Three weeks to go

Here are this week’s GB polls:

Opinium/Observer (9/4) – CON 36%, LAB 34%, LDEM 7%, UKIP 11%, GRN 6%
YouGov/Sun (10/4) – CON 33%, LAB 35%, LDEM 8%, UKIP 13%, GRN 5%
YouGov/S Times (11/4) – CON 34%, LAB 34%, LDEM 7%, UKIP 13%, GRN 6%
YouGov/Sun (12/4) – CON 33%, LAB 36%, LDEM 7%, UKIP 13%, GRN 5%
Ashcroft (12/4) – CON 33%, LAB 33%, LDEM 9%, UKIP 13%, GRN 6%
ICM/Guardian (12/4) – CON 39%, LAB 33%, LDEM 8%, UKIP 7%, GRN 7%
Populus (12/4) – CON 33%, LAB 33%, LDEM 8%, UKIP 15%, GRN 5%
YouGov/Sun (13/4) – CON 33%, LAB 34%, LDEM 8%, UKIP 13%, GRN 6%
TNS (13/4) – CON 34%, LAB 32%, LDEM 9%, UKIP 14%, GRN 5%
YouGov/Sun (14/4) – CON 33%, LAB 35%, LDEM 8%, UKIP 13%, GRN 5%
YouGov/Sun (15/4) – CON 34%, LAB 35%, LDEM 8%, UKIP 13%, GRN 5%
Ipsos MORI/Standard (15/4) – CON 33%, LAB 35%, LDEM 7%, UKIP 10%, GRN 8%
Panelbase (16/4) – CON 33%, LAB 34%, LDEM 8%, UKIP 16%, GRN 4%
YouGov/Sun (16/4) – CON 34%, LAB 34%, LDEM 7%, UKIP 14%, GRN 5%
Populus (16/4) – CON 33%, LAB 34%, LDEM 9%, UKIP 14%, GRN 4%
Survation/Mirror (17/4) – CON 34%, LAB 33%, LDEM 7%, UKIP 17%, GRN 3%

Voting intention continues to be pretty much static, with levels of Conservative and Labour support extremely close. There were sixteen polls published in the last week, nine had the two parties within a point of each other. The UKPR polling average is back to showing a tie – CON 34%(+1), LAB 34%(nc), LDEM 8%(-1), UKIP 14%(-1), GRN 5%(nc). It is hardly been an exciting election campaign anyway, but certainly nothing seems to have made any significant impact upon voting intention and we are running out of time for anything to do so. The first postal ballots will have gone out this week.

Other polls

This week we’ve also had ComRes polling of Lib Dem seats in the South West and Ashcroft polling of some seats in Scotland. I’ve written about both of them at length already – Ashcroft here and ComRes here.

Week fifteen

This week was manifesto week – the five main GB parties all published their manifestos, though the SNP are saving theirs for later in the campaign. YouGov have done two bits of polling on the manifesto for the Times and the Sun.

The most widely supported policy in the Labour manifesto was to reduce the deficit every year (76% support – largely because it got the backing of Tory voters too), followed by the promise to raise the minimum wage to £8 (71% support), freezing utility bills (65%) and the mansion tax (61%) – none of their main policy announcements got a thumbs down.

Looking at the Conservative manifesto the most popular policy was linking the personal tax allowance to the minimum wage (supported by 80%, again because it got wide cross party support), followed by stopping above inflation rail fare rises (67%) and lowering the benefit cap to £23k (65%). Unlike Labour some of the main Conservative policies got a thumbs down – opening 500 new Free Schools only got 26% support, the flagship announcement of extending right to buy to housing associations only got 28% support.

I shall make my usual caveats about overestimating the importance of individual policies. Despite Labour’s individual policies polling better, in the same poll the Conservatives had a narrow lead on having the best policies and ideas for the country (29% Conservative, 26% Labour). Neither do people pay much attention to these announcements – a separate YouGov poll for the Sun found that the right to buy policy was the only one of the manifesto policy announcements tested that a majority of people could correctly link to the right party – in most cases less than a third of people were able to say which party had proposed it.


