We all know, especially those of us who work in business or are interested in equal rights (or both), how unconscious bias is supposed to work. There is a tranche of rhetoric and dogma which says that people – by this we usually mean white men, are innately biased against women and people from ethnic and racial minorities.
Implicit bias was also brought up in the Presidential election when Hillary Clinton complained about it in relation to her rival, Donald Trump. The narrative has been with us for a long time and has spawned a $8 billion a year industry, but is it real?
Not according to one study.
The Australian government’s Behavioural Economics Team undertook a randomized study of 21,000 people divided into two groups. Their shared task was to review a bunch of CVs/resumes and decide who to hire and who not to:
Group A: The CVs contained names and other identity markers like gender, nationality, and race.
Group B: Just the career and education details.
Logic, if unconscious bias is true, would tell you that Group B would deliver a fairer share for women and minorities. After all, this is what many activists are calling for – blind selection based on ability and achievements alone to remove unconscious or implicit bias from job applications to allow more women and ethnic/racial minorities to get better jobs.
The Results:
- 2.9% more women were shortlisted in Group A
- 3.2% less men were shortlisted in Group A
- 5.8% more minority males shortlisted in Group A
- 8.6% more minority females shortlisted in Group A
The results were not what was expected. It’s a big study so there’s a lot of validity there – a smaller study would be easier to dismiss. I do not know if the study was open about what they were doing or whether the participants were just asked to assess job applications and make hiring or not hiring decisions. The latter would be most likely as it is more rigorous.
What is certain is that it is more useful a study than the traditional ones so enshrined on the Harvard University website. Hopefully more, similar tests, can be undertaken by people all over the world divided into these two groups to determine whether this is the case or not.
On reading it, though, it did make me think that there may be other reasons for underrepresentation and it is always a good idea to see what alternative reasons might be. It’s unscientific to just take one solution without testing others after all.
And it also, as a final point, reminds me of anecdotal evidence I’ve picked up from talking to business owners, former managers, and so on for the last decade and a half. That evidence, while not scientific, has shown many examples where women are far less likely to promote other women than men. Instead of focussing on the unconscious biases of white men all the time, let’s look at everyone as a whole and within more specific markers and see where the problems lie. Ideally we want to always be able to hire based on merit and to feel that whoever we are applying to, no matter their own status, is going to judge us fairly regardless of ours.
Further Reading:
Singal, Jesse, 2017, Psychology’s Favorite Tool for Measuring Racism Isn’t Up to the Job, New York Magazine
Young, Toby, 2017, The Trouble with Diversity Training, The Spectator Magazine