THE EXPERIENCE OF DISCRIMINATION

The experience of discrimination

Order Description

Imagine you are in heaven, and the angel in front of you says: “You are going to be born to the world. In America. But guess what? We give you choices. There are some groups in America that are in disadvantaged positions… for example, Blacks. So you can choose to be born White, or you can choose to be born Black with cash compensation. The cash will be deposited to your bank account when you are born.”

Question: If you choose to be born Black in America, how much compensation do you think is reasonable? The more specific the better.

Chapter 11
*
The Experience of
Discrimination
Oh, is there still racism?
-ANONYMOUS STUDENT, ON HEARING THAT A COURSE
ON RACISM WAS BEING OFFERED ON HER CAMPUS,
QUOTED IN TATUM (1997, P. 3)
I don’t think White people, generally, undmtand the
full meaning of racist discriminatory behaviors directed
toward Americans of African descent. They seen1 to see
each act of discrimination or any act of violence as an
“isolated” event. As a result, most White Americans cannot
understand the strong reaction n1anifested by Blacks when such
events occur. They feel that Blacks tend to “overreact.” They
forget that in most cases, we live lives of quiet desperation
generated by a litany of daily large and small events that,
whether or not by design, remind us of our “place”
in American society.
-ANONYMOUS BLACK PROFESSOR, QUOTED IN FEAGIN
AND SIKES (1994, PP. 23-24, EMPHASIS IN ORIGINAL)
Chapter Outline
Social Stigma
What Defines a Stigmatized
Group?
Stigma by Association
Tokenism
418
Responses to Prejudice and
Discrimination
Attributional Ambiguity
personaJ/Group Discrimination
Discrepancy
Consequences of Prejudice to the
Target
Stereotype Threat
Vulnerability to Stress
Threats to Self-Esteem
Coping with Discrimination
Psychological Disengagement and
Disidentification
THE EXPERIENCE OF DISCRIMINATION 419
Behavioral Compensation
Summary
Suggested Readings
Key Terms
Questions for Review and Discussion
As we saw in Chapter 6, many White Americans think prejudice is more or
less a thing of the past. It is certainly true that more blatant fonlLS of prejudice
have declined in the United States, because of both legislative and social changes.
It is also true, however, that the existence of prejudice and discrimination can simply
be iuvisible to many members of the majority group. It is sometiules difficult
for the majority group to accept that, for many people, prejudice and discrimination
are a “lived experience” (Feagin & Sikes, 1994, p. 15) and are not inconsequential
beliefs and actions that can siulply be overlooked while “getting on with
one’s life.” Instead, for members of stereotyped groups, these experiences are
woven iuto the fabric of their lives. Much of this book has focused on theories about
and research on prejudiced people. In this chapter, we tell the story of prejudice and
discrimination from the poiut of view of those lived experiences, focusiug on the social
psychological research that describes and explains them.
As we have seen in earlier chapters, prejudice and discrimination can take
many fonns, depending on the actor, the situation, and the historical time period
in which a person lives. These factors similarly affect those who experience prejudice,
creating a dynamic interchange between those who treat others unfairly and
those who are the recipients of this injustice (Dovidio, Major, & Crocker, 2000).
This chapter focuses on the consequences of this exchange as they affect every
aspect of the stigmatized person’s life, including their academic and economic
achievement and their physical and mental well-beiug.
SOCIAL STIGMA
To fully understand what it is like to experience discrimination, it is important to
know what factors set others apart from the dominant group, increasing the likelihood
that they will be discriminated against. Recall from Chapter 1 our discussion
of group privilege. This privilege is defined as membership in the dominant
group, a status that is seen as nonnal and natural and is usually taken for granted
(A. Johnson, 2006). Dominant group membership is sometimes referred to as
majority group membership, but this is somewhat of a misnomer. Privileged status
often comes from being in the majority; however, it is not defined simply by
420 CHAPTER 11
a group’s numerical advantage. For example, the British rule of India lasted more
than 300 years; during that time, Indians faced severe racial discrimination from the
British even though the Indians greatly outnumbered the British (Dirks, 2001),
Similarly, although Blacks in South Africa outnumber Whites four to one, until
1994 Blacks were subjected to apartheid laws that enforced their segregation
from Whites, governed their social life, and limited their employment options
(Beck, 2000), The vestiges of apartheid continue to affect Blacks in South Africa.
Privileged status, then, is defined less by a group’s numbers and more by its power
and influence. We begin our discussion by outlining the factors that delineate a
group’s privileged or disadvantaged status.
What Defines a Stigmatized Group?
~Whether they are consciously aware of it or not, individuals with privileged status
define which groups do or do not share this status. In social psychological tenns,
those groups that do not share this status are stigmatized or deviant. Stigmatized
-groups differ from the privileged or dominant groups in terms of appearance or
behavior. Members of stigmatized groups violate the nOn11.S established by the
dominant group on these dimensions and, as such, are lllarked by the resulting
social stigma (Jones et al., 1984). Because of this, members of stigmatized groups
are sometimes referred to as the marked and those who are the actors, or the ones
who stigmatize, are sometimes referred to as the markers. Marked individuals are
“devalued, spoiled, or flawed in the eyes of others” (Crocker, Major, & Steele,
1998, p. 504). The consequences of this devaluation are fur reaching and can include
dehumanization, threat, aversion, and other negative treatment, including
subtle forms of discrimination (Dovidio et al., 2000).
Which groups are stigmatized by the privileged or dominant group? The
answer depends on the culture and on the historical events that led to the current
cultural context. As we saw in Chapter 1, for example, the Irish and Italians were
once considered non~White and were targets of discrimination in the United
States; today, they are accepted as part of the White majority (Rubin, 1998).
Returning to our earlier examples, India is now governed by its own people and
is not subject to British dominance and Blacks in South Africa have made significant
strides toward undoing the effects of apartheid. Hence, historical events and
changes in laws and social nonns affect cultural beliefS about who can or should be
stigmatized, even if it sometimes takes many years to see their effects. More generally,
dominant group members detennine which individuals are stigmatized, based
on any number of characteristics, including membership in an underrepresented
basic social category, such as ethnicity or old age, or in a socially deviant category
defined by physical or mental disability, weight, socioeconomic status, or sexual
orientation. People also can be stigmatized because of their acne, their mother’s alcoholism,
a speech impediment, or illness, among many other things (Jones et al.,
1984). To be stigmatized, then, individuals must have a characteristic that is devalued
by the dominant group and that sets them apart from that group. Regardless of the
source of the stigma, in all cases, there is shame associated with being nurked
(Goffinan, 1963).
THE EXPERIENCE OF DISCRIMINATION 421
As you read this list of stigmatized groups, you might have concluded that
almost everyone has had the experience of being different from the majority and
has suffered because of it. It is true that being different from the group is often part
of normal human life. If you have had such experiences, it may give you some
insight into what it is like to be a member of a stigmatized group. But for majority
group members, nlany times these experiences are short-lived or othenvise benign.
Benign stigmas, such as acne, a correctable speech impediment, or a short-tenn illness,
differ in important ways from the more harmful stigmas social scientists most
often study, such as those based on ethnicity, severe mental illness, or sexual orientation.
Because these latter stigmas typically have more negative consequences,
ranging from depression to extreme violence against the stigmatized group, they
are the focus of this chapter. Edward Jones and his colleagues (1984) have identi–J
fied five dimensions that are particularly helpful in differentiating between harmful
and benign stigmas: course, concealability, aesthetic qualities, origin, and peril.
1.
2.
Course. Benign stigmas are often temporary; that is, the course of the stigma’l
is short. For example, acne is usually outgrown or can be cured by a J
dennatologist. In contrast, the course of many negative stigmas cannot be -J
changed. An individua1’s ethnicity is typically part of his or her lifelong
identity, for example. Another tenn that is sometimes used is stability; some
stigmas are perceived to be stable, or pennanent, whereas others are
perceived to be unstable and so can change over time. In general, people
believe that physica11y based stigmas, such as blindness or cancer, are stable
and that mental-behavioral stigmas, such as drug abuse or obesity, are
unstable (Weiner, Perry, & Magnusson, 1988). In general, stable stigmas
have more negative consequences for the stigmatized person.
