More Miliband polling

The cause of today’s polling excitement are MORI’s questions asking respondents to compare Ed Miliband and David Cameron, results here. Briefly put, David Cameron enjoyed leads over Ed Miliband on most measures, often by a long way. He led on being eloquent by 59% to 15%, on being Prime Ministerial by 57% to 17%, on being tough enough for the job by 54% to 18%, on being smart enough for the job by 54% to 22%, represents Britain by 46% to 26%, on being fun to meet in person by 34% to 21%, likable by 38% to 29% and a good person by 35% to 30%.

Miliband lead on understanding people like me, by 36% to 26%, and protecting British jobs, by 37% to 31%. The two men were pretty much neck and neck on having the right values.

The fact that people think David Cameron is better suited to the job of Prime Minister than Ed Miliband is not particularly new. Cameron has a consistent lead on best Prime Minister, as we saw in July on PM preference Miliband trails a long way behind Labour’s position in VI. There was a Populus poll earlier this week showing even a majority of people who thought David Cameron was doing a bad job as PM would still rather have him in the role than Ed Miliband. All this new poll helps us to understand is some of the reasons why… and again, the picture is in line with other polling about Cameron and Miliband’s respective strengths and weaknesses. Ed is better on understanding ordinary people, but trails badly on being Prime Ministerial or being strong.

What does it actually mean though? As I wrote in July, people’s answers to this are very much coloured by what they would like to be true. I see an awful lot of Labour supporters trying to convince themselves that how voters see the leader is an irrelevance, and an awful lot of Conservative supporters trying to convince themselves that it is impossible for people to actually vote for Ed Miliband and he will be a fatal block to Labour’s chances. As ever, I expect both ends of the spectrum are wrong in their own ways.

Unfortunately, the evidence on which one is closer to the truth is not cut and dried. The last three British Election Studies (the major academic study of why people vote at British general elections, based on extensive parallel face-to-face and online polling and key driver analysis of the data) have consistently shown that voters’ opinions of the party leaders is a significant factor in deciding how they vote. It certainly convinces me, and I would have thought it almost a statement of the bleeding bloody obvious that perceptions of the party leader colour people’s perceptions of the whole party and, therefore, influences votes. However, it would be wrong to say that all academics agree on this – it is a controversial subject and some argue the opposite.

What causes me more pause for thought is the fact that opinions of party leaders are, as it were, already factored into the price. People don’t rate Ed Miliband highly as a potential Prime Minister… and yet they are telling us they would vote Labour. Clearly it can’t be putting them off that much. The question here – and again, it is one to which there is no good answer, is whether the issue will become more important as we get closer to a general election. It is a reasonable hypothesis that people answering opinion polls mid-term (and voting in mid-term elections) are largely registering a protest against the incumbent government, whereas once we approach an actual general election it becomes more of a comparison between two alternative governments, parties and Prime Ministers. If that were the case, Ed Miliband’s ratings now wouldn’t necessarily matter much, but could become increasingly important as the election approached.

I don’t particularly expect to cause many pauses for thought here, I’ve read enough comments to know it is one of those issues where people believe what they would like to be true. I shall leave, therefore, with the historical example that is nearly always cited in discussions like this.

In any conversation about this issue, the topic of Margaret Thatcher is brought up. Mrs Thatcher wasn’t particularly popular as Leader of the Opposition, while Jim Callaghan was comfortable and avuncular and likeable. He pretty consistently outpolled Margaret Thatcher on who would make the best Prime Minister. It certainly shows that people can and have voted for the less popular “Prime Ministerial candidate”. It does not follow, however, that it doesn’t matter. How much better would the Conservatives have done with a more popular leader than Thatcher in 1979? How much worse would Labour have done with a less popular candidate than Callaghan? We can’t tell.

157 Responses to “More Miliband polling”

1 2 3 4
  1. Ed stood for Labour leadership and won. He didn’t stab anybody in the back. You might as well say that David was attempting to stab Ed in the back. It makes no sense and loyalty has nothing to do with it.

    It’s a just mud slinging.

  2. @TOH

    `Don’t get me wrong I am delighted EM is Labour leader`

    Am sure you`ll be even more delighted when he is PM

  3. @ Tinged Fringe

    “They uploaded a video to Youtube/I believe it was also shown on TV of Nick Clegg apologising for tuition fees.
    But after the video was shown, the LibDems made it clear that he wasn’t actually apologising for tuition fees but he was apologising for making a promise that he knew before the election that he couldn’t keep.

