Two polls tonight –

YouGov’s daily poll for the Sun continues to show the two main parties neck and neck, with CON 40%, LAB 40%, LDEM 10%.

Meanwhile a ComRes telephone poll for the Independent has topline figures of CON 36%(+1), LAB 40%(+3), LDEM 12%(-4). Changes are from ComRes’s previous telephone poll a month ago, rather than their online poll a week or so ago (these figures suggest there isn’t a vast difference between the two – last week their online poll was 37/38/13 – but I’ll wait for 6 months parallel data at the very least before treating them as a single series). I believe this is the lowest Lib Dem score since the election from a telephone survey.

155 Responses to “New YouGov and ComRes polls”

1 2 3 4
  1. Ian from L:

    There are indeed very strong correlations between saying you’d vote Conservative and approving of the government, and saying you’d vote Labour and disapproving (and a weaker correlation between saying you’d vote Lib Dem and approving). There are not vast differences in any of these cross breaks in those two polls, so I’d expect there was a significant difference in the answers others/don’t knows or won’t votes gave.

  2. @ Roland,

    Thanks :-)

    I missed you, when Anthony sent you to the naughty step. 8-)

  3. amber

    apart from howards passing reference, no one has mentioned the easing of credit controls or the easing of reserve requirements as contributery causes in the whole “lending” bubble thingymejig

  4. Garry K –

    Buggered if I know, I’ve got statisticians for that sort of stuff. I use the calculator.

  5. @ Neil A &Howard

    in deference to Anthony won’t continue but I will point out what you are saying is TINA

  6. @ Roland

    “This seems like an Anglicised version of an old Magyar proverb about love and marriage!”

    Indeed :/(. I tried to do it, and failed as the different parts of the sentence did not come together. But, I suppose it comes through in spite of the esotheric metaphore and the positioning of the parts of the subjunctive sentence.

  7. @ Anthony

    Previous posts discussed the different approaches different polling organisations adopted to the re-allocation of ‘Don’t Knows’. How they voted last time seems to be the main criteria for re-allocating a ‘Don’t Know’ to a given party. Would the response to the goverment approval question also be considered in deciding the party of a ‘Don’t Know’?

  8. Anthony


    That’s why I love this site!

  9. Ian from L:

    ICM and Populus use past vote for it. I think ComRes use party ID. In their eve-of-election polls MORI use newspaper readership (i.e. if they are finding 70% of Mirror readers are voting Labour, 10% Lib Dem and 10% Conservatives, then they’ll allocate any Mirror reading don’t knows in the same proportions).

    I think in the distant past Gallup did use to reallocate them looking at other questions like best PM, but government approval probably wouldn’t be a particularly good choice as it’s a binary question when we’ve got at least three parties they might be voting for.

  10. Anthony

    Thanks for your reply.

    Presumably another binary question is “Who would you next like to see in 1010 Downing Street?

    While I am on a roll, one last question. Like most of the posters here I think I know exactly why the Lib Dem vote has dropped so sharply – it is because they disagree with what I think. The problem is that we all seem to think differently and try and make the polls fit our own partisan opinions.

    Has there been any poll that has investigated the reasons for the drop in Lib Dem support rather than just quantifying it. Has anyone asked ex-lib dem voters what issues changed their minds, and who they now intend to vote for?

  11. @Amber,

    People choose a house based on what they can afford to pay each month, more than the total they are allowed to borrow. A 2% difference in interest rates makes a huge difference to what a set house will cost per month. If we’d been in the Euro, and our interest rates were 2% lower, the women you speak of would have calculated that they could afford an even bigger mortgage and still afford new shoes. Hence the interest rate directly affects house price growth. As people’s houses grew in value, they felt more confident about borrowing for even more shoes, knowing that if their income wasn’t sufficient to cover it then they could always remortage and dip into all that magical new equity that was accumulating faster than their actual salary.

    Shoeconomics in a nutshell (IMO).

  12. neil

    the banks would have been unable to lend so much money if the reserve requirement had not been relaxed

    regulation of the banking industry has been aweful since the early 80s. all the 1930s laws which were designed to prevent a repeat of that crash have either been removed or watered down or badly/rarely enforced

  13. @ Neil A

    People choose a house based on what they can afford to pay each month, more than the total they are allowed to borrow.
    Let me stop you right there; I don’t believe that is a fact or even a reliable assumption.

  14. @RiN,

    I agree with that. There’s no doubt that the problem originated under Tory rule, and continued to get worse and worse throughout the 1990s into the 2000s.