Concealability. Some stigmas are concealable, which means they can be I
hidden or controlled by the stigmatized person. Such stigmas can be avoided J
simply by keeping the stigma private, such as by not ta1king about one’s
alcoholic mother, or can be hidden, such as by wearing makeup to cover a
scar or birthmark. Moreover, some individuals can and do choose to “pass”
for a member of a different ethnic group, thus concealing their group
membership. However, as John Pachankis (2007) explains, concea1ing a
stigma does not reduce the guilt and shame associated with that stigma.
Moreover, the need to continuously monitor behavior so that the stigma
remains undisclosed can be anxiety provoking. fu he notes” [i]n every new
situation that is encountered, such individuals must decide who among the
present company knows of their stigma, who may suspect this stigma, and
who has no suspicion of the stigma” (p. 328). Many gay men and lesbians,
for example, are not open about their relationships out of fear of social
rejection, loss of employment, or the threat of physical violence; as a result
they often find themselves lying about or hiding an important part of their
life and they feel guilt and shame because they must do so (Meyer, 2003).
Similarly, people often fail to seek treatment for menta1 illness because of the
stigma associated with revealing their problem (Corrigan, 2004). People who
have stigmas that cannot be concealed have a different set of problems; they
422 CHAPTER 11
3.
~.
‘
realize their membership in a stigmatized group is apparent and this, in tum,
affects their thoughts, feelings, and behavior. They must always directly
cope with the prejudice and discrimination associated with their group
membership (Crocker et al., 1998).
Aesthetic qualities. Aesthetics refers to what is beautiful or appealing.
As we discussed in Chapter 3, many stereotypes are triggered by physical
appearance cues (Fiske & Taylor, 1991) and many stigmas are based on
this dimension as well. In general, less physically attractive people are more
likely to be stigmatized (Eagly, Ashmore, Makhijani, & Longo, 1991). One
reliable indicator of physical attractiveness is facial symnletty, or the degree
to which the left and right sides of the face are mirror images of each
other (Langlois & Roggman, 1990). Inclividuals with facial disfigurement
typically do not meet this standard and are likely to be stigmatized. In North
American culture, slimness is emphasized and overweight people become
the targets of cliscrimination (Crandall et al., 2001). Similarly, a central
component of the old-age stereotype is a decline in physical attractiveness
and mobility (Slotterback & Saarnio, 1996).
Origin. This tenn refers to how the stigma came to be and whether its
onset was under the control of the stigmatized individual. Stigmas perceived
to be controllable include drug addiction, acquisition of HIV, and obesity;
those perceived to be uncontrollable include cancer and heart disease
(Weiner et al., 1988). Physical characteristics that one is born with, such as
race or many disabilities, also are perceived to be uncontrollable Gones et al.,
1984). People’s beliefs about the controllability of a stigma have important
implications for acceptance of the stigmatized other. When people believe
that a stigma is uncontrollable, they feel nlore pity and less anger toward
the stigmatized individual compared with when the stigma is perceived
as controllable (Dijker & Koomen, 2003; Weiner et al., 1988). This
viewpoint is evident in this excerpt from a letter to the editor that appeared
in the Chronicle Review: “Race is something that a person has no control
over; hence racism is wrong. Homosexuality is a choice a person makes,
and therefore it is not wrong to disagree with it” (Colvin, 2003, p. B4).
Research suggests that others share Colvin’s viewpoint. For example,
Bernard Whitley (1990) found that people who believed that sexual
orientation was controllable had more negative attitudes toward lesbians
and gay men than did people who believed sexual orientation was not
controllable.
Peril. Members of some stigmatized groups are perceived, correctly or
incorrectly, to be dangerous. Persons with a mental illness, for example, are
stereo typically perceived to be dangerous, even though statistically they are
no more likely to commit violent crime than people not so diagnosed
(Corrigan & Penn, 1999). As we saw in Chapter 3, people stereotypically
assume that Blacks are more dangerous than Whites (Duncan, 1976).
Especially in the early years of the AIDS epidemic, the stigma associated
with HIV infection was found to be related to the belief that persons with
THE EXPERIENCE OF DISCRIMINATION 423
AIDS were highly contagious and therefore dangerous (Triplet & Sugannan,
1987). In general, groups assumed to be more dangerous are more stigmatized
than groups perceived as less dangerous (Jones et al., 1984).
Stigma by Association
So far, we have discussed what sets individuals apart from the dominant group.
One underlying assumption is that the dominant group generally rejects members
of stigmatized groups. But what happens when a member of the majority
group associates with a stigmatized person? Erving Goflinan (1963) proposed
that such an association would result in a “courtesy stignla” whereby the nlajority
group member would also then be stigmatized. In the past, mainly anecdotal
data supported this possibility. However, recent research suggests that Goffinan’s
hypothesis was correct. For example, Steven Neuberg and his colleagues (Neuberg,
Smith, Hoffman, & Russell, 1994) asked male research participants to watch a
social interaction that they believed was between either two friends or two
strangers. In the course of the conversation, one of the men (Person A) discussed
his relationship as being with either a woman or a man, which also revealed that
he was either heterosexual or gay. Person B, the other man, was presented as
heterosexual. Results supported Goffman’s hypothesis: there was a “courtesy
stigma” or a stigma by association with the gay man. That is, male research participants
were less comfortable with Person B when they believed he was a friend
of, rather than a stranger to, the gay Person A. When Person A was described as
heterosexual, Person B’s evaluations did not depend on how well he knew
Person A. Janet Swim and her colleagues (Swim, Ferguson, & Hyers, 1999)
also found that people fear stigma by association with gay people. In their study,
heterosexual women behaved in ways that socially distanced themselves from a
lesbian, even when doing so required agreeing with socially unpopular positions
or making sexist responses.
Additional research suggests that simply interacting with an obese person can
produce a courtesy stigma. Research participants were less likely to recommend
hiring a job applicant who was shown interacting with an overweight person at a
social gathering, regardless of how well tbe applicant knew the overweight person
(Hebl & Mannix, 2003). Similarly, children as young as 5 years old dislike
girls more when they are pictured next to an overweight rather than an average
weight child. However, this courtesy stigma did not emerge for boys who were
pictured with an overweight boy (Penny & Haddock, 2007). Finally, individuals
who are dating a person with a disability are subject to stigma by association,
including the perception that they are less intelligent and sociable than those
dating a nondisabled person (Goldstein & Johnson, 1997). Yet some aspects of
this stigma by association were positive, including the perception that those dating
the disabled were more nurturant and trustworthy than those not doing so.
Even so, these positive perceptions are consistent with the idea that those associated
with stigmatized others are different. As the authors note, even respondents’
positive comments focused on this difference, pointing out, for example, how
424 CHAPTER 11
much a person had to give up to date someone with a disability. In DUllY cases,
the comnlents indicated sympathy for the nondisabled person. Taken together,
these studies suggest that Goffinan’s idea has merit; there are social consequences
for associating with a deviant.
Tokenism
We noted above that being a numerical minority is not, in and of itself, sufficient
to produce stigmatized status. That is, power and status are important cOlnponents
of defining privilege and nonprivilege. This does not mean, however,
that being in the minority produces no negative effects, particularly in certain
situations or settings. That is, one can be in the majority or near majority in a
larger population, but still have stigmatizing experiences from being a minority
within a particular context. Women, for example, are now represented in the
labor force at numbers nearly equal to men. Many, however, still have negative
experiences that result from being in the minority in some environments, such as
[
being the only woman in a particular work group (Yoder, 2002). When individuals
are a statistical minority within a particular setting, they can be treated as
tokens and can be stigmatized because of it. In general, token status occurs when
there is a preponderance of one group over another, such as when one gender or
ethnicity is in the majority and only a few individuals fronl another gender
or ethnicity are represented (Kanter, 1977).
Rosabeth Moss Kanter (1977) pioneered the research on tokenism in her case

-‘stUdY of a multinational Fortune 500 corporation. Kanter highlighted three perceptual
tendencies that affected the daily lives of tokens: visibility, contrast, and
assimilation. Visibility refers to the ,tendency for tokens to get attention or, as she
put it, “capture a larger awareness share” (p. 210). Consider, for example, this vi-
sual field containing a series of 9 XS and ouly 1 0:
XXXXXXXOXX
Notice that your eyes tend to be drawn toward the 0 and not to any individual
X. As we saw in Chapter 3, the perceptual process is similar in social situations;
people’s attention also tends to be drawn to the novel or unique person
rather than to members of the majority group (Fiske & Taylor, 1991). Members
of the minority, or token group, are slll1ply noticed more than are other group
[.
members. Contrast refers to the polarization or exaggeration of differences between
the token and the dominant group. A White person in a group conlprised
‘. only of Whites, for example, might not think much about her or his racial identity.
