The YouGov/Sunday Times poll this morning showed a four point Labour lead, interpreted in some quarters of the commentariat as showing an advance for Labour after the Paxman interviews. As ever, it was only one poll. Now we have a second post-Paxman poll, a ComRes telephone poll for the Daily Mail & ITV, and this one shows the complete opposite – CON 36%(+1), LAB 32%(-3), LDEM 9%(+1), UKIP 12%(+2), GRN 5%(-2) (tabs).

If the four point Labour lead in this morning’s YouGov poll equalled their best this year, this poll is the biggest Conservative lead ComRes have delivered since 2010. Where YouGov showed Miliband’s ratings improving, ComRes show Cameron widening his lead as best Prime Minister.

There is no great mystery here, I expect we’re just seeing normal sample variation. I said this morning we needed to wait for some more polling to have any idea whether the Paxman interviews had really had any effect, whether there was a consistent trend. With two polls now showing movement in opposition directions there certainly isn’t yet. It could still be that the rest of the week’s polls show a similar movement to YouGov and the ComRes poll was just a blip… or that the rest of the week’s polls show a similar movement to ComRes and the YouGov poll was a blip. I’ve a sneaky suspicion though that we’ve just happened to see two outliers in opposite directions, and we’re going to see lots of polls showing no clear movement. Time will tell.

442 Responses to “ComRes/Mail/ITV – CON 36, LAB 32, LD 9, UKIP 12, GRN 5”

1 6 7 8 9
  1. RAF.
    Thanks for that.

    If Labour can get the SNP to 30 seats then Ed M will be in Downing Street, I think, as PM

  2. James Peel

    Former labour person,

    “Assuming the SNP get 40 seats”
    I think you are assuming a lot here James.
    A lot of people in Scotland are sick tired of hearing theSNP/Salmond/ Stugeon
    continually banging on about breaking up the UK.
    You may be more accurate if you assume an SNP total of 30 seats max.

    This is a big assumption but it’s based on current constituency betting and projections….

    If the SNP underperform, and get say 30 seats, labour will need fewer seats to be the largest party.

    If labour hold 25 seats in Scotland, which SNP at only 20 implies, it’s obviously easier for them to get to be the largest party.

    I am sticking with 40 as I am just using information from projections and markets, but you’re perfectly right that labour could defy these expectations by holding more seats in scotland than anyone expects

  3. @former labour person

    The trouble is the unionist vote is split. Is there much talk of tactical voting to deny the nationalists?

  4. chrislane1945

    You’re absolutely right. If the SNP only get 30 seats, it is highly likely that Ed M gets into number 10. scottish seats are a zero sum game between labour and the snp….

    labour anecdotally aren’t even bothering to put up a fight in seats where the snp are challenging the liberals.

  5. I still remain puzzled as to why there is this incessant concentration on GB VI percentages by the pollsters.

    They were a useful tool, when combined with UNS, in those far off days when there were two dominant GB parties and opinion moved in much the same way across GB.

    Such a quaint 20th century way of thinking! :-)

    There is no GB UNS that would make any comparison of current and previous percentages meaningful.

    Even within Scotland (which needs to be looked at separately) while there is a national swing in the same direction, the Ashcroft and Curtice analyses show that it certainly isn’t uniform in any meaningful sense for seat projection.

    Whether in England, or England & Wales, there is a useful swing in the same direction at a reasonably uniform rate, I don’t know.

    My guess would be that if there is not, then the different swings will not operate conveniently within geographic regions, and a new, more sophisticated, way of modelling voter behaviour is required for future elections.

  6. @oldnat

    I guess the people who pay the pollsters, i.e. the press, want a simple way of communicating what’s going on for their headlines and such. What we need is more Lord Ashcrofts to spend insane amounts of money on more constituency polling. But that’s very expensive.

  7. PAUL AD

    P.Cymru are a minor party even in Wales, never mind the rest of the UK and their 3 MPs (could fall to 2 at GE2015) could not hold the balance of power at Westminster even if they wanted to as the only UK party they could possibly support would be Labour.

    Per Ofcom: At present in Great Britain, major parties are defined as: the Conservative Party; the Labour Party; and the Liberal Democrats.
    In addition, major parties in Scotland and Wales respectively are the Scottish National Party and Plaid Cyrmu.

    See this short [2 pages] PDF dated 16 March 2015, and note that UKIP are added only for England & Wales.

    The DUP’s Nigel Dodds has far more right to be there than her.

    The PDF also says: The major parties in Northern Ireland are: the Alliance Party; the Democratic Unionist Party; Sinn Fein; the Social Democratic and Labour Party; and the Ulster Unionist Party.