    I’m not making a party political point. I am absolutely in favour of lending being regulated by sensible rules.

    The origins of this debate are about why I am glad the UK didn’t join the Euro in the 2000s and my belief that our boom and bust would have been worse if we had done so.

  15. @Neil A

    … and if in their old age they can no longer afford the mortgage and the house is repossessed, then all is not lost. I sure I read somewhere that “there once was and old lady who lived in ……..

  16. @AMBER
    On interest rates.
    I actually agree with much of what you have followed up with. However that isn’t what you said in your original post, or at least it wasn’t clear that’s what you meant.
    Of course easy credit led to a housing boom. However, it’s easier to cope with a 4% mortgage rate than one at 18%. That’s one reason why re-possessions are low in this recession compared with the last.

  17. @Amber,

    Go to any mortgage lender’s website and you will see a repayment calculator widget so that a potential borrower can work out what a given advance would cost them per month at varying levels of interest.

    If your income was £2000 pcm and the widget said your repayments would be £1900 pcm, I submit that you’d have to rethink the size of loan you wanted (whether it was technically available to you or not).

    If your income was £2000 pcm and the widget said your repayments would be £200 pcm, I submit that you’d rub your hands together and start looking at houses twice the size or in twice as good an area.

  18. @Amber

    I was going to launch into a lengthy rebuttal, (it had witty word-play, metaphors and everything) but thankfully sanity (and Anthony’s comments policy) prevailed. However, I do have to point out that unless and until my mortgage is paid by the Great Pooh-Bah of Zarg and his many-tentacled minions (henchbeings?) and my boiler repaired by Aslan, I *do* inhabit the real world… :-)

    (Incidentally, were you making a somewhat oblique “Shoe Event Horizon” reference?)

    @Ian from Lichfield

    You said “…Has there been any poll that has investigated the reasons for the drop in Lib Dem support rather than just quantifying it. Has anyone asked ex-lib dem voters what issues changed their minds, and who they now intend to vote for?…”

    I haven’t done a poll but I did do pretty graphs, so I’ll take a punt on this. During the Coalition negotiations, Yellow peaked. Following its formation and the emergency budget Yellow dropped dramatically but then stabilized so that by about week 24, Yellow had a solid minimum of about 11-12%. This stabilization period lasted for about eight weeks and included the party conferences and the Comprehensive Spending Review which, counterintuitively, did not disturb the polls overmuch. This led to much confusion amongst Reds, who couldn’t understand why people might *like* Coalition economic policy. But then came the tuition fees kerfuffle and Yellow started heading down again, currently minimuming at 9%. So my guess is:

    * Phase 1 of Yellow slump: (weeks 4-16?): ABT voters peeling off from Yellow
    * Phase 2 of Yellow slump: (weeks 24-present?): Student voters peeling off from Yellow

    Regards, Martyn

  19. @Ian from Lichfield,

    Tell that to the Old Ladies of the Irish Republic who are about to get repossessed and be personally bankrupt all in one go, as their homes are worth 40% of what they paid for them and they are “underwater” (to borrow the US term) by 100,000 euro.

  20. Garry K (and Anthony)

    The usual Margin of Error used in opinion polls is one to give a 95% confidence estimate. This is equivalent to 1.96 standard deviations on the normal distribution, though pollsters seem to prefer the term “standard error” rather than “standard deviation” (the people they work with think them odd enough already).

    In practice if you are sampling from a large enough population (and both YouGov’s 300,000 panel and the UK’s 40 million electorate are big enough) a sample of 2,000 will give you a margin of error of + or – 2.18. This means if you estimate a Party’s vote to be 50%, based on this sample, you can be sure that, 19 times out of 20, it will be between 47.82% and 52.18%.

    This is based on something called “simple random sampling”, where the only thing interfering with your sample being a perfect representation of the population, is the random chance of which ones you have picked out from the population. Unfortunately life isn’t the National Lottery and your sample will have all sorts of biases in it which you will do your best to correct. But you will never do as well as you do with simple random samples, so your real margin of error will always be greater than say the 2.18 above for a sample size of 2,000.

    I did a longer comment on margin of error back here:
    ht tp://

    There’s one extra complication (actually there’s loads but only one I’m mentioning now). The margin of error also depends on what you are estimating* and the 2.18 above is based on the “true value” being 50%. This makes no real difference if you’re looking for a margin of error on 40%, but if you’re going as low as 10% (where have I seen these figures before), the margin of error is theoretically only 60% of the 2.18 ie +/- 1.31. (It varies as sq rt{p(1-p)/n})

    *Yes I know that if you know the true value of what you are estimating, you don’t need to estimate it. Just bear with me.

