It is year since the 2017 general election. I am sure lots of people will be writing a lot of articles looking back at the election itself and the year since, but I thought I’d write a something about the 2017 polling error, something that has gone largely unexamined compared to the 2015 error. The polling companies themselves have all carried out their own internal examinations and reported to the BPC, and the BPC will be putting out a report based on that in due course. In the meantime, here are my own personal thoughts about the wider error across the industry.

The error in 2017 wasn’t the same as 2015.

Most casual observers of polls will probably have lumped the errors of 2015 and 2017 in together, and seen 2017 as just “the polls getting it wrong again”. In fact the nature of the error in 2017 was completely different to that in 2015. It would be wrong to say they are unconnected – the cause of the 2017 errors was often pollsters trying to correct the error of 2015 – but the way the polls were wrong was completely different.

To understand the difference between the errors in 2015 and the errors in 2017 it helps to think of polling methodology as being divided into two bits. The first is the sample – the way pollsters try to get respondents who are representative of the public, be that through their sampling itself or the weights they apply afterwards. The second is the adjustments they make to turn that sample into a measure of how people would actually vote, how they model things like turnout and accounting for people who say don’t know, or refuse to answer.

In 2015, the polling industry got the first of those wrong, and the second right (or at least, the second of those wasn’t the cause of the error). The Sturgis Inquiry into the 2015 polling error looked at every possible cause of error, and decided that the polls had samples that were not representative. While they didn’t think the way pollsters predicted turnout was based on strong enough evidence and recommended improvements there too, they ruled it out as cause of the 2015 error.

In 2017 it was the opposite situation. The polling samples themselves had pretty much the correct result to start with, showing only a small Tory lead. More traditional approaches towards modelling turnout (which typically made only small differences) would have resulted in polls that only marginally overstated the Tory lead. The large errors we saw in 2017 were down to the more elaborate adjustments that pollsters had introduced. If you had stripped away all the attempts aimed at modelling turnout, don’t knows and suchlike (as in the table below) then the underlying samples the pollsters were working with would have got the Conservative lead over Labour about right:

What did pollsters do that was wrong?

The actual things that pollsters did to make their figures wrong varied from pollster to pollster. So for ICM, ComRes and Ipsos MORI, it looks as if new turnout models inflated the Tory lead, for BMG it was their new adjustment for electoral registration, for YouGov it was reallocating don’t knows. The actual details were different in each case, but the thing they had in common was that pollsters had introduced post-fieldwork adjustments that had larger impacts than at past elections, and which ended up over-adjusting in favour of the Tories.

In working out how pollster came to make this error we need to have closer look at the diagnosis of what went wrong in 2015. Saying that samples were “wrong” is easy, if you are going to solve it you need to identify how they were wrong. After 2015 the broad consensus among the industry was that the samples had contained too many politically engaged young people who went out to vote Labour and not enough disinterested young people who stayed at home. Polling companies took a mixture of two different approaches towards dealing with this, though most companies did a bit of both.

One approach was to try and treat the cause of the error by improving the samples themselves, trying to increase the proportion of respondents who had less interest in politics. Companies started adding quotas or weights that had a more direct relationship with political interest, things like education (YouGov, Survation & Ipsos MORI), newspaper readership (Ipsos MORI) or straight out interest in politics (YouGov & ICM). Pollsters who primarily took this approach ended up with smaller Tory leads.

The other was to try and treat the symptom of the problem by introducing new approaches to turnout that assumed lower rates of turnout among respondents from demographic groups who had not traditionally turned out to vote in the past, and where pollsters felt samples had too many people who were likely to vote. The most notable examples were the decision by some pollsters to replace turnout models based on self-assessment, with turnout models based on demographics – downweighting groups like the young or working class who have traditionally had lower turnouts. Typically these changes produced polls with substantially larger Conservative leads.

So was it just to do with pollsters getting youth turnout wrong?

This explanation chimes nicely with the idea that the polling error was down to polling companies getting youth turnout wrong, that young people actually turned out at an unusually high level, but that polling companies fixed youth turnout at an artificially low level, thereby missing this surge in young voting. This is an attractive theory at first glance, but as is so often the case, it’s actually a bit more complicated than that.

The first problem with the theory is that it’s far from clear whether there was a surge in youth turnout. The British Election Study has cast doubt upon whether or not youth turnout really did rise that much. That’s not a debate I’m planning on getting into here, but suffice to say, if there wasn’t really that much of a leap in youth turnout, then it cannot explain some of the large polling misses in 2017.

The second problem with the hypothesis is that there isn’t really that much relationship between those polling companies who had about the right proportion of young people in their samples and those who got it right.

