ComRes’s poll was supposed to be embargoed until 10pm, but given both ITV’s Meridian Tonight and Conservativehome have already published it it seems somewhat pointless to wait. Their rolling poll for ITV and the Independent has figures of CON 36%(+3), LAB 29%(nc), LDEM 26%(-3). ComRes’s fieldwork would have been conducted between Monday and Tuesday, so wholly prior to the Mrs Duffy incident (though at the same time as Populus’s poll yesterday, which also showed 36% for the Tories).

365 Responses to “ComRes/ITV/Indy – 36/29/26”

1 2 3 8
  1. I’m not convinced by such a rapid 3-point swing from LD to Con. 1 or 2, maybe, but 3 seems to be too big too fast.

  2. I’m not convinced by any of them such a range, all over th place. Just confused.

  3. Another Comres with a heavily Tory sample on a Tuesday? What is it about Tuesdays, Tories and Comres?

  4. Con and Lab holding up well as LibDem fall back. After all this fuss we will probably get back to Con and Lab being just a couple of points apart. Refreshing to hear members of public on tv saying ‘bigot-gate’ will not make any difference to their vote and it just shows Brown is human, everyone has done it. Media still mistakenly believe they can decide this election. Don’t think this will harm Lab in polls at all

  5. The irony is that this “rolling” approach is supposed to smooth things out, but if anything Comres is more volatile than the other polls!

    I do note that the LD share seems more variable across poll to poll? Possibly this is because it’s not distributed as evenly as we might have guessed, and the pollsters haven’t caught up with their geographic weightings to compensate.

  6. Totally unreliable polling company in my opinion. Have these guys ever been close to what everyone else publishes?

    I’m surprised that The Independent (and ITV but not so much) still use this polling company. It makes them look bad too.

  7. Is this a rolling poll? If it is, the second 500 has produced a very weird breakdown. Tories around 40% & LibDems 23% in the second 500. It doesn’t pass muster, if so.

  8. Anthony – Please tell us, is it true there will be a “gaffe-gate” poll from YouGov at 10?

  9. ComRes all over the place. Going to have to discount their results due to the wild fluctuations they keep having.

  10. With this and the Harris Poll on politicalbetting the WMA is 34:27:29.

    Still no stat significant trends. Though FWIW the WMA LD:L lead, which had been narrowing sharply from 2 to 0, has now started to go back to 2. But far too early to say if it is a trend.

    At the end of May 2009 the L WMA was under 22. I’d guess that post bigot-gate the utterly core Labour vote is about 15. My guess, but it’s only a guess, is that we may see exponential decay towards this figure, but it won’t remotely reach it by the election. Afterwards, however…

  11. Rolling polls make no sense whats so ever to me – can anyone give a positive point for them?

  12. Maybe, just maybe this saga may make us all realise just how bigoted we all really are?!

  13. @Christopher

    If they deliver headline-grabbing fluctuations, then I can see that a positive point for the TV news programme that partly commissioned them.

  14. This poll is bonkers.

    Last night, ComRes had an improbably sharp rise in the LD vote against the trend of every other poll. I didn’t believe it. Today it shows a sharp drop in LD support and a rise in the Con vote, again against the trend of every other poll. I don’t believe that either.

    This kind of volatility is not reflected by any other polling company, so be very suspect of any ComRes poll.

  15. @Christopher

    “Rolling polls make no sense whats so ever to me – can anyone give a positive point for them?”

    You have to do half the work to get your sample size up?

    More seriously, I can’t think of a reason why they should be worse than a single poll done over 2 days.

  16. ComRes is clearly a respectable company conducting respectable surveys. However, their results do seem very wierd compared to the two other reputable polling companies. ComRes seems to show wayward pro-Conservative swings after each weekend, which they then water down marginally for the rest of the week towards the steadier results of the other two! Very, very peculiar???

  17. Intresting stat:

    Book retailer Waterstone’s says sales are up 160% on 2005, and there are still eight days to go before election day.

    The biggest rise compared to five years ago is for the Lib Dems with curious voters driving a 250% boost.

    But figures show the Conservatives are ahead with 38% of the total manifesto sales so far at Waterstone’s, followed by the Liberal Democrats on 32% and Labour on 30%

  18. From a 1 point C – LD gap to a 10 point gap in just 2 days? I think that says a lot more about ComRes polling methodology than it does about voting intentions.