  21. @Neil A

    Sorry – I thought it was a gentle attempt to recapture some of this site’s usual wit and charm, but it turns out that I am obviously the heartless sort of person who likes nothing better than to mock the homeless, bankrupt and elderly of Ireland.

  22. @Ian,

    It’s me that should apologise. You caught me in full steam polemic mode and my comment wasn’t well considered..

  23. Martyn

    I don’t know if you spotted it, but there’s a YouGov poll of students gone up on the website today.

    ht tp://

    It’s a little old (fieldwork 16-19 November), but shows the Lib Dem vote among students dropping from 45% to 15% since the election. If anything I would imagine that if anything the percentage has dropped further since (the Guardian revelation about Clegg not seeing the student fee pledge as important only came out on 19th).

  24. C40 L40 LD40 Approval -8%

    More no change, no change.

  25. Latest YouGov/Sun voting intention – CON 40% LAB 40% LDEM 10%
    Latest government approval: minus 8 (Approve 38%, Disapprove 46%)

    Regards, Martyn

  26. Ian from L

    You’re OK flla, anmd don’t let these buggers intimidate you (including me).

  27. Ahem that should be LD 10 of course…

  28. Ian from Litchfield
    I think, as I have posted before that the engine driving down Lib Dem support is a belief amongst a large proportion of their voters thathey are being held up as fools.
    this is much worse than, for example, finding the party isn’t what you thought.
    Early advertising executives in the US thought it would be impossible to sell the VW Beetle in the US. Their research said Mr Average thhought it too ugly but their was a small following amongst academics. The slogan then became Your VW stays ugly longer.
    Huge sales folllowed because buyers felt they were being told they were clever. This time the Lib Dems are doing the opposite.

  29. Doug Stanhope
    Good man ..again!

  30. Ian from L
    Especially if they can’t type..

    A mini revival clearly on the way, I see a golden sunrise.

    It’s all down to VC between 1300 and 1600

  31. @ Martyn

    Thank you for taking time to reply.

    I share your belief that the tuition fees have much to do with the decline in Lib Dem poll ratings. With a 15 year old son and 13 year old daughter I am bound to feel that. I suppose I was trying to find polling evidence to back up what I feel are my entirely reasonable prejudices.

  32. how is the welsh asembly going to fund T fees for all welsh students whereever they study, will it be legal for them to fund welsh students in wales but not english ones

  33. @Neil A

    No worries – you wouldn’t be in “full steam polemic mode” if you didn’t care passionately, and there is nothing wrong with that.

  34. @ RIN

    Welsh should be okay, we do no fees in Scotland but charge English students.

  35. Thanks for that Roger

  36. Amber
    I think you know that my main reason for doubting a Labour SNP alliance is based on the independence issue. For as long as the snp mine every topic for a possible dispute with the UK government. In addition, I have to say that I would rather eat my own flesh than serve under A Salmond who I honestly believe to be a terrible man. I don’t have that aversion to SNP politicians in general. I think you will understand this a littl if you look at what, not just Labour politicians but folk like Tavish Scott think of him. However, I don’t think he will have much interest in Holyrood if he is not a SNP only First Minister.

  37. @RiN,

    Perhaps you should phrase it differently. What wonderful things will England be able to afford to do with the money we will have that Wales and Scotland won’t have?

    More police? More school sports funding?

  38. Martyn,

    Thanks for your explanation. I’d much rather have half the number of polls with nearly a whole point better accuracy but I guess I don’t pay the bills.

  39. Barney

    Sometimes you say sensible things. At other times you are simply silly.

    “For as long as the snp mine every topic for a possible dispute with the UK government.”

    That statement is transparently foolish. The demand from the UK Government for a $7m charge to maintain the SVR was an obvious case for the SNP to make political capital. In my view they should have done so, instead of respecting the SoS request for confidentiality. If you are going to make accusations, you should at least provide a semblance of accuracy.

    “A Salmond who I honestly believe to be a terrible man.”

    That you disagree with independence for Scotland, the SNP, fiscal autonomy and anything else unconnected with a Greater UK we understand. That you are so obsessed that you resort to personal vilification of an individual says a lot about you – and not to your credit.

