There is also a Harris poll for the Daily Mail out tonight – the topline figures there are CON 33%(+1), LAB 24%(-1), LDEM 32%(+2). No vast change there, though Harris are showing a much lower level of support for Labour than some other pollsters, and the highest Lib Dem score we’ve seen from any company for about a week and a half.

I’m not aware of any other polls tonight – there have been rumours of a MORI poll, but as far as I can tell this comes from a comment by Kevin McGuire talking about an old poll that has been misconstrued.

771 Responses to “Harris/Mail – 33/24/32”

1 9 10 11 12 13 16
  1. As a libdem I’m wondering whether I’d rather have:

    1) No overall majority so we can get rid of the ancient FPTP system and replace it with something more proportional.

    2) A small Tory majority, due to a collapse in the Labour vote, putting a significant number of points between Labours third place and the Libdems second. Let the Tories sort out Labour’s mess, and take the hit that comes with it… and give the Libdems a chance to mature a bit more and lay the ground work for the following election establishing the centrist party as the real home of modern progressive politics and leaving Labour to the old, hard left.

    Let’s face it, M King wasn’t wrong when he said (whether he said and when…) that the next government is getting a poisoned chalice.

  2. @ Pompey Observer

    That looks like a more realistic prediction

  3. Sue Marsh
    Re BBC
    There is more subliminal messaging by the BBC News website with a picture of the 3 leaders.
    Full side faced pictures of Cameron and Clegg looking directly at each other, through Brown who appears in the background.

    David Cameron ‘giving it everything’ -positive
    Illegal immigrant worked on Hazel Blears’ campaign – negative

    Video clips:-
    Cameron offers butcher ‘barbecue summers’ – Positive
    Paxman v party leaders – Negative (a strange title that entices you to find out how all the party leaders have done against Paxman, only to find out that the clip is exclusively about Paxman grilling Brown over ‘Flockingate’.

    They have been doing this sort of stuff for the whole campaign.

  4. @ SUE

    I hope you have a lovely time – I know you will say a ‘prayer’ for AL J while you are there.

    See you back here later :-)

  5. As I explained yesterday I think Clegg’s performance in the debate will cause a drop off in support.

    I was wrong if last night’s two polls are correct, but I’m being consistent!

    I should add that this is not the result I would want so it is not a partisan projection

  6. @GreenG,

    Yes- the BNP has been having quite a good campaign.

  7. I don’t see the Labour share of the vote shoring up. If history is anything to go by, their share of the vote is overestimated by 1-2% by most pollsters (perhaps not the newer ones who give them bottom %ages).

    The tories they got pretty much spot on last time, with around 1% underestimated. That’s why I think 35-36% for them.

    The libdems will get around 30% – the average of polls was spot on last time, I think they may underestimate them slightly this time due to a number of ‘new’ voters and traditionally low turnout voters taking their chance this time. Not a bit effect, but worth a 1% on the average off polls.

  8. I think the Lib Dem vote will fall by a few % in the next week as well. I thought NC wasn’t anywhere near as convincing as the first debate. Just my opinion, of course.

  9. Other than YouGov and Comres it is quite frightneing how little polling eivdence we have had.

    either a) the feildwork is always rather outdated (eg Harris/TNS)

    or b) The big comapnies like Ipsos and ICM do not produce polls very often…

    It makes it very difficult to compare like for like polls.

    Saturday night is the one night every week where we get that chance.

  10. @Eoin (and all)

    A generally reliable source has dropped a clue that suggests (but does not 100% confirm) our guess of 35/28/28 for ICM is correct.

  11. @Skudor,

    History has shown mixed results on estimating the reds share… in 1983 polls showed a 20% gap between blue/red with a week to go.

    In the final week the gap shortened by 5% in favour of red. :)

  12. Interestingly, More 4 isn’t showing The Daily Show from last night on its on demand web site.

    During the show, which was almost entirely about the UK election, Sir Michael Caine said that Nick Clegg wiped the floor with Cameron (and Brown) during the leaders’ debates.

    The Tories surely would want to keep that quiet.

    It’s a brilliant show and worth watching. You can download it on bit torrent, or you can watch it via the Daily Show’s on web site on firefox, provided you install an add-on first (more on how to do that here h t t p:// )

  13. I suspect that tonight’s polls will give us pretty much what the outcome will be on Thursday. The debates focussed the election coverage earlier in the campaign, and I think opinion has hardened since then. The undecideds will split largely according to the rest of the population.

    I do not think the past is a very good guide this time round.

  14. @EoinC
    “Yes- the BNP has been having quite a good campaign.”

