UK General Election - May 7th 2015

[YL]

A prediction:

http://www.belfasttelegraph.co.uk/debateni/blogs/bill-white/projecting-the-result-of-the-may-2015-general-election-30824967.html

I'm a bit doubtful that the SNP will do quite as well as 45 seats, but otherwise that looks quite plausible.

electionforecast.co.uk currently has a best guess of Lab 283, Con 280, SNP 34, Lib Dem 27, DUP 8, SF 5, UKIP 3, Plaid 3, SDLP 3, Green 1, Other 3 (Bercow and presumably Hermon and Long).

Stephen Fisher's model currently has a slightly less dysfunctional Lab 299, Con 291, Lib Dem 29, Others 31 as its best guess (though I think he's not handling Scotland separately).

Martin Baxter's Electoral Calculus (based on current polls) has Lab 316, Con 258, "Nat" (i.e. SNP and Plaid) 45, Lib Dem 19, Others 19 (the Northern Ireland seats and Wyre Forest; he treats Bercow as a Tory).  He has a decent track record but I don't think his model handles UKIP (or minor parties in general) well.

[YL]

A YouGov Scottish poll gives Westminster voting intention as SNP 47 Lab 27 Con 16 LD 3 UKIP 3 Green 3.

It also gives independence referendum voting intentions of Yes 52 No 48.

[YL]

Kerry Smith, the UKIP candidate selected for South Basildon and East Thurrock, then pushed aside for Mr Brown Envelopes, then reselected has himself quit after the MoS obtained a recording of him joking about shooting peasants and being rude about UKIP's LGBT group.

There was also the other candidate they had in between the first time for Smith and the Hamilton story, the one who resigned after that story about their general secretary.  Quite an entertaining selection, this one; perhaps it'll end up with Hamilton after all...

[YL]

Here's this week's YouGov summary.

12 Dec: Lab 34 Con 32 UKIP 14 Green 7 LD 7 SNP/Plaid 5
14 Dec: Con 32 Lab 32 UKIP 16 Green 7 LD 7 SNP/Plaid 4
16 Dec: Lab 34 Con 32 UKIP 14 Green 8 LD 6 SNP/Plaid 5
17 Dec: Lab 34 Con 33 UKIP 16 LD 6 Green 6 SNP/Plaid 5
18 Dec: Con 33 Lab 33 UKIP 14 LD 8 Green 7 SNP/Plaid 4

Lib Dems may prefer to look at ICM, which puts them in a tie for third rather than fourth: Lab 33 Con 28 UKIP 14 LD 14 Green 5.  (There's no headline figure for SNP/Plaid, but it looks like it must be 5 or 6.)

Why the difference?  ICM reallocate 50% of don't knows back to their 2010 party, which boosts the Lib Dems from 11 to 14.  That 11 is already high by YouGov standards, though, so there's more to it than that.  ICM's effective sample size is quite small, so some of it may be sampling variation...

[YL]

Last two YouGov polls of 2014:
19 Dec: Lab 35 Con 30 UKIP 16 Green 8 LD 6 SNP/Plaid 4
21 Dec: Lab 34 Con 32 UKIP 15 Green 8 LD 6 SNP/Plaid 5

In other polls, MORI still has the Tories ahead and has a very high Green figure:
Con 32 Lab 29 UKIP 13 Green 9 LD 9 SNP 6.

Meanwhile TNS-BMRB and Opinium have both come out with 7 point Labour leads; the latter's figures are
Lab 36 Con 29 UKIP 16 LD 6 Green 5
(no SNP figure in the Observer article).

[YL]

Here's another prediction site (take with as much salt as you feel appropriate):
http://www.thehustings.co.uk/

They're predicting Clegg to lose his seat...

[YL]

I tend to think that a noisy Labour campaign would be counterproductive. Of course activists and resources won't be directed there anyway; as fun as trolling can be, this is a seat where Labour only polled 16% in 2010. If a freak win happens a freak win happens, but its not worth trying for.

I'm not close enough to either party's local leadership to know what they think is really going on, but my impression is that the local Labour party are trying to win, and think that they can, while the local Lib Dems are worried.  Perhaps that is just because of the Ashcroft poll, which I don't entirely trust, but you have to remember that that 16% was the result of twenty years of a tactical squeeze which the beneficiaries of have pretty explicitly abandoned.

It is still of course a very middle class constituency, which doesn't make it easy for Labour, though there's middle class and middle class...

