It’s Saturday night, so I’d expect lots of polls for the Sunday newspapers. The first out of the traps are ComRes, ICM and MORI.

ComRes for the Sunday Mirror and Independent on Sunday has topline figures of CON 34%(-1), LAB 28%(+3), LDEM 29%(+2). This is an increase for Labour, but may well be a reversion to the mean after some rather odd ComRes polls in the week. The previous ComRes polls were their rolling polls for ITV news and, as we discussed at the time, they appeared to have included an extremely Conservative sample from Monday, which produced 9 and 8 point Tory leads that looked rather anomolous at the time.

The second new poll is for ICM for the Sunday Telegraph, and has figures of CON 35%(+2), LAB 26%(-2), LDEM 31%(+1) – so in contrast they have the Conservatives rising and Labour falling. As with ComRes, the Lib Dem boost remains healthy.

Finally there is an Ipsos MORI poll in the News of the World, which has the most surprising result. Their topline figures with changes from the last poll are CON 36%(+4), LAB 30%(+2), LDEM 23%(-9), so they have the Lib Dem boost almost entirely unwinding. Ben Page of MORI has has just been on Sky – and hats off to him for giving a responsible and measured account of the poll rather than claiming it shows something spectacular. Ben said they’d checked their figures very carefully, scratched their heads, but they have to publish them… but he did re-iterated that one in twenty polls are rogues. That’s about as close as pollsters come to warning that one of their own polls they’ve just released might be a rogue!

Then again, it might be the start of a trend. We should have more polls to come later tonight (at the very least there will be YouGov in the Sunday Times) so let’s see what they say.


210 Responses to “Sunday polls – ICM, ComRes and MORI”

1 2 3 5
  1. When did Mori do that poll …April 1st?

  2. Given the wide fluctuations, I am quickly losing confidence in opinion polls.

  3. Simply put, nobody can read anything in to thress three, as they are all so different. i.e. 1 encouraging for Labour, 1 awful for Labour & good for Cons, 1 awful for Lib Dems etc etc

    I need Eoin’s view on this I think! Where are you. You are like Merlin out of John Boorman’s excalibur, never around when your required!

    Looking forward to YouGov.

  4. was that last ipsosmori poll really 32/32/32 ?

  5. Averages (of all published polls) thus far today

    Con = 34.3
    LD = 28.8
    Lab = 26.8
    Oth = 10.3

    On Electoral calculus (without Tactical Voting tool switched on) that gives us

    CON 277
    LAB 246
    LIB 96

    Combined Lab-Lib total is 341: majority figure required (as SF don’t take seats) is 322.

    This result means a Lib-Lab coalition under PM Clegg and with Labour spending first three months electing a new leader.

    It may also have some nationalists in it specifically on things like voting reform (a referendum of *options* for which type of reform NOT the ‘do we or don’t we referendum which is what the Conservatives offer); plus also on sensible staggered spending cuts.

    For YouGove tonight to be considered good for Lab they need 27 or more; good for Con means 35 or more; good for LD means 29 or more

  6. The MORI poll is probably simply reflecting the fact that novelties tend to wear off over time.

  7. So, in summary

    ComRes: CON 34%(-1), LAB 28%(+3), LDEM 29%(+2)
    ICM: CON 35%(+2), LAB 26%(-2), LDEM 31%(+1)
    IpMORI: CON 36%(+4), LAB 30%(-2), LDEM 23%(-9)

    I hate to say this, but I think this *is* the start of a trend. I’ve said before that the “LDEMs are unilateral disarmers” meme has gained traction amongs the populace (I’ve heard a couple of people advancing it as a reason why they’re not voting LDEM) and, to be blunt, it’s killing them deader than a dead thing that’s been deaded twice.

  8. Anthony – Sky’s rolling summary has MORI as 34/30(LAB)/29(LDEM)/7. What is going on ? Polling industry losing cerdibility I think.

  9. I expect Eoin will highlight the clear Tory slide, with ComRes showing their lead over Lab down from 10 to 6 points.

    :-)

  10. @ Éoin (when he appears)

    IF, IF, IF Mori is anywhere near correct, your prediction of CON up, LAB chasing them & LD a spent force is so remarkably astute that we will demand you do a victory dance!

