Here is the figure comparison (I note it is by gender not sex) for women (unweighted by likelihood to vote)
28th / 29th Jan
CON 28%
LAB 34%
LD 4%
SNP 3%
Other 4%
Would not vote 7%
Don't know 20%
5th / 6th Feb
CON 29%
LAB 28%
LD 6%
SNP 2%
Other 4%
Would not vote 10%
Don't know 21%
If YouGov are right, then this certainly isn't a switch from Lab to the Cons as I suspected. Its about the Don't Knows and the strength of feeling to vote.
I do think we need urge some caution here. There is a margin of error on these figures. However a difference of 6% which seems to have mainly switched over to the will not vote column is bigger than that margin of error.
BUT
You should never just take two polls in isolation. Its whether it shows a long pattern that's the bigger deal.
Here are the previous yougov figures for 7th / 8th Jan which look like this:
CON 26%
LAB 31%
LD 5%
SNP 4%
Other 3%
Would not vote 9%
Don't know 22%
10th / 11th Dec
CON 29%
LAB 33%
LD 5%
SNP 3%
Other 5%
Would not vote 7%
Don't know 19%
4th / 5th Dec
CON 28%
LAB 29%
LD 3%
SNP 3%
Other 5%
Would not vote 12%
Don't know 20%
7th / 8th Nov
CON 25%
LAB 34%
LD 4%
SNP 2%
Other 4%
Would not vote 10%
Don't know 21%
Which puts something of a different angle on things. There isn't an obvious particular pattern.
Pollsters aren't chasing the voters with a clear intention, they are chasing the individuals about what they would do if it came to the crunch. How they calculate that through weighing is the key bit.
YouGov model on the basis of how likely you are to vote on a scale of 1 - 10. So is more
Here's what it looks like weighed excluding those who would not vote and those who don't know.
5th / 6th Feb
CON 44%
LAB 40%
LD 8%
Other 8%
28th / 29th Jan
CON 39%
LAB 46%
LD 6%
Other 8%
7th / 8th Jan
CON 39%
LAB 45%
LD 7%
Other 9%
5th / 6th Feb
CON 40%
LAB 45%
LD 6%
Other 9%
4th / 5th Dec
CON 43%
LAB 42%
LD 4%
Other 11%
7th / 8th Nov
CON 38%
LAB 49%
LD 6%
Other 8%
No obvious pattern here either. To a certain extent it looks like a shift from Lab to Con but its not.
I think what you are seeing is actually a distinct lack of conviction in Labour voters. The Conservative underlying figure is relatively stable. What is changing is how much Labour voters are convinced they will vote and how motivated they are about voting Labour.
And the answer, is not as convinced as the Labour leadership might like. In order to win an election, they will have to stir up feeling and conviction. They did this successfully at the last election with a message of hope and listening to the people when others were not.
That might well be a one trick pony that won't have the same effect a second time around.
Notably, one of the biggest shifts in voting was in middle aged women aged 25 - 49 which swung things last time around. If they are not convinced by Corbyn or think he isn't listening they won't turn out.
And that's where the danger is for Corbyn. He has to get three key groups out in order to win any election: The Under 25s, BAME voters and women aged 25 - 49.
The important thing to remember is that General Elections are won and lost on the Don't Knows and the Won't Votes. How likely someone is to vote is the key. Neither of these groups of figures really breaks that down. You still have 20% of voters classing themselves as Don't Know.
If you want to put the wind up the Labour Party this is why, saying you won't vote is really important. Make them chase your vote and not take it for granted or be classed as a soft 'I'll vote for them if push came to shove'. If you are asked, say a flat 'I will note vote because'.
Tell them why you can not vote for them as much as you can.
THEY NEED YOUR VOTE AND THOSE FIGURES PROVE ONE THING AND ONE THING ONLY CONCLUSIVELY: The Labour vote is particularly soft and not set in stone and they need to work harder than the Cons for every single one.
The Cons lost a majority by a total of 72 votes across the country. SEVEN TWO votes cast in the right constituencies cost them dear.
Your vote and more to the point your stated voting intention matter.