The Guardian has a new poll of Liberal Democrat target seats here. I think this is the first such study of these seats, which have suddenly become key to the election result. ICM polled the first 42 seats on this list, seats where the Liberal Democrats need a swing of up to 6% to win.

The share of the vote in these seats back in 2005 was CON 36%, LAB 23%, LDEM 35%. In ICM’s poll today they found support at CON 35%(-1), LAB 18%(-5), LDEM 39%(+4). That equates to a swing of 4.5% from Labour to the Lib Dems, and a swing of 2.5% from the Conservatives to the Lib Dems. This isn’t actually very good for them if you compare it to the national polls – the last ICM GB poll was showing the equivalent of a 7.5% swing from Labour to the Lib Dems, and a 3.5% swing from the Conservatives to the Lib Dems – in other words, the Lib Dems are doing worse in their target seats than nationally. On these figures, the Lib Dems would gain 10 seats from the Conservatives, and 11 from Labour.

It gets more interesting though if you look seperately at the Conservative-held LibDem targets, and the Labour-held LibDem targets. As Julian Glover rightly warns in his commentary, only 15 of these seats are Labour held so the sample size isn’t huge and some caution is necessary, but it appears to show that the Lib Dem advance in marginals is wholly concentrated in Labour held ones: taken separately, responses in Con-v-LD seats shows no discernable swing to the Liberal Democrats, but a swing of about 8 points in Lab-v-LD seats. That would result in the Lib Dems taking
about 28 or so seats from Labour, but few if any from the Conservatives. If this finding is at all accurate, it will be key to the result.

71 Responses to “ICM poll of Lib Dem target seats”

1 2
  1. Sue – read Julian Glover’s article in the Guardian. He says if you break the data between Lab-v-LD seats and Con-v-LD seats, there is virtually no swing to the LDs in the Con-v-LD seats, and an 8 pt swing in the Lab-v-LD seats.

  2. This what we have been waiting for – a poll in LD target seats.
    Your third paragraph is most interesting. Essentially this is what happened in 2005. LDs made no progress in Con held seats and actually lost ground. However they picked up some surprising Lab held seats which were not even regarded as marginal eg Manchester Withington.
    I can certainly see LD picking up 28 Lab held seats as article states whereas their net gain from Conservatives should be in single figures.

  3. “In the 42 seats in which the party came second in 2005 and which it could take on a swing of 6% or less, Lib Dem support is now 39%. That is four points ahead of the Conservatives on 35% and 21 points ahead of Labour, on 18%.”

    Worth a read.

    I have done a very quick check of what the 42 LD targets would mean.

    27 Con/LD marginals. The lead here is only 4% which is only a 2%swing. Thus only 6 would theoretically fall.

    15 Lab/LD marginals. All could fall on the 6% swing ie if LD lead Lab by12 %. The currrent lead is 21% ie a swing of 10.5%. This would mean all 15 would fall. Incidently another 27+ Lab/LD marginals would also fall!

    Total losses would be in the order of:
    Con 6 maybe
    Lab 15 + 27 + more (my charts don’t go that high) = 42+

    With this sort of polling percentage quite a lot of Lab seats not in the top LD 100 targets would fall but no more Con.

    For posters above that talk of further gains in Con/LD marginals, this already includes those marginals and on this poll only 6 are vulnerable. There are unlikely to be many gains from Con.

    For further LD gains you will have to look to more Lab/LD seats of which I have already done about 27+ of them.

    At the very most if LD were to get an 8% lead over Con (4%swing), it still translates to another 8 seats max.

    By the way there are not 42 seats in which the LD came second that need only a 6% swing. To get this figure you have to include Scottish and 3 way marginals where LD came third.

  4. The reason why the LDs won’t win many Tory seats is because in many of their target seats the Labour vote is already very low and can’t go down much more, so if the Tory vote remains the same the LDs will find it hard to increase their vote.

    Only if the number of those elegible to vote is fixed and all elegible voters voted last time and will definitely vote this time.

    If turn out was sub 60% last time in a constituency (average) and 60% = 60 voters, then there are 40 voters who didn’t vote last time who could this time.

    Furthermore, what of those who have, since the last election, become elegible to vote?

    In this election, like no other in the last 65 years, to make an assumption, has never been more likely to make an
    ASS out of yoU and ME!

