The Guardian released their latest poll from ICM earlier today. Topline voting intention figures were CON 42%(nc), LAB 39%(-2), LDEM 8%(+1). Changes are from early April and fieldwork was Friday to Sunday, so once the Windrush scandal was in full throw but before Amber Rudd’s resignation. While the changes are well within the normal margin of error, three points matches the largest Tory lead ICM have recorded since the election (and bear in mind the last two weekly YouGov polls also showed the Tory lead at a post-election high). In other words, while I would still urge caution about reading too much into a couple of polls showing a similar pattern, it’s possible that the Conservatives are opening up a small lead. It is worth keep an eye on at least.

Voting intention polls don’t tell us that much at this stage of the Parliament, but if the Tories have improved their position relative to Labour over the last few weeks then perhaps Thursday’s local elections may not be so bad for them as they might have been (for what it’s worth, when the same council wards were last fought in 2014 the Labour party had a small lead in the national polls… but of course, the polls back then were likely over stating Labour, so that does not necessarily imply a swing to the Tories). On the subject of local elections, Survation have said they’ll be releasing some London local election polling overnight.

Turning back to the ICM poll they also asked about the impending visit by Donald Trump. 33% of people said they supported Trump’s visit, 31% were opposed, 33% neither supported nor opposed it. Full tabs for the poll are here.

UPDATE: As well as the ICM poll, there is also a new ComRes poll out for the Sunday Express. As with ICM fieldwork was Friday to Sunday, so at a time when Windrush was all over the news, but before Amber Rudd resigned. Topline figures with changes from mid-April are CON 40%(nc), LAB 40%(-1), LDEM 9%(+2). The parties are neck-and-neck, rather than the slight Tory lead we’ve seen in other recent polls, but it does not suggest that the Windrush scandal has had any impact.


183 Responses to “Latest ICM and ComRes voting intention polls”

1 2 3 4
  1. That’s a disappointing poll. I would have expected Windrush to have knocked Con down a point or two.
    I still think that will happen and evidenced in forthcoming polls.

  2. MP

    Don’t worry somebody will be along in a minute to say it’s an outlier or in England Labour are ahead ????

  3. Looking at Britain Elects tweet and the last ICM, shouldn’t that be Lib Dem +1?

  4. Mike – we’ve now had several post-Windrush polls and none have shown a Labour lead. I absolutely agree the Tories do not deserve still to be ahead but it is unreasonable to expect the polls to turn now, when the news agenda seems finally to have moved on to other things. Largely because of the way the Westminster press pack operates, Amber Rudd’s resignation represents closure on the story as far as many of them are concerned. I do hope Amelia Gentleman keeps going with this story but I think it has mostly passed in terms of pubic recognition.

    It’s an unusual situation for Labour: they have absolutely triumphed inside the Westminster bubble, with even the much-maligned Diane Abbott getting well-deserved praise for her powerful interventions, but they have fallen somewhat flat outside it. Totally the opposite of what happened last June.

  5. I did, of course, mean pubLic recognition. That’s one unfortunate typo…

  6. @Turk

    OK, in order not to disappoint you, I think this poll is an outlier and Labour are ahead in England!

    :-)

    More seriously, and I’m surprised Anthony thinks slightly otherwise, these national polls aren’t usually borne out in local elections with voters behaving in different ways to how they might if it was a GE. Local issues give rise to erratic results that buck national opinion polls. On the projected national vote share derived from Thursday’s elections (BBC usually do one), I suspect both Labour and the Tories will be some way below these 40% VI shares that are being shown for both of them in the recent national opinion polls. I have a feeling Labour will have a good night, and the Tories a bad one, but I don’t think this is going to tell us very much about a future GE result. Where they can have longer-term implications, however, is the effect they have on party morale and the increase/decrease in councillor strength for the respective parties. Councillors tend to be the basis of your local party organisation and the loss of them weakens a party’s campaigning potential. Incumbent governments usually suffer the worst attrition in this respect and can pay a price further down the line. Foot soldiers are still important, I think, especially when the parties are neck and neck and when getting the vote out becomes absolutely critical..

    Anecdote alert, I accept, but I’m getting quite surprised by the number of 2017 Tory voters saying that they’re going to vote Labour in these local elections. Forget Brexit, antisemiticism, Windrush, Salisbury etc; local council cuts, housing and the loss of public services are the big issues on the doorstep. I think the Tories will be punished.