The chart below shows the proportion of voters aged under 25 in each polling company’s final polling figures. The blue bar is the proportion in the sample as a whole, the red bar the proportion in the final voting figures, once pollsters had factored in turnout, dealt with don’t knows and so on. As you would expect, everyone had roughly the same proportion of under 25s in their weighted sample (in line with the actual proportion of 18-24 year olds in the population), but among their sample of actual voters it differs radically. At one end, less than 4% of BMG’s final voting intention figures were based on people aged under 25s. At the other end, almost 11% of Survation’s final voting figures were based on under 25s.

According to the British Election Study, the closest we have to authorative figures, the correct figure should have been about 7%. That implies Survation got it right despite having far too many young people. ComRes had too many young people, yet had one of the worst understatements of Labour support. MORI had close to the correct proportion of young people, yet still got it wrong. There isn’t the neat relationship we’d expect if this was all about getting the correct proportion of young voters. Clearly the explanation must be rather more complicated than that.

So what exactly did go wrong?

Without a nice, neat explanation like youth turnout, the best overarching explanation for the 2017 error is that polling companies seeking to solve the overstatement of Labour in 2015 simply went too far and ended up understating them in 2017. The actual details of this differed from company to company, but it’s fair to say that the more elaborate the adjustments that polling companies made for things like turnout and don’t knows, the worse they performed. Essentially, polling companies over-did it.

Weighting down young people was part of this, but it was certainly not the whole explanation and some pollsters came unstruck for different reasons. This is not an attempt to look in detail at each pollster, as they may also have had individual factors at play (in BMG’s report, for example, they’ve also highlighted the impact of doing late fieldwork during the daytime), but there is a clear pattern of over-enthusiastic post-fieldwork adjustments turning essentially decent samples into final figures that were too Conservative:

  • BMG’s weighted sample would have shown the parties neck-and-neck. With just traditional turnout weighting they would have given the Tories around a four point lead. However, combining this with an additional down-weighting by past non-voting and the likelihood of different age/tenure groups to be registered to vote changed this into a 13 point Tory lead.
  • ICM’s weighted sample would have shown a five point Tory lead. Adding demographic likelihood to vote weights that largely downweighted the young increased this to a 12 point Tory lead.
  • Ipsos MORI’s weighted sample would have shown the parties neck-and-neck, and MORI’s traditional 10/10 turnout filter looks as if it would have produced an almost spot-on 2 point Tory lead. An additional turnout filter based on demographics changed this to an 8 point Tory lead.
  • YouGov’s weighted sample had a 3 point Tory lead, which would’ve been unchanged by their traditional turnout weights (and which also exactly matched their MRP model). Reallocating don’t knows changed this to a 7 point Tory lead.
  • ComRes’s weighted sample had a 1 point Conservative lead, and by my calculations their old turnout model would have shown much the same. Their new demographic turnout model did not actually understate the proportion of young people, but did weight down working class voters, producing a 12 point Tory lead.

Does this mean modelling turnout by demographics is dead?

No. Or at least, it shouldn’t do. The pollsters who got it most conspicuously wrong in 2017 were indeed those who relied on demographic turnout models, but this may have been down to the way they did it.

Normally weights are applied to a sample all at the same time using “rim weighting” (this is an iterative process that lets you weight by multiple items without them throwing each other off). What happened with the demographic turnout modelling in 2017 is that companies effectively did two lots of weights. First they weighted the demographics and past vote of the data so it matched the British population. Then they effectively added separate weights by things like age, gender and tenure so that the demographics of those people included in their final voting figures matched the people who actually voted in 2015. The problem is this may well have thrown out the past vote figures, so the 2015 voters in their samples matched the demographics of 2015 voters, but didn’t match the politics of 2015 voters.

It’s worth noting that some companies used demographic based turnout modelling and were far more successful. Kantar’s polling used a hybrid turnout model based upon both demographics and self-reporting, and was one of the most accurate polls. Surveymonkey’s turnout modelling was based on the demographics of people who voted in 2015, and produced only a 4 point Tory lead. YouGov’s MRP model used demographics to predicts respondents likelihood to vote and was extremely accurate. There were companies who made a success of it, and it may be more of a question about how to do it well, rather than whether one does it at all.

What have polling companies done to correct the 2017 problems, and should I trust them?

For individual polling companies the errors of 2017 are far more straightforward to address than in 2015. For most polling companies it has been a simple matter of dropping the adjustments that went wrong. All the causes of error I listed above have simply been reversed – for example, ICM have dropped their demographic turnout model and gone back to asking people how likely they are to vote, ComRes have done the same. MORI have stopped factoring demographics into their turnout, YouGov aren’t reallocating don’t knows, BMG aren’t currently weighting down groups with lower registration.

If you are worried that the specific type of polling error we saw in 2017 could be happening now you shouldn’t be – all the methods that caused the error have been removed. A simplistic view that the polls understated Labour in 2017 and, therefore, Labour are actually doing better than the polls suggest is obviously fallacious.
However, that is obviously not a guarantee that polls couldn’t be wrong in other ways.

But what about the polling error of 2015?