    It’s quite a big shift in narrative – before this point, there’s been no hint at an apology because ‘the promise would only count if we had a majority but we’re in coalition’.”

    Thanks. I had no idea. I don’t think that’s a particularly wise move politically.

  4. @Chordata and Brian

    Thanks guys. I should perhaps add a few things…

    1) Obviously I accept, as I said, personality IS a factor. Why am I arguing it isn’t particularly a factor in the Ed vs Cam debate? Because Ed’s main battle is with Clegg.

    2) But there’s another reason. Of course, Clegg’s polling is through the floor, like Osborne’s, and that is much MORE of a factor. My main point really in this regard, is that big popularity isn’t as big a deal as real UNpopularity. Hence negative campaigning. Hence the popularity of Churchill, Wilson, Callaghan, Major and Cammers is no real guarantee of success. Real unpopularity is a bigger deal

    3) So in theory, Tories might hope that IF Ed proves unpopular enough, it might be of benefit, BUT, with Ed’s main battle being with Clegg, who crucially is already on the floor, that’s a lot to hope for.

    4) Worse for the Lib Dems, Clegg’s position is almost irredeemable. Because of the broken promises. It’s one of the worst things you can do, because even if you repent, people may not believe you. Osborne is in a better position in this regard, since he hasn’t been guilty of such flagrant betrayal.

    5) Of course, that’s all based on the current situation. LDs could change their leader to someone more popular, and then it may become a bigger deal if Ed becomes more unpopular, maybe through gaffs or a media assault. Additionally, coalition may pull some policy rabbits out of the hat, or… relax economic policy, or exports to nonEU may continue to grow significantly, or… events dear boy.

  5. And yet,and yet,remember the”curse of Cameron “at the Olympics and the
    booing at the paralympics.The financial times are reporting that Eds leadership
    Qualities(or lack of them)will be their focus leading up to the next general
    election.I think this may have already begun,hence Michael White warning
    “young Ed ” and”Milliband junior”that the present poll lead is nothing but mid
    Term blues!
    By”their” I mean the Tories not the FTs.

  6. Just watching the Scottish Budget Statement.

    It’s going to be a long one if he’s intent on reading out all items equivalent to 0.003% of overall spending; “£1m to the National Museum of Scotland”, “£1m to Scottish elite athletes”….


    Thanks for the responses on pollsters and marginal states.

  8. @JimJam

    I’m not sure, if you look at Cameron’s stand-alone personal ratings, that he’s as popular as you think. Admittedly, he’s coming out ahead of Miliband in like to like comparisons on most of the personal qualities that Populus and MORI put to their respondents, but it would appear that the public, just as they were in May 2010, are still fairly underwhelmed by him.

    Now, some would argue that his position vis-a-vis Miliband is all that matters considering that is the likely personality contest in May 2015, but I think the fact that he’s not a wildly popular or liked political figure in gross terms is an important factor here that a lot of people are overlooking. If he was trouncing Miliband AND proving hugely popular in his own right, then it might be game over but, as it is, he’s on fairly shaky ground himself when it comes to his own personal standing in the country.

    Accordingly, I think the situation is very fluid and, as events occur, could turn around quite quickly.


    Very interesting response and I don’t doubt most of what you have said and in fact agree with most of it but I do think public perception (rightly or wrongly) will either be key to who wins the keys for No 10 or at the very least determine if Ed can muster a majoirty if he indeed becomes the next PM.

  10. @Rob Sheffield

    You were equivocating about Miliband’s future at the start of the year, when you predicted that Labour would enjoy a steady poll lead of around 5% during 2012.

    Given that the poll lead has been double your prediction for a long time now, and might even be edging up a tad, I was hoping at some time to see a post from you gushing praise on Miliband for leading Labour to a polling position that far exceeds your expectations.

  11. @SMukesh

    I would not wish that on my worst enemy let alone this country.

  12. PHIL

    Careful now, Antony might moderate you if your head gets any bigger!! ;)

  13. the labour leader and the rest of the shadow cabinet must absolutely resist the language of the think tank when seeking to speak to the public. For example , the term predistribution ought not to be bandied about. Ed
    Miliband must speak to the television audience in simple terms , in ways that are cogent and straightforward. In other words he must keep it real at all times.

  14. @TOH

    `Calm down dear.`(in a DC voice).

