Talent Acquisition

How to Beat Unconscious Bias in Recruitment

Is your hiring process prejudiced? Recent research from Mckinsey shows that a gender and ethnically diverse workforce produces greater company return than a homogeneous one. So why are we still hiring the wrong people?

Well it’s not for the lack of trying. Western companies spend billions every year on trying to de-clone their workforce. In 2015, Google alone invested $150 million into diversification. Improvements, however, are slow – and recent reports of Facebook’s failing efforts to create a more diverse workforce isn’t good news. After all, if Facebook can’t do it, no-one can.

So what’s the problem? Is the talent really not there? Of course it is. The problem, my friends, is not the candidates. It’s the recruiters.

Jennifer or John?

We all know the basics. Men dominate tech firms, C-suites and finance. Women rule primary care, counselling and HR. But is this imbalance caused by a genuine gender divide in good candidates? Or is it a problem with company hiring methods?

In an experiment conducted by Corinne Moss-Racusin of Skidmore College, more than a 100 STEM (Science, Technology, Engineering and Maths) professors from across the U.S. were asked to evaluate two CVs: one belonging to a Jennifer, the other to a John. The CVs were identical, apart from the name at the top. And yet, in the feedback given from the professionals, John’s competency was considered significantly superior to that of Jennifer. Not only were the professors less willing to hire Jennifer as a lab manager, but Jennifer was offered an average salary of $4,000 per year (13%) less than John.

Chances are, few of the professors involved in Moss-Racusin’s trial would have classed themselves as sexist prior to the experiment’s results. It was their unconscious gender bias that tipped the scales in favour of John. Nor is gender bias unique to STEM fields. Across all fields, recruiters subconsciously expect men to perform better in examinations, offer women lower salaries and tailor high-paid job adverts towards males over females, according to Medreps.

Of course, the scales tip the other way too. Particularly in fields such as primary care, as this author has written in the Guardian, men are much less likely to be recruited into roles than women. Whether this is down to significantly lower male applications, and male expectations of higher wages over modest teacher salaries, is up for debate.

Lakisha or Laura?

Racial bias remains a strong issue in many fields – and not always because of recruiter preferences. A study by NBER, similar to the Jennifer/John experiment but only using ‘ethnic minority’ names, showed that candidates with minority names are 33% less likely to receive a call-back for a submitted CV.

In fact, studies like this one in Fortune, indicate that many companies shy from hiring ethnic minorities for a more difficult reason: they fear their customers prejudices. Unfortunately, a study by Cornell found that many ostensibly un-prejudiced hiring managers tend to employ people of their own ethnicity – not necessarily because they are racist, but because we tend to empathise most with people similar to ourselves.

‘Who’s hired?’

There are plenty of other preconceptions that affect hiring decisions. Recruiters are, like all of us, human. They make assumptions based on height, weight and race.

Studies show that, thanks to the ‘halo effect’, attractive people are more likely to both be hired and progress rapidly within an organisation. The ‘affinity bias’ ensures that recruiters show preference for candidates with a similar background to themselves, whether it’s the school they went to or their preference for techno music.

And then there’s the contrast phenomenon – where we pick a candidate purely because their CV stands out from the others, whether the reason makes them good for the job or not. And finally you’ve got simple confirmation bias – that is, when you’re more likely to hire someone about whom you’ve made a previous positive assumption.

In the end, it doesn’t matter what your recruitment team’s unconscious prejudices constitute; they’re causing the best candidates to be passed over, and that’s neither fair nor beneficial to your business.

Taking action

The first step to ensure that you’re hiring fairly is to look at your typical job ad. Job postings can, famously, be misleading – and Harvard  found that they are often gendered too. Popular words like ‘ninja’ and ‘dominant’ are likely to discourage female applicants, while ‘caring’ and ‘collaborative’ will do the same for the boys. By ensuring your ad appeals to men and women equally, you will be reaching out to the maximum available talent pool.

A simple answer to name-based prejudice is to knock it out at the source. In October 2015,  the UK government announced plans for UCAS to begin considering student applications only with the attached names erased. This ‘name-blind’ approach, intended to start from 2017, would ensure candidates were not discriminated against for racial or class-related reasons at the application-reading stage. Such an approach could, of course, be extended by recruiters to include other details that should be irrelevant to the hiring process, such as age and gender – as several companies in Australia have recently demonstrated.

Recruiting firm TMPW suggests video interviews as a possible solution. Their reasoning? “Because multiple individuals can view the video, the probability of unconscious bias is reduced.” Of course, videos themselves offer possibilities for potential discrimination, by revealing the race, attractiveness and accent of the candidate involved. But, as TMPW acknowledges, “with the right amount of education and monitoring, this type of discrimination can be resolved, and companies can ensure that managers focus on the relevant criteria.”

There is, of course, a simple solution to all these issues: robotisation. Already most companies use some form of computerised selection to narrow down potential candidates from the pool of applicants. Over the next few years, even the final stages may be decided by AI. Apps like Joonko and Blendoor are already seeking to combat the bias inherent in recruiting processes. They’re inexpensive to use and could be a valuable training tool, if nothing else.

So next time you’re hiring, honestly ask yourself: can I be confident that no bias will be present in this process? If the answer is no, remember that there are things you can do to buff up your hire. Your candidates – and business – will thank you for it.

About the author: Susanna writes for Inspiring Interns, a graduate recruitment agency.

By Guest

This post is written by a guest author. If you are interested our sponsored content options, check out the the Advertising Page - we look forward to hearing from you!