Time for a post-mortem of the European election polling. I’m not a fan of the sort of horse-race approach to these things – just because they are the final poll of the race, they still have normal margins of error, so if one pollster is a fraction more accurate than another it is often just the luck of the draw. Realistically the best a pollster can ever hope to do is get all the results within the margin of error. Better than that is luck. However away from the public “who won” stuff, comparing poll predictions to actual election results is an absolutely critical tool for pollsters – it’s our chance to compare our figures with reality, to improve and finesse our methods.

The final polls from each company are here. Note that I haven’t included Populus – they did conduct one European poll, but it was a fortnight before the election when just a week is a long time in politics! While I’ve included it in the comparison, one should allow some leeway for ICM for the same reason; their poll’s fieldwork finished a week before the actual election.

CON LAB LD UKIP GRN Average Error
ACTUAL RESULT 23.9 25.4 6.9 27.5 7.9
YouGov 22 26 9 27 10 1.4
(-1.9) (+0.6) (+2.1) (-0.5) (+2.1)
ICM 26 29 7 25 6 2.0
(+2.1) (+3.6) (+0.1) (-2.5) (-1.9)
Opinium 21 25 6 32 6 2.1
(-2.9) (-0.4) (-0.9) (+4.5) (-1.9)
TNS 21 28 7 31 6 2.2
(-2.9) (+2.4) (+0.1) (+3.5) (-1.9)
ComRes 20 27 7 33 6 2.6
(-3.9) (+1.4) (+0.1) (+5.5) (-1.9)
Survation 23 27 9 32 4 2.6
(-0.9) (+1.4) (+2.1) (+4.5) (-3.9)

The most obvious current difference between Westminster polls is the reported levels of UKIP – there is a big gulf between the levels of UKIP support report recorded by companies like ICM, MORI, YouGov and ComRes’s phone polls and polls from newer companies like Opinium, Survation and ComRes’s online polls. We don’t know what the reasons for this are – there are a couple of things like prompting and re-allocating don’t knows that we can account for, but mostly the difference is not easily explained. It may be something to do with interviewer effect, or the representativeness of different companies samples. We can’t tell.

The European elections were obviously an opportunity to check figures against reality. I half expected the polls to all converge together in the run up to the election, as they have a tendency to do before general elections, but in reality we got the same sort of contrast as we do in Westminster polls. Higher figures for UKIP amongst newer online companies, lower figures from YouGov, lowest from ICM… and when the votes were counted the YouGov figure was the closest.

Of course, European elections aren’t general elections. On the issue of prompting, for example, every company prompted for UKIP in their European polling, whereas only Survation do it for general elections. There were no telephone polls for the European election, so it can tell us nothing of them. European elections are low turnout elections, so some of the errors may have been down to too strict turnout filters (ComRes used a very strict turnout filter for Euros and would probably have been better if they’d used the method they use for general election polling. There was the issue of the Independence from Europe spoiler party on the ballot paper and so on. At a purely personal level though, getting UKIP right at the next election is the biggest challenge currently facing pollsters, so I’m relieved that in the first real proper national test we got it right. Phew!


440 Responses to “European polling post-mortem”

1 2 3 9
  1. .tsriF

  2. Euro election final vote share changes (from http://www.bbc.co.uk/news/events/vote2014/eu-uk-results): BNP -5.1%, English Democrats -1.1%, Con -3.8%, UKIP +11.0%.

    BBC editorial conclusion: UKIP taking a lot of votes off Labour as well as Tories. Okaaaaay….

  3. Hopefully with this, and all the previous polling successes, journalists and politicians won’t, as they did in the US, ignore the polling data – whatever it says.

    Republican commentators completely ignored the polls, deciding that Obama’s second term was impossible despite what the polls said.

    We’ll have to see who makes the same mistakes in 2015, whatever they point to.

  4. Good Afternoon All.
    Probably lots of ‘churning’ of the votes.

    UKIP picking up putative Labour voters, though, GORDONING.

