Sunday polls

I’ve been caught up with various family commitments this weekend, so a very brief summary of the polls in the Sunday papers. We have the monthly ComRes poll in the Independent on Sunday, the fortnightly Opinium poll for the Observer, two YouGov polls (one in the Sun on Sunday, one in the Sunday Times) and a Panelbase Scottish poll in the Sunday Times.

Opinium in the Observer have topline figures of CON 28%(-2), LAB 33%(nc), LDEM 7%(-1), UKIP 20%(+3), GRN 6%(+2) – a return to decent Labour lead after their poll a fortnight ago had shown things tightening up.

ComRes’s monthly online poll for the Independent on Sunday and Sunday Mirror has no such movement, with the race remaining very tight. Their topline voting intentions show virtually no change from last month’s, with topline figures of CON 33%(nc), LAB 34%(nc), LDEM 7%(-1), UKIP 18%(nc), GRN 3%(+1)

YouGov in the Sunday Times also show a one point Labour lead with topline figures of CON 31%, LAB 32%, LDEM 7%, UKIP 18%, GRN 7%. There is a second (and completely separate) YouGov poll in the Sun on Sunday but with fairly similar topline figures, CON 31% and LAB 33%.

Finally the Panelbase Scottish poll in the Sunday Times has topline Westminster voting intentions of CON 14%(-1), LAB 31%(+3), LDEM 3%(nc), SNP 41%(-4), UKIP 7%(nc). The SNP lead of ten points would still be pretty good for them by historical standards, but it’s a drop compared to the very large leads they’ve been showing in other Scottish polls since October, which have varied between 16 and 29 points. As ever, it is only one poll – it may be the first sign of that SNP lead narrowing a bit, or may just be random sample variation.

115 Responses to “Sunday polls”

1 2 3
  1. @ Anthony

    Is the huge difference in UKIP VI just a coincidence? No split methodology or anything?

  2. A lot of high UKIP scores here. When was the last time UKIP got a 20?
    Also, are there any methodological why ComRes have the Greens so low?

  3. The last Panelbase poll (for Wings[1]) was at the end of October and had Labour on 28% which is the highest we’ve seen till now. (SNP 45%, Lab 28%, Con 15%, Lib Dem 3%, UKIP 7%, Green 1%, Others 1%). So may be that Panelbase is quite good for Labour at the moment. Not that that means they’re wrong, though Amber’s now going to have to stop all that complaining about their methodology :D .

    It does imply a small movement SNP to Lab, though it’s small enough to be MoE. But certainly there’s some scope for Labour to improve its position in that No voters were more likely to be DKs in recent polls and so could go for them (or the SNP on a ‘best for Scotland’ basis, or not vote at all). Despite that the polls have actually been remarkably consistent since the end of October and we need to see more than one before we can be certain if there is any movement.

    Usually the tables for the Panelbase-ST don’t appear till Monday morning, though Curtice may have seen them, so we may get some detailed commentary before that.

    [1] This explains why the latest social attitudes Wings poll didn’t include any VI questions. It actually has the same sample size and fieldwork dates so was presumably done as the same survey. Is this the start of a whole new synergy between the media empires of the Rev and the Rupe?

  4. I suppose I’d better put up the links again.

    ComRes’s commentary is here:

    with link to the tables.

    They make great play with something they call the ‘Favourability Index’ which they ask about occasionally. I’m not really convinced about its usefulness, because 20-30% of people say ‘neither’, suggesting that a lot of people don’t really see it as a valid sort of question. Some of this will be a variant of Don’t Know (5-8% say DK in addition to that) as some things such as the Greens get a higher figure, but it looks like 20% are being refusnik, which might not make it as useful a measure as ComRes would like.

