The idea that Labour has gone down around a third and Lib Dem support has nearly doubled is just plain silly. This brings polling back to the bad old days of 1992.

Leaving aside that I am inclined to agree with Ralph, your statement “differences from one month to the next may be down to there simply being more or less Labour supporters amongst that month’s sample” - I do not quite see why this is different when you “weight toward past vote”.

Malcolm - there is a strong correlation between how you voted at the last election, and how you’ll vote at the next one. Lots of people change their minds of course, there’s a lot of churn, but the fact remains that the people most likely to vote Labour in 2005 were people who voted Labour in 2001.

In a poll that is weighted by past vote, x% of the sample will always be made up of people who voted Labour last time round. That percentage may actually be wrong because of the difficulties of estimating the level of false recall, but it will be same from one poll to another, so we know that if the labour vote goes up, it means that an increased proportion of Labour’s 2005 voters are sticking with the party (or more 2005 LDs, Cons or non-voters are switching to them) (of course, within the 2005 Labour voters, normal sample variation may mean you happen to have less loyal Labour voters. Nothings perfect, but you’ve clamped the variation down a bit).

In contrast, if you don’t control for the political make up of a poll it could just be that you ended up with more 2005 Labour voters in your sample this month, and their views haven’t changed.

The ICM and YouGov polls are relatively consistent with eachother. I suspect the 6-7% C lead over Lab is more accurate on that basis, especially if this poll here isn’t weighted.

It seems to me like Communicate have had two consecutive ouliers. Unlikely normally, but based on the numbers I’d suspect that. 29% does seem low for Labour - though after this weeks news I may not be surprised to see it more often again soon.

If I’m right putting these into the Weber Shandwick calculator you get:
Labour 42 seats (-8)
SNP 45 seats (+18)
Lib Dems 22 seats (+5)
Conservatives 18 seats (nc)
SSP 1 seat (-5)
Greens 1 seat (-6)

The Greens though are very close to the threshold where they’d get 5 seats rather than 1 so that’s quite a risky one.

The Conservatives on this also seem to be quite close to losing seats.

The result would mean any combination of two of Lab, SNP, Lib Dem works or Conservative coaltiion with Lab or SNP might just about hold together a minority government.

Rob,

Who are the big loosers if the Greens hit about 7%, Tory or LibDem?

Peter.

I would guess that both Labour and Lib Dem would be a hit by a rise in Green support, whereas the Tories wouldn’t be much affected.

Doesn’t follow, Andy. It varies from region to region. The party they deprive of the last list seat in a region may not be the party they won the votes from to win that seat. It’s all about quotas.
The regions where the Greens will look to pick up an extra seat are West and Central. West is their best chance. Here, the last couple of list seats could go either way. The SNP should make advances here and win some constituency seats, so the list seats should end up 2SNP, 2 Tory and 1LD. The last two should be a toss up between a second LD, the SSP and the Greens.
In central, if the Greens take a seat it will probably be at the expense of John Swinburne (SSCUP)

If you play about with the figures on the swingometer increasing Green by 1% leaving the major parties % unchanged would give them 5 seats with Labour and SNP down 1 each and Conservatives down 2 seats . Increasing by 1% more at the expence of SSP would give them 1 more seat with Conservatives down 1 more so it appears it is the Conservatives who would have most to lose .

Sorry about that - when I said the Tories wouldn’t be much affected by increased Green support, what I meant was the Tories wouldn’t actually be losing many votes to the Greens, whereas Labour and Lib Dem probably would do.

I think Andy, it is the imploding SSP factions that might boost the Green vote.

I think that on those poll figures Greens are very close to getting list seats and only a small increase in vote ( or poll error ) would give them extra list seats . On these poll figures it is the Conservatives who are clearly very narrowly ahead of the Greens and would lose out if the Green vote goes up even if there vote stays the same .

Yes, it’s true that while the Tories probably won’t lose many votes to the Greens, they could well be pushed off list seats by them if Green support increases from other sources.

Andy, Mark,

What might be interesting to look at is the Libdem vote, as they are both the most likely, on polling exidence, to split their vote and switch to Green.

It may be that the Tories list seats which are most vulnerable are where the LibDems are strong, such as here in the Highlands, where they are unlikely to get any list seats because they do well FPTP and where there is a strong green agenda.

Peter.

A straight 1% switch of list votes from LibDem to Green on these poll figures would give 4 extra Green seats with Lab and SNP minus 1 each and Conservatives minus 2 . A 2% switch from LibDem to Green would give the Greens 2 more seats with LibDem down 2

We’ve seen a handful of Scottish polls recently but I’ve yet to see one for the Welsh Assembly elections. The talk seems to be that the Conservatives will make modest gains and become the 2nd party of Wales. Do any opinion polls back this up?