ICM’s monthly poll for the Guardian is out. The topline figures, with changes from ICM’s previous poll at the beginning of the month, are CON 40%(nc), LAB 31%(+2), LDEM 18%(-1). Others are on 11%.

After conflicting polls from YouGov and ComRes over the weekend this one is very much in line with YouGov. Like their Sunday Times poll it shows the Conservative vote steady, but Labour gaining support at the expense of the Liberal Democrats and others. This is the lowest lead in an ICM poll since December last year, and on a uniform swing would leave the Conservatives narrowly short of an overall majority.

As with other polling, ICM didn’t find any particular enthusiasm for the PBR. While once again the tax on bankers budgets was found to very popular, overall the PBR was viewed less positively. Only 12% of people thought it would make things better, compared to 19% who thought it would make things worse. Cameron & Osborne retained a lead on who would do a better job managing the economy, 38% to 31%, but this down from 18 points two months ago.

109 Responses to “ICM too show the lead narrowing”

1 2 3
  1. ““As someone running a business in the private sector, I know that almost all of the people I encounter loathe Labour with a passion. However, regional variations will apply and I accept that there are areas dependent on welfare handouts and public sector jobs where Labour support will be high.””

    @ Hardpressedtqy
    I know people in the private sector who will vote Labour and people in the public sector who will vote Tory. Surely the point of having polling is that we don’t just rely on our small circle of friends/colleagues to guess the result of the next election?

    Also on the topic of public sector jobs: Maybe I am being overly idealistic here, but surly we cannot assign all the public sector support for Labour to the fact that they might have a better chance of keeping their jobs? Maybe some of this support is down to the fact that these people see the good that the public sector does on a day to day basis in hospitals, schools etc?

    ““I have read over the past few days comments on the recent polls. A survey of 1000 people is not representative of the nations views.””

    @ J A Brown

    Why are you on this site then? If you don’t think polls are valid and you are going to press on with your own opinions regardless?

  2. So in a nutshell, Lawrence, for you the impending election is focussing minds.

    I’ve always found the question “Who would you vote for if there were an election to-morrow” can’t really reflect party support, because we can all say we’d vote however we feel to-morrow, and then change our minds when an election comes.

    If the question were “how do you intend to vote at the next election?”, I suspect the “don’t knows” would go through the roof to start with, and slowly diminish until the actual date.

  3. @NBEALE

    My intuition says we are already at 40/30/18.

    This is based on excluding the Angus Reid poll and any other poll that appears way out of line with the others (e.g. Comres’s last poll, Mori’s 6% lead poll) .

    I think excluding AR from any averaging is reasonable, since its systematic error (of low Labour, high others) appears to be greater than the sampling error, and it is also consistently in the same direction. So whatever it is measuring, it isn’t the same thing as the other polls, hence an average seems inappropriate.

    Note, I don’t say I ignore AR, just that I don’t use it in any average. Trends in its movement will be interesting. Whatever its telling us, it doesn’t seem to be measuring the same thing the other polls are.

    I would be interested in your own (and other) views on this.

    In the grand scheme it makes only a little difference, a single % here or there on Lab score.

  4. The slight rise in September unemployment figures were due to the governement putting people on schemes that mean even though there unemployed they don’t appear on the figures.

  5. If you take the last 6 poll results and average them it comes out at just below 12% lead for the Tories. Surely this suggest an 11-13% lead for the Tories.

  6. @Paul B

    Yes, it does if you take an average of all these polls. If you exclude Angus Reid for the reasons I give above (essentially that it is not measuring the same thing as the others) , then you have Comres sitting outside the margin of error of the other 4 polls – highly suspicious, probably sampling error, so balance of probability suggests Lab around 30. Hence 10% lead.

    Mori will add a further sample. If it scores Lab low on 26, then I’ll accept the lead is probably higher.

  7. JC – MORI and Populus both include a break by public and private sector employment in their tables, so can be confident in saying that public sector workers are more likely to vote Labour. However, the difference is not that vast and neither group are monolithic.