The latest forecasts from Election Forecast, May 2015, Elections Etc, the Guardian and YouGov are below, as well as the less regular prediction from the Polling Observatory team who released some new numbers today. As ever, all show a hung Parliament, but most are now showing Labour with more seats in a hung Parliament – with the notable exception of Steve Fisher’s model, which has the Tories with about 30 more seats than Labour. On the subject of the differences between the models, Chris Hanretty of the Election Forecast team wrote a blog post earlier this week.

Elections Etc – Hung Parliament, CON 292(+3), LAB 260(-6), LD 22(nc), SNP 51(+2), UKIP 4(-1)
Election Forecast – Hung Parliament, CON 280(-2), LAB 277(+2), LD 27(-1), SNP 42(+1), UKIP 1(nc)
May 2015 – Hung Parliament, CON 268(+3), LAB 276(-3), LD 26(nc), SNP 54(nc), UKIP 3(nc)
Guardian – Hung Parliament, CON 269(-2), LAB 271(nc), LD 29(nc), SNP 55(+2), UKIP 4(nc)
YouGov Nowcast – Hung Parliament, CON 266(+2), LAB 279(+2), LD 27(-1), SNP 50(-5), UKIP 5(+1)
Polling Observatory – Hung Parliament, CON 268, LAB 278, LD 28, SNP 49, UKIP 3

670 Responses to “Three weeks to go”

1 2 3 14
  1. Three forecasts have changed today. The average of all seven forecasts showing a tie for the second day.

    See tables here:!474615&authkey=!AO8P7XYzPSSbRRs&ithint=file%2cpdf

  2. Another week when nothing happened.

    The Tories pull up to 34 largely on the back of THAT ICM poll, as they’ve been a smidgeon behind Lab most of this week.

  3. 292-260 is certainly bold from Fisher’s model.



    Can’t we have the election tomorrow please?

  5. RAF

    Agreed. Apart from the earthquake in Scotland the polls in England and Wales remain surprisingly constant.
    If it stays like this we will be in a position on the evening of Friday 8th May where some behind the scenes negotiations are going on.

  6. Prof Fisher is taking a gamble but will be hailed as the guru of gurus if he is provem correct.

  7. Anyone know anything about the midnight announcement Smithson mentioned earlier?

  8. Fisher’s numbers do work if UKIP supporters break to the Conservatives by about 3%. Its not impossible.

  9. Going for a 2 point Tory lead in YG, I can feel it in my bones.

    Swingback..Its happening!!

  10. @Catoswyn


  11. Who is making a midnight announcement?

  12. @Bluebob

    At times like this optimism certainly has its merits (much that I hope you are wrong!).

  13. @Gary Gatter
    Thanks for these graphs – very clear.

    It’s interesting that Lab+SNP is only about 325. I know that some of the odds-and-s*ds will usually suppport them as well, but there could be a lot of close votes if that turns out to be right.

  14. The midnight announcement is from Lynton Crosby. He is due to announce that swingback to the Conservatives will start from 6am on Monday.

    Source: Reuters

  15. @ BLUEBOB

    My prediction as well.

  16. @ RAF

    It’s been a long day

  17. England Only – Survation crossbreaks – Con-Lab swing of 5.05% (MS).

    We need to see dedicated England only polls.

  18. Millibounce following yesterday’s debate – lab 35, con 32, ukip 15, LD 9

  19. RAF

    I agree. I’d love to see England only polls right now.

  20. Cross-breaks:
    Applying the regional swings from the mean of the last 7 Yougov polls to my constituency database makes the following changes to the number of seats compared to my existing method of London, Scottish, Welsh and ‘rest of England’ swings (from properly weighed polls – roE is derived):
    LD +2
    others – no change.

    However, the fit to the Ashcroft polls was worse, in particular, too many Southern seats going to Lab where Ashcroft has shown Cons clearly ahead.

