Four weeks to go

Here are this week’s polls – slightly fewer than usual because of the Easter bank holiday, meaning we didn’t get the usual Ashcroft poll on Monday.

Survation/Mirror (3/4) – CON 31%, LAB 33%, LDEM 9%, UKIP 18%, GRN 3%
Opinium/Observer (3/4) – CON 33%, LAB 33%, LDEM 7%, UKIP 14%, GRN 7%
YouGov/S Times (4/4) – CON 34%, LAB 33%, LDEM 10%, UKIP 13%, GRN 4%
Populus (6/4) – CON 31%, LAB 33%, LDEM 10%, UKIP 15%, GRN 4%
YouGov/Sun (7/4) – CON 33%, LAB 35%, LDEM 8%, UKIP 14%, GRN 5%
TNS BMRB (7/4) – CON 30%, LAB 33%, LDEM 8%, UKIP 19%, GRN 4%
ComRes/ITV/Mail (8/4) – CON 34%, LAB 33%, LDEM 12%, UKIP 12%, GRN 4%
YouGov/Sun (8/4) – CON 34%, LAB 35%, LDEM 8%, UKIP 13%, GRN 5%
Populus (8/4) – CON 31%, LAB 33%, LDEM 8%, UKIP 16%, GRN 6%
YouGov/Sun (9/4) – CON 35%, LAB 34%, LDEM 8%, UKIP 12%, GRN 4%
Panelbase (9/4) – CON 31%, LAB 37%, LDEM 8%, UKIP 16%, GRN 4%
Survation/Mirror (9/4) – CON 31%, LAB 35%, LDEM 9%, UKIP 15%, GRN 4%

Most of the polls continue to be close (with a couple of notable exceptions towards the end of the week), but we’ve seen significantly more Labour than Conservative leads in the last week, putting the UKPR polling average back to a one point Labour lead – CON 33%(-1), LAB 34%(nc), LDEM 9%(+1), UKIP 15%(+1), GRN 5%(nc). It’s too early to be confident there has been any real change, but it’s worth keeping an eye on it.

Scottish, constituency and other polling

If it hasn’t been too busy a week for GB polls, there has been a wide selection of other polls. Last weekend there was a ComRes poll of Thanet South (the seat where Nigel Farage is standing) which showed the race as almost a dead heat between the Conservatives, UKIP and Labour. It provoked an attempt by UKIP to try and pooh-pooh ComRes’s methodology, which I wrote about at length here.

There were two Scottish polls, both of them showing no sign of the SNP’s lead fading away. Panelbase found an SNP lead of 16 points over Labour, up from 10 in their previous poll (though it’s possible that was connected to question order). Tabs for that are here. Last night YouGov put out a new Scottish poll conducted after the first of the Scottish leader debates and straddling the second of them – it found the SNP’s lead growing, up to 24 points over Labour. Tabs for that are here.

Lord Ashcroft released another batch of his own constituency polling mid-week, this time returning to some Con -v- Lab marginal seats where he had previously found tight races. These don’t tell us much about the national race (they were a selection of seats that were showing a smaller than average swing to begin with) but give us some details on the individual races. More details here.

Finally, I saw an unusual county level poll of Kent today – conducted by Facts International (the company who do the fieldwork for ComRes’s phone polls for the Mail). They found voting intentions in Kent of CON 39%, LAB 22%, LDEM 6%, UKIP 24%. That would be a 5% swing from Con to Lab, which on a uniform swing wouldn’t be enough for any seats to change hands (Labour’s closest target seat in Kent is Dover, which needs a 5.2% swing). Of course, the seats in Kent that are actually most interesting are not Labour targets, but UKIP ones – holding Rochester & Strood and trying to win Thanet South, and a uniform swing across Kent isn’t much good in predicting the distribution of UKIP’s vote.

Week Fourteen

This week doesn’t seem to have been a good one for the Conservatives. The start of the week was was largely dominated by Labour’s policy on non-doms, the end over arguments about Trident which became overshadowed by rows about the language Michael Fallon used to criticise Ed Miliband. How you tax rich people and whether or not the Conservatives are too nasty, to put it mildly, strong issues for the Tories. I expect we’ll get polling on all of these issues at the weekend.