    (Incidentally, my guess is that there is a tendency for people in the public sector to support Labour because they see it as more supportive of the public sector… but also a tendency for people who are more left wing or Labour supporting in their views to go to work for the public sector in the first place, so the correlation probably works both ways)

    John TT – nope, it doesn’t massively increase the don’t knows, but it does tend to make a difference to the overall figures (though I can’t say I’ve seen enough examples to draw any patterns)

    Paul B – I’m spending too much time moderating comments from you that are just why you don’t like the current Labour party. It’s not the place for it! :)

  8. Brown and the fact that Labour have been in office a long time are both, still, vote losers.

    But the economy – and the fact that Labour appear to have quickly dug the UK out of recession – are, potentially, vote winners.

    If the economic news continues to improve, Labour can close the gap to about 6/7/8 points and keep it there. In which case expect a hung parliament.

    If, on the other hand, the economic news worsens or stagnates, the Conservatives will still get that comfortable majority.

  9. With the regional breakdown in mind, it does look like this poll confirms what many have said in that the closing of the gap is largely due to Labour voters returning, mainly in the core areas. While this would help Labour survive a total meltdown and give them a base to move forward from post 2010, the huge Tory lead in the South and crucially the Midlands still shows them in a comfortable position. I would say however that while labour has stabilised Brown still faces key risks, primarily on the economy. The Tories are also creating problems for themselves, and the one area where Cameron could start losing support in more critical regions is on the trust issue. While ‘class’ is not going to be a big issue he’s made a big mess of the non-doms factor and the continuing Goldsmith storey is not good for them. I suspect they wish they could drop the IHT pledge but post Lisbon trust and judgement are significant issues and they dare not be seen to go back on another cast iron pledge.

    My long term view is unchanged. Labour need some economic good news, and Cameron’s leadership has not been half as good as some people have thought and the Tories have made long term strategic errors that will continue to hurt them. How significant this will be will largely depend on whether Brown can neutralise his negatives on the economy. Only then will the polls narrow where it matters.

    I agree in principal with your logic. However, I am not at all sure the economy has given much cheer. If as you say things go backwards, Labour will also go into reverse.
    @ ALEC
    Yes, once again I agree, it takes a better man than me to seperate Scotland and calculate the fact that Labours new hardening will probably make a big majority in South Yorks ect even bigger, whilst the party dies in the south, west, midlands and Wales. However, the economy will be the all important factor.

  11. David in france
    “fact that Labour appear to have quickly dug the UK out of recession”

    as far as I’m aware we are one of the only developed countries not yet out of the recession.

  12. ICM overestimated the Labour lead by 3% just before the last general election.
    This happened last Christmas – Tory voters from Surrey, Buckinghamshire etc have gone away for the week. I think all this talk of a hung Parliament is far too hasty. 9% is still an election winning lead and all the indications are that the swings will be bigger in the marginals.

  13. @Andrew

    You forget that all the poll takers weight their sample in some way to account for economic status. And not only that, but economic status, ie ability to get away on holiday in December, is not the vote indicator it used to be.

  14. @Roland Haines – you’re probably right about Labour disappearing in large areas of the south, but two issues could cause the Tories problems. Firstly in London, where I think I’m right in saying (please everyone correct me if my memory is incorrect) that the Tory performance is not as good as in other marginal areas. This could be Labour leaning voters unimpressed with Boris, but if Labour can squeeze a few more seats than expected in London that would help them.
    Secondly, the Lib Dems will play a big role in how seats break. People have observed they are better at defending existing seats than the polls often suggest, and I would tend to think this habit will be reinforced if there continue to be wobbles over the Tories suitability for office. So even if they stay on 40% nationally, they might find they are behind the game in key Tory/LD marginals. Anecdotally a number of Tories already appear to saying this is happening in the south west.
    It is very tempting to dismiss the lead reduction as purely a Labour core vote thing and carry on assuming the marginals will ensure a good Tory majority, but the marginals contain varying sets of seats in different regions and they might not all behave in the same way. Allocating an assumed trend to all marginals is as false as assuming a single national swing.

  15. The Guardian poll is rubbish, take a look at the ICM website and the poll results. The alledged responses were heavily weighted towards Labour VI’s and those who previously voted Labour. I canvass for my PPC in a current Labour seat and am finding a very different picture, more in line with the Angus Reid and Comres results. This is a marginal just within the top 100 required to produce a tory government. As the say in the trade LET IT BE SO.

  16. Well I tried eliminating all the AR polls and the WMA is still 40:29:18 – CLead 11. it only changes the last decimal place and that’s irrelevant anyway. The WMA is pretty robust. Of course on the principle you suggest we should also eliminate the Ipsos/Mori poll on 15-Nov which was 6 points below the WMA.

