UK General Election - May 7th 2015

[Clyde1998]

Here is the interview for those who haven't seen it: https://www.youtube.com/watch?v=5dFn8RIXOBE

[Clyde1998]

Scotland constituency polls to be released by Ashcroft next week - hopefully they'll clear up the Scotland situation.

[Clyde1998]

Constituency polls have never knowing cleared up any situation

No - but hopefully, they'll give us some sort of idea of where the SNP might be performing well and where Labour are holding on.

[Clyde1998]

But what if they're as accurate as all those Australian constituency polls in 2013

Then the Lib Dems might be leading in every seat

Ashcroft do ask two questions - which is quite confusing:
"If there was a general election tomorrow, which party would you vote for?"
"Thinking specifically about your own PARLIAMENTARY constituency at the next General Election and the candidates who are likely to stand FOR ELECTION TO WESTMINSTER there, which party's candidate do you think you will vote for in your own constituency?"

Surely there should only be one question on voting intentions?

[Clyde1998]

Poll shows hardly anyone knows about UKIP & Green policies in key areas.

[Clyde1998]

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

[Clyde1998]

Poll shows hardly anyone knows about UKIP & Green policies in key areas.

How does that compare with the major parties?

It's hard to tell - as most pollsters won't poll on the main three parties. I would assume that the average person would be able to know at least something about the main three parties and their policies in different categories though.

[Clyde1998]

YouGov poll of Wales: Labour 37, Con 23, UKIP 16, Plaid 10, Green 8, LDem 6

Usual caveats about the historic issues with polling in Wales, etc, etc.

No major change on the last Wales poll:

Lab +1; Con N/C; UKIP -2; Plaid -1; Green +3; Lib +1

Electoral Calculus' Wales predictor (where you sadly can't predict Plaid) says:
Lab - 36 seats
Plaid - 3
Con - 1
Lib - 0

Looks like the Conservatives won't do well in Wales then...

[Clyde1998]

Nationally, this is what Sky News are predicting at the moment:

[Clyde1998]

YouGov poll of Wales: Labour 37, Con 23, UKIP 16, Plaid 10, Green 8, LDem 6

Usual caveats about the historic issues with polling in Wales, etc, etc.

No major change on the last Wales poll:

Lab +1; Con N/C; UKIP -2; Plaid -1; Green +3; Lib +1

Electoral Calculus' Wales predictor (where you sadly can't predict Plaid) says:
Lab - 36 seats
Plaid - 3
Con - 1
Lib - 0

Looks like the Conservatives won't do well in Wales then...

?

I have it no change. They lose Cardiff North but gain Brecon.  Labour gain Cardiff Central.

It might be differences in the calculations, maybe...?

[Clyde1998]

Opinions on Ceredigion?

Traditionally one of the more difficult to predict seats in the union...

Lib Dems will collapse, Plaid will stay steady, Conservatives will fall, Labour will gain. I think Plaid might gain that seat, despite the huge Lib Dem majority.

[Clyde1998]

I don't think Respect will hold, but one seat is the Speaker. I think the ICHC is the National Health Action Party now - but they'll run the Conservatives close in Wyre Forest.

I think it's:
3 - Plaid
1 - Green
1 - NHA
1 - Speaker (unless he's listed as Conservative, then it'll be 4 Plaid).

[Clyde1998]

I've been playing around with seats in Scotland - ahead of the Ashcroft polling results for Scottish seats.

Eilean Siar    (SNP) - SNP majority to increase
Northern Isles (LIB) - LIB hold, but seat to become a three-way marginal between LIB, SNP and LAB.
Dundee West (LAB) - Reduction in LAB majority, with the SNP more than halving it.
Glasgow South (LAB) - Massive LAB majority.
Argyll & Bute (LIB) - SNP gain. LIB to finish fourth.

I've only done five seats so far, but I will look at doing more. Based on my calculations, the SNP aren't doing as well as the national polls show.

The calculations are based on the recent Survation poll of Scotland, which noted the change in support between parties. All constituencies are calculated using the same changes between parties.

2010 Labour Voters (1,035,528)
LAB - 76.0%; SNP - 9.4%; CON - 3.7%; LIB - 1.5%; UKIP - 1.3%; GRN - 0.0%; OTH - 0.0%

2010 SNP Voters (491,386)
SNP - 90.9%; LAB - 2.9%; UKIP - 2.7% ; CON - 1.5%; LIB - 0.5%; GRN - 0.5%; OTH - 1.2%

2010 Lib Dem Voters (465,471)
LIB - 47.7%; SNP - 21.6%; LAB - 19.1%; GRN - 7.1%; CON - 4.6%; UKIP - 0.0%; OTH - 0.0%

2010 Conservative Voters (412,855)
CON - 76.0%; SNP - 9.4%; UKIP - 7.9%; LAB - 5.2%; LIB - 1.5%; GRN - 0.0%; OTH - 0.0%

[Clyde1998]

Thing is, Scottish polls have shown everything from a 10 point lead to a 28 point lead this month. How on Earth can anyone project what's happening up there?

Of course, 10 points behind would still be a complete disaster for Labour, but I'm sure they'd take it considering the last few months that Scottish Labour has had.

Using the Survation poll - that had a 20 point SNP lead over Labour (http://survation.com/wp-content/uploads/2015/01/Cross-tab-Record-tables_1_26.pdf) which is probably a mid range poll.

