Categories
politicalbetting.com » Blog Archive » Great news for IN from ICM phone survey and for OUT from IC
politicalbetting.com » Blog Archive » Great news for IN from ICM phone survey and for OUT from ICM online poll
ICM's two referendum polls after excluding DKs
PHONE IN 55% OUT 45%
ONLINE IN 48% OUT 52%
0
This discussion has been closed.
Comments
Pro-cake, pro-eating it; pro-polling, pro-being right after the event whatever happens.
Is it an effect caused by more politically engaged online panels or by the relatively high non-response rate for telephone surveys? I guess we'll find out in a month or so.
FPT: It looks MoE to me, which is interesting by itself given the amount of eggs and chips being flung around.
I stand by what I said over the weekend - a true Remain lead of 6-8%.
I'd still my view that what the mass of floating voters in the U.K. want is the closest possible Remain result - ideally, 51-49% - and will try and conspire to get such a result, which may will miraculously materialise through psephological osmosis.
Beppe Grillo said he couldn't wait to see the new Muslim mayor of London 'blow himself up at Westminster' during a televised event in Padua on Saturday.
http://www.dailymail.co.uk/news/article-3592778/Italian-politician-says-t-wait-new-Muslim-mayor-London-blows-Westminster-following-Sadiq-Khan-s-election-victory.html
Hilarious.....
Prior to May 2015 every pollster wanted to speak to me, but since then, nada.
https://www.icmunlimited.com/wp-content/uploads/2016/05/2016_guard_may_mode_test.pdf
Unless he was double-bluffing..
Broken, sleazy Leave on the slide.
Now everyone can selectively quote me after the fact
http://www.theguardian.com/politics/blog/live/2016/may/16/trump-complains-sadiq-khan-has-been-very-nasty-about-him-politics-live
But here @DanHannanMEP backs EFTA model - with "free movement..of people." https://t.co/nJHguI47gb
Or astrology...
Con (-3.8) Lab (+3.8) LD (+0.9) UKIP (-1.9) Grn (+0.2) Oth (-0.7)
If they repeat that right/left error margin again....
How many Brexiteers are now wishing Farage was more prominent in the campaign than Boris?
Using the 10/10 certainty to vote in the Phone poll Remain's lead goes down to 7 points, in the Online poll Leave's lead goes up to 7 points.
All Boris said was that the business leaders and foreign visitors to Downing Street supporting Remain were reminiscent of hostage videos.
Boris being silly is par for the course.
https://twitter.com/BBCNews/status/732192698891804672
In the ICM Phone poll over 65's are in favour of Remain, in the ICM Online poll they are 2-1 in favour of Leave.
The entire difference is caused by that one category and it's consistent since the start.
No, can't see anyone taking offence at that. At all.
Brilliant Boris. Keep up the good work.
2/1 UKIP supporters infesting online polls
2/1 Phone poll respondents not wanting to express an opinion SJWs think is racism
5/1 Demographics of this are so different that they cant get a representative sample without about 10,000 responders
10/1 Phone pollsters have been overpolling bits of Bedford
25/1 Both polls are grossly wrong
100/1 Nige has put a virus on the online pollsters program
500/1 Brussels plot to confuse everyone.
[I think it's fine, not too fussed either way. Perplexed as to why people would actually think it controversial].
Doesn't quite work like that, particularly since the base of Remain support is from the Left. Centre-right voters are splitting 2:1 for Brexit and more Conservatives back Brexit than do not.
So I think the true picture may well lie between phone and online, but not quite as much in favour of phone polls as some Remainers seem to think.
It's something to do with the violent deaths. Maybe you can work it out from there.
after all the online poll got 2015 MORE wrong - and so should be more biased to the left???
https://www.youtube.com/watch?v=bADMad_jXBo
The ICM Online Poll has Lab 32, Con 33 for the GE.
Which one is likelier to be close to the truth, bearing in mind the results we saw last week - London identical to 2015, Labour routed in Scotland, treading water at best in Wales and net losers in England council seats?
Phone polls were nearer the mark in the GE, Online nearer the mark in the more recent London and Scotland election.
What is remarkable is the consistency of the difference.
Could it be
The more wealthy are out and therefore less likely to answer the phone - unlikely as the online polls had a bias to labour last time
The retired are more likely to answer the phone - unlikely as if that was the case the phone polls would give a brexit lead
Shy tory syndrome - people reluctant to admit voting tory/brexit on the phones but more willing to admit tory/brexit on the computer - you would expect the online to show closer result in 2015 if that was the case.
