Full tabs for the YouGov/Sun on Sunday poll are now up here. The slightly larger sample than usual was to make sure they had a good sub-sample of Sun readers, which the Sun used in yesterday’s analysis of the poll to look at what their own readers thought. The Express, however, has decided to report the Sun reader crossbreak as a national poll – obviously it wins the coveted UKPR crap media reporting of polls award. Just to be crystal clear UKIP are not in second place in this poll. The headline figures for this poll were CON 33%, LAB 34%, LDEM 8%, UKIP 15%. The figures quoted in the Express relate only to respondents who read the Sun.

160 Responses to “No, UKIP aren’t in second place”

1 2 3 4
  1. Valerie


    A lovely response & remark.

    Your experience brings to mind a similar incident -with a different outcome.
    My daughter was trying to negotiate a vehicle parked on the pavement. She steered her wheelchair into the road to do so, whereupon, a smartly dressed & well spoken young man insisted on “helping” her back onto the pavement.
    He then ripped her handbag from around her neck & walked off with it.

    Stereotypes eh !

  2. @Shevii

    “Also I tend to find these stories take longer than a weekend to affect polls anyway. In fact as Anthony keeps saying there are very few things that noticeably affect polls- the rest is a gradual drip drip.”

    I’m not sure that I agree with you there, although it may well be a point that is almost impossible to either disprove or prove (they tend to be very effective points of argument too!!). However, thinking about it rationally, wouldn’t an event like a top level political resignation and/or sensational by-election result have an immediate short-term polling effect rather than a drip-drip delayed one? I can’t quite get my mind around a voter who mulls over such things before making their mind up. Surely it’s much more knee-jerk and visceral than that, isn’t it?

    I seem to remember our old mate Cleggy having an immediate effect on the polls after his barnstorming first TV leaders debate in May 2010. Not much mulling and considered thought went on there, I don’t think. He moved opinion immediately and I would have thought UKIP’s win in Rochester might have accelerated their bandwagon in the initial post Rochester polls. Similarly, if Thornberry’s difficulties were going to damage Labour significantly, I would have thought we’d have seen some immediate colateral in the polls.

    Maybe tonight’s YouGov will tell us something different, but I can’t see any slow-burners being ignited here. Instant conflagration, if anything, more like.


    You don’t get to my age without accumulating a few stereotypes in your mind.
    But a couple of late teenage granddaughters make sure you don’t attach too much prejudice to them !

    I think it is prejudice rather than stereotyping which can be problematic.

    Mind you 5 years or so visiting UKPR doesn’t help in trying to shed stereotypes.

  4. Lord Ashcroft [email protected]
    Ashcroft National Poll, 21-23 November: CON 27%, LAB 32%, LDEM 7%, UKIP 18%, GRN 7%. Full details on @ConHome, 4pm.

  5. Two five point leads for Labour on the same day, so much for St George!

  6. Here we go.

    Tory meltdown by Christmas

  7. And no, that is not a new low for the Conservatives with Ashcroft, he had them on 25 on 1 Jun, and they have been on 27 with him at least 5 times this year, I think it just means his polls are more volatile than the others.

  8. Tabloids spin and both sides predicting the other’s meltdown…the silly season has definitely started.


    59% for Con and Lab combined…hmm. 9% for others? Hmm.

    I’ll continue to ‘hmm’ until another poll confirms it. :))

  9. Apparently the house owner turns out to be a “cage-fighting car dealer” (called ‘Dan’ to make headline-writing even easier) who has since given his ‘thoughts’ at some length to the Sun:


    which rather suggests that Thornberry was spot-on for once. Though I harbour a suspicion that the place and van actually belong to an Asian lesbian who votes Green, but the Sun didn’t feel its readers could cope, so Pressman has hired this gentleman.

    Thornberry has always given me the impression that she did have this slightly sniffy attitude towards most people. It may be a lawyerish thing (she’s not just a lawyer but the daughter of one and married to a judge) where most representatives of the working class you meet professionally are criminals, you may tend to stereotype. Of course most other people may feel the same thing about lawyers.

    This may have meant that the reaction from the media and Miliband’s people was “Oh, that’s typical Emily” and this informed their subsequent actions, which seem a little over-the-top to an outsider.

