I’ve only had a couple of hours sleep so this is a very short comment on lessons from the polls at the election. The two best performing traditional polls seem to be those from Survation and Surveymonkey. Survation had a one point Con lead in their final GB poll, Surveymonkey had a four point lead in their final UK poll. The actual lead is 2 or 3 points depending on if you look at UK or GB figures. Congratulations to both of them. While it wasn’t a traditional poll, YouGov’s MRP model also came very close – it’s final GB figures were a four point lead (and some of the individual seat estimates that looked frankly outlandish, like Canterbury leaning Labour and Kensington being a tossup, actually turned out to be correct).

Looking across the board the other companies all overstated the Tory lead to one degree or another. The actual share of the Tory vote was broadly accurate, rather it was that almost everyone understated Labour support. I have a lot of sympathy with Peter Kellner’s article in the Standard earlier – that to some degree it was a case of pollsters “trying too hard”. Companies have all been trying to correct the problems of 2015, and in many cases those changes seem to have gone too far.

A big gulf between pollsters that many commented on during the campaign was the attitude to turnout. The pollsters who were furthest out on the lead, ComRes, ICM and BMG, all used methods that pumped up the Tory lead through demographic based turnout models, rather than basing turnout on how likely respondents said they are to vote. This was in many ways a way of addressing an issue in 2015 polling samples that contained too many of the sort of young people who vote, weighting down young turnout (and turnout among working class respondents, renters, or less well educated – different pollsters used different criteria). This wasn’t necessarily the wrong solution, but it was a risky one – it depends on modelling turnout correctly. What if turnout among young people actually did rise, then pollsters who were replicating 2015 patterns of turnout might miss it. That may be what happened.

That said, one shouldn’t jump to conclusions too quickly. It may be a case of how demographic turnout models were applied (by weighting the whole sample to match 2015 recalled vote and then separately weighting different demographic groups up or down based on likelihood to vote there’s a risk of “double-counting”). Most importantly, the YouGov MRP model and the Surveymonkey survey both based their turnout models on demographics too, and they both got the election right, so clearly it’s an approach that has the potential to work if done correctly.

Personally I’m pleased the YouGov model worked, disappointed the more traditional YouGov poll had too big a lead… but that at least gives us something to learn from (and for most of the campaign the two showed a similar lead, so rolling back some decisions and learning from the model seems a good starting point).

And with that, I’m going to get some sleep.


2,448 Responses to “Post-election thoughts”

1 2 3 49
  1. First?

  2. I think the big winner does look to be the Yougov MRP model and I’d be interested to see analysis on how the seat-by-seat predictions matched up with the final result

  3. It seems that the problem facing pollsters is turnout. No amount of weighting and modifying your samples can account for this when the vast majority of people that take part in these polls are very politically engaged.

    Seems there needs to be a fundamental re-thinking of how these polling samples are sourced…

  4. Downing St confirm that the senior Cabinet members will remain in post. No surprises.

  5. And, if I might add my acknowledgement of our very own DrM who came out the GE in a league of his/her own on here.
    a few others agreed with the low Tory lead and seat counts but did not drill down in the way DrM did.

  6. ANTHONY WELLS

    Thank you for your commentary and information since 2015. It, and this forum and its regulars, are all greatly appreciated.

    Thank you all.

    P.S. At my work sweepstakes I bet on Tory minority with 320 seats, if only to play the contrarian (everyone else had ConMaj from 20 to 172).

  7. What are people’s thoughts on whether May will step down, and when? She’ll never lead Cons into another election, that much is clear; every day she stays on is a stain on the Conservatives’ reputation.

    Will there be any polling on this?

