Just the two regular polls in Sunday’s papers. The weekly Opinium poll for the Observer has topline figures of CON 34%, LAB 35%, LDEM 6%, UKIP 14%, GRN 6% (tabs), the weekly YouGov poll for the Sunday Times has figures of CON 34%, LAB 34%, LDEM 8%, UKIP 14%, GRN 5% (tabs). Both very much in line with the broader picture of Lab & Con almost neck and neck, Labour just a touch ahead.

YouGov asked whether people would consider voting for each of the main GB parties and their awareness of their policies. Of the two main parties, 40% would consider voting Conservative, 42% Labour – a slightly bigger pool for Labour but only just. The pool of potential voters for the other three substantial parties is pretty similar – 23% for the Lib Dems, 26% for UKIP, 25% for the Greens.

Asked about how aware of are of each party’s policies, 63% say they know a lot or a fair amount about Tory policies, compared to 59% for Labour, 45% for UKIP and 37% for the Lib Dems, 27% the Greens. Note how more people think they know about UKIP policies than those of the Lib Dems – a sign of how the Lib Dems have struggled to get a clear message out from within coalition.

YouGov also reasked the “protest party” question they asked about UKIP last year about the Greens. They found 15% of people think that the Greens are a serious party with workable policies, 56% a protest party for those unhappy with the main parties. These are very similar to the figures for UKIP, with UKIP 17% thought they were serious, 62% a protest party.

Moving onto other issues, 51% of people would support a ban on MPs having second jobs, but only 25% would support it were it to be offset by a higher salary. Asked about the current £67,000 salary for MPs and the appropriate level or reward for the sort of people they’d like to be MPs, 32% think the current salary is too much, 16% too little, 46% about right.

Finally there were some questions on defence and what sort of threats Britain should be prioritising. 16% of people think that Britain spends too much on defence, 49% too little, 20% about the right amount. By 52% to 18% people think we should be focusing resources on defending against threats from Islamist terrorism and insurgents, like Islamic State, rather than potential threats from states like Russia. 50% of people think that the West’s sanctions against Russia haven’t been strong enough, but on balance people are opposed to even the sending of British troops to help train and advise the Ukrainian army – 43% are opposed with only 36% support.

376 Responses to “Sunday polling round up”

1 6 7 8
  1. Allan Christie

    If the SNP don’t want a Tory government, their supporters here have a funny way of showing it.

  2. Assuming @Barnaby Marder is right about ICM, today’s polls are as follows:

    Populus: Con 32 Lab 34
    Ashcroft: Con 34 Lab 31
    YouGov: Con 35 Lab 32
    ICM: Con 34 Lab 35

    Overall average: Con 33.75 Lab 33.

    So oven on a good polling day for the Tories they’re still less than 1% ahead overall (and that with a likely YG outlier too).

  3. @Hawthorn

    The SNP don’t want a Tory government. On the whole. @AC does. He’s a Jacobite.

  4. RAF

    Not sure how you can put an average con VI of 33.75 and claim that 35 is outside a 95% MOE in the same post.

    Unless we’re using the alternative description of outlier today.

  5. Will events abroad have an impact? Interesting defence debate tonight. Greece surely taking EZ to the edge accusing Spain and Portugal of being “axis powers”.

  6. RAF @AC

    Ye Jacobites by name, lend an ear, lend an ear!
    Ye Jacobites by name, lend an ear,
    Ye Jacobites by name,
    Your fautes I will proclaim,
    Your doctrines I maun blame – you shall hear, you shall hear!
    Your doctrines I maun blame – you shall hear!

  7. @AC

    Chances of Conservatives replacing Labour as main opposition to SNP in Scotland?

    What do you think?

  8. @Pthiers
    Loved that George Carlin link – thank you

  9. Amber Star

    I dislike Ashcroft polls because he doesn’t state which firm did the polling for him. Does anybody even know for sure who did the Scottish constituency polling for him?

    As others have said, Populus did and are continuing to do Scottish polling for Lord A and interviewers identify themselves as ringing on behalf the company. Who Populus are working for is another matter. Certainly some interviews seem to follow the standard two question format, but others seem to contain items that suggest that they are working for another dubious client:


    the Labour Party.

    Would anybody like to comment on the strength & depth of their experience of Scottish polls & polling?

    How close were they to getting the referendum right?

    What was the VI of their Scottish poll that was done around the same time as their constituency polling?

    What was the overall VI of their most recent Scottish poll?

    Apparently the Scottish polling they did for the referendum was for another dodgy organisation called Better Together. So I’m sure that you’ll agree they are completely guilty by association.

  10. @Alan

    Of course I use the term “outlier” quite wrongly. It isn’t an outlier but towards the boundary of MoE.


    I’ve not made up my mind who i’m voting for yet.

    Tactical voting is in the air where I stay.

    Now about that birch polypore I removed from the tree in the grounds around me. Do I fry it or grill it?

