This morning’s YouGov poll has topline figures of CON 34%, LAB 40%, LDEM 10%, UKIP 7%. This is the second YouGov poll in a space of a week to show the Labour lead down to six points. After yesterday’s YouGov poll showed a Labour lead back to eleven points following the six point lead at the weekend I had rather dismissed the YouGov/Sunday Times poll as just an outlier. This poll suggests there may be more to it.

Of course all the YouGov polls over the last week or two have had the Conservatives within two points of 33% and the Labour party within two points of 42%. While two six point leads after months of nine to ten points leads is rather a big co-incidence, strictly speaking there isn’t actually anything that couldn’t be explained by normal sample variation. Let’s wait and see a bit longer before looking for explanations.

There is also a new Opinium poll out today which has figures of CON 31%(nc), LAB 42%(+2), LDEM 8%(-2), UKIP 9%(-1). There is certainly no sign of a narrowing Labour lead there.


167 Responses to “New YouGov and Opinium polls”

1 2 3 4
  1. The polls are like yo-yos all of a sudden.

  2. I don’t see that the polls are greatly varying. Con up two, lab down two, makes 4 narrower percentage lead. So what?

  3. “So what?”

    So the people who AW has told time & time again, not to read too much into one poll, are still doing it.

    Only now they are slightly concerned-rather than convinced of their invincibility.

  4. I still think when Labour are properly scrutinised over policy then the polls will change in the Tory direction.

    I’m not sure why the beast of Bodmin moor, oops Bolsover, was calling for David Cameron to call an election?? Gordon Brown and that poll anyone??

  5. This is all rather strange. YouGov have an 11% Labour lead sandwiched between two 6% leads, only a week after showing three 12% Labour leads out of a run of five polls. Now we have Opinium showing the largest Labour lead (11%) that they’ve recorded since June! Most other pollsters are more or less as you were.

    We probably need to calm down a bit, unless YouGov start to show a run of consistently smaller Labour leads, because I’ve seen no recent development that could conceivably have shifted public opinion significantly either way. In fact, the only two big events over the summer that might have drifted support the incumbent government’s way, the Jubilee and the Olympics, have come and gone without touching the polls at all. Unless I’ve missed something big, nothing has happened in recent times to move the polls in any meaningful way beyond the vagaries of statistical volatility.

    Expect a Labour lead of 9% in tomorrow’s YouGov!

  6. This YouGov six-pointer (and the previous one?) is showing similar proportions (but not exactly equivalent numbers) of Con 2010 voters and Lab 2010 switching between the two parties. Fewer 2010 Cons in the don’t know column as well perhaps?

    Most recent polls have shown more Con switching to Lab than the other way round.

    No big difference in the behaviour of LD>Lab or Con>UKIP that I can see.

  7. @ Colin

    Only now they are slightly concerned-rather than convinced of their invincibility.
    ——————
    Not me however; I am unconcerned. Conservatives aren’t on the rise. The movement is in & around the center-left parties. The future Con plan seems to be stop the rot to UKIP & win back some of their losses on the right. Lord Ashcroft’s polling & recommendations appear to have been largely ignored, if the shuffle is anything to go by.
    8-)

  8. @Amber Star

    Not me however; I am unconcerned. Conservatives aren’t on the rise. The movement is in & around the center-left parties. The future Con plan seems to be stop the rot to UKIP & win back some of their losses on the right. Lord Ashcroft’s polling & recommendations appear to have been largely ignored, if the shuffle is anything to go by.

    ___________________________________

    Yes, but don’t forget that Tories won’t necessarily think they need to compete with Labour in the centre. That’s the Lib Dems job. If Tories move to the right and LDs to the left (wealth tax etc.) then they potentially kill 2 birds with one stone: both differentiating themselves ahead of the election and eroding Labour support.

    It’s still quite a challenge, but at least having gotten the more, erm, controversial policies out of the way early, it’ll be easier to differentiate without new policies in practice seriously undermining all the time…

  9. @CB11

    “This is all rather strange. YouGov have an 11% Labour lead sandwiched between two 6% leads, only a week after showing three 12% Labour leads out of a run of five polls. Now we have Opinium showing the largest Labour lead (11%) that they’ve recorded since June! ”

    As opposed to polldrums. There’s no pleasing some people. :)

    If we get another 11% poll we can look at five poll @ 12,6,11,6,11 to be an average of 9.2% which is reasonable MoE I think.

