Changes are since the last YouGov poll about 10 days ago.

Typical! He always picks family… [/not the nine o'clock news]

Among the findings was that 20% would not vote for a leader who was an atheist (other categories: 72-year-old 43%; gay/lesbian 23%; Muslim 23%; woman 7%; divorced person 7%; Christian 7%; black 5%).

I’m tempted to say we should go for an octogenarian black lesbian atheist divorcee (the daughter of a Muslim/Christian mixed marriage) next time, just to spite the electorate. But sadly I don’t think we have an MP who fits that description, following the narrow defeat of Fatima O’Reilly in Tunbridge Wells.

But seriously, I’m surprised that an atheist leader would be almost as unpopular as a homosexual one, and I wonder if some of our opponents may be tempted to make capital out of Clegg’s pronouncements on the issue.

On these figures, online seat projection algorithms indicate a hung parliament. And that’s assuming more than 30 Lib Dem seats will be lost to the Tories. If that were scaled down to (say) 20 or so, for reasons discussed before, it would leave the Tories well short of a majority.

Westminster:
CON 20%(+3)
LAB 38%(+6)
LDEM 11%(-2)
SNP 29%(-5)

Holyrood (constituency):
CON 14%(+1)
LAB 31%(+5)
LDEM 12%(-3)
SNP 39%(-3)

Holyrood (regional):
CON 16%(+2)
LAB 29%(+4)
LDEM 11%(-3)
SNP 32%(-3)

Changes are since last month. For reference, the changes in Westminster voting intention since the 2005 general election are:
CON 20%(+4)
LAB 38%(-2)
LDEM 11%(-12)
SNP 29%(+12)

I say ignore them and let’s go canvassing ourselves!

Changes are since last month – though that was the weird one that had the Tories on 52% and the Lib Dems on 12%.

So not really any confirmation here of the jump in Lib Dem support shown by ICM.

YouGov, MORI and ComRes all dismiss don’t knows – making the assumption that would either be unlikely to vote, or that their votes would split in roughly the same proportion as other people (or that it isn’t a pollsters place to second guess what people will do).

ICM and Populus both reallocate a proportion of the don’t knows on the assumption that they will vote in the same way as they say they did in 2005. Populus reallocates 50% of former Conservative and Labour voters, and 30% of former Lib Dem voters (which seems unfair, but is on the back of how interviewees behaved in Populus’s 2005 polling). ICM reallocate 50% of former Con, Lab and LDem voters.

Phew. In short, ICM’s treatment of don’t knows is the most favourable to the Lib Dems (at the moment – at various times in the past the reallocation has helped the Tories and Labour). Not that it is all down to that, there will be sampling and weighting differences too.

The margin of error is 3% or so, these sort of variations just don’t mean very much at all.

The most recent polls from the other 5 polling organisations all show us at or below our low for the year so far.

Rather than just splitting the difference it would make more sense to attach greater weight to the 5 that agree. But only time will tell who is right – ICM or all the others.

If you assume we’re at about 17%, then virtually all recent polls fit within the margin of error. The Tories seem to be at about 42% and Labour at about 29%.

This ICM poll is obviously different (though it is normally one of the more favourable ones for the Lib Dems, and has shown us at or above 21% five times this year). Only time will tell whether it’s an outlier or a herald of a different trend.

The next indication will come quite soon. I gather a MORI poll is due in the next day or two, together with a YouGov survey of marginals (though I don’t know whether it will include Lib Dem marginals).

…

tomorrow morning.

I think it’s very egalitarian of the polling companies to produce three polls, one good one for each party

What worries me to an extent is that all the polling companies have developed a methodology to account for the “silent tories” since the mid 90s. As that trend is unravelling then they need to take account of that but essentially they are guessing at how to account for that (in a very educated and well tested way – but no-one really knows how accurate it will be untill we get to a real election).

The other possibility is that with a highly volatile electorate, minor differences in questioning are having big effects. Still doesn’t really explain why one company shows us up 4 and one down 4 though!

I suddenly feel like a right meanypants arsehole myself.