Gary Andres has a “nice piece”:http://www.weeklystandard.com/blogs/mad-men-how-gender-differences-may-shape-2010-election_511360.html at the _Weekly Standard_ discussing political science research on the gender gap in voting behavior and public opinion. One key tidbit comes from “some research”:http://departments.colgate.edu/geography/People/elgie/geog225/text225/gender%20gap.pdf (partial pdf) by Karen Kaufman and John Petrocik:
bq. They also show that much of the gap was caused by men becoming more Republican rather than women shifting toward the Democrats.
The graph above updates one of theirs through 2008 (data “here”:http://www.electionstudies.org/nesguide/2ndtable/t2a_2_1.htm and “here”:http://www.electionstudies.org/nesguide/2ndtable/t2a_2_2.htm).
The important finding is that the gender gap in party identification is entirely due to the changing preferences of men. There is no noteworthy trend among women. I think that finding, which is hardly new, still hasn’t fully filtered into the conventional wisdom.
Andres cites other research, including this “new paper”:http://poq.oxfordjournals.org/content/74/3/477.abstract (gated, alas) by Paul Kellstedt, David Peterson, and Mark Ramirez, that finds that the preferences of men regarding the role of government are also more labile.
One explanation for the increasing gap between men and women comes from this “research”:http://www.lmalla.com/storage/todd/Thesislit/Gendergap/Dynamics.pdf (pdf) by Janet Box-Steffensmeier, Suzanna de Boef, and Tse-Min Lin:
bq. We find that from 1979 to 2000, the partisan gender gap has grow when the political climate moved in a conservative direction, the economy deteriorated, and the percentage of economically vulnerable, single women increased.
See their paper for more.
Andres’s ultimate point is that “Mad Men” might be influential in this election, due in part to their greater movement away from the Democratic Party and toward the Republican Party. That may be correct, but it’s important to acknowledge that other subgroups of the population have moved in that direction as well. In fact, nearly every subgroup has moved in that direction, depending on how you capture their attitudes (e.g., party identification, presidential approval, generic ballot, etc.). So we can’t yet conclude the 2010 election will depend on trends among men specifically.