Gender inequality & overconfidence

Why are there so few women in company boardrooms? There are, of course, many reasons. But a new paper highlights one that’s overlooked:

Gender differences in overconfidence concerning their own performance explains a significant proportion of the lack of female leadership

They established this experimentally. First, they gave 134 MBA students 150 seconds in which to add up as many sets of 4 two-digit numbers as they could.
 Then, 15 months later, they split the students into 33 groups and asked them to choose a representative to do the same adding-up task, with the best representative winning money for the whole group.
Very few women were chosen as representatives; only 4 out of the 33 groups chose one, which is only half as many as would be expected.
This was not because the women were worse at adding up. Instead, it’s because men claimed to be better at the task, and so the group chose them.
This was not because the men simply lied; men and women lied roughly equally. It’s because the men misremembered their past performance. When the researchers offered the students $50 if they correctly recalled how many correct answers they got, men over-estimated their performance by an average of 2.4 answers, whilst women over-estimated by only 0.9 answers. This over-estimate of past performance led to an overconfidence about prospective performance, and hence a greater likelihood of being picked.
This is consistent with other research, which shows that overconfident people are perceived to be better than they actually are.
This suggests that gender inequality can arise not from simple discrimination, but rather from gender differences in a particular form of irrationality - a form which happens to be favoured by the market.
If this is inconvenient for those who would try to justify gender inequality, there’s something else that is awkward for those who would oppose it. As Virginia Postrel points out, overconfidence can have many beneficial side effects, so it mightn't always be a bad thing for groups to overly favour the overconfident.

Comments

[Rob]

"This is consistent with other research, which shows that overconfident people are perceived to be better than they actually are."

Is that necessarily true? If you nominate someone else ahead of the overconfident person, and your nominee fails in the task, the overconfident person will have been proved right all along, and may harbour resentment for not being given the chance they "deserved". The rest of the group may regret the choice too. Maybe it's just easier to let the loudmouth have their way? In this case, we give them what they want not because we believe they're actually better, but because we perceive a higher cost to opposing them.

In the above example, the chance of winning the actual competition (a one-in-thirty-three chance) seems slim with the very high probability of pissing off the group loudmouth.

[john malpas]

Doing a few sums is 'research'!
Selecting for management skills would be totally different. It may look easy but it aint.
It sounds like you are angling for yet more
affirmative action.

[Business directory]

I think that is an imaginary problem

[Jeremy Poynton]

Missing the point. The coalition are doing nothing that Labour would not have done. Indeed, it was Labour who implemented tuition fees, Labour who commissioned the report which the Coalition have in fact watered down.

YET - there was no violence from the Left when the above happened. The violence happened SIMPLY as a result of the Tories being involved.

The hypocrisy is staggering, but then that is no less than we expect from the Left after a disastrous thirteen years which have brought the country to its knees. What is clear is that the country now knows that the NUS are a bunch of jerks, who are happy to hide thugs in their numbers.

[MarinaS]

The "overlooked" fact that "gender inequality can arise not from simple discrimination" is hardly the step child of modern academia; it's called evo-psych, pulls down vast sums in research funding, and is overwhelmingly statist bullshit.

Kind of disappointing to see this reported unclritically, or at least without mentioning complementary research that shows that women who *do* estimate their performance highly, e.g. by asking for promotion in the workplace, suffer disproportionate penalties in perception, and can even be held back in their careers.

There is more to why men and women are prone to behaving in certian ways than the oft-cited but as-yet-to-be-demonstrated "hardwiring".

[Brett Anderson]

I can tell you firsthand that there is a big difference in the way that men and women are treated. My wife is a school teacher and she deals with it on a daily basis. I've even noticed it with my company Killer Bee Printing. We are a printing company that specializes in cheap brochures and occasionally we've had customers that have lodged a complaint. When they talked to a male (politely) they were completely different than when they talked to a female (rudely). Sad but a fact of life.