  10. Ashcroft’s poll suggests a slight drift from Conservative to UKIP, which is what people expected from the Rochester UKIP win, before the press got obsessed by Thornberry. I suspect that the Thornberry thing may not register in VI polls at all.

  11. @ Colin

    Talking of stereotypes- are they ever off those effing smart phones? :-)

    @ Crossbat

    Yes maybe you are right but I think it also is a case of reading Sunday papers for a semi delay and also straws breaking camels backs. So it’s hard to see that tweet alone changing anyone’s political opinion but combined with enough stories a fortnight before or a fortnight after might just.

    After that Ashcroft I wonder if I will be writing MOE all week…

  12. Hmmm. Populus and now Ashcroft. Something on the move, perhaps? Need YouGov tonight to either gainsay or endorse.

    Had a piece of nostalgia to chew on as I drove to work this morning. Outside an ambulance depot in Birmingham were gathered 150 or so hardy souls wrapped in thick high-vis jackets to ward off the sharp November morning frost. It was still dark and I could see the flames of the brazier lighting up a few ruddy faces who had gathered around its burning contents for some welcome warmth. Plenty of passing motorists parping their horns in support and of course, I’d happened upon some picketing ambulance drivers, striking against their meagre pay offer. Easily and casually demonised as grasping public sector workers and trade unionists, I felt more than a frisson of empathy with, and support for, them. One day, my life well depend on them. I saw no demons on these Birmingham streets this morning.

    What I did see, certainly in my eyes, was a group of poorly paid workers with families to support trying to fight for a better deal in the only way they knew how.. Like old times really and I’m with them, emotionally anyway, all the way. This is what real trade unionism is all about and I think the public are more well disposed to it than many on the right would care to admit.

  13. To whoever it was who hold to pull my head out the sand regarding the Tweetageddon and the influence of the press.

    I might not be working class, but most of my family are so perhaps I know a bit more about what might upset “normal working class” people than a bunch of patronising tabloid hacks.

  14. Ashcroft tables are here:


    SNP 5% (49% in Scotland), PC 1%, Another party 2%. Final adjustments knock a point of UKIP as usual but nothing else changes.

  15. Ashcroft poll….

    SNP…49% = 56 seats

    LAB…21% = 2 seats

    TORY..16% = 0

    LIB/DEM…8% = 1 seat

  16. @ Colin

    Shocking story by the way and my sympathies however long ago it was. I can’t begin to imagine the long term affect something like that could have on your daughter in terms of trusting any stranger.

  17. Sorry for the slightly garbled post above…rewriting posts and drinking black coffee do not go together. ;)

  18. Anthony,

    Can you please make that a real award, invite submissions, set up an award committee, and announce the winner of the eve of the PSA awards or similar.

    I’m sure there would be enough volunteers within the commentariat to crowd-fund a decent trophy!

  19. Allan C

    Ashcroft’s Scottish cross-breaks

    Based on 63 people, not selected (in that group) on a weighted basis. Doesn’t really reward or merit serious scrutiny.

    @Allan Christie

    “Papers are well aware of the need to reflect the views of their readers and so, for the tabloids at least, it makes sense not to damn UKIP entirely. I expect the approach from the Sun and Mail will be to suggest they understand why people might vote UKIP but ask readers to vote Tory to keep Miliband out/ get a referendum/ keep the economy on track, whilst the Express might go the full way and back UKIP”

    Absolutely agree with you. In 2011 the Scottish Sun came out with the headline (play it again salm) and even though it didn’t endorse the SNP’s flagship policy of independence in the editorial, it did take the softly softly approach to it presumably to keep hold of a large chunk of its readership.

  21. Still think Ahcrofts too small sample size to be meaningful.

    I think Richard’s breakdown of the strenght of commitment to given VIs is the most encouraging set of numbers for Labour in a few weeks.

    Allan C
    Ashcroft’s Scottish cross-breaks
    Based on 63 people, not selected (in that group) on a weighted basis. Doesn’t really reward or merit serious scrutiny

    That’s correct a very small sample with a big big health warning but the VI is still in line with what the current polling for Scotland is showing, although taking out the 52% STV poll it is at the slightly higher end.

    Yougov today had the SNP on 41%,, 18% ahead of Labour with a sample of 173 (I think?) so yes it does show the volatility of subs but they still show a trend.

