How the Unconscious Mind Can Act Out Our Prejudices
Stay up to date with the latest headlines via email.
The following is an excerpt from The Hidden Brain: How Our Unconscious Minds Elect Presidents, Control Markets, Wage Wars, and Save Our Lives by Shankar Vedantam (Random House, 2010).
Lilly Ledbetter's life followed a clockwork routine. When she worked the night shift at the Goodyear Tire & Rubber Company plant in Gadsden, Alabama, she came home from work around nine-thirty in the morning. She took a hot bath, laid out her work clothes for her next shift, and slept until the afternoon. At around five, she set off again for the plant. Her shift did not start until seven, but the route from her home in Jacksonville involved a stretch of about ten miles on a country road, where she sometimes got stuck behind a slow vehicle.
The rubber plant was solid, stolid work. During each shift, managers such as Lilly were given instructions that told them what they needed to get done. If the instructions-which were called a “schedule”-required Lilly's team to make tread belts one night, she had to make sure she had all the components and labor in place before the shift started. The shift ran for twelve hours, but Lilly usually got to the plant early and stayed late. Lilly had once worked for a group of gynecologists, but she felt she was not cut out for medical work. She joined Goodyear when she was forty, straight out of a position with H&R Block. The tire company was a man's world, and Lilly was the only female manager on the night shift. It didn't bother her. If she kept her head down and worked hard, she knew she would be treated the same as everyone else.
One evening in 1998, Lilly reached work around six o'clock. She went to an upper floor where the managers had their mailboxes. There were a number of documents in her mailbox, and inserted among them was a torn sheet of paper. On the fragment were four names-first names. One was hers. The others were area managers who worked with her, doing identical work. The four managers supervised identical crews, worked the same hours, handled the same responsibilities, and had the same level of experience. Lilly was the only woman in the quartet.
Next to the names were numbers. Lilly instantly recognized the number next to her name. It was her salary: $3,727 a month. She looked at the other numbers and instantly felt sick. The other managers' salaries ranged between $4,286 and $5,236 a month. Lilly made $44,724 a year. Her co-workers made $51,432 to $62,832 a year.
Lilly's cheeks flushed. She looked up to see if anyone was watching her. No one seemed to be taking any notice. Lilly rushed to the women's room and collapsed on a sofa. She stared at the paper. She did not feel angry; she felt ashamed, small, and humiliated. “What am I going to do?” she asked herself. “How do I do anything?”
She slipped the paper into her pocket, determined not to show her feelings, but all through her shift the realization that her company valued her so much less than her co-workers gnawed at her. She crossed paths with another manager, a man who was being paid much more to do the same work. Lilly said nothing, but she ached inside.
Lilly did not want to think of herself as a victim of discrimination. Over the years at the Gadsden plant, there had been people who'd been nasty to her, but there had also been plenty of people who'd been friendly. There had been times, for example, when her supervisors had failed to tell her what her team needed to get done during a shift, even as her colleagues received their schedules. She would fall back on personal relationships to get out of the jam-friends in the scheduling department would pass along the instructions. When she had run-ins with supervisors, she put it down to individual chemistry. When a department foreman told her, “God damnit, Lilly, your department looks like a whorehouse!” Lilly coolly told her supervisor, “I don't know. I have never been in one.”
Another time, when corporate bigwigs from Akron came down to Gadsden for a visit, Lilly learned that two of her colleagues had been invited to meet the bosses at a social event after work. She asked a manager what time she needed to be at the gathering, and was told she didn't need to attend. She stayed home and fretted, but there was not much else to do-it wasn't like she could barge into a social event uninvited.
After some poor performance reviews, a supervisor told her that if she would only go “down to a local hotel with him,” the reviews would start saying Lilly was a good worker. When Lilly complained about sexual harassment, the company changed her supervisor, but she felt she was thereafter branded a troublemaker. She got left out of meetings, which made it harder for her to do her job. Setbacks usually made Lilly more determined. She had grown up on old Westerns, where the cowboy gets spat on and cursed at but keeps his cool, and the world comes around in the end. Whenever she felt dispirited, Lilly told herself, “I'm not a quitter.”
Despite her determinedly sanguine attitude, Lilly did sometimes suspect she was being paid less than her co-workers. She heard rumors, for example, that some of her colleagues were making twenty thousand dollars a year in overtime. Lilly worked overtime hours that were as long as anyone else's, but she did not make nearly that much money. The fragment of paper she received-nineteen years after she started working at Goodyear-was the first piece of tangible evidence to support her suspicions. Lilly was working a shift with a complicated two-week cycle; the next time she got a day off that fell on a weekday, she drove an hour west to Birmingham, to the Equal Employment Opportunity Commission, and filed a complaint. The EEOC was backlogged but told her that she had a strong case and encouraged her to work with a private lawyer.
