UK General Election - May 7th 2015

[YL]

A lot of negativity on here towards UKIP. However their central policy (withdrawal from the EU as the EU is on an irrevocable course towards a federal superstate which is not in the UK's interests to be a part of) has widespread support among the British public.

I'm a strong supporter of Britain remaining in the EU, but my distaste for UKIP has rather little to do with that.  It has more to do with other reactionary tendencies in the party, especially the scapegoating of immigrants and the tendency towards climate change denialism.  See also "Poujade, Pierre".

[YL]

A Guardian/Observer article about Green ambitions.

The target list mentioned (other than Brighton Pavilion) is
Norwich South
Bristol West
St Ives
Sheffield Central
Liverpool Riverside
Oxford East
Solihull
Reading East
York Central
Holborn & St Pancras
Cambridge

I think they have a good chance of holding Pavilion.  They might have a chance in Bristol West if something goes wrong with the Labour challenge there, but I think it's more likely that they'll be squeezed, while I think they'll do well in Norwich South but catching Labour there is a tall order.

The others seem even longer shots.  They certainly have potential in Sheffield Central but are starting from a very low base having been squeezed in 2010.

[YL]

Again, 'a stereotypical LibDem voter' is very different today to what it was pre-Coalition.

They've lost a lot of their 'stereotypes' to Labour, the Greens, the SNP and Plaid it seems.

Its the regional breakdown thats interestig. Who typically votes LibDem in Sheffield, Cornwall, Highlands etc.

I think that even in Sheffield the Lib Dem vote has been quite heterogeneous.  Especially in council elections, they have sometimes been getting the votes of people who dislike Sheffield Labour for all sorts of reasons which might not even be very compatible with each other.  (UKIP are now filling this role more than the Lib Dems.)  In general elections, I think the Lib Dem vote contains/contained substantial amounts of following types of voter:
(a) the left-liberal middle class voters much as described by EPG;
(b) a sizable, mostly middle class, vote which doesn't like the "nastiness" associated with the Tories but thinks of Labour as not very good at running the economy/not very good at running the City Council/too dominated by the unions (delete as appropriate).  Think of Moderate Heroes, or the sort of people who were attracted by the SDP in the 1980s.  (On the other hand, a lot of the (a) and (c) voters who were old enough may have voted Labour in 1983; there are people I know who voted Labour in 1983 but Lib Dem in 2010, and I doubt they think their views changed that much.)  Nationally, this sort of voter mostly voted for Blair, but in west Sheffield they may not have done because of the Lib Dems' local strength and their dislike of the council.
(c) Tactical voters.  In the areas which have always been in Hallam this would mean people who were basically Labour supporters who voted Lib Dem because they thought Labour had no chance.  There were big campaigns to get Labour supporters to do this at least from 1992 onwards.  There's something of an overlap with (a): a lot of Hallam Labour supporters would be quite small-l liberal (in the British sense, not the Venstre sense) anyway.  (These days, the Lib Dems and their bar charts have switched to trying to get Tories and Kippers to vote for them to stop Labour.  In Sheffield Central, where it was Labour they were fighting, I think they did that already in 2010.)

NB Hallam and the stronger Lib Dem parts of both Central and Heeley constituencies are very much the middle class side of Sheffield: the Lib Dems have never done well in general elections in the strongly working class parts of the city, though they have sometimes won council seats there.

Addendum: I probably should have mentioned voters (probably mostly ex-Tories, looking at the figures) who voted Lib Dem in 2010 out of enthusiasm for Clegg...

[YL]

East Midlands

Labour top in Ashfield, Bassetlaw, Bolsover, Broxtowe (the seat of which would be a Labour gain), Chesterfield, Corby (Labour would gain the seat) Derby City, Gedling, High Peak (a Labour gain), Leicester City, Lincoln City (by enough to gain them the seat of Lincoln), Mansfield, North East Derbyshire, and Nottingham City. Erewash is evens between Labour and the Tories (which seat wise would translate into a Tory hold) That would translate into 4 modest seat gains by Labour from Conservative. The Lib Dems top nowhere (and hold no seats at present anywhere)

UKIP also do not factor. This is worth a look. They are a few % points behind the Tories in Boston but outside of that council area (the Boston and Skegness seat is larger) they are further behind in other parts of Lincolnshire. So while it is definitely a target seat, there’s not much to suggest they will challenge elsewhere.

IIRC Erewash the district is less Labour than the constituency, and the same is true of Amber Valley, so that could well be two more Labour gains, and I don't think there's any way you can get a good handle on Sherwood constituency from district figures.

