to be rogues!”


  16. I was looking forward to the next poll just to ease my mind, but i think this is understating Labour a little.
    Neil A predicts their unshiftable support is 22-23%, i think it is more like 25-26% so im not totally sure about this poll either.

  17. COLIN,
    I believe these polls lack credibility because:
    1.they have consistently pitched Labour lower
    2. they have consistently pitched the LibDems higher
    3. they have consistently pitched ‘Others’ higher
    than the more established polling companies.

  18. The hardcore Labour vote is 28%. Michael Foot got that in 1983. During a real election, it cannot go lower than that.

    I suspect Labour will get 31-32% in the elections since even disgruntled Labour voters will come out in the end.

  19. Yes, as Andy says. The picture over the last few months is fairly stable in terms of the Tory lead if you average it out. We’re most likely looking at a lead of 12-14 points and that doesn’t seem to be changing so far. Maybe it will in future but it doesn’t seem likely unless something huge, freaky and favourable to Labour occurs.

  20. I have recently often urged people to be cautious in crying rogue when a poll shows better news for Labour, and I think we should do likewise here. As AW always says, pollsters all follow some pretty strict guidelines and the data tables are published, so no poll is ‘wrong’ as such. Its also well over a week since the fieldwork for the last two polls that showed an improved Labour score, and as I said previously, we may see the improvement reverse as the afterglow from Labour’s better week wears off.

    Having said all that, while this poll adds to the data suggesting the Tories could be under 40% the extreme low score for Labour and high percentage of others does look odd, especially given that Angus Reid does seem to be developing a habit of underscoring Labour when compared to other pollsters.


    I think it might be worth remembering that when Labour polled 28% in 1983 the ‘Others’ amounted to little more than 2%. Had the ‘Others’ then been polling 12 -18% Labour would have dropped below 25%!

  22. On the assumption that the actual Conservative figure is 39% and the actual Liberal Democrat figure is 19%, then both Mori and Angus Reid can be accommodated within their margins or error for these two parties. The question remains over Labour because there’s now to reconcile the two polls or the other recent polls to each other.

    I think it’s possible that there was an upward blip in Labour support around the 15th, so it’s possible that neither poll is a rogue. However, I feel that at the moment Labour is even less popular than the Conservatives at John Major’s nadir in the polls. If NickR is right, I think that Labour’s support may hold up reasonably well in the marginals, only for them to face meltdown in seats that should be reasonably safe.


    I think Labour may poll 27% at the next election if things don’t go well for them in the next few months and during the election campaign. I agree with you that they won’t got down to 22-23%.

  24. Shades of Brian Blessed as the King in Black Adder: “I like not these polls. Bring me some other polls!”

  25. The hardcore Labour vote is probably somewhere 20-30% depending on how popular they are, and as the election gets nearer i think a lot of the “Other” support will fade away as many will think that a vote for an “Other” party would be a wasted vote (not that i personally think that).

    The polls have usually tightened when the election gets nearer, i think the Tories in 1979 had a 20% lead over Labour a few weeks before the election, but won by only 7%. Labour’s a few weeks before the 1997 Election was between 20-30% in a lot of polls, but they won by only 12%.

    So as unpopular as they may be, i can’t see any reason why this can’t happen again, especially considering that the governments of Callaghan and Major were at least as unpopular as the current government.

    My prediction is that the Tories will win, but narrowly.

  26. Imtiaz Kabir

    “The hardcore Labour vote is 28%. Michael Foot got that in 1983. During a real election, it cannot go lower than that.”

    Records are there to be broken. They got 16% in the recent Euro elections. I know that many people vote differently in Euros, but you’re saying Labour can’t fail to nearly double their support at the last real election?

  27. Bobby:

    Of course a narrow Tory win in seats would almost certainly mean a pretty large lead for them in terms of share of the vote.

    A narrow Tory win in votes would probably mean Brown staying in Downing Street thanks to the unequal numbers of voters in different kinds of seats.

  28. The difference between this poll and those of the 13th Nov discounting Mori is 3 once and 4 twice. Following the debacle and negative publicity regarding the Queen’s speech this is perfectly possible. I don’t know what time Anthony posted this information but i wonder if the media will make as much of this result as they did regarding the Mori poll. I don’t think The Guardian will be saying anything about this one nor the BBC. I do believe however that this result will not get half the coverage Mori got.

  29. Oops I should have said around 13th Nov (first sentence)

  30. It does appear that AR do have some kind of methodological difference that shows Labour support way down when compared to other pollsters. Three in a row with Labour over 3 points under the weighted average of other pollsters is no chance occurrence.