    If this carries into future polls, you’ll be the only chap who is not eating his shorts ;-)

  11. It seems that the pin is very close to the Whig bubble and is threatening to burst it.

    All the polls suggest a trend back to Labour and the Conservatives. Although the Mori one seems to be a bit of a rouge/outliner (delete as appropiate).

    I expected the ComRes poll to show a slight drop in Tory support due to the rolling polls they conducted last week.

    All in all, it is looking very exiting and the debate on Thursday should bethe game changer.

  12. These take the WMA to 34:28:28

    The 7-day trend increase in the CLead over Labour is now decidedly significant (R2=0.64) and running at 2% per week. This would take the CLead to 10 at the GE in the polls, and probably 11-12 in the ballot boxes when they are counted.

    The trend for LD overtaking L seems to have been reversed over the last 3 days, with LD voters going 2:1 to C rather than L.

    the MORI figues for LD are clearly out, they are 5 pts out on the LD-L position but this is perfectly normal, the MoE on a lead is about 5%.

  13. @Ky;e

    “I expected the ComRes poll to show a slight drop in Tory support due to the rolling polls they conducted last week.”

    I’d wait for their monday night poll- you are comparing apples and oranges.

    I also expect Comres to show a drop in the Mon-Wed ITN polls on monday night compared to last wednesdays rouge.

  14. On UNS (I know, I know) Electoral Calculus gives: C271, L 254, LD 93 for the first poll; 290, 219, 109 for the second; and 285, 268, 65 for the third.

    I suspect Conservatives would opt for poll number 2 at C 35%, L 26%, LD 31% if they had to pick one right now! However, it’s interesting that on UNS the polls right now make little difference to Conservative total seats but big swings between Lib Dem and Labour.

  15. @Kyle,

    I don’t think the Tursday debate will have much impact.

    I think Anthony had it spot on. The first debate was a one-off game changer, the other two are likely not to have much impact. I actually expect Cameron & Brown to be very close on the next debate. Brown will score with his knowledge of fiscal & monetary policy, but Cameron will score on levels of debt and state of finances. It will be stalemate. The perhaps slight unknown is whether the middle spot will favour Brown or not, as up to now it has been a poisoned chalice, but could work in a PMs favour maybe? Just not sure.

    Rich

  16. @Martyn

    “it’s killing them deader than a dead thing that’s been deaded twice.”

    So dead that two out of three polls show an increase in their support.

    I do find it odd how people claim to see trends in polls that show no trends. Face it, the polls aren’t offering anything consistent and any predictions are more hope than objective analysis.

  17. Here is a repost from the previous thread but it is more relevant here. There have been 4 polls today but with an unusually large spread between the lowest and highest figures.

    Con 32 – 36 = a 4 point spread
    Lab 23 – 30 = a 7 point spread
    LD 23 – 32 = a 9 point spread

    This doesn’t seem right to me. If the MOE is +/- 3% then results more than 6 points apart should raise suspicions. One Poll is responsible for the lowest LAB and highest LD. MORI for the lowest LD and highest LAB. Its a bit of an odd one. Mori have all but admitted that their poll is a rogue and One Poll, so I’ve gathered, isn’t the most popular on here. Still, the average of each party’s score for the 4 polls is 34, 27, 29 – very similar to ComRes’ poll. It all averages out after all.

  18. @ ROGERH

    You are making a funny about Éoin, right?

    ‘Cos he getting bullied on the other thread for predicting Tory up, Lab chasing Tory & LibDem a spent force.

  19. “The perhaps slight unknown is whether the middle spot will favour Brown or not, as up to now it has been a poisoned chalice,”

    Someone said he won’t be given the centre position because of his eye.

  20. @ Colin
    Agreed.

    I’m curious as to what Eoin will say. The Mori poll seems to back up his personal predictions best, but how much credibility will he give it considering what the other polls say?

  21. @ ROGERH

    My understanding: GB is to the left of DC & NC – for all the debates ;-)

  22. As a Lib Dem supporter its good to see so many unbaised Tory/Labour supporters in the comments .

    To be honest that MORI poll is worrying, especially as the Torygraph and The Daily Fail will be same saying this is the signs of things to come.

    On the other hand, the other two polls show Lib Dems going up slightly and MORI feel there could something wrong with their poll.

