Two new polls in the Sunday papers. This week’s YouGov/Sunday Times results are here – topline figures are CON 32%, LAB 32%, LDEM 7%, UKIP 16%, GRN 7%.

The YouGov poll also had questions on the end of the Parliament: the majority of people (56%) think that MPs have now started to concentrate on the election rather than concentrate on bringing in laws (7%). Not withstanding that there is little support for an early election – most people think the next election should still be in May 2015 as planned. The principle of having fixed term Parliaments has majority support (56% to 29%), though those who support it are split between agreeing with the current five year set up and preferring a fixed term election every four years. Asked about the fate of the coalition, 25% of people want it to end now (17%) or in the next few months (8%). 33% think it should continue up until the start of the formal campaign in April, while 28% want it to continue until polling day itself. The vast majority of Tory and Lib Dem voters want the coalition to continue until at least April.

MPs themselves continue to have a poor reputation. By 55% to 12% people think they are poor value for money and by 45% to 33% people think they are lazy rather than hardworking. 43% think that the reduction in Parliamentary business towards the end of the Parliament is just being used by MPs to do less work, rather than for constituency work.

Meanwhile a new ComRes poll in the Independent on Sunday has topline figures of CON 33%(+3), LAB 34%(nc), LDEM 8%(nc), UKIP 18%(-1), GRN 2%(-1). Changes are from their November online poll and tabs are here. A quick aside about that very low score for the Greens – as regular readers will recall, ComRes recently made a change to their methodology. They started including UKIP in the main prompt for voting intention, but also made some changes to their likelihood to vote weighting – this is not quite clear from the tables, but as far as I can tell from reverse engineering the tables in their online polls they now apply a more harsh turnout filter to UKIP and the Greens than for the Conservatives, Labour and Lib Dems. The end effect of the combined changes looks to me as if UKIP support is largely unchanged, but Green support will be decreased.

ComRes also asked people to put the parties on a left-right scale, with a surprising result. The average scores for Labour was 4.13, the Lib Dems 4.87, UKIP on 6.61 and the Conservatives 6.91 – so the Tories seen as more right wing than UKIP. This is in contrast to a similar exercise by YouGov earlier this year which found UKIP and the Conservatives the other way round. There are months between the polls, so opinion could simply have changed (especially since UKIP have been putting in an effort to appeal to Labour voters), but there were two significant methodological differences between the polls – YouGov asked people on a verbal scale, ComRes on a numerical scale and, probably more importantly, YouGov included a don’t know option and ComRes did not. In the YouGov poll over a quarter of people said don’t know to the questions (ordinary people don’t necessarily think of parties, policies and so on as being “right” or “left” wing!), so it could just be that lots of people said 5 when they weren’t offered don’t know as an option. That said both versions found people positioning the Conservatives and UKIP in a fairly similar place in the political spectrum, so probably not worth getting too excited over the difference.


157 Responses to “Sunday YouGov and ComRes polls”

1 2 3 4
  1. Mr Wells I see it says Projected Maj Labour 10 ,. How do you feel when everyone else say No overall and the bookies going 2/5

  2. Sorry my English was not very good “Says”

  3. Polly – it’s a pure UNS calculation on the poll average, no more, no less. It’s not a prediction of the result.

  4. When has UKIP come out and made a comment about white English people being scroungers?

  5. You say not worth getting too excited – I can already see ‘Tories more right win than UKIP!’ headlines being written….

  6. Polly – the concensus amongst the election predictors seems to be Cons 33, Lab 31 and LD 8 from current trends, and the two main parties at or just below 300 seats apiece. Put these numbers into the swingometer in the top right and that’s pretty much what you will get. Don’t expect them to be accurate down to the last seat though.

    I don’t think there’s much between 6.61 and 6.91, rather more between those and 4.13. Differences over the Europe question will cause some confusion here, I don’t think that issue can be so easily fitted into a left-right scale.

  7. For a prediction from “everyone else”, I suggest the betting markets on seat numbers are the best source, because most bookies offer the same odds above and below the same number, and the odds on offer will in large part reflect the views of the betting public. Current betting-based predictions are :

    Lab 287-289
    Con 281-285
    LDs 25-28
    SNP 25-27
    UKIP 6-8
    Greens 1

    That is very much minority government or three party coalition territory, with an outside possibility of a fresh election within a few months to try and sort out the mess.

