Personality & Partisanship in Congress
Really neat research here from Adam Ramey, on the interaction of personality and partisanship in Congress. The basic thesis is that partisanship isn’t everything – that personality also matters. Ramey used some “recent methods” in text analysis to analyze floor speeches and generate personality scores for each Member of Congress. He then uses the model to generate simulations of how these personality traits interact with partisanship to explain cosponsorship, absences, etcetera. For example, there’s one of his simpler charts – describing how measured conscientiousness predicts the number of absences. Much as we might expect, the more conscientious Congresspeople miss fewer votes.
One thing I’m very curious about is the alignment of Congresspeople with their districts. You’ll occasionally see these maps of personality traits as they vary across the country – New England is more neurotic and more open, Minnesota is more agreeable. It would be interesting to examine congruence between districts and their Congresspeople. Especially given historical data, there’s some very interesting work to be done. Perhaps with the ideological sorting of the late 20th century, Congresspeople came into better congruence with their districts. Or perhaps the personality traits of Congresspeople turn out to be independent of their districts.
One thing I would worry about with this study is the data source – Ramey is using floor speeches from Congress. There are many things that might best represent the true spirit or personality of a person, but I’m not sure that floor speeches are the best way. They’re formal, they’re pre-composed, and they’re mostly written by staff committee. Even if they genuinely represented a Congressperson’s personality, we might still have selection bias if Congresspeople choose when to speak or not to speak. For example, a particularly ill-tempered Congressperson might appear milder on paper if his Chief of Staff has any sense at all. I’m not able to comment on the specific methods Ramey is using, but this does seem like a large potential source of concern in drawing conclusions.