  23. To actually talk about polling:

    The Ashcroft polls are quite up-and-down so I think it is safe to draw the conclusion that Labour has suffered no damage from the Rochester-related tomfoolery rather than a genuine boost. Likewise, the Populus poll continues to show a good Labour position relative to other pollsters.

    So it comes to Yougov. If Labour show a lead, then it would be consistent with most other polls having Labour with a nose ahead of the Conservatives. So nothing to see there.

    If it continues to show level-pegging or a small Conservative lead, then there must be a methodological factor which explains the inconsistency with the other pollsters. So which pollster would be correct? (I am guessing AW would say Yougov).

  24. Not a great start to the week for Cons, it must be said. However, Populus seem generally much better for Lab, and Lord A very volatile, so while this would fit with the assumed outcome of another by election loss, it is contrary to the weight of media coverage post R&S, when we had the usual ‘Labour is finished’ type of coverage.

  25. Christian Schmidt –

    The trophy would be a golden 3D pie chart, as there is really no excuse for 3D pie charts.

  26. A couple of points of interest in that Ashcroft poll. UKIP’s regiuonal VI:

    South East 15%
    Midlands 24%
    North England 24%
    Wales & South West 23%

    Net England (and Wales, presumably): 22%

    Scotland 2%

    Well over 20% in three regions. Is there scope for UKIP taking seats then? Certainly scope for blocking in marginals (Con, Lab or both?).

    We’re into the realms of Con / UKIP alliances with their total in England coming to 50%. That’ll be popular in Scotland, where the same total is 18%.

    Labour on a low 21% to the SNP’s 49%, means 55 seats to the latter, and three to the former, while the Lib Dems still cling to the one seat. Conservative wipeout. All hypothetical, but thought provoking.

  27. @Statgeek

    “Labour on a low 21% to the SNP’s 49%, means 55 seats to the latter, and three to the former, while the Lib Dems still cling to the one seat. Conservative wipeout. All hypothetical, but thought provoking.”

    Is there any reason why today’s Ashcroft poll merited a comment on the Scottish crossbreaks, but today’s Populus poll didn’t?

  28. Looks to me that the Labour and SNP VI will basically be a swap around from the 2010 GE figures. Perhaps not quite a SNP steamroller, but they should get the most Scottish seats out of it.

    Just read the article on the Yougov website about how UKIP (and SNP) are making UK politics more European. What irony. But I’ve long been wondering whether we are in a period where overall majorities are much rarer than normal. Problem is the parties here do seem very unused to forming alliances, as are the electorate. Would they come to dislike these, and get behind a strong candidate and thus hand that person an OM whatever his party? Perhaps PR will come along and make that impossible before long.

  29. Alec is right, but 2 polls in a day with Labour 5% ahead hasn’t been seen in a long while & 2 have more credibility than one, though they clearly aren’t conclusive. These polls are grim for the Tories and if they turn out to be accurate they would mean that that party has made up no ground whatsoever since January 1. Probably YouGov won’t show quite the same figures, but it would be a surprise if the Tories are still shown as being ahead when the figures are announced later. Certainly, if these figures are even within the MOE, let alone completely accurate, there will be great relief in Labour circles, despite the party’s own rather low share of the vote, and despite the poor Scottish crossbreak – but the sample size of that crossbreak is far too small to be reliable.

  30. I have never been in a trade union as I haven’t had an appropriate job but surely the time is right for a revival.

    Now if Labour came out with some pro-TU policies – not a return to the 70s but for example TU representation on company boards. I think it might help them with the working class – but are they too timid?

  31. One wider consequence of a rise in the SNP and to a lesser extent UKIP ( as it will be harder for them to win seats) is that the recent UK red North Blue South Constituency map might become more Yellow Red Blue, with few if any Red or Blue in the North and little but blue in the South but maybe some purple.

    So as well as no party with an overall majority we could see no Party having a national spread of representatives.

    Labour will still have a cluster of seats in London but I wonder if the Tories might over time become more interested in the South and London as a target for seats than the North, particularly if UKIP eclipse them out with the South East as the main rival to Labour..

    If UKIP really are at 15% in the South East and 25% elsewhere the Tories in third North of Watford we could well see questions asked about Labour and the Tories being National Parties.


