When you sit down to do a marginals polls one consideration is where you draw the line: what is a marginal? The thing you want to avoid is under or overshooting the real battleground – the risk is that you poll lots of seats that need a swing of up to 5% and find a swing of 10%, enough to win lots of seats you didn’t bother polling.

That’s the sort of thing that’s happened in the third of Lord Ashcroft’s three sets of marginal polls – full details here. He polled the four most marginal LD -v- Lab seats, Norwich South, Bradford East, Brent Central and Manchester Withington. These need a swing of up to 2.1% to go from LD to Lab, which unsurprisingly Labour get easily. The average LD => Lab swing in these seats was 15%, confirming that the Lib Dems are doing much worse where they are up against Labour and easily enough to unseat almost all Lib Dem MPs with Labour in second place. In practice of course we can’t actually be that confident that voters in a tight LD-Lab marginal will behave the same way as in a seat where the Lib Dems have a 20% majority, so it’s a bit of a shame Ashcroft didn’t include some more challenging LD-Lab fights like Cambridge, Hornsey & Wood Green or Bermondsey.

An interesting thing to note is that the Greens are doing notably well in a couple of these seats. In Norwich South they are in second place on 20%, but that was one of their target seats anyway, they are also doing well in Manchester Withington, up 8 points on 10%.

While it’s hardly a LD-Lab battle, Lord Ashcroft also polled Brighton Pavilion, Caroline Lucas’s seat. Voting intentions there with changes from the general election are CON 18%(-6), LAB 33%(+4), LDEM 5%(-9), GREEN 32%(+1), suggesting an extremely tight race between Labour and the Green party.

193 Responses to “Ashcroft polling of LD-Lab marginals”

1 2 3 4
  1. New thread as predicted!

  2. “the risk is that you poll lots of seats that need a swing of up to 5% and find a swing of 10%, enough to win lots of seats you didn’t bother polling. ”

    In fairness, if you get a swing of 10%, then it’s likely that whatever your problem (or triumph) was, your polling of marginals probably wasn’t the hinge factor.

  3. Argh, otherwise intelligent people on Twitter are posting insane UNS calculations extrapolating the result from the 4 most marginal Lib/Lab seats to every Lib Dem seat.

  4. “@AW
    These need a swing of up to 2.1% to go from Lab to LD,”

    LD to Lab I presume ?


    *thicko alert*

    I’m confused by Ashcroft’s commentary yesterday when he said:

    Despite the small Conservative lead, voters overall would prefer to see Labour in office than the Tories. Nearly a third (32%) said they wanted a Labour government, and a further 8% a Labour coalition with the Lib Dems. One quarter wanted the Conservatives in government alone (25%), and a further 9% in another Con-Lib Dem coalition. Notably, only four fifths of Labour voters and three quarters of Tories wanted to see their respective parties governing alone.

    How does a Con poll lead of 2 equate to the majority declaring they wanted a Lab govt ?

  5. “How does a Con poll lead of 2 equate to the majority declaring they wanted a Lab govt ?”

    Presumably more of those from other parties gave a preference for a Labour government.

  6. I think the Greens will lose Brighton because of the Tories.

    People who want to see a Labour Government will back Labour and I think a general dislike of the government will swing it their way.

    Not really fair on the Greens but then if politics was fair I’d still be a Councillor!


  7. Thank you Roger !

    In my defence, I did at least warn that it was a thicko alert & I’m ashamed to say that it has been proven……

  8. “more challenging LD-Lab fights like … Hornsey & Wood Green “.

    Well, after the May Borough Election, the challenge is seeing how Ms Featherstone could conceivable hold on. She has already pillaged the natural Tory vote,

    Electoral Calculus extrapolate the 2014 outcome to (round numbers) Labour 22,500, LD 15,000, Con 6,000, others (and in 2014 this was mainly the Green vote) 11,500.

    In other words, for a LD hold in HWG, the overwhelming bulk of the Green vote has to go LD, and there is no slippage from LD to Con.

  9. These Ashcroft polls do throw up some strange findings.

    51% are either satisfied with David Cameron as PM or are dissatisfied with him but would prefer him to Milliband as PM (in Lab/Lib Dem marginals)?

    Doesn’t seem to correlate with the other findings in the poll.

  10. As we say in Hornsey and Wood Green ‘When we swing we swing big’

  11. Interesting to see the media reaction to recent polls. With some ambiguous polling, they have largely focused on the ones showing Con gains, slipping all too easily into the Juncker effect meme.