    On the whole I don’t like throwing away information if I can avoid it, and the more complicated one makes statistical filtering the more possibilities there are for glitches. It’s also pretty easy to analyse the distribution for a WMA but if you censor extreme values it becomes impossible to address analytically.

    Alothough I find it quite interesting, all of this is a bit like counting ships before Trafalgar in my view. The Combined Fleet had more ships and guns than Nelson, but everyone knew in their bones that Nelson would win. No-one (except perhaps for Gordon Brown himself) thinks he has a chance of winning the next election. And probably almost no-one really wants him to. So he won’t.

  17. @ALEC
    Indeed you are right, not all marginals are the same. Mind you, many should never ever have been lost by the Tories in the first
    instance and shows the dire state they were in.

    London is a different kettle of fish within the south. Labour has imported voters from all over the world in an effort to control the capital, but it did not keep Ken in office.

  18. @ David C

    ‘The cost of living is falling.’

    It is in fact up today from 1.5% to 1.9% according to the BBC.

  19. Growth of 1-1.5% next year means even if they were correctwhich I doubt looking at last years guess. Would only average a 0.25 % growth in the first quarter next year hardly an economic recovery.

  20. More like economic stagnation

  21. @NBEALE
    I read you comment with interest, sea battles are an excellent reference for General Elections. In the light of the fact Cameron
    is no Nelson ( and Villeneuve was to good to be compared with Dr Brown) perhaps Jutland is a good choice.

  22. Andrew – skews in polls are not as straightforward as that. All polls are weighted by social class, so if lots of ABs in Surrey and Buckinghamshire had gone off ski-ing (and I think it’s actually a bit early for the Xmas getaway to start), ICM would just weight up the ABs they did find to get things back into the right balance. There is no sign of them having had to do this – in fact, they weighted the proportion of ABs downwards.

    AttackDog – plain wrong. ICM weighted Labour voters DOWN and Conservative voters UP. It is standard practice to weight past vote to marginally more than Labour actually got at the last election to account for false recall. It is impossible for the difference between this poll and ComRes’s to be down to political weighting, since ICM weighted the poll LESS favourably to Labour than ComRes did.

  23. Anthony Just wandering if polls are weighted at all to level of education ? Also what is the main reason for Labour needing less votes to win the election than the Tories.

  24. @Paul B

    Traditional weighting is by ‘social class’. Which has become rather less useful these days.

    And the election is won on the number of seats that a party gets in the House of commons. This is complicated by having a first-past the post system. The current electoral landscape means that the Conservatives are at a slight seat disadvantage, due to their support being more concentrated, and having slightly more significant opposition from the Liberal Democrats.

    The PV+ proposal, preferential constituency voting plus a proportional selection from party lists to give larger regional seats, would go a long way to ‘fixing’ the seat-bias problems that are inflicted on which ever party the winds are against at the time.

    Of course, should they actually introduce PV+, then it’d result in a Lab/Lib coalition at the next election.

  25. Do we have any scottish breakdown for this poll?

  26. Paul – no one does at the moment. It’s the sort of demographic information that is regularly asked though, and it would have been noticed if it significantly diverged from reality.

    The normal weighting variables used are age, gender, region, tenure (i.e. owner occupier, renter, etc) and social class. You also get car ownership and foreign holidays sometimes (that’s a hangover from the early days of phone polling, to get over the bias against very poor households not having a phone. Uprating DE wasn’t enough, since you got richer DEs – foreign holidays and car ownership solved it). There are also public or private sector (used by MORI), newspaper readership and party ID (used by YouGov), and recalled 2005 vote (used by ICM, Populus, ComRes and Angus Reid)

    There’s a run down for the reasons behind the perceived bias here. Basically it’s a combination of low turnout in Labour heartlands, out of date boundaries, and tactical voting against the Conservatives.

  27. Paul M – Nope, ICM don’t produce one.

  28. I would have thought level of education was unlikely to show much difference.
    To generalise hugely highly educated people are probably more likely to be Labour – that metropolitan liberal elite so disliked by the Tories save that the privately educated are probably more likely to be Tories.

