A pollling quirk or are women really flocking to Dave?

The full dataset from May’s YouGov poll in the Telegraph shows a gender divide in voting intention on a scale that is almost unprecedented.

The 1,009 men in the sample split CON 33: LAB 35: LD 15: OTH 15 – Labour 2% ahead

The 1,093 women who were polled split CON 43: LAB 28: LD 16: OTH 13 Conservatives 15% ahead

Putting these shares into the Baxter calculator Labour would have a 72 seat majority if the electorate was compised only of men while the Tories would have a 154 seat majority if it was all female.

Normally you need to be wary about taking subsets from polls because inevitably the smaller the sample the greater the margin of error. In this case, however, the sample size for each sex is larger than for most ICM polls. However while the overall sample was adjusted to make it politically representative, using YouGov’s “political identifier”, this is not applied to sub-sets like gender.

What makes this interesting is that a similar picture has appeared in other recent polls – though not on anything like the scale of the latest YouGov numbers. The significance of a move by women to the Tories is huge for it was the move of the female vote to Labour in 1997 that played a key part in Tony Blair’s victory.

In the Observer at the weekend Mary Riddle noted “If women alone had voted in 2005, Labour would have won by 90 seats instead of 66. In a men-only ballot, Blair’s majority would have been down to around 20 seats. Should the female vote melt away next time, then Gordon Brown, as leader, would face catastrophe.“

This change seems to have been fairly recent. In a poll for the Sunday Times taken just after the leadership election in December the male and female CON-LAB figures were almost the same. By February, after Cameron’s initial honeymoon was starting to wear off, the pollster had the Tories with a 3% lead amongst women but 6% behind amongst males. In March the Tories had a 4% female lead balanced by a 4% shortfall amongst men.

Going back through recent ICM polls there is now a small pro-Tory margin amongst women voters but nothing on the scale of that seen in this latest YouGov survey.

Whether this is a quirk in one poll or a serious trend we need to see more evidence. As I have observed here before the YouGov methodology tends to pick up trends earlier and then magnify them more than the other pollsters.

If a switch in female voting intention has changed it is simply a return to normal. Right up until the time Tony Blair and NuLab came on the scene women tended to be much more Tory than men.

Mike Smithson