On poll movements

This morning’s YouGov poll for the Sun again showed a Labour lead of seven points – CON 32%, LAB 39, LDEM 11%, UKIP 13% (that’s Tuesday MORNING’S poll, btw, Tuesday evening’s isn’t out yet!). Five of YouGov’s last six polls have shown single figure Labour leads, whereas previously the average Labour had been consistently around 10 or 11 points. Put in the context of the falling Labour leads from ICM, MORI and Opinium it is pretty undeniable that something is afoot.

YouGov’s average figures in the first half of April were CON 31%, LAB 41%, LD 11%, UKIP 11%
The average over those last six polls is CON 32%, LAB 40%, LDEM 11%, UKIP 12%

So roughly speaking we appear to have had a small increase for the Tories, a slight knock for Labour. At this point we can normally expect lots of speculation about what has caused it… or more typically, lots of people claiming that the thing they personally care deeply about has caused it, the thing they think their party shouldn’t be doing has damaged them, or the thing they think their party should be doing has helped them. Normally such claims don’t bother with evidence.

The harsh truth is that we usually can’t really tell what has caused a movement in the polls. Sometimes there is an obvious event that coincides with a big shift in the polls which, while it doesn’t prove anything, does strongly imply a connection (after all, we can’t be sure that the big drop in Tory support in March last year was definitely due to the budget, but it would be a remarkable co-incidence if it wasn’t!). Other times there are all sorts of plausible explanations.

The most obvious explanations for the current narrowing relate to Margaret Thatcher’s funeral. That could impact the polls in terms of lots of positive retrospectives about Thatcher in the media… or could have an indirect effect in the sense that it interupted the normal flow of politics. David Cameron got to spend a week or two looking statesmanlike without the normal dirty business of politics and governing. However one could equally look at other underlying factors, the welfare debate for example, perhaps a generally more focused presentation by the government since Lynton Crosby returned, some figures from the Blair era apparently criticising Ed Miliband. All these things add up.

My own working assumption is still that is it is a Thatcher effect of one sort or another that will fade away, but it really is impossible to know. We shall have to wait and see if it lasts.

UPDATE: The Sun Politics team have tweeted tonight’s results – CON 33%, LAB 40%, LD 10%, UKIP 12%

482 Responses to “On poll movements”

1 2 3 10
  1. There’s not a realistic scenario where Labour (a) get more than 40% of the vote and (b) don’t end up the largest party and probably with at least a small majority.

    So in this respect the Labour VI is more significant than the Tory VI: the Tories winning over a lot of UKIP support is a necessary condition for Labour not winning in 2015, but not a sufficient one.

  2. Crap game generally and i think there will be a one percent shift to UKIP on the basis of too many europeans in the Champions League.

  3. Could, it be a non event which has shifted the polls?

    I.e its been a few weeks now without any government screw ups or scandals. Are polls merely reverting a little (albeit not in full) to how they would be without all the constant screw ups and scandals?

  4. Jolly good poll [playing it safe again]

    The number seven is considered magical.

    I think that’s ‘cos of Snow White but am not certain.

  5. Actualy AW I think that is a pretty fair analysis.

  6. First

    (to get moderated?)

    I just hope this thread isn’t as lousy as the last one.

  7. Oh sod it. I’ve just spent 20 mins mis-typing the following into my iPhone with my man-sized fingers, so I’m going to re-paste it here.

    AW. Given that you’ve just posted more or less identical comments in the post for this thread, we’ll now that you are on a mission to eliminate pinko opinion if you mod this. We’re watching you…~Roberto Di Niro, “Meet the Fockers” stylee two fingers pointed to eyes then to you~


    The Lab VI fall (such as it is) appears to have started back in mid-Feb. back then, the 5-poll rolling average was about 43 for Lab. Currently it’s about 40. I don’t recall anything horrendous happening to Lab back then, but I do have a vague recollection of a string of undisastrous economic news. Could be that, compounded by the more recent issues that you mention.
    As for the Thatcher effect, there’s no proof of causality, but there is a strong correlation between the media carpet bombing campaign a couple of weeks ago (The DT headline “I Vow to Thee My Country” on the day the funeral hymns were announced was just the most egregious example of the line between reporting and editorialising being violated by serious national dailies’ front pages) and a 2-3 point uptick in the Con VI.
    It could be other things that have caused the Con VI upward shift. It coukd be one of those eye-catching initiatives announced in the last two weeks, or that tsunami of excellent economic news.
    Me. As ever, I’ll be guided by Occam’s Razor. But I’ll be delighted if the Tories convince themselves that their VI uptick is due to some endogenous issue.
    We’ll see. We usually do, even though what we see never seems to be quite the same thing.

