I'm no judge of male pulchritude, so I can't say whether Ed Miliband is too ugly to become Prime Minister*. But I can say that his appearance matters. There's evidence around the world that better-looking candidates tend to get more votes (I stress 'tend' here because looks aren't everything, as Michael Gove's electoral success demonstrates). This shouldn't be surprising. Beauty pays in the labour market generally, so why shouldn't it pay in the political labour market too?

What's worrying, though, is: why do looks matter in politics? We can dismiss the possibility that it is because looks are correlated with ability, because most studies find a beauty premium for earnings even controlling for observable qualifications. Whilst it might be the case that Oxford PPE graduates are better looking than the population generally, it is not so obvious that the better-looking PPE graduate is more able than the less good-looking one.

Instead, I fear there is another reason why look matter. Rick Wilson and Catherine Eckel have found (pdf), from laboratory experiments, that people are more inclined to trust the better looking. Given that labour contracts, like the implicit contracts between politicians and voters, are incomplete, this means that employers and voters are inclined to hire the better looker.

But here's the thing. This trust is unwarranted. Wilson and Eckel found that attractive trustees were no more likely to repay that trust than others. And other research has found attractive people (measured by facial symmetry) to actually be meaner, in the sense that they are more likely to defect (pdf) in prisoner dilemma games, and more likely to make low offers (pdf) in ultimatum games.

Good-looking politicians might win more votes, then, but it doesn't follow that they behave better when elected; insofar as Nick Clegg is good-looking, he corroborates this theory.

In this sense, voters' preference for good looks is not just unfair, as Norm says. It is irrational. And it might be an expensive irrationality at that.