Do Snap Judgments Amount to Bias in the Workplace?
Stay up to date with the latest headlines via email.
The following is an excerpt from Giving Notice: Why the Best and Brightest are Leaving the Workplace and HOW YOU CAN HELP THEM STAY ," by Freada Kapor Klein (Jossey-Bass, 2007).
Stare at these words for a moment.
Did you read: THE CAT? Most people do. Now look again. Notice that the symbols for H and A are not actually letters -- they're identical, nonspecific symbols. This wasn't a problem at first glance. Our brains filled in the information we needed, using pattern recognition based on past experiences. This is an inherent part of being human. At one time, this pattern recognition was a survival mechanism: Red mushrooms make you sick. Red mushrooms make you sick again. Stop eating red mushrooms. In the starkest Darwinian system of natural selection, you either figured it out by recognizing the pattern or you died.
When people consider the patterns they live by, most would deny that they include stereotypes. And it's true -- after hundreds of years of the most egregious racism imaginable, most people in this country today are not overtly biased. It's true in the corporate world too. Ask a group of CEOs, and they'll tell you that they can find genuine talent regardless of gender, race or sexual orientation.
But consider the example (above) when your brain automatically filled in the information that was missing. And then consider this: What information is your brain implicitly providing when you walk into a conference room and see a person dressed a certain way, a person whose skin is darker or lighter, a person whose hair or size or style or age is different than what you are used to?
On a cool fall day, after spending the weekend in the office wrapping up an intense but creative project, Eric Johnson shrugged off his coat as he stepped into his manager's office to talk about his next assignment. He was hoping to be given a lead role with a new initiative that looked like a sure money maker. He was more than hoping, really.
Eric's patterns continued to be influenced by his childhood in Detroit, when his parents, both auto factory workers, taught him to live in shifts. He saw his life in strict time blocks for both his personal and professional goals. He knew life wasn't fair, and he'd seen how the ups and downs of the auto industry -- the epitome of big business -- impacted his own day-to-day existence while growing up.
But now he was playing that corporate game, fueled by career ambitions, working to understand it, tame it and win it. Eric knew that his success on his prior project, coupled with his track record for hard work and creativity, put him at the top of a small heap to head the new assignment. So he was stunned when his manager told him he had already decided that leadership spot should go to Eric's straight, white male colleague -- a man with slightly lower productivity and accomplishments than Eric, but a chipper man, a good worker, and a positive and friendly person.
Eric's manager, when he made this decision, wasn't blatantly thinking: "Eric's black and gay, so I can't put him in charge." In fact, he considered Eric to be a talented and motivated team player; it's just that he wasn't all that comfortable with Eric. There wasn't any perceptible tension or discomfort. Indeed, Eric's manager prided himself on his open-mindedness. He made sure that Eric was given assignments he could handle, and that if they slipped a day, it wasn't mission critical. Yet Eric's manager never stopped to reflect that Eric had never missed a deadline and was often completing assignments early and offering to help out his colleagues. Completely unconsciously, Eric's manager assumed that Eric had been an "affirmative action hire" (someone who wasn't as qualified as his peers), and acted accordingly.
"It's nothing against you at all, Eric. You're doing great. Just stay focused on your current projects. Mark is a better fit: He's worked with many of these business units before and went to the same school as Chip, the big boss. Since they're both such loyal alums, I thought it would actually help all of us," his manager explained.
What happened to Eric was neither overt bigotry nor an anomaly. Researchers have now refuted the notion that only racists use stereotypes and instead confirmed the uncomfortable fact that stereotypes are an inherent part of how we all relate to each other.
In the workplace, the science is clear: Unconscious bias routinely creeps into those "blink," split-second decisions in the office, impairing business leaders' ability to make intelligent, intuitive judgments.
[T]he Level Playing Field Institute, [which promotes innovative approaches to fairness in higher education and workplaces by removing barriers to full participation, has] an exclusive partnership with Project Implicit, a collaborative research initiative aimed at examining the thoughts and feelings that exist either outside of conscious awareness or outside of conscious control. The project's key tool is the Implicit Association Test (IAT), an online exercise (take it yourself at implicit.harvard.edu) that measures how quickly it takes a person to respond with positive or negative words to photos of certain types of people. This striking research, which is gaining widespread attention for its provocative findings on unconscious bias, is one of the most objective systems devised to date to quantify prejudice. We all have prejudices based on learned patterns -- that is, we prejudge a situation or a person based on cues from our past experiences. Not all prejudices are bad. Some are simply preferences, such as favorite colors, songs, or art.
