Two new polls today:

Survation for Good Morning Britain this morning had topline figures of CON 43%(nc), LAB 37%(+3), LDEM 9%(nc), UKIP 4%(nc). Fieldwork was on Friday and Saturday and changes are from the previous week. Tabs are here.

ICM for the Guardian had topline figures of CON 45%(-1), LAB 33%(+1), LDEM 8%(nc), UKIP 5%(nc). Fieldwork was from Friday to Monday and changes are from the ICM/Sun on Sunday poll at the weekend. Tabs are here.

Both of today’s polls continue to show movement in Labour’s favour, thought the overall lead is different. A six point Tory lead would represent a small swing towards Labour, a twelve point Tory lead would still give them a stonking great majority.

The key difference between polls showing large and small leads is, as I wrote at the weekend down to how pollsters are treating turnout. There are lots of differences between different polling companies methods: they sample differently, weight by different things, do different things with don’t knows and so on. However, right now the one really huge difference is turnout. Weighted with all its normal demographic and political weights, ICM would have shown a Tory lead of only 3 points – that was transformed into a lead of 11 points by the turnout model, which predicts how likely respondents are to vote based on the estimates of turnout by age and class at the last election (the change from 11 to 12 points was the reallocation of don’t knows). That’s a big change, but given the errors in the polls in 2015 that may be necessary. On the other hand, if Jeremy Corbyn has managed to enthuse young people and there is a higher rate of turnout among younger voters than in 2015 then it risks understating Labour support. We shall find out next week…

1,134 Responses to “Latest ICM and Survation polls”

1 2 3 23
  1. First

  2. Damn, second!

  3. HOORAY !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    No silly posts yet.

  4. No partisan posts yet, let’s see who’s the first.

  5. Given that these two polls are at the opposite spectrum in terms of methodology changes since ’15 is it fair to split the difference and say Tory lead is currently 8%?

  6. Nice to see no partisan posts yet.

  7. long day, I meant 9%

  8. @RedRich,
    Wouldn’t that make it a 9 point gap?

  9. Is there a YouGov tonight?

  10. Looking at the effects of weighting by dint of turnout in these polls, it seems to me that if Labour are to make a fist of this election then they have to confound the pollsters judgements on differential turnout amongst varying demographic groups.

    Maybe, one little glimmer for Labour might be their supposed vast army of foot soldiers working on the ground. If they can get the potential Labour vote out on June 8th, in ways that they haven’t been able to do in past elections, then the result might be closer than we think.

  11. First Partisan poster spends an hour alone locked in a room with naked JC.

  12. Richo. I would expect the YouGov tomorrow night

  13. If this trend continues where would that leave us come election day?

  14. Philtres,

    Is that meant to be a punishment or a reward?

  15. @Philotes

    Thats not much of a threat. Change it to Diane Abbot and you are talking.

  16. Rob,

    In the dark.
    Waiting anxiously for the first results to tell us whether even the exit polls are remotely correct.

  17. RR

    Probably where we started before TM called the election.

    Would you bet against all the political hot air changing nothing?

  18. @Paul H-J

    I think it would be partisan to comment.

    See what you did there! Trying to trick me.

  19. This General Election I’m going by the “smell test” for polls. Does it ring true?

    If Corbyn is 20 points behind on best PM, if he is 20 points on being able to manage the economy, if he is still -15 points approval and May about +15 approval (and these are all pretty accurate scores at the time of writing), then I am very much going with the larger Tory lead.

    I remember in 2015 I could get why Ed M would be behind in all these key factors (you know like, “Do you actually want him to be PM?”) and the polls were showing a dead heat.

    On so many objective benchmarks, including taking the economy into account, it feels like a 10 point Tory lead campaign so far …

    (and that is of course before Corbyn’s Governor Perry moment).

  20. On Corbyn’s ‘Women’s Hour’ performance, the ES has a long write up on it which was not too flattering. However, having just watched on line its not so damning. In this neck of the woods the former may have more influence on how this incident is viewed – but I very much doubt if it will have any noticable impact on VI.

  21. “If they can get the potential Labour vote out on June 8th”

    We’re probably getting to the point where long-term weather forecasts for the 8th might start to become believable. Given the reputed effect of weather on voter turn-out, should we start gathering weather reports now, as well as poll results?

    BBC says fine, but not too warm for Birmingham on the 8th. I may keep you updated.