The presence of a Black person, however, brings race to the forefront, raising awareness
of race for members of the dominant group. Similarly, adding a woman to an
all-male work group can raise awareness of gender issues. Often, both dominant and
[
Oken group members are uncomfortable when this happens. Assimilation occurs
when the token is stereotyped; in particular, the token’s characteristics are distorted
so that she or he fits the expected stereotype. A group of men, then, notice when a
-token woman behaves in a way that confinns their stereotypes about women and
THE EXPERIENCE OF DISCRIMINATION 425
often generalize from that confinnation. However, the same men tend not to notice
when the woman’s behavior does not confonn to their gender stereotypes.
These perceptual tendencies have important consequences for the token,
which Kanter (1977) illustrated with examples from her case study. She found,
for example, that whenever token women did something unusual, it stood out.
As she describes it “[t]hey were the subject of conversation, questioning, gossip,
and careful scrutiny … Their names came up at meetings, and they would easily
be used as examples … [S]01ne women were even told by their managers that
they were watched more closely than the men” (p. 212). This was a doubleedged
sword; their achievements were noticed, but so were their mistakes.
And, their actions were seen as representative of a11 women, not just of
themselves as individuals. Consequently, evel1 small decisions, such as what to
wear to a business meeting, became important. Most people find such situations
difficult to navigate, as the additional examples provided in Box 11.1 illustrate.
Tokens often feel isolated but, at the same time, must go on as if the differences
do not exist and do not affect their work. Solos, or people who are the only
minority member in a majority group, often feel alone and without support
(Benokraitis & Feagin, 1995). As one Black woman wrote, “the responsibility
associated with being the ouly Black female in my college and only one of a
handful in the university, was overwhelming. I have suffered several instances of
burn-out and exhaustion. & a consequence I have learned to maintain a less
visible profile as a coping and survival strategy” (Moses, 1989, p. 15). All told,
the negative effects of being in the minority can create what has been ca11ed the
“chilly climate” (Sandler & Hail, 1986). Tokens do not feel welcome or supported
in their environment and often their work and personal lives suffer
because of it.
Although Kanter (1977) defined token status as simply being in the numerical
minority, more recent work suggests numbers alone do not define token
status. For example, women who pursue nontraditional occupations are more likely
to experience the effects of tokenism than are women in traditional occupations
(Yoder, 2002). A survey of undergraduates, for exanlple, found that women in
male-dominated academic areas, such as math, science, and engineering, reported
higher levels of current sex discrimination than did women in female-dominated
academic areas, such as the arts, education, and social science (see Figure 11.1;
J. Steele, James, & Barnett, 2002). However, men’s perceptions of current sex discrimination
were not affected by their area of study. This pattern also emerged in a
measure of whether sex discrimination was expected in the future; women in maledominated
professions were most likely to hold this expectation and were most
likely to consider changing their major. As we discussed in Chapter 10, men in
female-dominated occupations, such as nursing and social work, rarely have the
same negative experiences as women in male-dominated professions and may even
be on the fast track to promotion (Maume, 1999; WillianlS, 1992), although there
may be exceptions in some settings. For example, Susan Murray (1997) found that
male child-care workers were pushed away from performing tasks that requITe
nurturing and received the clear message that child care was women’s work. These
426 CHAPTER 11
What does happen to the deviate? The deviate
can convert, but short of a sex change operation,
a time machine to age me, and a personality
overhaul, conversion seems out of the question
for me. Be isolated? That originally was all right
with me, but that surely does not make me a team
member. What can I do? Yet, the failure is placed
squarely on my shoulders. “What is wrong with
you?” “Why can’t you get along?” These
questions haunt me, undermining my self-image.
-JAN YODER (1985, p. 67)
It is difficult to document exactly what form a token’s
negative experiences might take. That is, the actual
events that comprise those experiences are very
personalized. Moreover, many of the individual instances
that lead to the isolation and loneliness experienced by
tokens seem harmless on the surface, especially to those
who are not directly living with them. As you read the
personal accounts described in this chapter, they too may
seem harmless. Keep in mind, however, that the research
evidence suggests that, over time, such experiences
affect those in token roles by isolating them from the
dominant group, lowering their self-esteem, and creating
loneliness (Sue et aI., 2007). As a respondent in Paula
Caplan’s (1994) survey of women in academe described,
their cumulative impact is similar to “lifting a ton of
feathers” (p. 9). Over time, their weight ‘IS unbearable.
This weight is illustrated by the opening quote in
this box, which came from Jan Yoder’s (1985) first
person account of being the first female civilian faculty
member at a United States military academy. Her
writings captured her dilemma about howto respond to
her interactions with the military officers who comprised
97 percent of the faculty. As she notes in her account,
no one event seemed overly traumatic. Yet, because of
their cumulative impact, she stayed only six months.
Here are a few of her experiences:
Because she openly questioned the sexism of
some exam questions, she was given a suggestion book
so she could quietly record her objections without
disrupting faculty meetings.
Her department chose to use “Macho Man” as its
theme song, a song few women would choose to
represent themselves.
Gossip about her ranged from “she’s a lesbian” to
“she is heterosexual, but promiscuous.”
Despite her efforts to clarify her position in the
academy, at social gatherings it was widely assumed
that she was the wife of one of the officers.
” ..
…. …. …’ .” :
.’ .•.• … ……….. …. ·<i
Jan Yoder is now a highly successful faculty
member at the University of Akron. Her study of Black
women firefighters (Yoder, 1997) shows how the
experience of being a token can threaten the safety of
both the firefighters and those they are protecting.
One Black woman in her study reported that, in
response to a request for help, she received no
constructive information, but instead was written up
for presumed negligence. A coworker directly told
another Black woman that when there was a fire, she
was not to touch anything, but rather to stay out of
the way. Many of the women reported receiving the
“silent treatment,” with the men literally walking out
of the room when they entered. One reported that,
during her formal testing, she was required to hoist a
hose onto a shelf that suddenly had been raised five
inches above where it was during training.
One of the ways tokens can be made to feel
alienated is through the conversations majority group
members initiate with them. Black managers, for
example, express frustrations with queries that seem to
hold them accountable for other Blacks’ behaviors,
such as “Why do all the Blacks sit together?” and the
relative lack of discussion about business-related topics,
such as how to make the company succeed (Caver &
Livers, 2002). Blacks often feel invisible as well.
Anderson Franklin (2004) describes the experience of a
successful Black manager who took a White business
client out for dinner in New York City. The maitre d’
ignored the Black manager, instead asking the White
client if he had reservations. And, after dinner, the
waiter returned the Black manager’s credit card to the
White client. After dinner, the White client easily
found a cab, but the Black manager was ignored
by cabdrivers for over 15 minutes, even as other
Whites successfully hailed a cab. Echoing the
sent’lments expressed by others ‘In this chapter, at the
individual level, such actions may seem harmless to
dominant group members, but to tokens “[ilt’s the
cumulative effect that wears us down” (Caver & Livers,
2002, p. 78).
Many others have written about these individualized
experiences. Researchers look across such events and,
based on patterns, draw conclusions about the short- and
long-term effects of being a token. On a positive note,
research suggests that when the group composition
changes so that, for example, several women become part
of an otherwise male-dominated group, these negative
experiences dissipate and job satisfaction improves
(Niemann & Dovidio, 1998).
2
1.5
0.5
o
Male dominated
academic area
THE EXPERIENCE OF DISCRIMINATION 427
Female dominated
academic area
II1II Women
Men
FIG U R E 11.1 Perceived Current Sex Discrimination by Gender of Respondent and
Academic Area
Female undergraduates in a male-dominated academic area reported higher levels of sex
discrimination than did female undergraduates in a female-dominated academic area or
male undergraduates in either academic area.
SOURCE: Adapted from Steele, James, and Barnett (2002).
men reported feeling under suspicion, especially about their sexual motives for
choosing a career in child care.
The majority of the research on tokenism has focused on women who
occupy nontraditional roles and remain the minority in those roles. Only a few
studies have examined the experiences of people of color (see Moses, 1989, for
one example). Jan Yoder (1997) studied the experience of Black women who
were training to be firefighters; these women were a double minority in that
setting, She found that their efforts sometimes were directly sabotaged by their
superiors and coworkers (see Box 11.1), Additional factors, such as one’s status
in an organization, also may affect one’s experience as a minority. Mary Kite
and Deborah Balogh (1997) found that untenured women faculty were more
likely than untenured nlen to report the kinds of negative interactions that are
typically associated witb the chilly climate, such as being excluded from social
events or having their comments ignored at meetings. Tenured women and
men did not differ in their reports about negative interactions, even though,
at that time, both tenured and untenured women were a statistical minority
at their university. This may be because their secure status or their experience
in the environment provided tenured women with a buffer from the effects of
tokenism.