In the lawsuit that followed, Lilly learned she was earning seventy-nine cents to the dollar of her male counterparts. The numbers on the piece of paper were only base salaries; they had other consequences. Her salary determined how much she got paid for overtime. It determined how much she could set aside in her personal retirement account, and how much she contributed to Social Security. Eventually it affected how much pension she received, and all the other sources of her retirement income. Lilly figured if she had been compensated fairly, her income in retirement would have been twice as large as it actually was.
Goodyear countered in court by producing a number of Lilly's performance reviews, which were below par. Lilly's salary had lagged behind her counterparts', the company argued, because she was an underperformer. Lilly argued that she had been evaluated unfairly because she was a woman. She pointed out that in 1996, Goodyear itself had given her a top performance award. The company put a manager on the stand who said the company had given Lilly the high rating because her salary had been lagging behind her peers and the company had been trying to justify giving her more money. Far from discriminating against Lilly, Goodyear suggested, the company had been biased in her favor, and had given her merit raises she did not deserve.
The case eventually made its way to the Supreme Court, which dis- missed Lilly's complaint on the grounds that the discrimination she alleged had taken place a long time before. A central tenet in the law, the Supreme Court ruled, is that people need to file complaints promptly. Chief Justice John Roberts openly worried that if the Supreme Court were to consider Lilly's allegations, the courts might be flooded by cases from people alleging discrimination in earlier eras. Goodyear had denied Lilly only two raises in the 180 days before she first reached out to the EEOC, and the court ruled there was insufficient evidence to show that those two decisions constituted discrimination. Lilly pointed out the obvious: She could not have filed a complaint earlier in her career because until she found that scrap of paper in her mailbox, she did not have evidence she was being paid less than her co-workers.
Gender issues were an undercurrent in the decision. Justice Samuel Alito, who had recently replaced Justice Sandra Day O'Connor, the first woman to serve on the Supreme Court, wrote the majority opinion. Justice Ruth Bader Ginsburg, who was then the only woman on the court, dissented along with Justices Stephen Breyer, John Paul Stevens, and David Souter. In an unusual move, Ginsburg read her dissent aloud from the bench. “In our view,” she said, “this court does not comprehend, or is indifferent to, the insidious way in which women can be victims of pay discrimination.”
Ginsburg personally sympathized with Lilly's long silence at Goodyear. Like Lilly, the Supreme Court justice had herself gone years being silent about the effects of sexism in her career. It was the norm for women of Ginsburg's age not to go to law school; the norm that if they did go to law school, they would have a difficult time getting good clerkships; the norm that they would be paid less than their male counterparts. The fact that Lilly had not complained for nearly twenty years about her treatment at Goodyear was unremarkable to Ginsburg. When the Supreme Court justice went to law school at Columbia in the 1950s, there were no women's bathrooms in the building. “If nature called, you had to make a mad dash to another building that had a women's bathroom,” she recalled as she discussed her feelings about the Lilly Ledbetter case. It was “even worse if you were in the middle of an exam. We never complained; it never occurred to us to complain.”
Lilly Ledbetter never found out who left the scrap of paper in her mailbox. Her case-decided on the timing of her complaint rather than on its substance-became a signature example of the capriciousness of discrimination laws, and the failure of such laws to account for real-world circumstances. One of the first bills signed into law by President Barack Obama was the Lilly Ledbetter Fair Pay Act of 2009. It was designed to give many victims of pay discrimination a fair hearing. The new law would have allowed Lilly's complaint to be judged on its merits, rather than on a technicality.
Lilly Ledbetter was widely hailed as a crusader for justice. She danced with President Obama at an inauguration ball. She got no financial recompense for her troubles, but she did get one of the pens Obama used to sign the Fair Pay Act into law.
As Lilly Ledbetter's case was being dismissed by the Supreme Court on the grounds that it was not timely, America was engrossed in the 2008 Democratic presidential primary race between Hillary Clinton and Barack Obama, the first female and African American presiden- tial contenders in history to each have a serious shot at winning the presidency. Clinton, the early front-runner, faltered in the late winter. As she raced to catch up with Obama, her campaign and supporters repeatedly argued that she was the victim of sexism.
