ComRes released their latest poll of marginal seats today. As regular readers will recall, ComRes’s marginal polls cover the 40 most marginal Con-v-Lab seats (25 Conservative held, 15 Labour held). Unlike Lord Ashcroft’s marginal polls (which are actually a series of individual constituency polls in seats that are marginals, which we can aggregate together to get an extremely large sample across a group of marginal seats) ComRes’s poll is a more traditional marginals poll – a single poll of a group of marginal seats, meaning it gives us a measure of those seats as a whole, but has far too few people to tell us anything about the individual seats within that group.

Latest voting intention figures in these marginals with changes from the last time ComRes polled them in September is CON 31%(+1), LAB 39%(-2), LDEM 7%(+1), UKIP 18%(+1). These seats had Labour and Conservative equal at the last election so an eight point lead here is the equivalent of a four point national swing and a one point Labour lead in national polls…pretty much exactly what the national polls have been showing lately (actually if you look at the crossbreaks of the poll they suggest a swing towards the Conservatives in the Conservative held seats, a swing towards Labour in the Labour held seats, but given the sample size of those two groups and that the poll is only weighted at the level of all forty seats I wouldn’t put too much weight on that).

Note also that, judging from the tables, ComRes have switched over to prompting for UKIP in their main voting intention question in this poll – as with their last national poll, it does not seem to have had a major effect (UPDATE – I think this is because ComRes have changed turnout weightings, so that there is a tighter turnout filter for the Greens and UKIP than for the main parties). Tabs are here.

There should be another batch of Lord Ashcroft polls of individual marginal seats later this week.


147 Responses to “ComRes poll of marginal seats”

1 2 3
  1. First – having been nearly last on the previous thread before being zapped (apologies AW)

    Seems to be polldrums in the marginals as well as the country as a whole.

  2. It’s the polldrums, or the view of Electorate who have made their mind up quite strongly and won’t change easily.

  3. With the SNP doing so well in Scotland, and the Libs holding onto more Con-Lib marginals than a UNS suggests, I reckon the Tories will need a lead of around 2.5% or so to be the largest party.

  4. So basicly what the polls and marginals are saying is that overall the national picture looks neck and neck between The Tories and Labour.

    Something I have been saying over and over again to thwart any premature grandstanding.

    The GE is now nearer 5 months away than 6 months away.

    .
    .
    .
    .
    .
    .
    .
    .
    .
    .
    .
    .
    .
    .
    .
    ………….Tictoc!!

  5. I’ve been knocking some figures around today, in a system based on probabilities, and I rate the chance of 3%+ Conservative at the GE to be about 30%.

    Overall I have the them at about 75% chance of being the largest party.

  6. Lord Ashcroft tweet
    Released tomorrow polling in Doncaster North showing UKIP 2nd so if Tories tactically vote for UKIP Miliband loses

  7. On these sort of numbers how far away are Labour from having a majority in England or England & Wales?

    A minority (UK) government that needs C&S for UK wide legislation but has a majority for enacting English only laws would be an interesting possibility if the numbers added up.

  8. CATMANJEFF

    Once [the papers] let loose the opening salvos of the campaign proper then I would put the probability of the Tories becoming the largest party at 100%.

    Unless the pro Tory media (News Corp and others) manage to pull off a spectacular then I can’t see a Tory majority.

  9. Northumbrian Scot

    “A minority (UK) government that needs C&S for UK wide legislation but has a majority for enacting English only laws would be an interesting possibility if the numbers added up.”

    That’s a situation that I have envisaged before.

    With the Smith Commission report going to the printers in 10 minutes time : widespread agreement in Wales & NI that they should have the same powers as Scotland : the reported panic in Westminster Labour : the post May 2015 UK Government having responsibility for implementing (or not) Smith recommendations – that would be interesting.

  10. Probably better news for the Conservatives than Labour, this poll.

  11. I think we’ll have Con + LD, or Lab minority, with SNP helping out on occasion this time in 6 months.

  12. If Ashcroft is right about Doncaster North, we could have an even more fascinating situation of the largest party having to find a new leader!

