One more final poll for the referendum tomorrow – YouGov for the Sun have topline figures of YES 40%, NO 60%. That’s in the same sort of ball park as YouGov’s AV polling over the last fortnight or so, a very substantial NO lead (unless the polls are horribly, horribly wrong a NO victory appears a certainty) but not as large as suggested by ComRes and ICM. The poll was conducted yesterday and today, with a sample of 5,725.

There is apparently also an Angus Reid poll on AV due out tonight – I do not know when or where.

To avoid confusion, people are also tweeting a poll from the Metro apparently showing a 4 point lead for Yes. I’ve no idea of the veracity of the poll, but it is of readers of Metro under the age of 44, so is not intended to be a nationally representative poll or a prediction of the overall result.

UPDATE: The final Angus Reid poll on AV is now out, and has YES on 39%, NO on 61%

206 Responses to “Final YouGov AV poll”

1 2 3 5
  1. Whoever tweeted the results for YouGov got the yes and no % round the wrong way. oops…

  2. I think YES still have an outside chance as long as their poll rating is on 40% or greater because differential turnout could have some very strange effects on the outcome.

  3. @Mark S
    Seriously, how could anyone have got them the wrong way around?!

  4. Mark S. Deary me – I can’t leave them alone for 5 minutes. I’ll have to login and send a new tweet for them now :)

  5. Anthony….I’m sure the #yes2av mob have already fired it around twitter many many times…I know I have! (sorry for the bias…)

  6. YG

    36 blue
    40 red
    11 yellow

    thats a 1% lead ICM, 3% ComR, 4% YG

    2.66% lead for reds in the month of May.

  7. @TGB
    We’ll see what the election results on Friday morning actually say. I reiterate Lab are arond 6% ahead. ICM/Comres always have Lab lower.

    And I think YG have it about right on the LD numbers.

  8. Tonight’s polls are good for the Tories as it would ensure that the local elections are not disastrous, as many on here (me included) predicted. Of course, local elections are notoriously hard to predict because of their reliance on turnout, but I’d say that the Tories would definitely take tonight’s polls.

  9. As I said in the ICM thread it’s worth noting for the Labour voters out there that if you look at the unweighted sample in tonights YouGov VI poll you’ll see the Tories ahead.

    May not mean anything but worth keeping an eye on at least.

  10. I haven’t been visiting this site recently so sorry if this is an obvious question: are the YouGov AV polls adjusted for likelihood to vote?

  11. I’d say anything around a 0-4% Labour lead in the local elections would be good news for the Tories and bad news for Labour; 5%+ Labour lead would be a bad day for the Tories and a good result for Labour.

    All of tonight’s polls predict the former, but I guess they could be underestimating the true lead or the turnout could also favour Labour tomorrow.

  12. @AmbivalentSupporter
    Absolutely. Having the elections on the same day is likely to be a boon for the Tories.

  13. Edward – yep. As far as I’m aware all the final polls from all the companies are.

  14. As a novice in these matters, can I ask: are these polls weighted for turnout based on likely turnout because of a respondees demograhics e.g based on my anecdotal knowledge older voters or tory supporters are mpre reliable voters, or do those surveyed indicate how likely they are to vote as part of the survey? is it possible for example that if the likely outcome is weighted on how likely a respondee of a particular age in a certain area is likely to vote in any election, might this be innccurate if, say, a huge number of younger voters turn out to vote yes for AV because we under-estimate hoe likely this group are to vote?

  15. Thanks, Anthony.

  16. Well even if they were adjusted as London doesn’t vote the % resluts will not be aporitcualry revealing of a GE result…..

    What is interesting to me is that the Conseravative vote remains near its GE values and that’s in my opinion both good and bad….good their GE base level is a strong position to be at ….12 months into a 5 year Parliament….bad it might actually represent a ceiling for Conservative vote that is lower than their previous periods in office….

    We will as ever know so much more shortly but probably also as ever, be no wiser…..