    Poorly educated but reasonably well off people – your classic Essex man or woman are Tories but then those in public housing who may also not be educated to a particularly high level are more likely to be Labour

  29. @Paul

    How many sweeping generalisations can you fit into two paragraphs! :-)

    Where does the cap doffing, deferential chimney sweep fit in? “You’re a scholar and a gent squire”

  30. Barnaby –
    Nope, but the glory that is


    still does. Rejoice! I have resisted redirecting some people there today.

  31. I posted yesterday, sadly Anthony muct have been having his tea because it took a long time to get through moderation.

    My reading of 2009, using Anthony’s chart is:

    Tories – ended here they started, 40 -> 40
    Liberals – picked up during the year, 15 -> 19
    Labour – bit of a nosedive, down about 34 -> 28

    My question to the statisticians who know is, how much time do we allow before we say there is a clear trend. 10 polls? 20 polls? 100 polls? Ad what is regression?

  32. Is anybody else sick of hearing the excuses of partisan Tory supporters on here? Apparently all Labour voters have been bought with benefits or jobs in the public sector, or else they’ve been “imported” to help the Labour party.

    Why is it that these “imported” voters aren’t willing to vote Tory? How is it that the Tories ever won an election if Labour can so easily buy off votes?

    The fact that the Tories are unwilling to admit to the mistakes that made them so unpopular is the reason why they’re faltering in the polls. The public can see that behind the shiny new veneer there is no change and no remorse.

  33. “You also get car ownership and foreign holidays sometimes (that’s a hangover from the early days of phone polling, to get over the bias against very poor households not having a phone. ”

    @ Anthony Wells. I seem to remember that in the recent US elections there was some talk about the polls getting skewed by the fact that some sectors of society are less likely to have a home telephone (the young i would presume). Do you have any thoughts on this?

    Can you difference in polling be attributed to the method of polling? YouGov needs internet users which might skew there polls against home phone polls. Are any polls done using mobiles at the moment? I seem to remember some talk of Yougov introducing text polling a while back but haven’t heard much of it since

  34. In response to an earlier suggestion…..

    An average of the last 12 polls – removing the two from Angus Reid – gives the following.

    Conservative 39.3
    Labour 28.9
    Lib Dem 18.5

    That’s a 10.4% lead for the Conservatives.

    A strong lead.

    But not what it has been.

  35. Oh.

    And for what it’s worth, that is a Conservative majority of 18

    (If you put those figures in the electoral calculus generator).

  36. Regional Breakdown

    Scotland and N England Lab 45% Cons 28%
    Midlands Lab 28% Cons 47%
    South East Lab 21% Cons 53%

  37. It strikes me that poll fluctation may be down to this simple fact: people are changing their minds a lot.

    Look at what’s stable, basically the Conservative figure, which appears to be pretty well stuck at 40. I’d read this as indicating that, barring a high profile scandal in the government between now and the election, they have hit their ceiling. This is supported by Labour’s fall being mostly to the advantage of the others, rather than the opposition.

    Labour’s vote is, I believe, volatile because people will make the call on whether or not to vote with the government based on their own personal experience of how the economy is recovering, which is a very unstable variable.

    Lib Dems are interesting. They usually gain during the campaign, but then last time out they had a very media friendly, charismatic leader. This time they have Nick Clegg, can he really make an impact? Or do they have to wheel out Vince Cable at every opportunity? This could be key in Lib/ Lab marginals, if the Labour vote holds up better they could be looking at small but significant gains here.

  38. Number of polls to make a trend depends on how consistent the trend is. If successive polls showed 42% 41% 40% 39% 38% this would be pretty convincing evidence of a downtrend. But if they showed 42% 39% 40% 41% 38% it would not be very convincing at all. Look up Regression in Wikipedia for more info.

  39. @NBEALE
    I think the difference you demonstrate has been the reason for hard words on the site. My understanding of a trend concurs with yours. Fluctuation is another matter altogether. The Tory lead being a good example.

  40. @JAKOB
    This site is about polling and voting in the UK. As a rule most people agree that public sector workers are more supportive of Labour than Tories. The Tories dont like big government.
    Another general assumption is that long term unemployed are better looked after by Labour. As for immigrants, one does not have to be a political genius to understand that if any main party is likely to restrict immigration, it is the Tories. These are the reasons for the comments you dislike.

    No remorse is shown because speaking for myself I have none.

