In what is presumably their penultimate general election poll (their final call poll is normally in the Standard on election day itself) Ipsos MORI have topline figures of CON 35%, LAB 30%, LDEM 8%, UKIP 10%, GRN 8%. It’s quite a shift from their previous poll, which had a two point Labour lead, so usual caveats apply. Full details and tables are here.

Panelbase meanwhile have new figures of CON 32%, LAB 34%, LDEM 8%, UKIP 17%, GRN 4%. I haven’t seen any tables yet, but I’ll update when available. UPDATE: Tabs are here

Still to come tonight we have the daily YouGov poll and a snap ICM/Guardian reaction poll following the Leaders Question Time special.

UPDATE2: The daily YouGov poll for the Sun has topline figures of CON 34%, LAB 35%, LDEM 8%, UKIP 12%, GRN 5%. Meanwhile the ICM/Guardian instant reaction poll following the Question Time special found veiwers thought David Cameron came out narrowly on top – 44% thought Cameron did the best, 38% Miliband, 19% Nick Clegg.

1,350 Responses to “Latest Ipsos MORI and Panelbase polls”

1 21 22 23 24 25 27
  1. I agree with Alec that the Election is slippling away from Labour, I’d be amazed if we wake up next Friday with EM in Downing Street in any shape or form

  2. @ John B

    All is sunshine & working hard here in Edinburgh. :-)

  3. Paul M

    I agree and have never said otherwise


    It does surprise me that so many otherwise sensible people are not prepared to admit the possibility that the polls may move a couple of points in the last week.

    The polls might move a couple of points in the last week. There. Satisfied?

  5. Catoswyn

    I shall add you name to my list of sensible people. Oh, sorry, it was there already.

  6. Getting the vote out may turn out to be what the turnout is all about in the marginals. I hate to keep on so – but with less than a week to go it bears repeating – if Labour gets its vote out in its marginals more effectively – a lower turnout in its safer seats and a higher turnout in safer Conservative seats will not in the end add up to less than a slight Labour advantage – as in February 1974.

    It is quite possible for Labour to squeeze the few extra seats which will make a Conservative government impossible whilst the conservatives may have more votes than Labour. The two jokers are – how well the LibDem vote will hold up in LibDem/Con marginals and how many Labour and LibDem seats will survive the SNP tsunami. Here again – we’ve seen an old governing party in NI disappear before our very eyes – UUP – but it did not stop ultimately progress on NI devolution and power-sharing.

    Sometimes political parties outlive the nomenclature of their original raison d’etre. Labour certainly is more than its Union origins and has happily been so since at least the 1920’s. whether this will be the future for the SNP only a clairvoyant can divine but power does something for political parties that makes principles highly mutable.

    That is not to imply politicians have no principles but it is a comment on the governing realities of political power – look only to Greece – ancient and modern.

  7. DAVE84

    Steve Fisher’s forecast probably does most to factor in shy Tories and award incumbency bonuses, and he’s just revised today up to C290.

    Once you start trying to project C290-300 into actual seats, you’re either arguing for holds that require a minimal swing to go L (e.g. Bedford) or gains from LD that do nothing to the path towards a working majority.

  8. Alan

    But the margin of error for a particular batch of polls only tells about that particular batch. They’re not necessarily representative of the country as a whole of even of all marginals.

    Indeed because Ashcroft is picking seats which look ‘interesting’ the odds are that they are an atypical bunch. If for example he is polling where the swing was smaller than expected (for local reasons) the odds are that they will still show small swings. It doesn’t mean that the swing can be applied to anywhere else – or indeed to any of the seats involved except where he is polling a homogeneous group (as with some Scottish seats).

  9. If ashcroft,opinium,icm and mori follow survation and comres back towards even steven then we have a clear trend.If not then its the prayer mat for tuesday and wednesday.

    I hear parking restrictions outside the hospital have been extended to tuesday -hint hint.



  11. Re: Ashcroft marginals.
    1. Ashcroft is using reallocation of DK, LTV AND SoS. Without these filters Labour are doing much better – including being level in Croydon Central (which I have no doubt Labour will win).

    2. Ashcroft uses phone polling. Usual caveats apply. There is no reason to believe Ashcroft’s phone polling
    would be any less Tory leaning than any other form of phone polling.

    3. Just because Ashcroft polls CVI often doesn’t make the. Any more or less accurate. YouGov also runs CVI polling and is showing bigger swings to Lab. Again, point
    2 above applies.

    Applying the above, I would say the real average swing in Con-Lab marginals is at least 4% and probably higher. I think we have to ignore random seats like Battersea. That would only fall to Labour in a landslide. Without it, even on Ashcroft, the Con-Lab swing is 4%.

  12. Labour rallying, I think, although the polls are still suggesting a dead heat.

    YouGov tonight will be interesting. This is career defining stuff.