428 CHAPTER 11
RESPONSES TO PREJUDICE AND
DISCRIMINATION
The personal experiences and experimental research we described in Chapter 10
provide a snapshot of the many and varied forms prejudice and discrimination
can take. We focus here on how these behaviors affect stigmatized group members.
The effects can be viewed on a continuum. At one end of the continuum
are discriminatory behaviors that can make stigmatized group members uncomfortable;
at the other end of the continuum are behaviors that cause menlbers of
stigmatized groups to lose job opportunities and that can affect their health and
well-being. Social psychologists have documented that stigmatizing experiences
can create uncertainty for m_embers of stigmatized groups, especially in how
to interpret interactions with members of the dominant group. We tum to this
research next.
Attributional Ambiguity
Most people find it difficult to talk directly about stereotyping and prejudice.
When members of dominant and stigm_atized groups interact, the topic of prejudice
can become the proverbial “elephant in the room.” Recall from our discussion
above, for exalnple, that when tokens are present, it increases the likelihood
the majority group members will think about their own group membership,
even if they do not discuss this awareness. In these situations, both dominant
and minority group members can become uncomfortable. One cause of this discomfort
is the ambivalent attitudes that dominant group mem_bers often hold
toward stigmatized groups. People who are not disabled, for example, often report
feelings of both sympathy and anger toward those who are (Dijker & Koomen,
2003; Fichten & Arnsel, 1986). Similarly, Whites often hold ambivalent attitudes
about Blacks, viewing the group positively on sonle dilllensions and negatively on
others (Czopp & Monteith, 2006). Moreover, as we discussed in Chapter 6, what
people are willing to say about stigmatized groups has changed; in the United
States, people today are much more accepting of the principle of equality for all
people and most people want to avoid the appearance of being prejudiced. This
does not mean, however, that prejudiced attitudes have disappeared.
Members of stigmatized groups are well aware of these mixed reactions. This
awareness leads to a situation that Jennifer Crocker, Brenda Major, and their
colleagues (Crocker, Voelkl, Testa, & Major, 1991) refer to as attributional
ambiguity. That is, members of stigmatized groups often find it difficult to interpret
feedback from dominant group members. Although such feedback may
-e based on the stigmatized group member’s actual ability or achievement, it also
may be based on feelings of sympathy or pity or on the desire on the part of the
Qominant group member to appear unbiased. Research shows, for example, that
Whites sometimes give more positive feedback to Blacks than to Whites for the
same poor performance, perhaps to avoid the appearance of being prejudiced
“(Harber, 1998; see also Chapter 6). Questions arise, then; for example, a Black
THE EXPERIENCE OF DISCRIMINATION 429
person might wonder whether his supervisor’s evaluation reflects his competence
or stems from the supervisor’s biases and prejudices. The answer to this question
is unclear, leaving the stigmatized person unsure about how to interpret the
feedback.
Crocker and her colleagues (1991) demonstrated the effects of attributional ambiguity
in a study that was supposedly about ftiendship development. Participants
were paired -with a White student in an adjoining room who was actually a confederate
of the experimenter. To manipulate whether the confederate knew the
participant’s race, either the blinds were drawn between the two rooms (so the
participant could not be seen) or they were not. The participants described
their likes and dislikes on a form that was allegedly shown to the confederate.
Next, participants received bogus feedback indicating that the other person had
either a positive or negative reaction to the information. When the other person
could not see them, Black participitants were more likely to attribute negative
than positive feedback to prejudice. In contrast, when Blacks knew their
partner could see them, they atttibuted both the positive and negative feedback
to prejudice. White participants’ attributions to prejudice were unaffected by
the valence of the feedback or by whether their partner knew their race.
These results suggest that stigmatized group members sometimes discouf}.t feedback
from the majority group because they believe it is based on factors other
than their ability or performance (Crocker & Major, 2003).
Members of stigmatized groups do not always discount positive feedback,
however, and instead may augment it, or conclude that the positive evaluation
was due to their own deservingness. In one study, for example, unattractive people
found positive feedback to be more believable than did attractive people in a
similar situation, perhaps because the unattractive participants did not think their
partner had ulterior motives-in this case, an other-sex attraction toward them.
Attractive people, in contrast, may have assumed attraction played a role and, as
such, discounted the feedback (Major, Carrington, & Carnevale, 1984). When
ulterior motives are not suspected, then, people are more likely to conclude
that the feedback is due to their abilities or characteristics, but when such
motives are suspected, members of stigmatized groups do not believe the positive
feedback and conclude that it was due to factors such as the evaluators’ desire to
appear unbiased or to their sympathy or pity.
What are the psychological consequences of receiving unclear feedback?
Research shows that such consequences depend on whether stigmatized group
members augment or discount the feedback and on the valence of the feedback.
Discounting negative feedback, for example, has self-protective consequences. In
the Crocker and colleagues (1991) study described above, Black participants who
could attribute negative feedback to prejudice were less depressed than those
who could not. In contrast, discounting positive feedback tends to produce
lower self-esteem, even when compared to those individual., who received negative
feedback (Crocker et al., 1991). However, atttibuting negative feedback to
prejudice is not always beneficial; stigmatized group members who frequendy see
themselves as a victim of prejudice in many situations have lower self-esteem
(Branscombe & Ellemers, 1998; Branscombe, Schmitt, & Harvey, 1999), a point
430 CHAPTER 11
we return to later in this chapter. Moreover, individuals who strongly identify as
members of their social group do not psychologically benefit from attributing
negative feedback to prejudice (McCoy & Major, 2003).
Interestingly, both stigmatized and dominant group Inel11bers are aware that
people are likely to consider another’s racial group menlbership when evaluating
thein. For example, Bruce Blaine and his colleagues (Blaine, Crocker, & Major,
1995) asked undergraduates to put themselves in the shoes of Blacks or women
and imagine how they would feel if they were offered a job for one of two
reasons: they were qualified for the job or the employer felt sympathy for past
discrimination against their group. Participants who imagined the job offer was
motivated by sympathy reported that they would have lower self-esteem, higher
depression, 1110re hostility, and lower motivation than those who imagined the
job offer was based on their qualifications. Follow-up studies showed that these
effects were quite general. For example, similar patterns emerged regardless of
whether the employer felt sympathy because of enlployment discrimination specifically
or felt sympathy for the stigmatized group in general, in this case people
with disabilities (Blaine et al., 1995). Research also has found that both Whites
and Latinos who portrayed Latinos in a virtual world discounted negative feedback
about their performance lllore than those who portrayed Whites in this
virtual world (Hoyt, Aguilar, Kaiser, Blascovich, & Lee, 2007).
Interestingly, self-esteenl can be affected even if the stigmatized group member
does not blame the evaluator for the negative feedback (Crocker, Cornwell,
& Major, 1993). Overweight women who received negative social feedback from
a male evaluator were more likely to attribute the negative feedback to their
weight than were normal-weight WOlllen, but did not dislike the evaluator for
providing this feedback. That is, they did not attribute his feedback to prejudice.
Even so, the overweight women who received negative feedback reported being
in a lllore negative mood, and reported higher levels of depression and anxiety
than did normal-weight women who received negative feedback or women of
any weight who received positive feedback. Overweight women are not getting
a boost to their self-esteem, but instead may shoulder the blame for their weight.
This may have harmful consequences. For example, the overweight may conclude
that they will fail at dieting or following an exercise program and therefore
not try or give up too easily. If, instead, the overweight had benefited from this
buffer, they might be more willing to try and succeed in these endeavors.
In summary, research on attributional anlbiguity shows that stigmatized people
consider the source when receiving feedback and, if they believe the source is
prejudiced against them, weigh that feedback differently. When the feedback is
negative and can be attributed to prejudice they discount it. They also sonletimes
discount positive feedback because they doubt its validity, believing instead it
stemmed from the evaluator’s ulterior lllotives. Other times, for example when
ulterior motives are not suspected, stigmatized individuals augment the positive
feedback, deciding it must be due to their ability or characteristics. For minorities,
discounting negative feedback has beneficial effects on self-esteem, but discounting
positive feedback does not. Overweight women, however, do not gain
a psychological benefit from attributing negative feedback to prejudice.