    Because (a) Conservatives tend to have more spare money to wager and (b) those favouring a particular party tend to be more optimistic as to their chances, there is a view in some quarters that betting markets might reflect a very slightly rose-tinted view on the part of Conservative punters.

  8. @CMJ (FPT)

    Could be. While Scotland might be great Election night entertainment, I think that the results in England will make or break the government. They always did, and I doubt that will change.

    @All (repost)

    Applying the recent Scottish poll’s 2010 VI changes to a 2015 scenario (caveats etc.):

    http://t.co/GqzUumZRJB

  9. the markets severely underestimated labour ahead of the 2010 election. a trader at one of the spread betting firms reportedly lost many thousands because he was convinced labour would get about 225 seats.

    I happen to think Labour is being oversold as the market is assuming the SNP will get 25 seats. I don’t see the SNP getting anywhere near that. I think the tories are over bid at 281. Granted they gain 10 seats, they have to lose fewer than 30 seats to finish up on 282. That doesn’t even factor in any losses to UKIP.

  10. @ AW

    Re UNS I suspect I know the answer but nice to have it confirmed that your UNS above isn’t really taking into account movements in Scotland as it is too small an area and grouped along with Greens, Welsh Nats and Respect.

    Would it be fair to say that the current UNS doesn’t show any SNP gains or are there a few seats that might be showing as changing hands?

  11. @Statgeek

    While the English seats are the biggest part of a GE, I think a big SNP vote would commence a massive Constitutional change for sure.

    So long term the Scottish seats perhaps are more significant.

  12. Talking of forecasting the election, this is an interesting site

    http://electionforecast.co.uk/

    it claims to be updated daily and currently gives a tiny plurality to the Cons. I expect a lot of you already know about it, but it was new to me.

  13. Another Conservative Tendring District Councillor has just jumped ship and joined UKIP. That’s one MP and three councillors now…..

  14. Sorry, I should have added that now means that UKIP have more Councillors on Tendring than the LibDems.

  15. @ Statgeek

    I was interested by the steps you take to use the recent polling data to make GE seat projections for the Scottish constituencies. Your overall VI estimates may provide an accurate snapshot of what you acknowledge in your piece as a fluid environment (perhaps changing even today as Murphy gets down to work). However, as you say, it is the final step – seat projection – that seems most shaky. Given the wide variation in regional referendum balanced, it seems reasonable to assume that the SNP advances may have been equally varied across the different constituencies. Any shift from uniform swing could see marked changes in the type and number of seats that look likely to change hands.

    Do we have information on how the SNP VI gain is manifesting itself in different parts of the country?

  16. I find it rather interesting that a poll can show the public to believe that the Conservatives are to the right of UKIP. The UKIP posters who bombard the comments sections of all the national newspapers insist that Liblabcon are all the same and that Cameron has led the party far to the left with a socialist agenda. The rest of the population must therefore believe that the Conservatives are well to the right of UKIP. None of this seems to make any sense and shows how confused the public are at the moment.

  17. Thank you for your help Mr Wells and Keith

  18. RMJ1
    I think it shows how inadequate and outdated the terms ‘left’ and ‘right’ are.

  19. From previous thread

    “The other point about the poll is Ed Miliband scores only 8% as best leader behind Sturgeon (43) Cameron(19) and Farage(10).
    IMO this is a big problem for Labour with their ‘Vote Labour to kick the Tories out’ line. If the voters actually prefer Cameron to Miliband.”

    The thing with that is – is Miliband the cause of Lab’s problem or is Lab’s problem something else and Miliband’s personal ratings are based on his inability or unwillingness to deal with them?

    If this was 2005 and still in the height of history’s biggest credit bubble (1998-2008) with everyone high on cheap credit then I think his personal ratings would be fine – not Blair levels but okay.

    //

    As to Miliband relative to Cameron Lab have only got themselves to blame because although they have a bee in their bonnet over toffs from their Oxbridge days most people don’t so all the “Toff” stuff they’ve been pushing for years has *helped* Cameron.

    If they want to damage the Cons they should portray them as spivs not toffs.

  20. “I find it rather interesting that a poll can show the public to believe that the Conservatives are to the right of UKIP.”

