Two weeks to go

A brief summary of this week’s polls before I have some downtime:

YouGov/Sun (17/4) – CON 34%, LAB 34%, LDEM 9%, UKIP 14%, GRN 5%
Opinium/Observer (17/4) – CON 36%, LAB 32%, LDEM 8%, UKIP 13%, GRN 5%
YouGov/S Times (18/4) – CON 33%, LAB 36%, LDEM 8%, UKIP 13%, GRN 5%
YouGov/Sun (19/4) – CON 34%, LAB 35%, LDEM 8%, UKIP 13%, GRN 5%
Ashcroft (19/4) – CON 34%, LAB 30%, LDEM 10%, UKIP 13%, GRN 4%
Populus (19/4) – CON 32%, LAB 34%, LDEM 9%, UKIP 15%, GRN 4%
ICM/Guardian (19/4) – CON 34%, LAB 32%, LDEM 10%, UKIP 11%, GRN 5%
TNS (20/4) – CON 32%, LAB 34%, LDEM 8%, UKIP 15%, GRN 5%
YouGov/Sun (20/4) – CON 34%, LAB 35%, LDEM 7%, UKIP 13%, GRN 5%
YouGov/Sun (21/4) – CON 35%, LAB 34%, LDEM 7%, UKIP 13%, GRN 5%
YouGov/Sun (22/4) – CON 33%, LAB 34%, LDEM 7%, UKIP 14%, GRN 5%
ComRes/ITV/Mail (22/4) – CON 36%, LAB 32%, LDEM 8%, UKIP 10%, GRN 5%
Populus (23/4) – CON 32%, LAB 35%, LDEM 8%, UKIP 14%, GRN 5%
YouGov/Sun (23/4) – CON 33%, LAB 35%, LDEM 8%, UKIP 13%, GRN 6%
Panelbase (23/4) – CON 31%, LAB 34%, LDEM 7%, UKIP 17%, GRN 4%
Survation/Mirror (23/4) – CON 33%, LAB 29%, LDEM 10%, UKIP 18%, GRN 4%

The UKPR polling average continues to show a tie – CON 33%(-1), LAB 33%(-1), LDEM 8%(nc), UKIP 14%(nc), GRN 5%(nc). Some of the individual pollsters are showing consistent leads for one party or the other (YouGov and Populus, for example, are generally showing small Labour leads, ICM and Ashcroft small Conservative leads) so it’s not the case that all the pollsters are showing an exact tie, more than the average of the different companies’ house effects is neck and neck.

Other polls

There was one Scottish poll this week – a new YouGov poll that confirmed their previous 24 point lead for the SNP (tabs), there was also a new YouGov poll of London for the Evening Standard with topline figures of CON 32%, LAB 44%, LDEM 8%, UKIP 10%, GRN 5% (tabs).

There were three constituency polls. Lord Ashcroft released two extra constituency polls with his weekly GB poll, showing the SNP ahead in Edinburgh North and Leith and Edinburgh South. Meanwhile Survation released a new poll of Thanet South, showing Nigel Farage nine points ahead. We should have some more constituency polls from Lord Ashcroft first thing tomorrow morning, including a poll of Rochester and Strood.


The latest forecasts from Election Forecast, May 2015, Elections Etc, the Guardian and YouGov are below, all continue to show a hung Parliament, but the models disagree on whether the Tories or Labour will be ahead on seats – Elections Etc, Election Forecast and the Guardian all have the Conservatives with more seats, YouGov and the Guardian have Labour slightly ahead.

Elections Etc – Hung Parliament, CON 286(-6), LAB 263(+3), LD 26(+4), SNP 51(nc), UKIP 4(nc)
Election Forecast – Hung Parliament, CON 283(+3), LAB 270(-7), LD 24(-3), SNP 48(+6), UKIP 1(nc)
May 2015 – Hung Parliament, CON 270(+2), LAB 273(-3), LD 26(nc), SNP 55(+1), UKIP 3(nc)
Guardian – Hung Parliament, CON 273(+4), LAB 268(-3), LD 28(-1), SNP 55(nc), UKIP 4(nc)
YouGov Nowcast – Hung Parliament, CON 270(+4), LAB 277(-2), LD 27(nc), SNP 50(nc), UKIP 3(-2)

339 Responses to “Two weeks to go”

1 4 5 6 7
  1. @NeilJ – Salmon was going to be the game changer a few days ago now it’s Libya – get a real sense of desperation and straw grabbing now by some

  2. Roger –
    Thanks for those figures.

    I think it’d also be important to look at the crossbreaks – with the obvious caveat that with small subsamples comes great responsibility.