  22. @RedRich,
    There is nothing particularly scientific about splitting the difference. Just because one poll says 6 and another poll with different methodology has a lead of 12 doesn’t mean the truth is somewhere in the middle. The truth could equally well lie at one extreme or the other.

  23. Does anyone think one explanation for some of the narrowing of the polls is that at the beginning, there were shy Labour voters and brazen Tory voters.

    Then the manifesto week…suddenly, some Labour traction and it was no longer necessary to vote Labour shyly. Meanwhile, the Tories who started cockahoop suddenly felt a little embarrassed over their badly communicated manifesto to admit their allegiance and metamorphosed into shy Tories.

    If so…. almost impossible to catch this in the polls which makes next Thursday’s exit poll all the more tantalising.

    Over to John Curtice!!

  24. It’s 8.83 if you average the latest poll from each of the 6 pollsters who have published in the last few days.

    I don’t suppose that figure is really that helpful as the differences between pollsters aren’t because of random error. If it turned out like that it would save them all embarrassment though. If either ICM or YouGov are correct the other is going to look pretty silly.

  25. Andy T,
    “Would you bet against all the political hot air changing nothing?”

    It would certainly be an interesting result… Wonder how it would affect the respective party leaders’ political survival?

  26. I have this sneaking feeling that the polls are grossly understating the Labour turnout as an over-compensation of the last lot of polls.. and as such we may see a better Labour performance than the polls predict…

    Either that, or the Survation one of turn-out has still over-estimated the willingness of the younger generation to get out and vote and as such will see the Labour vote collapse..

    Will there be any indications of how likely age groups are to vote in the run up to the 8th that may change the polls results, rather than relying on 2015’s turn-outs?

    Out of interest, how did 2015’s turn outs by age compare to previous years and is this the first year that there are any sort of “goodies” that would tempt younger ones to vote?

  27. The Survation poll shows a very high 10/10 certainty to vote figure – 81% (it was 76% a wk ago).

    Does anyone know why its so high when others are in mid 60%s..?

  28. I’m sorry, I meant ICM and not Survation in my last post

  29. Are there any more polls out tonight?

  30. RP

    The truth could equally well lie at one extreme or the other.

    Agree with you – its just that everyone seems to be flying a bit blind atm, and on balance any purely arbitrary assumption seems as good as another.

  31. @Crossbat

    That should certainly help them. From the members I know on the conservative side of things their local associations are dying out. One is on a mailing list that 15 years ago had 40-50 people on it. Now its down to 6. Thats what comes with depending so heavily on the pensioner vote and alienating the young. You lose your door knockers.

    Labour on the other hand have 500K+ people to do the rounds.

  32. “Are there any more polls out tonight?”

    I’m guessing no, given that AW has not mentioned it. But surely we’re due a whole load at some point soon, and would be nice to see some Wales and Scotland ones again.

  33. All very fascinating. Shouldn’t we be accepting Survation as the most accurate as it got the result spot on and picked up the correct trends the day before Election Day?

    The unpublished poll is in the link below. What is interesting is that the turnout figures of the young are slightly out but within 5 or 10% of the actual turnout of the 18 to 34 year olds.

    What is also very worrying for the Tories is that the gap between Lab and Con in England in the last Survation published last night is only 5% after weighting applied. This would imply a swing to Labour of 2% meaning they would pick up enough to be in hung parliament territory.

    As someone involved in Corbyn’s campaign it is foolhardy to say the man and his policies are making no difference on enthusing young voters. The amount of my friends in that age bracket who I implored to support Miliband either weren’t registered or voted Green or elsewhere. Even working class kids employed in the local Subway are voting and have registered to do so who wouldn’t otherwise. Same down the local pub and more than one pub at that.

    If the turnout of the 18 to 34’s hits 65% Theresa May could well be toast. There may not be enough time left to turn public opinion around the way we would like but there is some trend emerging.

  34. Tory vote just about steady now.
    Labour vote gaining some ground on Tories, but at the expense of all others, not Tories.
    Polls all saying battle is 18-25 (Lab) V 55-80+ (Con).

    Older generation swung Brexit, to the detriment of younger generation.
    Younger generation now hoping to have their say – but will they actually bother to vote?

  35. I agree it does feel like no poll currently gives that much confidence, and it probably not until the exit poll anybody can relax or be sure. I bet the polling companies are nervous.

  36. In previous campaigns as the GE got closer the polling companies results got closer but that was largely because of DK reallocation back by some and not others and/or at different ratios.