  5. “At a purely personal level though, getting UKIP right at the next election is the biggest challenge currently facing pollsters, so I’m relieved that in the first real proper national test we got it right. Phew!”

    Anthony, what was the average UKIP figure from the last few YouGov polls, though? Weren’t some of them a bit higher, or was that just the other polls? Maybe wishful thinking but I wonder if there was a bit of a case of cold feet among some prospective UKIP voters at the polling station.

  6. AW

    “At a purely personal level though, getting UKIP right at the next election is the biggest challenge currently facing pollsters, so I’m relieved that in the first real proper national test we got it right. Phew!”

    Yes, well done YouGov!

  7. Apparently Clegg is hiding out in Chevening House today. Must be nice for him.

  8. Guardian reporting Press Association on final results:

    Ukip

    24 seats

    27.49% (up 10.99)

    Labour

    20 seats

    25.4% (up 9.67)

    Conservatives

    19 seats

    23.93% (down 3.8)

    Greens

    3 seats

    7.87% (down 0.75)

    Lib Dems

    1 seat

    6.87% (down 6.87)

    SNP

    2 seats

    2.46% (up 0.34%)

    BNP

    0 seats

    1.14% (down -5.1)

    English Democrats

    0 seats

    0.8% (down 1.05)

    Plaid Cymru

    1 seat

    0.71% (0.13)

  9. Well done Anthony, you are to modest.

    A pity certain politicians who have had a very poor couple of days, don’t have a bit of your humility.

  10. Also, I wonder why YouGov overestimated the Greens, but the other pollsters all consistently underestimated them. I guess it’s hard to get small parties right though.

  11. @ Anthony

    Well done YG – except you should’ve listened to Chris Lane, you did have the LibDems too high. ;-)

  12. AW
    see PT for my congratulations. You always change it just after I post, but I am sure you read all 1100 odd, just for moderation reasons (hee hee)..

  13. @CHRISLANE45: “UKIP picking up putative Labour voters, though”

    I’m not sure what you mean by “putative”: if you just mean *potential* Labour voters, then I agree with you. Most of these people probably would have voted for Thatcher, and many would have voted for Blair. The question is, how far does Labour have to go to the right to appeal to these people? And does it want to go that far?

    I don’t think we should say that *everyone* in a low-to-median-income area “should” be voting Labour. Some will be right-leaning, others left-leaning. The proportions will just be different than in more affluent areas.

  14. Gordoning – Don’t know about the Greens. We’ve had problems before prompting for them and overestimating them.

    Penultimate YG polls were showing slightly lower UKIP, but that was because we only applied the turnout weight to the final poll. Not sure if they’d have fallen using consistent turnout approach. Other factor was An Independence for Europe – in the final poll we prompted using the full candidate list in each region, which correctly got about 1% for AIFE, and presumably took some of that from UKIP.

  15. @ Anthony

    BTW, much of the table is black text on a black background. I copied it into a spread-sheet so that I could read it. Would you be able to change the table if others are having the same issue?

  16. Congrats Anthony.

    Interesting table.:-

    5 of 6 pollsters underestimated Cons-only ICM overestimated them. Survation were closest.

    5 of 6 pollsters overestimated Lab-only Opinium underestimated them. YG & Opinium were closest

    Apart from YG-all other pollsters had big errors on UKIP

  17. No doubt there is give and take of voters between the different parties. However, I would have thought that UKIP was recruiting voters whom Labour might have hoped to attract (i,e, they are relatively poor and suffering from the kinds of things that labour’s policies are supposed to address)

    Why do these voters not go to Labour? My guess is that they don’t trust any politicians, they don’t think labour is competent to deliver, they don’t feel that labour really cares about their issues, and they do think that things aren’t what they used to be.

    Presumably there is evidence for and against some of this but I don’t know what it is.

  18. What I mean is, yes we could ask:
    (i) Why are these people in mainly whiteouter-conurbation low-to-median-income housing estates not voting Labour when they would probably have voted for Blair;
    but we could equally well ask:
    (ii) Why are they not voting Conservative when they would even more probably have voted for Thatcher.

