This week’s YouGov/Sunday Times poll is out here. Topline figures are CON 32%, LAB 32%, LDEM 7%, UKIP 18%, GRN 6%. YouGov’s average for UKIP this week has been running at only been 14%, so the 18% here looks unusually high – it could be an effect of the the events in Paris, or could just be a random blip.

Part of the rest of the poll addressed the attack on Charlie Hebdo – of course, these figures need to be seen in that context and people’s opinions may very be different in circumstances that are not so emotionally charged (it’s an issue I’ve sometimes commented on about polling about the death penalty – people only commission polls on the death penalty when there is a particularly heinous murder in the news, so polls are always influenced by a particular event).

Looking at the polling, a strong majority of people think the press should be free to criticise, mock and ridicule religion, but even in the current context a sizeable minority disagree. Around a quarter of people think the media should not be allowed to mock or ridicule religious beliefs or figures, 18% think the media should not even be allowed to criticise or question religion. More specifically, 69% of people think it was acceptable for Charlie Hebdo to publish cartoons of the Prophet Mohammed, 14% unacceptable. In the aftermath of the attack, 63% think that other newspapers should have reprinted the cartoons, 71% that the media in general have an obligation to show controversial items that might offend people if they are newsworthy.

Moving back onto party politics, YouGov asked about the two issues that dominated the first few days of campaigning last week – the economy and the NHS – along with expectations and preferences for the result.

A majority of people (58%) think that the pledges and promises that Labour have made mean they would end up having to increase taxes on people like them. However, people feel almost the same about the pledges and promises made by the Conservative party – 51% think they would end up having to increase taxes for people like them. Overall 37% think George Osborne has been a good Chancellor, 44% a bad one – a net rating of minus 7. This actually compares relatively well to people’s recollections of past Chancellors – Alistair Darling scores minus 19, Gordon Brown minus 18, Ken Clarke minus 8 and minus 19 for Norman Lamont.

Labour maintain their normal lead on the party most trusted to deliver NHS services – 31% would trust a Labour government under Ed Miliband more, 22% a Conservative government under David Cameron (there was a ComRes poll late last year that showed David Cameron more trusted than Ed Miliband on the NHS, which caused some comment. I think that’s probably just a salutory lesson of not paying too much attention to single polls with unusual results – the overwhelming majority of polls on the NHS show Labour are more trusted on it even if you do mention David Cameron and Ed Miliband in the question.

Asked about their own experience of GP services, 15% say their local GP service has got better, 34% worse, 40% that is has stayed about the same. 49% of people say they are normally able to get an appointment when they need one, 36% that they are often unable to. 8% say they have had to go to A&E when they were unable to get a GP appointment. Long waits at Accident & Emergency are mostly blamed on people turning up with minor ailments, rather than funding shortages from this or the previous government. 54% blame people turning up with minor problems, 29% blame immigration and health tourism, 28% not enough social care and 27% lack of GP out of hours service.

Looking towards the next election people are split down on the middle on their preferences – 38% would prefer the Conservatives to have the most seats, 38% for Labour to have the most seats. 52% would like one of the parties to win an overall majority, 24% would prefer a hung Parliament. Asked what they think the result will actually be, 59% expect a hung Parliament, only 18% expect a majority government. The Conservatives are seen as slightly more likely than Labour to be the largest party, 42% to 35%. Asked a more detailed question about coalition preferences, Tory voters would prefer another deal with the Lib Dems to one with UKIP (48% to 37%). Labour voters would prefer a Lib Dem deal to one with the SNP or UKIP (42% Lib Dem, 29% SNP, 12% UKIP).


289 Responses to “YouGov/Sunday Times – CON 32, LAB 32, LD 7, UKIP 18, GRN 6”

1 3 4 5 6
  1. Jayblanc

    Thanks.

    So what you meant to write was something like. ” I should note that the model’s revert to mean operates most strongly on Scottish vote share.”

    You aren’t actually applying a strong revert to mean on the Scottish vote share – which would be a very different thing.