    I wonder if YG asked group 1 then group 2, then group 1 again (6, 11, 6).

  10. This is quite an interesting site.

    http://istwitterwrong.tumblr.com/

    When there is a story going around, worth checking to see, if it is another fake Twitter story.

  11. CARFREW
    If Tories move to the right and LDs to the left (wealth tax etc.) then they potentially kill 2 birds with one stone: both differentiating themselves ahead of the election and eroding Labour support.
    __________________________________________
    Struggling to see how the Tories moving to the right will erode Labour support?

  12. CORCKSCREW
    Struggling to see how the Tories moving to the right will erode Labour support?
    ————————————————–
    Immigration and EU policy for starters. Probably crime as well…

  13. @Corkscrew

    “Struggling to see how the Tories moving to the right will erode Labour support?”

    That assumes that all Labour voters are left-wing. Past examples of Labour voters migrating to the BNP seems to suggest there are some who are not left or right wing as such, but vote according to the flavour of the month.

  14. If we get a few more like this one, I’ll begin to believe it . Then of course I will, especially if the same trend is evident in others (which it isn’t – yet).

  15. @Statgeek

    “Past examples of Labour voters migrating to the BNP”

    Going to need to see some evidence that shows this on a UK wide scale at a statistically valid margin.

    But no, the Conservatives moving to the right *does not help their national vote share*. It loses them their party-left-wing and floating centrists to either LibDem *or Labour*, and may not gain them back as much from UKIP. A right wing move only makes sense on the assumption of a right wing leadership threat within the party. (Teapartification?)

    The Conservatives still need to increase their total vote share to get into government again. Giving more to the LibDems does not help them, because it won’t get the LibDems back up to a level of threatening Labour. The Conservatives can not hope to govern the country on only a third of the vote, even with FPTP. Even in the notorious 1951 election, where the Labour won the popular vote but lost the house of commons, the Conservatives were at 48.0% to Labour’s 48.8%.

    It’s firmly my opinion that Lord Ashcroft is right in this regard. There is no net gain to the Conservatives of firming up their support on the right, they are only cementing themselves in place while the tide rises around them. The best bet for the future of the Conservatives would be to allow their right wing rump to eject it’s self to UKIP, and absorb the Orange Bookers and realign themselves as “Conservative Liberals”. This reshuffle makes it clear this is not going to happen.

  16. @Corkscrew

    CARFREW
    If Tories move to the right and LDs to the left (wealth tax etc.) then they potentially kill 2 birds with one stone: both differentiating themselves ahead of the election and eroding Labour support.
    __________________________________________
    Struggling to see how the Tories moving to the right will erode Labour support?

    =============================================

    Yeah, I was thinking about the overall impact. Of both the Tories and LDs. It’s the LDs moving leftwards that mostly threatens Lab votes… (Tories tacking right may assist them via differentiation though).

    The important thing from the Tory point of view is as Amber says, to recover some UKIP votes. This is liable to be limited, as a proportion of them will be voting UKIP over Europe and immigration, which they see as so fundamental as eclipsing most everything else, and some of them at least have twigged Tories are unlikely to deliver as business loves the EU and immigration.

    The slump in LD fortunes and rise in UKIP changes the game somewhat. If Tories move to the centre, they lose votes to UKIP. And at the same time, pulling LDs into coalition has stifled their effect on Labour’s vote. So it’s an interesting conundrum…

  17. @jayblanc

    But no, the Conservatives moving to the right *does not help their national vote share*. It loses them their party-left-wing and floating centrists to either LibDem *or Labour*, and may not gain them back as much from UKIP. A right wing move only makes sense on the assumption of a right wing leadership threat within the party. (Teapartification?)

    ____________________________________

    Not so sure about that. Tories are already far enough to the right to have cost them much of the central vote. Hence their polling. Moving to the centre may not GAIN them much either since competing with both LDs and Labour.

  18. CARFREW

    Thanks for the clarification, I agree with your analysis – it is a very interseting conundrum for the Tories…..