  5. Opinium tables are here:

    They also asked a number of questions about who should take part in the Party leader debates and their format:

    The leader of the Conservative Party 74%

    The leader of the Labour Party 75%

    The leader of the Liberal Democrats 66%

    The leader of the UK Independence Party (UKIP) 61%

    The leader of the Scottish National Party (SNP) 30%

    The leader of the Plaid Cymru 23%

    The leader of the Green Party 46%

    The leader of the British National Party (BNP) 18%

    Other 1%

    Don’t know 13%

    Support for the proposed 4-3-2 format was lukewarm at 37-31 with more strongly opposed than in favour. If Bennett isn’t invited people think Cameron should take part anyway by 52% to 25% – similar figures to YouGov and again not a partisan matter Tory voters were 52-29.

    If Cameron does refuse then people think the debates should still go ahead by 57% to 25% (Con 48% to 38%). And voters also say that in those circumstances he “is looking for excuses not to take part in the debates” by 51% to 26%. Even 32% of his own voters agree. And asked as a general principle there was overwhelming support that the debates should go ahead by 79% to 11% with a similar number believing that they will (77-10).

    They also asked how well people thought the various leaders would do:

    Cameron: Well 42%, Badly 25%

    Miliband: Well 27%, Badly 37%

    Clegg: Well 22%, Badly 41%

    Farage: Well 49%, Badly 21%

    Bennett: Well 33%, Badly 17%

  6. Lab are not going to feel too downhearted about the weekend polls, after a week of events which would normally benefit incumbents. The fillip in Scotland is a bonus but Labour need another 5% to draw level with SNP – not impossible?
    Sometimes it feels like a perfect storm for Ed with inflation at record lows, employment (admittedly including too many p/t,insecure and low-paid jobs) at a record high, fuel discounts fortuitously putting spare money in most pockets, grim NHS publicity in Wales to balance the mess in England, and a partial Labour meltdown in Scotland , not to mention a shadow Cabinet of shall we say uneven talent and loyalty — yet there he is ahead of them all in every poll. No one in the media gives him the credit he deserves for this – but it’s just beginning to seem possible that enough of the electorate will.

  7. Seems that Yougov for the Sun is very different from the Sunday Times Yougov on the UKIP figure

    Lord Ashcroft [email protected]
    YouGov/Sun on Sunday poll LAB 33% CON 31% UKIP 13% LDEM 7% GRNS 7%

    13% or 18%. Same methodology. Same timeframe. Very different result. 2 extremes of MOE? Will be interesting to compare the tables.

  8. It is possible that Labour’s resurgence in Scotland is down to Murphy’s honeymoon; or it may be a general ‘swing back’ to the status quo – though a modified version.

    What I find a little disconcerting is that the Labour VIs in Scotland seem to vary in inverse proportion to the UNS predictions for OMs at Westminster! When the SNP seemed to be getting higher VIs, Labour were being shown with an OM. Now the SNP VIs are down and the Labour OM has disappeared. Of course it’s nothing like as simple as such a superficial glance would suggest, but maybe Milliband would prefer a lower Scottish Labour VI and a Westminster majority……… if only to get some of the more dinosaurian tendencies off his back…… !

  9. We might also expect the drop in the oil price to have an effect on how people perceive Scotland’s ability to pay its own way, whether than be within or outwith the Union.

  10. I’ve solved the two YG polls released last night I think.

    The Sun on Sunday one, the fieldwork was done on the 14th-15th, a day prior to Sunday Times one.

    Also, the UKIP @ 13% is possibly a typo – according to the tabs UKIP polled 16%, the same as the other YG for that day.

  11. Here is YG this week vs last week, by 2010 voter ID:

  12. Just going back to the Panelbase VIs, those figures (on the Scotland Votes site) would leave Dumfries and Galloway (geographically the largest Labour seat in GB by far) with its incumbent, whilst many Central Belt urban seats would go SNP.

    Dumfries and Galloway is one of the three way marginals. Were the Tories to reach 16% (nationally) they would replace Labour. But if Labour drop again to below 30 and the SNP gets above 40 then it could go SNP.