    Can it all be down to the “in your constituency” element to the voting question? Certainly the Politics Home marginals survey suggest many Labour supports are happy to vote Lib Dem in areas where Labour has no chance, but for that to translate to such a large difference nationally seems unlikely.

    Do we know how the panel has been put together? Perhaps AR themselves would like to comment on why their polls show much lower Labour support and higher Lib Dem support than even other panels pollsters?

  31. jack r
    ‘Is it just me or do these Angus Reid polls produce some weird results compared to other polling companies (Labour 22%??)’

    No they are quite accurate and Labour are on 22 – although I accept it could be slightly less.

    Everyone said AR were wrong on ‘others’ in Canada – until the election. Then they stopped.

    Chris Partick Moore.
    The ICM was done on the same day as MORI and now this suggests that Labour got a bit of a bump from the by election and then it went down the tubes. And I expect the BBC would headline it for a week but I think a cat that looks remarkably like Hilter has been born in Rochdale and you just can’t compete with that.

  32. Excuse me if I ask an obvious question. Angus does an on-line poll. Do theothers poll on the phone? If so, could this cause a lower Labour count?

  33. Sorry I didn’t make myself clear. Would an on-line poll cause a lower Labour figure?

  34. YouGov usually do their polls online as far as I know.

  35. Ah, I have just found Anthony’s post from October 22nd, 2009,
    (http://ukpollingreport.co.uk/blog/archives/2321#comments), suggesting why AR polls may under-estimate Lab and overestimate Lib Dem compared to other pollsters. Its all in the weighting.

  36. You are correct Andy. YouGov do their polls on line. I used to participate.

  37. Thanks Andy

  38. Betting men are putting up real money to pay for Angus Reid. Behavoural economics say that people staking money on the result tend to take the question more seriously than the rest of us.

  39. I hope YouGov can clear the fog a bit when it reports at the end of this week. At the moment I have misgivings about three of the pollsters – Angus Reid – Mori and ComRes.

  40. Graham why ComRes?

  41. AL J,
    ComRes is far too hit and miss. Sometimes it comes up with results in line with other pollsters , but far too often its results are obvious garbage.They also – like Mori’s 100% certainty to vote figures – tend to be volatile..
    So the polls I tend to take seriously are YouGov , ICM and Populus – in no particular order!

  42. Graham:-

    “I believe these polls lack credibility because…etc”

    But the all important lead-gap between Cons & Labour- is consistent with other polls ( except MORI ) isn’t it?

  43. Gosh It can get confusing!!

  44. Colin,
    In so far as that is true I suspect it is by accident rather than design – though the lead is very much at the higher end of the recent range of findings. It does not seem too far out in respect of the likely Tory % share at the moment, but it is giving Labour a rating just over 55% ot the Tory share whilst ICM put it at nearly 70%.

  45. When most people (me included) make predictions we often can’t help being influenced by what we’d like to happen. But as Diversity points out when people are betting their hard-earned money on outcomes they tend to be more dispassionate and therefore more accurate. That’s why the betting markets are usually pretty good at predicting results. I know the Iowa Electronic Markets have been very good at predicting American elections.

  46. I don’t wish to show off, but I don’t find these polls confusing.

    After the by-election win by Labour they enjoyed a brief boost. But the ICM poll which showed Labour to be on 29 was much more accurate.

    The ‘in your constituency’ part of the Angus voting intention poll does actually make a difference. It encourages people to think about their tactical options, something that is likely to happen when the day comes. For this reason it seems to me that Labour’s situation is even more serious than the Yougov polls have suggested in recent months.

  47. Realistically, how likely is it that Labour is going to from 29% to 22% in a week?

  48. Imtiaz Kabir,
    Labour’s hardcore vote is 28%. Michael Foot got that in 1983.
    So why automatically assume an equally useless, dithering, unpopular leader is going to fare as much as 4% better?
    As Graham points out the ‘others’ now make up a far bigger proportion of voters than in 1983.
    You also state that even disgruntled Labour voters will turn out in the end.
    If they do not fall into the true ‘core vote’ category, they might not be intending to do any such thing.
    I still believe it is a serious uphill task to poll 31-32% for Labour, they need a few events working in their favour or my money is on 27-28%.

  49. “Realistically, how likely is it that Labour is going to from 29% to 22% in a week?”

    Very unlikely, particularly as there have been no sensational political events. However, Labour could quite easily go from 26% to 25% or from 27% to 24% and this would keep both ICM and Angus Reid within their error margins.

  50. Poll out in the Daily Telegraph Tuesday re Scottish Voting intentions – any update?

1 2