  8. Nice summary AW, thanks for sharing your thoughts, and enjoy your sleep!

  9. The reason the polls could not call it more accurately at the end is because without an official transparent launch a new party tried to hijack the Conservative brand and as the voters discovered this, so it added to the volatility of VI.
    The Theresatives seemed at first merely the Conservatives committed to Brexit seeking a vote of confidence to finish once and for all the diehard Remainers before negotiating as a united nation against the united EU opposition.
    However, it emerged that the policies, principles, norms of the Theresatives were not the Conservative party from 1983 to 2015 plus Brexit. They certainly were not the 2015 manifesto plus Brexit.
    Above all the Conservative voters of 2015 and the Leave voters became confused, unsettled, anxious about what the Theresatives would do (on Brexit) and much else if they were given a huge majority and blank cheque.
    Thus, many would be Brexit and Conservative voters once convinced the Theresatives would win big, then looked for ways to mitigate the victory. They were looking for a 3-2 win rather than a 7-0 win.
    The hapless front of stage performances by the leader made matters worse and lack of playing the full team puzzled many.
    The manifesto was uknown to Tory MPs, candidates, activists, Cabinet Ministers and a charge of the Titanic towards the iceberg.
    The other team of Corbynistas were attacked on associations 35 years ago and not forensically examined on the possibility that their promises of giveways would lead to no money at all to pay police, nurses, doctors, etc. To have no petrol, no food in the shops, no toilet paper. Spending is great if you can afford it and your creditors do not lose confidence and call in a £2 trillion national overdraft running at huge interest.
    Of course people wanting free gifts of £35k and big pay rises and even a house would turn to the party making the offer unless said party were challenged on ability to deliver.
    The late realisation that Theresa May had cut police numbers by 20% and had plans to continue (and defend) the trend did not play against the terrorist attacks.
    Above all the Government did not seem to have an effective plan to deal with terrorism nor the daring to brazenly appeal for votes as an act of defiance to the terrorists.
    The appeal of Theresa May in April and early May was a mile wide and inch deep. The campaign got through the inch.
    Pollsters had no real chance of calling this one.
    The confused electorate did not get a clear feel for who the parties were and the shapes shifted every few hours.

  10. Worth having a look at the opinion polls in April:

    Opinium/Observer 2017-04-28 47 30 8 7 3 17
    ICM/Sun on Sunday 2017-04-28 47 28 9 8 4 19
    YouGov/Sunday Times 2017-04-28 44 31 11 6 2 13
    ORB/Sunday Telegraph 2017-04-27 42 31 10 8 4 11
    YouGov/Times 2017-04-26 45 29 10 7 3 16
    Ipsos-MORI/Evening Standard 2017-04-25 49 26 13 4 1 23
    Panelbase/ 2017-04-24 49 27 10 5 3 22
    ICM/Guardian 2017-04-24 48 27 10 7 3 21
    Kantar/ 2017-04-24 46 24 11 8 4 22
    Survation/Mail on Sunday 2017-04-22 40 29 11 11 2 11
    ICM/ITV 2017-04-21 48 26 10 8 3 22
    YouGov/Sunday Times 2017-04-21 48 25 12 5 3 23
    ComRes/Sunday Mirror 2017-04-20 50 25 11 7 3 25
    Opinium/Observer 2017-04-20 45 26 11 9 3 19
    YouGov/Times 2017-04-19 48 24 12 7 2 24
    ICM/Guardian 2017-04-18 46 25 11 8 4 21
    ICM/Guardian 2017-04-17 44 26 12 11 4 18
    Opinium/Observer 2017-04-13 38 29 7 14 5 9
    ComRes/Independent/S. Mirror 2017-04-13 46 25 11 9 4 21
    YouGov/Times 2017-04-13 44 23 12 10 4 21
    YouGov/Times 2017-04-06 42 25 11 11 3 17
    ICM/Guardian 2017-04-02 43 25 11 11 4 18

  11. Dear everyone else, I promise to go through all the roughly 200 individual predictions that were made when I get the time (probably tomorrow), and will look for the closest on the various factors that were brought up.

