Published online by Cambridge University Press: 12 October 2016
In this article, I present results from a conditional logit model of vice presidential selection that predicts the selection of vice presidential candidates for both Hillary Clinton and Donald Trump in 2016. Examining contested major party vice presidential nominations from 1960 through 2012, the model finds media exposure, political experience, military service, age, and demographic (gender/racial/ethnic) diversity to be significant factors in the selection process. In the end, the model correctly predicts 15 of the 21 (71.%) contested major party nominations during this period. For 2016 the model correctly and convincingly predicts Mike Pence as Donald Trump’s selection, but incorrectly predicts Cory Booker as Hillary Clinton’s pick. This reduces the overall percentage of correct predictions from 1960 to 2016 to 69.6% (16 of 23), but the approach taken here still represents a more appropriate way for social scientists to think about what factors drive vice presidential selection.