Lord Ashcroft has released a new batch of Scottish constituency polling. Full details are here. As regular readers will know, national polls in Scotland suggest an SNP landslide with no sign of the SNP lead narrowing as the election approaches. In fact the most recent Scottish polls from YouGov and TNS showed the SNP lead growing. Lord Ashcroft has previously carried out two waves of Scottish polling, both showing the SNP winning in seats that were previously thought very safe.

Across the two previous waves Ashcroft polled nineteen Labour seats, finding the SNP ahead in all except East Renfrewshire and Glasgow North East. He has also polled four Lib Dem held seats (finding solid SNP leads in three and a narrow SNP lead in Ross, Skye and Lochaber) and the one Tory seat in Scotland where he found a tie.

In this latest wave Ashcroft has returned to the five of the tighter races he polled earlier, and polled three more Lib Dem held Scottish seats for the first time. Most of the close races are no longer close – in Glasgow South West (where Labour have a 46% majority) Ashcroft found the SNP 21% ahead. In Paisley and Renfrewshire South, Douglas Alexander’s seat, he found an SNP lead of 11%. In East Renfewshire he found a 9% SNP lead – this is the seat of Labour’s Scottish leader Jim Murphy and one of the two seats where he previously found a Labour lead.

People commenting on the Scottish election seem to fall into two groups – those who expect an almost complete SNP walkover, and those who expect them to merely do extremely well, to take “only” thirty or forty seats. I think the logic of most people in that second group is simply down to the scale of the SNP landslide looking unrealistic – the belief that they can’t be doing that well. I am often the first to advise caution towards polls showing huge shifts, so it’s fair to be sceptical – with a landslide of this scale perhaps the polls could be overestimating the swing… but they are consistent across national and constituency polling, online and telephone, and if they are correct the SNP are headed for an overwhelming victory and Scottish Labour towards something close to wipeout.

The polling in the sole Tory seat in Scotland remains extremely close. In his last poll Ashcroft found a dead heat between the Conservatives and the SNP, this time he found an SNP lead of two points.

Turning to the Lib Dem held seats, last time round Ashcroft found only a five point SNP lead in Charlie Kennedy’s seat in Ross, Skye and Lochaber, that has now grown to fifteen points. He has now additionally polled North East Fife (Menzies Campbell’s seat) where he found a 13% SNP lead and East Dunbartonshire (Jo Swinson’s seat) where he found an 11% SNP lead. Finally he polled Berwickshire, Roxburgh and Selkirk – here he found a 20 point boost for the SNP, which is enough to make it into a three way marginal but not to put the SNP ahead – the poll found the Conservatives ahead on 30%, SNP 29%, LDEM 28%.

Note that the East Dunbartonshire is one of the seat where the Lib Dems have recently released their own private polling (which may or may not be a co-incidence!). Their own polling showed figures of CON 13%, LAB 16%, LDEM 35%, SNP 32%. The SNP are 8 points lower than in Ashcroft’s poll, the Lib Dems 6 points higher. I wrote about the Lib Dem’s own polling at more length here. The primary differences are that the Lib Dems prompt using candidate name, which is fair enough, and ask about how favourably people see their local MP before they ask that voting intention question, a more questionable approach. We shall see which is the better guide. The Lib Dems have not released any of their private polling for the other three Lib Dem seats Lord Ashcroft released polls for today, from which one must draw one’s own conclusions.


164 Responses to “More Ashcroft Scottish polling”

1 2 3 4
  1. And the 2010 election delivered a lot of Con gains in urban English constituencies — often ones with a significant Lab-LD swing — that you’d expect to revert to Labour this time, not necessarily on direct Lab-Con swing but from a decline in LD support.

    So not mission impossible at all if you treat a chunk of 2010 LD support being on loan from “natural” Labour voters. The big question is how much of it gets returned.

  2. Hawthorn
    Fair enough on Norwich but I have to say of we count the “toss ups” (which in hindsight I think High Peak and Colne Valley should be included) as well as the six you mention there is also I feel
    Great Yarmouth
    Pudsey
    Peterborough
    Watford
    Swindon South
    Bristol North West

    Also another seat which is tipped to probably go Labour (which considering its importance I cant believe I forgot) is Wirral West.

    all in all it a big ask for Labour to be largest party if they are totally wiped out in Scotland but certainly well within current polling.

  3. Hawthorn
    Bugger I forgot we aren’t allowed to mention the North West, OK scrap Wirral West.

  4. Stardaz
    I don’t think Osborne needs rescuing.

  5. The Midlands are going to be absolutely key because Conservatives aren’t really strong enough in the North or in London (Boris clearly isn’t much of an asset…). It’s surely part of the explanation for that Greater Manchester devolution push, with extra revenue-raising powers and the NHS thing.

    Do we know how well the devolution plans poll in Greater Manchester?

  6. @AW

    It is difficult to dispute the evidence that there will be something close to a clean sweep in Scotland. However, I would query your statement:

    In fact the most recent Scottish polls from YouGov and TNS showed the SNP lead growing.