  17. What would all of tonight’s polls mean in terms of council seats? Anyone?

  18. Anthony, please delete my post on Stoke-on-Trent.

  19. YouGov have also done some polling on the local elections that’s been included in the AV poll tables……

    And thinking about the candidates and parties who are
    standing in local council elections in your area, who
    would you vote for in a local council election in your
    area on the 5th May?
    [Only those who live in areas with a local council election.
    Adjusted for likelihood to vote]

    Conservative 32
    Labour 33
    Liberal Democrats 14
    Another party or an independent 12
    Would not vote 1
    Don’t know 8

    Headline Local Council Voting Intention
    [Only those who live in areas with a local council election.
    Adjusted for likelihood to vote]

    Conservative 35
    Labour 37
    Liberal Democrats 15
    Another party or an independent 13

    And thinking about the candidates and parties who are
    standing in local council elections in your area, who
    would you vote for in a local council election in your
    area on the 5th May?
    [Only those who live in areas with a local council election.
    Not adjusted for likelihood to vote]

    Conservative 29
    Labour 31
    Liberal Democrats 12
    Another party or an independent 11
    Would not vote 6
    Don’t know 11

    Headline Local Council Voting Intention
    [Only those who live in areas with a local council election.
    NOT adjusted for likelihood to vote]

    Conservative 34
    Labour 37
    Liberal Democrats 15
    Another party or an independent 13

  20. Details of YouGov council and AV polls are now up here:

    Local Council percentages are:
    Conservative 34%
    Labour 37%
    Lib Dem 16%
    Other/Ind 14%

    NOT adjusted for likelihood to vote (which would help Tories most)

  21. Why am I constantly bombarded with NOtoAV ads when I check this site for poll figures? On this page, for example, there are three huge adverts exhorting me to vote NO which between them take up at least twice as much page space as the factual report about the YouGov poll.

    If “UKPollingReport is a site for non-partisan discussion of polls” then surely it should also be free of advertising for any particular side of the referendum?

    Shame on you!

    [Pass on your name and address and I’ll forward all the bills for hosting the site to you to pay. In the meantime, feel free to bugger off – AW]

  22. @Leetey

    That has to measured against the overall share of the vote in the same seats in 2007, when Lab and LD were on around 24% and Tories around 38% (someone will have the precise data).

    On the top set of figures thay would be:

    Con -6
    Lab +9
    LD -10

    Or a swing from Con to Lab of 7.5% (or more likely of LD to Lab of 8.5%.

  23. These would pretty poor figures for Labour, if indeed they are true and verify. Cameron would navigate his party through the local elections without the disaster many, myself included, were predicting.

  24. Sorry Leetays’s figures are correct (d*mn small YouGov figures)

  25. RAF

    That has to measured against the overall share of the vote in the same seats in 2007,”

    I think he’d forgotten that- to make an accurate rather than a party political point- you have to compare apples with apples :D

  26. Presumably the other polls are more favourable to the Tories?

    I reckon tomorrow might bring a few nice surprises for Labour though.

  27. Yup, it’s true that compared to an absolutely drubbing that virtually hollowed out Labour’s presence in local government, these figures are a big improvement.

    Context is everything….

  28. Utilising FPTP and assuming that the coalition holds (therefore we are in a two party system): Labour on 40 and Tories on 36 makes for a comfortable win under a 650 seat electorate (a 38 seat majority via best model we have available on electoral calculus) and a modest win- certainly largest party status- under a 600 seat HoC.

    I’ve said before that IMHO we won’t have the massive swings in VI in polls that we have had over the last 25 years (and had had by now in previous electorla cycles).

    Labour will be 2- 8% ahead on average (with a few outliers and rogues at either end of the spectrum from time to time) all the way through to the next election. Not even *when the cuts actually start to bite* which- it would be wise of some posters to remember- they have not yet, will Labour chalk up huge leads

    We won’t see any 15- 20 point lead like Kinnock for brief periods prior to 1992 had (and what seems to be what Tories on here grasp onto all the time) !

    Unless the Lib Dems pull the plug- now that would be a poser for Ed Miliband…..

  29. Predict vote shares in local council elections tomorrow will be about equal for Cons/Labour (which will still be about a 6.5-7% swing from 2007 – BUT remember there are also seats being contested that were fought in 2008-2010. Councils such as the one in my area are changing to putting up all wards simultaneously for the first time).

    Probably both about 36%. Not exactly a honeymoon for Cameron (except as compared to Libs) but hardly an endorsement for Labour and the Eds who like to imagine sometimes that most of the country agrees with them that the cuts are “too far, too fast”.

  30. @ Rob Sheffield,

    Problem is that I don’t think it will be quite as comfortable when you factor in the boundary changes with the reduction in MPs as well.

  31. @Neil A
    …and losing nearly 100 seats in last year’s GE. It may well be a low base, but they were at a very low ebb even this time last year (council elections in London last year excepted).

    And as Rob Sheffield says, we must compare apples with apples…

    In truth, it just goes to show there isn’t much enthusiasm for anyone really.

  32. Labour would be happy because they would make pretty significant gains.

    The Tories would be happy because the local elections would not be the disaster many said it would be for them.

  33. @Timothy
    Wow. AR and YG practically identical.