  23. Last saturdays average was

    Con = 32.3
    Lab = 28.5
    LD = 29.3
    Oth = 9.9

    Thus far today it is

    Con = 34.3 (+2% week on week)
    Lab = 26.8 (-1.7)
    LD = 28.8 (-0.5)
    Oth = 10.3 (+.4)

    The average over the week 17th – 24th is

    Con = 33.1
    Lab = 27.7
    LD = 29.8
    Oth = 9.8

    We still have You Gov tonight:

    given today and yesterdays averages anything less than 35% for the Conservatives in tonights YG poll would suggest that debate two had a negative impact on DC…..

  24. Anthony

    1. I understand that Harris is now only incorporating responses in their polls from those who are registered to vote. Do the other pollsters now do the same?

    2. Is it possible to explain a significant part of the drop off in the Mori poll Lib Dem vote by the fact that Mori’s figures relate to those certain to vote and it could be some of the LD new supporters are beginning to have 2nd thoughts about whether they will actually vote?

  25. Opinion Polls seem to be like ink spot tests – people see in them what they want to see. I shall wait for Eoin to pass judgement – Eoin being the Vince Cable of this site. (That’s intended as a compliment BTW – though with the slightest nod to the possibility of a “one-eyed man in the land of the blind”-type scenario.) From my very much “blind” ignorance these and last night’s polls seem to be fairly good for blue, not bad for orange and fairly bleak for red. Am I missing something?

  26. rogerh
    “The perhaps slight unknown is whether the middle spot will favour Brown or not, as up to now it has been a poisoned chalice,”

    Someone said he won’t be given the centre position because of his eye.
    ******************************************************

    arr ok. I didn’t realise. I assumed they would rotate again and he would finish in the centre. You could imagine a centre spot favouring a strong incumbent PM, say Thatcher or Blair in their prime.

    good point though.

  27. You are right Amber. This is because GB is blind in his left eye.

  28. @ CON 36%(+4), LAB 30%(-2), LDEM 23%(-9)
    That’s about as close as pollsters come to warning that one of their own polls they’ve just released might be a rogue!

    It should also be pointed out that Ben Page is the one and the same man who has gone on record to “run through the streets naked” if there is an LD majority on election day, so he should declare a vested interest in keeping the LDs on a low polling score!

    Or should that be a non-vested interest? :-)

  29. The chicken entrails are all over the place. We still have no idea what’s happening yet. The only clear indicator is that the LibDem surge has not continued. Very few people really expected their post-debate boost to lead to them surging into a commanding lead (although one or two expressed that hope), and we can be sure now that at least isn’t going to happen.

  30. @ Anthony,

    Please, please tell us – is it 10pm for YG tonight?

  31. was the last ipsos mori poll 32 32 32 then?

  32. Might be the consituency I am in, but what I gotten on the streets and doorstep is that the Lib Dems are still pretty srong. But then again, I am in a very odd consituency, which despite being a marginal, wont decide the next government in either way. Not unless there is a coalition.

  33. Labour have got to be eyeing 30% as their minimum target at the moment – especially if the Tories are struggling to get over 36%. That would probably be enough to leave them the largest party.

    It will be interesting to see on election night if the Lib Dem surge is ‘wasted’ all over the country or whether it is concentrated where they need it to make more significant gains than the uniform swing might suggest.

  34. @Amber Star

    “‘Cos he getting bullied on the other thread for predicting Tory up, Lab chasing Tory & LibDem a spent force.”

    I wouldn’t call it bullied – just mild bemusement at some of his deductions. He did identify a supposed LibDem slide from the ‘reliable’ ComRes polling, though.

    I think the only certainty from these polls is that the LibDems are the only one of the three in a better position than they would have hoped or expected at the start of the campaign. The actual result remains an enigma.

  35. @ Rob

    “a Lib-Lab coalition under PM Clegg”

    Rob I don’t understand your conclusion – I thought Clegg had said that if there were a hung Parliament he’d prefer to chummy up to “the party with the strongest mandate” (albeit all are weak). From your projections that would be Conservatives with more votes and more seats wouldn’t it ? So a Con/Lib coalition might be the answer ?

  36. There is a lot of static in the polls. That said the poll of polls probably isn’t far off and repesents a sort of consolidation of each sides support.