    Whatever, it will be an interesting night on Thursday. It always is when real votes are cast.

  7. Crossbat – interesting. Remind me where you live again?

  8. Thanks for the correction

  9. Turk

    Usefully, ICM also give an England crossbreak – but they show Con at 44% and Lab at 40%. You really mustn’t let your paranoia – even if everybody is out to get you!

    Quite why ICM have 8% of those in the North [of England] voting SNP isn’t clear.

    An ICM error? : Mis-keying responses? (though it’s hard to see how} : Having a vote in Scotland? (possible, but unlikely on that scale) : wanting to be able to vote for a party that they see delivering competent governance to their north? (sounds most likely to me! :-) )

    In so far as one can deduce anything from the wee Scots sample (which isn’t much!) I find the SGP VI interesting.

    I haven’t been tracking this, but my impression from both Scottish and GB polls is that they are securing a significant toehold.

    It seems unlikely that there will be many SGP candidates for Westminster, and a large chunk of that vote will end up voting SNP.

    What will happen with the Green Party of E&W VI in their strongest areas (according to this poll) – Wales and Northern England (both 5%) – I have no idea.

    The VI breakdown for Northern England may be wholly distorted (since it isn’t internally demographically weighted), but 23% choosing parties other than Lab/Con seems a tad unusual.

  10. While our ward is pretty solid Labour, Greens and Lib Dems used to contest it (also the Trots, Tories and UKIP). There hasn’t been a single leaflet apart from Labour (2, plus a bit very committed canvassing effort, but then they used their database – it was in their hand :-) ) – quite unusual.

    Anyway, I think the voting in the local elections will show too much fluctuation to draw anything from it to a GE voting intention.

    However, the earlier point of losing many councillors is a very important one (we have seen that we LibDems).

  11. I think VI is “decided” subliminally – obviously via the news/media – and feel that Salisbury/Syria and Corbyn’s response may have had the most effect.

  12. @Polltroll

    “Crossbat – interesting. Remind me where you live again?”

    Redditch. Labour controlled council but on a knife-edge with a majority of just one. Only needs one ward to change hands and it goes blue. Not much Lib Dem, Green or UKIP support in the town and it’s more or less a straight, old fashioned Labour v Tory fight. Over the years we’ve had both Tory and Labour councils, but never one, for either party, with such a small majority. Accordingly, it’s all getting a bit untypically nasty with some pretty unsavoury campaigning going on. The new local MP, Tory Rachel Maclean (she replaced a seriously ill Karen Lumley as the Tory candidate in 2017), is a bit of a strange one. Unlike Lumley, she looks to have carpet-bagged the seat and comes over as a serious careerist of no little ambition. She’d love to tell Brandon Miller about how she returned a Tory Council to Redditch! I’m hearing she’s got a team of young IT buffs cranking up the Tory social media campaigning efforts. We haven’t quite got to claims that Labour has introduced hepatitis to the town yet, but it’s all getting a bit unpleasant.

    My gut feel is that Labour will increase its majority on the council, but I wouldn’t bet the total house on it.

    Maybe the garden shed!

    :-)

  13. That of course should be Brandon Lewis, the Tory Party Chairman, not Brandon Miller.

    Brandon Miller was a very useful South African all rounder who played as an overseas pro for Studly and Astwood Bank for many years!

  14. Crossbath

    I try and put that smiley face thing in when I’m joking for some reason those question marks come up instead. I suspect operator error.
    As to the council elections given that the majority of seats being contested are already Labour held even if the expected Labour victory takes place it would be difficult to see how that could be extrapolated as overwhelming support for Labour in a GE I seem to remember Hague did very well in council elections but failed to gain power in a GE.
    Personally it looks at least in London the Tories will have a bad night I wouldn’t be surprised if the Liberals didn’t produce a few surprises as well. Of course the danger for Corbyn is if Labour don’t do as well as expected
    I would say that brexit is likely to be the thing that brings it all to a head May could carry on till 2022 but my guess is that’s unlikely and. GE will take place sometime during within the next year

  15. Good luck Crossbat, sounds like it will be a close one.

  16. Turk

    “my guess ……….GE will take place sometime during within the next year”

    Since we don’t have elections this year, but have had a couple of SNP leaflets hand-delivered and a postal “survey” from the SCon List MSP, I suspect that parties here are anticipating that your guess may be right.