  13. @Oldnat

    I think there’s a good chance that UKIP might pick up seats on the tactical voting of both Con and Lab. There’s an irony in there somewhere.

  14. According to your Swingometer Respect will gain 10 seats, the Scot Nats and Plaid no seats.
    This bell does not ring true!

  15. Allan Chistie,
    What on earth can you let loose about next?He is already according to you a
    looser,a nerd,out of touch etc,and yet annoyingly still keeps on polling ahead of
    the Tories..

  16. @CMJ

    Are you prepared to divulge any details of your “system based on probabilities”?

    As you have no doubt gathered I am interested in the overall challenge of making such predictions and I always prefer to be able to look under the bonnet.

  17. Good Evening All.
    Ann in Wales.
    Hello to you.
    I think in the GE campaign the exposure which Ed M will receive as part of the media coverage looks likely to be an advantage to the Labour Party; IMO, of course. Voice, looks and image count for a great deal in the ‘modern’ world.

  18. ANN IN WALES
    Allan Chistie,
    What on earth can you let loose about next?He is already according to you a looser,a nerd,out of touch etc,and yet annoyingly still keeps on polling ahead of the Tories/
    _______

    If you’re going to quote what I have said in the past then fine but please don’t make rubbish which clearly I have never said.

    I think an apology is called for.

  19. Oh, the MP for Doncaster North is one Edward Miliband. Now I get it….

  20. @Ambivalentsupporter

    “With the SNP doing so well in Scotland, and the Libs holding onto more Con-Lib marginals than a UNS suggests, I reckon…”

    Can’t really argue with the first of the premises on which you base your conclusion. But I would argue that the second is the *opposite* of the current state of play and so I wonder what source of evidence you are using as a basis for making this statement.

    My reason for challenging it are set out in a long post uploaded yesterday. Basically, calculations using Ashcroft marginal data indicate that the LDs are losing about half their support in each constituency on a proportional basis together with a further uniform reduction of about 3%. I won’t repeat the calculations here as they have already been set out very recently.

    If this is correct, then LDs will be suffering greater-than-UNS VI losses in their own seats: by definition this is where their share of the vote started out highest and so falls most under proportional vote loss.

    I should point out that Ashcroft has looked at quite a good sample of Con/LD and my LD vote-reallocation formula is basically just a summary of his own explicit findings.

    If you have a different source of evidence I’d be very interested to hear about it.

  21. OLDNAT
    “If Ashcroft is right about Doncaster North, we could have an even more fascinating situation of the largest party having to find a new leader”
    _______

    Ed Miliband Member of Parliament for Doncaster North.

    Uh oh!!

  22. @Unicorn

    I need to go out to buy some food, so I’ll get back in about an hour.

  23. Guys I know it’s kind of the point but you’re getting far to ahead of your yourselves!! Five months out in 2010 no one would have said the LDs would have held the balance of power. A major cockup or great debate performance from any of the five main parties could swing it in any direction. To quote Alex salmond (I think!) this is still the phoney war!

  24. On that Comres poll, very confusing like their last national poll.

    See page 8, bottom line

    142 Don’t know’s. Of those 25 intend to vote Cons, 29 intend to vote Labour, 18 Lib Dem, 6 UKIP.

    But I thought they didn’t know how they would vote?

    And why doesn’t the voting intention column on page 8 match the voting intention row on page 8?

    Example the UKIP row – 169 respondents said they would vote UKIP, but only 95% of those, 129 intend to vote UKIP in the column. Well who are the other 40 going to vote for?

  25. In 2010 ukip plus eng democrats plus bnp was 8.6 in Donc North so fairly large racist right wing base to start from.With ukip currently polling l5 plus on their own then there will be labour seats where they push tories into third.

  26. I still see some kind of “accommodation” between Tory and UKIP, before the GE.

  27. Help ma Bob…lock up your kids, the Scots are coming!

    http://www.bbc.co.uk/news/uk-scotland-south-scotland-30217248

    Peter.