A few months before the primary race began, I wrote a column in The Washington Post about the role that sexism plays in the way people perceive female leaders. I talked about how a pattern emerged in the way the online encyclopedia Wikipedia, which is the open-source creation of its users, described women leaders. Margaret Thatcher was “Attila the Hen.” Golda Meir, the first female prime minister of Israel, was “the only man in the Cabinet.” Indira Gandhi, India's first female prime minister, was called the “Old Witch.” And Angela Merkel, the first female chancellor of Germany, was the “Iron Frau.” The conventional explanation for why women leaders are widely perceived to be ruthless and conniving-and to have lost their caring, feminine side-is that politics is a tough sport and women who climb to the top must have rough elbows. But before these women became leaders of their countries, I found that they had been described in very different ways. Merkel had been nicknamed das Mädchen-“the girl.” Indira Gandhi had been called gungi gudiya, or “dumb doll.” How did “the girl” become the “Iron Frau”? How did the “dumb doll” become the “Old Witch”?
If these women leaders had wills of iron and attitudes like Attila, why had dumb blonde stereotypes been attached to them before they'd assumed power? I argued in my column that people have unconscious stereotypes about men and women, and also about the nature of leadership-which is linked in our minds with strength, decisiveness, and manliness. When a woman assumes a leadership role, our unconscious stereotypes about leadership come into conflict with our unconscious stereotypes about women. The hidden brain reconciles the conflict by stripping women of their feminine, caring side. Our hidden brain makes women leaders appear ruthless and dislikeable for no better reason than that they happen to be women leaders.
There is fascinating experimental research to back this up. Madeline Heilman at New York University once conducted an experiment in which she told volunteers about a manager. Some were told, “Subordinates have often described Andrea as someone who is tough, yet outgoing and personable. She is known to reward individual contributions and has worked hard to maximize employees' creativity.” Other volunteers were told, “Subordinates have often described James as someone who is tough, yet outgoing and personable. He is known to reward individual contributions and has worked hard to maximize employees' creativity.” The only difference between what the groups were told was that some people thought they were hearing about a leader named Andrea while others thought they were hearing about a leader named James. Heilman asked her volunteers to guesstimate how likeable Andrea and James were as people. Three-quarters of the volunteers thought James was more likeable than Andrea. Using a clever experimental design, Heilman determined which manager each volunteer preferred: Four in five volunteers preferred to have James be their boss. Andrea seemed less likeable merely because she was a woman who happened to be a leader.
Hillary Clinton's supporters had ample evidence that their candidate was treated in a sexist fashion. Cable news was chock-full of overt sexism. Conservative talk show host Rush Limbaugh, for example, said, “Will this country want to actually watch a woman get older before their eyes on a daily basis?” Hillary Clinton was compared to Lorena Bobbitt (of penis-chopping fame) and likened to “everyone's first wife standing outside a probate court.” Tucker Carlson said on MSNBC, “When she comes on television, I involuntarily cross my legs.” Clinton was widely described as unlikeable and untrustworthy. She endured numerous sexist taunts-“Iron my shirt,” for example, and the ever popular “bitch.”
But did Clinton lose the primary race to Barack Obama because of sexism? Millions of her supporters, men included, were convinced that was the case. But opinions tended to fall along lines of political allegiance. Obama supporters, women included, were much less likely to attribute the outcome of the primary race to sexism. (Many Obama supporters noted that Clinton benefited from a different bias-some of her supporters explicitly said they would never vote for a black man.) Many political consultants also argued that Clinton made critical political errors in the race. And Clinton did win eighteen million votes, more than any successful Democratic presidential primary candidate before 2008. Candy Crowley of CNN said it was not clear if the “attacks were being made because she was a woman or because she was this woman or because, for a long time, she was the frontrunner.”
The existence of unconscious sexism can be scientifically proven in laboratory experiments. We know that unconscious sexism caused the laboratory volunteers in Heilman's experiment to find Andrea the manager less likeable than James the manager, because two groups of volunteers, divided at random, reached different conclusions about the likeability of the managers. Since the only thing that varied between the groups was whether they were told the manager was named Andrea or James, we can confidently say the outcome was produced by that single difference.
Bias is much harder to demonstrate scientifically in real life, which may be why large numbers of people do not believe that sexism and other forms of prejudice still exist. Many people think we live in a “post-racial” and “post-sexist” world where egalitarian notions are the norm. Indeed, if you go by what people report, we do live in a bias-free world, because most people report feeling no prejudice whatsoever.
