I recently came upon an academic paper in which three UCLA poli-sci researchers concluded that more competitive election contests produce better-looking candidates than less competitive ones. For example, if the 'out' party has a good chance of beating the incumbent and taking the seat, somehow they come up with a good-looking candidate; if they have little chance of winning, they come up with someone more homely.
The authors do not use those exact words, naturally. They use fancy-pants terms like "facial competence" and "valence advantage," and they have some whole scheme for determining facial quality. But that's the gist.
But they have little to say in their paper about open seats -- and how likely it is that the 'out' party would end up nominating... well, Scott Brown.
So I emailed to ask. Lead author Matthew Atkinson responded:
I cannot claim that we would have predicted that the Republicans would nominate a candidate you describe as a "pretty-boy former model Republican." But given the circumstances -- a coveted office and an out-party that can only win statewide office with candidates possessing some sort of valence advantage -- perhaps it's not surprising that the Republicans have put forth a former model.
But wait! Atkinson had more to say -- which I will put in a follow-up post.