Self-confidence plays an important role in depressing social mobility. That's the message of this new paper:

Even small differences in initial confidence can result in diverging patterns of human capital accumulation between otherwise identical individuals. As long as initial differences in the level of self-confidence are correlated with the socioeconomic background [which they are]'self-confidence turns out to be a channel through which education and earnings inequalities are transmitted across generations.

The gist of their thinking is straightforward. Say you have two people of equal cognitive skills, but one is over-confident about his ability and the other under-confident. The over-confident one is more likely to stick with a subject during the early steep phase of the learning curve - believing that 'I can master this if only I apply myself' - whereas his under-confident is likely to give up, thinking the material too difficult for him. Alternatively, the over-confident student might choose 'difficult' academic subjects at high school, which qualify him for entry to some elite universities, whilst the less confident one would choose less academic subjects which disqualify him.

Important and powerful as these are, they are not the only ways in which class differences in confidence can affect outcomes. There's also:

- Overconfident people might select into occupations where there's a high pay-off to the lowish probability of success, such as management, law journalism or politics. Less confident folk, under-estimating their chances, might prefer occupations which yield less skewed rewards.

- People misperceive overconfidence as actual ability. The overconfident job candidate is thus more likely to get the job than the more rational one.

- 'Posh white blokes' can - perhaps unwittingly - manipulate the social awkwardness of others for their own advantage, and thus progress at work.

The bottom line here is clear. In a class-divided society, the very notion of meritocracy is incoherent, because merit in the sense of academic achievement or career success might be the product of an overconfidence which is, initially at least, irrational and unjustified.