  19. Colin – Events dear boy events – rolling polls can’t pick these impacts on polls. Polls should be over 2 days maximum, these rolling polls are covering 2 days per poll with figures added onto one another.

    Also, rogue polls; these are added onto valid polling figures and mixed together.

  20. @Colin Green

    You’re right, there’s no obvious reason why this sort of rolling poll should be unstable, but nonetheless they do seem to be wobbly (the ones during the party conferences last year were similar IIRC)

    Perhaps it’s nothing to do with the rolling-ness, but some other aspect of Comres’s methodology that is causing the volatility?

  21. If there’s no “Gaffe gate poll” either

    A) Peter Kellner didn’t say it at all
    B) It didn’t show a huge drop for GB (or even showed a rise) so the Sun won’t run it?????

    Please tell us Anthony, I’m on tenterhooks.

  22. @Greengrass

    “That’s 100%+ loan to asset value – not the same thing. Loan-to-income of 250% has always been considered prudent and we’re only talking 75% here.”

    The GDP is not equivalent to income when looking at national debt. The appropriate debt to look at in that instance would be the national debt + the overall debt of people + companies. If you’re only looking at the national debt then total taxation is the appropriate income to compare too, which is about 35% or so of GDP, which would put the loan-to-income close to 250%.

    And while that much may be prudent for a mortgage, I’m not sure the same applies for a country, especially one with (analogy) a few credit cards maxed out they didn’t tell the bank about…

  23. all margin of error stuff in all the polls.I suspect we’ll know the real picture on Monday after the debate and the “Brown disaster” work their way through the polls.After that i’m not expecting much movement.

  24. Taking into consideration Harris’ poll and the probabilty that Yougov tend to overestimate Labour’s position by a few points at the expense of the Lib Dems the situation is likely to before the Brown ‘bigot’ revelation Cons 33 Lab 25 Lib Dems 31.

  25. @William

    As a Waterstones employee I can tell you that 150% rise on not very many is still not very many. We sell very few manifestos so whilst it’s interesting to note I doubt it is much of an indicator.

  26. @Sue Marsh (10.08)


    BBC News at 10 have just said that the Sun interviewed the lady in the gaffe but decided not to continue with the interview. Tends to support your B)

  27. @ Sue Marsh

    The Sun posted this to their election feed on Twitter “About 7 hours ago”

    “YouGov/The Sun is polling on Bigotgate now – how much of a disaster is it for Labour’s campaign? Answers tonight.”

  28. William
    ‘Book retailer Waterstone’s says sales are up 160% on 2005, and there are still eight days to go before election day.’

    What they’ve sold 260 instead of the 100 last time?.

    Who on earth would by one of these things, when they are all on the Internet, other than manifesto collectors?

    I suppose that’s the answer.

  29. It went largely unreported. I am not sure why. But Nick Clgg got a standing ovation at the Royal College of Nursing. I think e spent the rest of the day back down in Southhampton. But it was a good day for him. Also te institute of fiscal studies showed the LD’s to have the most transparent mainfesto pledges on cuts. Perhaps we can say this was enough a) to stabilise the LDs b) to push them up a little. With YG having 4% rise but only 2% drop in big three it is likely there is a bit of rounding going on. Thus, I genuinely hold their poll to be more accurate.

    HAving said that, Populus, Harris, Opinium and now Comres have the Tories on 35/6. I believe them.

    Thus I think it is fair to read something like

    35/ 28. / 29 Into the polls.

    It will be interesting to see what YouGov’s poll has tomorrow night….

    Now please can somebody help me find a TNS-BRMB? :)

  30. Hmmmm. ANTHONY!!!!!!!!!! WHERE ARE YOU!!!!!!!

    Surely you must know, have you been “embargoed”

  31. @JamesB
    “a few credit cards maxed out they didn’t tell the bank about…”

    I think they did tell the bank…in fact it was the bank that helped to max them out with QE

  32. Con 36%
    Lab 29%
    Lib Dem 26%

    shock horrror, dont believe it etc….spare me, this is the exact GE prediction consensus here two nites ago..Peter Kellner et al,,,,,,the shock is it has come to pass so soon…

  33. Things are never what they seem.

    The Brown “disaster” is really no such thing-just a huge media operation to character assassinate Brown which almost certainly will give him a slight boost in the polls short term before the tory surge early next week which would have happened if Mrs Duffy hadnt left her house to buy bread and make an ass of herself today.