  2. @ JB

    Some useful clarification but I still don’t get it.

    You write as it reversion to mean is operating for the Tories in the model and ‘keeping their head above water’. If so, there must be some other influence in play to overcome this process and move the Tory VI *away* from the mean (down from 33% today to 32% in May).

    Perhaps what I don’t understand is what this other driver might be.

    You offer one or two hints:

    “…the Conservative VI has been coasting downwards on average for over a year,”

    So, are you suggesting that this trend might introduce sufficient momentum to reverse the polarity of the reversion-to-mean?

    Bear in mind that I have previously posted (e.g., on Nov 24) the result of regression analyses of all 2014 polls and for the Tories the best-fitting slope turned out to be a minuscule minus 0.00253 per day. So, the stable trend is for them to lose less than a single percentage point over an entire year. Are you suggesting that this tiny downward trend is sufficient in your model to nullify and reverse the normal regression effect?

    “.. economic confidence is falling, and Conservative VI rallies have been short lived”

    But you model doesn’t ‘know’ about these matters does it?

    I don’t think you can deny that even if you model does have the Tory VI benefiting from a small reversion-to-mean effect there is some other factor overpowering this and pushing the VI opposite direction – ultimately causing the drop from 33% now to 32% in May.

    If it’s not the snail’s pace of the current Tory VI decline, then what do you think might have produced this counterintuitive effect in the output of your model?

  3. @OllyT

    I can only find that quote in 4 places on the internet, where you’ve reposted it here and in lib-dem voice, the original Guardian comment and on the Spectator comments 3 days ago.

    Given that, and given that Dr Sami Alrabaa is basically someone who makes a living from demonising Islam, I reckon it’s probably just been made up.

  4. OllyT

    The legal right to freedom of speech does not allow libel.

  5. @Unicorn

    I think you forget that VI is a zero sum game. The conservatives can only make gains from someone else’s loss. Gains from revert to mean have to come from somewhere, and if other VIs are holding up because of momentum, there’s not a lot that the Conservatives can receive in benefit. It’s not only the momentum of the Conservatives that is considered, but the momentum of everyone else. Revert to mean is saving the Conservatives from being dragged down not by their own momentum, but by UKIPs.

  6. By the way folks, I had to tweak the leadership ratings charts yesterday to allow for the recent very low ratings:

    http://www.statgeek.co.uk/polling/approval-ratings/leadership-ratings/

    (You can see each region, or view each leader across all regions)

    Clegg has dropped to a net rating of -76, and Miliband is -75 in Scotland, so that perhaps gives an indication of how unpopular one is to the other in the minds of some Scots.

    Caveats of Con + UKIP in Scotland being in the realms of 29%, which seems unlikely, so the leadership ratings perhaps have been skewed. Not sure about Cameron…his higher ratings in Scotland have been constant, despite his party’s VI in Scotland consistently dropping over the past six months.

  7. @Statgeek

    I just think it shows how leadership ratings are irrelevant to parliamentary VI. Just because they think Cameron is doing a good job as a Conservative, doesn’t mean they’re going to vote for a Conservative.

  8. JAYBLANC

    Reversion figures for the SNP and other parties in Scotland may be different to expected because of the demonstrable increase in support for independence over the 3 year campaign. If pollsters keep asking the “Would you vote Yes or No” question and/or DevoMax/FFA questions in Scottish polls, those results may provide useful indications.

    STATGEEK
    Not sure about Cameron…his higher ratings in Scotland have been constant, despite his party’s VI in Scotland consistently dropping over the past six months.

    Just a thought, but that might just have something to do with the 2/3rds of the Scottish electorate who support FFA feeling that Lab and LDs are back pedalling on the issue whilst the Cons seem more open to it.

  9. @STATGEEK

    perhaps being called Cameron helps!