  32. @Phil Haines


    SNP 49
    Lab 34
    Con 18
    LD 5
    UKIP 2
    Greens 2

  33. 23% expect to see Ed Miliband as PM. Is that because 77% won’t vote for him, as Ashcroft suggests?
    The Conservatives will be saying do you want EM or DC to run the country?

    Another 5% Labour lead, despite no good news for Labour is surprising.

    LD revival look like its gone.

  34. @Peter Cairns

    Don’t count your chickens just yet!

  35. COUPER

    That adds to 110%?

  36. Ashcroft also tweeted that his next set of marginals polls will be released later this week.

  37. Ashcroft’s England only figures (per Mike Smithson):

    Con 28
    Lab 34
    LD 6
    Ukip 22
    Green 8

  38. @Phil

    I was aware of Ashcroft’s poll (tabs, that is) but not the other one. If you’re looking for partisan reporting, it’s not here. I’ll happily report on any Scottish cross break, but don’t report on them all. It caught my eye, as Lab’s 21% was lower than most :))

  39. Apologies

    When I said “Tory meltdown by Christmas” , I meant mental meltdown not, probably, polling collapse. I think both main parties are in a state of near frenzy and a succession of non- knife edge polls one way or they other could cause excessive and/or unpredictable reactions.

    Despite the money and coverage, actual GE campaigns don’t fundamentally change voting intentions and before anyone mentions 1970 and 1992: In the 1966 -1970 parliament, Labour had trailed the Tories by mostly double digit margins since the devaluation in 1967. Labour did close the gap to 6 or 7 points by Spring 1970, but I think Wilson was sucker punched by Gallup giving Labour a 7 point in early summer (probably a rogue). There weren’t many polls in those days. In 1992, the Tories didnt make dramatic headway in the campaign, the polls were broadly neck and neck throughout but it was the pollsters that.got it wrong. The Tories were always going to win.

    So, there are just over 23 weeks to go. But this Parliament ends on the 30th March. Two weeks for Easter and then into the campaign proper. Therefore, realistically, only 18 weeks for the Tories to establish a convincing lead.
    Where were they 18 weeks ago? Where they are today, 5 points behind.

    No wonder nerves are fraying

  40. This post is not about dodgy polls or the unusually large recent Labour leads.

    I would be pleased to find a more timely occasion to post, but unfortunately relevant threads don’t come up all that often.

    Anyway, in the hope that there is someone out there who shares my interests, I wanted to share some thoughts about making sense of long-term polling trends. (If I am just reinventing the wheel, then apologies for this: there are hundreds of thousands of comments on UKPR and no obvious way to search back and check the issues that have been explored in the past.)

    Modellers and contributors have used a variety of approaches to predict polling figures and seat tallies for the election. For example, some models make use of marginal polling data and Stephen Fisher’s Swingback model makes predictions based on the assumption that month by month swingback will track the swingback we have seen over the last few GEs.

    What I’ve tried to do instead is to see what happens if we assume that recent polling trends continue as they have in the past. Labour’s VI has been going down, Ukip’s has been going up and so on. Where will each of them be on May 7 if none of the trends are disrupted?

    To follow this through, I have taken the 433 polls tabulated on the Wikipedia UK Polling page (covering all listed polls from 1st January, 2014), and put the various part VIs into (linear) regression equations with observed VI on the y-axis and date-code as the x-value. (In my formulae today’s date-code is zero, Jan 1st was -322 and Election day is +167). The respective best fitting equations are as follows:

    Conservatives: Predicted VI on Day D = 31.97 – 0.00253 x D
    Labour: Predicted VI = 33.78 – 0.0138 x D
    LibDems: Predicted VI = 7.37 – 0.0079 x D
    Ukip: Predicted VI = 15.86 + 0.0113 x D

    That is, the best fitting straight line confirms that the Tories are basically flatlining (over the period examined). The negative slopes for Labour and LibDem capture the fact that both are losing VI as time passes. In contrast, Ukip VI I obviously going up.
    If you insert D=0, you get the best estimate for today’s VI for the four parties. More interestingly, in you plug in D=167 (i.e., Election Day) you get the following VI projections:

    Conservatives: 31.55%
    Labour: 31.47%
    LibDems: 6.05%
    Ukip: 17.75%

    (Note: You may not get exactly these values if you repeat the calculation. I have used slope and intercept values to five decimal places in calculating these value – with spurious precision.)