  19. @Jayblanc

    h ttp://www.channel4.com/news/articles/politics/domestic_politics/who+voted+bnp+and+why/3200557.html (2009)

    “But perhaps the most startling finding came when we tested anecdotal reports that many BNP voters were old Labour sympathisers who felt that the party no longer speaks up for them. It turns out to be true. As many as 59 per cent of BNP voters think that Labour “used to care about the concerns of people like me but doesn’t nowadays”.

    h ttp://iaindale.blogspot.co.uk/2009/02/labour-ukip-voters-turn-to-bnp-in-kent.html

    Swanley St Mary’s, Sevenoaks District Council (2009)
    BNP – 41% (+41)
    Lab – 34% (-21)
    Con – 25% (-)

    Also during 2010 election (picked from BBC map):

    Leeds East
    Scunthorpe
    South Shields
    Tyneside North

    In fairness, I can’t prove that the voters transferred directly, or if they came from a previous ‘didn’t vote’ bracket.

  20. Well I said it in the last thread and I’ll say it again: the most likely explanation is simply the rise in Tory media coverage over the weekend -pre-shuffle.

    Its what happens as we approach Autumn/Winter that will be much more important.

  21. Carfrew and Corkscrew

    Wow, and all this before the conferences!

    The YouGov ‘undulations’ are not repeated elsewhere, yet.

    Bit early (er, 3 years) to start GE tactics is it not?

  22. Statgeek

    Mrs Duffy alive and well in Sevenoaks?

    I just do not see minor parties getting anywhere but a little voice inside me wishes them well, as long as they don’t really get anywhere.

    Under PR, it has to be admitted, it can happen.

  23. I don’t see any way that Lib Dems will have credibility ‘moving to the left’ on policy issues. Sure they can produce a whole load of left leaning policies, however no-one will believe them without a new leader and some serious slagging off of Clegg in the process. The left leaning voters will take a while to feel they have changed. This is very unlikely to have time to work itself through before the next election.

    It’s a bit like the Labour issue of having to defend their handling of the economy- they have struggled with this in a ‘we got it wrong but have learned our lessons’ type way, but because of the current state of the economy Labour have so far got out of jail a bit because the economy looks just as bad now as back then.

    Incidently I do not believe these 2 x 6 pointers one bit- just nothing that would have caused such a move in the last week. The most that can be said is that probably there was never a 12 point lead.

  24. @howard

    Carfrew and Corkscrew

    Wow, and all this before the conferences!

    The YouGov ‘undulations’ are not repeated elsewhere, yet.

    Bit early (er, 3 years) to start GE tactics is it not?

    ___________________________________

    Nice try but I wasn’t particularly talking about the timing now was I?

    The fact remains that Tory polling is being eaten into by UKIP and LDs have had theirs eaten into by Labour. I was just pointing out it might rather be in their interests to do something about that if they are keen on being re-elected.

    Still, they’re not going to leave it till 3 weeks before the election are they.

    I mean, if you’re a Lib Dem the sooner the better really. It took a year for LDs to get the AV referendum that went nowhere, and two years to find out they weren’t going to get HoL. Tories do seem keener to get a move on than LDs to be fair. Now while LDs um-and-ah over whether to ditch Clegg, you may have noticed Tories have just had a shuffle and and Cameron doesn’t even like them…

    If that wasn’t enough, you don’t even know for certain if the coalition is going to survive another 3 years anyway…

  25. @statgeek

    You mignt want to read this (In short, the people the BNP seem to appeal to are actually “working class Tories”):

    h
    ttp://ukpollingreport.co.uk/blog/archives/2172

    @Jay Blanc – “Going to need to see some evidence”

    No need, it is quite simple…

    Labour voters read the Guardian.
    Working-class people vote Labour.
    BNP voters are working-class.
    Therefore BNP voters read the Guardian.

  26. @Howard
    My comment on the previous thread, using your own words about Clarke, was:
    “I can think of another five who also have the salary, the perks, and who sit in Cabinet to tell the others what to do, without apparently being taken much notice of.”

    You chose to interpret that repeatedly as a comment that such ministers “do nothing”, which is rather different.