THE EXPERIENCE OF DISCRIMINATION 431
Personal/Group Discrimination Discrepancy’
You have no doubt heard about serendipitous research findings that were at first
puzzling but later led to important new theories and research. Faye Crosby
(1984) stumbled across just such a phenomenon when she surveyed working
women who lived in a Boston suburb. Objective indicators showed that these
women were being discriminated against; for example, the women earned between
$5,000 and $8,000 less than men for equivalent jobs. Yet Crosby also
found that the women were just as satisfied with their job as the men were.
Perhaps even more puzzling was that the women were well aware that sex discrimination
existed in the United States and, moreover, they were aggrieved by
this state of affairs. They just did not believe this discrimination was happening in
their own lives.
Crosby’s (1984) surprising finding has led to a great deal of research on Wha~’]
is now known as the personal/group discrimination discrepancy (PGDD),
people’s belief that their group, as a whole, is 1110re likely to be discriminated
against than they, themselves, are as individuals (Taylor, Wright, Moghaddam,
& Lalonde, 1990). Researchers have reported findings consistent with this hypothesis
vvith groups as diverse as Black college activists, French Canadians in Quebec,
Canada (who live in a largely English-speaking country), English-speaking residents
of Quebec (where French is the dominant language), unemployed workers
in Australia, and lesbians (see reviews by Crosby, Pufal1, Snyder, O’Connell, &
Whalen, 1989 and Taylor, Wright, & Porter, 1994), Figure 11.2 illustrates the
pattern the PGDD generally follows; lesbians perceived higher levels of discrimination
for their group at a national and a local level than for themselves, They also
believed lesbians at the local level and the national level had a greater need to hide
their sexual orientation at work than they themselves did (Crosby et al., 1989).
Cognitive Explanations. There are two main categories of explanations fO~-]
the PGDD: cognitive and motivational. Proponents of cognitive explanations
suggest the personal/group discrimination discrepancy is simply a function of the
way people process information. For example, Faye Crosby and her colleagues
(Crosby, Clayton, Alksnis, & Hemker, 1986) found that when participants evaluated
information about discrimination in the aggregate form (that is, they read
about patterns of discrimination compiled over several individuals), they believed
that discrimination occurred. But when the same information was presented on a
case by case basis, they failed to perceive discrimination. Hence, the way in which
the infonnation was presented and processed either produced or inhibited the
perception of discrimination. Research also has demonstrated that the PGDD is
quite general, applying to domains unrelated to discrimination, such as the economy
and the threat of A1DS (Moghaddam, Stolkin, & Hutcheson, 1997). People
believe, for eXaIllple, that a good economy is more likely to benefit their group as
a whole than them as individuals. Such findings suggest a general process is operating
that extends beyond perceptions of discrimination. It may be that the
PGDD emerges because group examples more readily come to mind than do individual
examples or because group information is more easily processed than is
432 CHAPTER 11
6
5
4
3
2
Perceptions of
discrimination
II National
!iii Local
Personal
Need to maintain
a heterosexual faQade
FIG U R E 11.2 Lesbians’ Ratings of Perceived Discrimination at the Personal, Local,
and National Level
Lesbian respondents saw more evidence of discrimination at the local level than at the
personal level and the highest level of discrimination at the national level. They also
believed lesbians at the local level and the national level had a greater need to hide
their sexual orientation at work by appearing heterosexual than they themselves did.
SOURCE: Adapted from Crosby, Pufall, Snyder, O’Connell, and Whalen (1989).
infonnation about the sell. Supporting this possibility, research shows that the
PGDD is found for perceptions of positive events as well as negative events; people
believe, for example, that the group, overall, is more likely than they, as individuals,
to have warn’! and supportive friends or to benefit from the improved
efficiency of computers (Moghaddam et al., 1997).
~ Another cognitive-based explanation is that people are using different comparison
standards when judging their own versus the group’s level of discrimination.
That is, when deciding about their personal experience with discrimination,
people consider their experiences in comparison with their own group melnbers,
but when deciding about the groups’ discrinrinatory experiences, they compare
themselves to other groups (Taylor et al., 1994), Women, then, may believe that
they, personally, are better off than most won’!en, but that their group is doing
worse, on the whole, than men are. If this is the case, making the referent group
explicit should reduce the PGDD. Research supports this possibility (Quinn,
Roese, Pennington, & Olson, 1999). Ratings made in the absence of a referent
led people to use ingroup comparisons for judgments of personal discrimination
and outgroup comparisons for judgments of group discrimination. Moreover,
providing a specific referent to an ingroup, in this case by asking women to
compare their personal level of discrinunation to other women’s, reduced the
PGDD. Perceptions of how often discriminatory acts occur also affects the PGDD;
THE EXPERIENCE OF DISCRIMINATION 433
women see a smaller PGDD for discriminatory acts thought to occur frequently in
the workplace, such as being told to act in feminine ways, com-pared to events
thought to occur infrequently, such as not receiving the same raise as their male
colleagues (Fuegen & Biernat, 2000).
Taken as a whole, these studies show that the way people process infonnation
in general, and about discrimination specifically, affects their views about
their own and their group’s experience with discrimination. It should be noted,
however, that neither making the referent group explicit nor including information
about frequency or severity of discrimination completely eliminates the
PGDD. For example, making the referent explicit by asking women to compare
their level of discrimination specifically to men’s does not reduce the PGDD
(Quinn et al., 1999). Coguitive explanations tell us something, but not everything,
about why the PGDD occurs (Taylor et al., 1994).
Motivational Explanations. Motivational explanations assume that peOPle~]
have reasons for believing that they are not personally discriminated against,
even while recognizing that their group is. The motivational explanation that
has received the most support is Crosby and colleagues’ (1986) hypothesis that
people want to deny or minirnize their own experiences with discrimination.
Recall from our earlier discussion that, in Crosby’s studies (Crosby, 1984;
Crosby et al., 1986), people reported little personal experience with discrimination,
even though by objective indicators discrimination existed. There are
several reasons why individuals might deny their personal experiences with discrimination
(Taylor et aI., 1994). In some instances, individuals might take personal
responsibility for their situation, and thus not acknowledge that the poor
treatment they received could be due to discrimination. In other instances, people
deny discrimination to justify their failure to accuse a specific discriminator or
their decision not to take action against the unfair treatment. Finally, people may
view their own situation as relatively hannless compared to more dramatic examples
of discrimination, particularly those highlighted in the mass media.
Mauricio Carvallo and Brett Pelham (2006) have proposed another reasoD.Jpeople
deny personal discrimination: they have a strong need to affiliate and
bond with other people. These authors note that stigmatized group members
often are discriminated against by people with whon1 they have meaningful relationships,
such as mends and coworkers. Therefore, acknowledging this discrimination
would mean also acknowledging that they do not fit in with their
social group. Consistent with this perspective, Carvallo and Pelham found that
people who have a high need to belong were more likely to believe that their
group experiences discrimination but were less likely to believe that they personally
are discriminated against. They also found that when people were made to
feel accepted by their group, they were more likely to acknowledge personal
discrimination than were individuals in a control condition.
People also can recognize the social costs of claiming discrimination and, as a
result, try to avoid those costs by not making claims of unfairness (Kaiser & Miller,
2001a). The social costs of claiming discrimination include being viewed negatively
by members of the dominant group, such as being labeled a whiner or someone
434 CHAPTER 11
who takes advantage of possible discrimination for personal gain (Feagin & Sikes,
1994). To see if such outcomes occur, Cheryl Kaiser and Carol Miller (2001a)
asked introductory psychology students to read a description of a Black student
who failed a test that had been scored by one of eight White judges. The potential
bias he faced was lluuipulated: either none, four, or all of these White judges
reportedly had a history of discriminating against Blacks. The research participants
learned of this possible discrimination and that the student had failed the
test. They also learned that the Black student attributed his failure to either the
quality of rus answers or to discrimination. Regardless of how much possible
prejudice he had faced, participants were more likely to label the student as a
complainer and to evaluate him less favorably when he nude attributions to discrimination
rather than ability. Interestingly, however, the student who attributed
his failure to discrimination also was seen as truer to himself than the
student who attributed his failure to abiliry.