  5. @greengrass

    “the debt mountain behind it which will top £1TN by 2014 and we have no idea how to even begin paying it off”

    As against GDP of £1.4 TN – equivalent to a 75% loan-to-income mortgage

    And banks haven’t worked out how to take the whole of income in mortgage repayments – except for Lemming Brothers et al

    Not claiming that they have (and i take your point on my first incorrect response), but the income of the government is c40% of GDP so the tax take in 2014 would be c£560billion against debt of c£1000billion+ unfunded pension liabilities of c£600billion + additional off balance sheet (yes just like Lemmings) of c£60bn. So equivelent loan to income is 300% modest perhaps for a 20 something (Tiger economy) But not for a mature pensioner (Western European country) surely?

  6. @underdog

    there is very little evidence of ashcrofts millions being spent here in Eastbourne, all activity is from the LibDems, either they think theyve lost the seat or they dont think the LibDems will gain the handful of votes they need. Cons have a majority of 755 here over the LibDems

  7. sUY7

  8. sorry about that misposting

    YouGov: 1/10 less likely to vote labour cause of “bigotgate”, only 26% think Brown apology genuine

  9. I don’t know what that last post was -don’t get it i’m afriad.

    I agree with the poster (Statto was it) who was diisappointed that they mixed up lab/ LD and Con/LD.

    Just one poll of each type with a reasonable sample in each type would have been perhaps more instructive.


  10. A point was raised about the Lab/Con targets and how many Con would get on this sort of %. It amounts to a swing of 8.5% average. Discounting Scottish marginals it amounts to a staggering 120+ seats that would fall. I do not believe this will be the case as I think some of the heartland turnout for Lab will buck the trend. However it did occur then it would see Con in as a majority govt.

    LD +15Lab +27Lab +6Con = +48
    Con +120 -6toLD = 114
    Lab -42 -120 = -162

    Final figures would suggest

    Con 215 + 114 = 329
    Lab 344 – 162 = 182 + 3 SDLP
    LD 63 + 48 = 118

    Result Con overall majority
    LD still in 3rd place on seat

    These figures were done by looking at the targets seats individually – not using UNS or other prediction system, just hand-driven maths.

  11. I agree with the poster (Statto was it) who was diisappointed that they mixed up lab/ LD and Con/LD.

    Just one poll of each type with a reasonable sample in each type would have been perhaps more instructive.

    If you look at my figures you will see that I have separated them into Lab/LD targets and Con/LD targets. I have also said what it would be if these rest of the LD targets were considered. You have the figures you wanted.

    The only notes of caution I would take are:
    1. Individual seats have local factors which can shade results.
    2. If there are further poll changes especially post debate then these figures will be altered, however the scope to change things has shortened – only one week to go.
    3. This is a snapshot of what it could mean on the groound if the result was repeated on 6th May.

  12. My view is that the Conservatives will end up on around 325 seats, Labour around 200 and the Libdems just short of 100.

  13. Not surprised that Tory vote little down. It would be odd were it so, given national situation.
    A little surprised that LD not further up.
    Labour probably about right.

    However, I doubt if the full effect of tactical voting comes through in these polls. These targets seats will have had leaflets screaming various two-horse-race messages, but it does take time for them ti sink in.

  14. @david

    If that’s true in most of the Con/Lib marginals, as it looks like, then it raises a spectre that the Conservative Vote share has stacked up in the low-hanging-fruit of their Labour targets, and Conservative safe seats.

    It looks like any seat premium the Conservatives have gained from the Ashcroft marginal focus, has been eaten into by the Liberal Democrats.

  15. @David

    Your seat is not typical of Con/LD marginals. On the evidence of these polls alone, yes it would fall to LD. However there are local factors. There is a popular ex- LD councillor standing as an Indep, which is more likely to draw away more LD voters than Con voters. Eastbourne may stay Con against the National marginals trend, IMO there could be several recounts.

    The polls suggest that the Con vote has been increasing, but because the polls have been focussing on national % then they have not been reflective of more regional aspects. The only regional polls have shown that regions are behaving differently – much the same as the regional poll in Scotland is vastly different from the rest of the UK.

    This poll of LD targets shows LD doing much better in the North against mainly Lab held marginals, but significantly less well in the South against mainly Con held marginals. These % figures include the debates bounce and new voters. Yes, they may have margins of error, but MOE can go both ways not just in the favour of LDs. Yes, other factors such as ‘gaffs’ (current & future) and the third debate will factor in changes, but if things stay the way they are then we can estimate what the results would likely to be.

    Tactical voting cost the Cons about 33 seats last GE, without it Lab would not have got an overall majority at all. The Con/LD seat majorities already assume that the same degree of Tactical voting will occur and therefore the majorities give what EXTRA tactical voting and new voters are needed to take the seat. We know from polls of LD prospective tactical voters that instead of voting Lab, 25% of them will actually vote Con. That is new and reduces the effect of LD tactical voting. Of significance this year is the Con tactical voter. If Con have no chance of winning then 13% of the Con vote in that seat have said that they will vote tactically. That vote will probably be solidly against Lab. This will have the effect of seeming to boost the LD vote in Lab/LD marginals. It should allow LD to take more Lab/LD marginals and the polling showing the huge Lab/LD swing may actually be capturing a smaller Lab/LD swing and some Con tactical voting.