    Lab voters:
    Lab Maj: +80, Lab+Lib +23, Lab+SNP +10, Lab+Lib+SNP +12, Lab+Con -43
    All Labour options have positive results with Labour voters – except the grand coalition. Lab+Lib only marginally more popular than Lab+SNP.
    This should be the proof that puts to bed the argument that “Labour should form a grand coalition rather than work with the SNP”.
    I also wonder if the Lab+SNP figure would have been higher had Labour not spent so much time ruling out working with the SNP [1].

    Con voters –
    Con Maj: +93, Con+Lib +61, Con+Lib+UKIP +5, Con+Lib+UKIP+DUP -18, Con+Lab -42
    “Rainbow” coalition unpopular with Con voters – If they try to scrape through with 323 seats it may actually be more damaging than allowing Labour to scrape through and waiting for the next election.

    LibDems (Warning: Very small subsample)
    Con+Lib: +47, Con+Lib+UKIP -47, Con+Lib+UKIP+DUP -57, Lab+Lib +35, Lab+Lib+SNP -37
    LibDems slightly toward preferring Con over Lab but only if it doesn’t require parties to the right.
    LibDems stuck between a rock and a hard place if it comes to being kingmakers in any rainbow coalition, right or left. But the party did take the electorally difficult decision of Con+Lib coalition in 2010 so if it could manage the internal discipline again it shouldn’t be too much of a problem.

    UKIP voters far more positive about Con governments than Lab – which should be no surprise given the make up of their party support.
    Con+Lib+UKIP +17, Con+Lib+UKIP+DUP -8

    Surprised that YouGov didn’t think that Con+UKIP was likely enough to ask the question – if Con are a few short, they may ask UKIP for support (if only to try to win back deserters, if UKIP rejected it).

    All –
    Is it too late for either Labour or Conservative to switch electoral narratives?
    The Scottish referendum was won in part because a sudden change in electoral narrative at the end of the referendum – a switch from fear to love.
    IMHO, Lab and Con are making exactly the same mistake, but in the opposite direction, because they’re both running the same sort of classist campaign.
    Lab are running a campaign talking about protecting those at the “bottom end” of a class structure – tenants, workers, debtors, etc
    Con are running a campaign largely in defence of those at the “top end” of a class structure – landlords, bosses, creditors, etc
    Both parties seem to do best, electorally, when they ignore class entirely, and instead go for those with aspirations of living in a centre-right classless society – homeowners, self-employed, etc
    Cons have spent much more time on this – with help to buy, etc but both seems to be neglecting this group.

    I realise that both parties, when neck and neck are trying to get their core out to vote – but is it too late for either to switch toward focusing on that centre-right aspirational voter?

    [1] IIRC AW has mentioned polling that asks “Labour support this policy, do you support this policy?” massively boosts Labour support for the policy in polling. So partisanship plays a part, rather than pure ideology.

  3. I’m finding NC’s comments weird about legitimacy. If a grouping of parties gets the numbers then that’s legitimate under our constitution.

  4. NC is obviously trying to shore up his votes in the South East. This legitimacy argument is an incredibly flakey one – anyway it’s still a cointoss whether he’ll be about to stipulate any agreement rules

  5. In avance of Lord A,who blames DC personally in the Indy for the Tories poldrums,one stst from the latest yougov to give the blues hope.

    Undecided voters are much better disposed towards the coalition and its record than those who have made their mind up.

    However this may simply be becos the undecided contain a large jumber of 2010 lib dem voters ,often women ,who could simply sit on their hands on may 7.

    Ps kate is late then ,

  6. CloudSpotter reminds us from time to time that there are less LD voters to squeeze in Con/Lab marginal as they already have been in previous GEs.

    Cross-breaks are seeming to show that the Lab net gain over the Tories from the 2010LD vote dispersion has lowered to perhaps 15% from the 20+% that was eveident for a long time.

    This could be just tactical voters returning to LD in Con/LD seats but also imo some is ABLab moving to Con in Con/Lab seats.
    Question is then will the LD squeeze be less or more in Con/Lab marginal than UNS and will the split be different.

    FWIW, it seems to me that even if there is more 2010 LD squeezing in Con/Lab marginal and even if Lab benefit more in net terms this 2010 LD vote dispersion will not deliver enough seats to Lab if other things remain equal.

    Upshot Lab need a net gain from the Tories from some other vote movements or combination of movements.
    Probably better amongst first time voters and they may well win the GTVO contest in marginal but they need the UKIP to hold more Blue Kippers than Red Kippers by at least 2% and maybe more in the key seats.