Psychology professor Mahzarin Banaji, who helped develop the IAT, was surprised to discover her own biases when she took her test. But, in fact, her experience was typical. Almost all of the people taking this test describe themselves as unbiased at the onset; yet a whopping 88 percent of white people who take the test show some bias against blacks, and a majority of people who take the test show bias against photos of people who are overweight, gay, elderly and Arab/Muslim.
So how does this impact the workplace? Actually, the better question might be: How doesn't it impact the workplace? From letters of recommendation, resumes and hiring interviews to promotions, wages and job assignments, the unintended but inherent biases of our corporate leaders throw up barriers that are not the blatantly discriminatory practices that can be fought in the legalistic framework. These barriers are, perhaps, even more insidious, since they remain the largest impediment to success for people of color, women, and gays and lesbians in the United States.
It starts with resumes, before anyone even sees anyone else's face. In 2004, researchers at the University of Chicago Graduate School of Business sent close to 5,000 fictitious resumes in response to help-wanted ads in Boston and Chicago newspapers. They randomly assigned very white-sounding names (such as Emily Walsh or Greg Baker) to half of the resumes, while using African-American sounding names (such as Lakisha Washington or Jamal Jones) on the other half. The results? White names received 50 percent more callbacks for interviews, regardless of occupation, industry and employer size.
"Taken at face value, our results on differential returns to skill have possibly important policy implications. They suggest that training programs alone may not be enough to alleviate the racial gap in labor market outcomes," the researchers wrote in their conclusions. Theirs was one of many similar studies -- a growing field of sociological "resume studies" -- conducted in recent years. In 2003, for example, a sample of 236 undergraduates (most of whom were white women) rated resumes having equal qualifications in which gender, masculinity and femininity, and sexual orientation were apparent. Overall, the participants -- especially those who described themselves as "religious" -- rated lesbian and gay male applicants less positively than straight men but more positively than straight women.
The bias is at least as pervasive in academia. In 1999, in a study still widely cited, University of Wisconsin-Milwaukee researchers sent altered curricula vitae with the names Brian or Karen Miller on the top to 238 academic psychologists. Both men and women were more likely to vote to hire the man, rather than the woman, even though their records were identical. Similarly, both genders reported that the fictitious Brian Miller had adequate teaching, research and service experience compared to Karen Miller, who, they suggested, needed more work.
Letters of recommendation
In the case of our own Kristen Van Der Kamp, it was her letters of recommendation that showed inherent bias. Kristen's scholarly prowess had launched her from thrift-store poverty on a farm to an academic scholarship at Harvard University, where she earned her B.A. and M.B.A. Graduating among the top in her class, she was juggling multiple awards and job offers, but she still needed letters of recommendation, which she sought from several of her key advisors.
Their responses were sincere and supportive, but they were also patronizing. They described her as "an earnest farm girl" and said that she had a "traditional hard-work ethic." These were professors, both men, who knew Kristen well, so they referred to her by her first name: "Kristen is an intelligent young lady," and "Kristen works hard, but is able to maintain balance in her life."
Inadvertently, those professors were sending all sorts of unwritten messages in their letters, and managers who had already decided to bring this promising M.B.A. into their departments read between the lines while making their assignments. Key phrases, like "balance in her life," were translated as: "Watch out, she might want a family and may be less inclined to work long hours." "Intelligent young lady" translated to: "She's very smart, but not a leader." The phrase also signaled "not threatening." "Earnest farm girl" meant "naive, not a shrewd negotiator."
These professors wanted the best for Kristen, but their letters led her new manager to steer her toward smaller clients -- children's hospitals, nursing homes and retailers -- while several male counterparts were assigned to high revenue-generating accounts like oil and gas and financial services.
These types of gender-based assignments are not unusual and have prompted sex discrimination litigation in recent years, including claims against Morgan Stanley, Costco Wholesale Corp., Boeing, and retail giant Wal-Mart. In Kristen's case, as it is for many women, being assigned to less lucrative clients meant her chances of being made a senior manager were greatly diminished. At the end of her first year, as Kristen was given her first in-depth performance review, she tried to identify what misstep had set her behind several men in her hiring class. It never occurred to her that it could possibly be a few phrases in her initial letters of recommendation that had set her on an invisible, slower track.