UKIP got over 50% in Boston in the Euros (and close to that in South Holland too) so I'm a bit surprised that your model doesn't have them ahead there: can you give the calculations?

[YL]

YouGov puts the Greens ahead of the Lib Dems for the first time: Lab 34 Con 31 UKIP 17 Green 7 LD 6 SNP/Plaid 4

(Tables here.)

[YL]

This poll is actually even better for the Nats than the YouGov crossbreaks (which, of course, aren't weighted properly for Scottish polls) have been.  I tend to think of MORI as an OK pollster, although they can be quite volatile.

Don't take the seat predictions too seriously; in an extremely high swing scenario the uniform swing assumption might not work very well.  (Actually I think they used Martin Baxter's model, which at least avoids the problem of votes going negative, rather than uniform swing.)  Labour shouldn't take any comfort from that, but Kennedy might.

[YL]

The high volatility of the Ashcroft polls is probably simply down to having a fairly small effective sample size.  The "headline" sample size may be 1000, but the actual voting intention figures in this Monday's poll are based on just 522 respondents.

Speaking of Ashcroft, he has just released some more constituency polls, in the Tory held seats ranked between 40 and 52 on the Labour target list.  Summary of the results (using his preferred question, the constituency specific one): he has the Tories narrowly holding Blackpool North & Cleveleys, Kingswood and Loughborough, but Labour gaining Northampton North, Bury North, Erewash, City of Chester, Croydon Central, Worcester, Keighley, Cannock Chase (three way marginal, with UKIP actually ahead on the standard question and second, only two points behind, on the constituency one) and Wirral West.

[YL]

For comparison, here are the YouGov polls for the last week (the date is the date of publication in the Scum or the Sunday Times):

31 Oct: Con 33 Lab 32 UKIP 15 LD 7 Green 7 SNP/Plaid 5
2 Nov: Lab 32 Con 31 UKIP 18 LD 7 Green 6 SNP/Plaid 5
4 Nov: Lab 34 Con 33 UKIP 15 LD 8 Green 5 SNP/Plaid 4
5 Nov: Lab 34 Con 32 UKIP 15 LD 7 Green 6 SNP/Plaid 5
6 Nov: Lab 33 Con 32 UKIP 17 LD 7 Green 7 SNP/Plaid 4

[YL]

YouGov's headline sample size is also larger, for example 2041 in yesterday's poll.  (Presumably it's easier for them to get a large sample than it is for a phone pollster.)  It does indeed seem as if they have a weaker turnout filter, as the actual voting figures are based on about 1600 respondents, around three times as many as Ashcroft's.  This may have something to do with apathetic non-voters being less likely to sign up for a polling panel as well.

I have some qualms about internet panel polling as a methodology, but my impression is that YouGov have a fairly good track record, which suggests that they have found ways of dealing with its intrinsic flaws.  (I'm not prepared to say the same about any of the other internet pollsters...)

And yes, political weighting (such as party ID or past vote, though I have issues with the former especially as used by Populus) or newspaper readership (likely to be correlated with party ID) will reduce volatility too.  So it's not surprising that Populus and YouGov give relatively stable figures.

[YL]

Here are the last week's YouGov polls:

7 Nov: Lab 33 Con 32 UKIP 15 LD 8 Green 7 SNP/Plaid 4
9 Nov: Lab 33 Con 33 UKIP 16 LD 7 Green 6 SNP/Plaid 5
11 Nov: Lab 33 Con 32 UKIP 17 LD 6 Green 6 SNP/Plaid 5
12 Nov: Lab 34 Con 33 UKIP 15 LD 7 Green 6 SNP/Plaid 5
13 Nov: Lab 35 Con 32 UKIP 15 LD 7 Green 6 SNP/Plaid 4

Yesterday there was a MORI poll with a three point Tory lead which got a lot of headlines.  It also, rather bizarrely, had the SNP on 8% (which is not far short of Scotland's share of the population).

[YL]

If you're going to post single YouGov polls, and I'm not sure it makes much sense to do so given that they come out five times a week, please post them all, not just the ones with Tory leads.

Here's my weekly summary of the last five:
14 Nov: Con 33 Lab 32 UKIP 15 LD 8 Green 6 SNP/Plaid 5
16 Nov: Lab 33 Con 31 UKIP 18 LD 7 Green 5 SNP/Plaid 5
18 Nov: Con 33 Lab 32 UKIP 15 Green 8 LD 7 SNP/Plaid 4
19 Nov: Lab 34 Con 32 UKIP 15 LD 7 Green 6 SNP/Plaid 4
20 Nov: Con 34 Lab 33 UKIP 14 LD 7 Green 6 SNP/Plaid 5

[YL]

For what it's worth, there are only six seats where the electionforecast.co.uk model currently projects a chance above 10% of a UKIP gain.  These are

Clacton (99%)
Boston & Skegness (85%)
Thurrock (49%)
South Thanet (43%)
Great Yarmouth (22%)
South Basildon & East Thurrock (19%)

[YL]

A below 50% chance of Farage himself winning? what.