    For now, I’m going to post the entire final list at the end of the ‘Final eve-of-election polls’ thread, so they’ll remain easy for anyone to find and look through:

    http://ukpollingreport.co.uk/blog/archives/9909/comment-page-19#comments

    Cheers!

  12. Good Evening from Bournemouth East; sunny here. Labour is now in second place in this seat in the neighbouring Bournemouth West.

    Thank you to AW for fantastic work and to all UKPR posters and lurkers for great debate and insights.

    I have a deep interest in the North of Ireland, and it will be interesting to see how the Westminster Govt maintains its impartial role in the operation of the strands of the Northern Irish Agreements.

  13. This was a pretty decent guess for Lab-Con Richard!

    (124) RICHARD (#3 – 7/6)

    Con: 42% (312)
    Lab: 39% (263)
    LD: 7% (5)
    UKIP: 4% (0)
    Grn: 2% (1)
    SNP: 4% (47)
    PC: 0.5% (2)
    Oth (20)

  14. @Anthony,

    Congrats to YouGov on the new model, it was pretty spectacular in the end.

  15. As I see it May’s credibility within her party, let alone the country, is damaged beyond repair. Her ‘strong and stable’ position as PM is just a bad joke now. I cant see her remaining in post more than a few months at best. Probably gone by the party conference. Even with a DUP deal the majority will be wafer thin within a couple of years through by-election erosion and another election then unavoidable. So a GE within 2 years max. And there is no way the party will tolerate her leading them into another election given the debacle just played out. The need to give a new leader time to bed in in advance of the next election, not to mention the Brexit timescale, means the only sensible time to change the leader is basically right now. That would also give the new leader the political space to distance him/herself from the more disastrous elements of the so-called manifesto, and get in post prior to the really serious phase of Brexit negotiations which wont occur till after the German election this autumn.

    The Tory party has a long history of intolerance towards failing leaders. Its gets rid of them quickly. I see no reason why May will be treated otherwise. I expect murmurings are going on right now and a new leader chosen over the summer.

    Cant see it being BoJo. Davis would be my guess. Highly competent, more personable than May and we know he’s always coveted the job. And besides, he’s completely up to speed on Brexit.

  16. Everyone ready for round 2 in a few months time?

  17. Lowest estimate for SNP was Oliver on 38, after than a 40 and a 41, with the vast majority 45-50.

    Having said that he also had the Tories on 365, so the 3 out on the SNP (8%) is slightly undermined by the 47 put for the Tories (13%!)

    Peter.

  18. Yougov new model must be a HUGE problem for smaller pollsters. They just cant compete with the size of the op behind it.

    Overall this was a lesson in why herding is a terrible idea. Sometimes polls are wrong because there is a late swing to a party but perhaps there can be a late swing towards apathy. Weighting based on previous expectations or your own expectations just does not work.

    Seems Survation learnt this and clearly polls closure to the poll are more accurate. But for me it was pollsters listening to commentators and seeing a swing that didnt fit the narrative so they made adjustments.

    ———

    Cant wait for the next few weeks. May looks to be setting up a hospice government – too weak for any big changes just means shes waiting for someone to finish it off. Expect grumbles from Tories about how her negotiations are going…

    Lot of the press will have egg on their face for their May love in. Looks painfully out of touch. Clearly an anti May narraative will grow but it will be interesting if Corbyn starts getting more positive coverage.

  19. @AW

    Thanks for your reply on previous thread. Just to say, I was wondering if the shift towards Labour in votes per seat might have been down to more students voting and their being more tactical this time, voting where their vote will be more effective… unless there’s some hidden flaw in that idea I wouldn’t spot (not being a voter myself, you see…)

  20. Well, as I suggested in previous thread, protest marches are indeed to take place. Didn’t realise they’d start quite so soon!
    5pm today at Downing Street protesting against the ‘bigot coalition’. Will presumably be small scale as short notice but I expect the scale of these things will reach unprecedented heights over the coming weeks if May tries to cling on.