    Personally I think that the polling since Christmas will be well received by the Tory hierarchy. The UKIP boost, in these dreadful circumstances, was bound to happen but will probably be short lived. I would expect a small but significant drift back from UKIP to the Tories by the election date, some of it in the polling booth. There is then the shy Tories phenomenon which reliably adds about one or two percentage points. We could thus be looking at a Conservative vote share of 35% or36% which, with the expected SNP gains in Scotland, leave Labour well behind.

    All imo of course and dependent on “events”.

  10. @ Jayblanc

    Very important point about the zero sum nature of VI.

    It also means that the VI is not made up by the sum of the party preferences, but the total is distributed to the parties (hence the modification of the raw figures by various factors). From this it follows that the unit of analysis is the VI, and the analysis of preferences to individual parties are highly problematic as these are derivatives of the distribution.

  11. @Jayblanc

    Voting is only a zero sum game when you include the non-voters. This group are usually ignored in opinion poll headline figures. If you simply look at the headlines an increase in Tory vote share doesn’t need to come from Labour, LibDem, UKIP etc, it can simply come from the non-voter pile.

  12. @Jayblanc

    Indeed. However, there will come a point with every voter when they look at a given party’s leader and say “He’s a complete [insert uncomplimentary noun]”, and move their vote to another party. How many UKIP faithful would change their vote to Con if Thatcher rose from the grave, or how much VI would Labour get if Blair returned? (we might argue they would have negative effects, but in truth, I think they would get a 2nd honeymoon, if only to lead the preferred party to another victory)

    For NC and EM to get -75 in ratings means that many in their own party do not favour them. We can’t drill down to the party level of the regional CBs, but if you look at each party at the national level we see:

    Cameron: +89
    Miliband: +8
    Clegg: +31

    So Miliband is not looking too popular, or it might be that all the Conservative ‘No folk’ have gone, leaving only the Cameron faithful…while Miliband still retains voters that are less well disposed to him.

    I think what I’m saying is that when it comes to the election, if there are voters that vote more because of the party leader, rather than the party, they might well go with Cameron before the other two, while during the non-campaign seasons they might identify with a party other than Conservative.

    I don’t suggest this is likely, rather, I’m throwing an idea out for you all to bash around. :))

  13. STATGEEK

    I’ve bashed it around-and I quite like it :-)

  14. @JB

    Hmmm…so the contra-reversion downward pressure on the Tory VI comes from what is happening to Ukip?

    But as you pointed out to @Oldnat above, in your model Ukip themselves are subject to strong reversion effects and their VI is projected to fall from 15% today down to 14% in the current output of your model. As far as I can see this should free up VI so that the Tory’s reversion process can run to plan.

    In most accounts of Ukip, Swingback/reversion analysts assume that the Tories will be the main beneficiary – even if only by a modest margin. Your model seems to embody a rather different kind of churn: a process in which Ukip loses VI whilst at the same time dragging down the Tories so that the support moves instead to a third party.

    Also, you have not previously described the process by which parties compete for VI, nor given any indication of the algorithms the model uses to adjudicate between conflicting claims of this kind. It is proving very difficult to get to the bottom of how your model works and to form any judgement about the degree of credence to be given to its projections.

  15. @ RMJ1 if the Conservatives got to 35/36% that would be pretty much what they got in 2010; for a Government to retain its initial level of support is very unusual – but some degree of precedent is going to be broken in 2015 – the question is if this is the one

  16. @Unicorn

    It’s all the different parties, all having their momentum, all of which needs to balance out a single Vote Share.This isn’t tracking the momentum of multiple tennis balls, but the momentum of a single tennis ball moving through nth dimensional space. The directions of movement all need to have their momentum moderated by how much is available for them to take against every other directions momentum. UKIP are a drag on Conservative, LibDems claw back some protest votes from UKIP, Labour loses most of its lead… You can come up with simple narratives, but ‘Churn’ is not really a simple matter of easy to identify herds of swing voters moving around, but a hugely chaotic system of millions making individual choices.