    Let me be clear. I was not questioning the actions of your five Lib Dem Cabinet ministers, just their ineffectiveness, which is indeed reminiscent of the role of ministers without portfolio. Their inability to achieve little beyond what might have happened in their absence from the Government reflects their own mainly junior, limited and constrained portfolios as well as their inability to constrain the actions of others with more substantive roles in the Cabinet.

  27. @phil

    Their inability to achieve little beyond what might have happened in their absence from the Government reflects their own mainly junior, limited and constrained portfolios as well as their inability to constrain the actions of others with more substantive roles in the Cabinet.

    ______________________________________________

    It also reflected their lack of urgency. While Tories pursued their agenda at breakneck speed, LDs gave them what they wanted up front, content to wait a year or too for the voting reform stuff, only to find… they weren’t going to get it anyway.

  28. Phil and Carfrew

    Indeed, so add those LD examples to those getting minister salaries for nothing and — Carfew, what made you think I was trying to somehow stick up for the LibDems? I comment here as a non-partisan as the rest know full well.

    It is actually possible to support a particular choice without being partisan, as our host demonstrates and I hope i do.

    What makes me puke is the Newsnight hack politician apologist stuff we get on here such as ‘oh, good move that Buggins is now in x dept, that should transform our fortunes’. Ugh!

  29. @Howard

    “Mrs Duffy alive and well in Sevenoaks?”

    No idea what that means.

    “I just do not see minor parties getting anywhere but a little voice inside me wishes them well, as long as they don’t really get anywhere.”

    I was just making the distinction that some (mostly those on the left) assume that all BNP voters are from the right. I don’t think it’s quite as clear cut as that, and nor do I think that it’s a strawman (i.e. if they voted right, they were never left anyway).

    I’m inclined to believe some people are so disillusioned that they turn to almost any party which (claims to) represent their interests.

    Tbh, their 2009 successes didn’t materialise in 2010, so it’s probably nothing to worry about, but there may be potential for smaller parties in 2015, if it seems that there’s little changing (in peoples’ perspectives of parties). If that’s the case, it may very well affect the large parties’ fortunes.

    e.g. If we assume UKIP voters to be ex-Conservative voters, UKIP have done the Conservatives out of seats in the past, but have never won a seat themselves.

  30. A Tory MP remarked that the reshuffle indicates winding down of the coalition.With hawks in Environment and Justice and two minders to look after Cable,it is not difficult to see confrontations emerging.Clegg has achieved power for the Lib Dems for a few years at the cost of the tactical Labour voter for a couple of Parliaments I think

    Boundary changes are looking more unlikely and Labour need a small VI lead to get OM with the old boundaries.I hope Miliband has sent a thank you note to Tory backbenchers for scuppering Lord`s reform.

  31. Sigh. I can’t leave you lot alone for a moment can I? When I looked at the YouGov tables this morning I thought “Surely they’ll see what’s wrong straight away and not get all excited over that”. No such luck.

    What’s wrong of course is our old friend the YouGov Yoof Problem. Because under-25s are under-represented on the YouGov panel (particularly males and early leavers from education) this means that the views of those who reply are heavily weighted up.

    Furthermore the sample is so small that the VI figure often appear in italics because the percentages are based on a sample of less than 50 people. Even when this doesn’t apply it means the percentages vary wildly and because of the heavy weighting involved there can be a disproportionate effect on the topline VI figures.

    Normally Labour has a strong lead in the 18-24 age group – yesterday they led the Tories 55% to 29% for example. However this poll shows the latter leading 34% to 28%. I would estimate this enormous swing was based on just 10 less people choosing Labour and 8 more Conservative. If this hadn’t happened I reckon the topline figures after weighting would be Lab 42%, Con 32% – a 10 point lead similar to before rather than 6 everyone is getting excited over.

    The calculations are back of an envelope stuff but they give an idea how the low numbers and hence apparent volatility of the under-25s can move the YouGov figures about based on very little.

    If anyone needs any further proof of how those figures can be distorted they should consider the fact that 12% of this group are saying they will vote SNP or Plaid. Given that Scotland and Wales together only contain 13% of the population of Britain, I think even our resident Nationalists may consider this a little hopeful.