  8. Sun Politics [email protected]_Politics

    YouGov/Sun poll tonight: CON 33%, LAB 40%, LD 10%, UKIP 12%. Lab lead still 7 points. Is the Maggie bounce permanent?

  9. It’s a moment in time with quite a lot happening.

    To be honest, since this government has taken Blair’s advice to hit the ground running very much to heart, there is usually quite a lot happening.

    And quite often the breakneck pace leads to the odd snafu = even more happening.

    In the last few weeks we’ve had seasonal extras on top like the budget, welfare changes kicking in, and extraneous stuff like Thatch, and Miliband taking some friendly fire.

    Which naturally makes it hard to pinpoint what the key factors are in terms of affecting VI.

    I think overall you have to consider the possibility that much of what has happened in the last few weeks has been favourable to the government and Labour might breathe a sigh of relief the hit so far has only been around 3%.

  10. Government (dis)approval was consistently in the minus-mid-30s for weeks on end, until the last week or so when they suddenly improved to an average of around -30. Tory VI correlates extraordinarily well with government approval, so the narrowing of the Labour lead is no big surprise.

    I leave people to work out for themselves what event might have caused an abrupt shift in government approval.

  11. I still am unsure it’s a Thatcher bounce. Are people nervous about the talk, including from Blairites, about a left wing manifesto? Ironically I do think the New Labour split (if I can push to call it that) probably hasn’t helped. As its those voices that the papers publicise most.

  12. So the Sun also views it as being down to Thatcher. But to matter, “permanent” would need to be two years, not one or two weeks.

  13. I guess most of us here are fixed in our poitical leanings. Because of that I think its hard to figure out exactly why those who change their VI actually do so. I think its because there isn’t really a reason generally: they sort of pick up stuff from the airwaves, newspapers and change intuitively – although that’s probably too grand a word: thoughtlessly is probably better, if a bit dismissive.

    And I mean that about any change in their VI.

    This will take a while to settle in my view.

  14. What happened to OleNat? He seemed to take umbrage at something and sulk off.

  15. Colin
    Just spotted your point regarding the RCN.

    The situation between the Royal Colleges is not the same.

    The RCN is by far the biggest and for decades the only Nursing Trade Union a job it actually does rather well despite Francis’s comments, it is non affiliated and hasn’t called a strike since it’s foundation nearly 100 years ago.

    It is unlikely that it will turn round to its 500,000 members and say sorry we are going to be exclusively a professional body you can all go off and join Unison.

    The Francis report did not as you suggest say that the RCN should cease to be a trade Union it said the RCN should consider separating the Rolls of Professional Organisation and Trade Union which is not the same.

    The attacks on Nurses,Teachers and Police Officers Trade Associations I suspect is part of the Conservatives Policy in relation to diverting criticism.
    It is possible it might have had some benefit short term in relation to VI but how it will play at GE with members of these organisations and their families remains to be seen.

  16. Yougov poll for the Resolution Foundation. Dunno what that is but here’s a link to the article:-


  17. London Westminster seats: Correlation between White British % Census 2011 and Labour % GE 2010?https://twitter.com/Sunil_P2/status/326658870867013633

  18. @Spearmint

    Thank you for the graphs on the last thread.

    The fall off in the week-long peak of the LDs (much commented on here at the time) at the end of March is worthy of note. From just above 12% then to 10.4% currently (including tonight), on a rolling basis.


    Most of the drop predates Thatcher’s death.

  19. Steve

    The attacks on Nurses,Teachers and Police Officers Trade Associations I suspect is part of the Conservatives Policy in relation to diverting criticism.
    But not the Bar Association

  20. paulcroft

    What happened to OleNat? He seemed to take umbrage at something and sulk off.
    Perhaps if the yes vote picks up he will return

  21. Old Nat kept going into moderation; he thought it was AW but I think it was mostly auto-mod because it was often well past Anthony’s bedtime when it happened!

  22. The Sun Politics team have tweeted tonight’s results – CON 33%, LAB 40%, LD 10%, UKIP 12%
    Labour supporters ought not to mind CON clawing back a few points whilst Labour stays on 40+; dropping to 39 in Monday’s poll would’ve worried them, even though MOE suggests it will happen every now & again whilst Labour are in 40-42 territory.

