Sierra Davis

Why are women so underrepresented in American politics? Scholarship has generally preferred supply-side driven explanations (i.e., that few women run for office) to demand-side driven explanations (i.e., voters, and to some extent party gatekeepers, discriminate against women candidates in support). This focus on supply at the expense of demand is troubling for many reasons. First, the individual observational studies and meta-analysis that typically dismiss demand side explanations often bury significant partisan differences in demand, which find that Republicans do discriminate against women candidates. Secondly, voters do evaluate men and women differently on some aspects like parenthood. Such findings indicate that demand is not quite dead, though perhaps more complicated than originally conceived. Thirdly, completely separating demand and supply is problematic because demand at one political level can often affect supply at other levels. The actual picture of the candidate pipeline in general and by gender is thus unclear because the pipeline is a function of both supply and demand.

Finally, too few studies explicitly compare experimental findings (which typically provide high internal validity) with observational findings (which typically provide high external validity). Those that do compare tend to focus later in the candidate pipeline, where selection effects from earlier processes in the pipeline pose major problems to inference. Having both evidence types not only helps researchers draw more valid conclusions; it also contributes to debate about the usefulness of behavior in survey experiments in understanding and predicting real-world outcomes. If results from a survey experiment significantly differ from observational findings, this incongruence raises serious questions about either the observational model specifications or the external validity of (and thus real-world conclusions drawn from) experimental results.

As part of a larger dissertation project to better map gendered and partisan differences in the candidate pipeline as a function of both supply and demand for women candidates, I will run an experiment to investigate whether women and men are evaluated differently in state-level elections.