  41. It is going to come down to the economy. Having set up my own small business under John Major (who’s Government were the cause of my redundancy) – it thrived under Labour’s company tax policies that helps SMEs grow and invest in new products. Labour remains the party of small business and entrepreneurs. and is unlikely to lose that support. Large business may drift back to the Conservatives. As for those who work for their living ( I also have a part-time job) if the economy does rebound then they will drift back to Labour as well. So it is still Cameron on a slow decline, Labour on a slow climb and everything to play for next year still. Watch out for the final quarter GDP figures and the first quarter unemployment figures – they will tell us far more than any opinion poll.

  42. @Roland Haines

    As for immigrants, one does not have to be a political genius to understand that if any main party is likely to restrict immigration, it is the Tories.


    Although off topic, you ought to bear in mind that immigration increased, substantially, over the course of the last Conservative administration.

    Once in government, economic necessity makes parties (of either persuasion) do quite different things from what they promise in opposition.

  43. @ Eric Goodyear:-

    ” Labour remains the party of small business and entrepreneurs. and is unlikely to lose that support.”

    From the website of the Federation for Small Businesses :-

    The Federation of Small Businesses welcomed a calling off of the proposed rise in Corporation tax by the Chancellor in his Pre Budget Report and moves to extend the Enterprise Finance Guarantee for a further year which FSB had urged.
    However, the FSB said that the proposed one per cent rise in National Insurance is a serious blow to small firms wanting to grow and employ and warned that the rise could cost 57,000 jobs.

    As I understand it Osborne has said he will try very hard to avoid the NI increase. That would appear -in light of FSB’s comments-to put him within the ambit of your “the party of small business” status too.

  44. It seems that 40% is at the lower end of the Tory accomplishment spectrum: to get more than that, they have to either be putting their policies out there (eg. the party conference) or on the offensive against Labour.

    What puzzles me is how quiet the Tories are and what they’re saying when they do say something, eg. David Cameron saying that Gordon Brown would benefit from an early election. This is probably over-analysing, but it would be a conceivably strategy on the part of the Tories to stay quiet over the winter, tease an early election, then go for broken when the campaigning starts.

    On the other hand, that would be a very risky and unorthodox strategy, as well as one that would miss out on the potential of Labour embarassment post-budget.

    Most importantly, there’s no reason to suppose that this or any other trend will be permanent. The Labour vote may well go down again, the Tories might decline, the Lib Dems might even become relevant again etc. There are many modalities here and the easy way to smell out political preferences on this site is to look at which trends people expect to continue indefinitely.

    I still hold out hope for a hung parliament and a Liberal/Conservative “coalition of liberty”, but past history suggests that both are unlikely, which is immensely unfortunate.

  45. @ Rod

    I don’t believe that the Tories have hit their ceiling. There are a few factors that could add a point or two come election time:

    1) A natural gain (who’s to say it won’t happen)
    2) A competitive advantage once campaigning starts as their budget is a lot larger
    3) Drift back of UKIP support
    4) Historical under-reporting of Tory support in the polls (although that may not happen this time)

    That lot could see them at 44% or more come the GE.

    Who knows!?

  46. I think one reason for the tories being quiet might be clutter.

    With both the PBR and the Climate summit set to dominate the news any tory attempt to set the agenda might have ben crowded out and so where as they can continue to snipe and criticise trying to launch new policies when there were two big political issues coming up plus Christmas and the [email protected]£$%DY X Factor might have meant they didn’t get the impact they wanted.

    Just a thought ( I do occasionally have them).


  47. I have looked at education figures and Labour heartlands have the least educated people of all parties. Places like Glasgow where they had the by election had 52% of people with no qualifications in 2001 census.

  48. A lack of education normally means your likely to vote either for a working class party or be swayed by the media as you are less likely to question information your given.

  49. Peter – you thought too soon! Cameron’s gone quite loud on the Climate Change front, and slapped down the “greenosceptics” in his ranks, declaring that he drives policy and that’s how it is going to be.

    £20bn Tory scheme for Tesco to lag lofts and take a cut of the savings in return. Dismissed as an “unfunded spending commitment” by Labour.

    I look forward to more examples of Community Action policies from Cameron. it’s a risk, because people don’t necessarily buy the idea that Community Action is the answer. Interesting though.

1 2 3