    However an analysis of the YouGov crossbreaks shows that the SNP lead peaked at about 19-20% in the latter half of January and then drifted downwards to about 12% by the end of March followed followed by a modest rebound a couple of days later and flatlining since then. It depends on exactly how you define ‘recent’ but their lead so far in April is lower than it was for the two month period before mid-March.

    Not that it makes a blind bit of difference whether the lead is 19-20 (then) or about 14-15 on a rolling average (now). It is just that moving downwards in this way is a funy way of ‘growing’.

  7. MAD 25-poll time-weighted

    http://www.statgeek.co.uk/2014/09/here/ (set of ten)

    Basically Lab a touch in front in GB and a smidgen behind in England & Wales.

    Scottish data with UNS suggests seats of:

    SNP 42
    Lab 13
    Lib 2
    Con 2

    Ashcroft’s polling pretty much knocks UNS on its backside, I think.

  8. People saying CON will gain 20 LD seats?!

    Have they been looking at the Ashcroft Polling?

    *Chippenham
    Cornwall N
    *Devon N
    *Mid Dorset & Poole
    *St Austell & Newquay
    St Ives
    *Somerton & Froome
    *Taunton Deane
    Torbay
    Wells
    Portsmouth S
    Berwick Upon Tweed
    Berwickshire
    *Solihull
    Cheadle

    thats 15, the absolute MAX that the Tories will take off the LDs and of those the ones with * are the only ones you can say with any conviction..

  9. Ashman
    Your Tory loss figure seems pretty wishful to me. I’ve thought for a while that a good result for the Tories would be if they lose only about 20 seats to Labour, 10 just seems fanciful. I’ve personally got 19 Tory held seats in my “labour almost certain to gain” category and current polling suggests Labour winning an awful lot more.

  10. @CMJ

    No, I tried getting page two earlier. No show.

    I right clicked and view source, and if you are patient, you can find the relevant poll links in there.

  11. @Unicorn

    You do know that YouGov have changed there method and are re-polling people they polled in January so the cross breaks are not comparable to earlier ones. I would expected the xbreaks to look more or less identical to January xbreaks if no change.

    Old Nat has a spreadsheet but from memory the current xbreaks look similar.

    The dip you detected I contest that was a real dip but rather YouGov fiddling around with their methodology, there was massive down weighting of SNP identifiers at that time,

  12. @ rivers 10

    I would have followed the trend of conservative losing at least 20. Speaking to one of the people involved, giving him a hard time on the lack lustre performance, in my opi

  13. @Unicorn

    That was ‘NOT’ a real dip.

  14. @CMJ et al

    A css-free version of page two.

    http://www.statgeek.co.uk/wp-content/uploads/2015/04/yougov.html

  15. Survation Daily Mirror

    CON 34% (+4); LAB 33% (-2); UKIP 17% (+1); LD 7% (-1); SNP 4% (NC); GRE 3% (-1); AP 1% (NC)

  16. @ rivers 10.

    Sorry, typing in the train. Accidently pressed submit

    One seat that would be on the top of your hit list. Was expected to fall towards Con to Lab in dec. They are now confident of retaing the seat. I still have that seat down as Lab gain. On the ground different story.

  17. Survation:

    Con 34
    Lab 33
    Ukip 17
    Libdem 7

  18. John Smith
    Agree 100% and even that list seems to be pushing it, for example Liverpool University did a seat prediction for the North West (I study politics there and spoke to one of my lecturers and also the main author of the prediction) and he told me that their prediction was based on some heavy swingback to the Tories but importantly not to the Lib Dems. Consequently their prediction saw Labour only gaining two seats from the Tories in the region and yet they still predict Cheadle will remain in the Lib Dem camp.

  19. @ Colin

    I second that. Lol.

    Queen Sturgeon of Scotland.

  20. Comment on Survation poll:

    More polldrums. Another poll that confirms the neck-and-neck status of this election contest.

  21. Interesting throwaway comment in the May 2015 article about polling in Scotland:

    “Pollsters are running internal tests to try and see if they’re overstating the SNP share, but none can see any evidence they are.”

    http://may2015.com/featured/election-2015-stunning-ashcroft-polls-show-the-snp-could-win-every-seat-in-scotland/

  22. Couper

    I wouldn’t worry about JR.

    She is as well capable of sorting out “assholes” (as I understand they are called in her country) as her namesake in the Ewings (US TV, not Scottish political, family), :-)

    Still, the multiplicity of comments in this thread from our love bombing friends in southern climes, does demonstrate a remarkable lack of understanding of how their country – the UK – is actually governed.

    English schools need more Chris Lanes!

  23. Survation looks like an outlier, you would expect UKIP to go down with Conservative VI increasing by that much.

  24. Ashman
    Out of interest which seat are you referring to?

  25. @ProfHoward

    Yes, I think you were respectful and your comment was totally valid.

  26. Re-posting the poll because NC has different changes to what was posted above

    CON 34 (+3)
    LAB 33 (-2)
    LIB 7 (-2)
    UKIP 17 (+2)
    GRN 3 (-1)

    This looks correct after a quick comparison with the older poll. +3 for Tories and +2 for UKIP seems odd.