    Of course Labour and Cons will be trying to attract a few extra voters and GTVO in marginals but the key to the GE result (can Labour take 40+ Con seats) is how effective the vote UKIP get Lab/Tory messages are in marginal seats 20-60 on the Lab targets list of the Cons.

  7. @JimJam

    “Cross-breaks are seeming to show that the Lab net gain over the Tories from the 2010LD vote dispersion has lowered to perhaps 15% from the 20+% that was eveident for a long time.”

    True. But in the last Survation poll (4% Con lead) it was down to 2% which is reason to doubt whether that poll was drawn from an unusual sample.

  8. Phil,

    At the risk of being accused of cherry picking I dismissed that poll less due to the 4% lead which within MOE is possible for the Tories but due to the 29% for Labour which is way lower than anyone else.
    That 2% figure will be a large part of the explanation.

    Also the 6% ICM (Cons lead) and Panel base 6% Lab lead can be ignored imo.

  9. looking at various rolling averages of polls, I would venture to predict that the outcome in vote share will be something like:

    con 33.5%
    lab 33.5%
    lib dem 8%
    ukip 13.5%

  10. If past Liberal leaders had followed Clegg’s made-up constitutional “rules”, there’s a few Liberal governments that would not have been formed (e.g. Asquith 1910, Gladstone 1886).

  11. @Old Nat

    I suspect that your comment was more down to obtuseness than real misunderstanding, but let me spell it out.

    Potential SNP gain from Lab = 0 net gain in seats for potential Lab/SNP coalition/deal/whatever (Sturgeon’s professed outcome)

    Potential Lab gain from Con = 1 net gain in seats for potential Lab led government (whether it relies on SNP support or that from anyone else)

    The latter depends on English newspaper headlines, the former Scottish ones.

    Which is why I don’t give a “diddly squat” about Scottish newspaper headlines at the moment, but am quite concerned about Sturgeon’s efforts to give English newspapers an excuse to keep banging on about fears of the SNP influencing the government.

    Sturgeon’s actions are rational only if her intent is to promote either or both of the following outcomes (a) another Conservative government in order to lay the conditions for a “Yes” vote or (b) a seat total for Labour that would be insufficient for Labour to govern as a minority party with support from other minor parties but not the SNP. In both cases, that requires her to try and damage Labour in England, and I’ve reached the conclusion from her actions that she is intent to do so.

  12. @Jim Jam

    I don’t think you should ever dismiss a poll just because of the headline figures they generate.

  13. Correction (7.40 post)

    Should read “But in the last Survation poll (4% Con lead) it was down to 2% which is reason to doubt whether that poll was drawn from a typical sample.”

  14. James
    1923 wouldn’t have happened either – Conservatives were by far the largest party (67 seat lead), but Labour formed a (short lived) government with Liberal support.
    So if they were following his “clear mandate” from the largest party rule, the Liberals should have backed the Conservatives.

    I think there will be lots of faces covered in eggs come the election result – unless the Conservatives can get a clear majority or enough seats for a another Con/Lib coalition.

  15. @Jack

    Best not to take too much notice of any report along the lines of “Nick Clegg says that he will do x after the election”, based on prior experience.

  16. @Phil Haines

    There is a much more obvious explanation.

    The SNP’s complaint in previous UK elections has been a lack of coverage, which then leads to people who may be sympathetic voting for other parties (usually Labour). e.g. in 2010, the SNP’s only input to the televised debates was to take an action at the Court of Session in a vain attempt to stop them happening!

    Therefore if a major newspaper like the Times is offering front page coverage, with the opportunity to get their views across, the SNP are going to take it.

    It also influences their attitude to TV debates. If the Tories or Labour were ~20 points ahead in the polls, they simply wouldn’t be happening. The SNP are ~20 points ahead in the Scottish polls, yet Sturgeon has made more of these appearances than any other figure in the UK (2 GB debates and 3 in Scotland). If they don’t take the publicity when its on offer, the SNP think they will be ignored again.

  17. @Phil Haines

    “It’s not about you”

    Does it ever occur to you that Sturgeon is trying to win the election in Scotland? This sense of entitlement is why Labour is dying in Scotland, there is no obligation on the SNP to help Labour win an election, the SNP are fighting their own corner

  18. @ Couper 2802

    ….there is no obligation on the SNP to help Labour win an election, the SNP are fighting their own corner

    Okay, so Sturgeon can “lock David Cameron out of Downing Street” on her own, can she? Because that’s the election that she herself says she is fighting now.