Yesterday was also the close of nominations for the general election. I’ll be trying to get the Election Guide part of the site updated as soon as possible to reflect the full official candidate lists (so excuse me if I go a bit quiet!)


The latest forecasts from Election Forecast, May 2015, Elections Etc, the Guardian and YouGov are below. As ever, all show a hung Parliament – currently two are predicting the Tories to have more seats, two Labour to have more seats, and the Guardian’s model a dead heat.

Elections Etc – Hung Parliament, CON 289(-11), LAB 266(+8), LD 22(+2), SNP 49(+2), UKIP 5(nc)
Election Forecast – Hung Parliament, CON 282(-2), LAB 275(-1), LD 28(+2), SNP 41(+1), UKIP 1(nc)
May 2015 – Hung Parliament, CON 265(-13), LAB 279(+12), LD 26(+1), SNP 54(nc), UKIP 3(nc)
Guardian – Hung Parliament, CON 271(-5), LAB 271(+1), LD 29(+1), SNP 53(+3), UKIP 4(nc)
YouGov Nowcast – Hung Parliament, CON 264(+2), LAB 277(+1), LD 28(nc), SNP 55(-1), UKIP 4(-3)

463 Responses to “Four weeks to go”

1 6 7 8 9 10
  1. Opinium Research on twitter:

    “Stay” takes lead in #EUreferendum tracker for 1st time: 45% would vote stay, 41% would vote to leave

    That’s page 48 in the tables link given above.

  2. Is this a phone or online poll?

  3. Opinium is online

    I understand the Guardian have an ICM phone poll in the works that should be published on Monday.

  4. Not a surprising poll from Opinion. As I have written previously, whichever side (red or blue) seems to be suffering from bad news then shows a significant improvement in the following poll.

  5. Thanks @James.

  6. CYT
    Yes, interesting times, to coin a phrase.
    It’s made me think, how far out should we look from to judge the accuracy of the predictions? A day, a week, a month?

  7. I don’t bother tracking Opinium because, while they no longer seem to massively downweight rising parties, they haven’t given any explanation as to why their Scottish crossbreak no longer looks ludicrous.

    FWIW SNP 45% : Lab 25% : Con 17% : LD 5% : UKIP 3% : Grn 5%.

  8. RAF

    Prima facie it does. And if that does occur it could be the clincher.

    But one poll doesn’t establish anything.

    I think it was Jim Jam who postulated a LIFO / FILO breakup , if it ever happened.

  9. @crossbat

    Thanks for letting me know about Any Questions. I missed it but sounds like it’s worth a listen.

    “I have a feeling that the word has gone out from Tory HQ to call off the dogs. But will the right wing press listen, I wonder?”

    That is almost certainly the case. Will the press listen? The right-wing press aren’t always kind to Tories. When Trident was in the news, the Telegraph laid into DC about dodging the 2% GDP of defence issue (like I needed reminding..). The Mail will just print whatever gets people sharing links to the Mail regardless of whether it helps the Tories, such as the story about Ed’s girlfriends.

    Re: Opinium

    Polls really are in flux. Either the pollsters are diverging, OR we’re just seeing a more sensitive electorate, an electorate which is paying more attention to the election and responding to Events and Big Moments more than pre-short campaign.

  10. Good Evening from a cool but sunny Bournemouth.
    Maybe the attacks on Miliband do work? His numbers are back at pre debate levels.

    I THINK that negative campaigning does work, although people say they do not like it.

  11. Re: EU polls

    In case people didn’t already know, there’s a nice wiki page with all the polls:

  12. That’s a good poll for the Tories. There is a long way to go and neither of the main parties seem capable of taking a sustained lead.

  13. @JimJam

    Ben Foley who has not been around for a while

    Perhaps because he is himself a candidate in the election and so probably doesn’t have time to hang around here.

  14. @Exileinyork -“Watching the women’s boats race. The crew on the right is falling behind – a metaphor for the election?”

    No. It just means you are watching the race from the wrong end.

  15. Good evening all from sunny Surrey.

    A better poll for the Tories.

    Are there any other polls tonight apart from YG?

  16. Re. Opinium.

    It’s interesting that the best tory polls never show them more than 2 points ahead these days…by contrast labour polls have shown bigger leads…

    I would suggest that labour are a nose in front. I haven’t seen any real evidence, other than mere superstition, that the result will be all that different.