  13. Looking at the Dumfriesshire, Clydesdale, and Tweedale Ashcroft poll showing SVI, (CVI) and 2010 result. I no longer think there is any realistic prospect of Con win in any of the 3 south of Scotland seats.

    Cons 30 (31) [38]
    Lab 16 (17) [28.9]
    LD 5 (5) [19.8]
    UKIP 5 (4) [1.4]
    SNP 41 (42) [10.8]
    Green 2 (2) [ 1.1]

  14. Alec
    Last night I was on the panel with A Salmond though there were more than 20 people (he may have had a tea-time meeting as well). It was the Scottish Farmers Union so I had a great time. Originally A Salmond was planning to miss much of the campaign because of a book tour but plans changed!

  15. Are the SVI and CVI measures now very close in Ashcroft’s polls? One might expect them to be.

  16. @Bramley – 5.18

    That would be Monday evening? Yes. Thank you.


  17. The polls suggest a slight drift to Con but no more than that. However just because 1992 underestimated Tory support does not mean it will happen again and again. Labour support might be under- estimated this time particularly in Scotland and it may well be that Labour do better than the polls are suggesting.

    A little counter- balance is needed to the mass of posts suggesting the Tories will do better than the polls suggest.

  18. Holgate

    I struggle to see cons lose less than 30 to Lab, and pick up more than 10 from LDs. That’s 287 (based on 2010 not byelections). Beyond that there are probably 5-10 qhich Lab are favoured to take depending on methodology, and another 10 they could take if there is no serious tory recovery (I’m basing figures on an expected C lead on election day of around 2%).

    Which is depressingly in line with the herd, with the exception of the Fisher model, and the odd person on here and elsewhere who believes this will actually be 1992. It might be, but the evidence is against that right now.

  19. @Peter

    More than 1100 now.

  20. @ ChrisLane1945

    The Feast Day of St Joseph the Worker.”

    Then it was moved from 19th of March … Perhaps to coincide with the International Feast Day of the Working Class (except for the UK, US) :-)

  21. Is “career-defining” a game-changer?

  22. @Mikey

    Discount Battersea and notwithstanding that Ashcroft polled some seats well beyond Labour’s “need to win” range, the swing was just about 4%.

    If I were the Tories, I would be concerned. Not worried, yet, as there’s still time, but definitely concerned.


    Didn’t much the same thing happen in a Canadian election? I think it created a bit of a backlash!!!

  24. SHELTS

    A very good question. Personally I find it dubious that it might but some people have suggested there might be a ‘feel good’ poll bounce from the combination of the bank holiday weekend and a new Royal Baby.

    If nothing else Cameron gets to go on TV first to be all Prime Ministerial and congratulate the royal couple.

  25. @Crossbat

    Rallying is pushing it on wheels! But the extent of Labour’s push in the marginals is becoming apparent.

  26. According to NC in the CON-defended seats Ashcroft polled twice, the CON-LAB swing has changed from 4.4% to 3.1%

    The big question is – can this momentum be sustained? Is this a blip or something which can be neutralised with Labour’s ground game, or can the Tories keep it going?

    I guess that’s the last we’ll hear from the marginals unless ComRes do another of their aggregate battleground polls.

  27. @Alec 5.37

    AS isn’t standing in the @rse end. That’s the seat west of Dumbarton which looks likely to switch from LD to SNP. (Hope your Gaelic is up to that!)

  28. I am sorry if this is repetition. I did review and could not see where this had been commented on previously regarding Ashcroft’s Scottish marginals.


    Please note that Ashcroft continues to weigh his polls results by recalled 2010 vote. This may make some sense in English and Welsh polls as they are not seeking a political tsunami underway. However, in Scotland, political weighing by 2010 recalled vote definitely impacts the final numbers.

    For example, in the East Renfrewshire poll, Ashcroft uses the following weighting.

    Weighted base Unweighted base

    218 216
    22% 22%
    366 347
    37% 35%
    Liberal Democrats
    67 62
    7% 6%
    83 152
    8% 15%
    Another party
    8 14
    1% 1%
    Did not vote
    177 107
    18% 11%
    37 48
    4% 5%
    Don’t know
    45 54
    4% 5%

    As you can see, the SNP vote is downweighted by almost 50% while the Labour vote is upweighted by 2%.

    When you couple this weighting change with Ashcroft’s personal “shy voters” adjustment, the race is much closer.

    Unweighted results: SNP 340 (49%) Labour 279 (40%) SNP+9
    Weighted adjustment: SNP 281(41%) Labour 241 (35%) SNP+6
    Weighted plus
    “shy voter”: SNP 39% Labour 36% SNP+3

    To accept that Labour is close in the race, you have to accept that the 2010 weighting reflects the 2015 race AND that the “shy voters” adjustment is valid.

