There is a new ICM poll in the Sunday Telegraph which is – rather inexplicably – already up on their website here. The topline figures with changes from the last ICM poll are CON 40%(-4), LAB 28%(-4), LDEM 22%(+6). The poll was conducted on the 4th and 5th February.

The changes since the last poll are rather odd – Nick Clegg got some good media coverage launching his parties new education policy… but probably not 6 percentage points worth! As ever it’s worth being a bit wary of any poll showing a big shift in any party’s support. My suspicion here is that the previous ICM poll’s Lib Dem figure was an anomoly – it had them three points down at 16%, which was their lowest score under Nick Clegg and meant YouGov and ICM were showing the Lib Dems on the same figure, whereas normally ICM have them a couple of percentage points higher. Moving up to 22 is still very good for the Lib Dems, their highest level of support for a long time when most other polls show them in decline – for that reason, treat it with some caution till it starts showing up in other polls.

It’s the second poll since Labour’s recovery turned sour to show Labour below 30%, though ICM tend to get a lot more attention than ComRes. It’s the first time ICM have put Labour that low since July 2008, though they had been in an even worse position before that – in June 2008 ICM had Labour at 25% and 26%.

Other questions in the poll apparently include a question asking whether Alistair Darling should be replaced – 51% think he should, 43% think he should stay at the Treasury (though the Sunday Telegraph’s write up implies this is a largely a partisan response, with Tory and Lib Dem respondents saying they’d like him gone and Labour voters saying they think he should stay.The poll also included a question on “British jobs for British workers”, 50% said they would like to see a tighter regime for EU workers and 46% think EU citizens should have the freedom to work in the UK.

32 Responses to “Lib Dems up in latest ICM poll”

  1. so it official 2029 before the uk hopes to pay back
    gordon ‘ponzi’ brown and tony ‘i dont do god’ blairs debt.

    thats assuming the existing businesses and entrepreneurs who will have to pay this dont flee.

    the lib dems were in government in scotland during a good part of the big spend years,they cant not get away with that.
    alex’ iv got all the sweets’ salmond is just another spender.

    what a mess

  2. I agree the +6 probably says as much about the last ICM poll as it does about the current situation, but I think the picture from the other polling firms is more positive for the Liberal Democrats than your comment suggests Anthony. For example, the last YouGov poll gave the party the highest rating since September.

  3. Anthony, do you think this weeks snow could boosted the Lib-Dems somewhat, with more public sector workers, particularly teachers, being off work in the week than is normally the case?

  4. I was very impressed by the Lib Dems’ education policy statement by Nick Clegg. I am a committed Conservative but I can easily believe that for floating voters, this could swing them, particularly if education is a major concern. For me, school education policies are not very important because I’m single and not likely to sprout a sprog. But I really liked the idea of abolishing Child Trust Funds and using the money more constructively elsewhere. I think that Child Trust Funds are one of Gordon Brown’s worst ideas.

  5. WMA 42:31:17 CLead 12 (11.5+) pretty much as it says on the can.

    As Martin Wolf in the FT remarks, Gordon Brown’s fiscal reputation has been demolished and Sarkozy’s forthright comments will surely have penetrated all but the thickest skulls. Since “economic competence” is Brown’s sole claim to leadership (his oratorical and leadership skills are amply demonstrated by his calling the recession a “depression” by mistake in PMQs) we can expect further movement over the next few weeks.

  6. Not surprising. If you look at how some local byelection results have been going, the Lib Dems are on the up. On 29 January they won six extra seats.

  7. With the occasional poll like this I think it’s doubtful the LDs vote at the next election will fall to something around the 15% mark (as it did in 1979), which some Conservatives are hoping for. Something around 17-20% seems more likely at present.

  8. Makes a hell of a difference Andy.
    On Anthony’s Prediction thingy:-

    44/32/16 = Cons 359 seats/ 68 Majority

    40/28/22 = Cons 230 seats / 20 Majority

    The Conservatives should be more worried about this Poll than Labour ,who must be losing all hope of another term in power.

  9. The political talk in my club always centres around how good Vince Cable and Cleggie are, not suprised by the poll reults, they will keep most of their present Mps an dgain more from Labour.

