The historical record shows that prejudice has been rampant in the past. But what about now? If we look around the world, surely we would conclude it’s not just an ugly aspect of history. Prejudice based on ethnicity, gender, and sexual orientation continues to fuel violence and societal turmoil in many parts of the world. But what about in the United States? In the 21st century, aren’t Americans past all of that? Isn’t America a land of truly equal opportunity?

Although many European Americans believe that racial prejudice is a thing of the past (Norton & Sommers, 2011), African Americans are far from convinced. In a Pew survey in 2007, 67% of African Americans reported that they often encountered prejudice when applying for jobs, and 50% reported that they had experienced racism when simply shopping or dining out. In a 2013 Pew survey, 88% of African Americans said that their group still experiences “some” or a “a lot” of discrimination, compared with only 57% of European Americans who believe that African Americans still experience discrimination.

Of course there is no question that great progress has been made in improving attitudes toward and treatment of ethnic minorities; women; and gays, lesbians, bisexuals, and transgendered individuals in the United States. Dial back time about 60 years to 1954. The U.S. Supreme Court has just announced the historic decision Brown v. Board of Education, which struck down state laws enforcing racial segregation in the public schools. The Court ruled that “separate but equal” schools for Black and White students were inherently unequal. The ruling was met with stark and at times violent opposition in a number of states. President Dwight D. Eisenhower had to call in the National Guard to protect African American children heading into and out of school in Arkansas.

Contrast that situation with where America is now. About 60 years later, we have anti-discrimination laws and an African American was elected president of the United States in 2008. In some respects at least, the United States has made tremendous strides in race relations.

359

Similar changes can also be seen with other prejudices—although some changes have taken longer than others. Consider prejudice and discrimination based on gender. Although the 19th Amendment guaranteed women the right to vote in 1920, it was not until 2006 that Congresswoman Nancy Pelosi became the Speaker of the House and the highest-ranking woman in U.S. political history. As of 2015, the United States has not yet had a female president, but many Democrats are enthusiastic about having Hillary Clinton on the ballot in 2016. What about prejudice against gays and lesbians? Although same-sex marriage is hotly debated, one CNN poll showed that 2010 was the first year that more than 50% of those surveyed supported same-sex marriages (CNN Political Unit, 2012). And at the end of 2010, the U.S. Senate voted to repeal the “Don’t ask, don’t tell” policy and struck down the ban on openly gay men and women serving in the military. In 2013, the U.S. Supreme Court decided a landmark case that opened the door for same-sex couples to qualify for federal benefits previously only extended to heterosexual couples (U.S. v. Windsor, 2013). There are also signs of progress toward acceptance of transgender individuals. In 2014, we saw the first openly transgender person ever to be nominated for an Emmy, Laverne Cox, for her performance in the television show Orange is the New Black.

Are these social, political, and legislative strides in step with the prejudices that people express? To a large extent, at least when considering more overt expressions of prejudice, the answer is yes. FIGURE 10.3 shows the results of two polls. Looking at FIGURE 10.3a, you will see that in 1958, 94% of Americans surveyed opposed interracial marriage, a number that had dropped to 17% by 2007 (Carroll, 2007). Similarly, with regard to gender attitudes, you see in FIGURE 10.3b that in 1937, two thirds of Americans said they would not vote for a female presidential candidate, but by 2006, only 5% of Americans expressed such an absolute aversion to having a woman in the Oval Office (CBS/NY Times, 2006; Eagly, 2007). But before we conclude that problems with prejudice are all but solved, we must keep a few important points in mind.

360

Second, even if attitudes toward some groups have become more favorable over time, social and political contexts can bring about new hostilities that are simply targeted against different groups. Many Americans probably had few strong attitudes about Muslims or people from Arab nations before 2001. But the September 11, 2001 attacks on the World Trade Center and the Pentagon fueled anti-Muslim prejudice. Whereas Jews were the most salient target of religious prejudice in the United States when Allport wrote his book in 1954, today “only” about 15% of Americans surveyed report having even a little prejudice against Jews, compared with nearly 43% who report having at least some prejudice toward Muslims (Gallup, 2010). Anti-Semitism is far from gone however; it is currently quite prevalent in many European and Asian countries (Baum et al., in press). As human beings, we have the capacity to distrust and dislike those who are different, but the targets of this animosity vary according to current context and particular culture.

