The Sun have tweeted out tonight’s YouGov poll – topline figures are CON 32%, LAB 42%, LD 8%, UKIP 12%. This is somewhat at odds with the Labour leads of two, three and three points so far this week, though for what it’s worth all four polls would be within the normal margin of error of CON 34, LAB 39, LD 9, UKIP 12, the average of this week’s figures.

When an unusual poll comes along I personally rather discount it – more often than not it’ll just be a blip. When the same happens two days in row it gets my attention, but I wouldn’t conclude anything. When you get three in a row I normally take it seriously, it looks as though something is afoot.

But it can still just be random chance. Right now we don’t really know what the position is. It could be that tonight’s poll is an outlier and other polls will continue to show lower leads. Alternatively it could be that actually nothing’s changed and its all just been random variation around the six point lead we’ve had for months. As ever, time will tell.

392 Responses to “YouGov/Sun – CON 32, LAB 42, LD 8, UKIP 12”

1 2 3 4 8
  1. RHuckle

    Terrible result for the Tories in that Chadsmead byelection – apparently they dropped from 1st to 4th.

  2. Well, well, well, what a rather interesting poll to round off the week. I haven’t had my usual Friday night bath with my double digit bubble bath supplement (available at all good co-op retail outlets) for some time. I will luxuriate in it tomorrow night, listening to Any Questions and dreaming of three points at Goodison Park on Saturday.

    Litlle things, please little minds I suppose, but I’ll have a jolly little weekend now until Sunday’s You Gov poll brings me back to earth again.

    “Labour are back, Labour are back, hello, hello, Labour are back, Labour are……..”

    (apologies to all Gary Glitter aficionados)

  3. @AW
    “When an unusual poll comes along I personally rather discount it – more often than not it’ll just be a blip. When the same happens two days in row it gets my attention, but I wouldn’t conclude anything. When you get three in a row I normally take it seriously, it looks as though something is afoot.”

    I agree, but the problem is that we just had three unusual ones in a row, suggesting we take seriously some weakening in the Lab vote/strengthening of the Con vote, only for that to be followed by the highest Lab vote share in two months. So where we are now anyone knows.

    Anyway, anyone on either side who is disposed to wishful thinking can go to bed happy tonight.

  4. @Chordata – No doubt as usual every Lib Dem activist in the country had been parachuted into the ward, and UKIP have picked up the same trick too. That works extremely well in by-elections, but less well at general elections when those activists have to be diluted between dozens of constituencies.

  5. “No doubt as usual every Lib Dem activist in the country had been parachuted into the ward”

    What, all twelve of them? Seriously though, they won’t be spread that thin at the GE – the Lib Dems are going to be fighting a defensive battle, which means there are ca. 500 seats they won’t even be trying in.

  6. Someone says – ‘Still, it’s unusual for swings from one extreme to the other.’

    Why? If a poll is wrong its wrong. If both the narrow lead and the widest lead is within MOE then by definition we should not be surprised. Do not these results show the futility of day to day poling?

  7. @Phil Haines – Realistically, the true margin of error in a given poll is higher than the standard “+-3%”, which only refers to a limited statistical margin of error, without accounting for all of the procedural difficulties inherent to polling or the nature of psychological or sociological vagaries. Anthony makes that very clear himself, elsewhere.

    Honestly, the unpredictability associated with the unprecedented nature of four-party politics, especially involving an extremist party (of either flank – the SDP / Lib Dem rise was resolutely centrist), far outweighs the technical measures of statistical error that are used in the standard MOE calculation.

    I think that it’s wrong to say that “where we are now no one knows”. Nothing in the polls this week has been grotesquely outside the recent averages, plus or minus a realistic margin of error. Labour have a real but not devastating lead. UKIP remain persistent, and are hurting the Tories. The Lib Dems are down to their bedrock. The front page is not being held for such news.

  8. @ R & D

    “Have you seen their ray guns?”