  17. via Number Cruncher

    ComRes/Daily Express:

    CON 40 (=)
    LAB 40 (-1)
    LD 9 (+2)
    UKIP 5 (+1)
    GRN 3 (+1)
    SNP 3 (=)

    27th-29th Apr (Changes vs 11th-12th Apr)
    N=2,030

  18. Back in May 2014 the Tories were nearly dead in Scotland if I recall and Labour still hadn’t suffered the post referendum meltdown there, now they are ahead of Labour.

    So that would indicate we should see a better result than 2014 for Labour/ worse result for the Tories in the English local elections if the national polls are looking similar to what they did in 2014?

  19. Are Labour losing Remainers to LD?

  20. Richard

    “now they [SCon] are ahead of Labour”

    Not according to polling

    https://en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election#Scotland

  21. @Oldnat

    Well, nit picking here – in the most recent poll you linked to, Scon = 28, Lab = 27. That’s a lead. But basically even.

    Go back to 2014 and it was Slab = 36, Scon = 16%.

    So clearly a big swing from Slab to Scon, which means it all things are equal, there has been an opposite swing outside of Scotland…

  22. Looking at London

    2014 polls
    https://en.wikipedia.org/wiki/Opinion_polling_for_the_United_Kingdom_general_election,_2015

    Con 31, Lab 44, Ukip 10, LD 11

    And latest London poll last week shows
    Con 31, Lab 52, UKIP 2, LD 10

    So London should show very good results for Labour on Friday.

  23. Richard

    “So clearly a big swing from Slab to Scon”

    Can I gently suggest that you’re idea of “clearly” is somewhat opaque or, alternatively, representative of someone whose ocular performance is limited by Kalnienk vision?

    There has been huge churn in Scottish VI as a result of the two referendums. That much is obvious from looking at the actual voting in the 3 Westminster elections of 2010, 2015, and 2017.

    Party votes in these elections –

    SLab 42% : 24% : 27%
    SNP 20% : 50% : 37%
    SCon 17% : 15% : 29%

    You may be a better man than me, Gungha Din – but you need to do better than that! :-)

  24. @Oldnat

    Ok, sorry, I don’t get it. My definition of swing is net voters have moved from one party to another. I don’t care if some went from Lab to snp, and some from snp to cons, I am purely interested in total Lab and total Cons to see how they have changed and influence England results on Friday.

    Looking at Wales:
    2014
    Lab 43%, Cons 22%

    2018
    Lab 46, Cons 33%

    So again a swing to Cons, outside of England.

    Another data point indicating Friday is going to be a good day for Labour, as my understanding is that is limited to English local elections, where there has clearly been a swing to Labour.

  25. Richard

    The idea of “swing” was developed during the time that there was (largely) a two party system in a GB polity.

    In those circumstances, it is a useful construct.

    It’s largely meaningless in a three or more party system. Trying to impose a construct from one polity (much larger in the past, but still possibly useful in the reduced polity that England has become) to a different system of politics is a pretty worthless exercise.

    Since England continues to have a two party system, trying to draw analogies from other polities, with different structures and recent political history, also seems pointless.

  26. @ HAL

    “Are Labour losing Remainers to LD?”

    This is a good question, and one I was about to pick at. If we accept (as I think is reasonable) that gradually between August last year and now there has been a swing back to Conservatives, ie Lab were just leading after the GE, but Con are leading now. Then let’s look at the Tory share. Really it’s difficult to discern anything other than very stable* around 40%. The swing is essentially due to the Lab VI weakening a bit. Now maybe they’ve mostly gone to Don’t Know, but maybe there’s some hint also of leakage to LibDem. Small, yes, but it could be there, and if so, most likely it’s due to Remainers who have realised Labour are not going to help them much.

    So while this is very speculative, and of course Westminster VI, particularly for LibDem, is going to be muted by FPTP effects, it could be a hint of a rather more significant LibDem revival locally. Or not. We shall see.