  23. Thatcher, followed by Boston. News focus elsewhere and elsewhen resulting in a temporary alleviation of ardour in regard to expressions of voting intention, and a corresponding narrowing of the margins…

  24. Like Amber says, it’s all about the Labour VI.

    Playing around with the swingometer at the top, even if the Tories manage to retain all of their support from 2010 – which we have to agree seems unlikely – at 36.1%, Labour only needs 1.1% more of the vote to have a (very) slim minority of 2.

    Now, let’s say the Cons actually get 35% and Labour manage just 36%, they’re only 2 short of a majority.

    Unless the economy turns around quick time and/or Labour explode there just doesn’t seem to me a way that Cons can realistically expect anything more than a hung parliment, more than likely leaning towards Labour.

  25. @Stec

    The alternative theory is that oppositions need to build around a twenty percent lead at some point because things tend to swing back to the government towards the election. .. incumbency factor, plus generous pre-election budgets, policies starting to pay off etc.

    But at the same time, we are in new territory with the rise of UKip, with the second party in the coalition having alienated half their voters thus uniting the left, and so far, government policies don’t look like yielding as much fruit as planned and austerity ties their hands a bit on freebies.

    2015 will therefore potentially be interesting. .. it may support the 20% thesis, .. but if Labour hold the Lib Dems and Tories don’t advance on their 2010 vote (another thing that doesn’t often happen, governments improving their vote in the second term) then the 20% thesis gets blown away.

  26. If we take a closer look at the 20% thing, then what causes things to swing back to the government tends to be one if two things.

    Either the opposition adopting rather unpopular policies – eg the unilateral thing in the eighties, or Kinnock saying he’d raise income tax in 92 – or else the government pulling a rabbit out of the hat – eg Falklands.

    At other times the government screws up and never really receivers, eg winterif discontent, ERM etc.

    Generally one of those three things has been likely to happen, hence supporting the 20% thing. But if the opposition doesn’t screw up and there are no rabbits for the government. …

    Modern oppositions know better than to leave policy hostages to fortune. There is unlikely to be a suicide note. So it rather depends on the rabbits really.

  27. I should add that there is a big question mark as to whether the 20% thing applies to a coalition government anyway.

    Tories have had to compromise in some policies, shoving some suooorters into the arms of UKip, and perhaps limiting the Tories in policy terms in the run up to the election. Meanwhile lib Dems had to not just compromise but trash key promises.

    The difficulty here is that it’s not really tenable for Tories to win over the Lib Dem refugees… they need the lib Dems to win them back. But how do Tories enact policies to do that without alienating their own voters?

    The dynamics of coalition government, may mean 20% isn’t necessary. There are just too many disappointed people and if you assuage the supporters of one party you alienate the other party’s supporters.

  28. “Labour supporters ought not to mind CON clawing back a few points whilst Labour stays on 40+; dropping to 39 in Monday’s poll would’ve worried them, even though MOE suggests it will happen every now & again whilst Labour are in 40-42 territory.”
    My 7-Day weighted average now has
    Con 32.2, Lab 39.9, Lib 10.4, UKIP 11.6

    Before Thatcher’s death had:
    Con 31.1, Lab 41, Lib 10.7, UKIP 11.3

    So changes would be
    Con +1.1, Lab -1.1, Lib -0.3, UKIP +0.3

    I expect, if the lead has switched to 7, it’ll be closer to Con 33, Lab 40 by this Sunday’s poll (excluding any possible GDP effects), which would effectively mean Con +2, Lab -1 (obviously this estimate is complete conjecture on my part, wait for the figures).

    So Lab supporters should be a little worried that all of the transfer (so far) seems to be Lab>Con. So without regaining any of their lost UKIP voters.
    That potentially means that the Lab vote will be squeezed three ways (to UKIP, Lib and Con) as we approach the election, which would reinforce a ‘mid term blues’/’Ed Miliband is the problem’ narrative.

    But given that Lab scored just 29% in 2010, even if they get squeezed another 5%, +6% is a pretty good jump in the polls and potentially (thanks FPTP) enough to guarantee him position as PM (even if the Cons regain all of their 2010 vote).

  29. Carfew – also the 20% thing has a lower level of support for the Governing party than we are seeing for the Tories in this parliiament as the more liberal discontents who may give LD as a VI but never Labour or UKIP are unlikley to go to the LDs as they are in Government also.

    That is why as many of us say it is all about Lab VI which has clearly fallen 3% or so whichever pollster you look at over th last 6-7 weeks.

    I do think there was a small slippage to UKIP in early March after Eastleigh which I think will stick now right through to the GE itself but virtually all returning then adding 1% or so to Lab VI.