    Different LTV filters will mean that these varaiances will persist right up to polling day.

    There will be random sampling errors so which pollster gets closest will have some luck involved but I assume the experts will crunch the raw data to compare methodologies as best as they can.

    Problem is that by next time different factors skewing the raw date will apply yet the adjustments might be based on the skewers for this GE.

    Not easy being a pollster.

  37. 7.6 million postal votes have been cast in the last general election out of 30.7 million ballots overall, roughly 25 percent.

    How many postal votes are already being cast this time?

  38. Interesting table here:

    ICM’s turnout weighting is turning 3% lead into a 12% lead!

  39. Telephone polls tend to produce Certainty to vote figures much higher than the Online polls.

    I would believe the figures in the 60’s. We are not going to have an 80% turnout. Last 3 have been

    2005 – 61.4% 27,148,510
    2010 – 65.1% 29,687,604
    2015 – 66.4% 30,691,680

    2015 was the first over 30million turnout since 1997 which was 31 million. I would expect a Turnout of 30million to 31 Million so any where between 66-70% Turnout.

    Reckon it will be slightly up this time. For reference –

    EU Ref – 72.2% 33,577,342

  40. I like to check this site because of the convenience of having all the various polls that are published collated together, with changes, etc. reported. I do think the layout could be improved, however. E.g. the table on the top right could be updated more frequently, and an up-to-date poll of polls graph could be shown on the main page.

  41. ICM’s “shy tory” adjustment doesn’t look good to me.

    In the data in the ICM polls:
    Re-weighting makes little difference to the figures.
    “Turnout” adjustment makes a huge difference.

    Raw: 675 con v 613 Lab
    Population Reweighted: 653 con v 596 Lab
    Turnout adjusted: 672 con v 500 Lab

    1st May
    Raw: 692 con 466 Lab
    Population Reweighted: 682 con 430 Lab
    Turnout adjusted: 681 con v 366 Lab

    In the raw data Conservatives have held steady, Labour have picked up loads mostly from UKIP and don’t know.

    The big problem for me is the “Shy Tory” effect doesn’t make any sense in this adjustment when you look at the 2015 recalled vote.
    The recalled 2015 vote requires only a small population adjustment is small to give the correct ratios for 2015 vote.

    However, if it is supposed to be publicly embarrassing to admit you are going to vote conservative hence “shy Tories” then it should also be embarrassing to admit to voting Tory in 2015?? Applying ICM adjustment to the recalled 2015 vote gives the conservatives a much bigger advantage than they actually got in 2015.

    The assumption ICM are making is they will need to adjust their data by the same amount in 2017 that they would have needed for their opinion polls in 2015 to give the correct result.

    They are ignoring a useful piece of information – how “shy” are their 2017 sample being about how they actually voted in 2015. Seems like they are less shy than what is being assumed.

    In my opinion ICM will overstate the Conservative vote share.

  42. In the 1979, 83, 87 elections the turnouts were 76, 73, 75%. Perhaps we’ll head up towards that precisely because politics has become more polarised around left and right – more of a genuine choice, harder perhaps to argue that it doesn’t matter?

  43. @AARON

    There has not been a truly left wing labour leader and campaign in decades. That’s one of the things that make this election so difficult to predict. There is no historical precedent to go by.

    Moreover that is further muddied by the effect of the brexit vote and the extent to which the vote against the wishes of the young has now activated them to vote. I full well remember the anger exhibited about it and how they viewed and presumably still view brexit as a betrayal by their elders. If I was in their shoes I’d want revenge.

  44. Anecdotal – I’ve just got back from a walk to the village shop, and as I’ve been reading this site for a few weeks now, it made me realise something a little odd.

    Our village is one of two halves – the older erstwhile mining/ council house part, and the newer estates. Usually, we see lots of Labour posters in windows and on the lamp posts, especially in the older section of the village, even though this has always been a strongly Tory constituency.

    Within a few days of the election being called, I saw a couple of large banners supporting Labour appear outside two houses.

    We’ve received leaflets from both Labour and Conservative candidates,

    Tonight, however, I realised I had seen no Labour posters at all as we walked around, other than those two, depite our route taking us into the older part of the village.

    So I don’t know if it is voter fatigue at the third election since 2015; something to do with the fact that this constituency voted 65% to Leave at the referendum; or if usual Labour voters aren’t sure about voting that way this time, perhaps because of Mr Corbyn or perhaps some other reason. But it is certainly very different from any previous election which has gone on during the 21 years we have lived here.