  17. ukip with a very low score thus boosting tory numbers – not seen in other polls – so might be MOE/methodology quirk.

    Anymore polls tonight?

  18. @tim n
    “Are there any other polls tonight apart from YG?”

    I’ve been expecting an Ipsos MORI poll for about two days now…

  19. What’s all this talk of ‘right’ and ‘left’ boats?

    They’re called ‘stroke side’ and ‘bow side’ if you don’t mind.


    I have as well.

  21. Funty, as our rowing correspondent i bow to you. Mrs BH asked me to congratulate ‘your Midlands football correspondent’ on AV’s victory in North London.

  22. The Opinium poll shows the Blue kippers going back to the Tories. The Tories must get them back , if they are to remain the largest party

  23. Opinium’s party weightings are interesting

    Here are their party weightings this week vs the weightings they used in the beginning of Feb when they started doing their party propensity weightings

    Con 26 23.7 (+2.3)
    Lab 24.8 23.8 (+1)
    LD 4.5 3.5 (+1)
    UKIP 8.4 9.8 (-1.4)
    Green 6.5 7.1 (-.6)
    Other 3 2.4 (+.6)
    Floating 17.8 18.2 (-.4)
    Unlikely 9 11.5 (-2.5)

    You can see the Conservatives are now weighted higher than Labour. That ‘other’ is probably mostly SNP and the new higher weight probably accounts for OldNat noting that the SNP cross break is becoming more reliable.

    But I wonder if they have now introduced some kind of lag by weighting to previous polls – so the Conservatives have a higher weight because they were doing better a few weeks ago?

    They certainly had to upweight their conservative voters this week and downweight their labour voters (and UKIP voters)

    They describe this weighting as follows

    “With party propensity, we ask voters how they feel about all of the parties and, from their answers, put together a unique picture of each voter’s view of the whole spectrum of parties. From this we can divide the electorate into groups based on which combination of parties they would or would not consider voting for.

    To decide how big or small each of these segments should be in our sample we have tried to make the process of creating weighting targets as organic as possible. This is based on a combination of how respondents voted in the European Parliament elections last year and a rolling average of the results of this segmentation during our trials over the last few months.

    Taking January 2015 as our base therefore, the targets are updated week to week by a rolling average of recent polls. This makes sure that, while our samples will remain politically balanced, they are still able to take account of actual changes in the electorate’s underlying views. If party X suddenly catches fire or completely nosedives then we won’t be hamstrung in reporting this by how popular they were in December”

    Still not really convinced about this party weighting….it doesn’t seem to work for Populus, jury still open for Opinium.

  24. I’m surprised that UKPR has not picked up on this Or doesn’t appear to have done so):

  25. Not that the “Catholic vote” in Scotland has been solidly Labour for some time, but this certainly won’t help LiS..

    The Roman Catholic Church in Scotland dealt a blow to the Labour Party when its bishops came out against Trident as churches urged their members to make use of their vote.

    In a letter to all 500 Catholic churches in Scotland parishioners will be urged to vote as it says that “nuclear weapons represent a grave threat to the human family”.

  26. @johnogilvie1615

    The tactical voting in Scotland has been discussed on here. The discussion involved references to that Kellner article and another by Curtice (“what Scotland thinks”) on the same subject.

  27. Richard

    Thanks for the detail

    Another way to read that would be “We adjust our targets every week to make sure that our numbers aren’t much different from other pollsters – potentially leaving us looking really foolish.”

  28. @ Funty

    Before leaving the rowing theme, perhaps the best parallel to the election is a crew giving everything they have got while not looking where they are going, and hoping the guy with the loud voice in the stern is shouting the right instructions. :-)

    BTW my daughter who rows at bow will make my life a misery if she ever sees my right side/left side post.

  29. The fieldwork for this Opinium poll is older than last night’s YouGov and yesterday’s Populus I believe.

  30. @graham

    The Opinium fieldwork was done at the same time as the Populus and Survation polls that gave the Tories 31%, and the YouGov poll which gave them 35%

  31. On the Opinium poll.

    The t test showed the change … When will be some care shown to false negatives … It is in the T test, but people don’t do it …

    Anyway, as I said last night: no significant (for Unicorn: there is a huge difference between statistical and non-statistical significance. Unfortunately, stats textbooks leave this out or have a paragraph on it (jolly nice when the book is 500 pages long). Did you look at the magnitude of the change and the power effect when you did your test? You may want to look at the debate about the American Economic Review on this and the 19 question.) change for months.

