It's worth checking out Linda Hirshman's piece in the NYT Magazine this weekend, in which she makes 16 different observations about women voters. Hirshman (best known for her "Get To Work" manifesto) makes many strong points here-such as the overemphasizing of voter gender in press coverage of the Democratic primary, and the underemphasizing of socioeconomic and other demographic differences. Indeed, these are topics I've also covered here and here. Her takes on gender differences in news consumption and political knowledge, while beyond the scope of pollster.com, are also worth reading.
But two of the 16 "ways" also relate to some of our work here, and so merit a little further discussion.
Clinton's presence is driving women's turnout?
First, "Way 11" says "Sisterhood is Power," by showing that women's participation in the Democratic primaries has increased since 2004. Last week I wrote that women's participation has been mixed when compared to 2000, but we did not look at 2004, since with an incumbent in office it was a less comparable election. Below is a table that includes the 2004 exit polls to which Hirshman refers.
Indeed, she is right-women's participation has increased since 2004. But since the 2000 percentages were as high or higher, that makes me think it is perhaps not Clinton's presence on the ballot alone that is increasing women's participation. Of course, there are other factors at work. There is record turnout this year in the Democratic primaries. Perhaps new surges of all kinds of voters (younger, African-Americans, women) render the percentage exercise not particularly useful for the Democratic primaries. At any rate, I'll include the 2004 numbers in all future discussions of this topic.
How do we measure the influence of unmarried women?
Second, "Way 15" says "The Political is Personal," and in passing uses a quote suggesting that a swing in the vote among unmarried women from Iowa to New Hampshire shows how influential this group could be. While I'm not arguing that unmarried women are not or could not be influential as a voting bloc, I don't think that particular finding is sufficient evidence. It would leave us in an awkward position of declaring unmarried women unimportant if their vote was consistent, or comparable to married women's vote. The table below shows exit poll results among married and unmarried women for each of the states for which data are available (Edwards is included since they are also states where he did well).
It seems like Iowa was an anomaly. It is the only state where the two groups of women disagreed on the winner. In every other state, Clinton's margin over Obama (or Obama's over Clinton) is not that different across that marital status/gender grouping. Given what we know about younger voters trending toward Obama, I would imagine that older, divorced women vote differently than younger, never married women. Since they are all "unmarried women" it's harder to identify exactly what is happening. If we want to argue the importance of unmarried women's contribution, it should be based on general election predictions or turnout numbers, not the primary vote (or else separate the unmarrieds by age).
But these are both fairly minor points. To me, the more ways people are looking at the women's vote, the better.