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 17 18 19 20 21 23

    Wow yes ~35% turnout for the younger generations is certainly too low this time around.

    The Brexit vote activated the young towards voting and the evidence we have is that once you start voting it becomes a habit. So I’d certainly expect an uplift over 35% particularly in light of the vastly better ideological and personality fit between the young and Corbyn vs the young and Ed “stone” Milliband.

  2. The FT came out in support of the Conservatives, but very scathing about both major parties.

  3. im not trusting any of the polls tbh. you gov in particular seem to be going out on a limb.

    All we can reasonably say is that the gap between labour and tory has narrowed somewhere between “somewhat” and “a lot”. we cant even tell if this is still happening or if it has stabilised.

    Maybe it would be more telling to look at which seats the parties are targeting with resources – based on their own internal polling and local knowledge.

  4. For me, the most interesting thing about the YouGov modelling is this:

    If you look at seats in the Midlands and South, and compare to the 2015 GE, we see the CON vote largely stable.

    We also see the Green vote collapsing and the Lib Dem vote increasing, and we are not seeing the big switch from UKIP to Con which was assumed.

    I am guessing that YG have solid data to back this up. But it does go against peoples expectations. This is particularly true in areas with a significant youth vote.

    Take Canterbury as an example:

    CON: Stable
    LAB: +15
    UKIP: -13
    LD: +5
    GREEN: -5

    There must be, underlying the YG data, something very interesting at play which is not confirming to the assumption that UKIP collapse = CON gain.

  5. @RICH

    Living in the area is hardly evidence… I wasn’t saying that it will swing, just saying that there is no way you can know!

    Tendency/tendancy… I’m not the best speller, unlike you Rich who can spell weird no problem. At least that is something we can agree on. You are a proper troll in how you bait – how are you allowed to comment at all?

  6. The Mori Scotland opinion poll out this morning has some interesting findings besides SLAB and SCON level on 25%.

    The most dramatic has been Teresa May`s loss in approval rating. She had + 16% approval in September, now its -27% (33% satisfied, 59% dissatisfied). Ruth Davidson has also declined, from + 22% to – 3%, whereas Kezia Dugdale has slightly improved.

    6 SNP seat losses to SCONs are predicted on this polling.

    But I am doubtful, and now think SLAB will have gains e.g. East Lothian.

    Whether enough Scots watched Monday`s C4 debate to make a difference, I don`t know. But TM would have lost many potential Scottish Tory voters by her “no deal is better than a bad deal” insistence.

    And the Mori poll was done May 22-27, so before that debate.

    I feel SCON`s chances are evaporating both from the local press comments, and from the party tactics as shown by their leaflets and posters.

  7. I think the YouGov poll is more likely to be closer to the actual results than the others (it seems to mirror the general direction of the campaign to me). There is a movement towards Labour now and it’s going to be difficult for the Conservatives to slow it down.

    My guess is 20 to 30 seats lost for the Conservatives with mostly Labour gains, but also a few LD wins.

    What can the Conservatives do to change things round? I think they need to focus on the shape and composition of the likely government, but it might be too late for that: it’s not 2015 anymore.

    If there is a coalition of every non-Conservative GB party then the election debates next time will be rather odd: everyone on one side – government – versus the Conservatives on the other. It will be de-facto two-party politics; maybe we’re there already.

  8. @ALAN

    The MRP used by YouGov is a probability based estimate, the statistical measure they use is a 95% prediction interval i.e they have a 95% chance of getting their numbers correct between 39 and 44% for the Conservatives and 35 to 41% for Labour.

    It is now those figures I am using and comparing them to the MoE based on the VI surveys.

  9. PS I would be surprised if it swung too, but I don’t like your certainty.

  10. “I am just saying if voters think about it logically. Whether they do or not is another matter.”

    No what you are saying is that if voters think about it the same as you do.

  11. Has any polling company – or anyone else – produced polls giving a range of predictions for different turnout levels?

    I ask because so many people here are saying, “it all depends on turnout.”

    Obviously, national turnout could include significantly different regional and constituency turnouts, not to mention by age or other demographic splits.

  12. As I have been saying for a while, Brexit may not be such an advantage for the Conservatives as things stand at present. That said, Labour cannot really commit itself to anything about Brexit since areas of its geographical support have opposing views on Brexit.

    It is when Mrs May repeatedly claims that “no deal is better than a bad deal” that Labour may score. If Mrs May does not know the implication of her words, others do. Here is what Richard North says.

    “Anyone who doubts this merely has to look at Article 3 of the EU-EU Air Transport Agreement, where each party grants to the other the right to fly across its territory without landing; to make stops in its territory for non-traffic all points in the United States and to perform international air transportation between the United States “any point or points in any member of the European Common Aviation Area (ECAA)”.