  15. Anthony in his piece didn’t link to the Evening Standard article that the Ipsos-MORI poll fed into and for which the extra questions were commissioned. No doubt, as an old fashioned sort of guy, he believes that “If you can’t say something nice, don’t say nothing at all”. However the Manx are not famous for their tact, so here it is:

    And it’s a little love – especially the start of commentary by Joe Murphy headed: PM’s battered image given heroic shine by comparison

    Just as his own MPs are questioning David Cameron’s leadership, Ipsos Mori’s research today gives the Tory leader some grounds for hope.

    While some Tory MPs call him weak behind his back, voters think him tougher by far than his Labour rival. On almost every measure, his battered image takes a more heroic shine when compared with the Labour leader.

    This will reinforce what Mr Cameron’s own pollster, Andrew Cooper, has been telling him for months: “Don’t worry, Prime Minister, the voters won’t let Ed Miliband become Prime Minister.”

    To be fair both the article and commentary do have some of the points made above, but the general tone is basically fanboy because of the way the personal quality questions were posed. And nowhere does it seem to point out that the standard leadership approval ratings, as asked for decades and from the sample, are Cameron -24 with Miliband doing much better on -9

  16. @Nick P

    You need to follow Smukesh’s advice, not me!!

  17. Anyone know how much the new Nick Glegg CD will cost?

    Thinking of including it in my christmas shopping list!!


    There is no upfront ‘charge’ for the CD. You just have to sign a pledge to say you will make a donation to charity…….. you can guess the punchline – too obvious to type!


    Ha Ha Ha :)

  20. @Old Nat, SoCalLiberal

    There is a Wikipedia page for Rasmussen Reports… accurate polls in 2004 and 2008, inaccurate in 2010.

    They detail how Nate Silver (fivethirtyeight/nytimes) for instance, initially praised, then subsequently became critical of the Rasmussen methodology.

    Interestingly this site provides Rasmussen-free maps – the Presidential race state-by-state excluding all Rasmussen polls:


    At the opposite extreme there is a site called UnSkewedPolls which “unskews” polls, they say, using Rasmussen partisan data. Their average currently shows Romney with a 7.8% lead over Obama.

    The fact that Rasmussen Reports conducts a presidential poll every day means that their figures (Obama lead 2% as of today) will be always be included in running averages such as the RCP average (RealClearPolitics), which currently has Obama on a 3.1% lead.

    No doubt the controversy will be given fresh impetus when state/national polling is compared to the actual results in November.

    Incidentally Scott Rasmussen (who was a consultant for the 2004 Bush campaign) was interviewed on Today earlier this week when he made it clear that he thinks Romney is a poor candidate.



  22. Can someone please shoot nick!!

  23. @richard in norway

    You are the Lib Dem Nadine Dorries and I claim my five nicker.

  24. @BILLYBOB how can you be so sure? Is there only one left?

  25. Billybob


  26. @ Richard in Norway

    So glad to read you… I’ve been worried by your absence :)

  27. Sue

    I’ve been lurking for a while but the musical apology from clegg just had to be commented on, BTW what is omnishambles

  28. @richard in norway


    I claim my five pounds… it’s an old joke, nevermind :( .

  29. Billybob

    Ahh, well hopefully that’s the only thing I have in common with nadine.

  30. @ Old Nat

    “Thanks for the responses on pollsters and marginal states.”

    You’re very welcome. Here’s my long answer that I kinda forgot to give you last night.

    The conventional wisdom in this race is that all the states Kerry won in 2004 are safe Obama States (except maybe New Hampshire), any states McCain won in 2008 are safe Romney, and any states won by Bush in 2004 but won by Obama in 2008 are swing states (except Indiana which has been assumed to be for Romney). I don’t agree with the conventional wisdom.

    I believe that the achilles heel for Obama is the upper midwest (the rustbelt to a larger degree) and that includes states that Kerry won in 2004. Any state in the upper Midwest therefore, I have considered in play except for Illinois, which is safe Obama. Those states are Pennsylvania, Ohio, Indiana, Missouri, Iowa, Minnesota, Wisconsin, and Michigan. (Pennsylvania isn’t really midwest but it is part of the rustbelt).

    Why do I say this? Obama absolutely swept this region in 2008, only underperforming in Minnesota and Ohio. But the voters there never were really in love with him. They voted for him because the economy was terrible and Obama is a midwesterner like they are and so they felt comfortable with him even if they didn’t love him. Fastforward to 2010, Republicans absolutely destroyed the Democrats in these states. The economy was still bad, people were still unhappy, even many who had excitedly voted for Obama in 08′ were disappointed. Many simply stayed home.

