How to remove all bias from the hiring processRemoving bias isn't the goal, or even realistically possible. Removing the wrong biasses should be the focus.
One theme I hear a lot among hiring managers and interviewers is the quest to remove all bias from the hiring process.
There’s really only one way I know of to actually do that:
Take all of the résumés you’ve received and put them in a stack. If you got electronic copies, print them out. Now take that stack of résumés and throw them up in the air as high and as violently as possible. Once they’ve all settled on the floor, pick the one closest to you. If two or more are touching your feet, pick them up and repeat, until exactly one is clearly closest to you.
The name on that closest résumé gets the job! Ta-da! Zero-bias hiring practice!
Of course this is an absurd way to hire people. But then, when you think about it, so is the concept of removing bias from the hiring process. In fact, the entire point of a selection process is bias. The key is, we want to bias for the right things. Not remove bias.
Pretending that our goal is to remove bias makes us blind to the actual process, and the actual biasses we introduce.
So here’s the point I really want to make: Every aspect of a selection process necissarily, and intentionally, introduces bias. Be mindful about whether it’s a useful bias.
Which biasses are introduced by photos on résumés? Which are good, and which are bad? This seems likely to introduce racial, gender, and age biasses into the hiring process. I’m sure we can agree that unless you’re running a modeling or acting agency, such a bias is bad.
What biasses are introduced by a take-home coding assignment? Which ones are good, which ones are bad? Maybe these assignments bias toward people with certain technical skills. Is that good? Maybe. It may also bias against people with free time. Is that good or bad? It might be bad. It likely biases against those with families, for example. It probably biases against people who simply don’t like these sorts of assignments. Is that good or bad? You decide.
What biasses are introduced by requiring a degree? Which ones are good, which ones are bad?
Which biasses are introduced by asking for a cover letter? Which are good, which are bad?
Every filter, including the unintentional ones, introduce bias. That is, in fact, the very definition of a filter. So are they introducing the right biasses?
Why do devs want more devs?
Almost always, more team mates means slower output, not faster. So what's the proper solution?
How to boost productivity without boosting headcount
We need to clear this traffic jam, so lets put more cars on the road...