Additional research suggests that m_embers of stigmatized groups are aware of
such perceptions and this awareness affects their decision to report or confront
discrimination. Women and Blacks who received a failing grade on a creativity
test, for example, were more likely to attribute the failure to discrimination
when reporting their attributions anonymously or when their explanation would
be seen only by a stigmatized group member, compared to when the explanation
would be seen by a dominant group member (Stangor, Swim, Van Alien, &
Sechrist, 2002). For mem_bers of nonstigmatized groups, attributions were unaffected
by who would see the results. The stakes of the encounter matter, too.
Nicole Shelton and Rebecca Stewart (2004), for example, found that women
who were being interviewed for a competitive, high paying job were less likely
to confront a male interviewer who asked sexist questions than were women
who were being interviewed for a low paying, less competitive job. This awareness
also may playa role iu the PGDD; supporting the idea that members of
stigmatized groups want to distance themselves frOlU negative attributes associated
with their group, Gordon Hodson and Victoria Esses (2002) found that
women were more likely to think that negative attributes (including being the
target of discrimination) applied to the ingroup than to themselves, suggesting
that they wanted to distinguish themselves from the group on these attributes.
However, this distancing was not found for positive attributes; instead, they
were nlore likely to report that positive attributes applied to themselves than to
the ingroup. These effects were luore pronounced for women who strongly
identified with their group and, therefore, were luore invested in how they
and their group were perceived.
Perceiving Discrimination. It would be incorrect to conclude, however, that
people never recognize that they are personally being discrinrinated against. Donald
Taylor and his colleagues (1990) found that both Haitian and Indian iumligrants to
Canada reported significant personal experience with discrimination, even though
they believed that their group experienced more discrimination, as a whole, than
they did as individuals. Sinrilarly, single mothers receivillg government assistance
reported feeling that their lives were sonlewhat unfair and evidenced resentment
THE EXPERIENCE OF DISCRIMINATION 435
toward their situation, but still believed themselves to be better off than other
mothers in their situation (Olson, Roese, Meen, & Robertson, 1995). There also
may be individual differences in the tendency to minimize one’s own experience
with discrimination; the more typical of their group that Iranians perceived themselves
to be, the less likely they were to exhibit the personal/group discrimination
discrepancy (Verkuyten & Nekuee, 2001). Similarly, Don Operario and Susan
Fiske (2001, Study One) found that non-Whites who were low and high identifiers
with their group reported equal amounts of discrimination directed at their group,
but differed in their perceptions of personal discrimination: high identifiers were
more likely to report discrimination directed at themselves than were low identifiers.
Results of a second study (Operatio & Fiske, 2001, Study Two), suggest that
this pattern emerged because high identified minorities are more sensitive to possible
discrimination and, therefore, react to both subtle and obvious indicators of
prejudice, whereas those less highly identified reacted ouly to obviously prejudiced
actions. That is, those who identifY strongly with their group may simply be more
likely to notice and react to subtle fonus of prejudice.
Interestingly, research suggests that this heightened sensitivity might he
counterproductive. Elizabeth Pinel (2002) found that women who were high
in stigma consciousness-that is, who believe that they live in a stereotyped world
and that this affects their interactions with outgroups-were more critical of men
who they believed to be sexist. This criticism, in tum, elicited more negative
criticism from those men. The end result was that the women concluded that
they were incompatible with the sexist men. No such effects emerged for
women low in stigma consciousness or for women who believed they were interacting
with nonsexist nlen. It is important to note that these results emerged
independently of the men’s actual sexist beliefs; the experimenter controlled
who was described as sexist. Therefore, the differences in ratings were due to
the women’s expectations and how the interaction was affected by them, and
not to sexist behavior on the men’s part.
It may seem that there is a contradiction between the studies on attributional
ambiguity, described above, which suggests that members of stigmatized groups
are well aware that they personally might be discriminated against because of
their stigmatized status and the PGDD, which suggests that people deny personal
discrimination. As often happens in social science, bodies of literature address
questions in different ways, resulting in seemingly subtle differences that account
for such contradictions. In this case, studies of attributional ambiguity focus on a
single, specific, instance of discrimination, whereas the personal/group discrepancy
focuses on broad patterns of perceived discrimination. For example, studies
of attributional ambiguity that focus on one interaction show that attributions to
prejudice can protect self-esteem. This protection is not evident when individuals
make attributions to broader, more stahle patterns of discrimination; that
is, when they consider, overall, how prejudice affects them as an individual, people
report that their psychological well-being is adversely affected (Branscombe
et al., 1999). But, as we saw with the personal/group discrimination discrepancy,
they still may believe these effects are worse for their group than for themselves.
436 CHAPTER 11
We have more to say about the effects of experiencing discrimination on mental
health later in this chapter.
CONSEQUENCES OF PREJUDICE TO THE TARGET
During the 1990s, there;were impressive increases in minority group members’
and women’s participation in undergraduate and graduate education. Women,
for example, are now lllore likely to enroll in college than are men and Blacks
and Hispanics are enrolling in record llU1ubers. These gains, however, do not necessarily
translate into greater academic achievement for these groups. Minority student
attrition rates are higher than Whites’ at both the graduate and undergraduate
level, and both women and minorities continue to be underrepresented in science
and engineering (National Science Foundation, 2002). Moreover, college
entrance exam scores continue to differ by sex and ethnicity. Boys, for example,
score higher than girls on the math section of the SAT and Whites score
higher on both the math and verbal sections than do Blacks and Latinos
(College Board, 2003).
One explanation that has been offered for these differences is that women and
minmities are not as able or as well prepared as their White male counterparts
(Benbow & Stanley, 1980; Herrnstein & Murray, 1994). Yet, abundant evidence refutes
this clalill. For example, when women and minorities participate in programs
designed specifically for underrepresented groups, they can and do succeed (Fullilove
& Triesman, 1990; Grinnnett, Bliss, & Davis, 1998). Moreover, research indicates that
girls receive higher grades in math courses than do boys (Kimball, 1995) and that
males’ math advantage may be limited to certain types of standardized tests. Scores
on high school achievement tests in 10 U.S. states, for exam-ple, showed no gender
different in scores on the math portion of the exam (Hyde, Lindberg, Linn, Ellis, &
Williams, 2008). Hence, the accuracy of the stereotypic belief that wonlen and minorities
are less capable than men or White students is highly suspect, As we will discuss in
the next section, situational factors have an important influence on the success of individuals
who are underrepresented in a specific discipline (such as women in math
and science) or in an academic setting more generally (such as Blacks at most colleges
and universities).
Stereotype Threat
If situational factors can raise the achievement of women and minorities, can
they also hinder their performance? Research evidence suggests that they can.
Consider, for exalllple, that Blacks are well aware that a negative stereotype
exists about their academic abilities. According to Claude Steele (1997), this
knowledge produces a “threat in the air” (p. 617). Blacks realize that they can
be judged or treated in terms of this negative stereotype and can be fearful of
confirming that judgment. If this fear is strong enough and also is personally
relevant to the stereotyped group member, it can create a stereotype threat
THE EXPERIENCE OF DISCRIMINATION 437
that interferes with academic achievement (Aronson, Quinn, & Spencer, 1998).
As we will see, this phenomenon can affect a person’s behavior even though no’–”
discriminatory actions actually were directed toward her or him.
In one of the first demonstrations that stereotype threat affects Blacks’
achievement, Claude Steele and Joshua Aronson (1995) asked Black and White
undergraduates to take a test composed of the most difficult verbal questions
from the Graduate Record Exam (GRE). Half of the participants were told the
test was diagnostic of intellectual ability (the diagnostic condition); the other half
were told the test was simply a laboratory problem-solving task (the nondiagnostic
condition). Steele and Aronson proposed that the diagnostic condition induced
stereotype threat for Blacks because their exam performance could support or refute
the stereotype that Blacks have low verbal ability. Supporting tills hypothesi” in
two separate studies, Black participants in the diagnostic condition scored lower
than Blacks in the nondiagnostic condition or Whites in either condition. Figure
11.3 presents these results for the number of items solved correcdy, collapsed across
Studies One and Two. Results of a third study showed that Blacks who were told
the test was diagnostic also were more likely to complete word fragments in terms of
the social stereotype of Blacks (for example, completing __ Z Y as LAZY) than were
Blacks who participated in the nondiagnostic condition, or Whites in either condition.
Similarly, compared with their peers in other conditions, Blacks in the diagnostic
condition were more likely to complete word fragments in a way that indicated
self-doubt (for example, completing L a ___ as LOSER) and were more likely to
distance themselves from stereotypically Black activities, such as liking jazz or basketball.