  17. To explain better…

    The state of an election is not individual vote shares in isolation. The state of an election is described as a position in a potential space of all vote shares within a zero sum game. You’re tracking *that* location.

    A presidential election between two candidates can be seen as tracking a tennis ball rolling back and forth between the two. This is also mainly the same with a parliament dominated between two parties. The momentum is either in one direction, or the other. Let’s call them north and south.

    Add in a third party. Now you can go north, south or up. Or north-up. Or south-up. Or north-south.

    Add in a fourth party. Now you can go north, south, up or out. Or north-up, south-up, up-and-out, north-south-and-up…

  18. But having said that, there’s also a really super-easy way to approximate what the zero-sum movement results in, taking component momentum pulls. I want to see if anyone can independently realise what that is.

  19. In furtherance to what Steve said – in a hostage situation (or most situations where firearms officers are needed) they aren’t shooting to kill, they’re shooting to neutralise the danger in the situation. The chest is an ideal target because:

    a) it’s massive
    b) if you hit someone anywhere in the torso it’s likely to knock them off balance at the very least as you’re delivering a huge amount of force to their centre of gravity (this applies just as well to suspects wearing body armour)
    c) the torso is deep and full of organs and bones so the bullet is less likely to come out of the other side and hit a hostage or bystander.

  20. Here’s a hint…

    Sarah, Jan, Tina and Yovik all want a portion of pie. But they each want to stick to their recommended diet, so have a set amount of pie they want.

    Sarah wants the least, only wanting 20%.
    Jan wants 30% of the pie.
    Tina also wants 30% of the pie.
    But Yovik is running a marathon today, so wants 40% of the pie.

    How do you fairly distribute the pie?

  21. Peter Kellner’s latest piece:

    https://yougov.co.uk/news/2015/01/12/use-and-abuse-opinion-polls/

    “Cameron’s response – and Miliband’s and Clegg’s – should not have been to lose their nerve,”

    Interesting comment. Mrs. Statgeek, who is generally not inclined towards SNP-esc politics asked the other day why the SNP was being omitted from the national debates, but when the proverbial was going to hit the fan, the three UK leaders felt that they had to ‘get involved’.

    Or in other words, if Scotland had to make the decision alone, why did they appear at the eleventh hour?

    I told here that as with most things in politics they were being hypocrites. They tried to maintain a ‘not getting involved stance’ when the polls were in their favour, and panicked when they were not.

    So that’s coming from a non-political person (Mrs Statgeek almost never brings up politics and goes noticeably blank and bored when I do…it’s a marmite thing). If the not-getting-involved rule in Scotland’s affairs of the referendum is easily broken, then why can’t we have the not-a-national-party rule easily broken (not that it’s a rule, rather, a handy little get-out clause for other parties’ followers to use)?

  22. jb

    What sort of pie is it?

    If it’s rhubarb I’d probably eat it myself.

  23. JB Reduce each by 1/6th

  24. Chris in Cardiff,

    The Tory vote share fell only 2% from 1979-1992. We also have cases in the post-war period of governing parties increasing their vote share in 1951, 1955, 1966, and October 1974. It’s not THAT unusual for governing parties to retain their vote share.

    On the other hand, given austerity and the rise of UKIP, it seems unlikely that the Tories could do so, and I think they’ll decline over the next few months as UKIP get put into the spotlight. I’m still predicting a 15-30 seat Labour majority, on about 34% of the vote.

  25. @Steve/Funtypippin

    Interesting observations and I have to say, notwithstanding Neil A’s expert witness, I thought the police marksmen appeared to be attempting to shoot to disable the assailants rather than kill in that horrendous Woolwich incident. If so, I take my hat off to their professionalism and sang-froid in the face of a very immediate threat to their own lives. Extraordinary stuff.

