Big Ideas

We can't get rid of bias—but we can disrupt it by design

Cheer 0
Share
Bookmark
3 min read
Canadian physicist Donna Strickland poses with her Nobel Prize during the award ceremony in 2018. That year marked the first time two female scientists won the Nobel Prize.
Canadian physicist Donna Strickland poses with her Nobel Prize during the award ceremony in 2018. That year marked the first time two female scientists won the Nobel Prize.
Photo by Henrik Montgomery | AFP | Getty Images

I n 2017, all nine people who won the Nobel Prize in the sciences were men. Historically, 97 percent of Nobel laureates in the sciences have been men, mostly from Western countries. Then, in 2018, something happened.

For the first time, two female scientists won the Nobel Prize in the same year: the Canadian physicist Donna Strickland and the American chemical engineer Frances H. Arnold. In addition, the human rights activist Nadia Murad, a Yazidi woman from Iraq, received the Nobel Peace Prize. These women, like so many before them, had done amazing work worthy of such high acclaim, but they were among the first to be recognized.

Many things changed between 2017 and 2018, including how nominations for the prizes are solicited. Research that I presented to the Nobel committees in February 2018 suggests that diversity is more likely to result when people make what we call “bundle decisions.” That is, when nominators submit multiple candidates each for consideration, the names of women and people of color are more likely to appear than when they submit just one candidate. Nobel nominators are now invited to nominate three people (or groups of people) deserving of the prize, instead of just one.

Even though implicit bias has seeped into so much of the basic machinery of society, it’s possible to be creative in circumventing that bias.
Design tweaks like this matter, not just for the selection of Nobel laureates but for all kinds of decisions. That’s because, even though implicit bias has seeped into so much of the basic machinery of society, it’s possible to be creative in circumventing that bias.

For example, some of our early-stage research indicates that informing managers of their past track record of promotion by gender can change the way they hire in the future. Most managers do not consciously disadvantage women, but it is hard for any of us to be objective.

Similarly, research shows that female musicians are more likely to be selected to join an orchestra when there is a curtain between the auditioning musicians and the selection committee—because selection committees find it easier to base their decisions on the quality of the music when they do not see who is performing it. Blind auditions helped increase the fraction of women in our major symphony orchestras from about five percent in the 1970s to almost 40 percent today.
What Works: Gender Equality by Design
What Works: Gender Equality by Design
This kind of design thinking is also starting to impact the way business is done globally, which is great news. Redesigning HR procedures based on research evidence should be a core priority of any organization. An increasing number of hiring platforms, including Applied, pymetrics, and Textio (to name but a few) are putting research insights into practice. Some firms are not only blinding themselves to the demographic characteristics of job applicants but also using algorithms to debias the language they use in job postings, designing tests that measure future performance more accurately than unstructured interviews, and overcoming the “groupthink” prevalent in panel interviews by requiring evaluators to make independent assessments.

However, we have to push the envelope even further and tackle the informal practices that shape our workplace cultures. Opportunities should be assigned on merit, not informally, based on whom people know or associate with—or whom they’ve associated with in the past. For example, in many law firms, associates face what is known as performance-support bias, or the “thin file problem,” whereby some associates come up for promotion to partner without ever having participated in high-profile deals. If they don’t appear qualified, it may be because they were never given the opportunity to make a splash. Fortunately, technology can help with these informal barriers, too. For example, meeting apps are now being developed that remind people about desired behaviors and also measure who is interrupting whom, and who gives credit to whom.

We all are biased. Trying to debias our mindsets is incredibly hard if not impossible. Instead, we need to design systems that are unbiased, playing fields that are more level, and opportunities that are given to all. Equality by Design is both the right and the smart thing to do for businesses and governments alike.

Posted: March 5, 2019
Edition: Forward


The ideas and views reflected in this article are those of the author and do not necessarily reflect those of Evoke or Melinda Gates.