OYesterday I went to a talk on how to achieve Gender Equality by Prof. Iris Bohnet. I don't go to much of these and I was nervous as I took my boss along and it might not be great. I left with a page full of notes, wanting to tell everyone about what I just heard and my boss bought the book. Mission accomplished.
I never thought of my dad as a feminist, but I guess he was. I grew up wanting to be 'owner of Daddy's company' as in the boss and I never saw my gender as an impediment. Even though he imported construction machinery and there were no women in the business other than my mum, the assistants and the accountant, that never stroke me as much. When I gave him my grades, he always asked who had been better. At times I condemned him for it, but today I focus more on how gender agnostic he was. As I started looking for jobs he was keen to see me suceed and not settle for my first job (which I kind of did). He never took it that I would take anything else than the best for me. When my brother in law told him that I was going to work in the Barcelona of Investment Banking he rested I was ok to take it (Barcelona was at the top at the time). I went to football games as I grew up and had a real passion for it. Even though most of my friends did not I felt it as something special not just for me as a girl but as a father and daughter experience. Today I still go whenever I can even if he does not.
Even though I am not here to talk about my dad but about the talk, my dad came to mind many times as I heard the talk so I could not avoid the interlude.
The talk gave some well known examples of bias that we would swear we don't have. At Harvard, ai knew about the case study that has Heidi as a VC investors, and about how as an experiment the faculty changed the name of the protagonist for half the students. We are talking business school educated students. The name was Harry. Both were assessed as competent and successfull. But no one really wanted to work for or with Heidi, nobody liked her. Heidi did not fit the prototype. The experiment is mind boggling.
Whilst you could talk for hours on unconscious bias, the talk was about how you design for success. The NY (i think) Orchestra started auditions behind a curtain to focus exclusively on the music, and the rates of acceptances for women shot up. And this dates back to the 70s or 80s. An insurance company has concluded that claims are lower if people sign in the begi ning of the form rather than the end.
First, we talked about recruitment. The mere adjectives on the job description can determine the diversity of the applicants you get, so one can scan to remove them or make them inclusive. The interview process as pannel determines failure as people as a group in specific circumstances dont have the optimal outcome - is it the madness of crowds? Also, unstructured interviews make people less able to compare across candidates and look for someone that fits a role rather than for the best candidates. Structured interviews and forcing comparability of pre-defined attributes that are required for a role ensure a better processed.
We talked about promotions: how self evaluations seen by managers before they do their own evaluations create a bias. And the difficulty of assessing someone's potential - needed as the world is not exclusively backward looking but dangerous as potential is easiest to fullfill through the use of a stereotype,
She also talked about her studies of what works: diversity training? No evidence. Mentorship? Some evidence but only if it is long term and at significant investment by mentor and mentee. Most astonishing is the lack of metrics to know what works. If you give training to an already high potential female candidate to promotion and she gets promoted - are you really successfull? Would she have been promoted anyway?
Prof. Iris also spent quite some time on design. Design at school and at work. Changing the SATs deduction for wrong questions mechanism increases the chance that women will answer all the questions and do better. Practically, it was not legal to the remaining risk averse 50% of the population to guess, rather than just the men. The design of groups as companies is also key to achieve results of diversity, not just of gender or race but also opinion. I was surprised that political correcteness improves the functioning of groups, but I guess I then understand the constant definition of the HBS classroom as a safe environment - one can test an idea without being ridiculed or subjected to a joke, it probably makes people say more than they would. It is hard to define norms for groups to operate without restricting how they operate to the point it is detrimental. I would like to know more about this.
Next step: buy book and take practical steps to change now. We are all biased.
Comments