As you point out, it's interesting how this problem is amplified by the dunning-kruger effect. I've also observed this issue to be worse in environments when success or failure is difficult to measure (or is only measurable far in to the future).
And I very much agree with your last paragraph: good managers and leaders are people who recognise competence and use it to further an organisation's goals. I don't mean that in the "take credit for your subordinate's idea" kind of way, as that's clearly an unsustainable strategy (the uncredited start withholding information and ideas, eventually).
On the game theory/evolutionarily stable strategy side of things though, I think it depends on context and perspective. If the context is one with a ubiquity of good managers, the pay-off for 'unjustified confidence' is probably pretty low or even negative. Also, as this is a multi-round game, the pay-off for 'unjustified confidence' gets lower and lower with each successive round (as good managers will factor past experience into their estimation of subordinate competence).
On the flip side, in a context with mostly incompetent managers, I'd agree the dominant strategy for individuals, competent and incompetent alike, is to 'act confident'. Although, as mentioned before, the competent are hobbled by the dunning-kruger effect here. Even if they weren't (i.e. competent and incompetent subordinates display equal levels of confidence), the outcome for the organisation would be indistinguishable from random decision making. In other words, you could replace managers with a coin (or a n-sided die, for problems with n possible solutions). And you don't have to pay coins or dice six figure salaries, making it the better option...
From the organisation's perspective, it seems crazy to rely on getting the competent people to 'be more confident'. In fact, it seems impossible. If your organisation needs to rely on 'confidence' to determine who is competent, then by definition it cannot directly discern which employees are competent. So how would it know which employees to encourage to be 'more confident'? This seems to reduce back to the '(at best) random decision making' outcome in the previous paragraph.
Funnily enough, if you factor in the dunning-kruger effect, in the 'bad managers' context the best way to make decisions (on average) would be to follow the advice of the 'least confident sounding' employees.
The only viable (albeit less hilarious) solution from the organisation's perspective (and therefore dominant strategy) is to try very hard to only hire competent leaders and managers.
> If your organisation needs to rely on 'confidence' to determine who is competent, then by definition it cannot directly discern which employees are competent.
That's correct specifically because in the absence of the ability to judge competence (the original research by Dunning and Kruger showed that those low in competence overestimated the competence of others as well as their own), confidence becomes the stand-in. So over time if you don't start with the most competent people in positions of leadership, and maybe even if you do, you end up with the most confident (and probably least competent) people in them over several refresh cycles.
And I very much agree with your last paragraph: good managers and leaders are people who recognise competence and use it to further an organisation's goals. I don't mean that in the "take credit for your subordinate's idea" kind of way, as that's clearly an unsustainable strategy (the uncredited start withholding information and ideas, eventually).
On the game theory/evolutionarily stable strategy side of things though, I think it depends on context and perspective. If the context is one with a ubiquity of good managers, the pay-off for 'unjustified confidence' is probably pretty low or even negative. Also, as this is a multi-round game, the pay-off for 'unjustified confidence' gets lower and lower with each successive round (as good managers will factor past experience into their estimation of subordinate competence).
On the flip side, in a context with mostly incompetent managers, I'd agree the dominant strategy for individuals, competent and incompetent alike, is to 'act confident'. Although, as mentioned before, the competent are hobbled by the dunning-kruger effect here. Even if they weren't (i.e. competent and incompetent subordinates display equal levels of confidence), the outcome for the organisation would be indistinguishable from random decision making. In other words, you could replace managers with a coin (or a n-sided die, for problems with n possible solutions). And you don't have to pay coins or dice six figure salaries, making it the better option...
From the organisation's perspective, it seems crazy to rely on getting the competent people to 'be more confident'. In fact, it seems impossible. If your organisation needs to rely on 'confidence' to determine who is competent, then by definition it cannot directly discern which employees are competent. So how would it know which employees to encourage to be 'more confident'? This seems to reduce back to the '(at best) random decision making' outcome in the previous paragraph.
Funnily enough, if you factor in the dunning-kruger effect, in the 'bad managers' context the best way to make decisions (on average) would be to follow the advice of the 'least confident sounding' employees.
The only viable (albeit less hilarious) solution from the organisation's perspective (and therefore dominant strategy) is to try very hard to only hire competent leaders and managers.