    As I recall they fired themselves out of cannons in capsules the last time they pulled this s**t

    Still, so long as they have lots of dosh they should be fine

  9. @MrNameless – “Seriously though, they won’t be spread that thin at the GE – the Lib Dems are going to be fighting a defensive battle, which means there are ca. 500 seats they won’t even be trying in.”

    There are an average of about 15 wards per constituency in the UK. If the Lib Dems will be, as you suggest, campaigning in about 100 constituencies, then for every 1 activist in the ward today, there will be fewer than 0.0007 activists there in 2015. To me, that seems much thinner, and very likely to impact their chances of success, but of course that is just my private opinion.

  10. @Hookeslaw – “Do not these results show the futility of day to day poling?”

    I would say instead that the /response/ to them shows the futility of getting too worked up about a brief trend in a series of day-to-day polls, which in themselves, interpreted correctly, are a vital tool of the psephologist’s art.

  11. @Chris Martin

    I totally disagree with the way you interpret statistical variance in this context.

  12. @Howard

    “You have to await tomorrow’s demolition of the sample by Statsgeek and Roger Mexico, spiced with a churn analysis by Spearmint before you can claim anything here.”

    If 33 / 39 is the norm, 32/42 is within MoE. If Lab’s recent 37s are the new norm, 42 is an outlier.

    5th December – 29 / 41, with a six and five lead either side of it.

    12th November – 32 / 42, with a seven and eight either side of it.

    You have to go back to June to see consistent 10% or more leads.

  13. @Phil Haines – Yes, why so? Happy to hear your argument. :)

  14. …”f Lab’s recent 37s are the new norm, 42 is an outlier.”

    I should also qualify that with, unless that’s (42%) a completely new trend.

  15. We need to see the data set. Outliers often show something odd in the weighting.

  16. hookeslaw

    “Do not these results show the futility of day to day poling?”

    No. The opposite.

    They show how UNhelpful and confusing monthly ones could be.

  17. Good early morning all.
    Politics is interesting at the moment I think, with fifteen months left to the next GE and three months until the EU Parliament Elections.

  18. Off to Marseilles for son’s wedding.
    Hope this poll is a positive augury .

    All together now: ” Marchons! Marchons! ….”

  19. This big ‘leap’ from the previous polls merely indicates margin of error from one side to another. The gap is probably still about 6%, which would probably equate to the average of a poll of polls this week.

  20. It’s possible that polls become volatile before a (fairly) big switch (?)

    Back in 2010 the switch from LD to Lab spread out over the summer and autumn as I recall. Perhaps as the debate continues people shift their footing.

    I notice the Government are giving OfGem one last chance to fix the energy market and break up the cartel of six/monopoly which has been ripping us off. Or to put it another way, “admitted that Ed Miliband was right, but because they didn’t freeze the prices and act at the time, allowed the energy prices to be hiked despite no increase in wholesale costs”.

    That might sway my vote, if I was swayable.

  21. Looking at the actual poll crossbreaks I can see no real obvious imbalance in the sample…2010 Cons perhaps slightly over-represented.

    I imagine that,after with the shambles with his backbench MPs yesterday, DC et al will be watching the next few polls with a slight shiny sheen on his/their brow(s).

  22. Working backwards, unless I missed some, Lab voting intention in polls looks like:

    42, 37, 33, 37, 40, 39 followed by a procession of 38-40%.

    The 33% came from another company so might be out of line (also might be the only accurate one for the same reason!). But if you ignore it, Lab was pretty stable until this 42% which might well be 38% tomorrow!

    Or the 50% tax effect has kicked in and they are heading for 49%?*

    *this might be a bit of wishful thinking, please ignore

  23. Interesting Yougov overnight,

    I think there is still a large centre of gravity around 33 for the Conservatives and 38-39 for Labour.

    Occasional positive period headlines perhaps boosts one party or the other, but unless sustained the polls then go back to the centre of gravity.