    Phil Haines is spot on; it’s you who refuses to see that Sturgeon’s behaviour gives the lie to her core campaign promise.

  19. Phil,

    I will re-phrase.

    Give little consideration to those 3 polls.

  20. @Cloudspotter/Ray

    OK but the difference in excluding Ashcroft has nothing to do with Lab/Con marginals. Lab ‘lose’ 21 seats ‘because of’ Ashcroft but 13 of these are to the LDs and 7 to SNP, only 1 to Con.
    Thus to say Lab does worse in marginals is misleading IMO

  21. @Amber Star

    The SNP are fighting to win the Scottish election. It is that simple. The idea that they have any obligation at all to help Labour is bonkers. No one except desperate Labour folk expect them to.

    In Scotland the campaign is a media chimera worthy of the worst Tory tactics. LiS have stared into the abyss for too long they have become the monster.

  22. It would be sad if the apparent dearth of election posters led to an equal decline in witty defacement. I smiled at the addition of the postscript ‘nothing’ to my local MP Philip Dunne’s name. Anyone seen any other notable amendments?

  23. Ashcroft:

    Ukip up in Thurrock

    Tories up in Rochester

    Big Lab lead in Bristol West

    Something for everyone?

  24. Greens are actually second in Bristol West, but well behind Labour (Lab 38, Green 25, LD 20, Con 14)

  25. @ Couper 2802

    The SNP are fighting to win the Scottish election. It is that simple. The idea that they have any obligation at all to help Labour is bonkers. No one except desperate Labour folk expect them to.

    Okay, so the voters were watching a doppelgänger, during the debate, when Nicola Sturgeon offered to help make Ed Miliband prime minister & lock David Cameron out of Downing Street. No one except “desperate Labour folk” thought it was really Nicola Sturgeon standing there, saying those words.

  26. If Lab are polling that well in a 3 way marginal seat like BW (at the same time as Con doing that badly) then that suggests to me they are on their way to largest party status

  27. Labour friends of mine have long been a little jittery about Bristol West. This seems to indicate they’ve nothing to worry about. Must be disappointing for the Greens, but one cannot win a seat on lefty students alone (take heed, Sheffield Central).

  28. Guymonde,
    None of today’s 3 Con-Lab marginals show a swing of 5%.

  29. @AmberStar

    On Any Questions last night, Charlie Falconer, in an otherwise pretty ineffectual performance, gave what I thought was a plausible explanation as to why the emergence of a weak Tory led government after May 7th was the SNP’s best bet in terms of achieving their long term political goals.

    This is probably Sturgeon’s covert view too and she may well be playing this election with that view in mind. In fairness too her, those would probably be my covert objectives and tactics too if I was an SNP politician. EVEL, for example, is a political gift to Sturgeon.

    Talking about Charlie Falconer, why are Labour, as we enter the last 12 days of an extraordinarily tight election campaign, not fielding their big hitters and best media performers at every opportunity now? Charlie is a very nice fellow, but he’s bumbling performer at the best of times and Liam Fox and Stewart Hosie ran rings around him last night. I lost count of the number of open goals he missed and he disappeared from nearly every discussion as Swinson, Hosie and Fox dominated the debate topics throughout.

    In the same way that I can’t believe the Tories fielding people like Schapps, Soubry and Truss in the way that they do, the Labour election team need to think harder about this. Charlie Falconer is not your man as you go into the final laps of an election campaign.

  30. I think that’s pretty encouraging for the Greens in Bristol West. If it’s clear that they are in second place, I reckon they’ll get a lot of Lib Dem votes. Obviously it’s even more encouraging for Lab.

  31. But tories ahead in colne valley(labour 76 target-5.3% swing needed) and high peak(labour target 68 ,4.65 swing required) tho too close to call.

  32. My favourite defacing of posters occurred in Cambridge in the late 1970s.

    Greene King ran a huge advertising campaign across the city for Abbot Ale, with the strapline: ‘Everything They Say About It Is True’.

    To which a very determined graffiti artist added to every one: ‘Abbot is Yak’s Piss’.

  33. Charlie prob had a wee chat to swinson and hosie tho as he will lead labours negotiating team.

    EM doesnt have a john reid he can turn to and he will defo need one where he seems to be going.

    I thought wee dougie, thornberry ,benn and sadiq khan were the best of his close circle .

  34. Two articles in today’s Times about the Con leaders’ campaigns in Scotland & UK.

    Very contrasting views , & more complementary to the former than the latter.

  35. @Crossbat

    Totally agree about Charlie Falconer – likeable man, but always wrong in my experience.