  34. putting an average of C34% L27.3% LD29% into the swingometer gives Con and Lab equal on 263 seats. I so want it to be a tie. Can I have one? Pleeeeeeeeeeeeeez?

  35. Interesting that Julian Glover says Oxford East has not been polled because it is already a notional LD seat, because with the Rallings & Thrasher figures that will be used by the mainstream media it is still a Labour seat. It is with the figures used by this website that it is notionally a LD seat. So maybe that means ICM are using the UKPR notional results.

  36. The Sun has refused to publish their interview with Mrs Duffy.. Apparently Duffy’s word ‘flocking’ was mistaken for something else.

  37. @Dave
    I agree, I was being sarcastic about the “Brown disaster”

  38. The bank was an analogy too, not specifically the BOE, just everyone really. QE doesn’t earn interest explicitly though I guess inflation counts as indirect interest.

    Re bigotgate poll, possible that the sun also remembered back to lettergate and had second thoughts.

  39. Andy JS – ICM do use my notional figures for their marginals polling.

  40. My favourite way of estimating the situation is using the ICM polls by subtracting 3 points from Labour’s score, and adding one point to the Cons and Lib Dems score unless other polls suggest a lower score for the Cons or Lib Dems.

    This is based on an analysis of ICM’s last 6 polls before 2005 GE.


    ‘a tie. Can I have one?’

    I want a ‘Hung Party’ rosette – do you think I can locate one – rarer than hens teeth – or should I say than ‘converts to the Labour cause’

  42. I’m therefore looking forward to the next ICM poll very much which I imagine will be tomorrow night.

  43. Journalists do seem to be feverishly excited about bigotgate. It puts me in mind of how bloggers seem to be able to work each other up into a frenzy over nothing based solely on positive feedback loops reading each others posts.

    Hopefully we get a poll or two between now and before tomorrow’s debate, it will be interesting to see if the public agree.

  44. If the LD share is set to increase by say 10% in the all of those seats where Labour are in first place and the LDs are in second place, what would the increase in the LD share be in all other seats?

    Since the latest UKPR shows a 6% increase in the LD share (to 29% from 23% last time), maybe the answer is that the LD share increase in seats that are not Lab/LD constituencies would be about 3-4% if that 10% increase happens in Lab/LD seats.

  45. PeterBell: “BBC News at 10 have just said that the Sun interviewed the lady in the gaffe but decided not to continue with the interview. Tends to support your B)”

    Or she burst into tears? Or she was too decent to answer any of the questions – doubtless about what GB said to her? Or something not obvious?

    The Sun is doubtless self-interested. But when she woke up, she was more likely to be thinking about getting round to posting her vote in favour of Labour in a marginal constituency than working out how to humiliate Gordon Brown.

    Brown’s best play was to continue with his first response, ie: annoyed with what she said about immigration. He could have said that she had phrased badly a perfectly legitimate concern. Then apologised and said that it was a salutary lesson in so many ways.

    But he spun it out of control.

  46. I expect the Labour Spin Machine to launch a full all out smear campaign against Mrs Duffy. If not I can see Labour Polling <25%

  47. Sun paid 50,000 for an interview but decided it was “Too boring” to print!

    Were you joking Eoin, I didn’t get it?

  48. @ Colin Green

    Trying to make sense out of our three very different polls tonight, I tried to average them too.

    I think C34% L27.3% LD29% feels about right.

    A tie, well now that would fun

  49. @Sue,

    Not too boring…. GB mistook the word “flocking” to be something else. I must admit so did my son.

  50. Interestingly on the Jon Sopel show tonight there has been a lot of supportive texts for GB and his gaffe and also some interviews reflecting the same.

    We can all be guilty of hypocrisy and I am sure there are many slating GB tonight who have at some point done so albeit in lesser important situations and without the media frenzy all who have been waiting for this to happen.

1 2 3 8