    For example, the recent good news on unemployment and the economy got some broad coverage for a little while. Maybe some people responded to that, but as the newspapers went in the recycling bin and the news machine moved on, the effect wore off, as these people realised they feel no better off by personal experience this week that they did last week or last month.

    I think this will continue until something really decisive and permanent occurs (as yet unknown I may add).

    If Labour or the Conservatives want to gain some serious VI, they both need to do or prove more than they are now.

  24. @Carfrew, Chris Martin

    I can’t explain how the calculation should be done (at least not without apending all day trying) but a little bit of “Monte Carlo” analysis suggests Carfrew is nearer the mark. I.E. I a generated string of 100,000 random numbers (from 0 to 1) and looked for 3 in a row under 0.2. I came out with about 0.79 incidents in each string of 100. So ver 20 weeks there’s a good chance of a “3 in a row”.

  25. Paul – dead right monthly polls can be a problem.
    Personally, I like ICM and their adjustments (except LD 50% from WV/KN 2010 base probably too high) but when they report moe can produce results which some newspapers take as meaningful even if nothing much has really happened.
    Imagine if this mornings had been the only YG this week?

    I note that even though the lead is 10% (with this outlier/ extremity of MOE poll) approval still -21% which will give some cheer to the right.
    I expect some in the -17/8 range soon which the Tories will hope presages VI movement in their favour while Lab may think demonstrates that support is highly entrenched even among those viewing the Government more favourably now.

  26. Well let’s not get to excited just yet about the latest poll as it is currently an outlier in the last four the next poll will tell us more of where we are ……probably bang on six points!

  27. CROSSLANE1945
    May I suggest that you try this on your class:
    “Moments of this and that (Balls sounding like he’d be running his own austerity policy; Tory press tell the world about the economic miracle) and the Don’t Knows on the fringe waver and the weather seems to change, but in any kind of underlying way it doesn’t.” (Colin Davis)
    Contrast with the proposition that changes in the class system in the UK give rise to long-term trends in voting intention, in the light of Max Weber’s writings on Marxist conflict theory (CARFREW earlier thread)
    How might these influences on voting intention be affected by public attitudes to behaviour of the police in providing support to a private prosecution by Virgin in return for a share in the resulting compensation? (Guardian newspaper, yesterday)

  28. I do have an irritation that we had a bunch of articles during the week making commentary based upon the ‘narrowing of polls’ we had, and yet barely anything stated about this poll other than the “usual caveats” about it possibly being an outer (we know!).

    In various newspapers and website, we had commentators conflating these polls with the issue of 50p taxation, Ed Balls and so on – with the inferred message that this was the impact of the policy announcement. Will we see the same commentators make more analysis based on this poll? I doubt it.

  29. @PostageIncluded – I repeat, please look at the “200 polls” example, which on Carfrew’s “methodology” suggests an impossible 160% chance of the event occurring, and thus proves very clearly that on this occasion his calculation was not the correct one.

  30. R&D
    “Mars invades and they’re all bleedin’ Tories for example.”

    The proven election strategy in that case is to play Frank Ifield”s “I remember you” over public loudspeakers, upon which their heads will explode.

  31. ” Will we see the same commentators make more analysis based on this poll? I doubt it.”

    I think the blame for the currency crisis in emerging markets has been placed firmly with David Cameron.

    The weekend will be interesting- Populus, Observer (I think) and another yougov. We should know by the end of those.

  32. @NickP there are too few Midlands: weighted up quite a bit. They normally vote against Labour. Midlands/Wales has an unusually high Labour VI. Possibly an outlier.

  33. Several people have alluded to this poll and the 50p rate announcement, and I have to admit that this was my first thought.

    It’s far too early to make any judgements about this, either way, but I was very surprised at the linking of recent polls with narrower leads to this issue (even on here, surprisingly) as the weekend polls were sampled entirely before the announcement, so clearly could not have been influenced in any way by it, the early week polls would only have been partially completed after the policy was released, and really on the Wednesday poll could be argued to fully incorporate reaction to the new position. Even here, we know from previous events and subsequent poll movements, that a period of several days tends to be needed for voter reaction to show through in polls.