This is a much more pertinent question. The methodology changes that were introduced in 2017 were intended to correct the problems of 2015. So if the changes are reversed, does that mean the errors of 2015 will re-emerge? Will polls risk *overstating* Labour support again?

The difficult situation the polling companies find themselves in is that the methods used in 2017 would have got 2015 correct, but got 2017 wrong. The methods used in 2015 would have got 2017 correct, but got 2015 wrong. The question we face is what approach would have got both 2015 and 2017 right?

One answer may be for polling companies to use more moderate versions of the changes them introduced in 2017. Another may be to concentrate more on improving samples, rather than post-fieldwork adjustments to turnout. As we saw earlier in the article, polling companies took a mixture of two approaches to solving the problem of 2017. The approach of “treating the symptom” by changing turnout models and similar ended up backfiring, but what about the first approach – what became of the attempts to improve the samples themselves?

As we saw above, the actual samples the polls used were broadly accurate. They tended to have smaller parties too high, but the balance between Labour and Conserative was pretty accurate. For one reason or another, the sampling problem from 2015 appears to have completely disappeared by 2017. 2015 samples were skewed towards Labour, but in 2017 they were not. I can think of three possible explanations for this.

  • The post-2015 changes made by the polling companies corrected the problem. This seems unlikely to be the sole reason, as polling samples were better across the board, with those companies who had done little to improve their samples performing in line with those who had made extensive efforts.
  • Weighting and sampling by the EU ref made samples better. There is one sampling/weighting change that nearly everyone made – they started sampling/weighting by recalled EU ref vote, something that was an important driver of how people voted in 2017. It may just be that providence has provided the polling companies with a useful new weighting variable that meant samples were far better at predicting vote shares.
  • Or perhaps the causes of the problems in 2015 just weren’t an issue in 2017. A sample being wrong doesn’t necessarily mean the result will be wrong. For example, if I had too many people with ginger hair in my sample, the results would probably still be correct (unless there is some hitherto unknown relationship between voting and hair colour). It’s possible that – once you’ve controlled for other factors – in 2015 people with low political engagement voted differently to engaged people, but that in 2017 they voted in much the same way. In other words, it’s possible that the sampling shortcomings of 2015 didn’t go away, they just ceased to matter.

It is difficult to come to firm answer with the data available, but whichever mix of these is the case, polling companies shouldn’t be complacent. Some of them have made substantial attempts to improve their samples from 2015, but if the problems of 2015 disappeared because of the impact of weighting by Brexit or because political engagement mattered less in 2017, then we cannot really tell how successful they were. And it stores up potential problems for the future – weighting by a referendum that happened in 2016 will only be workable for so long, and if political engagement didn’t matter this time, it doesn’t mean it won’t in 2022.

Will MRP save the day?

One of the few conspicuous successes in the election polling was the YouGov MRP model (that is, multi-level regression and post-stratification). I expect come the next election there will be many other attempts to do the same. I will urge one note of caution – MRP is not a panacea to polling’s problems. They can go wrong, and still relies on the decisions people make in designing the model it runs upon.

MRP is primarily a method of modelling opinion at lower geographical areas from a big overall dataset. Hence in 2017 YouGov used it to model the share of the vote in the 632 constituencies in Great Britain. In that sense, it’s a genuinely important step forward in election polling, because it properly models actual seat numbers and, from there, who will win the election and will be in a position to form a government. Previously polls could only predict shares of the vote, which others could use to project into a result using the rather blunt tool of uniform national swing. MRP produces figures at the seat level, so can be used to predict the actual result.

Of course, if you’ve got shares of the vote for each seat then you’ll also be able to use it to get national shares of the vote. However, at that level it really shouldn’t be that different from what you’d get from a traditional poll that weighted its sample using the variables and the same targets (indeed, the YouGov MRP and traditional polls showed much the same figures for much of the campaign – the differences came down to turnout adjustments and don’t knows). Its level of accuracy will still depend on the quality of the data, the quality of the modelling and whether the people behind it have made the right decisions about the variables used in the model and on how they model things like turnout… in other words, all the same things that determine if an opinion poll gets it right or not.

In short, I do hope the YouGov MRP model works as well in 2022 as it did in 2017, but MRP as a technique is not infallible. Lord Ashcroft also did a MRP model in 2017, and that was showing a Tory majority of 60.


  • The polling error in 2017 wasn’t a repeat of 2015 – the cause and direction of the error were complete opposites.
  • In 2017 the polling samples would have got the Tory lead pretty much spot on, but the topline figures ended up being wrong because pollsters added various adjustments to try and correct the problems of 2015.
  • While a few pollsters did come unstuck over turnout models, it’s not as simple as it being all about youth turnout. Different pollsters made different errors.
  • All the adjustments that led to the error have now been reversed, so the specific error we saw in 2017 shouldn’t reoccur.
  • But that doesn’t mean polls couldn’t be wrong in other way (most worryingly, we don’t really know why the underlying problem in 2015 error went away), so pollsters shouldn’t get complacent about potential polling error.
  • MRP isn’t a panacea to the problems – it still needs good modelling of opinion to get accurate results. If it works though, it can give a much better steer on actual seat numbers than traditional polls.