Because these tasks were completed before the actual diagnostic test was taken,
~
12
In
~ 10
“C * 8 ~
‘6”
.!.
“.C. 6 Black participants
~
0 ‘” 4 ‘” • White participants
E
~
..c.. 2 ;:;;
0
Diagnostic Nondiagnostic
FIGURE 11.3 Mean Items Solved by Participant Race and Test Diagnosticity
Blacks’ performance on a test of verbal ability were affected by whether the test was
described as diagnostic, and thus produced stereotype threat, or nondiagnostic
(nonthreatening). White’s performance was unaffected by how the test was described.
These scores are adjusted for overall verbal ability, as measured by the SAT.
SOURCE: Adapted from Steele and Aronson (1995) Studies One and Two.
438 CHAPTER 11
these findings suggest that the mere expectation of taking a potentially stereotypecon£
nning test brought up stereo typic thoughts, self-doubt, and a desire to be seen
as different from the Black stereotype.
General Features of Stereotype Threat. There are several keys to understanding
how stereotype threat operates (Aronson et a!., 1998; C. Steele,
[
Spencer. & Aronson, 2003). One is that stereotype threat stems from situational
pressures that bring the stereotype to mind, not merely from internalization of
the negative stereotype. In fact, people need not believe the stereotype about
—their group, or even be worried that it applies to then’! personally, for it to influence
behavior. Claude Steele (1997) cites the example of a Black man waiting at
an ATM machine for a woman to complete her transaction. This man is likely
aware of the stereo typic belief that Black men are violent and, even though he
himself has no violent intentions, he might be still concerned that the wonlan
will fear him. The situations that produce stereotype threat vary widely and
range from the diagnosticity of the test, as we saw in the Steele and Aronson
(1995) experiment, to whether a person is a minority in a situation. Michael
Inzlicht and Talia Ben-Zeev (2003), for example, showed that women who
took a math test in a group of other women, and were therefore in the majority,
scored higher than women who took the same test in a group of other men and
were therefore in the minority. The effects of stereotype threat have been
demonstrated both in the lab and in naturally occurring environments. I A second key feature of stereotype threat is that it is a quite general process L that can affect any group for which a negative stereotype exists. Jean-Claude
Croizet and Theresa Claire (1998), for example, investigated whether stereotype
threat could result from the stereo typic belief that individuals from a lower socio~
economic status (SES) are less intelligent than those from a higher SES background.
Participants were French undergraduates from both high and lower SES who
completed the verbal portion of the GRE under one of two instruction sets:
The test was described either as an assessment of intellectual ability for solving
verbal problems (the diagnostic condition) or as an assessment of lexical memory
(the nondiagnostic condition). Results supported the stereotype threat hypothesis.
Lower SES participants in the diagnostic condition tried fewer items and
answered fewer questions correctly than did lower SES participants in the nondiagnostic
condition. Scores for the higher SES participants were not influenced
by instruction set.
Stereotype threat also has been demonstrated for women and Latinos and
has been shown to operate in a variety of academic settings ranging from middle
schools to private and public colleges and universities (c. Steele et al., 2003).
Evidence for stereotype threat also has been found in a number of perfonnance
domains, such as athletics (Stone, Lynch, Sjomeling, & Darley, 1999) and the
workplace (Roberson, Deitch, Brief, & Block, 2003). Despite this generality,
certain conditions are more likely to produce stereotype threat; for example,
more negative stereotypes generally produce stronger stereotype threat
(c. Steele et al., 2003). Moreover, stereotype threat has its strongest effects on
those individuals whose self-esteem is tied to their pelfonnance in a domain or
THE EXPERIENCE OF DISCRIMINATION 439
who have the greatest chance for success in that area of achievement. As Joshua
Aronson and his colleagues (1998) put it, “all other things being equal, the more
a person cares about being good at something, the greater will be his or her
distress about stereotypes alleging a lack of ability” (p. 87). Whites who believe
their athleticism is important to their identity are more threatened by the stereotype
that “Blacks are more athletic that Whites” than are Whites for whom athleticism
is unimportant (Stone et al., 1999). Similarly, women who strongly
identifY with being a woman are more likely to experience stereotype threat
on a test of their math ability than those who do not have a strong gender identification
(Schmader, 2002).
A third key feature of stereotype threat is that the nature of the threat varieU
by the specific context of the negative stereotype. The type of threat that would
affect women, for example, could be very different from the type of threat tha
would affect older adults or athletes. Women, for example, are likely threatened
in the arena of mathematical ability, older adults in the area of memory, or athletes
on the football field. Moreover, stereotype threat does not generalize to
other situations. Women’s pertormance on an English exam, for example, would
not be hindered by a threat about their mathematical ability. Stereotype threat
also can be produced in groups whose members are not normally threatened by a
belief. White men generally do not worry about their math ability, for example,
and are not stereo typically believed to do poorly in math. Yet White men experienced
significant performance drops when they believed the test was designed
to determine why Asian males outperform White males in math (C. Steele et al.,
2003).
Evidence suggests that stereotype threat operates hy changing the way information
is processed, specifically by reducing people’s working memory capacity
(Schmader & Johns, 2003). That is, stereotype threat is not just an emotional reaction
to the possibility of confirming a stereotype about one’s group but also a cognitive
reaction. In one relevant experiment, working menl0ry was assessed by the
operation span test, during which participants evaluated mathematical equations
while memorizing words for later recall. Male and female undergraduates were
told that this was a test of the ability to either remem_ber two different pieces of
information simultaneously (nonthreatening condition) or to solve complex mathematical
equalities (condition threatening to women). Those in the threatening
condition also were told that gender differences in this ability might explain gender
differences in math perfonnance. Results showed that men’s operation span test
scores did not differ based on how the test was described. Women, however,
scored lower under stereotype threat conditions than under nonstereotype threat
conditions (Schmader & Johns, 2003). Apparently, the added threat taxes cognitive
resources, resulting in lowered attention for those under this threat. Interestingly,
holding stereotypic beliefS also can impair the cognitive performance of nonthreatened
group members; research demonstrating such effects is presented in Box 11.2.
Reducing Stereotype Threat. While reading about research confirming that
stereotype threat can affect achievement, you lllay have noticed an important
point. That is, when participants believed the test was not indicative of ability,
440 CHAPTER 11
In this chapter, we describe the many negative
consequences of stereotyping and prejudice to those
who are targets of these beliefs and actions. An impli~
cation one might draw from this discussion is that
there are no negative consequences for those who
hold stereotypic beliefs or discriminate against mem~
bers of stigmatized groups. Research suggests, however,
that this is not the case and that, instead, holding
negative stereotypes or discriminating against others
can impair people’s cognitive performance. For example,
Jennifer Richeson and Nicole Shelton (2003)
examined the effects of interacting with a Black person
on Whites’ executive function. Executive function
refers to the ability to plan, organize, and strategize;
when this function is impaired, as when resources are
overextended, cognitive performance suffers.
Participants in this study completed the Implicit
Association Test OAT; Greenwald, McGhee, & Schwartz,
1998), which, as you learned ‘In Chapter 2, is an indirect
measure of racial prejudice. Then, at the request of ei~
ther a Black or White experimenter, they were video~
taped while commenting on two controversial issuesone
of which was racial profiling in post-September 11
America. Finally, they completed the Stroop test, which
requires good executive function. Results showed that
the more negative the participants’ implicit attitudes
were, the more likely it was that interacting with a
Black person reduced their executive function. In con~
trast, interacting with a White person did not affect
executive function, regardless of the participants’
implicit racial attitudes. A subsequent study (Richeson
et aI., 2003) used a similar procedure, but also assessed
neural activity in the brain regions that control execu~
tive function. This activity was assessed as participants
responded to familiar and unfamiliar photographs of
Black faces. Changes in brain activity were significantly
correlated with racial attitude, and these changes also
predicted performance on the Stroop test. No such
relationships were found when participants responded
to White faces. These results further support the
hypothesis that, for those who are racially biased,
interracial contact impairs execut’lve function.
The ability to cope with stress is also compromised
for White individuals who are racially biased and are
being evaluated by a Black person. Wendy Mendes and
her colleagues (Mendes, Gray, Mendoza~Denton,
Major, & Epel, 2007) had adult White women and men
between the ages of 18 and 55 complete the rAT
online. Later, these individuals came to their laboratory
and gave a speech that was evaluated by two
interviewers-a task most people find to be stressful.
