Sunday, September 14, 2008

Gender gap: no VP effect

In anointing Sarah Palin as his running mate, John McCain was gambling that women voters will be more likely to vote Republican with the prospect of seeing a woman in the vice president's office. Does history bear this out?

Of course, we only have one case to look at: 1984's election, with Geraldine Ferraro as the Democrats' vice presidential nominee. Still, we can ask, was the gender gap wider in 1984? Did women cross party lines to support Ferraro? It doesn't look like it. The graph below shows the percent of men and women who voted Republican in each presidential election between 1948 and 2000:As the graph shows, the gender gap is a surprisingly recent phenomenon. It's only been occurring consistently and with any real strength since 1980. As for Ferraro, it doesn't look like she had much of an effect here. The gender gap in 1984 (7.6%) was exactly the same as it was in 1980, and was considerably smaller than it has been in recent elections.

Does this mean that Palin won't have such an effect? Hard to say with only one data point. And maybe things are different in a close election. But history still doesn't testify to much of a VP effect.

6 comments:

Anonymous said...

Seriously, Seth, is this the quality of "research design" that you learned at UCLA?

Seth Masket said...

Is this the quality of "intellectual debate" you learned in college? Are these the "manners" your parents taught you?

Anonymous said...

What does "manners" have to do with it?

Clearly you have a problem disconnecting your social scientific side with the (very strong) partisan side. This lame research design is a perfect indicator.

Seth Masket said...

Well, you've now criticized my analysis twice, anonymously, without saying what's wrong with it or how it could be improved. Yes, that's bad manners.

Whether Sarah Palin's candidacy will cause more women to vote for the Republican ticket strikes me as a pretty central question for the current election. One way of addressing that question is to look at the one historical precedent we have: the 1984 election. The data come from the NES -- you could replicate this analysis pretty easily. If you think I did something incorrectly, you might point out what. If you think this analysis doesn't apply to the current election, I'd be curious to know why. Otherwise, if what I write bothers you, you are under no obligation to read it.

Eric Rubin said...

maybe manners aren't the problem. i think anonymous is just a douchebag

Anonymous said...

Congratulations on picking up a troll. Signifies relevance.