  27. Survation Scotland (small sample) – https://twitter.com/StatgeekUK/status/589130391547699200

  28. ASHMAN

    I don’t think she is having to add up lists of gains :-)

  29. BIGD

    I said nothing of the sort and, whatever I may think on the subject, you’re not going to hear it on UKPR. Argue about the polls if you like but leave your personal politics to the side.

    Please :-)

  30. If anyone fancies a bet on UKIP in Dagenham & Rainham, you can get 8-1 l believe. Labour are 14-1 on. Labour are much too strong in the 5 Dagenham wards to lose the seat – there are 3 other wards in Havering.
    Rivers 10 – while l am expecting a pretty large crop of Labour gains, l feel that expecting my party to win in Watford and Bristol NW is fanciful, and the others unlikely too, except for Pudsey & Swindon S which appear to be toss-ups.

  31. @Statgeek

    Thanks for the new page :)

  32. All joking apart ( though………who’s joking?)-AN on DP this morning was contrasting the time it took for the Conservative Party to implode in Scotland , with this truly extraordinary destruction of a Labour heartland of big majorities.

    The awful guillotine that is FPTP at it’s tipping point.

  33. @statgeek
    Survation Scotland – what are the odds of the tories out-polling labour in scotland but losing in England :)

  34. LAGAVULIN

    Fair enough. You are right I should limit my comments to polling and not get so worked up!

  35. General comment:

    Some of the recent posts here seem (to me) to violate comments policy. Please remember the policy before posting.

  36. Woah I think AW will soon be getting his scissors out.

  37. RICHARD
    @Catmanjeff
    I get the same as you, Windows 8 in both IE and Chrome

    Snap in both Win 8.1 [64-bit] and Win 7[32-bit] in Chrome.

  38. Crossbreaks:

    I’ve started analysing Yougov regional crossbreaks to look for patterns. I’ve taken the mean of the last seven polls.
    So far all I’ve done is calculate the Con-Lab swing in each region. Next, I want to put that into my constituency database and then compare it to constituency polls to see if they make any sense.

    I could well end up proving this site’s golden rule:
    Don’t look at the cross breaks!

    Anyway, Con to Lab swing:
    London…….. +2.7
    RoS…………..+9.1
    Mid/Wales….+3.4
    North…………+3.8
    Scotland…….-8.0

  39. PROFHOWARD

    When you say “recently”-what sort of timescale did you have in mind?

    Perhaps you have missed some previous threads ?

  40. @ John Smith

    Knowing something about Cheadle’s demographics (plus the political history – tactical voting), I would be very surprised if the LiBDems lost it.

  41. VALERIE
    Woah I think AW will soon be getting his scissors out.

    I think he’ll need a scythe.

  42. Barnaby
    I actually agree, I might have went a bit over the top on those two :)
    As for the others I personally don’t think Labour will gain Peterborough or Great Yarmouth but in the case of the former I’ve heard interesting rumours about demographic changes and the latter interesting rumours about UKIP doing disproportionate damage to the Tories, must be emphasised though nothing other than rumours.

  43. Its unusual for the Tories to be ahead in a Survation poll. When was the fieldwork done for this poll?

  44. rivers10
    he knows my pseudo, I promised I wont. mention.
    Lab thinks, its in the bag.
    They are scaring Middle England, to the nth degree.
    It was one of those Con then Lab and then Con seat.

  45. I am a Labour party activist and local government candidate in the north east of England. So I have a partisan interest in this election. But I am also a student of politics and a lover of Scotland. And regardless of what it may or may not do to the overall GE result you have to stand back and goggle a little at the revolution that appears to be occurring north of the border.

    Several of my No-voting Scottish friends are in absolute disbelief at the polls, saying that no voters aren’t going to defect en-masse to the yes party. But apparently they are – it is truly remarkable even if we do find that the SNP position is somewhat over-stated.

    As someone who believes the ultimate settlement in these isles is a Federal UK, I think we are seeing the dying days of the current incarnation of the United Kingdom. Regardless of who is Prime Minister.

  46. I had a look at Ashcroft’s poll of my neighbouring constituency – Paisley & Renfrewshire South.

    The SVI/CVI VI shifts are interesting, as always.

    I use the figures prior to the manual adjustment for DK/Refusers are applied, as they have more detail, and transparency.

    Party : SVI : CVI : Change
    SNP : 52% : 52% : 96% SNP, 4% Lab, 1% Other
    Lab : 36% : 38% : 96% Lab, 3% SNP, 1% Con
    Con : 7% : 5% : 71% Con, 19% SNP, 8% Lab, 2% LD
    LD :1% : 1% : 93% LD, 7% Grn
    UKIP : 2% : 2% : 77% UKIP, 19% SNP, 1% Other
    Grn : 2% : 1% : 40% Grn, 30% SNP, 17% UKIP, 1% Other

  47. @Colin

    While traumatic for those on the sharp end, the massacre of Labour and the Conservatives in Scotland gives hope that we don’t live in a particularly static environment.

    That gives hope for democratic change and evolution.

  48. @cloudspotter

    That Rest of South swing… wow.

1 2 3 4