    So, if past trends continue undisturbed, the Tories are projected to poll slightly ahead of Labour on election day. Interestingly, crossover occurs just 6 days before the election. OK it IS just a model!

    You can plug these figures into Swingometers and other models for deriving seat conversions. But basically the figures are not that different from the polling averages and projections that appear in the top right-hand area of the UKPR home page today.
    Now, obviously past trends are not going to be a perfect predictor of what happens in the election. But do we do worse by tracking what is happening this time round rather than (say) basing predictions a swingback data that go back decades, and to a time when fewer parties were in contention?

    Given the basic trends, it is interesting to speculate what events might occur to break the existing patterns. Last time round, the Leaders’ debates were a trend breaker. It needs something like that to shake things about a bit.

    Step back a little from by-elections, Twitter blunders and so on and you will see that what has been happening over the last year has been a steady Conservative/Labour VI convergence with no hint that the underlying process is going to change any time soon. The rate of convergence would have to show a sharp uptick for the Conservatives to participate in the next government.


    Notes: (1) To simplify the calculations I have made no adjustments for the sizes of the 433 individual polls. Ideally bigger polls should have had a heaver weighting but I didn’t have the energy to do that.

    (2) The starting date (Jan 1 2014) is clearly arbitrary, but the exercise could be repeated over any period of interest.

  41. I wonder if the Twitter Storm (#CameronMustGo at the top of the trend list for 3 days now) is having an effect?

  42. @Unicorn
    “Step back a little from by-elections, Twitter blunders and so on and you will see that what has been happening over the last year has been a steady Conservative/Labour VI convergence with no hint that the underlying process is going to change any time soon. The rate of convergence would have to show a sharp uptick for the Conservatives to participate in the next government.”

    Not so much a convergence more a Lab fall. Con has remained remarkably static all year. However, there is evidence that the Lab fall has bottomed out at about 32/33% and is starting to show anaemic signs of growth.

    I therefore disagree that a continuation of the existing trend would lead to a Tory majority of votes, although I do accept your premise that even if there is a swingback it is likely to be too modest to save the Tories. They appear to be pinning their hopes on the Press and the SNP.

  43. UNICORN.
    Many thanks for your brilliant post;

    Lib Dem figures are very interesting; maybe they will achieve double figures on May 7th/8th.

  44. SHEVII


    She is now an ordained priest-so I assume trust is default mode :-)

  45. HH

    @”. Is that because 77% won’t vote for him, as Ashcroft suggests?”

    I don’t think he does-and they certainly didn’t make it clear why they gave that answer.

    Ashcroft ponders whether it indicates a voting intent, but they could equally well have answered Q4 on the basis of their perceptions of other peoples voting intention.

    It would be nice to know on what basis they did answer that question.

  46. an excellent website:


    points out that of the last 47 poll this month:
    the tories have been ahead of labour in 9 (8 1% leads and 1 3%)
    there have been ties in 8
    labour have been ahead in the other 30 (12 1% leads; 8 2% leads and 5%)

    It’s a slight exaggeration to say the polls are tied. labour preserve a slight lead.

  47. Crossbat

    Very few people have sympathy with these strikers; they have been cosseted for years with the benefits that the public sector has bestowed upon them – it’s now time for the harsh reality of the real world.

    As far as the poll goes, it’s in line with [NewsUK’s] own findings. There is a lot of work to be done in [their] heartland areas to get potential kippers back. What [they ]won’t be doing is portraying DC as their mate Dave or hammering UKIP too much – the left wing social media brigade will do plenty of that for [them]

    There is however plenty of mileage in the figures – many people will not focus on the election until the weeks up until polling day are here, and [the papers] will present it as a straight choice between DC and EM – which is what it is. If you vote UKIP, [they will say] the net result of your decision could be to put EM in No.10 – you have to live with that decision and when your kids or whoever asks you why this guy is/became PM at all, you will have to tell them what you did.

  48. sorry labour have been ahead in the other 30 polls (12 1%; 8 2%; 5 3%, 2 4% and now 2 5% leads)

    when you follow the polls daily it’s easy to lose sight of this…labour I think are still about 2% ahead.

  49. Populus Scotland Crossbreak
    SNP 44 Lab 31 Con 16

  50. A look at the recent polls plotted by time shows a very slight widening in Labour’s favour.


1 2 3 4