  10. Hennessy in the Telegraph blogs about the latest poll but I’m guessing it was covered early as otherwise it would be lost in this teaser for what Darling is to write about in the Telegraph too…

    ‘Darling also has a piece in tomorrow’s paper in which he makes a significant announcement about bank bonuses – and indeed the management of banks as a whole. Full details on this website later’.

  11. It’s difficult to decipher what the correct Lib Dem figure is from the previous poll which gave them an abnormally low figure and this one which gives them an abnormally high figure.

    In the meantime the government and the opposition seem to have reached a bit of a stalement for the time being with the Con lead consistently in the 11/12% region. I think the electorate are tired generally with the respective Party lines.

  12. I don’t think we should we should be putting too much emphasis on this poll at this stage.

    I haven’t picked up anything in the press or TV that would account for the libDems having a big jump.

    If you take a look at Anthony’s Polling average that covers the last five polls the results were;

    LibDems ; 17%, 16%, 16%, 16% and Now 22%.
    Lab ; 30%, 28%, 32%, 32% and Now 28%.
    Tory ; 44%, 43%, 44%, 43% and Now 40%.

    The Others may be up too and I could see how as it’s dominated the news the Scottish budget might have boosted one party or another, even labour if it’s supporters thought it was showing leadership backing the budget.

    I just can’t find the narrative behind the equivalent of a 30% rise in libdem support mainly at the expense of the Tories.


  13. All this goes to show winning the next election for the conservatives, is not quite so simple as it may appear. Going from 44% (chunky majority) down to 40% (tiny majority) is quite a change. If cleggy can come up with some decent policies he could be in a decent position come the election. That’s if this poll isn’t just public opinion thrashing about randomly.

  14. “Scotland on a knife-edge
    POLL EXCLUSIVE: 38% support independence 40% oppose”

    “The latest TNS System Three survey for the Sunday Herald found support for leaving the union rose three points during the last quarter, while opposition to a separate Scottish state fell to its lowest level since the poll began 18 months ago… The poll was taken after opposition parties initially voted down the SNP government’s budget on February 28…The poll asked 971 adults how they would vote in a referendum on whether the Scottish government should open negotiations with Westminster on independence… When TNS System Three began polling on the question, shortly after the SNP entered government, opposition ran as high as 50%. The new survey shows the gap between opposition and support, which widened to eight points last October in the initial reaction to the banking crisis, has now returned to the two-point difference seen last June… the actual percentage opposed to negotiation is the lowest since the poll was first taken in August 2007… The poll also found support for independence was greater among men than women, the middle age ranges, and those living in the north of the country.”

    Unfortunately, the Sunday Herald do not publish the voting intention findings (if any were actually commissioned). Perhaps holding them back for next week?

    Note: TNS System 3 (a subsidiary of market research giant Taylor Nelson Sofres) is a member of the British Polling Council, and abides by its disclosure rules. TNS System 3 may be unfamiliar to many readers, but actually are one of the oldest and best established pollsters in Scotland.

  15. Anthony,

    The Sunday Herald has a TNS System Three poll showing support for Independence at 38% to 40% against.

    Before people go over the top or indeed claim this as me being partisan, i’ll say that I agree with Anthony’s health warning on these;

    It’s just one poll.
    Independence polls are extremely erratic
    The way the question is asked can have a big effect.


  16. I think there is something in Gin’s suggestion that the high LibDem figure may be partly due to some sort of differential response or respondent availability due to the severe weather.

  17. Peter it is of no use comparing the LibDem figures in a series of polls from different pollsters given their differing methodologies . What you can do to a large extent is to take the raw data from Comres Populus and Mori polls and apply ICM weighting to that data and get an equivalent LibDem figure which would vary from 16% lowest ( previous ICM poll ) to 22% ( current ICM poll ) with figures in between of 18% Comres to 20% Mori indicating a pretty stable average of around 19% with variation caused by the normal sampling errors .

  18. As a parent of comprehensive aged children I am not surprised that there is a lot of dissatisfaction with the heavyhanded, wasteful approach to schooling displayed by this government. For people between the age of 35 to 50 who have children between the ages of 11 to 18 this is a big issue.

    ICM does tend, rightly or wrongly, to give higher poll scores for the Lib Dems. And I will not be surprised if following polls give them scores in the high teens. It still demonstrates the potential for the Lib Dems to do a lot better than the typical 16% they tend to poll.

    If the next poll does show the Lib Dems above 18% at the cost of Labour and the Conservatives then I shall indeed feel rather encouraged.