    * Strong and stable, obviously

  27. Survation London poll out at midnight apparently.

  28. Looking at different polls seems Labour got London, probably Manchester,
    but less successful elsewhere in England.
    If Labour dominate London and do much less well outside London in English regions, then the Government would be wise to move every lever of influence out of London before the next GE.
    Wembley, Channel 4, everything. Even Parliament itself and Whitehall Departments. Sell off buildings and pay down deficit.
    Total reverse of three governing parties since 1982. London population was falling until 1984.
    Good idea for government to shove part of Channel 4 to DUP constituency.
    Punish London for voting left and Remain. Reward provinces for supporting Government and Leave. No more Mr or Mrs Nice Guy.

    These locals seem bigger than usual. Wrong result could take down Prime Minister or shore up Prime Minister for a few years. Could crush Labour Westminster rebellion to leader or lead to new de facto vote of no confidence in Corbyn. Even a split in party at Westminster and maybe locally.

    The locals may trigger an early General Election or second referendum.

    I blame the lack of weekly and bi-weekly opinion polls since last June which makes this seem a meaningful deciding ‘rematch’ of the 2017 bout in the way David Haye is having a rematch with Tony Bellow.
    Same polling as Labour. London going Haye. Regions nudging towards Bellow.

  29. For immediate release.

    London Survey on Poverty and Wages.
    Local Elections Poll
    Sample size: 1,005 Fieldwork dates: 27th – 30th April 2018
    Methodology: People aged 18+ in London interviewed online

    Survation. on behalf of 4in10

    As with much of the country, London goes to the polls on Thursday, with Labour leading the Conservatives 51% to 31% the Lib Dems on 12%, others 6% in online fieldwork conducted by Survation 27th – 30th April

    Parties should not take voters for granted however, with one in 10 voters telling Survation they are currently undecided on which party to vote for, 61% of Londoners say they would be more likely to vote for a party that made a commitment to improving the lives of the poorest children (Table 21)

    81% think local councils should pay all their staff the Real London Living Wage (Table 23)

    Only one third of Londoners polled think their local council is doing everything possible to tackle child poverty (Table 18), with 65% wanting their new Council Leader to make a strong commitment on the issue (Table 19)

    Those we polled are optimistic about what could be done: 75% say councils can take action to protect tenants from rogue landlords (Table 28)

    Three quarters of Londoners also want their council to increase the availability of social housing (Table 27), while 85% would feel positively about their local council if they were to take measures to improve the quality of the private rented sector (Table 29)

    63% believe parents in part-time work can become trapped in low-paid positions by a lack of quality flexible jobs (Table 20), and 71% want their local authority to promote flexibility in work patterns (Table 11)

  30. Seems May continued to enforce her rigid anti-immigrant policies, even after becoming PM

    https://www.ft.com/content/ab8a2cce-4d5b-11e8-97e4-13afc22d86d4

    FT says Theresa May rejected pleas from her former home secretary to exempt doctors from quotas on highly skilled workers coming to work in the NHS.

    Damaging the NHS, in order to continue with a policy that seems to have been part of her core beliefs, is not a good recommendation for the continuation in office of even the most junior Minister.

  31. CMJ

    “61% of Londoners say they would be more likely to vote for a party that made a commitment to improving the lives of the poorest children”

    Sadly for them, the SNP isn’t standing in London.

  32. @Catmanjeff

    Thanks, looks similar to the Yougov results from last week.

    Does anyone know how EU citizens are polled, how many are registered?

    http://www.bbc.co.uk/news/uk-england-london-43922510

    Lots of stories saying they could be what swings some of those mid London Tory seats red, but are these stories just trying to get them to vote, or are they really likely to come out in force to punish the pro Brexit parties in their first and last vote on the issue?

  33. CMJ
    “Methodology: People aged 18+ in London interviewed online”
    “61% of Londoners…”

    Shouldn’t that be 61% of Londoners who can be bothered to be interviewed online?

  34. No sign of Survation as yet. They tend to get terribly excited and then not deliver on time [insert own sex joke here].