    There is a momentum for Labour or rather stall for the Conservatives, but that’s not a significant change.

  32. I see Andy Murray has got married in Dunblane Cathedral.Apparently in Scotland it is being treated as the wedding of the year. Find it strange that a couple that have been ‘living in sin’ for over six years end up having a white church wedding. It seems rather hypocritical and it surprises me that the clergy agree to carry out such services the effect of which is to bring the church into disrepute.

  33. Graham
    It may surprise you but this has never been an issue in Calvinist churches as far as I know. Though they may have been asked rather more forcefully to repent in the past, they would not have been denied marriage.

  34. In today’s on-line Guardian

    “The Conservative party chairman, Grant Shapps, is facing an electoral challenge from a candidate called Michael Green, the name Shapps used as an alter ego to promote get-rich-quick products.”

  35. Graham

    This is one of these happy occasions when “I agree with Barney”.

    Additionally, Ministers in a Presbyterian Church are not “clergy”.

  36. New Poll from Populus out; on twitter. Labour two points ahead. Mad

  37. Chris is that not yesterday’s Populus?

  38. GRAHAM,
    Most of us live in sin, I think.

  39. WES.
    Just tweeted, so not sure

  40. YouGov: Con & Lab tied tonight

  41. GRAHAM

    Any polling data on the issue you’ve raised?

  42. Compare leader ratings in last Survation poll & new Opinion poll. Former very good for Mili. Opinion (as normal) very good for Cameron.

    I suspect Survation & another recent poll were dodgy.

  43. Barney
    Thank you for that. As someone from a nonconformist background myself I am surprised by what you say.Is it likely that a statement of repentance will have formed part of today’s service?
    It is, of course, an issue that has arisen before. I think of William & Kate for instance. The Archbishop of Canterbury – then Rowan Williams – had a golden opportunity to make a moral point by declining to marry them. I have no objection per se to couples ‘living in sin’ if that is what they want to do , but I do object to the hypocrisy inherent in expecting ‘the penny and the bun’. As a matter of principle I have always declined to attend invitations to church weddings when the couple concerned have been cohabiting. Civil ceremonies are different matter entirely.

  44. Just pointed out on Political Betting “The usual ABC1: C2DE ratio in a sample is c. 55:45

    Survation used 40:60.

    Lots more poor people in the sample = better than usual result for Ed/Labour.”

    I have not checked this but if true it casts a lot of doubt on last Survation poll.

  45. @Laszlo

    I have a simple question for you.

    You are presented with a batch of polls and you are asked whether there is reliable evidence that Party A is ahead of Party B.

    What would you offer and accept as a criterion for making this decision?

    In the batch at the top of this post you question whether Labour are ahead of the Tories. I assume (perhaps with some risk) that you would not quibble about (say) Labour being ahead of the Greens. If so, then somewhere between the two there is a test of some kind that permits you to draw the line.

    For you, what is the formal process that can be used to decide the matter one way or the other?

  46. GRAHAM.
    A link to Political History.
    The lives of the Labour Party leadership in the 1960’s were strange. Gaitskell, Jenkins and Crosland lived ‘bohemian lives’ My Dad, being a strict Irish Catholic would have been appalled, and these men were heroes of his.

    Giles Radice’s book is reminding me of it all.

  47. Chrislane
    Well I would agree with your Dad.!

  48. That YouGov statement isn’t my prediction – it’s the actual poll

  49. crossbat11 @ 6.07pm

    Nope the negative campaign has not even started yet.

    Majority of the discussions on TV by Conservatives do not involve the OTT negative attacks on Miliband.

    One poll on negative campaigning, has found that Labour is carrying out more negative campaigning than even the Conservatives.

    If you care to investigate the negative character assignations carried out by Labour party on Conservative leaders, since 2000, makes even the Conservatives look like angels.

    I believe this is payback time.

    Saying that it does move people away from politics, or towards UKIP, Green etc.

    And it muddies the polices, every one remembers the negativity, and not the polices.

  50. OLDNAT
    Not that the “Catholic vote” in Scotland ….

    Wilde on Little Nell would seem to apply.

1 6 7 8 9 10