    The second group of states in play are southwestern/sunbelt states (except for Utah of course). Those states are Arizona, Colorado, Nevada, and New Mexico. Obama won the latter three by double digit margins in 08′ with shocking swings given that they voted for Dubya in 04′. The economies in those states are terrible, absolutely devastated by the housing bubble. They’re vulnerrable and in play for those reasons alone. They’re also in play because he relied on Latino voters who he initially struggled with but improved his standing. But unlike black voters who are intensely loyal to the President, Latino voters are not as naturally loyal. Also, Obama won 67% of the Latino vote in 08′ compared to 95% of the black vote. Now, I am of the opinion that all minority turnout will be even higher in 2012 than it was in 2008 (people don’t believe me but we’ll wait and see). And, there has been a major swing to Obama amongst Latino voters (a plurality of Latino McCain voters in 08′ are now planning to vote for Obama in 2012). Still, those states are in play.

    Where Obama has held up well is amongst his nationwide Bradley Coalition voters. Affluent/upper middle class whites and African Americans have pretty much stuck with this President through thick and thin. So in states where those voters can comprise a majority or a plurality (even if they’re swing states), I think the President is in good shape.

    So, this sounds strange but I am far more confident over the President’s chances in North Carolina than in let’s say Minnesota even though the President has a fairly decent lead in MN and a 1%-2% lead in North Carolina. They are swing states and in play but I remain convinced that Obama will win Virginia and North Carolina. I also think that Georgia may be in play (depending on how well Obama does nationwide) due to demographic changes.

    Other states in play are New Hampshire (the crazy aunt of New England), Florida, and Montana (again, depending on how well the President does nationwide). And polling (after Romney’s disgraceful Libya and Egypt response) shows Obama starting to move into a double digit lead nationwide. The polls aren’t yet really starting to reflect reaction to the 47% comments.


    Thanks for that more detailed analysis. I’ve bookmarked it so that I can compare it with the results!

    Hope you are right about NC!

  32. It soon became apparent that many Parliamentarians who are able to influence tax laws have taken up positions as directors and non executive directors in major companies with offshore links.

    There are 28 Tories – six of whom are MPs -16 Labour peers, three Lib Dem peers and another 21 are either crossbench or non-affiliated peers.

    I expect this to have absolutely zero effect on polling – but attitudes to all politicians may sink even lower – along with turnout.

  33. Labour lead back to 6:

    CON 35%, LAB 41%, LD 9%, UKIP 7%; APP -38

  34. @ Richard in Norway

    Glad to see you posting again. I like Nadine Dorries (though as my mom has said of the Palestinians, “[she] probably [doesn’t] like you”).

    Btw, did you hear the good news? Elizabeth Warren is now leading in her Senate race with nearly every poll showing a trend in her direction. Her primetime speech at the DNC was spectacular and I think it gave her a real boost. And now word has broken that Tom Menino, the Mayor of Boston, has endorsed her (apparently took some arm twisting from both President Clinton and President Obama…oy vei what a mess). He will unleash his powerful political machine on her behalf which will dramatically help her.

    Apparently in a debate with Scott Brown tonight she did well.

    h ttp://

    h ttp://

    His attacks seemed whiny and desperate to me (but I’m never a good judge of debate performances).

  35. so the Tories bounce back to practically their 2010 level.

    What was it? Shooting of police? Negativity about Milliband?

    A blip?

  36. NickP,
    45 followed by 41 indicates a 42-43 for Lab, and 30-35 indicates Cons on about 32-33.
    But just wait for today’s gloating from the Tory commentariat.

    I found it amusing that the narrative for the Populus +15 lead was that “monthly polls are too infrequent to rule out outliers” from the same people who ignored the 10-12s last time there was a 6% lead to shout from the houses about it (even though clearly that was outside of the trend).

    Not that this only happens on the right – left-wingers jumped all over the 15% lead, declaring that Ed had reached the point he needed to in order to win 2015.

    Which to be fair, is partly in response to the ever-changing goal post of how far Labour need to be ahead ‘by this point’ in order for there not to be an epic Tory landslide in 2015.
    When the gap is 5%, it should be 10% – when it’s 10% it should be 15%.. when it’s 15%, it should be 20%, etc

    Very similar to the changing narrative of how we decide who is going to win in 2015 – when the Tories were ahead in VI, that was good enough. Then when Labour moved ahead on VI, it was all about approval ratings. Then when Ed moved ahead on approval ratings, it was all about who was Prime Ministerial enough.
    I assume if Ed moves ahead of Cameron on Prime-Ministeriality, there’ll be another criteria to show that Labour will get slaughtered in 2015.

