General Election 2001 : The Gender Gap

In Saturday's Guardian Yvonne Roberts report discussed the "gender gap" in voting, arguing that women are not served well by the political system, and quoting research by Harriet Harman and Deborah Mattinson.

There has been a lot of talk in new Labour circles over the last couple of years about this. Up to the mid 1980s, the gender gap was very wide -- women were much more likely to vote Conservative than were men, and much less likely to vote Labour. In the last three elections that gap (measured as the difference in the Conservative lead over Labour between women and men) has been narrower, though except in 1987 still perceptible. Between 1992 and 1997, the gap closed from six points to two points; put another way, women swung to Tony Blair 2% more than did men. Not unnaturally, those concerned with women's issues in the Labour Party considered this a triumph and looked to consolidate and build on it at this election.

Nevertheless, some of the arguments that have been produced to encourage greater concentration on women's concerns have been weak or mistaken. For example, it is suggested that women may be more easily swung by the election campaign, on the basis that more female voters say they are "undecided" when faced with the voting intention question. Not a very useful finding, since we know that the vast majority of the undecided respondents, male and female, don't vote. Nevertheless, there is more valid data pointing to the same conclusion: in our poll last week for The Times, of those who gave a voting intention, 29% of the men but 36% of the women said that they have not definitely decided how to vote and "there s a chance" that they may change their mind. But, in any case, is this really evidence that women are more likely to change their minds, or simply that they are more prepared to admit when their minds are not made up? It may be significant that in most polls, not just purely political polls, we tend to find more don't knows among the women than among the men.

If women are really more easily swung, shouldn't they be politically more volatile? Many commentators assume that they are. But the polls don't bear that out: looking at MORI's monthly polls since March 2000 and comparing the month-on-month changes in voting intention of men and women, it is men who prove marginally more changeable. (The average change in the lead among men was 6.7 points, among women 5.3 points; both men and women changed most between August and September -- the period covered by the petrol crisis -- but the change among men was 26 points, double that of the 13 point move among women.) Roberts quote Mattinson: "During last year's fuel crisis and pensions row the female vote just bled away". Well, yes, but according to our polls just half as fast as the male vote did.

In passing, my nomination for the most entirely meaningless statement masquerading as psephological analysis that I have so far seen in this election comes from Roberts reporting Harman and Mattinson: "Two million more votes were cast by women than men [in 1997], almost all to Labour". What is this supposed to mean? How does one identify which two million women were the "extra" voters? If they were all women, then the rest of the women must be far more Tory than the men. And, incidentally, that two million gap is not only because women were more likely to vote in 1997, but because there are more women in the electorate to begin with.

And the "gender gap" at the moment? Labour's lead is two points less among women than men, in both of our polls taken since the election announcement, exactly the same as in 1997. So much for the female vote "bleeding away".