138 Responses to “Why the polls were wrong in 2017”

1 2 3
  1. That Yougov/Times poll is an outlier.

    (Caveat: I said exactly the same about Yougov’s MLR model and that was bang-on.)

  2. @profhoward:

    On the N. Ireland poll, it suggests that the EU issue means support has gone from 69-27 in favour of the U.K. to 45-41.

    That seems an awfully big swing given the lack of change in Scotland, and that N. Ireland might be left in the EU regardless.

    It would be interesting to hear from AW on this.

    You’d feel a muppet if you voted to join ROI solely on the EU issue, and then the whole thing collapsed.

  3. @ ProfHoward

    Thanks for the NI poll link, hadn’t seen that. One thing I don’t understand (one of many probably) is the apparent inconsistency in answers to the (to paraphrase) “Has Brexit changed your mind on border poll” and the actual border poll question. I guess this is another illustration of the point that the phrasing of the question matters.

    What I mean is that if you take the “Changed your mind” answers at face value, you could add up the “always wanted NI to join RoI” and “Brexit changed my mind so I now want to join RoI” numbers to get 55%, and you might presume these 55% would vote to join RoI in a referendum. Similarly, you could add the other answers to get 42% to stay in UK. Thanks 55:42 (with 3% DK) in favour of joining RoI. But when asked the question directly, it goes completely the other way, 42:45 with many more DKs. The tables might help, there’s probably some fiddling with DKs and “would not answer”s, but even so, that’s a large reversal between the two questions.

    On the latest YouGov, the C2DE’s are worth looking at again. We discussed this a few weeks ago, it really does go up and down like a yo-yo. This time it’s 48:37 for Cons. If this is becoming a consistent trend, that’s got to be a big worry for Labour.

    Finally, on Richard Leonard, I don’t know much about him, but I heard him for the first time on QT a couple of weeks ago. I have to say I was most underwhelmed. He didn’t seem to know very much about anything, and while some people can bluster so that it doesn’t matter that they know nothing, he didn’t seem to be good at that either. Maybe that’s the problem with SLAB support. Come back Kezia, all is forgiven.

  4. From the YouGov/Times Poll:-

    In hindsight, do you think Britain was right or
    wrong to vote to leave the European Union?

    28/29 May 40/47
    4/5 June 44/44

  5. A couple of corrections for above. There’s a spurious “Thanks” instead of “That’s”. Don’t know how that happened. And I meant “Any Questions” not QT.


    The YouGov is not really an outlier, it is just within the margin of error of 42/39 which is about the current average. It is just at the edge of the normal results. It is more than a coincidence that the poll has seen a similar shift in “right or wrong” on Brexit. This poll just shows Con at the top of their current range and Lab near the bottom of theirs. Over the month I expect YouGov to average out as a more “normal” set of figures unless there is some major event to shift things.


    @” Over the month I expect YouGov to average out as a more “normal” set of figures unless there is some major event to shift things.”

    What do consider to be “normal” ?

  8. Joseph1832: “You’d feel a muppet if you voted to join ROI solely on the EU issue, and then the whole thing collapsed.”

    Like the people who voted to keep Scotland in the UK because they were worried about Scotland’s EU membership?

  9. Bazinwales

    That’s pretty much what outlier means when people talk about MOE. There isn’t an arbitrary line drawn somewhere which classifies a result as “normal variation” or “outlier”. There isn’t some mystical quality possessed by a poll lying on the 96th percentile which a poll on the 94th percentile doesn’t have. You shouldn’t necessarily throw these sort of “outliers” away. Doing so can distort the data and underestimate the natural “spread”.

    The other sort of outlier are caused by something going wrong in either the measuring or analysis, or some fundamental reason why the measurement was made in a non-standard state (such as measuring a particular feature of a component when the component was either broken or missing).

    These sort of outliers should usually be excluded. Running a comparison of the analysis between including and excluding these data points is usually performed so the effect of excluding them can be determined.

  10. As observed yesterday, the EU seems to have rejected the backstop plan on the grounds that it cannot be time limited and that if the whole UK wishes to stay in the customs area, as the plan suggests, then that is cherry picking the single market. Either the UK stays in completely, or there is a backstop aligned to NI only, to preserve the GFA.

    Beggars belief really, but we were posting on this yesterday, but somehow the entirety of the UK government can’t seem to remember what they have already agreed and what the EU has already said.

  11. DAVWEL,

    On a roundabout going into Leeds there is a handwritten sign on a lamppost saying “Polish plumber” with a phone number.

    I suspect this means that local people have found by experience that Polish plumbers are more likely to be good value for money than English ones.