  21. @Baldbloke

    It is an open secret that Davis was crucial in convincing May to go to the country. He has never been popular with the more pragmatic Tory MPs who see him as an overconfident, self-important prig and he will be only a little more popular than Nick Timothy right now.

    In short, no, he will not be the next Tory leader unless the backbenches are sure it is a suicide mission.

  22. Watching the BBC and reading the Guardian comments section, can anyone tell me how big the Labour majority is? Obviously it’s a landslide but I’m not clear just how big.

  23. Lesson on predictions -make it as late as possible ,do plenty of doorknocking or telephone canvassing ,dont predict as you want it to ,dont trust all pollsters.

  24. Also of note is that the exit poll was nearly spot on.

  25. @Fraser

    “Yougov new model must be a HUGE problem for smaller pollsters. They just cant compete with the size of the op behind it.”

    Actually I would be interested to see how well it performed on a smaller sample. I’m sure the estimates for individual seats would worsen, but that doesn’t mean it’s overall prediction would necessarily be a lot worse.

  26. @RICH

    Why are you reading the BBC and Guardian? Wouldn’t you be more comfortable with a copy of the Telegraph down the Masonic Lodge?

    Labour didn’t get a majority, nobody did. However I went to sleep this morning very happy. Did you?

  27. Props to Survation for not herding and being right. There is not an unfixable problem with polling. Survation final poll in 2015 and this year shows it. There is a problem with pollstsrs trying to get figures to meet consensus and conventional? wisdom. I note that Kantar came close to final margin too but, like YouGov, anxiously tried to match consensus with commentary that it would mean an increased majority.

  28. @lewbrew,

    No I was gutted. But I do apologise for my over reaction to you on the night!

    Of course I was joking above, it’s all about context and expectation I guess.

    rich

  29. @ SSSIMON – Richard was also the person to send out the link on demographics by seat from which humble modellers/betters like myself could understand the YouGov model and hence have to seriously consider the consequences and the assumptions we had made in our own models (and read of the polling info).

    I was highly cynical of the polls (due to the huge variance between polling companies and retrospective “fitting” to the outcome they thought would happen). I think that was a fair conclusion for most polling companies (but not YouGov)
    I was also highly cynical of YouGov’s model. I was completely wrong there and eating humble pie accordingly. As you said, even at a seat by seat level they nailed it!

    So, huge thanks to YouGov model (and Richard) for showing where the Youth vote issue would have an impact (and DRMIBBLES of course for highlighting to us what was one of most extreme examples of its success – challenged only by some London seats and sadly just missing out on seats like Southampton Itchen!). As I discussed with CMJ at the time YouGov should have published that model’s final predictions – the 95% confidence levels and some caveats would have given them the reputational insurance that, under the circumstances of the very unique nature of this GE, was completely acceptable. I had scenario analysis that covered anything from a 300-420 outcome but I subjectively skewed that to the upside scenarios – I was wrong!

    The story of this election was for sure the Youth vote and YouGov nailed that. It made me review my model, put in a demographic add-on (thanks to Richard for the data) and chuck on some butt covering hedges (also thanks to the Irish betting companies for facilitating those hedges).

    I hope I helped others that wanted to take a financial punt on the results (I know many did not want that info) – either by advice on the best way to express their view and/or the best place to find the best odds… or by encouraging a flutter with our Irish friends! My little network of “shocked” Tories certainly offsetting their misery of the outcome with a bit more cash in their pockets!

    I think Aaron got the wooden spoon on the “predictions” but I think I was a close 2nd worst!

  30. @Brilliant Smith

    Always talk to people who are not like you. It is this fundamental error by Tory activists and commentators that has led to this huge mistake.

  31. In the unlikely event anyone wants to take a punt on the “next government” that market is still live:
    https://www.betfair.com/exchange/plus/politics/market/1.131095612

    10% return if you think CON can do a minority deal with DUP (seems a good bet to me but I am talking my book!)