    Seems to me (and a lot of other people) that the media is doing (i) much more than (ii) at the moment. Yet in the medium-to-long-term, (ii) is a far bigger problem for the Tories than (i) is for Labour. If Tories are not the biggest party after the next election then where exactly is their recovery going to come from? I reckon they will have to massively swing to the centre to attract the residual support of the former Lib Dems, now that the latter arelikely to be annihilated over the next decade or two. Not much room for growth on the right for the Tories.

  19. I did predict Labour to be ahead of UKIP by 2-3% but that was before their campaign started. I’m thinking that it did not impress voters. But maybe this sort of result was coming anyway. Labour have to sharpen up their campaign for the 2015 GE if they want that majority.

    Apparent results of the 1st count in Northern Ireland expected at 2pm. As for Ireland itself, full results expected tommorow, the Midlands-Northwest constituency is having some trouble with invalid ballots and also very close counts (ie, less than 10 votes in it).

  20. Well at least Cameron and Miliband are acknowledging that voters right across Europe want the Euopean Union to work differently. No point insulting UKIP voters or saying it’s just a protest, we can see right across Europe that people have serious concerns with some of the aspects of Federal Europe and the lack of transparency in areas.

    Where as the council result for UKIP was well over done by the media, this is a stunning result really. 23 MEPs to the Liberals 1, wow!

  21. @ Jack Sheldon
    Thanks for the Barnet data. Remarkable that Labour did so relatively well in the Chipping Barnet wards & so much less well in the Finchley ones.
    You didn’t do Hendon because of the non-elections in Colindale: however, if one estimates the latter’s 2014 results on the 2010 data, assuming a 35% turnout & 5% swing from Tory to Lab, then Labour emerge with a fairly small percentage advantage over Tories in Hendon.
    However, the complications of differential turnout, UKIP, local v general factors, potential for further Lib-Dem collapse, etc makes one realise how that any extrapolations from 2014 local, to 2015 general, election results are fraught beyond belief!

  22. Sky News: Nick Clegg to speak at 2pm.

  23. Actually, is it zero MEPs for Liberals? I thought they were heading for 1?

  24. Rich – Catherine Bearder held on in the South East

  25. Amber – I’ve taken away the shading, does that help?

  26. @Charles

    It may be that many who in past days would have voted Labour are still feeling pretty numb after the financial crash which occurred under Labour. “Labour had thirteen years in power and look where it got us” is what some may well be saying.

  27. Thanks. And I see UKIP is 24, so I was wrong there. Just catching up! I honestly can’t get head round UKIP with 24 MEPs and the pro European Liberals with 1. You just can’t brush this aside as a protest vote or disaffection, surely?

  28. @John B

    I think that’s it. Feelings are still very raw.

    EM has to communicate his working class friendly policies better.

  29. Quite a narrow squeak for LD’s to get that MEP, more due to having 10 seats on offer in the South East. Anywhere else I don’t think they would have got a seat. They got a higher share of the vote in the south west, but lack of available seats kept them out. D’Hondt does make getting seats difficult for parties with that level of support.

  30. I guess that about 1% of the 1.5% average vote for “An Independence from Europe – UK Independence Now” was intended to be cast for UKIP, and that pollsters couldn’t have been expected to pick up that error.

    Adding 1% to the UKIP vote to reflect voters intentions would add 0.2% to the average error for YouGov and ICM, and deduct 0.2% for the rest.

    Even that would still leave YouGov in front, so congratulations are definitely in order for AW.

  31. I thought there was an ICM phone poll for The Guardian that had European Election voting intentions?

  32. John B

    “Labour had thirteen years in power and look where it got us” is what some may well be saying.”

    I’m sure that’s right.

  33. Well done Anthony.

    Extrapolations to the GE are pretty pointless and angels and pins sprung to mind [the answer is three by the way] with all the yah boo stuff on the previous page.

