Imagine coming home after work each night to a cute 20-year-old woman who’s so excited to see you that she calls out, “Missed you, darling!” She’s always immaculately dressed, and the two of you never fight. If this sounds like a creepily retrograde, heteronormative male fantasy, well, it is. The designers at the Japanese software company Gatebox might just have had a certain target market (born of the country’s unique social circumstances) in mind when they developed their holographic virtual assistant, Azuma Hikari. But they might also have revealed their own unexamined gender biases in the process.
There’s nothing intrinsically bad about that; bias—the tendency to favor one thing over another—is just a name we give to some of the mental shortcuts our brains use to help us navigate complex environments that bombard us with far more information than we can consciously cope with. “The problem comes,” says Tiffany Jana, coauthor of Overcoming Bias: Building Relationships Across Differences, “when you allow unconscious biases—or blind spots—to influence your behavior and the way you treat others.” This is a reality that businesses know all too well as they struggle to craft more inclusive workplaces. And it’s one that millennials—the most diverse generation in the workforce right now (at least until gen-Z enters offices in greater numbers)—have grown up hearing about.
But one thing that gets lost in conversations about unconscious bias is that it can run it any number of directions. It’s not just about a majority group holding implicit, negative beliefs about a minority group. If millennials are going to successfully make the modern workforce a less biased place, the first step will be grappling with that complexity.
How? One good place to start has been around for nearly two decades. Launched at Harvard University in 1998, Project Implicit offers a series of “association tests” (IAT) designed to uncover participants’ unconscious biases about other demographic groups. So I asked four volunteers, aged 28–30, to go online, chose any of the 14 topics offered, and answer a series of questions Harvard researchers designed to probe “thoughts and feelings outside of conscious awareness and control”—then talk about the results.
Professional Stereotypes Can Get Personal
Eddie, a Cuban-American law student, picked a test dealing with body weight. His results showed a “strong preference” for thin over obese people, a bias shared by three-quarters of the IAT’s sample of web participants. Having recently (in real life) met a “rather rotund” real estate broker, Eddie now found himself wondering whether “being overweight and rather disheveled in appearance will translate into some kind of professional disorganization.” But he said he didn’t consider this assumption biased—it was just a way for him to “read people quickly and navigate the world effectively.”
Eddie’s choice of test was personally relevant in another way. About to start law school himself, Eddie confessed that he’s especially concerned these days with “maintaining a professional image” because “being overweight doesn’t fit the image of the law, which is a very conservative field,” and could be “a hindrance in, say, trying to sway a jury.”
It doesn’t matter whether or not Eddie is right about this—it’s a belief at least partly informed by his own career ambitions that influence his perception of others. But while Eddie’s results squared with IAT’s test sample, the other three millennials proved to be outliers, underscoring how “bias” isn’t as monolithic or one-directional as it’s commonly understood. It also appears to highly influenced by one’s family circumstances and upbringing.
“Overcompensating” For Attitudes About Skin Color
After taking a test focusing on skin tone, U.S.-born Priti, who’s from a Southeast Asian family, learned she has a “slight automatic preference for dark-skinned over light-skinned people,” according to the IAT, a result found in just 7% of the test’s web sample.
This didn’t surprise her, though. Priti is only too aware that Indian standards of beauty typically favor lighter skin. Because her dark-skinned mother was teased as a child in Mumbai and still talks about it, Priti was brought up to treat everyone the same. “It seems I’ve overcompensated to the other end of the scale,” she says.
But it wasn’t until we discussed her IAT results that Priti realized she actually finds it easier to visually recall and describe people of color than those with lighter complexions.
Your Gender Alone Doesn’t Determine Your Gender Biases
Both Heather, a white woman, and Matt, who’s African-American, took Harvard’s “gender career”–based IAT and wound up with different but similarly contrarian results. Heather, who works in data management for a research firm, was found to hold a “slight automatic association with male and family, and female with career,” typical of just 5% of IAT takers. While this is the reverse of the stereotype, Heather says it makes sense to her “because my father stayed home to be my primary caregiver while my mother worked.”
Matt, on the other hand, showed little or no automatic association between family and professional spheres by gender, a lack of gender bias that’s common to just 17% of IAT’s web sample. Matt was raised by a single mother and strongly influenced by his aunt, “both of whom were working professionals and incredibly entrepreneurial.”
Now a diversity educator, Matt considers his IAT results to reflect his belief that women are “very capable leaders” as well as his commitment to helping others understand the systemic, institutional nature of bias so they can begin to eliminate it for themselves.
Making Sense Of Bias In All Its Forms
Matt’s colleague at Wake Forest University, Shayla Herndon-Edmunds, has been using the IAT for six years in her role as director of diversity education. She asks students, faculty, and staff to take the test so that participants can process their results privately before joining group discussions—which she calls a “high-risk activity.” Herndon-Edmunds sees this awareness as a crucial first step, because you can’t work with others to break down biases if you haven’t first broken down your own.
Tiffany Jana agrees, adding that what my four volunteers did easily and quickly is crucial: Trying to make sense of their biases by tracing their roots through personal experience. Usually, she says, this takes time. Many people react to their IAT results dismissively, while others just feel anxious about less-than-positive beliefs they didn’t know they held.
But Jana says all this is normal. “Numerous studies show that levels of bias are consistent across generations when controlled for age. But to be fluent in talking about and owning biases openly, as these four volunteers have done, appears to be more common for millennials compared with older generations.” And given the extent to which biases still pervade the tech community especially, that has got to be a good thing.