    On to another subject, poor old Rupe’s having a bad Charlie Hebdo crisis, isn’t he? Rather like a few posters on UKPR are wont to do, he took to his twitter keyboard a bit too quickly, and inadvisedly, to say that “All Muslims” were responsible for the Paris killings, and now one of his news outlets, Fox News USA, is in hot water too. Watch this and I dare you not to be truly frightened at what comes out of a supposedly mainstream and major broadcasting news organisation:-

    https://www.youtube.com/watch?v=hwq8Bwr9wUg

    So apparently, I was totally imagining walking around the streets of Birmingham the other day because it is, according to Fox, a no-go area for non-Muslims.

    I don’t watch Sky News, but is it as bad as this? Surely not.

  26. Tell Sarah she can’t have any this time ?

  27. JB

    Yovik sounds like an immigrant to me, so don’t give him any.
    Alternatively, give it all to him, because he is the most enterprising and will create lots of pies to trickle down to the others. Mind you they’ll be very soggy.

  28. Give them all 25% each & say this is an egalitarian pie.

  29. Populus:

    CON 32
    LAB 37
    LD 10
    UKIP 13
    GRN 4

    It’s been a funny week for polls.

  30. @CB11

    http://en.wikipedia.org/wiki/Demography_of_Birmingham

    59.1% Christian
    14.3% Islam

    “Birmingham’s percentage of Christians is below the national, regional and county average. The percentage of Muslims in Birmingham is above the national, regional and county average, as are the Buddhist (0.3%) and Hindu (1.98%) percentages.”

    No mention of Birmingham being a ‘no-go area’ for non-Buddists. Maybe ‘Gouranga’ doesn’t get the Fox readership worried enough.

  31. 27th November was the last time Lab was on 37 and Con was on 32 also with Populus

  32. Thanks for the link to Peter Kellner’s article, STATGEEK.

    The big problem I find with it is that Peter doesn’t seem to admit the possibility that the YouGov poll might just have been wrong. I believe that was the real problem with it and that was actually the real reason why the three leaders shouldn’t have panicked. So, in that way, the polls do bear a lot of blame. (Or at least YouGov does!)

  33. mr N

    You’ve spelt POPULUS correctly !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    Let this be an example to others – it CAN be done folks.

    Re the polls, if there is a parliamentary majority’s worth of voters who want to see the Tories govern alone for the next five years, taking into account the last three or four years of polling and also thinking back from 1992 onwards, then I will be as surprised as a very, VERY surprised person.

    My instinct also tells me that the Tory election campaign may well be more of a turn-off than a turn-on to the majority of undecided voters.

    Of course that will be true of every party in terms of the electorate as a whole as none are supported by more than a quarter of them anyway.

    Any news on the pie?

  34. Yep – I agree with @ RosieandDaisy

    16.7% for Sarah, 25% each for Jan and Tina and the remaining 33.3% for Yovik. This is based on the argument that the collective demand is for 120% of supply, so each gets scaled back by a constant fraction to offer limited shares of the limited resource.

    However, this assumes that the pie-consumers’ individual desires are all that count in making the allocation. The allocations would be different I f you assume that – separate from this – there is a fundamental right to equal shares and that this second influence pulls back against simple rescaling as above.

    So, enlighten us. What precise bearing does all this have on VI-sharing in your model?

  35. 20% for Sarah, 26.7% for the other 3.

    The initial presumption should be for equality. When Sarah doesn’t take her full share, the balance is shared between the others.

    Otherwise, the more you want the more you get. The share out should not be influenced by the extent of someone’s greed.

  36. Looks like a classic case of swingback to the governing party, with the Lib Dems on 10% and rising.

  37. @CB11

    http://www.bbc.co.uk/news/uk-england-30773297

    It seems it was just a violent disturbance of the atmosphere with strong winds and usually rain, thunder, lightning, or snow in a cup from which tea is drunk.

  38. Scottish sub sample:
    SNP: 41% Labour: 25%

    One of the best Populus cross breaks for the SNP as far as I know. That’s a surprise considering the Labour lead on the headlines.

  39. @Statgeek

    “No mention of Birmingham being a ‘no-go area’ for non-Buddists. Maybe ‘Gouranga’ doesn’t get the Fox readership worried enough.”