    Multiple left/right axes.

    Pointless to even discuss right and left without at least taking two of those axes into account: economic and social.

    It’s like those online tests where the scores from being more right wing on social and more left wing on economic cancel out to “centrist” instead of sorting people into four categories: left/left, left/right, right/left, right/right.

    In normal times there might be more votes in a centre position because of weighting but there are no actual voters in the centre.

  21. Alex
    Welcome to Scotland! The message is that anyone who makes the slightest comment on economic rationality will be monstered. SNP projections did not countenance even the possibility of oil priced below $113 per bl. A Salmond said that even $140 was a very conservative estimate. The SNP are still firmly in denial portraying the fall as a short term fluctuation which is not a view af anyone in the industry.
    If Scotland had voted yes then they would be facing a deficit per head double the UK and equal to NHS spending in Scotland. (I’m going by figures from Prof Ronald MacDonald.)
    Oil is not a bonus for Scotland but the heart of the economy. The last figures I have seen said that most profit made in Scotland was made by the biggest 70 or so oil companies based in Aberdeen.

  22. @ Shevii

    That UNS is GB-wide, it doesn’t account for Scotland.

    If I plug the UKPR numbers and the latest YouGov Scottish poll into my UNS+Scotland sheet, I get these seat totals:

    LAB 304
    CON 250
    SNP 52
    LIB 20
    PC 3
    GRN 1
    OTH 2

    Of course this still isn’t perfect because if you compare marginal and national polling, the swing in marginals is close to the swing across GB (so much smaller than the swing in England and Wales). If you adjust for that (and the differential swings in Lib Dem seats) then you’re looking at LAB 280 CON 270 and the rest pretty much as above. Take a few UKIP gains off the Tories and that’s about what a nowcast (based on those polling numbers) should be.

  23. @MrJones

    “As to Miliband relative to Cameron Lab have only got themselves to blame because although they have a bee in their bonnet over toffs from their Oxbridge days most people don’t so all the “Toff” stuff they’ve been pushing for years has *helped* Cameron.

    If you’re talking about the pantomime portrayals that have emerged in one or two by-elections, then I think I’d agree with you, but I’m less sure than you appear to be that there hasn’t been real damage done to the Tories by their lingering image as a party of wealth and privilege that is incapable of understanding the problems and everyday lives of ordinary voters. This is why Mitchell’s problems and Lady Jenkin’s recent silly comments were so politically toxic because they played right into this image of aloofness and self-satisfied hauteur.

    It is a line of thought that almost defies the description counter-intuitive to think that any of this has “helped” Cameron.

  24. Crossbat
    But isn’t the Labour leadership also largely from a privileged background? e.g. Miliband, Balls etc went to Oxford and one of their female MPs recently sent a disparaging tweet about Rochester voters. I think to many voters there is little difference between Labour and Conservative in their remoteness from ordinary people.

  25. @ Pete B

    I think we can certainly agree that Electionforecast provide an interesting site, and it *is* updated very regularly. As they say under their FAQs:

    ‘..each day we drop in new data, and then a master script re-estimates the model, re-generates the report you are currently reading, and uploads it to this web site.”

    However, I do think they could be a little more informative about how the ‘master script’ does its job. How is Ashcroft constituency polling data processed in contrast with national polling crossbreaks? How are by election voting tallies handled by the script?

    Given the relative lack of historical polling information on Ukip, what basis do they have for projecting the steep drop in this party’s vote share down to 11.2% by the time of the General election? The model assumes a form of swingback or regression to the mean, but it is not clear how this should be applied to ‘new kids on the block’. In fact, given the present failure to match the size of historical swingback trends, one might expect to see some careful examination of this aspect of the model.

    Why does the ‘nowcast’ underestimate the current Conservative (rolling poll average) VI share by about 2-3% and overestimate the Lib Dem VI by about the same amount?

    It would also be interesting to see some discussion of how well the model does in anticipating results before they are published. On the day before the last batch of Ashcroft constituency polls, the Nowcast would have posted the VI percentages for the 5-6 main parties in each of these seats, providing an ideal glimpse of how the model is succeeding in capturing trends.