    This could be real, or it could be a bit of myth making in the process.

  12. It is interesting to look at the results for the difference between the generic “which party would you vote for” vs “thinking specifically about your constituency which party would you vote for”

    A reason often given for not prompting for the smaller parties is that results in overestimating their support, and you can see that in the different results to these two questions. In most cases the support for the smaller parties drops, unless they are a contender.

    A reason for the different levels of support is often given as the incumbency bonus, but it also seems to be a case of people will only vote for parties who stand a chance of winning in their seat.

    It would be interesting to see a national poll where they ask both questions to see what happens to the UKIP and Green vote and see where that goes.

  13. @Malcolm Redfellow

    I have made a point of talking to Green supporters and members while canvassing and at the local election count in Hornsey and Wood Green and most have said in the GE they intend to vote for the Labour candidate Catherine West.

  14. @MikeB

    Ahem! — I stood over the counting tables for the HWG poll, and watched the split votes. As you say, when the Green votes split, Labour was unfailingly the other preference. In the Wards I was monitoring, there were a lot of ’em (heart and head, perhaps).

    So I would concur with your impression.

  15. @Malcolm Redfellow

    The Greens also seem to be the people who dislike the LDs even more than Labour people do. If Featherstone is counting on Green votes to save her skin she is finished.

  16. Electoral Calculus updated state of the parties for June

    Current Prediction: Labour majority 48


    Lab 35%
    Cons 31.5%
    UKIP 17%
    LD 7.5%


  17. a world where labour pick up 17 seats of the lib dems, is a world where labour get an overall majority…

    can’t see it myself.

  18. Greens always semi-identified with the Liberal Woolies and SDP side of the LibDems, and even considered them political allies. But as the LibDem party has now all but transformed into the Orange Booker party, the Greens feel a great deal of antipathy towards them and even a sense of betrayal.

  19. Featherstone will be booted out of parliament in Hornsey…the labour results were remarkable in the local elections, and their activist base in that neck of the woods is young and hungry, by all accounts.

  20. Very disappointing poll after what I expecting we might see today. Fits into the category of “no carp Sherlock” (hi Mr Croft!)

    I am guessing Lord Ashcroft will be doing some more before the election and maybe this was just a baseline to work out where he will poll next time but like AW says I wish he had gone for a few that will be closer to call and might give some idea of the scale of Lab gains from LD.

  21. Surely what the Brighton Pavillions data shows is a strong anti Labour tendency – where people who do not want Labour look around and see the Greens as their best bet.
    This does not paint a good picture for Labour

  22. On an output per hour basis, UK labour productivity decreased by 0.1% in the first quarter of 2014.


    “There is still no sign of the rebound in productivity growth expected in most economic forecasts,” said Work Foundation chief economist Ian Brinkley. “With productivity growth flat, there is little scope for many employers to significantly increase wages, and the pressure on living standards, for many in the labour market, will continue.”

    source The Work Foundation

    How long can living standards keep falling…6 years and counting now

  23. @Peter Crawford

    I think Hornsey and Wood Green has the highest membership of any CLP. Also it may have had the highest contact rate in the recent canvas returns This means there is a lot of information as to where Labour support is in the constituency. The ‘ground war’ is something opinion polls do not measure.

  24. The only time Brighton Pavilion had a Lab MP was between 1997 & 2010.

    Prior to 1997 it had been a Con stronghold since it’s creation in 1950.

    The Green council is not popular with a large number of constituents so in 2015 it will probably be a fight between Lab & Green with probably UKIP in 3rd.

  25. I was talking to a former mayor of Wood Green today. Now 95 and still active in our party. They do breed them tough in Wood Green.

  26. Have I got a comment in the queue or has it got lost? I’ll be annoyed if it’s the latter – it took about twenty minutes this morning.

  27. Wouldn’t surprise me if the GE is not as close as most predict.

    I think there is just too much anti-Con resentment for them to do well.

    [Hi Mr. Shev]

  28. “So if ISIS has called for all Muslims to take a pledge of allegiance with essentially the new Caliph of the new Islamic State, what will Islamists in Britain do? Will they give their oath as their doctrine implies, and continue to support the Jihad either by travelling to the region or by staying here and undertaking further activities; or will they go against their principles and decide, after a period of self-reflection, that their ideas are not as robust as they once thought they were. After all, is the “Islamic State” not what they have been waiting for?