    I was equivocal about euthanasia until he came out in favour. I immediately concluded that it must be a bad idea, and have not regretted that decision.

    And why go for Soubry and Shapps when Hammond and Hague are available? Madness.

  36. @Amber Star

    It is NS’s strategy to win and keep the ex-Labour voters in Scotland and of course she will keep to her word she has a party full of ex-Labour members and supporters.

    Politicians go all out to win elections and to maximise their parties vote, it is that simple.

    LiS was running with ‘vote SNP get Tory’ so it was a very obvious position for her to take. The fact the Tories picked it up to damage Labour, is because they are looking to maximise their vote, that is how democracy works.

    I have said before by giving into pressure and ruling out SNP deals EM played into Tory hands. EM has accepted the Tory claim that SNP are illegitimate and so has helped the with the demonising of the SNP. This will also cause him a massive headache on May 8th when he finds he does have to do a deal with the SNP.

    So, if you are looking for anyone to blame for the Tory SNPBad campaign look to: LiS and EM

  37. CB11
    “In the same way that I can’t believe the Tories fielding people like Schapps”

    Better perhaps than fielding Shapps?

  38. I don’t think my crossbreak aggregations are telling us much, but I’ll complete the series up to the election anyway.

    Scottish crossbreak in today’s YouGov

    SNP 42% : Lab 28% : Con 15% : LD 9% : UKIP 3% : Grn 2%

    This week’s mean of YG Scottish crossbreaks

    SNP 43% : Lab 27% : Con 17% : LD 8% : UKIP 3% : Grn 3% (Range – SNP 40-47 : Lab 23-28)

    Mean of Scottish crossbreaks in last 3 week’s YG polls

    SNP 43% : Lab 27% : Con 16% : LD 7% : UKIP 4% : Grn 3%

    Mean of Scottish crossbreaks in last 3 weeks Populus polls

    SNP 46% : Lab 27% : Con 15% : LD 8% : UKIP 3% : Grn 1%. (Range – SNP 41-49 : Lab 24-32)

  39. Latest Ashcroft marginals look interesting.

    I also note that even when prompted to think about the candidates standing in their own constituency some people still plan to vote for a party that is not standing (e.g. Greens in Thurrock). Some people are going to have to make a last minute decision in the polling booth.. (That could affect the result in a few tight seats)

  40. Abbot is alright.
    But Greene King IPA is the devil’s dishwater!
    I wouldn’t use it to shampoo my worst enemy’s dog.

  41. Maybe having Shapps on TV keeps him away from his computer keyboard?

  42. @OldNat

    They are very useful and are telling us that nothing is changing, if anything the SNP are slightly up on January, Labour slightly down but I’d need Unicorn to confirm and last time I thought the SNP were slightly up she promptly disabused me.

  43. I read the two northern Con /Lab marginals as discouraging for Labour.

    (Bristol NW has a sophomore MP who is never out of the local paper, so not so hopeful for Labour)

    They are not lost but clearly a huge effort on turnout is their only hope IMO.

    LDs totally but totally doomed in the SW.

  44. The UK-Elect latest forecast will be changed today to include the new Ashcroft data.

    If anyone wants to compare the existing seat predictions (of what is actually a fairly mainstream forecasting methodology) with the Ashcroft polls they can do so [email protected] April 19th UK-Elect predictions

  45. Does anyone think that those marginal look poor for Labour? They only win one and no large swings. Is there a particular reason Labour are not winning these?

  46. Bristol NW now backed into a certainty, ticks up for the Tories in 5 of the 6..
    .Farage admitting he has ignored medical advice & now struggling under a Dr twice a week could make Thanet South voters think about his 5 year commitment?

  47. @Couper

    They are only three CON-LAB seats and the polls are pretty much in line with what Election Forecast has for them. But shows all is to play for – if the CONs can hold on to other seats like High Peak and Colne Valley they have every chance of being the largest party.

    Rochester is encouraging for CON but still a close a race. Thurrock shows UKIP can still make some gains beyond by-election seats and Thanet. Bristol West isn’t really a surprise but makes Stephen Williams look doomed. I guess Greens will feel it was always a long shot for 2015 – leaves them well embedded for 2020.

  48. Oh dear… Cameron gets West Ham and Aston Villa mixed up. Game changer?

  49. Am I right in thinking that based on Labour losing 45-50 seats in scotland and winning approx 10 from the LDs, they need to win approx 50 from the Tories. to be the largest party.

    How are labour doing in those 50 seats?

1 4 5 6 7