But Kristen's situation is one that is shared by many women and people of color who depend on letters of recommendations to swim up their professional streams. The leading research on recommendation letters, published in 2003, studied 300 letters submitted for faculty positions at a large American medical school in the mid-1990s. At the time, despite a greatly expanded pool of female applicants and students, women accounted for only 32 percent of the assistant professors, 21 percent of the associate professors and 10 percent of the full professors. The researchers, who were from Wayne State University, found that recommendation letters for women were consistently shorter, and were far more likely to include what they termed "doubt raisers:" phrases like "lacks confidence at times" or "has been limited by personal issues."
Their findings made it clear that if you're writing letters of recommendation, take care not to include biased or stereotypical language. And if you're reading them, consider the implicit biases within. Take a specific expression from a letter of recommendation, imagine changing the gender or race of the applicant and then ask yourself, "Would that same expression still be used?" For example, are men ever described as "perky?" (This term found its way into a significant number of reviews for women at a top-tier professional services firm that was a client of mine.) Similarly, how often do we use "qualified" as the adjective in front of anything but "minority"? Do we even mention race at all if the person is Caucasian?
For those who do make it into the job interview, the implicit biases are even more of a challenge. There is a great deal beyond education, experience, eloquence or even clothes and makeup that impact the hiring decision. And only some of these factors are within an applicant's control.
Studies of orchestra auditions have repeatedly shown that women are more likely to be chosen if the conductors doing the hiring can't see them. One such study indicated that blind auditions have accounted for 25 percent of the increase in female orchestra musicians. Because of this, most orchestra hopefuls now audition behind curtains. An accomplished violinist could be dressed in pajamas at her audition and still be hired. All that matters is her music.
In most workplaces, however, the hiring decisions are far less objective. Research conducted at the University of Toledo shows that individuals formulate opinions about a candidate within the first 20 seconds of the interview and that these first impressions will likely determine one's final evaluation. In the study, naive observers watched the first few seconds of 59 job interviews, which included the interviewees as they were greeted by the interviewers and escorted to a seat. The clip ended before the first prepared interview question was even asked, and yet the ratings given after watching this "thin slice" of behavior were similar to those made by interviewers after a 20-minute structured interview. The researchers suggested that our immediate, snap judgments become self-fulfilling prophecies that influence our behavior toward an individual and cause them to appear in a manner consistent with our initial impression. It appears that these snap judgments are "the most obvious threat" to the legitimacy of the interview process.
Miguel Rodriguez, our business school success who had fought his way out of a crowded barrio apartment toward a corporate career, was ready for snap judgments during his first job interview, but he faced an unexpected challenge. His awkward moment came when the interviewer switched to Spanish for a friendly question.
"Â¿DÃ³nde naciÃ³ Usted?" she asked.
Miguel wasn't fluent in Spanish, but he understood enough to know that she was asking where he was born. But was there more to this question? Was she wondering if he was an immigrant? Or testing his minimal Spanish language skills? Perhaps she was just being friendly, or trying to develop a bond. If he told her that he was born in a barrio apartment in Harlem because his mother didn't have medical insurance, would that work against him? Wasn't it equally honest to say "Manhattan"? Or perhaps this was an opportunity to explain that he wasn't Mexican. Miguel himself assumed that Mexican Americans were considered less ambitious than Cuban Americans.
Miguel was quiet for an awkward moment and then answered with a polite smile in English, "I'm from New York." The interviewer didn't delve further. She made her own assumptions, never asking if he spoke Spanish. Nor did she ask if he would be interested in working on Latin American projects. Those unwanted assignments simply came his way once he was hired.
Copyright 2007 Jossey Bass
Freada Kapor Klein is the author of " Giving Notice : Why the Best and Brightest are Leaving the Workplace and HOW YOU CAN HELP THEM STAY" (Jossey-Bass, 2007) and founder and board director of the Level Playing Field Institute , which promotes innovative programs focused on revealing and removing barriers to fairness from the classroom to the boardroom. She has worked on issues of diversity for 30 years, having co-founded the first group on sexual harassment in the United States.