    On November, 27, they tweeted: “We will have a post soon comparing this latest round of @LordAshcroft polls to where our model had those seats before we saw the polls.” I have looked back regularly in the hope of reading that post, but I have to admit that I haven’t been able to find it.

    So, interesting in the sense that they take a position about what the voting and seat tallies would be in an election conducted right now, as well as projections about these figures for the General Election itself. But not as interesting as it would be if they were to shine more light on the workings of the model itself and in particular on whether some of its basic working assumptions are standing up as the evidence continues to accumulate.

  26. @Unicorn
    Thanks for the detailed and thoughtful reply. I notice that they give quite afew clues about their methods in the FAQs and the intro to the Seat Predictions tab, but you’re spot on in saying that it would be useful to see how they have performed compared to pollsters against real events including the recent by-elections.

  27. It looks more and more like the outcome of the next GE largely depends on the numbers of those former Labour voters switching to the SNP in Scotland and on those former Tory voters switching to UKIP in England.

  28. Both election forecast and Stephen Fisher’s models show the tories slightly ahead of seats than labour.

    The actual live exchange betting markets we have- Betfair and Sporting index- show a slight advantage to labour.

    The models are relying on swing back models which the market discounts…the market already factors in the “this time is different” flavour of next May’s election.

    I think both Sporting index and the betfair odds are quite sensible actually….labour may be short of 300 as a consequence of losses to the SNP, but they are likely to be over 285.

  29. @Pete B and Unicorn

    Is their model meant to predict what would happen if an election were to take place now or what will happen when it actually does take place?

  30. Crossbat

    “real damage done to the Tories by their lingering image as a party of wealth and privilege that is incapable of understanding the problems and everyday lives of ordinary voters”

    Thing is…

    “Toff” is plus and minus (think Downton abbey)

    “Spiv” is all minus

  31. Charles
    As far as I can see it does both – the table on the right is for the election in 2015, but there is also a tab called ‘Nowcast’ which says what it would be if the election was now.

  32. Crossbat

    missed a bit

    “It is a line of thought that almost defies the description counter-intuitive to think that any of this has “helped” Cameron.”

    The general “If it’s not hurting (you) it’s not working” attitude is what does the damage.

    adding Toff spin as the reason for it *reduces* the damage (slightly) (downton abbey effect)

    spiv spin would *add* to the damage.

  33. Pete B

    Just seen your election forecast web site info.

    Very interesting . Thanks.

  34. @ Charles

    Yes, @ Pete B is right.

    I’ve spent a bit of time trying to work out the relationship between the Nowcast and GE projection VIs. The adjustment seems to be close to adding or subtracting a constant to/from each Nowcast figure for each constituency. For example, when I looked in almost every constituency the Conservative GE VI was either three or four points higher than the Nowcast figure. The swingback adjustments for the other parties are almost – but not quite – as uniform as this. Presumably the scale of the adjustment will change as the election approaches.

    It would be good to see an explicit statement on how the adjustment is implemented.

  35. Nice of YG to organise a Scottish poll to celebrate my being released from th HD Unit.

    Nytol

  36. @Oldnat

    Welcome back :-)

    This poll must be a great tonic for you!

  37. @Barney C – Thanks, but I didn’t feel I was monstered – just that @Statgeek was (unusually) completely wrong and totally misunderstands the issue. I expect that from party press releases, but from certain people on here I generally anticipate some level of rationality.

    I sometimes hear you talk of the poisonous atmosphere up there, which I’m sure is true in places, and it certainly sounds like you have had to out up with a lot, but for me, it’s less the antagonism but more the refusal to consider basic facts.

    Both sides have had their share of this, but the denial that an economy with 15% of GDP coming from oil would be adversely affected by a 45% fall in oil prices when there is no stabiliser fund in place is just barking and takes the biscuit, as far as I’m concerned.

    My initial post on this wasn’t meant to contain any controversy – the numbers are the numbers, after all. What I was really asking was whether this gives Labour a chance to undermine SNP credibility, as the picture now is not one they projected in September. Judging by the response, I guess it’s not possible to convince some, as black must be white.

  38. @Oldnat

    Welcome back, and good to hear you are on the mend.