    IMHO, I think the unweighted results are far closer to the reality.

  29. Alec

    Forgot to add: in every other respect I am in entire agreement with you. On this occasion…..

  30. I am betting literally on a 1970 and 1992 result and expect to get a good holiday out of the winnings.

  31. @Omni

    I think you have to take a look at the seats Ashcroft has chosen to poll twice and also the change in methodology. Why SoS now and not before?

  32. @David Colby

    The Scott Trust was wound up and it’s assets passed on to The Scott Trust Ltd, registered as a company in England, not as you seem to think in the Caymans. You can look it up on Companies House.

  33. Laszlo

    “The Feast Day of St Joseph the Worker.”

    Then it was moved from 19th of March … Perhaps to coincide with the International Feast Day of the Working Class (except for the UK, US) :-)

    Only by the (Roman) Catholics. Who you must admit have a long tradition of stealing other people’s feast days and calling them their own.

  34. @LASZLO

    Feast of St Joseph the worker created by Pius XII to counter communist use of May Day but Solemnity of St Joseph remains 19th March

  35. For me the key issue in the GE now in the seats Con v Lab marginals beyond the first 20 is can the Cons squeeze the Blue Kippers.

    The albeit unsophisticated churn numbers are suggesting that the 2010 LD vote dispersion is giving perhaps 15% Lab over Con of the LD vote per seat.
    Although, it may be a tad higher in Con/Lab marginals as the average is lowered a bit by a higher ABT retention in LD/Con marginals.

    The LD split then is only sufficient to bring perhaps 15-20 seats to Labour from the Tories (partly because as CS points out there are less LDs left to squeeze on average in these seats).

    This means that Labour need to gain over the Tories on top.
    They may do better amongst first time voters and may GTVO better but absent any Con-Lab or vice versa direct switching the Tories losing some votes to the UKIP is at the moment the decisive factor in current polls suggesting Lab taking 40 seats or so after incumbency is factored in.

    It seems to me that this is what Tory private polling is saying and they are pitching to the Blue Kippers (EU Referendum Red Line and EVEL plus anti SNP rhetoric) even if it risks a few centre voters, as they probably rightly calculate that they won’t lose any to Labour now ABLabs not keen on the anti-Eu stuff will stick with cons in Con/Labs seats rather than ‘waste’ their vote on the LDs.

    What CB11 would call my Pet Theory is that Blue Kippers in key marginals (as well as Red Kippers and Greens) will be squeezed hard and break in favour of the Tories seems to happened a bit already but not enough.

    It is possible the pro-Tory polls are the right ones otherwise the outcome of the GE imo has become dependant on a few thousand Kippers in the marginal and how the end up voting.

  36. @ David

    “Why would there be just shy Tories but not shy Ukipers?”

    This is a much more interesting point than many realise I feel.

    If any party has people who are vehemently or vociferously against it, either because the party is hated by some, or ridiculed by some, or both – then there will be shy voters of such parties.
    In this context I suspect there are shy voters of ALL parties, dependent upon what regional or social situation the followers of these parties find themselves in. For example, it might be slightly embarrassing to be staunch Labour and a member of a Surrey golf club. It may be embarrassing to be a Tory and a member of an ex-mining area working man’s club etc etc.
    However, it terms of even more significant numbers, parties which are “social death” or “social ridicule” to supporters within their social circles, whatever they may be, are likely to be those that attract strong responses.
    I suggest at this election this applies to
    a) SHY Tories (for fear of being seen as fusty, selfish or imbued with grasping mean wickedness amongst many on the Left)
    b) SHY Ukippers (for fear of being portrayed as closet racists etc)
    c) SHY LibDems (for fear of being ridiculed for their ‘hapless’ role in the coalition and for having the ‘dishonesty’ of Clegg)

    The last of these three I suspect have been underestimated more than the first two?

  37. @JimJam

    But when is a marginal a marginal. Labour are ahead in some in the 70s/80s plus on their target list. It’s not just about the top 40. And as I said above Ashcroft’s methodology may turn out to be right or wrong but it tends to show lower Con-Lab swings than other polling. Why is he necessarily right? And even if he is right that still means Lab will pick up around 40 Tory seats.

    Yes – it’s close. But the Tories need a pretty decent swing to prize Labour out of poll position now.

  38. DAVE84

    Agreed. There’s a point at which you start hitting hard boundaries on the constituency level either side that don’t map to Ashcroft or battleground polls or national projected E&W swing, and at that point you’re either pulling rabbits from a hat or shrugging.

    We know that there’ll be the odd surprise on a constituency level, because there are always surprises, but the reason they’re surprises is that they don’t show up in any kind of polling.