    My conclusion: don’t look at the cross-breaks!

  21. @ Omnishambles

    “The midnight announcement is from Lynton Crosby. ”

    Does he know that he has to finish it by the break of the dawn?

  22. Labour think they can beat farage ,giant mass mailing organised in thanet south likening ukip to the bnp(see guardian website)

  23. @07.05.2015

    Maybe Labour have imbibed @Bluebob’s optimism serum.

  24. Don’t think the Conservatives would benefit from being ahead. I think that the more it looks like Labour might pull it off, the more chance that UKIP voters and others who are ‘anti-Labour’ rather than pro-Tory will switch to prevent this.

    It would be too much perhaps to ascribe this as a deliberate policy by Crosby and Co. but it certainly looks from the outside as if the Conservatives are playing it all low key. Cameron has even been accused by the press of looking ‘disinterested’ in whether he is going to win or not. Can’t believe that so I can only think the style is deliberate and the only thing I can think of is that they don’t want to pull ahead too early.

  25. 07052015

    Well Labour are certainly putting up a fight in Thanet South. I think the Conservatives had it on their ‘104’ list of areas not to put too much effort into.

  26. Crikey Catsowyn. I think you are giving the Tories a huge amount of the benefit of the doubt. I’m quite sure they would rather be ahead in the polls at this late stage than neck and neck.

  27. If CVI does measure tactical voting (among other things) then it’s hard to believe that it won’t already be showing through in the CVI measure already.

    As the largest of Scotland’s minor parties, it’s worth having a look at the Tory SVI -> CVI changes in Ashcroft’s latest Scottish polls.

    SVI : CVI : % loyal : % to Lab : % to SNP : % to LD : % to others : Constituency
    32% : 34% :97% : 2% : 0% : 1% : 0% : Dumfriesshire etc
    31% : 30% :91% : 0% : : 0% :8% : 1% : Berwickshire etc
    27% : 24% :88% : 7% : 5% : 0% : 0% : E Renfrewshire
    19% : 16% :78% : 1% : 2% : 17% : 3% : NE Fife
    15% : 11% :75% : 21% : 1% : 23% : 1% : E Dunbartonshire
    11% : 7% :59% : 3% : 2% : 34% : 1% : Ross etc
    7% : 5% :71% : 8% : 19% : 2% : 0% : Paisley & Ren S
    6% : 6% :89% : 5% : 4% : 1% : 0% : Glasgow SW

    As you would expect from constituencies with different dynamics, the extent and direction of tactical voting varies significantly. That suggests that generalised comments such as “Scottish Tories will do X” are very wide of the mark.

  28. Since you all obviously have nothing to do I might as well throw this link up again for those interested.
    I highly recommend all you politico’s give it a whirl.
    The constituency data is particularly amusing. If this quiz was replicated the results would be….
    Labour 583 seats
    SNP 49 seats
    Greens 14 seats
    UKIP 4 seats
    If only people voted with their brains….

  29. @MIKEY

    I’m just scared Mikey!

  30. @omnishambles

    Will he announce which Monday? (My guess is May 11th)

  31. Can anyone remember 40 30 20 10? I am quite sure the Cons would prefer those to be the polls today (Con first %)..

    I travelled down from Bristol towards the south coast today and passed through a number of political domains. Of course field posters are mainly Con but displayed a variety of designs. One area had the ones with what I understand the PM now thinks is ‘green crap’ on them, so perhaps the environmental vote was being sought there. Others went for Union jacks on them so one assume Nigel’s supporters were the target in that constituency.

    Speaking of first names being sufficient to identify politicians, in Shepton Mallet I espied ‘Tessa’ posters. I think her surname has been of influence in that case.

  32. @Catoswyn
    I’m just scared Mikey”

    And with that comment I just KNOW you cannot possibly be a Tory! Lool!