    The point is that the key words are Great Britain. None of the NI parties stand in GB and none of the GB parties are considered major in NI.

    Apologies if someone else provided an answer, but I didn’t spot one.

  8. Omnishambles

    My guess is that it’s a two way thing.

    If pollsters offered a better system, the press would probably buy it – but if the press are mug enough to carry on paying them for useless numbers, why should the pollsters bother?

  9. Oldnat

    It’s possible that if the 2015b or 2020 elections saw only small swings then UNS would be a “good enough” model.

    The reason why they where relatively successful was due to the fact that the marginal seats did actually roughly follow UNS.

    UNS is wrong now due to the sheer size of the swings, coupled with a huge variation geographically of the size of the swings.

    While VI itself? Is less important, movements in VI are still the most important factor, if you want to reduce the state of play to a simple model.

    If we could agree where we stand now, I’d suggest a 1% swing could be applied as a UNS on top of the “now” position and a relatively accurate picture would be generated. A 10% swing and we’d need to recalculate a baseline position.

    F=Ma is a useful model most of the time, we don’t throw it away just because better models exist.

    Thanks for that; I think Labour will pick up ten-or-so Lib Dem seats in England and about thirty five to forty Tory seats, which takes Labour to 302 to 307 Seats.

    All this before the SNPvLabour ‘zero-sum game’ fight.

  11. There’s been some discussion about the implausibility or otherwise of voters moving from Lib Dem to UKIP, but this is nothing new. For example the latest ST YouGov shows 11% 2010 Lib Dem to UKIP – 9% of all 2010 Lib Dems and 2% of all 2010 voters. Spearmint is the expert here, but these seem fairly standard figures to me.

    However the movement to UKIP may be even stronger in Wales. Unlike in England it may be that they are attracting most of their vote from non-Tory sources, in part because there were fewer Conservatives to start with. Getting information on changes to 2010 vote is difficult however[1] and the only recent one I could find is from January: [2]

    As with all telephone polls this shows a small effective sample of around 450 once you take Will Not Vote (11%), Refused (10%) and Don’t know (25%) into account. Even before LTV is taken into it m,akes percentages only roughly indicative at best. Still looking at the 89 who said they voted Lib Dem in 2010[3] a whole 33% go to WNV/Ref/DK 22% to Labour, 14% each Lib Dem and UKIP only 2% Conservative. This suggests that the Lib Dems have even bigger retention problems in Wales than elsewhere and that UKIP are more attractive there than the Tories – something noted more generally.

    Perhaps as in Scotland there is an anti-Establishment vote that will never pick the Tories, but does not find Plaid an alternative (because of language associations) in the way the SNP is. With Labour the local establishment UKIP may be the obvious choice.

    Looking at the source of UKIP’s votes in the same poll (again a small sample) suggests only a quarter of them say they voted Conservative, less than you would find in a GB poll. So again Wales may be subtly different.

    [1] YouGov’s tables for the ITV/Cardiff University poll monitor don’t show the breakdown in their cross-tabs, presumably for space reasons. The most recent ICM-BBC poll doesn’t give voting intention (though bizarrely it does show likelihood to vote), presumably because the BBC isn’t supposed to commission VI polls during the election period.

    [2] I assume this is a telephone poll, based on the methodology for the (VI missing) Feb/Mar poll, but the January tables aren’t on ICM’s own website and Scully only has the actual tables without preamble.

    [3] I’m using pre-LTV figures as I’m looking at individual choice rather than forecasting. As has been noted current Lib Dems generally also have lower LTV as well than Lab/con/UKIP.

  12. @oldnat

    I think the problem is that the press can’t afford to pay the pollsters to do more sophisticated work on a regular basis. According to the article linked below, marginal polling (necessary for better polling) is very expensive – around £10,000 per poll. They estimate Lord A has spent at least £750,000 “so far” = October last year.

    One way around the problem is for the press to publish/produce models that take into account UNS, marginals, regional differences etc – some are starting to do this as we’ve seen. Problem is you get seats, not VI, with various inbuilt assumptions that may be flawed. But it’s an improvement. I don’t see how the press can afford to pay pollsters for extra regional detail, and how many readers will it steal from the paper that continues publishing vanilla UNS? Not many.


  13. @Rich,

    I’ve said before that I think a Labour government supported by SNP C&S is pretty much the nailed-on result of the election. There is a small but real chance that the Tories and LDs combined could still scrape across the line, but a) I’m not sure the LDs would want to participate in another DC government, especially a horribly fragile one and b) even if they did, it probably wouldn’t last very long.