[YL]

What constituencies are the UKers here in? Mine (the one I'll be voting in) is Bath. Sources tell me that the Tories are confident. Can't see it; the Lib Dem share will fall more sharply than in most of their seats in which the Tories are second due to A. Foster hugely increasing his share in 2010 and B. Student backlash, but it'll still be a 3-5,000 hold imo. The Labour candidate is a 20ish year-old student btw. Will be interesting to see how that plays out.

I'm in Sheffield Hallam (if that wasn't already obvious).

Bath strikes me as a tough nut for the Tories on current boundaries, although with Foster retiring they might see an opening.  Some outlying villages got transferred to Jacob Rees-Mogg's seat in 2010, which helped the Lib Dems in Bath while harming Labour there.

[YL]

The Times has a front page story this morning about Labour's "secret" plot to win Sheffield Hallam, and an editorial, endorsing Clegg, entitled "Clegg in Peril" and claiming that CrabCake and I will be voting in the "most important constituency to be contested in May’s general election".  It's nice to have my vote valued so much, not that I'll be using it in the way they want me to...

[YL]

First YouGov of the year: 34/31/14/7/8

Also SNP/Plaid 4.

No sign of anything very different from before Christmas.

You must be logged in to read this quote.

I'm not convinced it was ever going to be any different.  None of the recent election dates have exactly been a surprise this far out (except for the month's delay in 2001 because of foot and mouth disease).

IMO it was right to move to Parliament needing to vote to dissolve itself rather than it just being the PM's choice, but I don't agree with some of the other details, and if we were going to go for fixed terms it should have been 4 years (as for all other fixed terms in the UK except MEPs' ones) rather than 5.

[YL]

This week's YouGov polls:

6 Jan: Lab 34 Con 31 UKIP 14 Green 8 LD 7 SNP/Plaid 4
7 Jan: Con 33 Lab 33 UKIP 13 Green 8 LD 7 SNP/Plaid 4
8 Jan: Lab 33 Con 32 UKIP 15 Green 7 LD 7 SNP/Plaid 4
9 Jan: Lab 33 Con 33 UKIP 13 LD 8 Green 7 SNP/Plaid 5
11 Jan: Con 32 Lab 32 UKIP 18 LD 7 Green 6 SNP/Plaid 4

[YL]

Helen Pidd visits Sheffield Hallam: http://www.theguardian.com/politics/2015/jan/11/nick-clegg-sheffield-hallam-general-election

(BTW I'm sure there are more students in Fulwood ward, where there are two University student villages, than there are in Ecclesall.  But I suppose I shouldn't expect a journalist to get that sort of thing right.)

[YL]

It scarcely matters who Labour (or anyone else other than the SNP) put up in Gordon, anyway.

[YL]

Why are 2001 and 2005 omitted from that graph?

[YL]

Why are 2001 and 2005 omitted from that graph?

Because it's to do with Conservative governments and Labour was in power then. Labour's 2010 'swingback' is there for reference in red.

Not so.  I've just found the link and the justification for excluding 2001 and 2005 is that the government wasn't trailing.  (Which is pretty dodgy: if you exclude those two for having a government lead there are serious questions about cases like 1997 where there was an enormous opposition lead.)

[YL]

LOL

Is it time to raise the deposit to £1000? I'm starting to think so.

The Electoral Commission have actually just recommended abolishing it altogether...

I don't see any evidence of a problem.  If the Haltemprice & Howden by-election circus were typical of parliamentary elections in the UK, then I might support making it harder to stand (though I'd probably prefer increasing the number of signatures required to raising the deposit) but it isn't.  There's nothing wrong with the odd obvious joke candidate.

[YL]

Don't forget the Respect party!

[Snide Voice]

Why, everybody else has

[/Snide Voice]

Ah, but have the voters of Bradford West?

[YL]

This week's YouGov polls:

13 Jan: Lab 33 Con 32 UKIP 17 Green 6 LD 6 SNP/Plaid 4
14 Jan: Lab 33 Con 32 UKIP 14 Green 7 LD 7 SNP/Plaid 5
15 Jan: Lab 34 Con 32 UKIP 15 Green 7 LD 6 SNP/Plaid 4
16 Jan: Lab 32 Con 32 UKIP 16 Green 8 LD 6 SNP/Plaid 5
18 Jan (Sunday Times): Lab 32 Con 31 UKIP 18 LD 7 Green 7 SNP/Plaid 3
18 Jan (Sun on Sunday): Lab 33 Con 31 UKIP 16 LD 7 Green 7 SNP/Plaid 5

[YL]

SNP at 52% in Scotland with Ipsos Mori

What has been their best polling so far?