    It’s actually far more logical to argue that this high poll is the first reaction to the apparently popular 50p announcement, but I don’t think that would be a sensible position to take at this stage either.

    What this polling episode does demonstrate, however, is that there are still significant sections within the media that are not reconciled to the option of a Labour win in 2015, and are desperate to seize on data that suggests a Tory revival.

    This is, in itself, a problem for Labour, as winning often depends not so much of getting people to actually like you, but getting them to believe that you can win. Unless Labour can overcome the expectation that there will be a ‘normal’ swing back to the Tories between now and 2015, their job will be that little bit harder.

  34. 50p effect shoring up Labour waiverers I reckon.

  35. Could well be back to Labour on 39%-ish next, hopefully an end for now to all this flailing around.

    There’s things to be said for daily and monthly polls. One risks missing things, the other maybe sees too much.

  36. For the moment, I am assuming that “nothing happened” this week. We have become used to a relatively stable oscillation around a long term lead of around 6-7%, with a slight reduction over time (glacially paced but discernible). Until something (ie 10 more polls) convinces me otherwise, I think this is continuing but we’ve just seen one of those slightly larger than average wobbles in the oscillation.

  37. Alec – the so-called attack by FTSE CEOs (Martin Sorrell arch Tory the only one on record as far as I can recall) around Davos trying to drive a wedge between CU&EB on one side and EM on the other was taken in the PLP as corporate Tories believing an EM premiership is now a possibility.

    Soon the media will catch up but there could be some nastiness initially.


    “No doubt as usual every Lib Dem activist in the country had been parachuted into the ward”
    What, all twelve of them? Seriously though, they won’t be spread that thin at the GE – the Lib Dems are going to be fighting a defensive battle, which means there are ca. 500 seats they won’t even be trying in.”

    Classic sour grapes of the bitterest vintage.

    “the Lib Dems are going to be fighting a defensive battle”

    The Lord Ashcroft polling data show that defensive battle is going rather well.

    Presumably you’d rather Lib Dems fought in areas they can’t win?

  39. FWIW – my considered take.
    Most of us agree that Economic data good or bad has little real impact as it is peoples actual experiences that guide their VI.
    The exception can be at GE time when individuals future prospects (and their views about the Economy going forward) will play a stronger part.

    Occasionally, though, some stats or announcement can have a temporary impact due to sheer scale and it may be that the positive Economic numbers were so significant and ostensibly positive as to produce a decent Tory lift and a small Lab decline.

    Notwithstanding my observations re the approval rating this morning it may be that the ‘Economy may be improving but its the rich who are benefitting the most’ line from Labour (whether fair or not) has produced some unwind with the rest of the change being pure moe.

    We need more evidence but like Neil I expect that come the beginning of April we will have seen a modest narrowing mainly imo due to a Tory increase from DK/WV and some UKIP.

    The Euro Elections will mess up polls for several weeks as the UKIP will get a GEVI bounce so we may have to get to July or even September for a clearer picture…and then the conferences.

  40. I watched part of Question Time last night and both the Tories and Labour took a battering over tax but Ken Clarke and Emily Thornberry did give a good account of themselves and despite the polls both parties haven’t really convinced the majority of the public that either part has the answers.

    I know the Lib/Dem Peer on the program was a stand in for Charles Kennedy but I couldn’t believe how he was playing Labour off the Tories and making out the Lib/Dems were somewhat innocent.

    He spent more time attacking the Tories like so many Lib/Dems do on the TV than attacking the opposition so if the Libs are innocent as they try to make out then why are they sitting on between 6% & at best 9& in the polls?

  41. @Chris Martin

    Bit perplexed by your exclaiming that you weren’t telling me off. I hadn’t even suggested that you might have been. All I did was say I hadn’t read my stats book and stuff. Admittedly, going back to read it again, your “even to non-mathematicians” comment etc. might be considered to be rubbing it in a little bit, given it’s nit like I’d had a pop at anyone or anything, but no biggie… frankly I was just happy someone had got involved.