    More detailed polling will be needed in the run-up to the UK GE if we are to have any idea what to expect in Scotland.

  13. Good Morning All.


    Many thanks for very interesting links, early today.

  14. CMJ – thanks for the Cloud link. I found the Scottish responses to the supplementary questions interesting, especially the huge differences regarding the SNP as ‘king makers’ and the importance of further constitutional change.

    Also, slightly worrying that questions were asked north of the border on education and health, neither of which will be the least bit influenced by the result of the GE. This may be just ignorance on the part of the pollsters. On the other hand, if might be that the Scots are genuinely concerned for the wellbeing of those who live south of the border…

  15. Yougov scottish crossbreak -snp 33,labour 31,tories over 20

  16. 07052015

    Any link for that?

  17. Yes

    Only one poll of course but also six point improvement in EM approval ,tho he has a long way to go.More passion in his speeches means he is connecting a bit better.

  18. Electronic voting next time ?

  19. Political parties won’t like it, it makes telling difficult. I think it could be a useful way of driving up turnout slightly (although again I think that has to do with messaging and the parties rather than the public), but there’s something authentic and proper about walking down to the polling station and crossing a box.

  20. @07052015

    I hope we get eVoting.

    Given modern security systems, it would be as safe, if not safer than other methods.

    I really hope it gets a trial.

  21. Political parties won’t like it, it makes telling difficult

    Is that a good reason to do it?

    Our system should be made to suit people, not parties. Don’t let the tail wag the dog!

  22. Today’s YG Scottish crossbreak

    SNP 33% : Lab 31% : Con 21% : LD 5% : UKIP 5% : Grn 4%
    Mean of last 10 YG Scottish crossbreaks (not including Sun on Sunday)

    SNP 42% : Lab 26% : Con 17% : LD 5% : UKIP 6% : Grn 4%

  23. Whether the movement in Scotland is genuine or MoE, the sort of percentages and seats in this poll look close to what my guess is of the final Scottish outcome: namely a narrow lead in both seats and votes for SNP with some wide regional variations.

    I wonder if some of the change is not so much swing back to Labour as No voters starting to firm up their voting intentions and moving out of the previous high Don’t Know score they were showing? Will be interesting to see the tables.

  24. CMJ,

    That’s not my view or a reason not to do it – just an observation!


    If anyone wants a chuckle you can have a look at Ian Dale’s loony tunes predictions here:

    The man thinks the Tories will gain Hampstead and Kilburn, while Labour lose Heywood and Middleton to UKIP, but not Grimsby. At least he doesn’t have the Lib Dems gaining Sheffield Central, as his regional breakdown hinted at.

  25. Sorry, previous post should have said CON 33% LAB 33%!

    Silly me, a 35% Lab vote would result in a small majority for LAB.
    Con would need a good 40%+ for a majority!

  26. I retain my confidence that Labour will outpoll the SNP in May – though it remains to be seen whether Murphy’s prediction of not losing a single seat in Scotland comes true.

  27. Graham

    “I retain my confidence that Labour will outpoll the SNP in May”

    Mebbes Aye, Mebbes Naw (as in fitba, not referendum!)

    I retain my confidence that whichever comes out on top will depend on what proportions of the SNP/Lab floaters choose to see the election through a UK or a Scottish prism.

  28. GRAHAM

    Good Morning to you, I also think Labour are stronger than many people think, but ‘the party’ has had a big shock.

    Mr Murphy is a strong campaigner I think, and seems to have survived leftist and also sectarian jibes.

  29. 81% of patients are seen within 4 hours in Wales NHS and 90% are seen in 4 hours in England but Labour says the crisis is in the English NHS and never refers to the crisis in the Welsh NHS?

    A very odd state of affairs.

  30. @Robert Newark.