Lab - 42 seats (+1)
SNP - 14 seats (+8)
Con - 2 seats (+1)
Lib - 1 seat (-10)

That's a shock.

EDIT - Map

[Clyde1998]

Thing is, Scottish polls have shown everything from a 10 point lead to a 28 point lead this month. How on Earth can anyone project what's happening up there?

Of course, 10 points behind would still be a complete disaster for Labour, but I'm sure they'd take it considering the last few months that Scottish Labour has had.

Using the Survation poll - that had a 20 point SNP lead over Labour (http://survation.com/wp-content/uploads/2015/01/Cross-tab-Record-tables_1_26.pdf) which is probably a mid range poll.

Lab - 42 seats (+1)
SNP - 14 seats (+8)
Con - 2 seats (+1)
Lib - 1 seat (-10)

That's a shock.

EDIT - Map

I don't mean to be rude (because I called you out on your Wales projection) but how on earth are you getting these results? The figures for Survation on a crude swing would give the SNP 51 seats. How are you calculating, even for fun, that the SNP with a 20 point lead over Labour would only gain 1 seat from them?

My Wales one was a random guess, btw.

I don't think these look right - I'm just going through the calculations again.

Survation published figures on how voters are moving from each party to another - I may have used the wrong numbers... (I hope I have - because that'll mean the Survation tables are very wrong...)

[Clyde1998]

I think it just re-emphasises that you can't read tables and are really bad at making basic calculations.

I've corrected something (and I'm not 100% sure what) - but that seems to have solved it. I'll upload the correct map in a moment.

[Clyde1998]

OK... Here we go again...

Based also on an estimate of how many new voters there are for each seat.



SNP - 37
Lab - 21
Con - 2
Lib - 1

Seems more realistic.

[Clyde1998]

OK... Here we go again...

Based also on an estimate of how many new voters there are for each seat.



SNP - 37
Lab - 21
Con - 2
Lib - 1

Seems more realistic.

Maybe underestimating the tide against Labour in Glasgow, but then the SNP have been overestimated there before.

Still think Charles Kennedy's seat is at least leaning Lib.

There's still a few things that I will probably change - such as a more regional look (using the regional sub samples) and the incumbency bonus in certain seats.

[Clyde1998]

OK... Here we go again...

Based also on an estimate of how many new voters there are for each seat.



SNP - 37
Lab - 21
Con - 2
Lib - 1

Seems more realistic.

I think this is still wrong.

We'll see how it ends up, but it's simply based on redistribution of voters to new parties and a small amount of new voters (which I may have not added enough - which may affect the vote shares in each seat).

My calculations don't take into account regional factors either, but a lot of seats only have the SNP behind Labour by a few hundred votes.

[Clyde1998]

OK... Here we go again...

Based also on an estimate of how many new voters there are for each seat.



SNP - 37
Lab - 21
Con - 2
Lib - 1

Seems more realistic.

There will be no Tory gains in Scotland. 

The gained seat - Berwickshire - the Conservatives are second to the Lib Dems and the equivalent seat in the Scottish Parliament is held the Conservatives. It wouldn't be a shock gain for the Tories there.

[Clyde1998]

Another one on Scotland.

Of course turnout won't be as high as the referendum, no where near, but is it a fair bet that it'll have the highest turnout for all the regions? 70%?

I think the turnout will stay roughly the same as the previous election, but the actual number of people voting will increase - due to increased numbers of people who are part of the electorate.

[Clyde1998]

If I'm not mistaken May2015's forecasting model uses Lord Ashcroft's constituency polls to predict the results in some 100  seats that are marginal or otherwise of interest (your Thanet Souths and Sheffield Hallams), so that there is a considerable part of their prediction that is completely independent of the national polling figures. Very dubious, I'd say.

It's possible to turn those polls off when using the calculator.

[Clyde1998]

https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/a7awj68e8x/Final_Times_Results_150202_Website.pdf

New full-scale YouGov/Times poll shows SNP 21% ahead of Labour in Scotland - an all-time record.

Voting intentions:
SNP - 48% (+1%)
Lab - 27% (N/C)
Con - 15%
Lib - 4%
UKIP - 4%
Grn - 3%
Others - 1%

Leadership trust ratings:
Nicola Strugeon (SNP) - +42%
Jim Murphy (Lab) - -10%

Should Scotland be an independent country?
Yes - 49% (+1%)
No - 44% (-1%)

Excluding Don't Knows
Yes - 52%
No - 48%

If there was a referendum on Britain's membership of the European Union, how would you vote?
I would vote for Britain to remain a member of the European Union - 52%
I would vote for Britain to leave the European Union - 29%

[Clyde1998]

Should Scotland be an independent country?
Yes - 49% (+1%)
No - 44% (-1%)

Excluding Don't Knows
Yes - 52%
No - 48%

Why a Yes majority now, when it is too late? Just protest votes or a genuine swing?

I think it's a genuine swing - as it's the third consecutive YouGov poll showing a "yes" lead. I think that the undelivered devolution promises would have had a effect on future voting intentions (should there ever be a second referendum).

[Clyde1998]

Of course had Yes won, No would now doubtless be leading

Who knows. Also, the events that followed the referendum would have been different - so the polling figures would have changed accordingly.

Salmond resigning may have, ironically, moved people towards yes.