  28. From my perspective, what catches the eye most about this poll is the Ukip VI figure of 18%. This is three points up on their current UKPR polling average. But what is most interesting is that the Ukip 2010 voting share was significantly lower than average in the Con/Lab marginal seats of this kind. (This come from analyses I described a few days ago.)

    In my earlier post on this I interpreted this low 2010 Ukip vote-share as an indication that their supporters had opted to vote tactically in these seats (presumably with the aim of keeping out the Labour candidate wherever they could).

    Various conclusions can be drawn from the fact the high Ukip VI has been built up from a lower-than-average base. The main one is that if this is their genuine voting intention the Ukip supporters have stopped using their votes to block another party and are instead following the leader’s advice to ‘vote UKIP and *get* UKIP’.

    In short, in the face of all the advice being promulgated by [the press], this polling evidence suggests that they are moving in precisely the opposite direction..which could obviously have serious implications for the GE outcome.

  29. Allan

    [NewsUK] are aiming for the spectacular. It was made clear to Cameron from the highest level that a majority must be the aim, as only this can deliver the necessary changes that [they want] to see. I’ve no doubt the largest party will be the Tories but [people at NewsUK] want and most of us, expect more. Particularly as we will see the most brutal and intense campaign in British electoral history.

  30. So taking just the poll figures for Comres from page 8 without any turnout adjustment or adjustment for don’t knows we would have a headline of

    Cons 29
    Lab 36
    LD 5
    UKIP 23
    Green 6

    So while they may now be prompting for UKIP, whatever they are doing in their turnout adjustment/ adjustment for don’t knows sends the UKIP figure down 5%, same as what we saw in their national poll.

    So the figures on page 8 before all their strange adjustments look like the better figures to use in my opinion, and if we look here at the regional cross breaks we see UKIP 2 points behind the Tories in South East and Northern Marginals, giving Labour a clear lead there, and UKIP nowhere in London marginals, which are now tied between Tories and Labour.

    Lets see if that matches with Ashcroft’s figures tomorrow.

  31. Richard – that’s the squeeze question. If people say don’t know, ComRes then ask them how they would vote if they were legally obliged to do so.

    So in this poll 142 people said don’t know. ComRes then asked those people how they would vote if forced to. 25 said Conservative, 29 said Labour, etc, etc.

  32. @07052015
    The Tories should be happy to be in 3rd place in some area’s of the north. It used to be the LD’s who put them there, now it will be UKIP.
    I am not a kipper, but I well recognise the only force likely to take Labour on in these places, is UKIP.

  33. Richard – Ah, I think I’ve spotted what ComRes have done. They’ve gone back to only taking UKIP & Green voters who are 10/10 certain to vote.

    So for Con, Lab, LD voters, they include people who say their chances of voting are 5/10 or more (but weighted, so that someone who says 5/10 is worth 0.5, someone who says 6/10 is worth 0.6 and so on). For UKIP they only include people who say they are 10/10 certain to vote. That’ll be why the prompting hasn’t increased their reported level of support.

  34. @PRESSMAN
    I really don’t think that even the average SUN reader is going to be swayed by “MILIBANDS GOT BUCK TEETH” on the front page.
    Nor will “BULLINGDON POSH BOY” in the MIRROR do the trick.

  35. @Anthhony

    Thanks, that explains the don’t knows who now know how to vote, but what about the difference in the rows and columns on page 8?

    Conservative 212 in row and column – matches, that’s ok
    Labour 263 in row and column – ditto
    LD 39 in row and column – ditto

    BUT
    UKIP 169 in row, and 128 in column – why the difference?
    Green 47 in row and 29 in column – why the difference?

    There seems to be a ‘small party’ adjustment to downgrade them?