  32. So, anyway. TDKR passed the $1billion worldwide mark FOUR WEEKS EARLY thanks to a storming opening in China, grossing $32mill in China in one week (cf $17mill for the Avengers over a similar period). This has thrown my estimates out of whack: I figure it’d level off at $1-1.05billion by end of Sept, but – hey – we’re now looking at ~$1.1billion. Which is nice for the Nolanses (Nolii?).

    Anyway, why is this important? Because unsuspected popularity in an obscure but fast-growing market led to a significant underestimate. Which got me to thinking, and here’s my question:

    “Which is the fastest-growing sector of voters?”

    Split it any way you want: age, gender, geographical, political. I’d say (judging from UKIP’s rise from 2-3% to 7-8% in two years) it’s Eurosceptics. I’d also accept Lib defectors, judging from the drop in Lib votes from 24-26% to 9-11% in the same period. Possibly also Unionists after the Olympics, and that’s the most recent fillip.

    So, I assume the present volatility is down to differential attraction in one of these groups. Does anybody else agree, or is there another group of fast-growing voters who I’m too thick to notice?

    Regards, Martyn

  33. @Roger Mexico

    You said “…Sigh. I can’t leave you lot alone for a moment can I? When I looked at the YouGov tables this morning I thought “Surely they’ll see what’s wrong straight away and not get all excited over that”. No such luck…”

    Dude! We pay[1] you and Statgeek and Virgilio et al. to do all the hard work and thinking so the rest of us can pretend we know stuff…:-)

    Regards, Martyn

    [1]: We pay in esteem. It’s the new algamic economy: all the rage, apparently… :-)

  34. ROGER MEXICO

    Additionally, samples in the cross-breaks aren’t demographically balance within themselves.

    All that was needed to produce the 12% for SNP/PC was for more Scots in that age group to have responded than would be proportional.

  35. howard

    Phil and Carfrew

    Indeed, so add those LD examples to those getting minister salaries for nothing and — Carfew, what made you think I was trying to somehow stick up for the LibDems? I comment here as a non-partisan as the rest know full well.

    It is actually possible to support a particular choice without being partisan, as our host demonstrates and I hope i do.

    What makes me puke is the Newsnight hack politician apologist stuff we get on here such as ‘oh, good move that Buggins is now in x dept, that should transform our fortunes’. Ugh!

    _____________________________________________

    Well I find it’s even easier to avoiid being partisan if you’ve never been much impressed with any of the parties, frankly. Or with the ideology thing.

    But it’s another red herring that’s a distraction from debate. I didn’t accuse you of being partisan, I simply used the LDs as an example of what happens if you don’t seek to get a bit of a move on in government. They’re the only current example as Labour are in opposition and Tories are in fact getting a move on and, let’s face it, doing rather better than LDs for it.

    Blair did say he wished he’d moved faster, of course. Given the omnishambles, maybe Tories should have been a bit more careful… but they achieved most of their objectives before LDs realised they werre going to get pretty much sod all.

  36. Well said Martyn and good work Roger.

  37. @Roger M

    I make it 42% to 34%, an 8% gap, if we just include the under 25 sub-sample without any extra weighting, thus preserving the original whole sample without exaggerating the impact of the dodgy bit. Effectively that takes 112 (=205-93) people out of the sample used for calculating the % figures.

  38. Looking at the Opinium poll.

    According to the “house effects” graph, we should expect Opinium to be one of the polling companies which generally marks down the Labour lead (~ minus one), whereas YouGov would be marking the Labour lead up (~ plus one) relative to a notional midpoint.

    What does strike me about Opinium is that when other polling companies were giving Labour double digit leads at certain points in 2011, the Opinium Labour lead was always fraction of that.

    Their graph points makes the period between the budget and the local elections this year as being even more significant:

    h
    ttp://news.opinium.co.uk/survey-results/political-polling-31st-august

    @Roger Mexico

    I did look at the18-24 xbreak this morning, but… a) I’m not well equipped in the statistical department, b) I got sidetracked, and c) I was waiting for your magisterial summary.

  39. Well,I am sure that I will be moderated just for being alive,but the Tories have
    Had a great deal of media coverage in the last few days,wheras labour have
    Been invisible.Ergo ,Cons doing well in the polls.