    The interesting question is whether some of those same people voted Leave to stop EU immigration…

    Similarly, I had my car serviced and MOT’d yesterday by a Lithuanian guy who was recommended by my neighbour (who I am pretty sure voted Brexit). He has an attitude that I should only pay if the repair is good..

    I should add that I have found English tradesmen with a similar attitude, and who are good value for money. But I tend to see it as rare…

  12. @Alec:

    What the EU wants should not be accepted as the baseline for normal and reasonable.

    Barnier is still pretending that putting N Ireland significantly under EU is somehow normal and in no way harms U.K. territorial waters integrity. N Ireland will on the EU sid of disputes, and Ireland will be the arbiter of N Ireland’s interests in the EU…

    But I agree it is predictable.

    May is still mistaking it for a negotiation between friends. The EU sees it as the Versailles Treaty, except the defeated party gets to turn up and be laughed at.

    May should resign so that either someone can try to stand up to the EU, or a Remainer can sue for peace. At least let a Remainers take responsibility for their victory. They have worked hard for it.

    But on any view May has failed completely.

  13. @ A Wells
    “The actual details of this differed from company to company, but it’s fair to say that the more elaborate the adjustments that polling companies made for things like turnout and don’t knows, the worse they performed.”

    It is surely a wonderful irony that the straight-forward weighted samples provided the “correct” result & that the application of polling “expertise” created the “wrong” result.

    Lest we forget. This is the BBC on the botched polling in 2015.

    “The fact that most polls suggested a hung parliament shaped the entire campaign. Detailed interrogation of party policies took a back seat to endless discussion about who would go into coalition with whom. And the final weeks before polling day were so dominated by talk about the SNP that one could have been forgiven for thinking it was a Holyrood election rather than one for Westminster.”

    The fear of the SNP in a coliation was v damaging to the Lib-Dem vote in the many seats they lost?

  14. @ Peter Cairns
    Thanks for yr reply: belated response

    “Not NI but I suspect that the idea that Central Government shouldn’t legislate in devoted areas is pretty universal across the UK and most polls tend to show that.”
    You quote not NI but Scottish evidence to support UK-wide speculation. It seems pretty self-evident that Scottish respondents would support the autonomy of devolved administrations.

    I “suspect” that most people in England who consider NI think it’s ridiculous that major elements of devolved government have been suspended by its own representatives; that SF MPs never sit in London; that DUPs sit in London & prop up London government while their colleagues don’t sit at all in NI; that the DUP want Brexit & a soft border with rest of Ireland.
    In general Briish people are v keen on NI autonomy to the extent that “most” would prefer NI to leave Union than lose Brexit.

    “I seem to recall we had direct rule for decades and never brought Abortion legislation into line.”
    Times change: except in the minds of NI politicians, who are now behind their public on this issue.
    I can see that a (desirable) referendum on abortion in NI would have to be scheduled by the executive, but looks like that will never happen despite polls = “there is overwhelming public support for liberalising the anti-abortion regime in the province.”

  15. @Alec – i think the UK government know what they are doing even if it looks like they aren’t. Regarding why they are doing it. Maybe it’s to try and draw attention to the problems/big issues with Brexit to get as many leavers as possible, including in their own parliamentary party, to not be too fussed when we have a soft Brexit. Maybe also to get as many remainers/soft brexiters agitated and call for a soft brexit – have seen a few new names (to me anyway) poke their head above the parapet in the last day or so.

    Or maybe its to try and say that we did what we could but the EU were being horrible, etc, and leave with no deal.

    Seems to me that the second option is unlikely given the short time frame, so maybe its the first option.


    I agree. There was actually two elections in 2015 Lib Dem versus Tories in the South West and Labour versus SNP in Scotland. The rest of England and Wales was a wash.

    The LD in line with their complete ineptitude did not realise that was the Tories game plan until it was too late. They believed that Labour supporter would turn out and vote Liberal Democrat in the South West if you look at the figures the Labour party vote did quite well in South West, it lead to LD losing seats. The big issue on the doorstep was not letting the Scots have any influence (mos people did not put it so politely I was told). So it was essentially the Liberals lost their anti Ttory credentials with protest voters and Tories galvanised their voters. If you look at the numbers there was not a explosion of support for the Tories. LibDem either did not show or the protest vote went elsewhere.

    If you look at the Tories gains in the South West that is where their majority was made. Even Ed Balls losing his seat to my mind was less fascinating than what happened in the South West. What really has been interesting about everything is that the same thing happened in Scotland this time. May is only in power because of Scottish Tories and most probably Labour and Liberal Democrat intransigence in seeing the SNP as the big enemy.

    I fear that much of the navel gazing on the polls is because the electorate is so volatile these days. It is not about policy but about tribe

  17. @FROSTIE

    One could argue both the Tories and Labour have two tribes within their party that are unable to agree a path forward. I think one side feels that if it goes bad they will get the blame and the other side is suggesting a gung ho approach

    In terms of the Parties, I believe BoJo is agitating for harder approach because he know it will not happen. but the others I think genuinely believe in their coolaid.