  36. Anthony, thanks, just saw your update after my last post.

    So I am right then, there is a ‘minor party’ adjustment! To correct for the prompting…

  37. @ANN in WALES
    In fairness to Allan, I think the point is, Labour have a miniscule lead over the Tory party in the polls. But the publics view of their respective leaders is a very different matter.

  38. Richard – there clearly is in this one (look at page 5, where it’s crystal clear), but the likelihood to vote question doesn’t show the same clear contrast between the C/L/LD crossbreak and GRN/UKIP crossbreak in their last GB poll, so I’m uncertain if the same approach was taken.

  39. Re: Doncaster north… The first part of the Ashcroft tweet isn’t surprising at all – Sophie Brodie polled 21% for the Conservatives in 2010, compared with 4.3% for UKIP’s Liz Andrews, so on a uniform swing it’s no surprise that UKIP are second. But a big two party leader being defeated in their own seat? It hasn’t happened since Arthur Henderson (Labour) lost Burnley in the Tory landslide in 1931.

    A couple of things jump out from the 2010 result. Firstly, Miliband took 47.3% of the vote in 2010, so in order to be vulnerable to tactical voting, he must by definition still be under 50% and probably a few points below it. So becoming Labour leader hasn’t helped him outperform the party, in fact the opposite seems to be the case. He may well even be polling a lower vote share than at the last election.

    Secondly, the BNP and English Democrats between them managed a 12% share in 2010. UKIP seems to have been effective at taking votes from far-right parties, so it might have done so here. David Herdson pointed this out on PoliticalBetting six months ago – the linked piece also talks about local factors which could be relevant.

    It’s also worth noting that Tory voters aren’t known for voting tactically. Would the opportunity to decapitate Labour change that? We might not know until May, but tomorrow we’ll at least have some idea how much tactical voting would be needed.

    http://www.ncpolitics.uk/2014/11/marginals-update-plus-doncaster-north.html

  40. New GE2015 Forecast (NOV 26th)

    LAB 284
    CON 284
    SNP 33
    LD 24
    UKIP 3
    PC 2
    GREENS 1

    http://generalelectionuk.blogspot.co.uk/2014/11/ge2015-forecast-nov-26th.html

  41. Roland

    I think you probably know that the campaign will be much more subtle and multifaceted than that. The primary aim for those not yet switched onto the election, and that is many millions, will be to present it as a straight choice between DC and EM. The fear of the other guy. The acquiescence if YOU vote UKIP to EM winning. The message that YOU have to do your duty for your country on May 7.

  42. @NUMBER CRUNCHER
    It’s also worth noting that Tory voters aren’t known for voting tactically. Would the opportunity to decapitate Labour change that? We might not know until May, but tomorrow we’ll at least have some idea how much tactical voting would be needed.

    This statement is very true, we just cannot bring ourselves to put the cross anywhere else. However, do we want Ed beheaded?

  43. Deal tonight on devo max in scotland so that with new labour leader will change lab/snp swing -but by how much.?

  44. @ ROLANDGATINOISE

    We shall see…

  45. 07052015

    And in which direction?

  46. 07052015

    Incidentally “Devo Max” means the maximum amount of devolution. We’ll see tomorrow how far in that direction the Smith Commission has travelled.

  47. @”And in which direction?”

    You are an old tease Oldnat :-)

  48. William hill will give you 16/1 on ukip in donc north.Sounds like they see it as a walkover with labour on 1/100.

  49. Ukip aren’t stupid, they aren’t going to target safe seats where Labour or Cons have a big majority, money will be spent on marginals with favourable demographics.

  50. @Anthony

    I agree that their last national poll doesn’t show the same turnout cutoff, but they are still doing something to downgrade UKIP and the Greens in their national poll

    Table 3 (before turnout and don’t know adjustment) had

    Cons 27
    Lab 32
    LD 7
    UKIP 23
    Green 6
    SNP 4

    And their headline has
    Cons 30
    Lab 34
    LD 7
    UKIP 19
    Green 3
    SNP 5

    So the impact is the same – UKIP reduced by 5, Greens cut in half.

1 2 3