    Not influenced by my prediction of 5 Green gains then? :-)

  15. @skudor

    I certainly think they will be a good indication

    However I think you have to factor in that if the polls tonight show a clear lead for CON (which is quite likely) then this in itself may lead to a stronger focus on tactical voting by the pre-LAB/LIB element in the media – “keep the tories out” may increasingly be the message. Some LAB – LIB shifters particularly in the north may shift back in particular.

  16. @ Matt

    I don’t actually think performance in the debates has much effect on the polls. The LD surge was shown in polls which were conducted before the first debate, but they were reported as having shown the impact of the debate because they were released just afterwards. This then lead to the media massively hyping up the impact of the debate and as a result they gave Clegg a massive amount of extra exposure. So the combined exposure of the debate and the media reporting of it probably added a few points to the LD surge, which was already underway.

    I don’t think the performance of the leaders in following debates or in debates in future elections will have any significant impact (unless someone makes a big mistake). Not enough people watch them, and in future I think the viewing figures will be much lower. So basically I think the first debate was a one off.

  17. @ South Londoner

    “ComRes do not weight according to imagined viewership but according to what they call voting population profile. The sample watched on Thursday lined up as follows.
    Con 35, Labour 24, LibDem 36.
    Now look at how they voted in terms of who ‘won’ the debate. Cameron 35 Brown 26 Clegg 33.
    Get the picture? Indeed, based on that you could argue that GB ‘won’ (as indeed in my view he did) as he was the only one to score above his party rating”

    Not at all true, because starting off with an obviously disproportionate sample, ComRes weighted the poll results accordingly. In other words they already made the adjustment you are trying to make.. This is something which YouGov did not do, and why their poll appears to be so out of line.

  18. @Scotty Dog,

    Don’t agree with that. Tactical voting will be far far smaller this time round, as the Liberals are much stronger, so the ‘wasted vote’ syndrome won’t apply anymore.

    Also, you could argue that people like to pick winners, so leads in themselves lead to momentum and more votes.

    All in all, the result tonight will be withing 1% of Thursdays result. No further change now, 99% of people are enjoying the bank holiday weekend, not thinking about the election.

  19. I’m interested in this often – on here – quoted statistic that labour’s rating in the polls is “always” over-estimated.

    How accurrate a statement is this, and what does “always” mean?

    Surely there is a point before which opinion polls were far less accurate, numerous, or useful, so I’m assuming “always” means in recent elections?

    But surely if this is true its only useful information if its known why it has previously been overestimated. The implication that its a built in folly in the polls or trait in the voters doesn’t seem to be elaborated on much. What we DO know surely is that in the last 3 general elections the outcome was generally known and assumed beforehand and in labour’s favour. If that itself is the cause of fewer labour voters on the day then its hard to see how its relevant to this election.

    Wouldn’t it be just as correct to state that the party who look like winning are always over-estimated in the polls?

  20. @Rob A,

    The LibDem surge partially showed up in a single poll that mostly sampled before the debate.

    I think there is some truth that the boost wasn’t just the debate (the manifesto launch and the build-up to the debate may have given Clegg enough oxygen to start appealing to people) but there is no doubt in my mind that the debate itself was the thing that pushed them above 30%.

  21. There are so many people clutching at straws today, I am giving this a break until at least tonight!

    If ICM shows Tories on 35% against 28/28, I will be very happy to go in to the election with this.

    Beats the dark days of the last 12 years by a long way!

    Fair point – extreme movements in any direction, implyling a change to the game could have that effect. But if the polls confirm the basic narrative of a hung parliament I’d expect the status quo to continue.

    I’m not expecting to see a major improvement for the tories though, I think we’ve already seen the return to the fold of those that will return (of those who switched from Con to Lib following the debate) following the first debate surge. I didn’t see anything in the debate that would change this.

  23. @Richard O,

    “If ICM shows Tories on 35% against 28/28, I will be very happy to go in to the election with this. ”

    Me too. It would show that a Tory majority would still be possible going into the last week of the campaign, especially if other polls today show something similar.

  24. @richard o

    I agree there wil be less tactical voting but I think there would be some if it was clear the CONs were headign for an overall majority. In LAB/CON marginals in the north do you not think some voters who have shifted from LAB – LIB will not reconsider as an expediency?

    This depends on neither LAB or LIB being above about 28 in the polls.

    If they are 35 27 30 or similar then I would agree no tactical voting

    37 28 27 and I think there would be an awareness that in LAB/CON marginals in particular tactical voting could play a significant part in the outcome

  25. Average of tonight’s polls will be 33 – 30 – 25

  26. @ Neil

    You may be right. But I still think that first debate was a one off and we will never see any future debate have an impact of more than 1 or 2 points in the polls. Unless someone has a shocker of course.

