Phil reminds us of the answer David Cameron once gave to the question, why do you want to be Prime Minister? "Because I think I'd be rather good at it."

Well, that didn't work out too well did it? We've had plenty of PMs whose policies we didn't like. But Cameron was, as Owen Jones says, "a failure on his own terms". He wanted to eliminate the government deficit and keep us in the EU, and failed on both. I'd add that this his counter-productive pursuit of austerity contributed - at the margin - to Brexit by creating a climate of insecurity and xenophobia. In this sense, Cameron's failure was down to deliberate ineptness rather than to facing insuperable odds. For this reason, he is probably one of our worst-ever Prime Ministers.

This poses the question: what, then, was the basis of his belief that he'd be "rather good" at the job?

It's not that he had a great grasp of economics or politics. His mindless drivel about a "global race" and the "nation's credit card revealed utter ignorance of the former. And I've long said that Cameron's government failed to grasp the nature of politics - that it consists in ameliorating problems of collective action.

Instead, his confidence was rooted not in facts but in class. It's a cliché that going to a top public school often gives one confidence and arrogance which others lack. Here's Jimmy McGovern (16'20" in):

Particularly in the working class north, the one thing you do not want is arrogance...so the slightest sign of arrogance in your child you knock it out of them. But in knocking out the arrogance you knock out the ambition, the self-respect, the self-esteem.

So much is obvious. But this poses a question. We are all cognitively flawed. Only some such flaws, however, are penalized. Public school overconfidence, however, seems to be actively selected for: 19 of our 54 PMs went to Eton, and public schoolboys are over-represented in most professions. What are the mechanisms that allow such irrationality to thrive? Here are a few, in no order.

First, there are competence cues. Cameron Anderson and Sebastien Brion show that overconfident people give out more "competence cues" - such as fluency and body language - than others and that these cues are mistaken for actual ability, with the result that the overconfident are more likely to get jobs, regardless of ability. This might have been true of Cameron: in 2005 ConservativeHome said his leadership campaign was "transformed" by "two compelling performances" - exercises in fluency and confidence.

The converse of this is that the under-confident but able don't apply for jobs in the first place: this might mean there's a case for positive discrimination or quotas.

Secondly, there are overconfidence bubbles. Once you've gotten a job through no merit of your own then your overconfidence will tend to grow, perhaps causing you to become dizzy with success and so making a fatal error. We know this happens in financial markets: lucky traders take more risk. It might well have happened with Cameron: his success at winning the Scottish independence referendum made him think he could repeat the trick with Brexit.

A third factor is deference. I'm not thinking here merely of underlings deferring to bosses out of self-interest, but to a tendency identified by Adam Smith:

We frequently see the respectful attentions of the world more strongly directed towards the rich and the great, than towards the wise and the virtuous. We see frequently the vices and follies of the powerful much less despised than the poverty and weakness of the innocent. (Theory of Moral Sentiments,I.III.29)

Contrast, for example, the way the media indulged "Boris" with its othering of, say, John Prescott.

A further selection effect is the desire to overcome agency problems. It's natural for hirers to want people they can trust, and we trust people like ourselves. Public schoolboys will therefore hire public schoolboys, even if they have no deliberate nepotistic intentions. This can lead (pdf) to a lack of cognitive diversity and to groupthink and the Dilbert principle.

My point here is that selection mechanisms - in politics and in markets - don't necessarily weed out irrationality. Quite the opposite. They sometimes select for it. It is these mechanisms that must be dismantled. In this sense, if we are to be better governed in politics and business, we need more class war.