    And I’ll take a bet now (for £0.00) that if that day comes, the most masculine qualities will be bandied about as the ones which prime ministers have (just something I’ve observed generally of the commentariat).

  37. Con lead in 18-24 group!!!

    Doesn’t look right at all.

  38. Colin
    Slight Lab lead in the over 60s too.

    A quick question on gender VI differences – has anybody been tracking it?
    AW once did a post (when the Tories fell behind at a previous time) and at that time, the Tories were behind because female voters were largely against them.
    The latest YouGov poll has a similar gap, but I was wondering if it was a trend or not.

  39. CON 35%, LAB 41%, LD 9%, UKIP 7%; APP -38

    A couple more polls like this and YouGov might become the new Gold Standard of pollsters.

    Along with ICM, of course! lol

  40. 30-poll average prior to (this poll):

    Con: 33.3% (35)
    Lab: 43.1% (41)
    Lib: 9.2% (9)
    UKIP: 7% (7)

    At a UK level it’s within MoE. Larger than average sample as well. Not touching the crossbreaks, as I know they will be way off (e.g. Con on 21 in Scotland). Three in a row, and we can believe it.

  41. YG polls have been all over the place recently, after a period of them being pretty consistent.

    Why ? What is making them eratic.

  42. If I was the Sun I would find another polling company. MOE etc is an OK excuse once in a while but 4 times recently I don’t think so. How likely is it that the extremes of the margins are hit 4 times in a few weeks? The poll is telling me nothing about the state of the parties it is too unreliable.

    I believe you gov uses panels? Maybe they need to look at the make up of their panels.

    Also regarding this methodology – Doesn’t being asked questions about politics change the persons behaviour so that they pay more attention to politics and politicions and therefore results do not reflect the general public especially mid-term.

  43. There is a tendency for people to underestimate the side of normal margins of error. Looking at YouGov’s fourteen polls so far this month, the average Conservative score is 33. All fourteen polls have been within 2 points of this. Eleven have been within 1 point of this.

    The average Labour score has been 43. All fourteen polls have been within 3 points of this, and thirteen of them have been within 2 points of this. Eight have been within 1 point.

    The average Lib Dem score has been 9 points. All fourteen polls have been within 2 points of this, thirteen have been within one point of it.

    Therefore, assuming that actual underlying support hasn’t changed at all during September, there has only been one single poll at the extreme margin of a 3 point margin of error, and the majority have been within 1 point of the average… which is the pattern you would expect.

    Quite simply, you are expecting precision that polls of 1500 or so people are mathematically incapable of delivering.

  44. Anthony

    Fair points made. It was just that for a period earlier in the year, the polls were pretty static, with poll leads for Labour being about the same every day. We currently have polls within a few days, showing 6 and 12 point leads.

  45. R Huckle – just luck! The current sort of spread in the polls is statistically what we should expect.

  46. It’s possible that when leads get a bit bigger, things fluctuate more?

    So things were pretty solid when the lead was 5/6 points but more erratic when 9/10?

  47. I take it that this poll has a MOE of +/-3. So today that parties could be tied or Labour is 12 pts ahead. But two days ago – Labour was accoding to MOE 6 – 18 pts ahead. So Labour is somewhere between tied and 18 pts ahead?

    So what can we take from that ‘Labour is probably not behind’?

  48. Arguably…14 % points of the LD VI at the last GE has moved to Lab, and some 4 % points of Con VI has moved to UKIP. (I realise this is a massive simplification.)

    So, it is worth considering what effect the following will have:

    – an electoral pact between UKIP and Cons, as being promoted currently by Farage, and
    – some of the former LD voters returning to LD for TV reasons.

    We could potentially see Cons and Lab having the same share of VI – broadly 38 – 39%.This broadly results in hung parliament with Lab five seats short of an OM.

  49. NickP – statistically that is partly true. Margins of error are bigger the closer a figure is to 50, so there is actually a bigger margin for a party on 45% than there is for a party on 31% (and hence, you are going to get more volatility when the party split is 43-33 than you would if the party split was 36-40).

    However, while mathematically true, the effect of that is so small that I doubt very much we’d actually notice the higher volatility due to it.

    Couper – there is no point looking at the polls in isolation. If we only had today’s poll to go on then we could only conclude that Labour were between 0 and 12 points ahead. But we don’t. We have lots of polls to go on, so we can be fairly confident that Labour are around about 10 points ahead.

1 2 3 4