  12. dubious client…another dodgy organisation…

  13. Unicorn

    I’ll try again with the name altered to protect the innocent:

    @WB The relevant models do make exactly the kind of drift adjustments that you mention. This is true of Electionforecast and May2015 and is probably true of the Guarding model as well. (I don’t know precisely how it works so I can’t say for certain.)

    It’s true of the Guardian model too, according to their methodology:


    For example if we have a poll for Pendle, but it was conducted in October 2014, we continue to use that poll as a starting point, but impose the swing in this seat which we register in our rolling average of polls since that date.

  14. TSEofPB

    “Oooh @LordAshcroft marginals polls out on Wednesday is both Scottish and English seats.”

  15. Don’t get too excited about YouGov. It’s a Monday. Wait until Wednesday before jumping to conclusions.

    I say this largely in jest, but there is a bit of a pattern in YG polls early in the week emerging, although I’d be perfectly happy to declare this simple statistical noise.

    Since the start of the year, the average Lab lead in all YG polls is 0.64%. Average Lab lead excluding Mon/Tues YG polls is 0.78%.

    Average Lab leads on YG Monday polls is 0.22%, and on Tuesdays 0.11%.

    Yougov have shown 6 Con leads this year so far. Three of them were on Mondays, and three on Tuesdays.

    Whether these results are statistically significant I know not, but some of you out there (@Unicorn – I’m thinking of you) may know.


  16. PS – do please check my numbers. This was done in my head, so could well be wrong.

  17. @Alan @RAF

    Of the data posted at 10.55 pm the only outlier is the Con VI of 32.

    Outlier upper and lower limits

    Con 35.4/32.9
    Lab 38.0/29.3

  18. OLDNAT

    Got the ole feet tapping. :-)


    Chances of Conservatives replacing Labour as main opposition to SNP in Scotland?
    What do you think?

    I’m thinking it would be a cheeky little intimation.

  19. Anthony Wells [email protected] 3m3 minutes ago
    New post: Monday’s polls http://bit.ly/1AxZZlk

  20. Alec

    YG polls for the ST – SNP below 40 on 5 of the 7 since January (71%).

    YG polls for the Sun – SNP below 40 on 5 of the 47 since January (11%).

    (Hadn’t noticed that, till I saw your post)

    That is George Carlin, not Bob Newhart. Both were great deconstructors of words, but Newhart would never have used language like that.

    I believe you’re right. Thanks for that. I don’t believe I had heard Carlin before, but the voice is right on other clips of his to which I have just been listening.

    I found the a low quality version of the lecture on Napster in the late 90s, when I was initially sceptical, but could find no other source. It was certainly out of sync with his 60s & 70s stuff. The link I gave is a better quality recording from someone who was obviously as mistaken as I was – https://dysonology.wordpress.com/2011/02/07/the-guy-quote-bob-newhart/

  22. @Oldnat


    YG polls – SNP above 40 on 44 of the 54 since January (81.5%).

    (It’s the way that you tell em)

  23. @CatmanJeff
    @Alan @RAF
    Of the data posted at 10.55 pm the only outlier is the Con VI of 32.
    Outlier upper and lower limits
    Con 35.4/32.9
    Lab 38.0/29.3

    In my defence I was referring specifically to the YG score which had the Tories at an unusually high 35 and Labour at an unusually low 32. Given Anthony’s (albeit still not updated) rolling average has the Tories at 32, today 35 is on the outer limits of Tory YG MoE.

  24. @Guymonde

    ..there is no real evidence that marginal swings are different to UNS, unless December is the start of a trend.

    There are obviously some morsels of evidence we need to pick through before we get to the bottom of this. Your calculations suggest that if UNS were the model of choice it would be immaterial whether we used the model unmodified or whether we replaced the polled constituency profiles with Ashcroft data.

    What started me off on this tack was noticing that if you ripen the May2015 calculations without using Ashcroft data, the Labour tallies jump up by 31 seats. So, whatever the situation might be with UNS, it is far from immaterial what happens in this particular model (May 2015). Using Ashcroft figures rather than their more regular Strong Transition calculations has the effect of boosting Tory seat tallies relative to Labour. It follows that for this particular model , though not necessarily for UNS the routine prospects may be understating Tory showings. It may be that other models (like UNS) don’t suffer from this weakness. But I did also point to the fact that three different models understated the Labour margin in the December batch – two reliably so. It is possible that this was a flash in the pan. Other batches need to be tested including the new lot arriving on Wednesday.

    It may turn out that this is just a criticism of the Strong Transition calculations used by May2015 (and also by Electoral Calculus). If so, it may be useful information to take into account in evaluation the seat projections generated by these models.

  25. @ Roger Mexico

    It makes no difference to me which organisations Populus do private polling for. If they only do private polling then how can we assess whether their track record is any good? In short, we can’t.

  26. @Amber Star

    Presumably Populus wouldn’t still be doing private polling for Labour in Scotland if their private polling for Better Together (run by some of the same people, e.g. Blair McDougall) had been complete rubbish.

1 6 7 8