    The more interesting question for me is, if Labour and the nationalists were in a position to rule the country without the LDs, would Ed M still try to incorporate a deal with the LDs into his arithmetic? It would make things more stable, but would also mean accommodating even more viewpoints into the equation. Personally I hope he does, both because it may restrain some of the more left-wing impulses but mainly because it would give us a stronger government. Same with the DUP although I think the difficulties of including them are probably greater than for the LDs.

  14. UNS isn’t appropriate because there are now so many parties capable of winning at least a few seats.

    I wonder if there would there be any mileage in calculating swings between every pair of parties and then combining in some way? i.e. In England, calculate Con/Lab, Con/LibDem, Con/Green, Con/UKIP, Lab/LibDem etc etc

    Add in SNP in Scotland and PC for Wales.

  15. Correction: the models do give VI as well as seats, what the hell was I saying

  16. Neil A

    I think the LD/DUP are an exclusive or in relation to the SNP. As understand it both have ruled out dealing with the SNP and I’m not sure how labour can please the “borrow more” calls and the “borrow less” calls at the same time.

    Unless of course labour ignore the SNP on the basis they are pretty much forced into giving them confidence. Then a minority government while giving LDs a few crumbs in exchange for stability might work. Whether or not Lab + LD + DUP > 323 is irrelevant, the SNP have nothing to gain from a new election.

    Would the Holyrood elections have to move if a general election coincided with them? I imagine it might be very confusing to run both concurrently.

  17. Omnishambles et al

    If constituency polling was the only way to improve data, then it wouldn’t be a practical option (though polling the Scottish political system separately would be).

    Ashcroft is simply taking a system established for GB polling, and applying it in constituencies.

    The microdata from the pollsters already includes the constituency for each respondent. I’m simply suggesting that it should be possible to create groups of constituencies that behave in a similar way, and to handle that data that produces useful percentage VI figures.

    Nothing will stop the more ludicrous papers from misrepresenting everything anyway!

  18. Alan

    The FTPA specifically requires that the Holyrood elections be moved to accommodate Westminster.

    It happens all the time of course. Holyrood elections should have been in 2015 – 4 and 5 year terms don’t fit together.

  19. Oldnat

    Thanks. I wasn’t sue if there was any specific legislation.

    I guess if we have enough general elections it might be a while until the Holyrood elections happen.

  20. @ OldNat

    “Ashcroft is simply taking a system established for GB polling, and applying it in constituencies.”

    Very important point. It could have many ramifications.

  21. @Roger Mexico

    Agree with all of this. I’ve never thought of the move from LibDem to UKIP as implausible, indeed with the Wales-only polls showing limited movement for Labour and Tories and a dramatic increase for UKIP, even with churn there would have to be an element LibDem to UKIP to make sense of the shifts.

    Points on antipathy towards the Conservatives (Wales has never had a sizeable representation, going from Liberal to Labour without much in between) and the ongoing perception of Plaid Cymru as a language pressure group (despite party efforts to the opposite) are well made. Labour is clearly doing reasonably well, but may be held back by being a party in government in Cardiff Bay.

    In passing (and I wish I could remember where) I saw that someone mentioned that UKIP were particularly strong in’North Wales’, I’m guessing given the paucity of Wales polling means this was anecdotal or based on Euro results, but could reflect slightly different demography in those areas where there is migration from the North of England by retirees.

    The only other thing is whilst this effect may be more pronounced in Wales doesn’t the BES data from the weekend show that there are indeed a noticeable number of ex-LibDems among the UKIP ranks nationally?

  22. Roger Mexico “However the movement to UKIP may be even stronger in Wales. Unlike in England it may be that they are attracting most of their vote from non-Tory sources”

    There is ample evidence that UKIP are taking the Labour vote in North East Wales and also on Ynys Mon (Anglesey). Ynys Mon in particular is one to watch. In the Assembly by-election 18 months ago UKIP took 14% of the vote finishing just over 1% behind Labour and Plaid taking over 50% of the total vote with a near 17% swing ( ) .

    This is a Labour seat at Westminster level (Albert Owen), but the incumbent is largely viewed as a grinning idiot and nodding dog yes man. In GE2010, Plaid suffered because of the presence of a very strong local independent, local farmer Peter Rogers ( ), who basically scuppered them in 2010 and also in 2005 as his vote was almost exclusively Plaid territory. He is not standing this time and a combination of factors make it highly likely that this seat will fall to Plaid in May. Certainly Labour at local level are extremely worried and UKIP believe their presence – which has little impact on the Plaid vote, will topple the Labour MP and gift the seat to Plaid. In turn, Plaid locally believe that they are making inroads into the Labour vote as well as UKIP, that their Assembly Member local boy Rhun Ap Iorworth is extremely popular at street level and their Parliamentary candidate – again a local boy and also very popular make this is very much a winnable seat for them.

