Thursday’s polls

Earlier on today we had a new poll from Panelbase, with topline figures of CON 31%, LAB 34%, LDEM 7%, UKIP 17%, GRN 4%. Full tables are here. I’m expecting quite a few polls out tonight – Survation’s regular poll for the Mirror is due, we have the daily YouGov poll for the Sun and there may well be others to boot.

I’ll be out this evening, but will do a round up of them all once I return. In the meantime feel free to discuss them here as they appear.

UPDATE: On top of the Panelbase poll we have three other GB voting intention polls, from ComRes, Survation and YouGov:

  • Survation’s latest poll for the Mirror has topline figures of CON 33%, LAB 29%, LDEM 10%, UKIP 18%, GRN 4% (tabs).
  • ComRes’s latest telephone poll for the Mail & ITV has toplines of CON 36%, LAB 32%, LAB 8%, UKIP 10%, GRN 5% (tabs)
  • YouGov for the Sun have topline figures of CON 33%, LAB 35%, LD 8%, UKIP 13%, GRN 5%

Looking at reactions on social media there are lots of people getting excited or dismayed by getting two Conservative four point leads in short succession. There is always a temptation to look for movement in the random variation of polls (especially when there has been so little genuine movement to get excited over!). However, there are four polls today – two Conservative leads, two Labour leads. The time to pay attention would be when the balance of the polls consistently starts showed one party or the other ahead, right now they still seem pretty evenly balanced.


451 Responses to “Thursday’s polls”

1 6 7 8 9 10
  1. FallenTower

    If that was the share and seat result then Ed M would probably be PM on 4% less of the vote. I’m sure that would be controversial.

  2. Colin
    I was responding to in-chilled comment by you.
    I of course will oblige you
    The recent Guadn headline “the day the polls turned” looks as foolish as …. Fill in your own favoirite fool.

  3. Farage ahead in South Thanet according to new Survation poll

    Survation. @Survation · 4m 4 minutes ago
    NEW POLL:
    South Thanet. Fieldwork April 22
    Farage/UKIP 39%
    Mackinlay/CON 30%
    Scobie/LAB 26%
    Driver/GRE 2%
    Timpson/LD 2%
    Others 1%

  4. ALEC

    No one understands me like you do Alec :-)

    I shall miss you when this is all over & the site has become English Polling Report .

  5. Un not in

  6. That Thanet poll IS interesting.

  7. Ukip also appear to be winning the ground war in South Thanet. Bit higher contact rate by post, much higher contact rates by more direct canvassing (phone, door knocking). 19% say they have met Farage personally?!?

    http://survation.com/wp-content/uploads/2015/04/Final-South-Thanet-Bown-Tables.pdf

  8. 3s 3 seconds ago

    YouGov/Sun poll tonight – Labour have a two-point lead: CON 33%, LAB 35%, LD 8%, UKIP 13%, GRN 5%

  9. MOE all day long, just about

  10. Lololol

  11. @Mikey – when you’re hot you’re hot

  12. Back to the old certainties.

    Al is right with the world of UKPR :-)

  13. I still think that there is a problem in asking the same people two or three times. Once they have answered one poll they are most unlikely to want to change their answer in subsequent polls. The first answer they give becomes their position and they will tend to defend it. They are thus not in the same position as the general public who have not yet been forced to make their minds up. It would be interesting if yougov published data on how many people actually answer differently the next time.

  14. colin put the champagne back on ice,false alarm

  15. Smithy. I’m impressed!

  16. 8,13,5 again

  17. I”ll give TND of the Sun his due – he’s reliable, even though he almost certainly doesn;t know it, Nobody warn him now, will you, please!

  18. Amazing that 4 polling companies have the Tories 4% ahead and then YG are on a daily basis having Labour slightly ahead, rather than ever having one poll with Con a few points ahead….

  19. Smithy
    I don’t know why we bother with all these polls. We could just ask you.
    Pray tell us your seat projections for May 7th.

  20. Panic over.

    Today is like yesterday.

    I’ll make a bold prediction – tomorrow will be the same as today.

  21. @ RMJ1

    “I still think that there is a problem in asking the same people two or three times. Once they have answered one poll they are most unlikely to want to change their answer in subsequent polls”

    Isn’t this what YouGov always do? Or is this a new methodology?

  22. FALLENTOWER
    “This gives the impression that any Tory advance is coming mainly in Tory-LibDem marginals.”

    Interesting. As the last poll (Survation) seemed to imply a potential Labour to Lib Dem switch the whole picture is getting very murky.

  23. @Colin

    Something for everyone.

    It has been a good week for the Tories though; no question about it. Their campaign has recovered and it’s all to play for.

    I think the Tories should be offering federalism. Well, I think all parties should (!).

  24. Looks like Farage has nailed Thanet. Wonder then if UKIP is firming up elsewhere as well?

  25. RMJ1 Re YouGov panel

    I posted earlier on this, I wonder if there are any doubts over at YG HQ.

    However, another poster did say they are still sampling from the whole panel for Nowcast purposes so presumably if there was a divergence they will see it.

  26. MIKEY
    Smithy there will be words if your YouGov prediction is wrong. I do agree if its released at 10:30 its usually a better one for Lab.

    That’s ones for you – does still worry me – that the approach you gov have taken may not be picking up recent trends like others maybe doing. No faith in Ashcroft or Panelbase and Ashcroft is well Ashcroft – Com res and Populus are the ones to watch for.