361

Third, although overt expressions of discrimination and racial injustice are certainly declining, they are far from absent. In the fall of 2014, protests cropped up throughout the U.S. in response to police killings of African Americans that many people viewed as outrageous and unwarranted (Huffington Post, 2014). The most publicized of these cases were: the shooting of Michael Brown in Ferguson, Missouri; the shooting of 12-year-old Tamir Rice, who was playing with a pellet gun in a park in Cleveland, Ohio; and the chokehold killing of Eric Garner in Staten Island, New York, who died soon after telling the five police officers holding him down “I can’t breathe.” As of this book’s printing, the police officers involved in these actions had not been charged with any crime or fired. One reaction to these events has been to assert something that shouldn’t need to be asserted, that “Black Lives Matter.”

362

But in addition to subtle and indirect forms of prejudice, extreme acts of cruelty are still evident as well. Consider the highly publicized case in 1998 in which a gay University of Wyoming student, Matthew Shepard, was beaten, tortured, and left to die (Loffreda, 2000). This horrendous crime stimulated further blatant anti-gay actions that continue to this day. A pastor and founder of the Westboro Baptist Church, the Reverend Fred Phelps, brought several of his followers from Kansas to protest at Matthew Shepard’s funeral. They carried signs that read “God hates Fags” and “Fags deserve to Die.” Reverend Phelps repeated such proclamations in the wake of the 2007 collapse of a bridge in Minnesota, claiming that the bridge gave way because of Minnesota’s tolerance of homosexuality. Although Phelps’s behavior seems to be a particularly offensive and rare exception to contemporary tolerance, hate crimes in the United States have by no means stopped. In 2011, law enforcement agencies reported over 7,700 hate-motivated offenses (Federal Bureau of Investigation, 2012). And it was not until 2009 that perceived gender, gender identity, sexual orientation, and disability were added to the federal definition of hate crimes through the Matthew Shepard Act.

Beyond these extreme acts, there is still ample evidence of discrimination, even though the majority of Americans believe otherwise. During the late 1950s, the civil rights campaign brought into public awareness the problem of institutional discrimination, unfair restrictions on the opportunities of certain groups of people by institutional policies, structural power relations, and formal laws (for example, a height requirement for employment as a police officer excludes most women). The more closely Americans looked at their systems of health care, criminal justice, education, and the media, the more they discovered how these systems are arranged in ways that systematically disadvantage certain groups over others. In fact, this form of discrimination is so deeply embedded in the fabric of American society that it can take place without people intending to discriminate or even their awareness that institutional practices have discriminatory effects (Pettigrew, 1958).

Despite the heightened awareness of institutional discrimination, many examples continue today. Particularly in higher-paying jobs (e.g., hospital administrator), equally qualified women may earn only about 78 cents to every dollar that men earn (Eagly & Carli, 2007; Semega, 2009). Are women paid less than men for the same jobs? Sometimes that is the case. But even when it is not, keep in mind that women are more likely to be represented in jobs that have lower earning potential (see FIGURE 10.4). At a broader cultural level, our society has assigned less economic value to occupations traditionally held by women (nurse, teacher, administrative assistant), with the net result that women earn less than their male contemporaries (Alksnis et al., 2008).

Clear signs of racial discrimination remain visible, whether in employment, housing, credit markets, or consumer pricing (Pager & Shepherd, 2008). For example, studies have found that pricing among fast-food chains varies according to the demographics of the patrons likely to frequent a given establishment. A Big Mac may cost a couple of dollars in a predominantly White neighborhood, but is likely to cost 15 cents more in neighborhoods with larger Black populations (Graddy, 1997; Moore & Diez Roux, 2006). Or consider how society conducts its “war on drugs.” If you are caught in possession of cocaine in the United States, you could earn an all-expense-paid visit to the state penitentiary. But cocaine comes in different forms, and the Anti-Drug Abuse Act of 1986 specifies different mandatory minimum sentences, depending on the type of cocaine in your possession. Whereas 50 grams of crack cocaine will get you at least five years in prison, you would have to have 5 times that amount in powder form to get the same sentence (Sklansky, 1995). What does this have to do with prejudice? At the time this act was enacted, African Americans were more likely to use crack, whereas European Americans were more likely to use powder. As a result, prison sentences for essentially the same crime were disproportionately longer for Black Americans.

363

Disparities that implicate prejudice run even deeper. Analyses of inmate records show that, for comparable crimes, Black males who have “Afrocentric” facial features—in other words, who are particularly likely to look like Whites’ stereotypes of Blacks—receive harsher prison sentences than Black males who have less stereotypically Black facial features (Blair et al., 2004). They are also more likely to receive the death penalty for capital offenses when the victim was white (Eberhardt et al., 2006).