  27. I would be surprised if the Tories didn’t get a majority – they always get their vote out

  28. @Gattino,

    I think “inevitability” is definitely a factor in the turnout, especially for Labour voters (unlike “rain or shine” Tories). That could mean a more solid turnout this time, but only if Labour can stave off the impression of “inevitable defeat” which I would imagine could depress their turnout just as much as “inevitable victory”. In that sense this weekend’s polls are critical – if Labour is in the mid 20s a lot of support may decide it’s all over and lose interest, a bit like football supporters trudging to the train station 10 minutes before full-time when their team is 3-0 down.

    There is no doubt that more often than not the Labour vote has been lower on the day than it had been expected to be based on the polls in the days before the vote (as opposed to weeks/months before). I expect that it is purely the “spiral of silence” / Shy Tory phenomenon at play. I don’t actually think that this is always to do with the overall situation in the polls. I think there is a general perception in the population that the Left is “well meaning” and that the Right is not. Perhaps some people go to the polls with the intention of “doing the right thing” and then decide, in the privacy of the booth, to support the “nasty party” anyway.

  29. Yellow tactical voting is on e the most overhyped aspects of this election.

    Red tactical voting may happen and contribute to yellows holding seats from blue and perhaps taking three (Oxford, Watford, BNW). Ironically, soem of these will be at a red loss but they were heading that way anyway.

    I see very very little scope for yellow vote saving reds.

  30. Electoral Calculus( Hope not Voodo) up to 30th April.

    Con 35.81 305 seats
    Lab 26.14 220seats
    Lib 27.69 93 seats

    others 32

  31. Just anectodal evidence:

    talked to my mother today – a life-long Labour supporter apart from a short Plaid daliance – in the first debate she was really impressed by Clegg, but less so in the second debate and turned off by him in the third. Will now vote labour again.

    Vale of Glamorgan marginal

  32. twitter rumour: Sun Tel/ICM national poll shows significant moves for 2 of the 3 main

    Labour down 2 Lib Dems up 2

    33/33/26 ICM = significant movement for two with nc tories.

  33. @Rob A,

    Yes I think the first debate of 2010 will go down in psephelogical history as unique. I don’t necessarily think it couldn’t have pale imitations in the future though. I could imagine a situation (although I hope never to see it) where a far-right party with a very eloquent leader managed to get themselves a berth on a TV debate and make massive headway amongst the “I’m not racist, but…” section of the electorate. PR of course makes that much more likely.

  34. @GATTINO

    All I have are the 2005 results – out of five polling companies, only one got Labour spot on (NOP), the other four over-estimated by 1-2% (ICM, MORI, YG, Populus).

    They were mostly accurate on the tories, 3/5 got their share right and the other two underestimated by 1%.

    On average they got the Libdems spot on, but with variances between the orgs ranging from underestimated by nearly 3% to overestimating by nearly 2% – but the average of the five was spot on.

    That doesn’t mean it’s the way things will pan out this time, but it’s the only guide we’ve got. All the differences were within the MOE, so not necessarily due to methodoligy – just normal sampling error.

  35. what would a 33 33 , labour 26 leave us with

    of course, accepting variations ?

  36. I find the Guardian decision a strange one,even more so the timing.

    Labour voters who were thinking of voting tactically will now have second thoughts,the survival of their party in second place may cause open warfare between the two centre-left parties.

    I am bemused by the labour support over this campaign for the LIb-Dems,turkeys voting for Christmas springs to mind.

    Vince Cable quote’i don’t agree with industrial action’ (2009) how would this go down with the Labour Party? even more so the unions!

    Labour deliberately under instructions from Peter Mandelson got in bed with ‘Nick’ until he didn’t want to be the junior member of the partnership & now Clegg is the Liberal ‘toff’.

    Labour have been played like a harp and it is all self inflicted.

    The Tories would much rather have a main opposition that is pro EU,pro Euro and pro immigrant amnesty than dealing with the Labour Party in official opposition.

    DC is looking so happy today and i see why.

    I stick to my prediction.

    Con 40-43%
    Labour 25-29%
    Libs 23-25%

    I wonder if the Labour support would settle now for the old 2 party tussle,with the tories winning in 2010 and them getting rid of Brown/Mandy.Millibands,as some see it getting their party back and challenging under a new leader in 2014-2015.

    I think the effects of the LIb-Dems has begun to dawn on the labour support.

  37. “what would a 33 33 , labour 26 leave us with

    of course, accepting variations ?”

    Not good for Labour, if true. I’m still sceptical though about rumours.

  38. @FrankG,

    I’m not sure that you understood what I am saying. I don’t know how you translate the regional swing on individual seats, but if you just deduct 12% from labour and add it to the Lib Dem % from 2005 I am sure you are doing it wrong.