    Anyways, your explaination of the probability thing was perfectly serviceable, for those who may be ok with some “A” level maths, but for others it probably rather assumes too much. Still, even I can follow it, so that must count for something.

    As for the time series thing, well… I think you probably know that “clumpiness” possibly doesn’t really do it. You didn’t explain what a residual is, or how it differs from error, why it’s of significance, why dependence matters, why we check for “clumpiness” (or whether the residuals are randomly distributed), or what sort of things can give rise to residuals, and how residuals can frustrate the search for trends.

    On the plus side, it did establish the directionality of time thing. Which apparently is quite important for cause and effect and stuff…

  42. Comared with the previous Poll I make Labour’s gain of 4% pts a shift from :-

    Lab 2010 WNV/DK ( around 1.5% pts effect)
    Non 2010 id-ers – other parties , but particularly UKIP & Others( around 2.5% pts effect)

    Assuming the samples/base data etc in the last two polls is comparable-I would put this down to 50p appeal.

  43. JIM JAM

    Reading your comment I don’t think anyone could argue with what you have written.

    Polls for everyone will go up and down for various reasons and are only snapshots of public opinion at that time so any big policy announcements by a party could see a shifting of the polls and within a few weeks we’re back to where we were.

    On the European elections, I think we all know UKIP will make the headlines and after all this is their stage but if Labour do well then it will give them a momentum boost which is not a bad thing but as you say I think we need to wait until after July to get a clearer picture of where everyone stands.

  44. ” it could be that actually nothing’s changed and it’s all just been random variation around the six point lead we’ve had for months”
    and if you apply Occam’s razor as you should, there is no need to look for another explanation.

  45. “@Colin

    50p effect shoring up Labour waiverers I reckon.”

    I doubt that the 50% rate has made any difference. The polls on the 50% rate have been mixed. Personally I am against the 50% rate, because the amount extra raised does not appear very much and it just sends out the wrong signal. The UK should want wealth creators to base themselves here. Personally I would also increase the salary band where people start paying at 40%, as the amount is currently too low.

    Labour should concentrate their efforts against corporations who avoid UK tax, by offshoring their profits to cheaper tax countries within the EU. They should be looking to get the EU to make relevant changes to the rules. This should net the UK many billions in extra corporation tax.

    As for the Tories, Labour are running a very good smear campaign against them, that the Tories only look after big business and rich people. The response by the Tories has not been very clear and effective. Whoever is responsible for the governments communications on current UK financial conditions should be sacked. They need to present clear information and not some dodgy data, which Andrew Neil ridiculed recently.

  46. @ ALLAN

    Bit unfair to have a go at Matthew Oakeshott (the LD peer on QT)
    He resigned or was fired from the LD front bench over coalition policy and has been consistently critical of the coalition in general and Osborne (whom he called ‘the work-experience chancellor’) in particular.

  47. Given the popularity of Miliband’s campaign against the utility companies perhaps criticism of the 50p rate by business leaders has only benefited Labour further.


    I didn’t know all the background stuff on Matthew Oakeshott and probably explain his hostilities towards the Tories on QT.

    Okay yes it was a bit unfair to have a go at him and I know he was a last minute stand in for Kennedy but if they (BBC) are going to have panel members representing parties then at least they should be a true reflection of that party so people in the audience and at home watching the program can make a better judgement on each party and its policies.

  49. I think the 10% lead is down to random variation and that the Lab VI mean remains pretty constant at around 38%. It is tempting to look for explanations for big daily changes, but the temptation should be resisted unless there has been some truly dramatic political event..

    Also sensible to look at VI itself rather than the lead, which tends to be exaggerated because of the negative correlation between Lab and Con VI.

  50. So are we still in the polldrums or not?

1 2 3 4 8