    There also appears to have been a further 2% or so slippage in the last 2-3 weeks and that is where speculation depending on our outlook, as per Anthony, becomes interesting.

    Finally, Labour has never reached 45% on a series of polls (or similar with the adjusted ICM) but 43% was a steady level on YG for a while.

    This is why imo we will be a little short at the GE of an OM and the most votes will be close. I think there is every chance of Labour having most seats with 35-37% of the vote share.

  30. Good Morning All. I think that things seem to be a bit easier on the car fuel price front. The Conservatives seem to have ‘won’ the argument on welfare, against the protests from Labour.

  31. @jimjam and tinged

    Another way of looking at it is… what is Labour’s core vote. The vote that would only be shifted by a really significant rabbit.

    Conventionally one might say it’s 29%, given that’s what they polled in the election despite being on the floor.

    But since then, Labour have received an influx from LibDems. A goodly chunk of these are not easily shifted to Tories given it was coalition with Tories that saw them flee, and hard for Lib Dems to get them back given the betrayal.

    In effect, a chunk of the LDs have become part of the Labour core vote. And Labour are currently not far off that core vote, which may be somewhere around 36-37%.

    It’ll be increasingly difficult to shift Labour below that and more likely to need a rabbit .

  32. Yes Carfew,
    This is where the misrepresentation of the so-called 35% strategy comes from.
    Hodges et al have suggested it is Ed target but it is the base line and it is right that unlike Blair/Brown the core is not alienated; I know 2 former voters who stay away is only worth one swicther but losing touch with our base cost us in 2005 and 2010.
    So whilst we reach out to try to get a few extra we don’t over triangulate as we the new core has to become entrenched.

  33. Good Morning All. I think that things seem to be a bit easier on the car fuel price front. The Conservatives seem to have ‘won’ the argument on welfare, against the protests from Labour.
    If we see coverage of families made homeless and hungry by the move, while the government pushes on with it’s spare home subsidy , which seems a distinct possibility over the medium term I think that the argument will then have been lost.

    However, time will tell.

    40% for Labour isn’t actually a bad result spot on in the middle of the average for the last 2 Years

  34. I think I am right in saying that the UK opinion polls do not include Northern Ireland, therefore it is wrong to compare them to UK-wide GE vote shares (C36% L29%). They should be compared to the mainland UK GE vote shares (C37% L30%).

    It doesn’t make that much difference, I grant you, but the cognitive dissonance is killing me…

  35. In other news, I filled up my car yesterday and petrol was actually 1p dearer than diesel. Has that ever happened before? Is it something to do with the demand for heating fuel on the continent or something? (I’ve heard something like that said about diesel supplies in the past).

  36. In other news, I filled up my car yesterday and petrol was actually 1p dearer than diesel.

    -It’s 7p cheaper (well less expensive) here which I think is the national average.

    Maybe it’s the filling station having a moment?

  37. I honestly believe thectory vote started to firm up just before thatchers funeral the reports on mick philpott seemed to focus people mind. It became a firm yes I will be voting conservative. I know this is unscientific but this began to happen just before thatchers funeral the tories core vote had to be about 33% so this would correlate

  38. @Katie
    “Is the Maggie bounce permanent?”

    If it is permanent then it isn’t a Maggie bounce.

  39. My sense is that this is more than just the funeral. Labour struggled a little on the welfare front, which I would agree seemed to mark the start of a slight poll lead narrowing.

    The really interesting one to watch will be to see if the current score on welfare stays in the blues favour. If we see significant negative impacts (some have predicted for example that the housing benefit changes may actually cost more if it leads to evictions etc) then Labour may feel easier.

    On the wider political front, this could in the end prove a good thing for Labour. Complacency is a killer in politics, and sitting on a solid lead must tempt oppositions not to rock the boat. A reminder 2 years out that they need coherent policies and a clear message may be no bad thing for them, and for the political debate in general.

  40. I’m not sure I would consider Dan Hodges to be the most reliable guide to Labour strategy.

  41. Today’s poll looks a lot more “normal”, i.e. pre-Thatcher. The gender imbalance is back as Labour more popular among women than men, Cons more popular amongst men than women. And the age and class breaks also look “normal” again.

    The difference over three weeks ago is that the Labour 2010 retention is lower than it was: 7% loss to Con and 7% to UKIP are both higher than “normal”.

    Just one poll, of course. The key thing is to watch how this pans out over the rest of the week.

  42. The one way in which a Maggie bounce might permanently help the Tories is if it boosts the notion that they’re the only party that knows how to sort out the economy.

