en.wikipedia.org/wiki/Opinion_polling_for_the_United_Kingdom_general_election,_2017
List of the opinion polls for the GE.
There is something striking for the last week or so (since break for the bombing). Here are the Cons:
YouGov/Times 42%, Survey Monkey 44%, Kantar 43%, YouGov Model 41%, ICM/Guardian 45%, YouGov Model 42%, Survation 43%, YouGov/Times 43%, ICM/Sun 46%, ComRes 46%, Orb/Telegraph 44%, SurveyMonkey 44%, YouGov/Times 43%, Opinium/Observer 45%, Panelbase/The Sunday Times 48%.
UKIP get 4% in just about everyone.
There isn't a lot of movement here.
Its just the balance between the LDs and Labour that is changing.
Here is the Lab and LD breakdowns:
YouGov/Times 39%/7% (46%), Survey Monkey 38%/6% (44%), Kantar 33%/11% (44%), YouGov Model 38%/9% (47%), ICM/Guardian 33%/8% (41%), YouGov Model 38%/9% (47%), Survation 37%/8% (45%), YouGov/Times 36%/9% (45%), ICM/Sun 32%/8% (40%), ComRes 34%/8% (42%), Orb/Telegraph 38%/7% (45%), SurveyMonkey 36%/6% (42%), YouGov/Times 38%/10% (48%), Opinium/Observer 35%/7% (42%), Panelbase/The Sunday Times 33%/7% (40%).
That is something that is important. Polls are going to have great difficulty representing this if Lab / LD thing is pretty much localised by constituency. They do so many representatives per region instead. If they happen to pick more people in more LD seats, then what happens? I don't think they adjust for this in polls. But I would suggest they would in a model based on polling.
This is where if a lot of people do decide to tactically vote its going to have an impact; and this is where a model might pick it up but not a poll. If I'm right and this really is shaping up to be an 'anti-May' campaign rather than a vote for Lab / LD, then something which looks like a volatile LD/Lab vote would appear in the figures. Especially if voters really were being volatile in behaviour. I'd also say that there looks to have been a 'squeeze' on the LD vote. Again this probably would fit with this pattern of an anti-May vote, with the LDs simply being the smaller party and the campaign being modelled more as a Corbyn v May battle.
The YouGov/Times figures are interesting in respect of that:
30/31 May - Con 42% Lab 39% LD 7%
25/26 May - Con 43% Lab 36% LD 9%
24/25 May - Con 43% Lab 38% LD 10%
Compare with the YouGov Model
23 - 29 May - Con 42% Lab 38% LD 9%
24 - 30 May - Con 41% Lab 38% LD 9%
The detail of the YouGov/Times polling seems to show a slight up tick in the transfer of vote from the Cons to both the LDs and Lab in the last few they have done. That split seems a little all over the place though.
I think that's probably where they are headed with the logic. Whether that's right remains to be seen. I think I'd go so far as to say, that's its a really important consideration and its not being shown in the polls, but whether its effect is significant enough I'm yet to be convinced.
The Lab/LD split would need to be quite a bit higher than the Con vote alone for a hung parliament to make up for the problem of the split vote. (Remember the Con share also covers Scotland and the Lab/LD share of the vote there is a lot lower). The SNP would need a good showing. How UKIP affects things with where they are standing is also important. (again something that perhaps wouldn't show well in polls but might be represented in model).
A 'lock down' on immigration and socially conservative is a pitch for those SE voters and a pitch for those Leave Labour. It might backfire in remainy Con seats regardless of whether Lab or LD is second. But there is more seats to be won in the SE and with Labour leave.
Interestingly, that's what my Constituency is. And that's exactly what YouGov's model is seemingly suggesting. All these are Remainy Con: Hastings and Rye -> Lean Labour. Oxford West and Abingdon -> Tossup. Lewes -> Tossup. Twickenham -> Likely LD. Richmond Park -> Tossup Bath -> Tossup. This is where the YouGov poll might fall down. Its tipping these LD I think on its count up. They are so finely balanced.
Of course Lord Ashcroft has every single one of these as a Con Win (including mine). Some quite comfortably.
The Cons have been trying to scare liberal Cons with Corbyn. Yet have decided to stop doing this apparently, because its not working (or perhaps is having an adverse affect). If there is a strategy shift to Brexit in the last week will this help or hinder with Remainy Cons? Or do they really think they have them in the bag?
And hello, what's this?
(((Dan Hodges)))*@DPJHodges* (Mail on Sunday)
Being told LDs will definitely lose seats overall. Farron now in real fight to save his seat. If they can't come back now, when will they?
Its not a case of that though. Its not that people don't like the LDs. Its Brexit plus May v Corbyn. Its a polarisation and collapse of the centre as people have been told its one or the other. The centre still exists, but its voice is not being heard nor can it get it out.