  23. So I think Labour are doing well in the southeast, southwest and London, but the problem is that when I look at the Con-Labour marginals 15 of them are in the Midlands, plus another 7 Labour-Con marginals if things swing the other way there.

    So that leaves me to think that I am going to have to pool the crossbreaks to try find some kind of sensible mean, given that UKIP is all over the placeamong the pollsters.

    Other than that it is way to early to start talking about who is going to win this election and offer seat count prognostics.

    Maybe by around sometime after April 16th, when we have had two sort of debates will we be able to spot some trends.

    I know what Clegg is up to by being interviewed around April 30th, he is hoping to be in with the last bounce – but that will be one hell of a mis-calculation if Farage and UKIP have twice as much support as LD and he is runnng neck and neck with Green as they were in the European election in England.

    Voters who are looking to stop Farage and UKIP are not going to be clinging to LD, but trying figure whether to vote Conservative or Labour.

    That’s what happened to the NDP (Labour) between 1988 and 1993 in Canada, they got squeezed down from 20.4% to 6.9% caught between a resurgence of the Liberals, who went from 31.9% to 41.2%, and the rise of Reform as the new protest Party rising from 2.1% to 18.7%.

    The Bloc in Quebec mostly took their votes from Conservative. But in addition during the 1988 Canadian election the multi-party “Other” vote went from 3% to 3.6%, a rag bag assortment of Engish Canadian nationalists, etc.

    In this UK election the Green represents another protest or alternate poll of opinion on the left of Labour and LD, that is allied with the nationalist parties of Scotland and Wales.

    What a fascinating mix.

  24. Assiduosity
    Yes. I pointed it out earlier.

    I wonder what happens if we have (say) 3 UK General Elections in a couple of years? Would even Con and Lab be persuaded to go for some form of PR?

  25. @Andy Shadrak

    The Tories polled a highly improbable 40% in E&W in today’s Ashcroft phone poll. It was largely on the strength of this that the Tories gained a 2% lead. The Tories only polled 39% at their recent zenith in 2010.

  26. @ Stat Geek

    So do your graphs show the LD vote breaking to Con/Lab 6.3 to 13.5 to UKIP Green, which is basically 2:1 to UKIP/Green as compared to Lab/Con.

    That’s a 71.1% decline for LD in England.

  27. Tonight’s YouGov will be highly interesting, telling us whether the recent ComRes and Ashcroft polls, plus the Populus to some extent, are signalling a real strengthening in the Tory position. I know the Sunday YouGov poll showed a movement towards Labour, but the last three published polls are making that look a little like the proverbial sore thumb.

    My hunch is that we’rte still in a dead heat, but if tonight’s YouGov echoes today’s three polls then we may be witnessing a Tory revival. However, what’s still intriguing is the lack of obvious correlation between the Labour and Tory VIs. Both up in today’s Ashcroft, albeit a bigger increase for the Tories and Populus points to an improving position for the Tories, but not at the expense of the Labour vote. ComRes has Labour down three but only a 1 point increase for the Tories. Sample variation noise or the smaller parties beginning to get squeezed? We need more polling evidence but there’s still no great sign of any significant traffic between Tory and Labour, in either direction. Until we see that, then it’s deadlock really.

  28. Someone is certainly “continually banging on about breaking up the UK” but anyone who listened to Ms Sturgeon’s speech Saturday at the SNP conference would have to tell you that it certainly wasn’t her. That is unless you consider a single mention of Scottish independence to be “continually banging on’. :)

  29. CROSSBAT11

    “onight’s YouGov will be highly interesting, telling us whether the recent ComRes and Ashcroft polls, plus the Populus to some extent, are signalling a real strengthening in the Tory position.”

    Caveat: I don’t think a single poll can tell us such a thing definitively.

  30. @Crossbat XI

    I think we really have to look at why the phone polls from Comres and Ashcroft continually show a better position for the Tories than other polls. Is there a shy Lab/Ukip effect or are people more honest over the phone? We simply won’t know until the big day.