    :-)

    @MrNameless

    Thanks for beating us all, yet again, to that Populus poll. As for the poll itself, as the delightful female Olympic boxer, Nicola Adams, would say, “this has really made my day”.

    By the way, big coverage in the Guardian today of Coppard’s campaign to unseat Clegg in Hallam. There’s reference to his campaign manager being a young man called Harry who works in a bicycle co-operative by the day.

    Harry Nameless, by any chance??

  40. On a totally unrelated story British nurse Pauline Cafferkey, who contracted Ebola in Sierra Leone is showing signs of improvement and is no longer critical.

    A little good news for today

    If you want a bit more relief from misery might I suggest #foxnewsfacts

  41. CALUM

    As far as I can see the Lab from Others gain in the Populus Poll was mainly Greens.

  42. ROSIEANDDAISIE
    jb
    What sort of pie is it?
    If it’s rhubarb I’d probably eat it myself.

    -So we now know who ate all the pies!

  43. UNISORN

    @”What precise bearing does all this have on VI-sharing in your model?”

    I was going to ask that.

    The electorate don’t say “I would like x Party to have y%..

    But I am probably missing the whole point of JB’s little puzzle.

  44. Interesting that some people regarding Yovik as being greedy. Failure in skim reading the problem, or underlying reflection of not truly understanding fair allocation of resources. If someone’s recommended diet is higher, for the purposes of running a marathon, is it really fair to give them an ‘equal’ slice of the pie to someone who wanted less?

    But anyway… Replace ‘want a specified slice of the pie’ with ‘have a momentum that is heading towards a certain percentage of VI’.

  45. in birmingham muslim births outnumber christian births so its only a matter of time really, saying its a “no go” area now though is silly

    (as is forgetting non religious births!)

  46. @R&D

    Perhaps one day the whole populus will learn to spell the Populous pollster correctly. The ranks of those celebrating the day will be populace indeed.

  47. @CrossBat11 no Sky News is nothing like that; seems always to be a little similar to Channel 4 news in terms of being slightly less reverent than the Beeb but not showing any obvious partisan bias. I’ve certainly been given a fairly easy ride when giving quite anti Government interviews in the past!

  48. Happy New Year to everyone.

    Been watching but not commenting ’til now.

    @Crossbat11

    I meant to e-mail about the cup draw, a large number of Bradford fans were desperate to be drawn against the Villa again but we got Chelsea away instead, that’s if we beat Millwall in the replay! We’ll take it though.

    Assume, like myself, you’re happy with double figures for the LD’s, for once?

  49. Re the polls, if there is a parliamentary majority’s worth of voters who want to see the Tories govern alone for the next five years, taking into account the last three or four years of polling and also thinking back from 1992 onwards, then I will be as surprised as a very, VERY surprised person.

    the tories’ failure has been the split right wing vote which is obvious to everyone. ukip is taking thousands upon thousands of old C1/C2 white voters who would have voted tory all through the 80s and early 90s.

    Dave’s metropolitan, patrician toff hug a hoodie stuff has turned off these natural conservative voters who, rightly or wrongly, believe Farage’s know-nothing (“i went to public school but was a trader who got a real job and didn’t ponce about at university”) persona to reflect their attitudes. Everything he says about multiculturalism, with the beer and fag in his hand is cultivated to appeal to these people.

    The right is split. Labour is not remotely popular. Kinnock got about 34% of the vote in 1992. Mili will go well to equal this in May. a higher portion of the electorate supported kinnock in ’92 than is likely to support mili in this year’s general election.

    The difference with ’92 is that the right was wholly united behind the tories. ukip didn’t exist and the bnp and nf got barely 12,000 votes between them, on a 77% turnout.

  50. I don’t know if this has been photo-shopped or what but apparently it’s Europe’s glorious leaders at the head of the Paris march

    https://twitter.com/ianbremmer/status/554633676785729536/photo/1

1 3 4 5 6