I am personally convinced that Lilly Ledbetter was the victim of discrimination and wasn't just a mediocre employee who got her just deserts, but if you pressed me to prove my case scientifically, I would have to say I do not have conclusive proof. The jury that first heard Lilly's case concluded that she was the victim of discrimination. But a legal conclusion is very different from scientific certainty. We do not know what really happened inside the Gadsden plant, year in and year out. We cannot tell whether the manager who spoke insultingly to Lilly, or the manager who excluded her from a meeting with the big bosses from Akron, did so because she was a woman or because of some other factor. Even if sexism is endemic, it is conceivable that the insulting manager regularly insulted both men and women. That doesn't make him less of a jerk, but it does raise the question of whether he insulted Lilly because she was a woman. Similarly, it is not as if Lilly never got a merit raise. For any year in which she was de- nied a raise, we do not know whether Lilly was passed over because she was not up to par or because of sexism. Sexual harassment is obviously disturbing and illegal, but the connections between harassment and the wage gap Lilly suffered are far from clear. It is possible that the supervisor who propositioned Lilly was an isolated bad apple who in no way reflected how the company evaluated Lilly's performance and determined whether she deserved raises.
It may sound like I am going out of my way to give Goodyear the benefit of the doubt, but what I am trying to explain is how so many people can plausibly argue that bias is merely in the eye of the beholder. I am intuitively sure that Lilly suffered discrimination, but I cannot prove this scientifically because, for every outcome I see, there is always more than one possible explanation. Discrimination would certainly explain the pay disparity, but I have to concede it is also possible that Lilly was not as skilled as her colleagues.
This is true even though there is very clear evidence that women on average are paid less than men for doing the same work. Women working full-time earn about seventy-seven cents to the dollar as compared with men working full-time. Many people argue that this disparity arises because men and women choose different kinds of professions, work different hours, and are likely to take different amounts of time off to raise families. But even if you take all the confounding variables into account, women who work full-time and have never taken time off to raise a family earn about eighty-nine cents to the dollar as men in identical professions. While it is possible that some of those women are paid less than men because they are not as good, it is implausible that women on average are less competent than men.
The problem, however, is that every time we are asked to think about a situation that might involve discrimination, we have to deal with a single case, not averages. Judges might feel that sexism is real and that women on average are more likely than men to face an uphill climb, but that is not helpful in any individual situation.
If we wanted to settle the question scientifically, what we would really like to know is how Lilly Ledbetter would have been treated if she'd been a man. Would the supervisor have compared her work area to a whorehouse? Would she have received mediocre performance reviews and been passed over for raises? Would her colleagues and co-workers have been more helpful? If we could turn Lilly Ledbetter into a man and have her go back to the Gadsden plant in 1979, we would instantly have incontrovertible proof about the absence or presence of sexism. If (she) received all kinds of privileges, if (she) were given access to networks and mentoring, we would know- scientifically-that the reason she'd been excluded from those things in her real life was because she was a woman. And even the most skeptical among us would want to take Goodyear and nail its ass to a tree.
The same goes for Hillary Clinton. I can confidently say that women leaders on average are perceived to be tougher and more ruthless than they really are, and less warm and caring. But while we have intuitions about the role of such biases in any individual situation, the data cannot conclusively tell us whether a particular woman candidate failed because of sexism or because she was an inferior candidate. In any individual case, it is impossible to tell apart accurate perceptions of ruthlessness from unconscious beliefs that link female leaders with ruthlessness, because they both produce exactly the same perception in our minds. A woman leader can appear ruthless because of our own unconscious bias or because she actually is ruthless. As we will see in a chapter devoted to the issue of unconscious bias in politics, the fact that conscious and accurate assessments, and unconscious bias, can both produce the same perceptions in our mind is a major reason unconscious biases flourish and can be actively fanned by those who profit from them. Those who encourage our unconscious biases can plausibly tell us our views are the product of careful analysis, not bias.
The reason unconscious bias is so insidious and so powerful is that it can influence voters without their being aware of it. Explicit bias- Limbaugh and the other openly sexist commentators-got the headlines during the Democratic primary, but implicit bias may have determined the outcome. Overt sexism may have ultimately had very little effect. Millions of Democrats were repulsed by Limbaugh and the other misogynists, and the overtly sexist comments may have even driven up support for Clinton. Unconscious bias, on the other hand, may have prompted millions of voters to reject Clinton for reasons that ostensibly had nothing to do with her gender. How can we tell, with scientific certainty, that unconscious bias caused Hillary Clinton to lose? Real life does not provide us with control groups. If Hillary Clinton were a man, (s)he might still have lost.
Hillary Clinton and Lilly Ledbetter are examples of a pervasive challenge in applying the scientific research into bias to the context where it matters most-the lives of individuals. When a woman is passed over for a job or a raise or a plum assignment, when a presidential candidate loses or an employee gets fired, fair-minded people cannot say for certain whether the outcome was because of bias or because of other factors. Neither can bosses, voters, or the victims themselves. The experimental and observational data are very good at telling us that sexism exists, and that it may play a role in an individual case, but we have to acknowledge that other factors are invariably involved in any individual situation. There are certainly weak women political candidates and inferior women employees, and it would be a serious error to attribute every defeat and demotion of women to sexism.
