I was just rereading The American Voter and noted with interest their theoretical justifications for controlling for sex in voting analyses (see ch.17). These essentially boil down ot:
- The lingering effect of the belief that women should not play a political role
- The fact that this effect is stronger at lower levels of education (which is amplified by the fact that women are less educated than men).
- The fact that women have lower levels of political efficacy than men
- The fact that women have lower levels of political sophistication than men
For those unfamiliar with The American Voter, it is important to note that the book was written 50 years ago. So if we are willing to conclude that the “lingering effect” of the belief that women should not play a political role is no longer lingering today, then what’s the theoretical justification for continuing to include sex in voting analyses? Even if, as the American Voter authors suggested, sex is a proxy for feelings of political efficacy and/or sophistication – and that’s obviously a big if these days – wouldn’t we be better off just including direct measures of efficacy and sophistication in our analyses? Similar logic should apply to using sex as a proxy for policy preferences: why not just include the policy preferences themselves in the models?
The one justification that seems clear to me is when sex conforms directly to an “identity party”. So similarly to the fact that in a multi-party system featuring an agrarian party (or minority party), we want to control for being a farmer (or ethnicity), we should probably control for sex in elections featuring women’s parties. But beyond that, I pose the question again: what’s the theoretical justification for including sex in voting models?
{ 7 comments }
Surely the theoretical justification is that women be different than men.
Gender is an important form of identity. Even if there aren’t explicitly gender based parties in the electorate it’s important to control for possible effects of this identity. This is particularly true in the US where the two parties have to combine a number of different identities. The Republican party is at the same time a class based party, a racially based party, a religiously based party, and yet only weakly any of these things.
To put it another way, the fact that there is no explicitly “white” party in the US doesn’t mean that controlling for racial identity isn’t useful in many analyses. The same goes for gender.
Doesn’t the fact that there are differences between the way men and women vote, between their partisan commitments and between their views on a host of issues from the safety net to foreign intervention provide more than enough justification?
Like other demographic variables–race, religion, income, etc.–it can serve as a useful proxy for unobserved issue preferences. Even in the US, where there’s no women’s party (or black party, christian party, or worker’s party, at least not officially), there are certainly issues on which the parties and their candidates differ, where the cleavages at least partially split along gender lines.
For example, running some data from the 2004 NAES, women:
While in the NAES we have this data directly-measured (of course) and could thus use it on its own in vote models, in many surveys we don’t. Or, when we do, we have such fine-grained measures that aggregation is problematic. In either case, gender serves to capture some of this variation, so it’s useful for keeping vote models simple but meaningful.
Of course, there’s also the path dependency side of the equation; because everyone uses gender, it’s much easier to include it than exclude it. That’s a problem, of course, when we do have preference data, because the correlations between the data often drown out the significance of the issue preferences in regressions.
(Maybe that’s why the issue voting lit is still fairly primitive–social pressure to include demographic confounders makes the burden of proof that issues matter so much higher? Interesting topic for another time.)
Not sure what the status of the lit is on this question, but I imagine somebody’s tackled it before–seeing whether gender matters when everything else is controlled for as well. For what it’s worth, in my most recent paper (presented at APSA, on campaign effects), I found gender to be highly-significant for predicting 2004 presidential vote choice, even after accounting for partisanship, ideology, issue salience, and aggregated issue preferences. Didn’t run it with each issue separately, however, so that might have changed the results.
What The American Voter deems to be “political sophistication” is really a facet of internal political efficacy, as the original poster notes. It’s worth noting that the gender gap on this particular indicator is as large now as what it was when the AV was written.
Thus, if one is interested in voter turnout, a model that takes efficacy into account without taking gender into account would probably miss important conditional effects.
Would someone ask what the theoretical justification is in controlling for race? The conditions of life, employment, political representation, education, and attitudinal position of women and men in the U.S. are still pretty different, even 50 years out from the AV. Sex matters in the same way that a host of other demographic characteristics and experiences still matter.
In re the relevance of the American Voter for contemporary politics, see:
http://www.press.umich.edu/titleDetailDesc.do?id=92266
Comments on this entry are closed.