    On the Labour side I think there is a partial fear that Brexit is important to a section of their voters and whilst I see Tories hanging on to remainers, I don’t see Labour hanging on to leavers as easily.

    The whole issue with EU is that I would not agree a backstop which involved the UK because I don’t think it would work or May would be able to sell it. I don’t think that either party could sell it and be sure

  18. @Joseph1832 – I think we are now at the interesting point. The EU have (quite rightly) said that May’s plan leaves the UK with the benefits of customs free trade without any of the responsibilities of regulatory alignment, and so would be unacceptable cherry picking. This also seems reasonable as May has already agreed to maintain regulatory alignment in NI in the December Agreement, but then fails to deliver on this in yesterday’s backstop plan.

    The issue really comes down to whether the EU is acting fairly by demanding a specific NI only arrangement that excludes rUK. It is worth noting that the EU proposal (deemed unacceptable by May) would create a division between NI and rUK, but I don’t think the EU have definitively said that the backstop must inherently only apply to NI – only that an arrangement that distorts the four freedoms wouldn’t be acceptable. The EU are prepared to distort the four freedoms for NI, to respect the GFA, but not to do so for the UK as a whole. This is arguably an element of cherry picking in itself, albeit justified by trying to avoid any disturbance to the peace.

    This leaves open the idea that a UK wide backstop that accepts regulatory alignment might be acceptable to the EU, but effectively means staying in the EU.

    I think there’s a bit to quibble about here on both sides, but my overall impression is that HMG are struggling to understand what is needed to secure a deal, and I don’t think they fully understood the implications of what they had agreed to in December, or perhaps they did, but May just hoped she could wing it and persuade Boris et al. to go along with it.

  19. JOSEPH1832

    @”May is still mistaking it for a negotiation between friends. ”


  20. JOSEPH1832

    From This Week-BBC

    ” “The nature of our future relationship with the European Union, will not now be determined until we enter the transition period.
    It will probably go beyond the transitional period because we have now have added another year with this backstop.
    But we still have to pay the £39 billion immediately after March of next year. I mean we have been outmanoeuvred on this.”

    Andrew Neil

    “: “That is an understatement. Textbooks will be written on the incompetence of the British negotiation.
    But to call it incompetence is to be too kind, I think this has been a conspiracy.
    As you said a moment ago, all the senior civil servants are Remainers, about half the cabinet are Renmainers, and probably more than half of the Conservative parliamentary party are Remainers.”

    Michael Portillo

  21. @COLIN
    ‘Andrew Neil

    “: “That is an understatement. Textbooks will be written on the incompetence of the British negotiation.
    But to call it incompetence is to be too kind, I think this has been a conspiracy.
    As you said a moment ago, all the senior civil servants are Remainers, about half the cabinet are Remainers, and probably more than half of the Conservative parliamentary party are Remainers.”

    I am loathe to bring Godwins law into this but the statement very much reminds me of nazi propaganda that Germany did not lose the First World war but were sold out by the politicians.

    I am not a May supporter but all the problems she is having now were clearly to be seen prior to the Brexit referendum. But for some reason more ardent Brexiteers were blind to it, using ‘they need us more than them’ mantra stupiaims such as the trade deal should be “easiest in human history”

    Yes she has been inept and so was Davis, Fox and Johnson. But we were always on a hiding to nothing the moment we triggered Articlce 50. Any re-writing of history blaming it on remainers is just silly. As my mum used to say you made your bed now you lie in it

  22. Alec,
    “@Joseph1832 – I think we are now at the interesting point. The EU have (quite rightly) said that May’s plan leaves the UK with the benefits of customs free trade without any of the responsibilities of regulatory alignment, and so would be unacceptable cherry picking. This also seems reasonable as May has already agreed to maintain regulatory alignment in NI in the December Agreement, but then fails to deliver on this in yesterday’s backstop plan.”

    This just highlights that the whole Labour Party thing of backing “a Customs Union” but not the Single market with the 4 Freedoms never made any sense. It is a cake which the EU was never going to give us.

    If the UK would say “we want EEA plus CU” I think the EU would be happy to negotiate with us on that basis, including what would and would not g in our EEA deal. Because there is a price for an EEA deal it will take time to negotiate (the timescale of a transition deal, approximately)

    This is in fact the only thing that is consistent with the agreement on NI in December, which is why I said at the time “May has just agreed to a soft Brexit”. HM government has now wasted another 6 months arguing internally about which unacceptable proposal will be put to the EU, and the Labour Party has confused the issue in a quite irresponsible fashion by pretending they could get some deal which is equally impossible…

    I think frustration in both Leave and Remain camps is really building up, but the Remain camp in particular sense that the tide may be turning in their direction

  23. Survation

    CON 41 (-)
    LAB 40 (-)
    LDem 9 (+1)
    UKIP 2 (-1)
    GRN 2 (-)

    “May is only in power because of Scottish Tories and most probably Labour and Liberal Democrat intransigence in seeing the SNP as the big enemy.”