    Also if you have a view on next GE date then:
    https://www.betfair.com/exchange/plus/politics/market/1.132100025
    or
    https://www.betfair.com/exchange/plus/politics/market/1.132099836

    Next PM after May, etc, etc…

    The markets never sleep!

  32. “The electorate got it wrong’, says Conservative MP CrispinBlunt. ”

    I can only apologise for the small part I played in this collective error.

  33. Colin

    I see m to remember you being the only person here to pick up on my comment – at a time when what people were mostly worrying about was whether the Tory majority would be closer to 100 than 200 – that it seemed likely to me that a hell of a lot of people would have looked at the Labour manifesto and thought:

    “Ooo, I’ll have some of that, ta very much.”

    I realise that voting rationale is slightly more complex that that but I find it incredible that, at that stage, the Tories didn’t feel it necessary to revise their own, before publication.

  34. HENRY

    RE: political engagement

    That’s true if you have an unrepresentative panel, but pollsters invest a lot of money building up panels and it’s easy to measure things like political engagement and account for it.

    It was, to my eyes at least, always obvious that there was a serious problem with the turnout modelling as the campaign progressed;

    — Self-reported turnout in young voters started at 2015 levels, around 43%
    — It rose significantly during the campaign, to around 66% (which matched feedback on actual youth turnout)
    — Therefore, the RISE in self-reported youth turnout was a measurable, clearly noticeable phenomenon

    What happened is then;

    — Some pollsters completely denied this phenomenon existed (ICM, ComRes, BMG) and assumed their panel members were all lying to them
    — Some partially accounted for it but downweighted through other means (YouGov, Opinium)
    — One pollster took people at their word (Survation).

    I am honestly quite shocked at the level of stupidity some of the pollsters displayed. Particularly ComRes/ICM/BMG.

    On what possible rationale, assuming they haven’t had a massive influx of politically engaged people into their panels (which they can measure), can they just ignore such a blatant change in self-reported voting behaviour? It is clearly crazy. It makes absolutely no logical sense.

    — It puts an artificial hard ceiling on voting intention figures, at the expense of LAB.
    — You could see this because all pollsters started off in roughly the same place, with regards VI at the start of the campaign
    — Then, as young people became engaged and more inclined to vote LAB, pollsters more inclined to believe self-report showed a big change in LAB VI, while others couldn’t get it above ~35%

    It was staring us in the face and it’s sheer lazyness or stupidity on the part of some of the pollsters not to bother noticing or correcting what was clearly a giant clusterf-ck waiting to happen.

    All of that said – the YouGov model was sensational. A brilliant piece of work. That is the future of polling.

  35. Please can we have another election so I can use the YouGov model to place a load of constituency bets at nice juicy long odds, now that he know how good it is.

  36. CHRIS RILEY

    “It is an open secret that Davis was crucial in convincing May to go to the country. He has never been popular with the more pragmatic Tory MPs who see him as an overconfident, self-important prig and he will be only a little more popular than Nick Timothy right now.

    In short, no, he will not be the next Tory leader unless the backbenches are sure it is a suicide mission.”

    I read that David Davis is attracting criticism for pushing for the GE.
    Given the advantages in achieving five years to conclude EU negotiations and the fact that I doubt if he thought May etc would run a kamikaze campaign that seems rather unfair to me.

    Yes, I read that David Davis is attracting criticism for pushing for the GE.

    Given the advantages in achieving five years to conclude EU negotiations and the fact that I doubt if he thought May etc would run a kamikaze campaign that seems rather unfair to me.

    Even more so when you consider the local election results and the April polls quoted by NickP above.

  37. @Bantams
    ‘You’re doing a Thomas Dolby, blinding yourself with science! 326 is needed for a working majority with 650 MP’s so if 7 on one side of the equation don’t vote it reduces the figure to 319.’