    The ComRes is for the Daily Express rather than the Sunday version, though unlike other brands (especially the Mail/MoS) there’s not the same tension between the Daily and Sunday. The commentary from ComRes:

    http://www.comresglobal.com/polls/daily-express-voting-intention-and-life-in-britain-poll-may-2018/

    which reflects the questions that the paper wanted asking, is worth quoting in full because it is just so Express:

    The Conservatives and Labour are neck and neck, with 40% of the vote each

    Despite the two main parties polling neck and neck, those who disagree that the economy would be stronger if Corbyn was PM outnumber those who agree by more than two to one (51% to 24%)

    Just over one in five (22%) overall say international confidence in Britain’s economy would be stronger with Corbyn as PM, with more than half of voters (52%) disagreeing

    By a ratio of more than two to one, British people feel there has ‘probably never been a better time to be alive’

    ‘The beauty of the British countryside’ tops poll of what people most value about Britain
    85% think the NHS ‘rightly a tremendous source of national pride’

    Also, by more than two to one, people feel Britain can be proud of the impending Royal Wedding and birth of Prince Louis

    The only thing that is missing is a question about a miraculous new cure for Alzheimers/Arthritis/Cancer/Alzheimers.

    Tables are here:

    http://www.comresglobal.com/wp-content/uploads/2018/05/Daily-Express_May-2018_02052018.pdf

  35. Pete B

    “Shouldn’t that be 61% of Londoners who can be bothered to be interviewed online?”

    This is a polling site. Almost everyone on here understands what a “representative sample” is.

    How come you seem to have missed out on what almost everyone else here knows?

  36. OLDNAT

    Quite why ICM have 8% of those in the North [of England] voting SNP isn’t clear.

    Because ICM are firm believers in One Britain and their North British region includes North Britain (and Wales is included in the Midlands. If you look at the sample sizes, you’ll see that the three (N, Mid, S) add up to the total sample.

    Given the separate political landscapes in Scotland and Wales, I have no idea why they do this, but they always have. Possibly since 1603.

  37. ON
    As usual, you are being deliberately provocative. Can you or anyone point to research that shows what weighting is appropriate to compensate for obvious potential distortions in the online audience – e.g. potentially differential responses by age, ethnicity, gender etc etc?

    I can’t be bothered to have an endless nitpicking debate with you, so I’m off now. I may look at any response tomorrow.

  38. Roger Mexico

    Thanks for the clarification.

    Perhaps 1603 is a bit late, though? 1292 might have been the start of ICM’s imperial ambitions to include Scotland in the North..

    It’s really sad that they don’t also include France in the South. That the Crown of England gave up any pretensions to also rule France in 1800, would appear to be a very lame excuse!

  39. On the basis of Survation, I’d like to cede South London to Surrey, or better still to Croatia.
    I think (based on doorstep) that Lab will lose a smattering of votes to LD, entirely over Brexit.
    The Tories round here are a shambles, struggling to find candidates and with ‘Out’ cards delivered wrapped in leaflets. They are dumped in piles in the foyers of blocks of flats, making it perfectly plain to anybody who cares to read them that the claim they knocked on the door and got no response is an obvious lie.

  40. Pete B

    If you consider that all the polls run by Anthony and his colleagues as well as many other pollsters – are so flawed, as to be dismissed, I do wonder why you bother to log on here.

    However, Anthony has helpfully provided a guide for those who ask daft questions about polling.

    Here, for example, is his article on sampling.

    http://ukpollingreport.co.uk/faq-sampling

    If you want to engage in a debate with him as to why you know more than he does about polling – its benefits and pitfalls – far be it from me to dissuade you.

  41. @Guymonde

    Any indication of EU citizens coming out in force? Who are they voting for?

  42. OldNat

    I suspect he likes “nitpicking fights” far more than he lets on!

    Doesn’t seem to be that interested or informed about polling though.

  43. Richard

    Thanks for the link. Survation always used to put things up first on Twitter, but it’s clearly going out to the mailing list first now.

    It’s an online poll weighted to ‘all adults’ in London, though they don’t say whether many EU citizens are included in their panel – but then knowing how many would vote is another issue.

    Rather than the usual Inner/Outer divide (itself a moveable feast), tehy use five sub-regions, which I assume are these:

    https://en.wikipedia.org/wiki/List_of_sub-regions_used_in_the_London_Plan#From_2011

    On the whole things look plausible, though the Lib Dem figure of 20% in Central looks high. The London VI after LTV is:

    Con 31% [31]

    Lab 51% [52]

    Lib Dem 12% [10]

    UKIP 2% [2]

    Green 4% [3]

    Other 1% [1]

    [] = YouGov earlier the same week, so very little difference. There is a definite (though not enormous) swing to Lab from Con since 2017, which we don’t see in national polls, which reflects what some are saying from the doorstep.

