April 5, 2017

Is your resume screening biased?

TL;DR: Yes, it’s very likely that your resume screening is biased. But don’t worry. There’s an app for that!

You might wonder: “Why would recruiters or hiring managers be biased in screening resumes? It’s in their best interest to hire the best person for the job.”

That’s true. But the pressure to hire the best person often leads recruiters and hiring managers to make bad decisions.

Resume screening lends itself to bias

People under pressure to avoid a bad hire look for clues that a candidate will “fit in” with the team.

How do recruiters do that? A trove of research confirms that we feel most comfortable with people that are similar to us (or the majority of our team). So we unconsciously look for clues that the person will “fit in.” Unfortunately, this often creates unconscious bias against women, people of color, and immigrants.

By often, I mean a well-documented, statistically significant, shocking percentage of time. We know this from decades of resume audit studies. In such studies, researchers manipulate resumes so that they are identical except for a key variable that they want to test. Here are just a few examples:

  • In a now classic study, economists at NBER found if they replaced the name “Greg” with “Jamal” on otherwise identical resumes, Jamal got 50% fewer callbacks than Greg. Researchers had to give Jamal 8 more years of experience than Greg to get the same number of callbacks.
  • In a field experiment with 13,000 resumes sent to job openings across a number of industries, researchers found that resumes with foreign-sounding names were 40% less likely to receive call backs than those with English names. Language fluency, multinational firm experience, education from highly selective schools, and active extracurricular activities did not reduce the negative impact of having a foreign-sounding name.
  • Sociologists at Northwestern and University of Toronto sent thousands of resumes to over 200 top law firms. They found that resumes from men who signaled “elite” status with extracurriculars like sailing and polo were advantaged in the resume review process. At the same time, women who signaled “elite” status were effectively penalized. “Elite” women received fewer callbacks than elite men. In fact, they got fewer callbacks than women and men with lower socioeconomic status. Again, the resumes were exactly the same except for these extracurriculars.

Of course, most recruiters and hiring managers don’t discriminate intentionally.

But if the majority of your team is white and male, it’s not surprising that you might make a subtle calculation that a candidate who has a male, white-sounding name would be an easier “fit.” (In reality, “fit” has more to do with shared values than superficial factors like your name).

Other stressors are in play. When a hiring manager is looking to hire, it’s likely because the team is already overworked. Recruiters want to fill roles quickly. For some positions, sheer volume makes fast resume screening a necessity. We get it.

The problem is that when we move quickly, we miss things. We skim, we generalize, we fill in the blanks with past patterns. I often see names on resumes that I can hardly pronounce. In the past, I’d wonder whether that person was fluent and comfortable in English. It doesn’t really matter if 2 seconds later I see that person has a PhD in English Literature from Harvard. (Or that my own legal name is completely unpronounceable! Yelizavetta?) When we see a name on a resume, we quickly make all kinds of assumptions.

Luckily, there’s a simple solution for the persistent problem of unconscious bias during resume review: review resumes blindly! It’s 2017. You don’t even have to use a sharpie, we have a fancy machine-learning based app that will take any resume you throw at it and redact identifying information.

If you’re alarmed by the shocking statistics on resume bias, there’s literally no excuse any more! Give Unitive a try.