    I don’t know why things went so badly wrong with the SNP in 2017 & our usually voluble SNP posters had little to say about it.
    Two inexpert points.
    I. N. Sturgeon made a mess of it, muddling the GE & a 2nd ref when she should have concentrated on preserving the SNP seats & fighting for the 2nd ref on another battle field. The Unionist vote was fortified.
    2. The bit I don’t understand is why the Brexit Tory Unionists benefitted most. Lib-Dem fortunes again significant. There was a shift of Lib-Dem votes to Tories in 2015 & in 2017 this shift continued in say half the seats the Tories won? Eg Salmond won a Lib-Dem seat in 2015 because Lab NOT Lib-Dem voters switched to him; he lost in 2017 because of massive Lib-Dems shift to Tory. Puzzling.

  25. NEILJ

    @”I am loathe to bring Godwins law into this”

    ………but you overcame your loathing :-)

  26. Thanks to Andrew for the latest briefing from George Street.

    A customs union would have easy movement of people not free movement and an ECJ hybrid for oversight as a sop so minimal change and potentially acceptable to the EU, the other freedoms aren’t an issue.

    Stamer has been in touch with the EU throughout and taken seriously and the call for unilateral rights for EU citizens in the UK very soon after the vote banked credit for Labour which HMG does not have; and more respect in ROI.

    There are cakeish elements to Labours starting position but they are clear there would be a negotiation.

  27. Andrew111
    “the Labour Party has confused the issue in a quite irresponsible fashion by pretending they could get some deal which is equally impossible…”

    The opposition’s purpose is to oppose, like it says on the tin.

    2015 demonstrated to both both Labour and LibDems what happens under FPTP in the UK when supposedly left of centre parties collude with the Tories, whether in the Indyref campaign or by entering coalition.

    Labour don’t need a coherent position on Brexit. Apart from being impossible, since Brexit is such a transparently insoluble conundrum, it’s in their political interest not to have one.

    Cameron failed to lose in 2010 by studiedly having no policies until about a fortnight before the election, and if Labour had taken a leaf out of his book and refused to get involved with Indyref they might well have been back in government today, albeit with one of the right-wing moderates in charge instead of Corbyn and McDonnell.

    As long as they can send Keir Starmer onto the telly or radio once in a while to drone on about customs unions in a way that sounds reasonable even if no-one really understands it, and the few that do know that it’s no more stupid than what the Tories are coming out with, they remain in a position where they cannot be tarnished by the Tories’ toxic Brexit.

    I fail to see how any of this can be placed at the door of Labour when the Cabinet have just spent six weeks coming up with a fudged combination of two proposals which the EU already rejected last year, it’s a good job Barnier has the patience of a saint and even when speaking in a second tongue a grasp of diplomatic language which puts our Foreign Secretary to shame.

  28. Thanks Frosty.

    Survation consistent with a 2-3% Tory lead or so average of pollsters (NB average of polls lead bigger Con lead as more YG which show bigger leads presently which may be right of course).

  29. Oh, that was short-lived. New thread.

    Bet this one contains the highest proportion of non-B-word posts in recent times.

  30. AndrewIII @ 2.41 pm

    Thanks for the supporting info on attitudes to Polish workers.

    Our gang have just left at 6 pm, they still were harling at 5.45 on a hot day, having started at 7 am. I said as they walked to their van “will you be back 7 am tomorrow?” Yes, they replied smiling.

  31. One important point.

    You said it is possible the issues that mattered in 2015 (namely sampling bias) may not have mattered as much in 2017, hence no significant difference between polling companies that corrected well for it and those that did less well.

    If true, that still means the same sampling bias can very easily impact future polls. Those that took less care were therefore simply lucky it did not put them too much out of whack. And it seems those that did take care did not suffer at all from doing so via some kind of overcompensation.

    Hence the sampling fixes applied to 2015 should be applied to all future polls regardless of whether it is known to be an issue.

    Alternatively, one can track comparisons between full effective sampling and more lax sampling methods to see if you can detect bias arising. If not, then you know sampling bias is not arising in this election so can save resources or dedicate those resources in focussing elsewhere.

  32. The one exciting thing I thought MPR tried, and I may be wrong (in which case one should give the below a go) is that aside from essentially just doing local polling (which we saw in the US didn’t work very well), it also used big data tools to apply signals it found in once constituency to other constituencies with similar demographics, possibly weighted by those demographics.

    If so, then there may be a systemic reason why MPR could produce better results.

    If you think about it, the data you get from a local poll is far richer than simply predicting one region. It can determine the impact of localised issues and demographics and project how these issues will come across on similar populations with similar characteristics. And the best thing is, to some extent, the machine learning algorithm can detect and apply the signals for you without you having to identify them.