  44. PETE B

    Can you or anyone point to research that shows what weighting is appropriate to compensate for obvious potential distortions in the online audience – e.g. potentially differential responses by age, ethnicity, gender etc etc?

    Pollsters all have their different methods and tend to guard the details of them carefully. There are actually two different sorts of ‘differential responses’ and they try to cater for both.

    The first is in responding to the survey. So if women were twice as likely to respond in the expected time frame as men were, you would send out twice as many surveys to men as women hoping to get the same number of responses from each. (Obviously I’m simplifying the figures here).

    The second is reflecting how likely people are to vote. Pollsters cater for this by adjusting the targets they weight to. So if under-25s make up 10% of the voting population, but are only half as likely to vote, rather than weighting to a target of 100 in a sample of 1000, you would weight to a sample of 50.

    For various technical reasons, pollsters prefer to try to match the sample they actual get to what they want by the first sort of adjustment rather than relying to fix things by using a lot of weighting to get the sample to match the population.

  45. Trigguy,
    “, most likely it’s due to Remainers who have realised Labour are not going to help them much.”

    Despite arguing that local results might not be easy to predict from a poll which asks question about national results, it is possible voters minds are concentrating on the upcoming locals, not answering about some hypothetical future national election.

    So it might be a swing to libs represents peoples decision about how they will vote in the local election, not how thy will vote if there were a national.

    If your intent was to express dislike of Brexit, then a protest vote directed to the libs might be logical, to give a warning shot to both tory and lab. The logic of the nationals was that only the labour party could do anything about Brexit, so a vote for libs would have been wasted in achieving this goal. But at a local level this doesnt really matter. It is classic protest territory.

    And then, there might be more places in a local election where a vote for lib makes sense if you want to displace a tory (or indeed labour) candidate. I’m not sure how the proportion of seats where a lib is challenger compares in local elections rather than in nationals?

    Having said that, I think the lib brand image remains tainted by their coalition with tories, which indeed resurfaced during the windrush thing.

    “let’s look at the Tory share. Really it’s difficult to discern anything other than very stable* around 40%”

    That was a feature during the recent election too, I seem to recall, tory share (more) static while labour share changed.

  46. pete B,
    “Shouldn’t that be 61% of Londoners who can be bothered to be interviewed online?”

    Of course, but you can only poll people willing to talk to you and then do your best to compensate with weightings. These people will show changes of view which reflect society as a whole.

    But also, it doesnt very much matter what people think who care so little about politics they never bother to vote. They will exclude themselves from the vote as well as the polling.

    Except if a party suddenly manages to motivate the politically apathetic, and then it can upset the polling. Brexit has done this (on both sides)

    Roger Mexico,
    “So if under-25s make up 10% of the voting population, but are only half as likely to vote, rather than weighting to a target of 100 in a sample of 1000, you would weight to a sample of 50”

    The danger though is if pollsters use plan A, and only ask 50 people instead of 100, they are in a poor position to catch a change from historic behaviour.

    But they know that, polls are small because numbers cost. But it is a potential risk when we end up arguing whether turn out is changing in certain groups, and this methodoloy biases a sample against spotting such a change.

  47. Hi all – couple of questions about the locals, if these have already been discussed then please accept my apologies in return for a quick precis!

    Firstly what’s the feeling on how to gauge the extrapolated national figures? Presumably as these elections are rather more in Labour areas than Tory ones then the evaporation of the UKIP vote should give more boost to Lab than Con relative to 2014? So how much improvement over 2014 is a good night for Labour?

    Secondly are the London councils generally counting overnight or on Friday?

  48. Hi all – couple of questions about the locals, if these have already been discussed then please accept my apologies in return for a quick precis!

    Firstly what’s the feeling on how to gauge the extrapolated national figures? Presumably as these elections are rather more in Labour areas than Tory ones then the evaporation of the UKIP vote should give more boost to Lab than Con relative to 2014? So how much improvement over 2014 is a good night for Labour?

    Secondly are the London councils generally counting overnight or on Friday?

1 2 3 4