Despite the many encouraging declines in overt prejudice in the last few decades in the United States, these findings indicate that prejudice is still very much a problem in contemporary society. But because of progress in laws and cultural norms combatting overt prejudice and discrimination, the problem has also taken on a somewhat different character. Less prevalent are the explicit signs of prejudice, such as Jim Crow laws, formerly used to mandate various forms of racial segregation, or policies that overtly prevent women from pursuing certain careers. In the contemporary world, we often see more subtle—or what are termed modern—forms of prejudice. Because of America’s sordid history with slavery and explicit discrimination against African Americans, it is not a coincidence that the study of prejudice in the United States traditionally focused on racism in the form of prejudice against African Americans. Theories developed more recently to understand contemporary, more subtle forms of prejudice have also focused primarily on attitudes toward Blacks.

364

In the early 1970s, a new understanding of prejudice emerged in social psychology. Overt demonstrations of prejudice in the United States were dissipating, yet it was clear that many people were still harboring prejudiced attitudes. Two related concepts that psychologists proposed to explain contemporary prejudice were ambivalent racism and aversive racism. Although there are some differences between these perspectives, they have much in common. Each in its own way emphasizes the need to understand how and why people might explicitly reject prejudiced attitudes but still harbor subtle biases. Let’s briefly consider each of these perspectives and what they have taught us.

Ambivalent Racism

Because of changing social norms and values associated with the civil rights movement, contemporary prejudice against African Americans in the United States is often mixed with ambivalence (Katz & Hass, 1988). The term ambivalent racism refers to racial attitudes that are influenced by two clashing sets of values: a belief in individualism, that each person should be able to make it on his or her own; and a belief in egalitarianism, that all people should be given equal opportunities. The core idea of ambivalent racism is that many Whites simultaneously hold anti-Black and pro-Black attitudes that are linked to these contrasting values.

Consider Steve, who is White. He likes to think of himself as an accepting person and readily admits that Blacks have been disadvantaged in various ways. He supports affirmative action and votes in favor of using tax dollars for aid to the poor to help level the playing field. But while believing this, Steve also thinks that Blacks generally don’t try hard enough. Social psychologists would characterize Steve as having racial ambivalence. What happens with this ambivalence?

Believing that each person should be able to make it on his or her own conflicts with trying to justify giving certain people (e.g., Blacks, women) special affirmative action opportunities. Depending on which set of values is primed, ambivalent people are likely to respond more strongly in one direction or the other. This response affirms one of the two sets of values, which at least temporarily reduces the ambivalence.

So how do you know which value people will affirm? Well, it depends on which value is currently most salient, or active (Katz & Hass, 1988). If people are thinking about values related to individualism, they tend to be more prejudiced, but if thinking about values related to egalitarianism, they tend to be less prejudiced. For example, let’s take a look at FIGURE 10.5. When White participants were led to think about the Protestant ethic (a belief that emphasizes the individualistic value of hard work), they were more likely to report stronger anti-Black attitudes, but these individualistic thoughts did not influence their pro-Black attitudes. The pro-Black attitudes were no different from those of participants who received no priming. Thus, ambivalence remained, but the individualism prime shifted participants toward anti-Black attitudes. Conversely, when White participants were led to think about the value of humanitarianism, they were more likely to report pro-Black attitudes, but these humanitarian thoughts did not influence their anti-Black attitudes (again, as compared with the no-priming condition). Thus, again, ambivalence remained, but the humanitarian prime shifted participants toward more pro-Black attitudes.

Aversive Racism

A paradox of American society—a society founded on the principles of equality and egalitarianism but built on slavery and racial injustice—is the centerpiece of another perspective on modern prejudice. Sam Gaertner and Jack Dovidio’s concept of aversive racism (Gaertner & Dovidio, 1986) proposes that although most Whites support principles of racial equality and do not knowingly discriminate, they may at the same time possess conflicting, often nonconscious, negative feelings and beliefs about Blacks. Although they will consciously try to behave in line with their egalitarian values, in more subtle situations when decisions are complex and biases are easily rationalized, they may fall prey to the influence of prejudice, often without awareness that they are doing so.

365

Consider some of the earliest work on aversive racism, inspired by observations and news reports that Whites were less likely to help Black motorists in need. Gaertner and Dovidio (1977) presented White participants with what the participants thought was a situation where a Black or a White person needed help. However, the researchers varied whether a White participant thought she or he was the only person to hear the plea, or whether she or he thought others were available to help. The results of the study—shown in FIGURE 10.6—indicated that the White participants did not discriminate when they thought they bore sole responsibility for helping. They were equally likely to aid Black and White victims. However, if White participants thought that other people also had witnessed the emergency, they were less likely to help a Black victim than a White victim. Presumably, this occurred because, believing that others might help, they could rationalize their decision in nonracial terms (e.g., “There are plenty of people to help”).