    Not so. If 7 SF MPs abstain by declining to take their seats we are left with 643 MPs who vote. Ergo, 322 required for a bare majority.

  38. Neil
    The only reason why I could think there should be any protests is if Labour tried to seize power when they failed to convince the public that there view of how things should be done was rejected by the public ,of course you could say nor did the Tories convince the public but the truth remains that the Tories “won” by capturing more of the vote and 58 seats more than Labour.

    If Labour try mass protests I would think that would play to the Tories advantage .

    Lewblew

    Actually although I supported May and would have liked a clear majority I’m not to unhappy with the result I voted remain in the brexit vote and was not happy with crashing out the single market as I think it will seriously effect our ability to trade at least with such a small majority in the house a hard brexit seems more unlikely.

  39. KEN

    Crush the Saboteurs, hey Ken?

  40. One thing I couldn’t get a grip on was the swings on the night. Don’t think I’ve ever seen an election where the swings varied so much. Obviously this must have been down to the size of the UKIP vote and remain/leave splits. I can totally understand the Battersea and Canterbury movements (even if I didn’t expect it) but I didn’t get how those Northern and Midlands towns varied so much.

    So many seats that I expected to go Tory on the basis of that Sunderland swing and yet the more marginal the seat the less the swing. Eg- Halifax with a healthy 13% UKIP vote where the Tories only needed 500 pickings more than Labour out of the 5,000 UKIP haul and yet the swing there was in favour of Labour.

    This seemed to be repeated all over the place- Labour seats with huge majorities often had those swings to Tories. The marginals often had none at all or in favour of Labour and every vulnerable Lab hold or Lab gain did exactly what they needed it to do to stay/turn red.

    Maybe it is just there is a Tory ceiling in some of those places and the UKIP vote was made up disproportionately with former Lab voters in those tight marginals but disproportionately Tory in the areas with big Lab majorities?

  41. PAUL CROFT

    I can’t really remember what I actually posted here.

    But I think I was very concerned at the poor presentation in the Con. Manifesto. It really was an error of epic proportions to allow an intended demonstration of Cons’ determination to bring “fairness” into the payment of WFA , and personal contribution to Care in Old Age , to be portrayed as an attack on “pensioners” !

    So far as the Lab Manifesto was concerned-my fear was that in ( reasonably imo) dismissing it’s Tax & Spend approach-not one jot of sympathy was voiced for the 7 years of belt tightening which would cause it to be attractive to millions of people.

    Indeed when TM finally appeared on TV & that nurse complained about the long constraint on her pay-TM told her it was still necessary. I was astounded at that lack of empathy. Just a simple thank you -and I will put this right just as soon as I can would have been something.

    It was crystal clear that the “non-nonsense-get on with it” TM who had started the Campaign with those sky high personal ratings was beginning to be seen as wooden & lacking a common touch.

    How many times have we all agreed here on the importance of Leader’s Ratings?

    Finally, when that YouGov experimental Poll thingy came out with a Hung Parliament implicit in it I became very concerned. I have pretty much lost all confidence in OPs-but have some respect for YouGov-and they were clearly trying to get a handle on granular detail. So it cemented a pretty deep depression & fear.

    Now we just have a sodding mess. I can’t see the outcome-except that the Conservative Party has now asked the UK public to vote twice -arguably unneccessarily on both occasions-and misjudged both outcomes. That would be there tough luck-but The Country is significantly disadvantaged by both & I think this will hang around their necks for a long time.

    If-God Forbid-another GE results from this Hung Parliament I wouldn’t want to bet against a Labour Win.

  42. “when that YouGov experimental Poll thingy came out with a Hung Parliament implicit in it I became very concerned.”

    To be fair, even without the experimental regression model, just the standard poll also showed a 4% Con lead at the time, which surely meant a fair chance of a hung parliament if it was true.

