Tuesday, August 07, 2007

Mad Or Sad?

A recent study by Victoria Brescoll suggests that women in management might do better if they don't show anger. This is what she did in the study:

She conducted three tests in which men and women recruited randomly watched videos of a job interview and were asked to rate the applicant's status and assign them a salary.

The people in the job interview were either male or female but followed the same script. Each of them expressed either anger or sadness over losing an account due to a colleague being late for a meeting.

According to Brescoll, anger gave the men higher status and higher salary estimates. The reverse was true for women. It seems that it's better to be a sad CEO than a mad CEO if you are of the girly persuasion.

The usual popularized take on this study seems to be that anger helps men get ahead but keeps women back. Strictly speaking, this is not quite what the study found, as the study compared anger not to, say, general pleasantness, but to sadness as a possible reaction to one particular type of event (the loss of an account).

I couldn't get hold of the actual study which limits my ability to judge how reliable the findings are. As an example of the kind of data I'd love to see is the number of subjects who were asked to rate the videos. But supposing, for the time being, that there are no hidden bugs in the study, what do the results mean?

Brescoll herself says that angry women are seen as "out of control". This is fascinating, because the proper antonym for that in this context might not be "in control" but "under control". If anger is viewed as a legitimate reaction by someone who is entitled to dominate others (as the article from which I quote suggests), then those who are not seen as entitled to dominate others would indeed be "out of control" when expressing anger. This doesn't apply only to women who have traditionally not been allowed to show much anger, but to almost anyone in a subordinate position.

That sadness works better for women than anger fits into that framework pretty well. Besides, sadness is not directed to anyone outside the person. It is non-threatening. On the other hand, women are not supposed to be sad in public if they have power, because crying is seen as a sign of weakness. A Catch-22 indeed.

I was astonished by the salaries men and women were assigned in this study. Men were awarded much higher salaries than women (remember, the applicants were objectively identical), and the resulting differences are much larger than the actual average differences in full-time pay by gender. Both male and female reviewers assigned women lower earnings, and I have seen similar results from other studies.

One interpretation would be that the study subjects are trying to guess what a person with particular characteristics would earn in the real world, and assign sums accordingly, thus taking into account their knowledge that women, on average, earn less. Another one is that people in general are a lot more sexist than we perhaps have thought and rate and rank women lower than men on that basis.