    The point something % differences are really not worth arguing over.

    My feeling remains the same and has nothing to do with recent events: the Tories still have too many people who will never vote for them to get to the threshold they need to form a government on their own. The very thought will worry too many.

    Whilst the same is true – obviously – of the Labour party the difference is that they are a more broadly national party and have the system working in their favour.

  34. An interesting analysis of the common EU identity & background of UKIP type voters.

    http://www.economist.com/blogs/blighty/2014/05/eu-election-britain

  35. Dave – yes, there was, I should change the original article. As with Populus though, the ICM phone poll was too early to really compare fairly to the final result.

  36. Phil Haines – FWIW, amongst the long tail of others our final call poll did have AIFE on 1%!

  37. Talking of extrapolating EU elections to GE elections, I had a go at it, basing it on three EU elections. As you might notice, there was an GE 1-2 years later so you’d assume there’d be some link between the two. So with yesterday’s results input, the prediction is

    Lab 39%
    Con 27%
    LD 12%
    UKIP 11%

    The bottom two have a decent chance of being about right, but for the two main parties – no, I don’t think so. If there is link between EU and GE vote shares, it’s fairly weak.

  38. @AW

    Yes, taking AIFE into account it does look remarkably accurate. Congratulations!

  39. R&D

    Aye, fair points. And the distribution of seats issue must include the fact that the ‘popular’ vote is not often reflected above 75% in terms of seats. Both Tories and Labour pile up votes in their strong areas without winning more seats in a proportional way. Predominantly, the Tories will continue to be the party of the south of England, and Labour the north of England (though obviously there are enclaves for both in the other’s ‘territory’.

    But this, surely, puts the spotlight on London – which appears to have been the real decider between the two. London may determine who wins the GE, although UKIP’s effect on Tories and Labour in their respective heartlands will be a difficult one to calculate in advance.

    And, although hardly something AW was holding his breath for, my belated congrats to YG on a good call!

  40. Colin

    Very interesting – I haven’t really yet had the time to look over the European results (except the big French result for the NF).

    I did read an amusing headline elsewhere which effectively read ‘Centre-right EPP remains in first place, 15 years on’, in response to the headlines about the rise of Euroscepticism.

  41. John B – It may well be as you say. It wouldn’t be fair in my opinion but after four years of ‘the mess Labour left us’ it would be surprising if it were otherwise.

  42. Sky News has corrected their UK results.

  43. Kay Burley: We have actually tracked down Nick Clegg and we’ll be speaking to him shortly.

  44. Charles,

    Blaming the foreigners is an attractive narrative for some and they fell for it. Never mind that the economic problems are not the foreigners’ fault. I don’t think UKIP’s “offer” is going to stand the heat of a general election, though.

    Another point worth noting is that half of the UKIP advance (compared to Euro’ 2009) is accounted for by the collapse of the BNP. There’s no way Labour is going after these votes.

  45. Anybody watching Clegg right now on Sky? He looks awful. Like a man in total shock.

  46. Nick repeating his mantra. Same old.

  47. @Rich

    The wife has just agreed with you.

  48. He’s done a Blair and aged thirty years in four.

  49. @ROSIEANDDASIE:

    “The point something % differences are really not worth arguing over.”

    My point in posting those was to criticise the media line that all these potential Labour voters have suddenly been “lost” to UKIP. No, actually about half of them were already “lost” to the BNP etc. in 2009 – around the time of the expenses scandal. What’s new is the consolidation of the far-right vote by UKIP: it doesn’t seem that they have won many new “converts” in ideological terms.

    And I doubt that Ed Miliband’s leadership has much to do with this, as Nick Robinson et al. would have us believe. Sure under Alan Johnson or someone else these voters could imagine having a pint with, they might have been able to claw back 1 or 2%. But most of this base surely has a deep dissatisfaction with the main parties in general – as, to be fair, they have been telling us themselves!

  50. Is it just me, or did Kay Burley say – “here’s Nick Knight(!)”

1 2 3 9