  16. Living in Guildford, I can personally vouch for the crazy amount of Lib Dem support here. Whilst not an entirely precise methodology, going down The Chase there’s one blue sign and about fifty orange ones. The student vote will likely hold the balance of power, as the registration rate has gone up considerably since last time around, so the Lib Dem support is up considerably.

    On a larger scale, the idea that the Lib Dem support has grown less in their targets than nationwide is a great demonstrator for the inaccuracy of polls – larger samples are needed.

  17. A good performance of any of the leaders in tonight debate will throw the whole polls up in the air once again.. and this time i guess about 14 million if not more will be watching

  18. @Will B

    Your points about the very Young voter are valid for where there are peaks of new LD voters. These peak areas are normally associated with university student areas. But that is not valid for the majority of constituencies. You should be careful of judging by the enthusiasm that LD young seem to have and show in the number of posters. Posters does not mean votes in the polling booth. It would not be seen as ‘trendy’ for young Cons to advertise their new found allegiances in the same way as young LDs.

    As for the views of the 18 – 34 age group, that has been surveyed. In 2005 the breakdown of this group in a YouGov poll only 4 days ago gave voting intentions of:

    2005 Con 26.5, Lab 38, LD 26.5, Others 9
    26Apr10 Con 34, Lab 26, 35, Others 6

    This shows that the young voter is splitting roughly evenly between Con and LD, with far fewer going to Lab, a complete change from 2005. This means that generally the number of young LDs is beiing matched by the number of young Cons, but in the particular area of recent new young student registrations the LDs have a distinct lead. The problem is that this difference is highly pinpointed and not spread region-wide, let alone nationally. Hence in specific constituencies the total LD surge will have a profound effect, whilst in others the effect will be dissappated or ofset by young Con in this agegroup. Unfortunately for Lab, there would seem to be nomajor offset effect. Few Con/LD or even LD/Con marginals are likely to be in these LD peak areas. It is Lab in its university seats/areas such as Newcastle etc. that will feel the full brunt.
    I hope this help explain why Lab/LD marginals seem to be at far greater risk than Con/LD marginals in this poll.

  19. Agree that the list of marginal seats is not the real list of Lib Dem targets. (I’m not very involved this time but used to be on the Exec of the Lib Dem Parliamentary Candidates Association and stood in Brentford and Isleworth in 1997 and 2001).

    The real list of Lib Dem targets is based on a number of factors:-

    – Votes last time
    – Likelihood of taking votes from incumbent and 3rd party
    – Progress in terms of council seats etc and local memberhsip
    – Quality of local organisation
    – Geography (you don’t want all your targets close together if the intention is to get people from neighbouring constituencies to help out)

    There are also some seats with very good local organisation which are not official targets but the LIb Dems could win (this is in addition to the student points from other comments). e.g. I don’t think Lynne Featherstone’s seat was officially a target – she put together most of her impressive campaign organisation herself with little support from the central party.

    This all leads me to believe that the Lib Dems have a good chance of winning a few more seats than you expect from the polls you refer to

  20. I think what we are now seeing is an “anyone but Gordon” swing. This is benefitting both the Tories and the Libs. If the polls are reflecting peoples intentions of tactical voting already then this would explain the LibDem surge. It would also explain the odd “rogue” poll. Yes, they claim that a poll showing the Tories on 38% is a rogue poll – but it is in line with what the polls were saying at the beginning of April, so they are not that rogue. A slightly different phrasing of the pollsters question would result in different results depending on how people are intending to vote or their preferred outcome. So I would say today’s calculations showing both LibDems and Tories taking a lot of seats from Labour are probably true. In fact I would say that the deteriorating weather and the “Duffy” effect together with general apathy amongst Labour voters is likely to see a bigger collapse in Labour support than the polls reflect at present.

  21. Looking into the detailed figures:

    Cons. held Cons/LD marginals:
    Conservative lead holding up, if anything increasing slightly; Labour vote collapsing, but shared between all parties.

    Labour held Lab/LD marginals:
    Labour losing a quarter of their vote to the LibDems; Conservatives & others staying the same.


    Look at one of the other questions:
    “does the fact that Gordon Brown leads the Labour Party make you more or less likely to vote Labour?”

    In Labour held seats, 23% of people who voted Labour in 2005 say “less likely.”

1 2