    Without an agreed carry-over, on Brexit Day, the UK ceases to be part of the ECAA, and the Agreement no longer applies to UK-registered airlines. Nor indeed will US registered airlines be able to fly their aircraft to the UK. The chaos we’ve been witnessing after the British Airways IT failure will be looked upon with fondness, as a mere dress rehearsal.

    Interestingly it isn’t only the Financial Times which is waking up to this potential disaster. We also see the New Statesman point out that, if Britain leaves the EU without a deal, its right to participate in Open Skies will also end. It is hard, says Stephen Bush in the magazine, to see how for anyone in Britain who likes flying to Europe or America … ‘no deal is better than a bad deal'”.

    It looks to me that Labour may not be up to speed on what are the implications of “no deal is better than a bad deal”. If it were it could be extremely difficult for the Conservatives. ( I am not a Labour or Conservative supporter).

  13. What I don’t really understand is how Lord Ashcroft appears to be using a similar MRP model and although he is not giving a Voting Share % nationally he does at local level (updated to last Friday), and yet the figures compared to YouGov on a similar model are miles out.

    My constituency of Gedling has the following:

    YouGov: Con high 43 low 32 Lab high 58 low 46

    Ashcroft: Con est 45 Lab est 39

    Go figure!!!

  14. @Jamie,
    Simple answer, don’t claim to be an expert on people you know nothing about, I wouldn’t dream of saying that about you. That’s real trolling I am afraid.


    Noone is 100% certain to vote, creating (and testing) a model for how likely they are to vote based on data (rather that trying to get there from gut opinion) seems quite possible.

    I suspect their stated likelihood to vote will be a significant variable
    I suspect age will also be a significant variable

  16. Rob

    There are plenty on here who are partisan. One poster referred to Corbyn supporters as “loony” yesterday which is far more partisan than Dr M defending the YouGov poll.

  17. @DrMibbles

    Yes, as far as I can tell all the pollsters are getting pretty much the same data. The only difference between them is how they choose to build their model to fit those data.

    What evidence do we have about likely turnout? Very little. Perhaps there are many who failed to vote in 2015 and 2016 who did not do so because they thought they didn’t need to (misled by the polls), and who won’t make that mistake again. Or maybe the lower turnout in the young will continue. It is quite plausible that the likelihood to vote adjustments for some pollsters are completely wrong.

    If I were a pollster, I would right now be seeking whatever data it is possible to gather to be better able to model turnout in the future. It’s obvious that simply asking people their likelihood to vote is not good enough.

    Incidentally, it would be really useful in YouGov integrated different turnout models into their large seat prediction algorithm, to give a flavour of the resulting projections. It might also discourage the all too frequent, incoherent, unevidenced gainsaying by the Tory shills on this site.


    “No what you are saying is that if voters think about it the same as you do.”

    -Labour increased immigration significantly.
    -Migrants need housing
    -With increased demand and constantly supply, the price of anything increases

    Which of the above is incorrect?

  19. JonM . Totally agree. Ashcroft was closer to the truth than YouGov in 2015. Too many polls are showing much bigger leads for this YG to appear accurate in my opinion

  20. Increased immigration will lead to rent and house price increases. No amount of realistic house building can suppress this at current levels. Also, the smaller rural, suburban areas with good schools, communication links etc will just continue to drive up. And remember, when any govt gets in trouble, be it Conservatives or Labour, the bog standard response now is put a hot poker under the housing market and get people to feel better off and spend equity. I don’t see this changing.

  21. JonM and Alan

    The prediction interval is indeed 95%, so JonM is correct.

    The 95% interval gives very large range for seats.

  22. I get likelihood to vote for 16-24 year old on comres as 58% not 35% as quoted above.

    I’m confused where the 35% comes from. help!

  23. @Mike Pearce

    My point is that both Ashcroft and YouGov have gone for this MRP probability estimate type model Ashcroft has 40000 in his pool and YouGov a 49000 rolling pool (pool not poll!) and yet they are producing wildly contrasting results.

    I have absolutely no idea which of them is correct, but at least one of them isn’t, even within their statistical probability.

  24. Rich

    Try Wolverhampton south west and Birmingham edgbaston for starters. Not a chance of labour holding them

  25. If only people could see things how I think they ought to see them, they would vote the way I think they ought to vote. The poll results I don’t like are plainly because people are misleading them and they are too stupid to know better. I am right and everyone else is wrong. The only possible model for the economy is the one I adhere to, and anyone who says otherwise (even if only pointing out that other viewpoints are possible) is posting offensive partisan nonsense.

    I on the other hand only ever post balanced thoughtful arguments.