  43. Our Scotland contributors have gone unusually quiet, apart from single messages here and on the previous thread from Peter Cairns.

    But I have been struck by how other regulars from outwith Scotland have not learnt much from last night`s Scottish results, and have made shallow judgements about some of the politicians here.

    There have even been two or three folk saying Ruth Davidson was the star of the show and a candidate to be a future PM. IMO she has not the flair or intelligence to lead a very divided Tory party, and simply hasn`t experience or understanding of key issues in rUK.

    One reason why there is this lack of understanding is that SNP-favouring messagers have been the majority of the Scotland correspondents. Centrist messages have either been moderated out as irrelevant, or their authors have transgressed in too-strong wording.

    Also Scotland is treated as one region by the London-based media, and they cannot comprehend, or don`t wish to, that there are strong regional differences in occupations, outlook and social class. YouGov`s new model likely took this into account, and hence its predictions were pretty accurate.

    And for future polling, the other companies will have to improve their stratifications if they want to remain in political business.

  44. @PETER CAIRNS (SNP)

    “Lowest estimate for SNP was Oliver on 38, after than a 40 and a 41, with the vast majority 45-50.
    Having said that he also had the Tories on 365, so the 3 out on the SNP (8%) is slightly undermined by the 47 put for the Tories (13%!)”

    So please I got at least Scotland mostly right. Who thought that your nation would save the CON/DUP government and prevent a “coalition of chaos”.

    But thanks for coming up with percentage numbers, my CON/LAB miss looks suddenly pretty tiny. Hey, at least I predicted this election more accurate than ICM and BMG.

  45. Haven’t heard the exit poll mentioned once today. We heard it mentioned quite a lot in 2015.

    Funny how it only ever gets mentioned when it significantly underestimates the Tories. ;)

  46. Coiln

    If there was another General election called now for a month’s time, is there any sane person here who doesn’t think Corbyn would win a landslide?

    Obviously we’ll never know for sure … but we might find out later in the year whether another Con leader would do better.

  47. Shevii

    Here’s the map
    https://twitter.com/josephmdurso/status/873199145930149889

    YouGov’s predictive model suggested four or five different ways of the churn of the UKIP vote depending on the constituency.

  48. On UKIP’s vote it’s so tempting when we see,say, their vote drop 10% with Labour rising 7 and the Tories 3 to assume that we now know how their votes divided but it’s probably seldom anything like that.
    Labour gained a great many votes from new voters and recent non-voters.I’m willing to bet a chunk of UKIP votes abstained (The ex-BNP element say?) whilst a much larger proportion than is apparent in the basic figures simply voted Tory. I suspect the Tory vote shifted very sharply rightwards this time and they lost some of the Remain and ex-Lib Dem votes from 2015 to Labour.
    It’ll be a while before the studies enlighten us and I wonder if they’ll reach useful conclusions before the next election comes along.

  49. PAUL CROFT
    Davis’ argument that a GE now was right so as to push the next one to 2022 and beyond brexit timetable was sound especially given their poll lead when the decision was taken. No way could he have expected May to cripple the campaign.

    Noticed that w. hill today has BoJo on 6/4, followed by Davis on 3/1, then Davison 4/1, Rudd 7/1, Hammond 16/1.

    Cant see BoJo as a PM despite his popularity, just too verbally gaffe prone.

  50. London today was like a cloud had been lifted. People were smiling again, looking at each other with the silent acknowledgement that they shared something in common. Almost a spiritual moment. This country will come out all right in the end.

    In terms of the polls, my Lab/Con prediction came close, but the credit goes to the Yougov model which I firmly trashed on first reading, but as I looked at it compared to the cross breaks started making more an more sense.

    And also to Nate Silver for his first rule of polling and Anthony for so clearly spelling out the extent of herding in that last week, making it quite clear that the pollsters were doing what Nate said they always do.

    Thanks Ssimon for all the work compiling the predictions.

1 2 3 49