Their previous best was also 52%, also in a Mori poll, back in October:
http://en.wikipedia.org/wiki/Opinion_polling_for_the_2015_United_Kingdom_general_election#Scotland

[YL]

Well Cameron played a blinder on the debates

Depends who you believe.  If you believe the stories that he didn't really want them at all, then his bluff has rather been called.

[YL]

This week's YouGov polls, including the one the Sun hyped:

20 Jan: Lab 33 Con 32 UKIP 15 LD 8 Green 7 SNP/Plaid 5
21 Jan: Con 32 Lab 30 UKIP 15 Green 10 LD 8 SNP/Plaid 4
22 Jan: Lab 34 Con 33 UKIP 14 Green 8 LD 6 SNP/Plaid 5
23 Jan: Lab 33 Con 31 UKIP 17 Green 8 LD 7 SNP/Plaid 4
25 Jan: Lab 32 Con 32 UKIP 15 LD 7 Green 7 SNP/Plaid 5

[YL]

2. There aren't many marginal seats in Wales (hey on paper even Cardiff Central isn't). From a Labour perspective (and ignoring the Cardiff marginals) there's a low Plaid majority in Arfon and a comparatively low one in Carmarthen East & Dinefwr, there are the two Pembrokeshire seats (which may be harder to gain in reality than on paper), there's the Vale of Glamorgan (which is a strange socially polarised constituency), and I guess there's Aberconwy, but I'll believe that when I see it. Incumbency can matter a lot in all of these seats. Here's the fun part though; most of Arfon hasn't had a Labour MP since the 1970s, Carmarthen East & Dinefwr was lost in 2001, and one of the Pembrokes was lost in 2005 (as was Cardiff Central).

Aberconwy and the two Pembrokeshire seats are all dominated in electorate, if not area, terms by their more Anglicised areas, aren't they?  (Which makes "Preseli Pembrokeshire" a rather odd choice of name, but there we go.)

Does anyone have a feel for why Ceredigion swung so much to the Lib Dems in 2010?  If I had a feel for that I might feel more comfortable guessing whether or not it'll swing back to Plaid...

[YL]

Survation poll:
CON - 31% (+2% on December poll)
LAB - 30% (-2%)
UKIP - 23% (+3%)
LD - 7% (-4%)
SNP - 5% (+2%)
GRN - 3% (+1%)
OTH - 1% (N/C)

May2015 says with those results:
CON - 283 seats
LAB - 252 seats
SNP - 50 seats
LIB - 25 seats
UKIP - 17 seats
GRN - 1 seat
Others - 4 seats
NI - 18 seats

Depressing.

How can a virtual tie in the popular vote give the Tories a clear plurality of seats? That seems impossible, when considering Labour's recent large advantages in the vote/seats distribution (even when taking a possible Labour disaster in Scotland into account).

As recently as 2005, the Tories received (barely) more votes than Labour in England, yet received only 194 seats there, compared to Labour's 286. I'm probably missing something here, but I really can't understand May2015's model.

I wouldn't necessarily endorse May2015's model, but a Labour disaster in Scotland really does have the potential to make a mess of their advantage from the way votes are distributed.  It'll probably take at most 2 percentage points or so off their UK vote share, but at worst it could cost them nearly 40 seats, which is quite a bit more than a similarly sized uniform national swing.

For comparison electionforecast.co.uk (in its central forecast) currently gives a 283-283 tie between Labour and the Tories with the Tories having a 1.9% popular vote lead, but it only gives the SNP 33 seats, with Labour holding onto 23 in Scotland.

The way the very high UKIP figure (a Survation speciality) is being treated might possibly be hurting Labour too, but I haven't investigated how the model treats UKIP.

[YL]

This week's YouGov polls:

27 Jan: Con 34 Lab 33 UKIP 15 Green 7 LD 6 SNP/Plaid 4
28 Jan: Con 34 Lab 33 UKIP 14 Green 7 LD 7 SNP/Plaid 5
29 Jan: Lab 33 Con 33 UKIP 16 Green 7 LD 6 SNP/Plaid 4
30 Jan: Lab 34 Con 34 UKIP 14 Green 7 LD 6 SNP/Plaid 4
1 Feb: Lab 35 Con 32 UKIP 15 LD 7 Green 6 SNP/Plaid 5

[YL]

Ashcroft is finally going to release his Scottish constituency polls tomorrow, and is ramping them on Twitter.

FWIW (possibly not very much) in the YouGov poll the SNP voters are noticeably more confident that they will actually vote that way than the unionist parties' voters.  (The Lib Dem sample is tiny, of course, so is even less informative than the others.)