    Who am I?

  26. Robin

    While I’m sure that polling companies experiment with different turnout patterns, they can’t use these without being accused of covering their back – they have to say what they think correct, or the right approximation.

    By the way, the YouGov prediction model uses the same as their polling: demographics+BES 2010 and 2015.

  27. Owning a home is out of reach for increasing numbers of people because wages have increasingly lagged behind housing costs.

    a significant investment in municipal housing would lower rents overall – and take some of the heat out the housing market.

    And there is no hard evidence that immigration is not driving up house prices –

    But anyhoo – this is not a forum for debating reasons for the housing shortage but i wasn’t the one posting up the UKIPesque “COMMON SENSE” logic.

    See also – immigration causes traffic jams on the M4 (farage).

  28. Rob,

    “-Labour increased immigration significantly.”
    All types of migration in and out increased after the single market and both Labour and tory were fairly relaxed about it as the economy as rapidly expanding.

    “-Migrants need housing.”
    Everyone needs housing and smaller household sizes and longevity were bigger causes than immigration and both Labour and tory had plenty of resources and time to increase supply.

    “-With increased demand and constantly supply, the price of anything increases”
    Both Parties are happy to watch house prices rise and leave supply to the market and the public weren’t complaining either.

    It’s easy to play the “Fix the Roof!” card and point it at Labour but since Thatcher every Government has effectively followed the same free market Light touch private sector orientated approach.

    the fact that the forty year boom ended on Labours watch makes it easy to blame them for all our current problems but the facts don’t bare it out.

    Between the 75 and 05 UK debt ran at roughly 35-45% depending on the state of the general economy and the economic cycle with little difference over the medium to long term between Labour and tory governments.

    After the crash it went to over 100% we gave the banks £1t, started printing money, slashed interest rates and the cost of servicing our debt soared and the tax receipts from the financial sector dropped.

    The light touch pro city policies Labour followed caused that but the tories were in favour of all of them.

    It was the uniformity of approach between the two parties in government that got us here no Labour Bad/Tory Good!


  29. Rich,
    “No amount of realistic house building can suppress this [rent prices] at current levels”. Isn’t that the whole argument for radically increasing the rate of house building? This seems a pretty strong area for Corbyn.

  30. @JonM

    The sort of model that YouGov and Ashcroft are using is highly sensitive to the inputs, not just the data but the parameters of the model. I don’t know how Ashcroft has trained his model, but YouGov report having validated theirs.

  31. @ TERRYP

    That sounds much more plausible, considering that the British Election Survey put turnout among 18-24 year olds at 58% in 2015.

  32. Birmingham Edgebaston is likely to stay LAB without a larger swing to CON in England than most polls report.

    Not least because LAB will pick up a few extra points from the collapse of the Green vote, and there isn’t too much UKIP vote there. Final result likely to be something like 47 LAB 43 CON on current polling.

  33. @Robin,

    I get a hard time from, strangely, all the left biased posters.

    A few of your highlights from previous posts I checked. So these are all non partisan are they?

    ‘YouGov dataset simply has to be better’

    ‘May fears Labour victory’

    ‘With Corbyn, many more will watch, giving the opportunity to change some more minds’

    ‘Labour have got their media right’

    etc etc

    And the list goes on. Left wing sly partisan comments are simply not moderated very much here. And, if I stray partisan, I apologise, but I am jumped on by you, drmibbles, smithy and all the other clear Labour voters.

    Just for the record, why does YouGov, ‘simply have to be better’? Are you honestly saying it isn’t any confirmation bias?

  34. @Steven,

    Yes perhaps it is, but practically I don’t know how it’s achieved.

  35. Why is it that so many people here seem to be so exercised about the opinions of others? Even if they could be regarded as 100% flawed what on earth does it matter?

    The site is just descending into playground squabbling and it feels really unnecessary. I’d like to know more about different technical considerations with regard to polling, not read posts offering abuse because someone thinks Labour/Tories may or may not win.

    In the absence of that information [or more possibly, understanding on my part] my best guess is a narrower Tory win than first envisaged. But I won’t be hurling abuse at someone for thinking it will be much larger, or non-existent – or whatever.


    -Between 1997 and 2004 EU migration was relatively low, because the only countries in the EU were comparatively wealthy like us. Then in 2003 Labour signed the Treaty of Accession which allowed 8 (and 10 by 2007) of the poorest countries in Europe to join a free movement bloc. To make matters worse, Labour didn’t implement the transitional controls.

    In 1997 non-EU migration numbers rocketed.

    “Everyone needs housing”

    But the British birth rate has been constant or even declining for a while, therefore the native British population don’t require new houses to be built. When you invite people to your country, you’re requiring demand for more houses.