    It really is down to the circumstances in each of the 200 or so seats that can change hands and i am really unsure how the poll of polls conducted in the last 36 hours reflect.

    We were as good as told by Mori in live comment to\their own poll some seat numbers that reflected a Lib Dem that reflected (154?) at over 30%.>

  37. The Ipsos MORI figure should show Labour +2, as the previous figures were 32/28/32.

  38. “CON 36%(+4), LAB 30%(-2)”

    Am I missing something? Doesn’t that mean their previous poll must have said labour and tory were neck and neck on 32?? I don;t remember that.

    It would have made news, surely.

  39. @ Rob

    “anything less than 35% for the Conservatives in tonights YG poll would suggest that debate two had a negative impact on DC”

    Sorry Rob, another conclusion I can’t quite follow. Wouldn’t we need to be looking at polls conducted after the debate to conclude what effect the debate had on Cameron’s performance ? So far he’s polled 34, 34, 35, 36….

  40. @ Mikep
    “I thought Clegg had said that if there were a hung Parliament he’d prefer to chummy up to “the party with the strongest mandate” (albeit all are weak). From your projections that would be Conservatives with more votes and more seats wouldn’t it ? So a Con/Lib coalition might be the answer ?”

    Certainly a possibility. I do wonder though what would happen if LD’s managed (somehow) to get the most votes? Would they consider this themselves getting the strongest mandate and therefore ally with whomever was most willing to let them have their way?

    The LD surge has made things so much more dynamic, I hope it remains fairly significant.

  41. Keith

    Last Isos/MORI (F/w 18-19 April)

    CON 32 (-3); LAB 28 (-2); LIB DEM 32 (+11)

    h tt p://www.ipsos-mori.com/researchpublications/researcharchive/poll.aspx?oItemId=2594

  42. @ ROGER H

    I didn’t mean to accuse you of anything – but there was a ganging up against É which, taken all together, was unusually aggressive for this forum.

  43. The raw data is now up for MORI (just go to their site) and it is clear that (looking at past vote) LDs are underrepresented in the sample. I’m not enough of a statistician to tell you what a “past vote weighted” MORI would look like, but there is an explanation for the very low LD figure.

  44. @All

    Can anyone tell me how the Lab vote in Scotland is holding up %-wise compared with GE2005 of 39.5%

  45. Sky news showing old mori as

    lab 30
    con 34
    lib 29

    so why do the new figures say CON 36%(+4), LAB 30%(-2), LDEM 23%(-9)

    ?

  46. There is an error in this page. The Ipsos MORI site confirms that today’s poll actually had Labour +2 NOT -2. The last Mori Poll was not 32,32,32 That would stick in the mind. It was 32, 28,32, see 19th April, posted 21st April, page2.

  47. I would take MORI’s 23 with a pinch of salt. It is too much too soon. I said the British electorate are fickle but they are not that fickle.

    Best stick to ComRes and YG…

    28/9% seems more accurate.

  48. @ Amber
    I didn’t really see that. A bunch of people happened to be disagreeing with his interpretations. The accusations of bias against Eoin towards one party or another are amusing, though personally I’m starting to suspect he’s a little biased towards his predictions when he looks at polls now ;)

  49. @ rob sheffield

    Last saturdays average was

    Con = 32.3
    Lab = 28.5
    LD = 29.3
    Oth = 9.9

    Thus far today it is

    Con = 34.3 (+2% week on week)
    Lab = 26.8 (-1.7)
    LD = 28.8 (-0.5)
    Oth = 10.3 (+.4)

    The average over the week 17th – 24th is

    Con = 33.1
    Lab = 27.7
    LD = 29.8
    Oth = 9.8

    We still have You Gov tonight:

    given today and yesterdays averages anything less than 35% for the Conservatives in tonights YG poll would suggest that debate two had a negative impact on DC…..

    Rob this is interesting stuff – unlike most of the rest of the partisan rubbish in this blog.

    Q for you why do you think under 35 from you gov will show -ve outcome from thursday for DC surely todays average is both up on week and on last weekend already so a 33 or 34 would keep it that way – what am i missing?

  50. “Keith
    Last Isos/MORI (F/w 18-19 April)
    CON 32 (-3); LAB 28 (-2); LIB DEM 32 (+11)”

    So Lab went up 2 not down 2 as written above

1 2 3 5