  19. I’ve looked at the deviations between the LibDem polling value and the Weighted Moving Average going back to Nov 05 (292 polls). The mean deviation is 0 (naturally) and the Standard Deviation is 1.8. There have been 6 occasions where the deviation has been >4.5% (this one is 4.6%) and 4 of these have been ICM (1 YouGov 1 Mori) but in every case these have been bogus ie not borne out by subsequent polls.

  20. I noticed that there is a majority of people who agree there should be some kind of protection for British workers or at least a level playing field

    The loop hole of foreign employers only having to pay their skilled employees the minimum wage should be looked at

    What is the point of training for 3 or 4 years only to find you will get the same money as stacking shelves at Tesco

    There will be a lot of skilled workers out there worried if this becomes the norm as their standard of living will plumment only the foreign employers will gain from this


    I am interesred to know if any of the pollsters poll the same people say every 3 months or so or are all the polls taken at random every time

    I know this may prove very difficult to do but you would see at first hand whether there has been movement amonst the dont knows or between parties and their leaders

    Thanks Glenn

  21. It’s got to be a rogue. The LibDems have hardly been setting the political world alight of late with bold policy initiatives, etc.

  22. Not a rogue, but an example of the level of volatility which is catching the herd of self-appointed pundits napping.

    The economic climate is having divergent effects on different sections of society so I explain this as a sampling problem. Larger samples in this environment are required for the polls to maintain their integrity.

  23. I am a Lib Dem councillor. I think it is a combination of Nick Clegg getting more news coverage, and Vince Cable being on Desert Island Discs twice in 7 days on Radio Four !

  24. Having just analysed the last few weeks of actual local by elections….Voters not guess work.. show the Liberal Democrats holding their own and gaining from both the other major parties. Real Votes in Real Elections.

  25. In every single other case where a poll showed the LibDem vote c. 5% higher than the WMA it turned out to be a “blip” ie almost certainly a sampling error. In 292 polls you’d expect about 3 with this much error if they were normally distributed, and in fact sampling errors have “fat tails” so 6 is perfectly reasonable.

    We should never try to read over-detailed results from a single poll.

  26. NBeale,
    I don’t think it can necessarily be described as an error, but that there are clumps of LibDems which aren’t picked up by general polling. This fits the pattern we see when it comes to elections where localised hotspots for the party become apparent.

    Aggregated polls tend to smooth out such clumps and analysis therefore consistently underplays the actual state of their strength.

    If this were only a ‘blip’ in the sample, as you describe, it would be mirrored by equal numbers of ‘blips’ showing significantly lower LibDem support.

    It would be indeed be interesting to measure the scale and frequency of these ‘blips’ to provide evidence which could be looked at rather than just dismissing them with distribution predictions.

    My feeling is that there is a correlation in type between LibDem local campaigning and poll ‘clumping’ which enables them to float under the radar.

  27. I think everyone above including the POLLSTER ICM are wrong in this case with the Liberals.

    This time round – the British public know that there is only one way to oust the government & that is not to vote Liberal – no matter what the POLLS show at any time up to the GE – the Liberals will suffer along with Labour at the next GE.

    As i have predicted previously – it’s not going to be all doom and gloom for the Liberals – once Labour lose the next election and they break up as a party for good – the Liberals will come back as the 2nd main party to fight future elections.

    Cut and paste – as usual you heard it first from “the oracle”.

  28. Thomas: That would only be true if the distribution were symmetrical, and it’s not. I’ve looked at this: the mean LD error is indeed zero but the median is -0.36. Small under-recordings of LD support are more likely than the “Normal Distribution” would suggest. There have been 5 cases of an under-recording of >3% but none of >4%. However that’s statistics for you – you very rarely get exact Normal Distributions in practice.

  29. Mike “the Oracle” makes a good point. Public opinion seems very polarised re: Labour and its quite possible that there’s an element of “ideally, I’d vote Lib Dem …” that polls pick up but which would be absent in a general election (when the luxury of “ideally” is replaced for some by a “lesser of two evils” approach to political reality). My impression is that a lot of people, not only Tories, simply want this Labour govt gone at the next election and will decide how to cast their vote with that as their primary objective.

  30. NBeale,
    I hear you, but with an economic crisis on our hands it’s possible the trend for poll variability is increasing in scale or frequency (or both) – how is it possible to measure?