  37. @Rich

    Taking extracts from longer posts which were discussing *whether* various things were the case just makes you look stupid.

    Are you seriously suggesting that a dataset of 50,000 responses isn’t better than a dataset of 2,000?

    “I get a hard time from, strangely, all the left biased posters.”

    That’s because you’re an annoying right-wing shill without whom this site would be a whole lot better.

    @AW Happy for this post to be removed once Rich has seen it.

  38. Waiting for polls from Friday to know where we are.
    The Theresatives attack on Corbyn and Abbott over 35 year old support for IRA was misguided when they should have been going on the narrative that Corbyn will bankrupt the country, have the visa card credit limit be closed, and have literally no money to pay pensions, nurses, doctors, police, teachers, social security, housing benefit. That there would be no petrol, no heating and no food in the shops, Indeed like Venezuela no toilet paper.
    This may all be untrue (or impossible to verify) but it paints a picture with the type of voters who like the cut of his policy jib. Pensions, police numbers, NHS, school class sizes, etc
    This would be proper Project Fear 3.0 not stuff which to many voters ceased to be a direct physical threat to them 20 years ago.


    The crux of the matter is Labour intentionally increased immigration. We all know a senior Labour advisor admitted recently it was done intentionally to affect the ethnic makeup of the country.

    You cannot invite 3-4 million migrants, the majority who are low-skilled and then wonder why you have a housing shortage, high rent, house prices and similarly education and NHS crises. Whats the common denominator? More people!

  40. why are there people debating about migration? Talk about polls people for god sake!

  41. One thing i can guarrantee about the yougov poll. If it starts to move back to the tories those posters who today praise it will suddenly find fault and those attacking will suddenly see the merits.

    There is no truth merely evidence of truth.

  42. @ROB

    It could be just as equally argued that the issue is that the conservatives purposefully limit the building of new housing because their home owning elderly voters don’t want new housing built near them. After all, if we were capable of building 300-400K housing per year in the 70′,s given vastly improved productivity levels there is absolutely no reason we cannot reach at least this level now.

    But either way such arguments are utterly irrelevant. The blame for this issue can be laid at the feet of one group or another and one party or another, but that matters not in terms of political effects. We 100% know that renters vote labour and not conservative and that has long term consequences for the future of the conservative party.

  43. I don’t think we’re going to get an resolution to the “who is right” argument till next Friday. Dark days for polling when variance is this far out this close to an election.

    And please! Whoever it was that said because Yougov had a bigger sample size it’s better get off this site!! Go and hang on twitter! It doesn’t matter how big a sample is its accuracy that counts.

    Let’s talk polls not politics!!! Please! Please!!!

  44. Edgbaston is a funny seat/area, I know it very well.

    Very close to the city centre, yet not the normal slightly run down areas that often categorise edge of city centres. Some elements are very posh, am guessing Harborne is also in this, again has some very expensive housing and clear Conservative demographics.

    What might work against Cons is the very high student population in Edgbaston too.

    This will be a very interesting seat to watch.

  45. @Rich
    That’s real trolling I am afraid…
    Accusing other posters of trolling is a bit erm, rich?


    “Owning a home is out of reach for increasing numbers of people because wages have increasingly lagged behind housing costs.”

    What happens to wages when you increase the supply of labor?

    “a significant investment in municipal housing would lower rents overall – and take some of the heat out the housing market.”

    But its not just housing, you have to build more schools, more hospitals, more GP surgeries, more dental surgeries, more roads…… all because we admitted millions of low-skilled migrants. What exactly was the benefit? Its certainly not visible from observing wages.

    “And there is no hard evidence that immigration is not driving up house prices”

    I assume you didn’t mean the double negative. Any person who understand economics will tell you 3-4 million migrants will increase house prices and rent. Its GCSE Economics.


    Might be of interest. Not sure if posted previously by anyone else.

  48. @robin,

    Pot, kettle. why don’t you stop masquerading as an informed poster.

    As you know, I just read your last 10 posts and every one was why Yougov is right, Corbyn is right and May is awful….boring..

  49. Laszlo

    I’d like to know how they calculated that as +/- 3% can’t be due to sampling alone on a sample of that size.

    If it’s the effect on the results due to the range of biases expected to be seen in the model (in this case the errors in the regression coefficients themselves) then that might be plausible, in which case we’re measuring different things.

    Of course if the data is not representative of the voting population then the model won’t be centred on the true value.

  50. @ TERRYP

    The key phrase in ComRes tables document is “Voting intention figures are calculated using the ComRes Voter Turnout Model”

    It doesn’t matter what people tell them, they apply their own turnout assumptions based on demographics.

1 17 18 19 20 21 23