I don't think there's been a constituency poll there since Farage was selected, so the model probably doesn't really "know" that he's the candidate.  How good their model actually is at predicting seats which behave differently from UNS very much remains to be seen.

[YL]

Here are this week's YouGov polls.  There are six this week, because Murdoch's two Sunday rags had separate polls.

21 Nov: Con 34 Lab 33 UKIP 15 LD 7 Green 6 SNP/Plaid 4
23 Nov (Sunday Times): Lab 33 Con 33 UKIP 16 LD 7 Green 6 SNP/Plaid 5
23 Nov (Sun): Lab 34 Con 33 UKIP 15 LD 8 Green 5 SNP/Plaid 4
25 Nov: Lab 34 Con 30 UKIP 18 Green 6 LD 6 SNP/Plaid 5
26 Nov: Lab 33 Con 32 UKIP 16 LD 7 Green 6 SNP/Plaid 5
27 Nov: Con 33 Lab 32 UKIP 16 Green 7 LD 6 SNP/Plaid 4

The Sun on Sunday poll had a separate crossbreak for Sun readers (which had the Tories first and UKIP second).  In a piece of reporting bad even by their standards, the Daily Express reported these figures as if they were the full figures from a normal national poll.

In other polling news, Survation have just released a constituency poll of Camborne & Redruth showing UKIP 33 (+28) Con 30 (-8) Lab 22 (+6) Green 7 (+6) LD 6 (-31).  And yes, I realise that the words "Survation" and "constituency poll" may cause some people to run a mile.

[YL]

LordA releasing one from EdM's Doncaster North with Ed ahead, UKIP a strong second. His teaser was that there's enough Tories who could tactically vote to get rid of Ed. Not happening though, obv.

No numbers yet.

He also did Sheffield Hallam and South Thanet.  These are strange polls: there is some re-allocation of don't knows which is having a drastic effect in them.  So I'll give four sets of numbers for each.

Doncaster North, standard question, no re-allocation: Lab 52, UKIP 30, Con 12, LD 2, Green 2
Doncaster North, standard question, re-allocation: Lab 51, UKIP 27, Con 14, LD 3, Green 1
Doncaster North, constituency question, no re-allocation: Lab 55, UKIP 27, Con 11, LD 3, Green 2
Doncaster North, constituency question, re-allocation: Lab 54, UKIP 25, Con 13, LD 4, Green 2

Sheffield Hallam, standard question, no re-allocation: Lab 33, Con 23, LD 17, UKIP 14, Green 12
Sheffield Hallam, standard question, re-allocation: Lab 28, LD 27, Con 23, UKIP 11, Green 10
Sheffield Hallam, constituency question, no re-allocation: Lab 32, LD 26, Con 18, UKIP 14, Green 11
Sheffield Hallam, constituency question, re-allocation: LD 31, Lab 28, Con 19, UKIP 11, Green 9

South Thanet, standard question, no re-allocation: UKIP 35, Con 32, Lab 26, Green 4, LD 2
South Thanet, standard question, re-allocation: Con 33, UKIP 29, Lab 27, LD 6, Green 3
South Thanet, constituency question, no re-allocation: UKIP 36, Con 31, Lab 26, Green 3, LD 3
South Thanet, constituency question, re-allocation: Con 34, UKIP 29, Lab 26, LD 7, Green 3

Edit 2 December: the Doncaster North figures are now the corrected ones, not the incorrectly weighted ones originally released.

[YL]

There are also some more of his marginal polls, mostly in the LD/Con battleground.  There isn't a single seat where the standard voting intention question gives a Lib Dem lead, but once people are prompted to think of their constituency the Lib Dems are ahead in Brecon & Radnorshire, Carshalton & Wallington, Cheltenham, Colchester, Hazel Grove, Kingston & Surbiton, Lewes, Southport and Thornbury & Yate, but the Tories are ahead in North Devon and Portsmouth South.

There are also two LD/Lab seats.  Burnley, unsurprisingly, looks like an easy Labour gain, but Birmingham Yardley has the LDs three points ahead on the constituency-specific question (19 points behind on the standard one).

These polls don't seem to have anything like the same re-allocation effect as the leaders' seats ones.  The tables look different too, so I think there's some different methodology.