Similar results are observed in employment scenarios and college application processes. For instance, discrimination is less likely to occur when applicants have especially strong or weak records (because the applicants’ records do the talking) but is more likely to occur when applicants have mixed records. That is, discrimination occurs when a gray area allows race to play a role, but it can be justified through nonracial means (Gaertner & Dovidio, 2000). For example, when a Black applicant had higher SAT scores but lower high school grades than a White applicant, high-prejudice Whites decided that high school grades were an especially important factor in college admissions decisions. But if the Black applicant had higher grades and lower SAT scores, then the high-prejudiced Whites valued those scores instead as the best indicator of future success (Hodson et al., 2002).

So prejudice is more likely to occur in situations when it can be justified by some other motive. But you might be wondering, Do all people do this? Or are particular individuals most likely to act this way? To answer this question, a number of theories about prejudice suggest that it is critical to consider not just the attitudes that people are consciously willing or able to report but also those that might reside beneath their conscious awareness. Since the 1990s, there has been an explosion of interest in this concept of implicit prejudice.

The term implicit prejudice refers to negative attitudes toward a group of people for which the individual has little or no conscious awareness. Some people may choose not to admit their prejudices, whereas others may not be aware of them. But as we noted in our discussion of implicit attitudes in chapter 3, measures of implicit prejudice tap into attitudes that lie beneath the surface of what people report (Nosek et al., 2011). And indeed, whereas a majority of White Americans don’t report being prejudiced on explicit measures, most do show signs of having biases when their attitudes are assessed implicitly with either cognitive measures of implicit associations or physiological measures of affective responding (Dovidio et al., 2001; Cunningham et al., 2004; Hofmann et al., 2005; Mendes et al., 2002).

366

Physiological Measures of Bias

Measures of implicit prejudice tap into people’s automatic affective response to a person or a group. Some measures do this by indexing an immediate physiological reaction that people are unlikely or find difficult to control. For example, when Whites are asked to imagine working on a project with a Black or a White partner, they often report a stronger preference for working with the Black partner. But their faces tell a different story. Electrodes connected to their brows and cheeks pick up subtle movements of the facial muscles that reveal a negative attitude when they think about working with a Black partner (Vanman et al., 2004). Similarly, when Whites are actually paired up to work with a Black partner, they show a cardiovascular response that is associated with threat—their hearts pump more blood, and their veins and arteries contract (Mendes et al., 2002).

The brain also registers the threat response. The amygdala is the brain region that signals negative emotional responses, especially fear, to things in our environment. Whites who have a strong racial bias exhibit an especially pronounced amygdala response when they view pictures of Black men (Phelps et al., 2000). Interestingly, however, if given more time, this initial negative attitude tends to get downregulated by the more rational dorsolateral prefrontal cortex (Cunningham, Johnson et al., 2004; Forbes et al., 2012). This suggests that even people who acknowledge their prejudice may feel some pressure to control it. We’ll discuss why and how people go about controlling their prejudiced attitudes and emotions more in our next chapter. For now, the primary point is that automatic negative bias leaks out in people’s physiological responses.

Cognitive Measures of Implicit Bias

Although physiological measures provide insight into attitudes and affect, cognitive measures also tell us something about people’s implicit attitudes. These measures take different forms, but they all rely on the same assumption. If you like a group, then you will quickly associate that group with good stuff. If you don’t like a group, you will quickly associate that group with bad stuff. With this assumption tucked in our pocket, we can present people with members of different groups and measure how fast it takes them to identify good and bad stuff (Fazio et al., 1995; Dovidio et al., 2002).

For example, in developing one of the first studies of implicit bias, Fazio and colleagues (1995) reasoned that if Whites experience an automatic negative reaction to Blacks, then exposure to photographs of African Americans should speed up evaluations of negative words and slow down evaluations of positive words. To test this hypothesis, they presented participants with positive words (e.g., wonderful) and negative words (e.g., annoying), then asked them to indicate as fast as they could whether each was good or bad by pressing the appropriate button. Each word was immediately preceded by a brief presentation of a photograph of a Black or a White person. The results revealed substantial individual differences in White participants’ automatic reactions: For many White participants, being primed with Black faces significantly sped up reactions to negative words and slowed down reactions to positive words. Other White participants did not show this pattern, and some even showed the opposite. More important, the more closely these people seemed to associate “Black” with “bad,” the less friendly they were during a later 10-minute interaction with an African American experimenter.