    Perhaps they are waiting for half term holidays or for Baghdadi to introduce the welfare state – child benefits and job seekers allowance and free healthcare…or maybe they are just waiting for him to instate the concept of freedom of speech that they are used to?”

    Haras Rafiq
    Quiliam Foundation.

    :-) :-)

  29. Re: [Lord Ashcroft]’s “Constituency Voting” question.
    It looks as if he asks every respondent this question after asking the standard VI – and this produces changes. It seems to me that asking these 2 quesions this way is implicitly prompting the respondent to consider changing his/her vote. Would these changes be the same if he asked them in reverse*, or if he asked one set of respondents one question and another set the other? My suspicion is that in a real vote a fair number of voters won’t think about their constituency when voting, because they are not prompted, and that Ashcroft’s question overstates the local change.
    * sounds mad, but you could ask “your constituency” first and “if you moved to another constituency…” afterwards.

  30. I’m thinking that polling in 4 seats is not enough to extrapolate nationwide. Especially when what happens with LD seats is currently highly inconsistent with UNS.

    Norwich South is near to being a 3-way marginal, although I do not really expect the Conservatives to win there soon. I wonder why that was chosen? It might be considered to fall into a different class of seat.

  31. agreed those four ld/labour marginals were bizarre polling choices…they are all practically guaranteed labour gains…

  32. @Keith, Peter

    We have to remember why Ashcroft spends his “hard earned” money this way – to help the Cons! An exhaustive survey of all LibLab marginals wouldn’t do that – in fact it would do the opposite, help the Libs and Labs!

    Why poll them at all then – especially these very obvious Labour gains? Two reasons. First he has to look even handed – if he only polled where it helped Tory strategists his money would look like a direct donation. Second, he is temperamentally a Jeremiah – he’s warning the Tories that Labour will be winning these seats, and they can’t just rest on their laurels.

  33. COLIN
    “Perhaps they are waiting for half term holidays or for Baghdadi to introduce the welfare state – child benefits and job seekers allowance and free healthcare…or maybe they are just waiting for him to instate the concept of freedom of speech that they are used to?” Haras Rafiq

    Irony may not meet the case in a movement which has developed so fast, and in which we do not know the further direction of a British ISIS membership. If Rafiq can speak elsewhere, as he has, of long term entryism as a fundamentalist strategy, he may also think of an ISIS jihad as xiliastic, one in which martyrdom and massacre are desirable ends and any t errorist act as an acceptable means. No room there for irony, I suggest.

  34. @HOOKESLAW: “Surely what the Brighton Pavillions data shows is a strong anti Labour tendency – where people who do not want Labour look around and see the Greens as their best bet. This does not paint a good picture for Labour.”

    I think it’s just that people there know that a Green vote isn’t likely to be a wasted vote. That may be because they support the Greens or because they’re seen as the best choice for an anti-Tory vote.

  35. @ Bramley,

    Ashcroft asked a separate “What kind of government would you prefer?” question after the standard VI question.

    The table of the results is actually quite interesting:


    Tories have slightly worse retention that Labour, but the big driver of the preference for Labour is the 3:1 Lib Dem split for Labour majority over Tory majority. Also notable: Kippers prefer a Tory majority to a Labour one by 35%: 25%- ie. not by much. There are two ways to look at that- both parties are losing voters to Ukip, but if they come home at an equal rate they are only going to break 7:5 for the Tories.

  36. Another point of interest from Lord Ashcroft, this time from the Lib/Lab marginals: even here in the Labour landslide zone where the Tories have absolutely no hope of winning Ed Miliband’s ratings are abysmal and a majority of voters prefer Cameron for PM.

    This supports the “priced in” hypothesis of Miliband carpness, at least provisionally- Labour voters are not impressed with him but they’re voting Labour anyway.

  37. the brighton pavilion of 2014 is completely different from the seat represented by Julian Amery and Sir Derek Spencer, between 1969 and 1997….many, arguably most, seats don’t change that much…but Brighton Pavilion I think has…labour could win it back again.

  38. A fair number of Green voters in Hallam. I’ve not spoken to a single one who prefers Clegg to Labour or would vote that way as a second preference.

  39. Greens must be disappointed with having to fight hard for Brighton next year.

    Normally when you get a high profile, active MP who doesn’t get a say in government (rather like the situation LD’s used to have between 1997 and 2010) you’d expect to be more entrenched than that by now and looking at an increased majority even if it was a highly marginal seat. I’d have expected the hard part to be to win in the first place and the easy bit getting established.

