A more finely tuned bullshit radar

In today's column, Polly Toynbee writes:

at the 2005 election women and men voted identically.

Of course, the only way of getting to this is through polls, which are numerous and often contradictory.  But that should call therefore for greater precision when discussing results, not less.  There is indeed a source which claims that men and women intended to vote in almost identical ways -- the table at the bottom of this report by the BBC.  However, read it carefully, and you'll see that it does not talk about how people actually voted -- it is a sum of all campaign polls, of intentions in other words.

However, if you look at the polls which ask about how people actually voted, you'll see the big differences.  Take this:

One major surprise for Labour was contained in the final campaign polls for ICM and Populus which showed that women were crucial to Labour's victory last night.

Both pollsters showed that women were far less likely to vote Conservative - only 27% - than men, 33% of whom backed the Tories. Instead, 25% of women turned to the Liberal Democrats and 39% backed Labour. This compared with 20% of men voting for the Lib Dems and 37% for Labour.

Which came from the, er, Guardian.  The Fawcett Society found something very similar:

At the general election 2005, Labour polled higher support among women than among men, with support at 38% and 34% respectively.
[p2 of "Fawcett/ Ipsos MORI briefing on women’s votes, September 2006".  Word document available here]