There has been a recent dust-up between Linda Hirshman on one side and Mark Schmitt and Ann Althouse on the other. (Count me as a fan of both Linda's and Ann's scholarship.) If I am reading correctly, there seems to be an impression that women's votes seldom determine the outcome of presidential elections. Usually both men and women favor the same candidate, so generally neither men nor women control the outcome.
According to exit poll data reported in the New York Times, a plurality of women have voted for the candidate getting the most votes in every presidential election since 1972 except for the last one, 2004 (when they favored Kerry). And men have voted for the plurality winner in every election since 1972 except 1996 (when they favored Dole) and 2000 (when they favored Bush). So women's first choices have lost 1 election in the popular vote and 2 elections in the electoral vote, while men's first choices have lost 2 elections in the popular vote and 1 election in the electoral vote.
Bottom line: The assumption that women's preferences for president seldom determine the outcome of a presidential election is indeed true. But then men's preferences are seldom dispositive either. On the other hand, in each of the last three presidential elections, men and women have differed in their first choices for president. We may have entered a new era in which the winners of presidential elections don't receive a plurality from both genders.