  31. @ Albert’s Nemesis; you obviously have different info on Ynys Mon to me! Firstly the Holyhead people I work with who are probably normally Lab/PC swingers are pretty positive about Albert Owen….. and a PC AM of my aquaintance was bemoaning the selection of John Rowlands for YM and Vaughan Williams for Llanelli, reckoning that given their backgrounds a swap would have been ideal! You are right though that this is a seat where UKIP will damage Labour (and the Tories) and Plaid have some momentum from their startling Assembly seat win

  32. @Neil A; I’m not sure which “left wing impulses” you feel will need reining in; the SNP are, when assessed in totality surely just left of centre, PC a little more so but both very capable of striking agreements on policy with the LDs as well as Labour. the only real far left party in this is the Greens

  33. Chris in Cardiff

    Nicola Sturgeon has issued voting advice to English people to vote Green.

  34. Chris In Cardiff. I live in Holyhead and know all the players from both Labour and Plaid. Believe me, even in their heartlands of Holyhead and Amlwch Labour are looking distinctly shakey.

  35. Ashcroft says he will release polls of “the most marginal” seats on Wednesday. 8 in total.

  36. I know what I want to happen
    I know what I think will happen
    They are not the same
    I know what the polls are saying -its too close to call

    Anyway seven dimensional chess on Thursday and the lineups out

    Dave on the right and natalie on the left ,ed bang in the middle.

    Clegg and Farage on eds left as we look at the screen, ms sturgeon next to dave with ms wood next to her.

    I think they get four questions each and then debate four themes.

  37. I always thought that telephone polls were slightly biased against the Conservatives because they are more likely to be on tps lists with a corresponding propensity to refuse to answer polling questions. In the past, when most polling was by phone or face to face, this led to underestimates of Tory VI. If this has somehow turned round I would be very surprised.

    On the other hand, I have yet to encounter a shy kipper. They are usually quite evangelical.

  38. @Prof Howard; I’m well aware of this but if you look at what the SNP have done in Government it’s rather more centrist….. they after all have a lot of very affluent supporters to keep happy

  39. @Chris in Cardiff,

    I suppose I should congratulate you for calling me out. On reflection I was using “left-wing impulses” as code for “following SNP policy on spending”.

  40. Cameron going to “all 4 corners of all 4 nations” in this campaign.

    I wonder who he’ll be talking to in Foyle, Fermanagh, South Down and Antrim North?

  41. @ Albert’s Nemesis I’ve no doubt your local knowledge is better than mine which is limited to periodic work on the island; PC do have form though, for let’s say over enthusiastic forecasting

  42. @RMJ1

    Re:TPS – do calls from polling organisations count as targeted direct marketing? I would not have thought so.

    I really don’t know whether phone polls per se “favour” any party. It’s just odd that Comres and Ashcroft phone polls this year have rarely shown any Labour leads. ICM has shown some so it’s not as simple as just phone polls.

  43. Oldnat,

    To be fair, he didn’t necessarily say which nations. He could mean England, Scotland, Wales and Djibouti.

  44. C in C
    You are of course right about the political position of the SNP. They have no “left wing impulses.” But they do have a left wing tactic for the moment. On the surface, it must be strange that they advise people to vote Green in England when not only do they have almost nothing in common in terms of policies but also at the same time say they want a Labour government. (Vote SNP in Scotland and Green in England to get a Labour government?)

  45. England, Scotland, Wales and Cornwall?

  46. @RAF

    No they don’t but if you have ever tried canvassing people who are on the list you will often get a flea in your ear.

  47. “Cameron going to “all 4 corners of all 4 nations” in this campaign”

    And don’t forget Ynys Mon in the far corner of North west Wales where he will no doubt help Albert’s Nemesis help defeat the utterly loathed Albert Owen, an MP so unpopular he only managed to win the last 3 elections against all the odds!

  48. @ Old Nat….all four corners of all four nations is a good way of being unavailable for debates. With waiting for planes on Bara and lost in lanes near Aberdaron that’s most of his 6 weeks gone

  49. @RMJ1
    No they don’t but if you have ever tried canvassing people who are on the list you will often get a flea in your ear.”

    Lol! I can certainly believe that!

    I do often wonder who would want to get involved in telephone polling. How many people do you actually have to call before you get 500-1000 people willing to participate, let alone sufficiently weighted numbers for a representative sample of 500, let alone 1,000?

  50. BOOOOOMMMMM!!!!!!!!!!

    Tories outpolling Labour in Scotland.

    ComRes Scotland %, seats (change):
    SNP 39%, 39 (+33)
    Con 24%, 7 (+6)
    Lab 21%, 9 (-32)
    LD 13%, 4 (-7)
    Green 2%
    UKIP 1%

1 6 7 8 9