    My method says that you should use the swing whether regional or national to work out the proportion of labour voters who are switching, and then apply that on a seat by seat basis. So if labour nationally start on 36% and are now on 30%, you would move 1/6 of their voters (17%) elsewhere in each seat. In a seat where they start at 50% this means that you have to decrease their % in that seat by 8.3%, not 6%, and you add 8.3% to the Lib dems if that is where the swing is going. A swing of 6% nationally leads to a swing of 8.3% in that particular seat without any tactical voting or marginal boost. Similarly, in seats where Labour have <36% in 2005, the local swing will be less than 6%. It follows that the regional swings in part reflect the starting point of each party in that region, and hence it will be a mistake to use the regional SE swing which is based on a 2005 Labour vote of only 24%, in an individual seat like Watford where Labour start on 34%, or Bedford where they start on 42% and which I now identify as a 3-way marginal and a Lib Dem gain with a (quite big) following wind in the polls!

  39. @fingerbob69

    That would fit the tweet, but a reliable source has stated the ICM gives the Tories a 7pt lead over 2nd place. That fits 35/28/28 and 37/24/30 – both of which fit the tweet.

  40. @eoin

    my theory would be that it would be red tactical voters – having changed from red to yellow then back again when they realise their actions may cede a seat to blue

    It only works if it is clear yellow isn’t going to make significant gains – ie if they are back in the 23-27 area – and only in those seats which are LAB/CON marginals where it is clear LIB cannot be the main competition to CON

  41. “what would a 33 33 , labour 26 leave us with ”

    apparently it would be 254 / 120 / 243

    Still, we’ll see if it’s true, eh?

  42. Eoin

    I think your faith in the Labour vote is way too optimistic .

    C 35, LD 30, L 26, Oth 9

    second that.

  43. LibDem preference in the polls has been almost uniformly suppressed because of the softer likelihood of those indicating a LibDem preference to actually get out and vote. In comparison Con support has been uniformly firm. This has made Con %vote the most stable, and the LibDem %vote the least stable, with Lab somewhere in between. Different methodologies applied by different polling companies highlight this volatility.

    But it also means that as we head towards the booths on Thursday, there is far more potential for higher LibDem polling should their support start firming up.

    Of course, whether it does firm up, shears off or falls way is another question entirely. But I think that potential is injecting a great deal of uncertainty into the eventual outcome.

    Not that the percentage vote really matters. With a large proportion of the voters effectively disenfranchised, and many voting negatively, the percentage vote is highly distorted anyway. The very existence of tactical voting in an election is an abomination, but it does make for interesting elections.

  44. On the UKpollingreport swingometer, 33-33-26 would make the Tories the largest party in a hung parliament-

    Con- 254

    Lab- 243

    Lib- 120

    Oth- 32

  45. @ wanderingwelshman

    what would a 33 33 , labour 26 leave us with

    of course, accepting variations ?


    BBC seat calculator:

    Con 255

    Lab 235

    LD 131

  46. @Scotty Dog,

    Given that the Tories dropped from 39-40% to 32-33% when the LibDems surged, I would imagine there is quite a lot of scope for the opposite happening. Someone who has for months/years been looking forward to evicting Labour by voting Tory, and then switches to “nice new Mr Clegg” as the best hope of doing the job, might well swing back to Tory if his seat is a Labour-Tory marginal.

  47. @FingerBob, Dotski, George Gardener

    There are many insults which can be levied at a man, optimism is not one of them. :) Well not if your left leaning :P

    Good luck with your forecasts!! :) :) :)

  48. There should be a rule against posting unsubstantiated rumours on here. Don’t just post “a reliable source says this”. Unless you have actual evidence to back up your claim don’t bother. Most of the time these twitter rumours turn out to be untrue anyway.

  49. WMA 34:27:29.
    The 14-day trend in the CLead is now (just) signiifcant (R2=0.51) and increasing at 1% per week. However the 7-day trend is not. Amazing that there is no sign in the WMA of a Labour vote collapse. Their WMA is almost exactly where it was before Bigotgate and the last Debate.

    From a satistical PoV the only thing that has changed substantially in the campaign is the LD surge which began on/around the 15th and seems to have peaked on/around the 20th.

    I still think there will be a CLead over L of 10-15 points on the night, but what the CLead over LD will be is much less clear. 5-20 would seem a reasonable range.

    FWIW my guess is that the results on the night will be +3:-2:-1 or thereabounts from the final opinion polls. Tories will be working very hard to GOTV, L will be demoralised and LD just don’t have the organisation on the ground.

    But we shall see…

  50. @Rob A,

    Chill………… :)

1 9 10 11 12 13 16