Gender, identity and competition

Why is there a gender pay gap among senior managers? It’s tempting to blame evil patriarchal capitalists. But there's more to it than this.
There’s good evidence that women are - on average - less aggressive negotiators than men, and so prone to get worse deals. Women are also less overconfident and so less likely to get top management jobs. And they are also less likely (pdf)  to want to enter into the sort of high-stakes competitions that allocate top jobs.
But why do these differences exist? A neat experiment sheds light.
Economists got some MBA students to do some mental arithmetic, being paid according to the number of sums they got right. They were offered a choice of payment methods. They could either be paid simply according to the number they got right, or they could be paid much more, but only if they got more answers correct than three rivals. So they could choose between piecework and competition.
Before they were given this choice, the subjects were asked to fill in a questionnaire. One group was given a questionnaire about their career, the other about family and gender issues.
And here’s the thing. Among those given the career questionnaire, men and women were equally likely to chose the competition - a quarter did so. However, when given the gender questionnaire, 37% of men chose competition, but only 7% of women did so.
This suggests that an aversion to competition among women is not some biological given, but is instead a product of their social identity. And a very simple reminder of this identity induces women to shy away from competition, and men to seek it. This is consistent with research by Uri Gneezy and colleagues, which has shown that, in a matrilineal society, it is women (pdf) rather than men who choose to compete. It’s also consistent with George Akerlof’s and Rachel Kranton’s under-rated book, Identity Economics, which shows how our identities and the social norms attached to them shape our economic behaviour.
And it’s also consistent with the feminist claim that gender is, at least in part, a social construct.

Comments

[David Mills]

Can you publish the figures?

37:7 when the gender questionnaire was given.

What was it for the other questionnaire?

Did the career questionnaire make men more likely to choose competition?

[jameshigham]

There’s good evidence that women are - on average - less aggressive negotiators than men, and so prone to get worse deals. Women are also less overconfident and so less likely to get top management jobs. And they are also less likely (pdf) to want to enter into the sort of high-stakes competitions that allocate top jobs.

There are many reasons they're less able - their personas are different and that affects how they do business and the nature of that business. As managers, women make great administrators.

[Jim]

Just remind me again which gender has the children?

[chris]

@ David - the career questionnaire led to 25% of men and 25% of women choosing competition.
@ Jim - not obvious why having children matters in my context. Having to look after others could in principle make one more greedy/competitive.

[Dom]

I think Jim meant that (although he can speak for himself, I'm sure, and never having met him) that there are obvious biological differences between the genders, such as childbearing. Right?

"And it’s also consistent with the feminist claim that gender is, at least in part, a social construct."

Hmmm. Let's say you find that chickens are less likely to solve a maze than dogs, but if you take chickens out for daily walks, their maze-solving behavior improves. Does that mean genus is a social construct?

[Mat]

Dom; I can't see that analogy. The women in the study Chris talks about are not given assertivness training (or somesuch) along with the career questionnaire, nor denied it with the gender questionnaire.

Surely a closer analogy would be;

Say that the dogs and chickens are lead through a hall of mirrors before entering the maze.

In the first instance, all the mirrors are rigged to show a dogs reflection (regardless of the genus of the animal in the hall); and in this instance there is no difference between the maze solving ability of dogs and chickens.

In the second instance, the mirrors are real mirrors showing the true genus. And in this instance chickens ability to navigate the maze falls well below that of Dogs.

Then that would imply that genus (or at least the maze-solving abilty thereof) was a social construct.

[Jim]

Exactly as Dom surmised. The primary reason for there being a lack of senior (and thus well paid) female managers is that the pool of potential applicants is rapidly thinned when women reach their late 20s and early 30s. The recent report on the gender pay gap showed that female managers in their 20s were actually paid more than the men, so any subsequent drop in female participation is not discrimination but a natural consequence of women leaving the workforce/downsizing their careers in order to have and care for children. When they re-enter the workforce/resume their career they will be competing with men who have anywhere up to 10 years extra experience. Is it any wonder that fewer women become senior managers?

[chris]

@ Jim - the fact that the gender pay gap exists only for senior managers is consistent with this experiment. Once they become mothers, women also become more aware of their traditional gender role, which induces them to be less competitive.
The way to test this against your theory is to see whether a gender gap among managers exists, even controlling for the experience which women lose by taking maternity leave.

[Recusant]

Well there is a control, Chris. Women who never have children and lesbians suffer no gender pay gap at all. At all levels. In fact they benefit from a gender pay gap in their favour.

[Tony Woolf]

Would be interesting if only the research was referenced.

[barry]

@Recusant Do you have a source for this? It seems to back up the evidence, in that women who have had to construct a different gender identity to that of the "expectation" would benefit.

Very good piece Chris, I'm reminded of a study done with US college students in which researchers showed how when a group of male and female students ate together, the women ate differently to a group of only female students (in fact I probably read that here). I've been trying to find the link, but I think it's another interesting example of socialy based gender construction.

[chris]

Sorry chaps. A link went AWOL. Here is the paper I'm referring to:
http://ideas.repec.org/p/cbt/econwp/11-31.html

[BenSix]

"...women also become more aware of their traditional gender role, which induces them to be less competitive."

No, they just transfer their competitiveness to parenting. "Your child can read the alphabet at 3? Well, mine is studying Dante at 4..."

[Torquil Macneil]

Gender is entirely social construct, that's what 'gender' means, the social designation of sex identity. That's why we talk of French nouns having 'gender' but not 'sex'. It is a pity that this distinction has got lost with nonsense talk of gender equality' and the like because it is useful and it is why feminism latched onto this word in the fist place. Ironically, I think the rise 'gender' and the attendant confusion comes at least in part from a prudish reluctance to mention 'sex' in mixed company.