  40. Ann in Wales:

    You should jolly well get moderated for repeating what I’ve been writing, ad nauseum, for the past three threads.

    Disgraceful !!

  41. I am new to this site but felt i needed to say something on Statgeeks comments about Labour/BNP voters and especially his drawing attention to the Swanley St Marys by-election result from 2009. I was involved in that campaign and the one in 2011 when Labour won the seat back. The change in voting behaviour was mainly centred on who turned out. Turnout is extremely low historically in this ward and in 2009 people voting BNP were usually non-voters, allienated from the system, and not previous Labour voters, at least not Labour voters from the resent past.

  42. Watching PMQs today, I still think it might be Ed Miliband that costs Labour. We shall see.

  43. I’ve never been convinced that there is much in the way of an electoral dividend for the Tories in veering right, despite some ambivalent polling evidence suggesting that the British electorate appears to be firmly right of centre on a number of issues like welfare, immigration, law and order and Europe. The simple truth is that whenever the public has had a chance to vote for such a platform, they’ve more often than not declined to do so. Hague and Howard will testify to this and Cameron, to his credit, recognised this electoral fact of life when he steered his party, as he best he could anyway, to the centre ground in the latter part of the last decade.

    Now, there may be some short term benefit to be derived from tilting the government rightwards, but this will be for wholly internal Tory party purposes. Essentially, Cameron is leading a party that is to the right of the electorate on most issues, certainly if voting rather than polling evidence is to be believed, and this is why I think he is in an enormous political pickle. He could end up a bit like poor old Michael Foot; the more he pleases his party, the more he alienates the electorate. Put in layman, non-political terms, that’s a bit like being up a creek without a paddle.

  44. Roger M
    ‘on a sample of less than 50 people’

    Whilst equally in awe – ‘fewer’ Roger, ‘fewer’.

  45. BALTUS

    Always good to see new blood on here!

    Since it’s a civilized place, we seldom spill any.

    Unfortunately, that doesn’t seem to have been true in Quebec.

    http://m.theglobeandmail.com/news/politics/elections/one-dead-as-shots-fire-disrupt-pq-victory-celebration-in-quebec/article4518887/?service=mobile

  46. All these events today masked an appalling news coverage of ‘starving Britain’ an example of which was aired this morning on Radio 5. Mrs H was doing the chores while listening to the Derbyshire lass on Radio 5. These programmes are significant, in that, unlike PMQ, which I note is avidly followed by people here, it is likely to be listened to by significant quantities of voters. While discussing the chores (hah!), I listened to a couple of interviews. They were with single mothers and not once, not once, were the respondents asked what income was being received from the other parent. I wondered whether the BBC has a policy of ‘don’t go there’ on the issue, as I am sure hundreds of thousands of listeners were shouting at the radio, as Mrs H and I were.

    I think these issues are more likely to produce a move in VI than PMQ.

  47. @Baltus

    Good info. It’s always difficult to get to the facts in some of these cases. Several sources of info are always good, as they can add to the overall story.

    Low turnout is always a bad thing (imho).

  48. @Baltus

    Another poster from the “fantastic county” of Kent?

  49. Phil

    You actually have upweight the under-25s sample as I did to get comparability with previous polls, which gives you the 10 point lead (probably). It’s all a bit rough because I’m working back from rounded percentages and I don’t know exactly how YouGov do the weighting etc (I know they weight men and women separately in each age group for example). But it does give an idea how this group can throw out the topline figure.

    I should have mentioned that there’s yet another confounding[1] problem with the under-25s. They are usually the group least like to give a definite VI – often over 30% will be Will Not vote or Don’t Know as opposed to the low to mid 20s figure you generally get from the panel[2]. This gives the preferences of those who do have one even more weight.

    Martyn

    Thanks for the compliment … I think. ;)

    Howard

    Possibly so in that case, but remember that once you start weighting the figures it’s ‘less’, even if you then round to the nearest integer.

    [1] Both mathematically and as a very mild swear word.

    [2] As it happens that wasn’t true in this particular poll, possibly due to all those Scots. As we know Old Middle-Class Scotsmen think they know everything.

  50. ROGER MEXICO

    We do!

    It stems from having such close links to the Manx (or possibly lynx to the manks). :-)

1 2 3 4