Lets take a look at 2015 more closely...
https://pbs.twimg.com/media/Cg4ONFBWMAAfrUz.jpg
and the findings the BPC inquiry
https://pbs.twimg.com/media/Cil0nsAWwAEfwIn.jpg
I remember that the result on Nojam was 70-30 for Leave.
On that basis I declared victory for the online polls, of course the Nojam poll was an online poll.
Every single one of the final phone polls understated the Kippers.
Political Polls @PpollingNumbers 3h3 hours ago
Utah General Election:
Trump 43% (+5 since Apr)
Clinton 30% (-8)
Undecided 26
What's the betting that they will find that phone is more accurate for some demographics.
And online will be more accurate with others? We'll be no further forward.
Regarding the previous ICM Phone Apr 12-15 - the table on page 13 for certain to vote shows LEAVE as 40 and not the 41 in the figures above. Was there some adjustment?
2016_guardian_april_ICM modetest.pdf
This is a stunt.
https://www.youtube.com/watch?v=EDeUzB12ln8
https://www.youtube.com/watch?v=q0Y3A-zib-Q
Most of the older folks I meet are "reluctant leavers" - admittedly I tend to associate with SE/metropolitan ABs. Obviously not representative.
UKIP/Con Right Wing Bloc May 2015 Final Phone Polls
Actual 50.6%
Ipsos 47% (-3.6%)
ICM 46% (-4.6%)
Ashcroft 44% (-6.6%)
Comres 47% (-3.6%)
I think there is also a category "Don't care very much but will vote". Previous polling shows that most people are not that interested in Europe or the EU, but they will have to make a binary choice on the 23rd June. Will "Don't care" translate into "Fed up" or "Stick with what we know"?
C 37.8, L31.2, LD8.1, UKIP 12.9 Grn 3.8, Rest 6.3
Last Polls
Populus (online - I think) 5-7 May 34/34/9/13
Survation (phone) 6 May 37/31/10/11
Surveymonkey (online) 30ap 6 may 34/28/7/13
Ashcroft (online I think) 5/6 May 33/33/10/11
Ipsosmori (Tel) 5/6 May 36/35/8/11
Yougov (online) 4/6 May 34/34/10/12
ICM (Phone) 3-6 May 34/35/9/11
Panebase (online) 1-6 May 31/33/8/16
Opinium (online) 35/34/8/12
Conservative/LAB - nearly all polls underestimated conservative and overestimated Labour but by nowhere near as much difference as the ref polls are splitting between online and phone
UKIP - Online polls were spot on or not far out (exept one), phone consisently underestimated UKIP by two percentage points (11 against 13).
LD- generally overestimated a little.
So we can conclude that whatever else they got wrong, online polls were on the mark with UKIP voters. This rather demolishes the meme that in the referendum polls leave is ahead because it is more vulnerable to the Kipper Tendency.
It is clear that the polls undersampled Tories and oversampled Labour, nearly all of them did but it was more noticable on online polls.
Reasons seem to be:
1) Lack of traditional conservative voters
2) Misjudgement of turnout
3) Made a demographic error.
Undoubtably the pollsters will have tried to correct their conservative demographic. This seems to have caused a split in results between online and phone.
There are two Conservative camps. Public school inners and Grammer school outers. Getting that balance right would be crucial and very difficult. As indeed would be working out turnout.
What you wouldn't expect though is the split to be reflected consistently across phone and online polls which have all independently adjusted since the election.
And the split is not minor, it is huge, far bigger than any difference between the polls at the 2015 election.
I dont know the answer but I wonder this.
I suspect more affluent Conservative voters are probably easier to get hold of on a landline phone with working class labour voters easier to get hold of on a mobile - which would make such Labour voters harder to phone poll.
I suspect the same factor is going to be apparent in this referendum and with the lower down the social scale you are the less likely you are to have a landline or be able to be got hold of on one this might explain why online polls show more brexiters. On the other hand retirees are more likely to be at home, and so phone pollable. They will be pro brexit but not as much as the WWC.
This leads to the question. How well are phone pollsters getting hold of the white working class versus online polls. Could this be the difference. Who has the accurate ear of white van man?
By the way, for anyone interested, I've left my job as Director of Policy for Cruelty Free International to do a more varied mixture of consultancy (partly for CFI but also a variety of others) and translation. This frees me a bit in political terms, since the former job was strictly non-partisan, even though I was free to do political stuff in my free time.
Went to a farm shop and saw canned turnips for the first time ever
Couldn't help but think of malc