Taking the bias out of recruitment

When recruiters sift through job applications, they take less than 10 seconds to decide whether someone will be lucky enough to get through to the next round. If your CV doesn’t grab their attention immediately, you’re done.

“Millions of people are getting hired and fired every single day,” says Kate Glazebrook, CEO and founder of the Applied recruitment platform.

“And if you look at your average hiring process, it involves usually 40 to 70 candidates applying to a particular job.”

Decisions made at this speed require shortcuts. They rely on gut feelings, which essentially are a collection of biases. Without even being conscious that they are doing it, recruiters and hiring managers discriminate because they are human, and they are in a hurry.

Glazebrook says most hiring decisions are made in the “fast brain”, which is fast, instinctive and emotional. “That’s the automatic part of our brain, the part of the brain uses fewer kilojoules so we can make hot, fast decisions,” she says.

“We’re often not even aware of the decisions we take with our fast brain.” The fast brain/slow brain concept references work by Nobel Prize-winning psychologist and economist Daniel Kahneman, who studied cognitive biases.

The “slow brain” refers to thought processes that are slower, more deliberative and more logical.

“And I think it’s clear that when you’re 69 candidates down, it’s 5pm on a Friday, you’re definitely less likely to be using your slow, deliberative part of your brain, and much more likely to be using your fast brain,” Glazebrook says.

‘We all overlook people who don’t look the part’

The inherent biases in traditional recruitment practices go some way to explaining the slow and limited progress of diversity and inclusion in our organisations.

“There’s, sadly, lots of meta-analyses showing just how systematically we all overlook people who don’t tend to look the part,” she says. “And there’s evidence to suggest that minority groups of all kinds are overlooked, even when they’re equally qualified for the job.”

Bias against people with non-Anglo sounding names was famously demonstrated in an Australian National University study of 4,000 fictitious job applications for entry-level jobs.

“To get as many interviews as an Anglo applicant with an Anglo-sounding name, an Indigenous person must submit 35 per cent more applications, a Chinese person must submit 68 per cent more applications, an Italian person must submit 12 per cent more applications, and a Middle Eastern person 64 per cent more applications,” wrote the authors of the 2013 study.

Lack of diversity is a business risk. According to Applied, diverse teams bring different ideas to the table, so that teams don’t approach problems in the same way. This tends to make diverse teams better at solving complex problems.

Consequently, an increasing number of employers are committing to “anonymised recruiting”. Also known as “blind recruitment”, this process removes all identifying details from a job application until the final interviews.

In the initial candidate “sifting process”, recruiters and hiring managers do not know the name, gender, or age of the applicants. They can also not make any judgements based on the name of the university or high school the applicants attended or their home address.

When the State Government of Victoria trialled anonymised recruiting for two years, it discovered overseas-born job seekers were 8 per cent more likely to be shortlisted, women were 8 per cent more likely to be shortlisted and hired, and applicants from lower socio-economic suburbs were 9.4 per cent more likely to progress through the selection process and receive a job offer.

According to academics researching the trial, “… at the Victorian Department of Treasury and Finance we found that before de-identifying CVs men were 33 per cent more likely to be hired than women. After de-identification, this flipped and women were eight per cent more likely to be hired than men”.

Connecting to brain function

Glazebrook is an Australian-born behavioural economist working in the UK’s Behavioural Science Team, when she co-founded Applied in 2016 with Richard Marr. They aimed to use their understanding of how the brain works to offer a beginning-to-end anonymised hiring.

Applied runs the whole process, from crafting bias-free job specifications and advertising, to candidate testing and selection. Beyond removing identifying details, the company also breaks up assessment tasks among a team of people and randomises the order in which elements are looked at – to minimise the impact of other cognitive biases.

Applied’s clients include the British Civil Service, Penguin Random House and engineering firm the Carey Group. In the past three years, the company has dealt with more than 130,000 candidates.

“We’ve seen a two to four times increase in the rate at which ethnically diverse candidates are applying to jobs and getting jobs through the platform,” she says.

More than half of the candidates who have received job offers are women and there have been “significant uplifts” in diversity in other dimensions, such as disability and economic status.

Glazebrook says US companies spend $US8 billion annually on anti-bias, diversity and inclusion training. However, even with the best intentions of everyone involved, it seems to have limited effectiveness.

“The rate of change is quite slow,” she says.

There is even some evidence that anti-bias training can backfire. Glazebrook says a concept called moral licensing is a concern: “Once you do the training, you tick a box in your brain that says ‘Great. I’m de-biased. Excellent. Moving on’.

“And, actually, you are free to be more biased than you were before because we’re led to believe we have overcome that particular bias,” she says.

Studies show companies openly committed to diversity are as likely to discriminate as those who aren’t.