UK General Election - May 7th 2015 (The Official Campaign Thread)

[YL]

More Ashcroft polls.

Bristol West: Labour 13 points ahead of Green, 18 ahead of incumbent Lib Dem
Bristol NW: Tories 9 points ahead of Labour
Rochester & Strood: Tories 3 points ahead of UKIP
Thurrock: UKIP 4 points ahead of Lab, 5 ahead of Con
High Peak: Con 2 ahead of Lab
Colne Valley: Con 2 ahead of Lab

All margin of error stuff except the two Bristol seats, where West is horrible for the Lib Dems (that's a 19% swing from LD to Lab) and NW disappointing for Labour.

Caveats as usual.

[YL]

I'm just imagining a conversation going something like this:

"Which football team did we decide he was going to support when we were trying to make him a man of the people?"

"Wasn't it that one who play in claret and blue?"

"West Ham?  Yeah, that's it."

[YL]

Is there any chance the Tories and Lib Dems will try to continue their Coalition government or would they want to avoid such a scenario again?

My assumption is that if they have a majority together (and the Tories haven't done well enough to go on their own) they will continue it.  Andrew George, the Lib Dem defending St Ives, recently said it wouldn't happen, but I expect that the voices in the Lib Dems in favour would outnumber those against.

An "interesting" scenario is if the current coalition loses its majority, but Lab+SNP+Plaid+Green+Respect+SDLP+Hermon don't get one either.  The current electionforecast.co.uk prediction is very close to that: it gives Lab+SNP+Plaid+Greens+SDLP exactly 323 between them, and Con+LD 311.

NB (a) 323 is a majority in practice unless Sinn Féin lose Fermanagh & South Tyrone, in which case it's 324.  (b) The Speaker should be counted as a Tory, even though he isn't one, because of the way the Deputy Speakerships work.

[YL]

Cashcroft's national poll is as bouncy as hell.

It's not.

First off, in a close race opinion polls should be bouncy. Secondly, Ashcroft's polls have generally shown Conservative leads since January (with two exceptions). Polling companies have different trends (and I know you know this!)

ICM has shown Tory leads since January.

YouGov (who are experimenting with calling back the same panel) has shown a consistent Labour lead since their methodology change in early April.

ComRes has shown a Labour lead to March and a Tory lead since then.

Opinium has shown the same.

MORI has shown a consistent Labour lead

Populus has shown the same.

The thing about all the phone polls (including both Ashcroft and ICM) is that their headline sample sizes of around 1000 get substantially reduced (often to around 500) by excluding don't knows, won't says and won't votes.  This increases the margin of error, and I don't think they're any more volatile than you'd expect given that.

Re ComRes, the difference is between online ComRes (Labour leads, no poll recently) and phone ComRes (mostly Tory leads, with one tie).  MORI are the only phone pollster with Labour leads recently.

Opinium have changed their methodology more than once, and I think the change to Tory leads may be associated with that.

My feeling is that there has been very little movement this year, and the differences are mostly down to methodology and sampling variation.  Obviously I hope YouGov are right; Tories can hope that ICM and Ashcroft (national) are right.

[YL]

BTW Ashcroft has "Others" at 5% (BNP 1%, other others 4%) in Castle Point, although there are no candidates there other than Con/Lab/UKIP/LD/Green.

[YL]

There's a possible void election in Hull East. There's photographic evidence apparently of the postal vote ballot papers cutting off the Labour and Green candidates names.

This is the picture tweeted by BBC Radio Humberside:

[YL]

Red racing horses has a seat by seat analysis of the election prospects for all the constituencies outside of England, including a bunch of pretty maps. You guys might enjoy evaluating how much you agree with its analysis. While the site is partisan, I find the hosts' political handicapping pretty "fair and balanced."

As far as the predictions go, I didn't notice anything obviously ridiculous, but not anything obviously original either.

However:

You must be logged in to read this quote.

Firstly, LOL at "the actual city of Strangford".  Secondly, the village of Strangford has always been in South Down; the constituency name refers to Strangford Lough, most of the shoreline of which (except for the bit around Strangford village) is in the constituency.

[YL]

More Ashcrofts.  All have been polled before, and all changes in leads are tiny (as are the leads themselves).  I knew one of these was coming.

South Swindon: Con 1 ahead of Lab
South Thanet: Con 2 ahead of UKIP, 8 ahead of Lab
Sheffield Hallam: Lab 1 ahead of Lib Dem

[YL]

My phone's caller ID and a quick Google reveal that Survation tried to call me three times today.

[YL]

Final round of Ashcroft constituency polling:

Battersea: Con 12 ahead of Lab
Croydon Central: Con 4 ahead of Lab
Norwich North: Lab 2 ahead of Con
Peterborough: Lab 2 ahead of Con
Pudsey: Con 1 ahead of Lab
Stourbridge: Con 2 ahead of Lab
Wirral West: Lab 3 ahead of Con

North Cornwall: LD 2 ahead of Con

Dumfriesshire et al: SNP 11 ahead of Con
East Renfrewshire: SNP 3 ahead of Lab

Take with as much salt as you want, but there's something for everyone there.

(And also remember that even if you trust the sampling, there's still a margin of error, as conventionally defined, of about 6 percentage points for the lead.)

[YL]

Full figures for the ICM Hallam poll: Clegg (LD) 42, Coppard (Lab) 35, Walker (Con) 12, Jenkins (UKIP) 7, Garbutt (Green) 3, Other 2.

There are a number of possible reasons for the differences from the Ashcroft poll (which are, as the headline suggests, maybe Con down, LD up).  As well as the naming of candidates (the Lib Dems' favoured reason) note that the sample size is only 500, and that ICM's turnout and don't know adjustments seem to have made more difference than Ashcroft's (it was a tie without them), and who knows who was getting a better constituency sample.

ICM does, unfortunately, suggest that Clegg may be getting enough Tories to vote for him to save him.  Tories on 12% in Hallam would have seemed incredible 20 years ago.

[YL]

The Sun really is a pathetic rag.

Speaking of which, the Daily Express has endorsed UKIP.

[YL]

ICM gives a tie: Lab 35 Con 35 UKIP 11 Lib Dem 9 Green 3

[YL]

Just finished my last round of deliveries. Convincing UKIP-leaners to vote Labour instead is a wonderful feeling. No idea exactly what's going to happen tomorrow.

Tomorrow I'm up from 7am for the next 22 bloody hours. Wish me luck.

Good luck, though I think "Labour gain Kensington" is far too wonderful a thing to actually happen.