    Hardly odd I’d say, very predictable! Interesting polling on the NHS recently suggests that the public are starting to get much more sophisticated in their analysis of what’s going wrong & that blaming one party or another is missing a lot of the point

  31. There has also been the release of an ICM poll for British Future – rather old, as the fieldwork was 28-30 Nov!

  32. @ Mr Nameless

    OK, ID’s predictions are not that plausible, but let’s put you on the spot.

    You are woken early tomorrow morning by an urgent phone call. Turns out it’s Ed and, amazingly, he’s asking you to take over the job of masterminding your party’s national campaign. (No slur on Douglas: just fantasy politics.)

    Being an ambitious type you accept without hesitation.

    An early task is to establish plans for distributing your campaigning resources, and in particular to set up a list of constituencies to send your troops. Let,s suppose that the target list (seats to GAIN) presents no great problem, then here’s my question for you:

    In which seats do you deploy your defensive resources?

    (Bear in mind that funds are tight and big-name constituency visits have to be carefully rationed. Any resources you move from attack reduces your chance of making some of the more marginal gains. But nor do you want to lose positions behind the front line.)

  33. PK has a article in ST today :-

    “May prediction”

    Swing back in last four months on precedent.

    Incumbency bonus =12 seats for Cons.& 10 seats for LD.

    UKIP + 4 seats-losing Reckless.

    SNP to fade -gaining 17 seats to 23.

    Seats overall:-
    C 293/L 277/UKIP 5/ LD 30/ SNP 23

  34. The Guardian apparently doesn’t know how to count. Check out Farages approval ratings.

    if it adds up by the time you read this, it means it’s been fixed

  35. “Is the huge difference in UKIP VI just a coincidence? ”

    The political class have balkanized the conurbations. You can have two areas separated by a main road with exactly the same SES and the Ukip score could be 25% in one and 5% in the other.

    If pollsters want predictive polls rather than guess-your-weighting then they need to include ethnicity in polling.

  36. Just about to venture out into a glorious Worcestershire winter morning for a little seven miler through the frost-bound lanes (much better than anything Bournemouth can offer, Chris L!), but I thought I’d share some thoughts on the Cleggster. He was on Marr this morning and, as much as I think he’s the sort of bland and technocratic sort of politician who gives mainstream politics a very bad name, I thought he acquitted himself quite well.

    He has the mother and father of all atrocious hands to play in May, but he might well play it better than some think. For example, he seems to be discovering, albeit very late in the day it has to be said, a visceral dislike of Toryism again. How genuine this is, of course, is for the voters to decide in May.

  37. Times Red Box:

    In an exclusive YouGov poll for Red Box, we asked: “How do you think you will actually cast your vote?”

    71 per cent said it would be “for the party or candidate that I most want to win”
    17 per cent said “for a party or candidate that is not my first choice, but who can stop a party I don’t like from winning”

    We also asked: “Which of the following is most important to you when deciding who to vote for in a general election?”

    Top of the list was “values and priorities”, 44 per cent
    “Specific policies”, 21 per cent
    “Best PM”, 9 per cent
    “Local candidates”, 7 per cent
    “Local issues”, 6 per cent
    “Don’t know”, 13 per cent

    The Conservatives are banking on the “best PM” factor so will will hope that 9 per cent is a huge under-estimate, as they know that on values they are a long way behind, but that on perceived personal qualities, Cameron beats Miliband just as strongly.

  38. Peter Kelner predictions are baffling – he sees the tories losing only 2% vote share on 2010 – yet UKIP getting 13%? On his figures that would mean that UKIP would be getting their +8% increase on 2010 mostly at the expense of the lib dems – rather than the tories.

    I think he is applying pre-2010 election behaviour to the very very different scenario we have now.