  39. The Duffy/ Prescott punch moment of this election has finally happened

    Clegg handled it quite well

    Nick Clegg [email protected]_clegg · 5m5 minutes ago
    Some people may not have been heeding my warnings about the need to tighten our belts.

  40. RAF

    your assumptions in point 2 and 3 of your post are on the basis that phone polls are less accurate than online polls. Just because phone polls seem to lean Tory doesn’t mean they are any less accurate than the online polls that seem to lean Labour.

  41. @raf

    A few of the pollsters seem to be switching on spiral of silence adjustments now. I don’t know why they didn’t do it before.

    @jim jam

    Pretty much agree

    @tony dean
    “SHY LibDems (for fear of being ridiculed for their ‘hapless’ role in the coalition and for having the ‘dishonesty’ of Clegg)”

    Ugh, yeah, I’ve stayed quiet about my voting intention since the election came up in conversation with some family they asked who I was voting for, and they expressed shock and despair at my choice then tried to talk me out of it for ages. Conversation got heated.

  42. Even with swing, the results are not uniform. It will depend on how well the LibDems hold what is left of their vote and how much UKIP takes in individual seats.

    However, the results from 2010 may provide a guide. With a 5.1% swing from Labour to Conservative, the results were Conservative +91 seats and Labour -97 seats. Also keep in mind, that no Labour or Conservative seats changed in Scotland so all this was in England and Wales.

    Thus a 4% swing from Conservative to Labour (a minimum swing at this point) could see the Labour gains range from 30 to 60.

  43. @ Roger Mexico, @ John Murphy

    Thank you both.

    I sometimes get confused. It is better not continuing it, but it is good to learn something (every day here).

    As to feast days. The 2nd Internationale chose this day not quite accidentally.

  44. @Nick Shaw

    Correct. I may be wrong about phone polls. We shall see.

  45. Jimbadger

    I just wonder how many shy UKIP voters there are, especially in traditional Labour areas. I recall two polls shortly before the Heywood and Middleton by-election both gave Labour a 19 point lead, but UKIP almost won on the day!

    It’s an interesting point. The last poll before the H&M by-election was for Ashcroft[1] with fieldwork 30 Sep to 4 Oct – so ending 5 days before polling (you could probably say a week as more polling is done towards the start of the period):

    what is interesting is the big gender gap for UKIP 23% of men but only 11% of women. Some of this is genuine – women are a bit less likely to support UKIP and the Labour candidate had 28% of the female vote but only 20% of men. But women are more likely to say they won’t vote or are unsure and it may be some of these are ‘shy’ UKIP voters who may see saying they will vote UKIP as socially unacceptable or disloyal.

    Now there will have been other reasons for the UKIP ‘late surge’ – differential turnout in a low poll in particular. But there could well be some shyness as well.

    [1] Incidentally Ashcroft did use the candidate names for his by-election polling, so it seems odd not to do so now for constituencies after nominations have closed.

  46. @ Tony Dean

    I still am of the belief that the Lib Dems will do much much better than what the polls are currently saying.

    Nick Clegg has done well to get his message across and I think a loss of 20 seats will be the outcome.

    If I were not such a staunch Tory I would probably vote for them as the last 5 years have proved they are not a bunch of tree huggers with ridiculous policies(that prize now goes to the Green party) and will indeed make tough decisions for the good of the country, not for the good of the party.

  47. Whilst it makes perfect sense to average all the polls to iron out the discrepency of outliers etc., there is an inherent danger in doing this in a “dynamic” situation. What would be a good deal more reliable would be to have a series of snap-shot polls throughout a campaign of the same cohort of people who were found to be genuine swing voters living in the various categories of marginal seat at the opening of a campaign. A thousand of such people would give us polls that could really determine who was winning where.
    Without such refinement, I suspect all pollsters are taking a massive gamble, to which they all apply their best scientific methods, a bit like gamblers who have developed highly sophisticated methods of playing the roulette wheel. In an election like this, with so many new and unclear factors I don’t envy them……let’s hope our host and his friends are lucky!!!!

  48. Also people I think we’re starting to see polling convergence around a statistical dead heat (©, @Crossbat)).

  49. Another ComRes phone poll tonight. Maybe we won’t see convergence after all :(

  50. Labour MP Ian Davidson publicly criticises Jim Murphy (“he is not a particularly stimulating leadership figure for us”) and calls for Ed Miliband to take over the Scottish campaign.

    ” Arguing that Mr Murphy’s tactics are not “breaking through”, he attacked the decision to focus the final week of the Scottish Labour campaign on warnings that the SNP will use a landslide next week to push for a second independence referendum.

    Mr Davidson said this would not work in his constituency of Glasgow South West, where he said the majority voted Yes in last year’s referendum and would like another chance.”

    Those tactics might work in East Renfrewshire, though…


1 21 22 23 24 25 27