    I guess not having your bins emptied must have made the difference?


    That’s probably true but it could still be a very potent factor with the current “don’t knows who mean to vote.

  41. @Postage Included

    Given how much importance that Ashcroft gives to the constituency and candidate question, I do wonder why he bothers with asking the first one at all. I agree that asking a second question focuses unduly on the subtle difference with the first and as such implicitly prompts a change of response in favour of the incumbent. That is, it’s almost like following up the first response with “Really? Even in this specific constituency and with the candidates standing here?”

    The issue is crying out for some polling research to resolve the question. All it needs is a split panel, with one half being asked the constituency and candidate question alone, and the other being asked it as a second question (as now). I think the differences in the two sets of responses would be statistically significant.

    The other problem is that in terms of name recognition I think the incumbent has an advantage over the challenger at this stage. Few people will know the name of the main opposition challenger at this stage. That will change during the election campaign itself as local campaigning takes effect.

  42. Phil – the thinking behind it (since it’s something I came up with originally) is that the first question essentially cleanses people’s palate, it allows people to get their national preference off their chest before we ask the constituency question.

    I don’t think it unduly makes paople answer the question differently when their national preference is actually the same as their local voting intention as in run of the mill Con-Lab seats it makes hardly any difference at all. It’s only Lib Dem seats where it seems to make a big difference.

    I did mean to test it in our final YouGov poll before the last election, but alas I was off doing BBC rehearsals and didn’t get round to making sure it went on the survey!

  43. Anybody else noticed that Angel di Maria, the left [foot] inclined Argentinian who had a carp game against the Swissies then redeemed it by scoring the only goal right at the end….has a decided look of Ed Miliband about him?

    Could this be a portent?

  44. R&D – yours of 11.43 on the previous thread

    You will remember that I specified that the general law has to be upheld (I gave the example of burning religious books, but yours of incitement to murder is, of course, even more valid). A religious veil being likened to a mask is a bit misleading, as usually a mask, by its very nature, intends to portray an alternative ‘person’, whereas a veil intend to hide the person entirely. Of course, the same could be said of stockings hiding the head of a bank robber – but then we’re back into the realm of criminology, not religious beliefs.

  45. @AW

    ‘run of the mill Con-Lab seats’ seem to me to be fewer in number now than at any time since the 1920s. Many people have Green or UKIP or SNP or Other preferences and are having to calculate with great precision how to vote in the upcoming GE.

    And does anyone have a ready reckoner on how many three-way or four-way marginals (define!) there are?

  46. @ T’Other Howard,

    Well, I did say provisionally.

    The Grand Unified Theory of Carpness posits that the major anti-Miliband swing against Labour won’t begin until the election campaigns are underway, so we’ll have to see what these marginals look like next spring.

  47. @AW

    I’m not denying that it’s a good question, just that the format of asking it as a supplementary is questionable because I think that format will potentially exaggerate whatever incumbency effect is genuinely there.

    It’s generally accepted that the LDs gain more from incumbency, but you can’t use that to discount the possibility that all incumbency effects may be exaggerated through the secondary nature of the question.

    Anyway I’d like to see someone – YouGov? Lord Ashcroft? – conduct some polling with split panels and publish the results to try and put the issue to rest. Until that’s done we can debate it theoretically to our hearts content and be none the wiser.

  48. Phil – my point is it wouldn’t answer the question unless you had an election result to compare it too.

    The first question is deliberately there so people can state their national preference in the hope it will stop them giving their national preference in the local question. People sometimes answer the question they’d like to answer, rather than the question you’ve actually asked.

    A difference between asking it straight and asking it as a two stage question could mean it’s exaggerating incumbency OR it could mean it’s working as designed, and allowing people an outlet for their national preference that would otherwise be wrongly reported in the local option. Asking the first question is supposed to make a difference – but in a way that makes the second question more accurate – or I wouldn’t have put it there!

    We can test if it makes a difference… but it’s supposed to. We can only test if it makes a positive difference and makes it more, not less, accurate if there’s an election to compare it to.

  49. Brighton Pavilion…. That oddity of a seat for me where I will be cheering on Labour to the rafters….

  50. Did anybody else notice that Tory MP proposing a directly elected PM last week ? Not sure if it was meant seriously or just fly ing a kite for Cammo, along the lines that Ed M would obviously be too carp to be elected directly and can only get in by commanding a majority in the HoC. Cheating really !

1 2 3 4