  33. I think the Nicola Sturgeon factor might turn out to be the Conservative match winner (if they are to have a winner). There is too much arrogance in her continuing use of the phrase “let’s lock out the Tories” and she forgets that around 70% of UK voters in England and Wales actually want to see one or other of the two main parties win a majority. She is an impressive debater but she may have gone a tad too far.


    lol – “don’t want to pull away too early!”

    Must be the Labour strategy in Scotland too.

  35. And I think she (NS) may also have gone too far in telling EM that his supporters will never forgive him. (she sounds too desperate for power for its own sake)

  36. Haha. We are all scared! As a Reading fan travelling to Wembley tomorrow I’m terrified!

  37. Rivers10

    That quiz was hopeless. I chose two areas and then in the first area, I ended up having to agree to one set whilst I agreed with nearly all in every set (environment was the area).. So i could not choose (it was all or nothing for each party). Useless.

    Mrs BH gave up too.

  38. BristolianHoward
    I think the idea is to choose all the ones you would consider voting for initially and then your favourite, but well it obviously wont be to everyone’s taste.

  39. Rivers10

    Perhaps I am not sufficiently partisan. With just one or two little exceptions, nearly all the policies on environment from all the parties were to my taste.

    I’ll bet I am the only person here who picked environment.

  40. I did that test before with everything selected.

    I got CON: 40%, LD: 30%, LAB: 30%

    Not that helpful but it was interesting


    @”And I think she (NS) may also have gone too far in telling EM that his supporters will never forgive him”

    Actually, I didn’t understand any of that stuff.

    Miliband said he “rejected” a Coalition, when we know he can’t get one because of Trident.
    NS does all this staring & threatening about EM not turning her support down, when EM can’t stop her MPs voting in support.

    So it seems to me that EM must be pleased that he WILL receive SNP support, unconditionally if he wishes-and NS is not going be the cause of an EM administration being defeated in the HoC., whether she gets concessions or not.

    But I’m probably missing some finer points :-)

  42. BristolianHoward
    I did them all because I’m sad :(
    But seriously I agree with you it is a bit all or nothing but isn’t that kinda how our electoral system works, we can only place one x on the ballot paper. I personally struggled for a good while on the defence question on what I could tell where the Greens and Labours policy sets. I agreed with all of the Labour points bar Trident which is a biggie for me so I had to weigh it all up, ultimately I went with the Labour set.

  43. Omnishambles
    I cant remember if I put my results up but in case your interested I got
    Green 50%
    Labour 40%
    Lib Dem 10%

  44. Suppose the alleged midnight announcement was that in the event of a Tory majority Boris would be Deputy PM and that he has taken himself out of any 2019 leadership battle to replace DC. What effect would we expect that to have on current standings if any?
    A knockout blow For The Lib Dems?
    UKIP reduced to well under 10%?
    No Labour gains in London but instead ,say, half a dozen Tory ones?
    Are any, or indeed all, of these possible?
    One thing’s for sure it would certainly make for some poll-watching excitement over the next week we haven’t really had so far.

  45. @ Simon K

    292-260 is certainly bold from Fisher’s model.

    I have it much the same.

    Yet two things must be borne in mind….

    1. I got it right in 1992
    2. But I am utterly hopeless in terms of both statistics and mathematics.

  46. David West

    Sturgeon “forgets that around 70% of UK voters in England and Wales actually want to see one or other of the two main parties win a majority.”

    I’m the first to admit to a lack of specialist knowledge about the motivation of voters in England.

    However, I’m surprised that Labour voters in England don’t want to “lock out the Tories”.

    It is one of the great things about this site that you learn something new every day.

  47. @Colin (10.17)

    Agree, except for a couple of issues eg. Trident, I see no liklihood of the SNP voting against EM. To do so would put the SNP back in the position they were before the referendum.

  48. @ Pete B “It’s interesting that Lab+SNP is only about 325”

    Remember to discount Sinn Fien.

  49. SEN5C
    I think there’s no chance of that being the announcement, but I can only think that it would make people less likely to vote Tory. Boris is popular, but if he’s never going to be leader it could lose them votes.

1 2 3 14