  27. @cloudspotter – that I’m afraid is changing by the hour

  28. “However, another poster did say they are still sampling from the whole panel for Nowcast purposes so presumably if there was a divergence they will see it.”

    The Nowcast has been updated

    Lab 277
    Cons 270

  29. CMJ

    “Today is like yesterday.”

    Works for weather forecasting about 70% of the time.

  30. Does anybody know how far these rather large differences might be explained by different ways of weighting ‘likelihood to vote’? As a guess I would expect this to be less influential in panel polls since their members will arguably be more interested in politics.

  31. I really do see the distinct possibility of an arrangement with UKIP and Lib Dems for the Conservatives to become the next government being a real possibility.

  32. @ AdamB,

    I can see you are depressed about a poll you don’t like, but there was a 1 point Tory lead three nights ago. What more do you want.

    There was an ICM Tory lead of six a while back, but then a Lab lead of 4 – surely this is going to be expected if the true picture is neck-and-neck.

    FYI – last time on election night I took as “the real” picture the average of ALL the polls but taking away the one MOST favourable for the Tories and the one MOST favourable to Labour and rounded up for Tories and down for Labour and got them spot on.

  33. Do YG use a panel. Are they polling the same people day after day?

  34. We have
    31 34 7 17 4 panelbase
    33 29 10 18 4 survation
    36 32 8 12 5 comres
    33 35 8 13 5 yougov

    So (with massively less sophisticated methods than the various statters on here):
    ~33 32 8 15 4
    for them all to be within moe?

    Still something for cons to cheer about then, they’re a point down…but lab are TWO points down.

    Yes I know my analysis is so crude as to be worthless :)

  35. Ofc the poll by [insert_here] is just another blatant example of how the [insert_here] media are trying to change the narrative by scaring people from voting [insert_here]!

    In fact, [insert_here] NEVER ASKED ME!!!, and therefore, their polls can’t be right anyway!!!!!1!!!!121!!!

  36. @catoswyn

    When distinct possibilities are real possibilities anything that can be seen is possible.

  37. catoswyn

    I agree.

    With a 9% lead well above MOE, and Con/Lab split within MOE it’s hard to see anything stopping Farage other than a serious own goal by UKIP.

  38. I wonder whether the polls are all over the place because the country is all over the place. Imagine if it’s about as far from a uniform national swing as it’s possible to get. Given that pollsters only interview about a thousand people per poll, wouldn’t that mean that there are quite high chances of the polls being out nearly all the time?

  39. SUFFOLK JASON
    Do YG use a panel. Are they polling the same people day after day?

    Well I’ve been polled twice this week, once last week. However though there may be some affect from using a panel and on line polling the point is that it is meant to be somehow representative of the country/electorate. So shifts that do or do not occur in a panel group are a valid measure of shifts in the whole.

    It could be that everyone says something different every day but that, like ants, we are all crawling around such a small area of choice that it appears to never change at all.

  40. @AdamB
    Amazing that 4 polling companies have the Tories 4% ahead and then YG are on a daily basis having Labour slightly ahead, rather than ever having one poll with Con a few points ahead…

    Someone is wrong. They could all be wrong! I guess poll of poll averages is the way to go in this situation.

  41. @ Wood

    “Yes I know my analysis is so crude as to be worthless :)”

    I rather like it. Single unifying theory. explains everything.

  42. Good Evening All.
    Professor Rallings seems to have confirmed my view that the Cons will take a vast number of LD seats as will SNP, while Labour will still do OK in thirty-or-so tory seats.

    For a time UKPR was ‘down’ and I missed it.

    Tories, I think, have edged into a VI lead.

  43. I am biased, but I don’t think any of the four point Tory leads pass the “eyeball” test.

    36% for the Tories (as per Opinium and Comres) is not implausible but it’s the very top of their likely range. They’d have to have a very good day to reach that level based on the other polling evidence we have.

    The other two (Ashcroft and Survation) have Labour on 30% and 29% respectively. No other polls through the campaign have had them this low. These polls also put Con on 34 and 33 respectively which IMO is an implausibly modest big two share.

    FWIW I think the same of Panelbase. 31% is just too low for the Tories. To my mind, only YG of today’s polls looks plausible as a result though a tie or one point either way seems even likelier.

  44. Feature of the past 10 days or so: apart from the one point leads, Lab with tendency for 2/3% margins; Con with larger margins.

    I dont see too much change from today’s polls.

  45. Chris,

    He hasn’t confirmed your view. He has agreed with it. Not quite the same thing.

  46. WOOD

    Absolutely

  47. Chris,

    But I agree. Today the Tories are slightly ahead. IMO.

  48. Labour nust be relieved with that Yougov 2 point lead.

    The poll of polls is tied 34% Labour/Con..Today 2 and 3 point labour leads 3 x 4 point Tory leads….looks like a slight Tory lead emerging this week…SNP scare stories working slightly?

  49. catoswyn

    “I really do see the distinct possibility of an arrangement with UKIP and Lib Dems for the Conservatives to become the next government being a real possibility.”

    The Guardian have an article on Tory “Wargames” for post election minority government / deals. According to this piece the Tories see the magic number as Con > 284 seats.

    Link below http://www.theg**rdian.com/politics/2015/apr/23/tory-election-war-games-under-way-to-reach-284-seats-for-minority-rule

  50. Suffolk Jason

    Answers are yes and no.

    Anthony once told us that you would be polled about once every two years, IIRC. Whether this has changed with this Jan /Feb business I don’t know.

1 6 7 8 9 10