  39. @Pete B/MrJones

    I don’t think the Oxford connections are the problem, although passage there via an expensive public school doesn’t help, it’s the image of aloofness which is the damaging one. I accept your point that Labour are implicated heavily too in the current disillusionment with mainstream politics, but I think their problem is more to do with many of their leading figures talking focus group generated political corpo-speak rather than any overt association with wealth and privilege. The “Toff” label is a bit of a red herring anyway, and people like Boris Johnson and Jacob Rees-Mogg make a comedic virtue of the Wodehouse stereotype, but it’s when the perception lingers that politicians are living parallel lives to those they purport to represent, that it becomes very damaging. That’s the factor that I think hurts the Tories most and opinion polls seem to support this contention.

    If you listen to people like Alan Johnson or David Davis you are reminded that it isn’t impossible for politicians to appear to be fully paid up members of the human race and it baffles me why so many of them allow themselves to come over as dessicated speak-your-weight automatons, seemingly forever frozen by a fear of being authentic. Of course, by doing so, they create space for demagogues like Nigel Farage, more a pound shop Oswald Moseley than Enoch Powell in my eyes (sorry, Russel!), to assume the mantle of a “say it like it is” man of the people. The fact that Farage is a monstrous fake appears to be lost on many voters and he’s getting away with blue murder.

  40. “I don’t think the Oxford connections are the problem”

    The point I made about Oxbridge connections was about why I think Labourites are so personally hung up on toffs when most people aren’t.

    If I thought they could change I wouldn’t mention it.

    “he’s getting away with blue murder”

    lot of that about

  41. The telegraph has a ‘leaked’ Labour document explaining the UKIP vote

    http://i.telegraph.co.uk/multimedia/archive/03138/CampaigningAgainst_3138005a.pdf

    Very interesting reading. It sounds like they are trying to figure out who the UKIP voters are who will revert to Tory, and leave them well alone, and only target the UKIP voters who may be likely to vote Labour.

    But it contains a wealth of information on who the UKIP voters are, etc..

  42. @Pete b and Unicorn Thanks!

    @Unicorn I am old and forgetful. Was it you that was doing really interesting things with a model of how individual constituencies were going to vote? If so, you were going to do some more work on it. Did you do this and if so how is it working out?

    (If it was not you, apologies for taking your name in vain)

  43. @Crossbat
    “Of course, by doing so, they create space for demagogues like Nigel Farage, more a pound shop Oswald Moseley than Enoch Powell in my eyes (sorry, Russel!), to assume the mantle of a “say it like it is” man of the people. The fact that Farage is a monstrous fake appears to be lost on many voters and he’s getting away with blue murder.”

    Isn’t this sailing a bit close to the wind re the comments policy (to say the least!)

  44. @Alec

    “completely wrong and totally misunderstands the issue”

    In your opinion. The fact that Barney agrees with you gives me much confidence. :))

  45. Here’s an irony: Labour were presumably worried about the impact of the loss of their Scottish seats if the recent referendum had been “yes”.

    And now it looks they are going to lose quite a chunk (ie at least half) of those seats anyway, to the SNP. I suspect DC knew his party would benefit either way.

  46. @Alec

    The reason no one cares about the falling oil price is that it is totally irrelevant. If Scotland had voted Yes and the oil price fall continues until March 16, then it would matter. The talk of the oil price is just a continuation of the ‘too wee, too poor’ campaign, and annoys folk.

    Similarly, the fact the SNP’s prediction was too high doesn’t matter for the same reasons as above. No one predicted the crash so not predicting the oil price slump, when it is totally irrelevant is not much of a story.

  47. I suspect DC knew his party would benefit either way.
    ——
    I suspect he “knew” no such thing because there’s scant evidence that the Tories will benefit in any way from the change in politics wrought by the referendum.

  48. @Amber,

    Probably not in Scotland itself, but I think partisan Tories will be quite glad of Labour’s weakness on the SNP flank as there is only room for a certain amount of bad news in the media each day, and any day where the focus is “Labour going to get murdered by the SNP” is a day when the focus is not on the slightly shakier economic picture of recent weeks, or other dangerous ground for the Tories.

  49. An outline of Miliband’s speech later today on immigration can be found here.

    http://www.newstatesman.com/politics/2014/12/miliband-turns-focus-immigration-labour-continues-drive-address-weaknesses

    []

  50. Oldnat,
    Good to know you are back on UKPR. You’ve been missed.

1 2 3 4