Since the late 1990s, the most commonly used measure of implicit attitudes has been the implicit association test, or IAT (Greenwald et al., 1998), which we introduced to you in chapter 3. The basic logic of this test is that if you associate group A with “bad,” then it should be pretty easy to group together instances of group A and instances of bad stuff, and relatively difficult to group together instances of group A and instances of good stuff. This basic paradigm can be used to assess implicit associations with any group you can think of, but it has most commonly been used for race, so let’s take that example. You can take this and other versions of the IAT test yourself by visiting the Project Implicit web site at www.projectimplicit.com.

In the Race IAT, people are instructed to do some basic categorization tasks (FIGURE 10.7). The first two rounds are just practice getting used to the categories. For the first round, you are presented with White and Black faces one at a time, and you just need to click on one button if the face is Black and another button if the face is White. In the second round, you categorize positive (“rainbow,” “present”) and negative (“vomit,” “cancer”) words by clicking on one button if the word is positive and a different button if the word is negative. In the third round, the task becomes more complicated as you are presented with both faces and words, one at a time. Deciding as quickly as possible, you must then click one button if what you see is either a Black face or a positive word but a different button if you see either a White face or a negative word. When your cognitive network links “Black” with “bad,” it’s relatively hard to use the same button to categorize Black faces along with positive words without either slowing down or making lots of category errors. In contrast, if “Black” and “bad” are closely associated in your cognitive network, you should find it much easier to use the same button to indicate that you see either a Black face or negative word, the task required in a fourth round. The average difference in time it takes to make these dual category judgments (measured in milliseconds) is the measure of cognitive association.

367

Research using this kind of measure has shown that most Whites and a substantial proportion of Blacks (though usually not greater than 50%) show an implicit association of “Black” with “bad.” What is less clear is what this association means. Some researchers have criticized the measure for confounding the tendency to associate “Black” and “bad” with the tendency to associate “White” and “good” (Blanton & Jaccard, 2006; Blanton et al., 2006). However, other evidence suggests that IAT scores do reliably assess responses that are predictive of behavior (Greenwald et al., 2009; LeBel & Paunonen, 2011). Even if we grant that the IAT is a reliable measure, some dispute continues about what it taps into. For example, people actually show stronger racial biases when they know the measure is supposed to reveal their racial biases (Frantz et al., 2004). Anxiety about being labeled racist might actually make it more difficult for people to perform the task. In addition, some researchers have noted that an association of “Black” with “bad” could mean a variety of things, such as the acknowledgment that Blacks are mistreated and receive bad outcomes, or simply cultural stereotypes that might have little to do with one’s personal attitudes (Andreychik & Gill, 2012; Olson & Fazio, 2004).

Even if we set aside the debate about the IAT in particular, a broader pattern emerges from the literature examining both implicit and explicit measures of prejudiced attitudes. Most notable, although they can be correlated, they are often quite distinct. In other words, people who have an implicit negative attitude toward a group might still explicitly report having positive feelings. But even more interesting is that people’s implicit attitudes seem to predict different kinds of behavior than their explicit attitudes. Let’s consider one study that has shown this most clearly.

Dovidio and colleagues (2002) had students participate in what they thought were two unrelated studies. In the first, the researchers measured implicit prejudice using a cognitive association task like the one Fazio developed in 1995. The researchers measured explicit prejudice by having White participants complete a questionnaire asking them outright how much prejudice they felt toward African Americans. After this, the participants proceeded to what they thought was an unrelated study in which they would have a brief, videotaped interaction with a Black person (actually a confederate). After the conversation, the participants were asked to judge how friendly they were toward their partners. The confederates made their own judgments of how friendly the participants seemed. Finally, the video and the audio soundtrack of the interaction were separately coded by a different group of raters whose only job was to judge how friendly the participant seemed during the interaction.

368

The patterns of relationships among these six variables were very telling. People’s implicit “outgroup = bad” attitudes and explicit “I value being a nonprejudiced person” attitudes predicted very different things about the interaction. White students who explicitly reported being nonprejudiced were coded (only on the basis of the audio recording of what they said) as coming across in a friendly way during a discussion with a Black student, and they came away from the conversation feeling good about the interaction. But their consciously endorsed values did not predict how coders rated their nonverbal cues (on the basis of only the video without sound) or how their Black partners perceived them. Instead, Whites who had a more negative implicit association with Blacks showed more nonverbal signs of discomfort and anxiety—which was what their partners picked up on. Effects such as these pave the way for misunderstandings. Explicitly well-intentioned Whites try to say the right thing, but their body language can communicate discomfort and avoidance that turns their partners off (see also Amodio & Devine, 2006; McConnell & Liebold, 2001).