  39. Unicorn,

    You’re on. After giving the Dear Leader a bollocking for waiting until three months before the GE to select defensive marginals, I would pick the following (limiting myself to ten) in no particular order:

    Falkirk (retiring disgraced incumbent, selection mishaps, good SNP second in 2010)

    Glenrothes (SNP held at Holyrood, retiring incumbent)

    Stirling (Ditto)

    Great Grimsby (1% Lab lead in Ashcroft poll, retiring incumbent)

    Dudley North (Lab and UKIP tied in Ashcroft)

    Southampton Itchen (Cons doing well there for some reason)

    Penistone and Stocksbridge (Three UKIP seats on Sheffield City Council, albeit two of them won on a split left-wing vote share, they topped the poll in the locals)

    Rotherham (It’s an obvious UKIP target, albeit under less risk than people thought before the PCC by-election)

    Rother Valley (Ditto)

    Inverclyde (by-election in 2011 showed a big swing to the SNP, further swing since could make it vulnerable)

    There are probably more in Scotland but I wanted a decent geographic spread and I don’t know Scotland that well.

  40. @Reggieside.

    You are assuming direct movement between parties rather than churn.

    Lib Dems were the protest party, now it is UKIP.

  41. @Mrnameless

    Glenrothes as a defensive marginal?

    That sounds like trouble for Labour if Glenrothes is considered a marginal.

  42. @Bramley.

    One must be suspicious of such polls.

    The electorate always believe they are more sophisticated and discerning than they actually are.

    One of the polls today shows that the majority believe they understand what the ‘deficit’ means while only a minority do.

  43. …ie gut reactions on being ‘prime ministerial’ are understated and ‘policies’ overstated

  44. Reggieside/Colin,

    At the start of the GE campaign in 2010 (so 4-5 weeks out) various polling company priniciples were invited to forecast the result.

    All (except one who eged his bets IIRC) forecast a modest Con OM.

    I have tremendous respect for PK and he offers a great insight; 2-3 months ago his piece suggesting that recent developments in Scotland plus incumbency may wipe out Labour’s FPTP post advantage so for example the same Lab and Con vote share would also be very close in seats was I thought particularly good.

    What he and other pollsters (and probably most of us on here who enjoy politics and/or make a living from it)) are bad at imo is predicting how voters will react to specific events and move their votes (after churn) and as a result fall in to the ‘historical’ trap a little.

    Bit like weather forecasters saying that the last time we had weather like this 2 weeks later this happened so it will again when other meteriological variables are in play.

    As you suggest his implied premis is that UKIP wil take most of it’s new votes from 2010 LDs (and 2010 DNVs) which is probably true as much of this is a ‘protest/NOTA’ votes but like you I think 35/31/13/10/6 unlikely.

    I hope (no polling evidence) that he is giving insufficient weight to a Green squeeze to Lab in CON/Lab marginals negating some incumbeny but we shall see.

  45. Kelner seems to be assuming a degree of “swingback” from UKIP to tory – but not allowing similar swingback from green, SNP and UKIP to labour.
    Basically if hes saying that a chunk of UKIP voters will ‘hold their noses” and vote tory to keep labour out – then surely the same will apply for the anti-tory vote currently expressing support for green etc?

  46. Reggie – agree hence my last sentence.

    No polling evidence for either of course.

  47. @ Roger Mexico

    So may be that Panelbase is quite good for Labour at the moment. Not that that means they’re wrong, though Amber’s now going to have to stop all that complaining about their methodology :D .

    I didn’t complain about their methodology; I complained about the composition of their panel. i.e. they didn’t have enough No-leaning responders, which Panelbase pretty much admitted to themselves.

    I think that all Scottish panels are a little bit suspect now. e.g. YG was a bit light on No-leaning responders too.

  48. Re swingback:

    Midterm all governments are widely detested so people will instinctively claim opposition support, whoever they are.

    Only nearer the actual vote, will the election become crystallised in people’s mind as a CHOICE between parties.

    That is why swing back generally favours the government – they are always more hated than the opposition.

    In the 80s and 90s Labour was getting poll ratings of near 60%.

    Not only was Neil Kinnock never going to get 60%, NO party has got anywhere near that figure.

  49. …ie swing back occurs because mid term opposition support is over inflated.

1 2 3