  45. I’ve noticed a lot of people wondering why there aren’t as many polls this time (compared to ’15) and the simple reason for this is that it was a snap election this time.

    Last time all the media and polling organisations had five years to plan for the election and the polling organisations will have worked with them to provide a full package for the weeks and months in the run up. This time they haven’t had that opportunity hence much less polling taking place

  46. I am surprised by the extremity of the 18-24 pro-Labour vote. I appreciate the young tend to be more socialist, but 60-70% is a little extreme, even for university students.

    At 18, you are barely past the age where much children’s politics follow their parents, as their main influence to date.

    I looked up this article from the Independence debate: Support for independence in young people was above 50%, but that for 16-18 year olds was not.

    I know that about 20% will be ethnic minority – a strong predictor for Labour – but also a lot will come from working class backgrounds of the sort that have voted heavily Brexit. They will hardly of grown up so uniformly lacking in the relevant concerns, particularly those who enter the job market to compete for entry level jobs with experienced Eastern Europeans.

    In short, whilst I do not doubt that the young are more socialist, the present polls seem to suggest an extremity of difference that could only be explained by schools being socialist education camps.

  47. RP

    I see the EU ref 33.5 Million as the absolute max turnout we could possibly see. Even that is unlikely. Percentage depends on the size of the Register of course. Hence using the Raw figure is very useful.

  48. I wonder if people could just stop making jokes etc about Diane Abbot’s appearance? no problem about her views and pronouncements. But I was brought up to think these sort of jokes very bad form,especially about women, and it leaves a nasty taste.
    This is not partisan. i don’t like jokes about how Tory women look either.

  49. Oh dear.
    It would seem that the conclusion of this election is going to be that some polling companies got it systematically wrong.

    I have no idea who those companies are, and presumably they themselves think they are right and the others wrong. Since they are professionals, the industry must be divided.

    I’m not a professional, and generally follow the yougov polls, because of the sponsor of this site, and because arguably they all seem to agree on trend, just not on absolute values. So my further comments should not be taken as uniquely picking on yougov, thats just the ones I have followed.

    I have posted before that the ‘will not vote’ percentage seems to be too low considering typical turnouts at elections. This might be a completely unfair criticism if we get a very high turnout this time. But if we don’t, then I think there are not nearly enough undecideds being asked their views.

    I recognise they do not affect the question the pollsters are trying to answer, if that is how people feel now. But it is a problem if they are trying to forecast the real result. They have time yet to decide.

    There are indications of how some of these people may vote from their past voting, both on party split and Brexit view, assuming they follow the established pattern of labour=remain and Conservative =leave.

    Possibly a number are those where these two main deciders are opposed. Where con/leave coincide they are presumably happy to have declared for conservative, but if they are con/remain or indeed lab/leave, then they would perhaps be undecided. Moreover, if this is their difficulty, then they are likely to be a group which DOES vote, rather than the habitual abstainers.

    Are enough of these people really being included in samples for them to be representative? I am arguing it is predictable they will vote, and they are sufficiently numerous to decide the outcome.

    I would think most people do not think the result will stay static from now to polling. But there might be evidence that people who have already made up their minds will remain fixed. Their polling answers seem pretty convinced this is so. But that again transfer attention to the groups who dont know/will not vote, who might in fact decide and vote.

    So what will they do? Will conservative remainers prefer to stick with a party which has just produced a rather atypical conservative manifesto? Or will they think that labour has some generally well received policies, so on that score they have no objection to lend their vote?

    I suggest that maybe the unusual manifestos had more impact than usual, precisely because they have impacted on the larger than normal group of politically active undecideds. The conservatives perhaps thought they had a big lead secured, but in fact they still had to convince the undecided, and they convinced them to go labour. It is unlikely this process is over, as people wrestling with their political consciences continue to do so.

    The manifesto is an example, this group is still deciding.What do labour/leave people choose? Are labour really remain – no. More remain inclined perhaps, but is their offer ultimately simply a more cautious leave? That leaves wide open the possibility of leave supporters with reservations about the process voting labour.

    Mate of mine suggested this could end up as a 1945 landslide, against the government which apparently had just been the victor.

  50. Danny

    The polls always put Labour Ahead in 1945, its just no one took any notice of them as they were a new concept.

1 2 3 23