  40. Ian from Lichfield

    are you going to be moving to wales

    i know it might be a bit of a comute but it might be worth it, you could save £36,000

  41. RiN

    I presume that the Welsh have the same residency conditions in place as we have, to stop English carpetbaggers from doing exactly what you suggest.

  42. Richard in Norway

    I might consider Wales – but I think moving to Norway is taking a bit too far!

  43. To newcomers, may I point out that that YouGov’s panel that give us these figures every evening have not responded to the survey for probably at least 12 months, because they will not have been invited.

    If this is a misunderstanding on my part AW, i should be very pleased for a correction.

    The point I am hoping to make is that any thought that repeated input by fanatics is unlikely.

    This of course goes also for the telephonist as they do so at random (across geographical areas – this is proved by their similar results for ‘nationalists’ as the panellists method returns.

  44. Re the debate about lending.

    Neil A said this – “I’m not making a party political point. I am absolutely in favour of lending being regulated by sensible rules.”

    Apart from general thoughts about capital ratios being raised and BoE interest rate setting looking at asset inflation as well as RPI/CPI, I would also look to institute a simple law of responsibility on lenders.

    Lending money carries responsibilities, as well as conferring legal entitlements to the lender. The primary responsibility is to the borrower, but there is also a collective responsibility to the wider society, as we have seen clearly what happens on a global scale if lending is uncontrolled.

    Put simply, I would enact a law that provides a legal route to complete cancellation of all liabilites if the lender was deemed to have failed to take reasonable meaures to ensure the debt could be serviced and repaid.

    It would ensure we avoid the moral hazard we now have of taxpayers bailing out irresponsible banks, and woud put a stop to destabilising lending practices. Mortgages are a particular problem as the lenders assume there is always a capital asset to claim if the contract fails. They need to be told that this asset might be removed from them if their behaviour is unacceptable and the poor practices will stop overnight.

  45. @Roger Mexico

    Good catch, thank you.

    @Ian from Lichfield, @Colin Green

    You’re welcome.

    Regards, Martyn

  46. alec

    we don’t need a new law, we just let them go bust

  47. Howard

    YouGov must be sampling people more often than that. They’re getting responses from approx 2,000 different people 5 times a week – 10,000 from a panel of 280,000. So allowing for even a small no response rate, it’s probably every six months on average. If you happen to be in a demographic group with low representation on the panel- Northern, male, UKIP-voting, Star-readers in their early 20s say – you probably get asked all the time (and you probably don’t reply as you’re down the pub).

  48. Howard – the repeat rate is certainly more than once every 12 months, especially if you are in a demographic group that is relatively underepresented on the panel.

    There is an automatic lock out preventing people doing the survey more than once a fortnight or so, but in practice most people will do it far, far less frequently than that (you can do the maths of 2000 people into 300,000 – even people who respond every chance they get and are in rare groups are not going to get picked that often)

    Nevertheless, random chance will mean some people do end up in it quite regularly. Sometimes we have cause to merge together the data files for a week or a months worth of daily polls, and we’ll normally find a few hundred people who’ve been polled more than once. Equally, some unlucky (or lucky, depending on one’s point of view) people can be on the panel for years and never get invited to a political survey.

  49. @Alec,

    You’ve articulated something I’ve actually toyed with a lot in the past couple of years. A sort of “Lender Beware” principle for the banks. To be enforceable they’d have to show “due diligence” in the processing of applications.

    It would probably narrow the amount of credit available, but decent prospects should have little to fear from it. And it would remove the justification for punitive interest rates on credit cards (the current rates are said to be caused by 10% default levels on outstanding amounts).

    In my work I’ve come across quite a few “liar loan” cases where organised criminals launder their cash through mortgages. They take out “self-cert” mortgages declaring an imaginary, and unchecked, income in the £40k-£100k range. Their real legitimate income, if they have one at all, is tiny but they know that regular cash inflow from drugs or brothels will more than plug the gap.

    Once the ruse is uncovered (usually because we’ve smashed their illegal operation and they can’t keep up the payments) then it is child’s play to convict them. But we only catch a tiny fraction of the whole.

    This whole process could be made virtually impossible for them by some pretty basic checks by banks.

  50. AW
    I think I was on a bit of a troll, but a genuine one, if you understand me. I too thought it was unlikely that anybody could be asked just once for a twelve month. I think i remember you saying that people had to serve a probation for a similar period and that’s what I probably had in mind.

    The point of that of course was to prevent people joining in just before an election to influence the poll.. I am right about that am I not?

1 2 3 4