  33. Thanks for the answers on Corbyn and Scotland. That is a huge drop in his personal ratings. The latest YouGov showing a Con lead of 7% is extra ordinary given that even Tory supporters know their party has been a mess over Brexit negotiations.

  34. Joseph1832,
    “What the EU wants should not be accepted as the baseline for normal and reasonable.”

    Mrs May goes off to Brussels, makes an agreement for a non time limited no border special deal for N. Ireland. Then six months later she sends off a proposal for a time limited arrangement to apply to the entire UK. Its really not the EU which is behaving unreasonably here. A pack of chimpanzees would make more sense than our negotiating team. Ar least they might all sit about serving tea.

    ” on any view May has failed completely.”

    On my view she is doing a good job of making Brexit impossible and bringing remain closer. Complicated, innit.

    “my overall impression is that HMG are struggling to understand what is needed to secure a deal”

    I dont think that is true at all. In fact, I cannot conceive how it could be true. The government is not trying to accomplish the relatively simple task of negotiating a deal with the EU, but of squaring a soft brexit or even remain with its voters.

    I still think May’s crowning moment would be if she could annouce that Brexit is cancelled and the UK will remain a full member of the EU. She would then have achieved something Cameron failed to do.

    “This just highlights that the whole Labour Party thing of backing “a Customs Union” but not the Single market with the 4 Freedoms never made any sense”

    Well yes, theyre at it too. Everyone is manoevering to demolish Brexit item by item.

    The exterminatingdalek,
    ” Barnier has the patience of a saint and even when speaking in a second tongue a grasp of diplomatic language which puts our Foreign Secretary to shame.”
    ‘backstop means backstop’

  35. Following on from AW’s thread it should be relatively easy for folks to do post-GE multiple regression analysis on the seat level data. I’ll avoid the stats jargon and simplify but for each seat you can create a formula for the votes each party should have got in each seat and then run the numbers to solve for least error.


    Party17 = aParty15 + bBrexit + cAge +dHouse prices +…

    You can run lots of different variables (commons library has a database where you can get the info).

    B4B have done this analysis just using Party15 and Brexit (they split Brexit out to Remain and Leave)

    There results gave the following ‘best fit’:

    CON_17 = 0.97*CON_15 + 0.23 * Leave – 0.1 * Remain

    (ie high loyalty with a small loss from GOTV, a boost from Leave and a hit from Remain)

    LAB_17 = 1.04*LAB_15 + 0.26 Remain – 0.01 * Leave

    (ie high loyalty with a small gain from GOTV, a boost from Remain and minimal lose due to Leave).

    However, one needs to look at the R2 to see if this is a good ‘model’. They don’t state the R2 (as it doesn’t suit their purposes to do so) but by simple eyeball you can see it fits fairly well for CON but is quite a poor fit for LAB. Personally I’d also focus heavily on the marginal seats and/or the gains v loses – if you do that you’ll see LAB’s R2 is very low where as complacent CON stays quite high.

    Any model folks can redo the analysis using other variables (best to add in LDEM, and UKIP where they stood as well IMHO but for simplicity lump Green in with LDEM). You should find that Brexit was actually a fairly low predictor, especially in the marginal or won/lost seats (after ‘loyalty’, age was the highest factor, then house prices*)

    B4B do mention the tactical vote success in 2017 and that can partially explain a lower R2 for the LAB formula.

    If you want to go one step further you can model a ‘Party19’ vote by seat and see which variables are most important to ‘prioritise’, especially in marginal seats that you need to defend and marginal seats you hope to gain!
    (LAB folks have done this in the final chapter of ‘Corbyn Effect’ by Mark Perryman)

    P.S. Tip. Remove Scotland and London* from the analysis and run those separately. OK, OK, that is not ideal and smacks of torturing the data to confess to what you want but if you keep Scotland and London in then you’ll struggle to get high R2 readings.

  36. @Danny:

    I am not sure how you deny the concept of away legal supremacy? It is EU legal orthodoxy.

    Now, you can talk about the position of the German Constitutional Court. There is a theoretical possibility of EU law being rejected as unconstitutional. But it has never happened. It is as best a rather theoretical exception to the supremacy of EU law.

    So, how can we leave? Obviously the right to secede is different from the Rights enjoyed whilst a member. There was a lively discussion about whether member states had a right if UDI. But, since 2009, EU law confers a right to leave – although the EU does not look at is a “no fault” divorce.

    Anyway, no more comparisons with NATO. There is an obvious argument that it is just too difficult to leave the EU. But it destroys sovereignty comparisons with other Treaty arrangements.

  37. ‘Disinterested’ doesn’t mean uninterested, it means impartial. So you have ‘uninterested’ young voters. If anyone is disinterested in the process, it is (hopefully) the vote counters.

1 2 3