  34. TGB you have done it again YG 3rd May 5% not 4%.

  35. @john Murphy – “What is interesting to me is that the Conseravative vote remains near its GE values and that’s in my opinion both good and bad….good their GE base level is a strong position to be at ….12 months into a 5 year Parliament….bad it might actually represent a ceiling for Conservative vote that is lower than their previous periods in office….”

    I think that is a very good way to express it. There may well be something in @Eoin’s idea of the LALA score, with a stable/stagnant Tory vote and the Labour lead oscillating depending on where the Lib Dem level is.

    This would suggest that there could be an anti Tory majority (of course there is if the Tories are on 37%, but I hope people get what I mean) and if Clegg goes into Tory attack mode, as he has done for the last week or so, some of those voters move from Labour to Lib Dem. This would fit with the fall back in the Labour vote, although we need to be a bit careful as this isn’t shown across the board – YG had Labour nicely on 42% yesterday.

    Looking at AW’s graphs shows a very striking pattern across all the posters he lists, with a remarkable long term improvement in Labour poll numbers stretching back to April/May 2009. This was only interrupted by the Clegg TV debate surge, but in retrospect now shows up as a sustained long term uplift over a very long time span. At some point this has to stop, and we might be near that point now.

    I’m still of the opinion that the Tories haven’t done enough to turn generally favourable voters into cast iron supporters. People are still happy to give them a go, but they don’t love them. I suspect the failure to gather sufficient high octane support at the last election is a sign that long term they are in some difficulties. The last time they held any kind of majority they had 11 seats in Scotland, and although we should never say never, it looks like Scotland is lost forever to them. They failed to make any significant inroads into northern England and the Wales 2010 performance looks like its unwinding.

    I think there are currently too many areas where the Tories are excluded from for them to get a GE majority,especially with the likely geographic impacts of the cuts, with the result largely being defined by the lab/Lib Dem balance. Tomorrow will be an interesting set of results.

  36. Raf

    For technical reasons, it’s quite difficult to assess what the exact percentages for the main Parties are even in the areas that did have elections, never mind extending it to the whole of Britain. Wikipedia gives the 2007 figures as Con 39%; Lab 27%; Lib Dem 26%, though I’m not sure if that’s the estimate for the areas fought or the whole of the country.

    However the four year gap between ‘equivalent’ elections will be complicated by boundary changes, local government reforms, changes to electoral cycles etc.

    In particular this year, since 2007 five counties have become unitary authorities (in whole or part) and so the district councils that comprised them will not be holding elections as they no longer exist (and county elections aren’t this year). So by my reckoning, instead of 2007’s 10,479 seats up for election we will only see 9,390 (down 1,089). I reckon Northumberland has ‘lost’ 239 councillors; Cornwall 249 and Co. Durham 312.

    It’s worth pointing out that this reduction has proportionately hit the Conservatives lightest. They have dropped from 5315 seats won in 2007 to 5026 defended now (-289 = -5%) and opposed to the Lib Dems 2171 to 1866 (-305 = -14%); Labour 1877 to 1620 (-257 = -14%); and Others 1116 to 878 (-238 = -21%) presumably lots of Cornish independents.

    Phil’s already made the point very well about how looking at the number of councillors (or councils) always looks good for the Conservatives, at least in seat terms because safe rural Tory wards are usually much smaller than their urban Labour equivalents and so need many fewer votes to elect a councillor.

    In fact this year of the electoral cycle is actually the most Tory-friendly because Wales, Scotland and London are not voting for local elections and some Labour-friendly counties have also gone out of the equation.

  37. The last 4 years have told me that there isn’t many in the way of ‘core supporters’ in modern politics. Most change their mind according to how the government is perceived ti be performing, how the economy is going etc.

    I think either Labour or the Tories could quite conceivably be scoring in the 40s come 2015 – I see little evidence or logical argument to suggest otherwise.

    If anything, these latest polls to me suggest that most people do not like the government i.e, Tories or LIbs much at all, but they have an equal disregard for the opposition i.e. Labour. Hardly a ringing endorsement of modern British politics, is it?

  38. @Roger Mexico

    Thanks for that. So what are your seat projections?

  39. Lowest YG Labour figure since early December if I’m right(?), but I noticed on tonight’s tables – and something I’ve noticed recently – the Don’t Knows/Would Not Vote combined figure has dropped a fair bit from where it was (tonight’s 17%, from what I remember it used to be anywhere between 20%-25%); so could it be disillusioned voters from their respective coalition parties returning, and thus diluting the Labour figure?

