Political leaders by nature or by competence?

We at Rigotnomics are very much intrigued with the issue of political leadership. Our discussions and views (here before rigotnomics ,here, here ) although not undisputed, claim that leadership matter as much as institutions in enhancing the welfare of a country, but we have unsatisfactory empirical evidence on that so far. There are a bunch of interesting questions, like how societies choose their leaders? One day, me and buddy PL wanted to make the following experiment: if we ask people to choose among two candidates they don't know, will they choose the one who has actually won the real election result?

We were surprised we had not found that study yet, and yesterday, I found that some guys from Lausanne just did that. John Antonakis and Olaf Dalgas asked some swiss undergrads to choose among two candidates for french parliamentary election they did not know about on the basis of their look. In 70% of the cases, the chosen candidate was the actual one who had won the real election. Interestingly, when they asked children between 5 and 13 years old, the same question as a game (who should be the captain of the boat sailing from Troy to Ithaca), they got the same result. The conclusions seem to be that electoral decision maybe are "naive": not exactly driven by appearence, but maybe voters give too much weight to their "tastes". This is not surprising: as Bryan Caplan explains in his entertaining book, the belief that voters choose rationally their best candidate (and hence policies) on the basis of competence, it's simply a myth. But are these winning candidates better or also in terms of competence? There is some evidence on that but it seems that looking good does not correlate with competence...although anecdotal evidence may dismiss empirics...