    Black Wednesday destroyed the Tories credibility in this regard and Brown was able to establish a strong positive Labour reputation.Harking back beyond Black Wednesday to Thatcher may rekindle the idea that the Tories are the safer bet for running the economy–more or less the message they’ll be trying to convey in 2015

  43. I can’t see any logic for a Thatcher effect on Lab VI.
    And certainly not a week after the funeral is over.

    I thought Maggie was the Devil Incarnate Labour voters?

    I think it much more likely that the feature ROBIN identifies, and indeed all the “performance approval” ratings , represent the driver for VI change.

    ie a feeling that the Government are doing something right for a change.

    But if AW doesn’t know-what chance do I have?

  44. As the fallout from the Reinhart & Rogoff fiasco settles, I’ve heard plenty of people saying that it doesn’t really matter. It was just one piece of work and its debunking doesn’t undermine Austerity. Policymakers were basing their policies on more than just one paper from a couple of pointy-heads.

    Well look at this gem from the indomitable Olli Rehn, just 2 months ago.


    This was the infamous letter in which he said that open discussion on the fiscal multipliers was “unhelpful”. Now have a look at the last 2 paragraphs of page 3. Direct reference to R&R, with a direct conclusion that their findings mean Austerity Now is essential for survival.

    The fiscal policy for 450 million people being driven by findings from academics who can’t operate Excel. A farce too ridiculous to laugh at.

  45. Maybe Joe public is just awakening to the fact that for all the criticism of Coalition policy by Labour, at least there is a policy to criticise. Labour has no policy on anything other than a recently announced plan to undertake a massive re-organisation of the NHS.

    Several issues in the news at the moment play well. Benefits are being cut slightly, to stop the outlandish claims. I can’t afford to live in London, why should someone on benefits? That resonates as common sense, as does why should people be subsidised to live in bigger houses than they need. Nurses should nurse, that’s common sense and they do themselves no favours with the public by giving the impression that giving bedbaths etc, is beneath them.
    Teachers should be competent to teach & be rewarded accordingly. That has happened in the private sector for decades. Longer days in school? Benn the norm on the continent for decades.

    As dor fuel, it must have fallen generally as it has fallen in France to 1.26 euro (c. £1.14) per litre. Down from 1.40 euro.

  46. Looking back at Polls in 1991 and 2003 and the subsequent elections it is difficult to tell what the result will be in 2015. I think we might see the unusual result of Tories maintaining the 37% they did in 2010 especially if the economy improves by then. My gut feel is that they will struggle to get a majority but then so will Labour. My guess would be a Hung Parliament again. I don’t think a 7% opposition lead at this stage is enough.

  47. If Sir Bob is right about UKIP come the GE, then 33/40/10/12 will translate to 36/40/10/9, to put UKIP at his maximum-all other things being equal-which they won’t be.

  48. My humbled take in the combo of Maggie bounce which will have motivated conservatives and the tug of the welfare debate. It may not be fair; it may not be right; it also may not even deal with the problem but the hard words of government plays well with some critical swing voters and it clearly contextualises a problem for Labour – which interestingly finds itself boxed in by an issue in opposition on which it had a pretty good record in government in being successfully tough.

    It’s time for EM to consider a big hitting politician to take on this welfare brief. Maybe to include link to housing & house building programme in the dept – giving priority to those who work etc for access to new homes programme & make a single approach to the vexed issue of “work”; “contributions” and “benefits” on a sort of points based system. Yvette Cooper would be effective…’tough women’ often resonate well with wider public & older voting demographic in policy areas which require an edge of steel!

    Not often commented upon but Mrs T did a bloody good job for Heath in Environment between Feb74 and Oct74 – it helped her get the leadership.

    Wishing things different will not make them different…but this actually may be an issue capable of being turned to labour’s advantage with a fresh approach and a new face.

  49. Colin

    We went through this last week. When the mood changes, voters tend to change their opinions of a party/politician’s attributes across the board.
    As I pointed out last week, in the immediate aftermath of the Falklands victory, the number if people saying that Michael Foot was “inexperienced” doubled. In the immediate aftermath if the 2000 fuel protests, the number of people saying the Blair Cabinet was divided spiked.

    If you decide you like a party, you often go in lock, stock and barrel. You’ll decide you like everything from their economic policy to the way they stroke cats. Not everyone of course, but enough to smear the results and make cause and effect impossible to determine.

  50. I think the reference to Linton Crosby is correct. The latest pronouncement from IDS to the effect that there is one million unemployed on needless benefits has his stamp all over it.

1 2 3 10