  40. @ Rob Sheffield

    It pains me to have to complement such an anti-socialist, but your analysis on why there’s lack of leads etc is consistently spot on.

  41. If we just take Con/Lab/LD councillors up for reelection (from the ALDC website):

    Mets…………… Con 184, Lab 392, LD 177
    Unitaries……..Con 912, Lab 484, LD 374
    Districts…….Con 3,943, Lab 722, LD 1,285

    Total: Con 5,039/ Lab 1,598/ LD 1,836/ Grn 59/ BNP 12/ Ind 306/ Other 554.

    Total 9,404.

    An increase in Ukip candidates this time round could cause one or two headaches perhaps, but other than Labour gaining control of some significant powerbases in local government, Phil’s and Roger Mexico’s qualificatory remarks should be borne in mind when the results come in.

  42. Raf

    That did make me laugh out loud. And after the unexpected results from even predicted percentages in this week’s Canadian elections, I suspect that no one will ever be tempted to a seat projection again.

    Actually I suspect that the raw percentages may even see the Conservatives very slightly ahead of Labour. As I said it’s their best year out of the four year cycle. YouGov’s are the only ones we have to go on and the weighting looks quite heavy in boosting up Labour – often a sign of some uncertainty (and of course political weightings are based on national party identification).

    I would be surprised to see Labour win much over a thousand seats – in the 700 -800 range would be more likely. The Conservatives may do better than expected by balancing losses to Labour with gains from Lib Dems – even where the Tory vote actually falls. UKIP at 4% may nibble a bit at the Conservatives vote, but they probably don’t have enough candidates to make a significant effect – those elected will probably be those who would have been elected as independents anyway.

  43. Though not exactly the election that never was… a snap May GE could possibly come to be seen as a missed opportunity.

    Today’s mention of the police officers being invited to reapply for their jobs on a voluntary/expenses only basis has been the first item of domestic news for a fortnight. The RW (possibly the Olympics) are coming too early on in the electoral cycle for the Conservatives, and calling a “pause” or backtracking on major reforms will not be a realistic option for much longer.

  44. It’s worth comparing the YouGov GE poll directly with the YouGov council, poll, as the two are taken from exactly the same sample of 5725 voters, despite appearing in different tables on the YouGov site.

    The three figures that follow for each party are:
    – GE voting %
    – Council voting % (not turnout adjusted)
    – Council voting % (turnout adjusted)

    Lab 40 37 37
    Con 36 34 35
    LD 11 15 15
    Oth 14 13 13

    Looking at the council figures from a red perspective, I see nothing that gives cause for concern when compared with the YouGov GE figures of 40/36/11 because:
    – London, Wales and Scotland aren’t included in the Council figures and including them would surely widen the Lab lead over the Cons
    – While it’s clear that the LDs will do better in the local elections than their GE poll ratings would suggest, it was ever thus. In fact YouGov is confirming that you need to take 4% off whatever national equivalent vote share the LDs do achieve tomorrow before predicting their equivalent score in a GE. And I would expect a fair chunk of that 4% to go to Lab, given what we know about the past allegiances of LD defectors in GE polls.
    – While the turnout adjustment reduces the Lab lead over the Cons by 1%, a narrowing due to turnout of some degree was again entirely predictable and I am if anything relieved to see it at only 1%.

  45. Aaargh – The LOL comment was to Raf inviting me to make a prediction after I warned against them (and then like a fool made one anyway).

    With regard to the Daily Mail piece he pointed to, it seems another example of the strange way Cameron conducts foreign affairs, but more of that another time.


    The Non-Voters figure is usually pretty constant at around 20%, though it varies a lot depending on how people voted last May. About 25-30% of 2010 Lib Dems were NVs but only around 12% of Tories and 8% of Labour. Obviously the percentage of those who didn’t vote then and are NVs now is much higher.

    There has seemed to be a few points drop in the percentage of NVs recently, in particular due to the Lib Dem percentage going down. I suspect some people, forced to make a choice for local and other elections, have become a bit less coy. It’s still not a massive change though and not all in one direction.

  46. @Billy Bob
    Thanks for those figures – I’d been searching for them for ages and didn’t realise that they had been published by the ALDC. (I take it it’s correct to assume that they are in fact the political allegiances of councillors in seats that are up for election, whether the same councillor him/herself is standing or standing down.) In the absence of them I had to derive some rough estimates for the unitaries and shire districts. It looks like these estimates were about 200 out in total but that’s not enough to alter the warnings stated earlier about seat counts.

  47. Phil

    great work and very persuasive.

  48. Rob – thanks.

1 2 3 5