Academic Promotions

Suppose a university, or one of its components, for example a school, faculty or department, collects data on promotions and discovers that the rate at which women are promoted is statistically indistinguishable from the rate at which men are promoted. What could this mean?

Possibilities include:

1. There is no difference between men's and women's experience of the promotion process.
2. There is a difference but it is too small to resolve with the data available. (Attempts to aggregate data from different years need to take into account that the outcomes in year N are not independent of the outcomes in years N-1, N-2,... .)
3. There is a difference but it does not manifest itself in differences in promotion rates but in some other way, for example, increased stress.
4. Men and women tend to have different expectations of a promotion procedure. For example, it may be the case that women tend to believe that if they meet the criteria then they will be promoted while men tend to believe that if they are better than the other guy then they will be promoted.
5. Men and women find the promotions procedure equally unfair but women interpret their experience within a paradigm of women's disadvantage while men use a different paradigm, for example, managerial incompetence.

Well constructed staff surveys would help distinguish among these possibilities.

A particularly interesting situation occurs when it is widely accepted that women are less likely to be promoted than men, even though the evidence that this is the case is statistically unconvincing. Not only may it be accepted that women are less likely to be promoted but it may also be accepted that the reason is that women are less likely to apply for promotion, even though the evidence that this is the case is also unconvincing. In fact people may be so strongly convinced that women are less likely to apply for promotion that they propose research projects to investigate why this is the case. To a scientist this is an interesting concept: a proposal to investigate the reasons for an effect that you can't demonstrate actually exists.

When there is a widespread perception that women are disadvantaged by the promotions procedure then it is necessary to be seen to be addressing the problem, even if the data do not support its existence. In this case the 'women are less likely to apply' explanation has a political advantage. It throws the responsibility for the perceived situation onto the women involved which means that proposed remedies can focus on modifying the behaviour of the women rather than on altering the procedures or the way they are implemented. It is generally easier to persuade people of the case for modifying someone else's behaviour rather than their own. Why is this an advantage? Well, it is if it means something happens rather than nothing. Of course there are women who find the 'here is our special help for the ladies' approach offensive but in my experience they are out-numbered by those who will take whatever assistance is offered regardless of the motives of those who are offering it.