It may seem quite obvious to you that the outcomes Lilly Ledbetter and Hillary Clinton experienced were the product of sexism. But the fact that that argument cannot be made with scientific certainty is, I believe, why many people dismiss bias as being merely a matter of opinion. The distinction between personal opinion and scientific fact is important: It doesn't matter whether anyone personally agrees with Heilman's data about volunteers finding a female manager less likeable than a man with identical qualifications. The powerful thing about scientific proof is that it renders personal opinions irrelevant. A scientific fact does not depend on our belief in it to exist, and it does not vanish because we disbelieve it.
What would be remarkably instructive in real life is if women in various professions could experience life as men, and vice versa. If the same person got treated differently, we would be sure sexism was at work, because the only thing that changed was the sex of the indi- vidual and not his or her skills, talent, knowledge, experience, or interests. Individual human beings could become their own control groups.
As it turns out, there are men who were once women, and women who were once men. Transgendered people allow us to scientifically apply the research on sexism to the lives of individuals because when a man becomes a woman or vice versa, the person's educational background, professional expertise, and life experience remain the same. If a woman who becomes a man suddenly finds himself privileged in all kinds of subtle ways, or a man who becomes a woman suddenly finds herself shackled, we can unhesitatingly-and scientifically-say sexism is to blame.
There is compelling empirical evidence to show that when men transition to becoming women, they experience all kinds of disadvantages that they did not experience when they were men. Their incomes, on average, fall. When women transition to becoming men, they find they have all kinds of new privileges. Their incomes, on average, rise. Transmen-people who transition from female to male-often report aspects of their professional lives getting easier. Transwomen-people who transition from male to female-often report the reverse.
The sociologist Kristen Schilt has tracked this phenomenon. Between 2003 and 2005, she followed the lives of twenty-nine transmen in Southern California. The transmen were white-collar and bluecollar workers, professionals, and retail salesmen. They ranged in age from twenty to forty-eight. They included people who were white, black, Latino, Asian, and biracial. Eighteen of the twenty-nine were open, meaning their co-workers knew they had once been women. Eleven of them were “stealth” transmen.
Overwhelmingly, the men told Schilt that they were being treated better than they'd been treated as women. Some enjoyed their newfound privileges, others felt uncomfortable. One thirty-nine-year-old white man who worked in a blue-collar job told Schilt: “I swear they let the guys get away with so much stuff! Lazy-ass bastards get away with so much stuff, and the women who are working hard, they just get ignored. . . . I am really aware of it. And that is one of the reasons that I feel like I have become much more of a feminist since transition. I am just so aware of the difference that my experience has shown me.”
Carl, a thirty-four-year-old “stealth” transman, told Schilt about the hardware store where he worked after he made the transition: “Girls couldn't get their forklift license, or it would take them forever. They wouldn't make as much money. It was so pathetic. I would have never seen it if I was a regular guy. I would have just not seen it.” A Latino attorney told Schilt that an attorney at another law firm had complimented his boss for firing an incompetent woman and hiring a new lawyer who was “just delightful.” The attorney at the other firm did not know that the incompetent woman and the delightful new lawyer were the same person.
One transman told Schilt that he was not asked to do different work after the transition, but doing his work suddenly became much easier. He recalled that before the transition, he would often be told that crews and trucks were not available when he needed some help. “I swear it was like from one day to the next of me transitioning. [I would say,] 'I need this, this is what I want,' and-” The man snapped his fingers. “I have not had to fight about anything.”
Another study that Schilt conducted with Matthew Wiswall analyzed the salaries of forty-three transgendered people after the volunteers made transitions from male to female or female to male. Schilt and Wiswall found that men who became women reported a decline of 12 percent in their earnings. Women who became men reported an increase of 7.5 percent in their earnings.
“While transgender people have the same human capital after their transitions, their workplace experiences often change radically,” Schilt and Wiswall wrote in a paper they published in The B. E. Journal of Economic Analysis & Policy. “We estimate that average earnings for female-to-male transgender workers increase slightly following their gender transitions, while earnings for male-to-female transgender workers fall by nearly 1⁄3. This finding is consistent with qualitative evidence that for many male-to-female workers, becoming a woman often brings a loss of authority, harassment, and termination, but that for female-to-male workers, becoming a man often brings an increase in respect and authority. These findings . . . illustrate the often hidden and subtle processes that produce gender inequality.”