[YL]

No way is Tory support up from 2010 in Donny North; the government is less popular than ebola in the area and the Tories have performed hideously in local elections. Duff poll.

It isn't, because this poll was conducted in some strange universe where Ed Miliband lost his seat to the Tories in 2010.  Look at the 2010 vote, after weighting.

Con 240
Lab 169
Lib Dem 78
(absolute numbers)

File straight into the nearest dustbin.  (Sorry for not noticing this earlier.)

The Hallam and Thanet polls have more reasonable 2010 figures, but I'm suspicious of them too, partly because of the crazy re-allocation.

[YL]

Did they only poll Sprotbrough or something? LOL

They actually massively upweighted the 2010 Tory respondents and downweighted the Labour ones.  The unweighted figures in the same table (these are from Table 2, but the pattern is the same in the others) are Lab 368, Con 114, LD 47.

[YL]

Ashcroft has retracted his Donny poll of EdM's seat and corrected it to

54 Lab (+7)
25 UKIP (+21)
13 Con (-8)
4 LD (-11)

Should've smelt a rat when it was showing the Tories UP on 2010 here.

Safe as houses. Terrible news for Ed Miliband.

I've changed the figures in my original post to the corrected ones.

Of course, the thing about the original figures was that to journalists who couldn't be bothered to check the details they fitted in quite well with their "Ed is carp" narrative, and indeed with their "UKIP are taking lots of votes from Labour in white working class communities" narrative.

[YL]

Perhaps I'm biased, but opt-in internet panels tend to be junk.

By default they are, but my feeling is that YouGov have a good enough track record (in the UK anyway) to suggest that they know how to sample from their panel (which is huge) and weight their sample in a way which gives reasonable results.  However, the possibility remains that a major change in British politics (and the rise of UKIP or the collapse of the Lib Dems might qualify) might mean the assumptions which have worked well for them don't any more.

None of the other online panel pollsters (Survation, Populus, ComRes, Opinium) have much of a track record.

By the way, I didn't mean to suggest that UKIP weren't getting any 2010 Labour voters, just that the scale of the movement of formerly Labour voters to them in historically strong Labour areas like Doncaster North is exaggerated.  I suspect that most "white working class" voters who fit the stereotype have never been particularly reliable Labour voters.

[YL]

This week's YouGov polls:

28 Nov: Con 31 Lab 31 UKIP 17 LD 8 Green 6 SNP/Plaid 5
30 Nov: Lab 34 Con 32 UKIP 15 LD 7 Green 6 SNP/Plaid 5
2 Dec: Lab 32 Con 32 UKIP 15 LD 8 Green 6 SNP/Plaid 5
3 Dec: Lab 33 Con 32 UKIP 16 LD 7 Green 7 SNP/Plaid 5
4 Dec: Con 32 Lab 31 UKIP 17 Green 7 LD 6 SNP/Plaid 5

[YL]

Not exactly the world's best kept secret, but Alex Salmond is now officially trying to be SNP candidate for Gordon.  (I was hoping for Inverness, Nairn, Badenoch & Strathspey...)

[YL]

Anyone have any idea why electionforecast.co.uk thinks UKIP have a 9% chance, higher than all but six other constituencies, of winning Dwyfor Meirionnydd?

[YL]

UKIP have suspended their general secretary. Something to do with candidate selection.

Looks like it was to do with this.

And apparently it's looking like Neil Hamilton may fill the gap it's left as candidate in one of their more winnable seats (South Basildon & East Thurrock).

[YL]

If I were in UKIP (stretching my imagination for a few minutes) I wouldn't touch Hamilton with a bargepole in a winnable seat.  OK, a lot of voters won't remember the scandal, but you can bet that the local Tories (and/or Labour) would remind them of it at every opportunity.  I might let him stand somewhere like Hampstead & Kilburn, or he could try to get his old seat back from Osborne...

Anyway, this weeks' YouGov polls, featuring lots of ties for first place and the Lib Dems in fifth more often than not:

5 Dec: Lab 32 Con 31 UKIP 15 Green 8 LD 7 SNP/Plaid 6
7 Dec: Con 32 Lab 32 UKIP 17 Green 7 LD 6 SNP/Plaid 5
9 Dec: Con 34 Lab 33 UKIP 15 LD 6 Green 6 SNP/Plaid 5
10 Dec: Lab 32 Con 32 UKIP 15 LD 8 Green 7 SNP/Plaid 4
11 Dec: Con 33 Lab 33 UKIP 15 Green 7 LD 6 SNP/Plaid 4