The interviewers were either both Black or both White.
Both physiological measures (assessed by neuroendocrine
responses) and behavioral measures (assessed by
interviewers’ ratings of the speakers’ anxiety level)
indicated that the racially biased White participants who
were evaluated by Black interviewers had the most
difficulty coping w’rth this stress, In contrast, White
speakers with egalitarian attitudes evidenced healthy
coping regardless of the interviewer’s race.
Although the research we have described found
larger deficits for individuals with negative racial
attitudes, evidence for more general effects also exists.
Specifically, priming non~Blacks with the Black
stereotype can lower performance on standardized
tests (Wheeler, Jarvis, & Petty, 2001). Participants who
wrote an essay about a day in the life of a Black college
student, and thus had their stereotypes about Blacks
primed, subsequently scored lower on the math section
of the GRE than did students who wrote about a White
student, and thus did not have their Black stereotypes
primed. These effects emerged regardless of their
scores on the Modern Racism Scale (McConahay,
Hardee, & Batts, 1981). Results of a second
study showed that these effects were stronger for
participants who included stereotypic content in their
essays, supporting the argument that stereotypic
beliefs played a role in producing them.
In this chapter, and throughout the book, we
have provided many reasons why stereotyping and
prejudice are harmful. As social psychologists know
well, personal involvement increases people’s attention
to persuasive messages, making it more likely that high
quality arguments will be accepted (Petty & Cacioppo,
1979). Perhaps raising awareness of findings such as
these can produce such increased involvement, leading
reluctant individuals to recognize the harmful effects
of prejudice.
those who would otherwise be threatened by group stereotypes performed
well. Thus, women, who believed a math test did not show gender differences
performed as well as men on that test (Spencer, Steele, & Quinn, 2001),
Findings such as these suggests that the way in which achievel11ent tests are
THE EXPERIENCE OF DISCRIMINATION 441
described to test takers may affect their scores. It is possible, for example, that
the combination of describing IQ tests as diagnostic and the awareness of a
test-related stereotype (for example, that those from a lower SES will achieve
lower scores) actually produces those lower scores. If so, SES differences might
disappear if the IQ tests generally were presented as non diagnostic (Croizet &
Claire, 1998). Yet, as Steele and his colleagues (2003) note, the diagnostic purposes
of standardized tests are well known and it seems unlikely that simple instructions
would override this effect outside the laboratory. Certainly, however,
taking care that instructions are as neutral as possible is important, especially for
tests that are not already labeled as diagnostic. Another way to reduce stereotype
threat is to teach students about the possibility that their perfonnance may be
affected by it. Women’s math tests scores were higher when they were told in
advance about the conditions that produce stereotype threat Oohns, Schmader, &
Martens, 2005).
Research also has demonstrated that the effects of stereotype threat are most
likely to be seen when the task at hand is difficult or frustrating and least likely to
emerge when the task is easier (Spencer et al., 2001). This may explain why
women do better than 11len in math courses hut not on the SAT; course grades
are based on previously studied material and so tests of those skills may be less
threatening. Interventions that reduce the stress associated with a test, then, may
reduce stereotype threat. Other promising interventions include encouraging students
by letting them know that intelligence and achievement are improvable or
by having them participate in well-designed programs for high-achieving (not
remedial) students (Aronson, Ftied, & Good, 2001). It is important to note that
such strategies have produced long-tenn improvements in achievement (see
Aronson et aI., 1998).
Providing role models also reduces stereotype threat. College women, for
example, perfonned better on a difficult math test after reading about successful
women, compared to women who read about successful corporations (McIntyre,
Paulson, & Lord, 2003). Stereotype threat also can be reduced by changing how
feedback is given. As we saw in our discussion of attributional alllbiguity, Blacks
have reasons to distrust feedback. But research shows that individuals who are
told both that standards are high and that they can achieve those standards accept
the feedback and are motivated to respond to it (Cohen, Steele, & Ross, 1999).
As Steele and his colleagues (2002) point out, such statements de-emphasize
negative stereotypes and affinn ability, thereby reducing stereotype threat. As
Joshua Aronson and his colleagues (1998) put it “there is nothing special about
the personalities, the belief systems, or the values of women and minorities that
undennines their perfonnance. Rather, we argue, they fall victim to a situation
that undermines their performance. This situation, which we have labeled
stereotype threat, arises when negative stereotypes are available as a possible
explanation for performance. What is hopeful about this analysis is that situations
can be changed” (p. 99). As researchers continue to discover ways to change these
situations, and the threat that accompanies them, the negative consequences of that
threat can be eradicated.
442 CHAPTER 11
Stereotype Lift. As we noted earlier, people are well aware of the diagnostic
purposes of intelligence and achievement tests; cultural expectations about which
groups tend to perfonn better on such tests also are well known (C. Steele,
1
1997). Research on stereotype threat documents how this knowledge can hinder
achieveluent for melubers of those groups for whom expectations are low. But
research also suggests that this same information can provide a perfOlmance
boost, or stereotype lfft) for members of nonstereotyped groups (K! alton &
-~Cohen, 2003). This lift can occur if members of nonstereotyped groups engage
in downward social con’lparisons; that is, if they are evaluating their abilities by
comparing themselves to others who are doing worse than they are (Fein &
Spencer, 1997). The boost that accompanies this comparison theoretically occurs
because these downward comparisons alleviate the doubt or anxiety associated
with possible failure in achievement-related domains.
This idea was tested by Greg Walton and Geoffi:ey Cohen (2003) who
reviewed 43 studies that assessed the test perfonnance of members of stereotyped
and nonstereotyped groups. Results showed that, overall, members of nonstereotyped
groups perfOlmed better when the test situation produced stereotype
threat for the members of the stereotyped groups. That is, what was a drain on
performance for the threatened group became a boost in perfomlance for the
nonthreatened group. Moreover, the effects of stereotype lift emerged even
when threat was introduced through situational cues, and thus was not made
explicit in the study. Such outconles suggests that members of nonstereotyped
groups autOluatically link negative stereotypes and intellectual tests and, therefore,
receive the resulting perfonnance lift.
The real-world implications of this are significant. Walton and Cohen,
for example, note that stereotype lift results in a 50-point advantage on the
SAT for White men compared to stereotype-threatened groups, a difference
large enough to make the difference in college admission decisions or awarding
of scholarships. Awareness of, and advocacy against, such inequities can result in
positive changes. The outcome of a legal challenge illustrates how this can
happen. Fair Test filed a legal complaint against the Educational Testing
Service and the College Entrance Examination Board (Test makers to revise national
merit exam to address gender bias, 1996), charging that the Preliminary SAT
(PSAT) was gender biased, resulting in girls being undenepresented in the group
of National Merit Scholarships. This bias resulted in boys receiving millions of
dollars lllore in scholarships than girls. As palt of the settlement, the PSAT was
revised; the new test has significantly increased the nmnber of female National
Merit Semi-Finalists. Consistent with research on reducing stereotype threat,
these gains were achieved by adding a writing component, an area of the test
that is not threatening to girls. Doing so should also theoretically reduce stereotype
lift to boys on the PSAT.
Vulnerability to Stress
Research on stereotype threat and stereotype lift addresses how prejudice and
discrimination can affect the economic and academic SLlccess of stigmatized
THE EXPERIENCE OF DISCRIMINATION 443
group members. Experiencing discrimination also can have a profound influence
on an individual’s physical and mental health. Such effects are linked to the stress
associated with chrome exposure to discriminatory actions and can have both
short- and long-tenn effects.
How Prejudice and Discrimination Can Produce Stress. It is well-established
that stress can produce psychological and biological changes that result in disease
(Lazarus, 1993). Kevin Allison (1998) has argued that either the chronic experience
of prejudice, such as repeatedly being stared at in social situations, or the experience
of a major individual incidence of prejudice, such as being threatened with physical
violence, can produce the chronic stress conunonly associated with disease. The impact
of subde prejudice is less intense than the impact of major incidents, but both
can have mental and physical health consequences (Sue et al., 2007).
Experimental evidence supports Allison’s (1998) assertion. For example,
Blacks who completed an achievement test under stereotype threat conditions
had elevated blood pressure levels compared to Blacks who took the test under
nonthreatening conditions or Whites under either condition (Blascovich,
Spencer, Quinn, & Steele, 2001). This finding suggests that Blacks’ higher incidence
of hypertension compared with Whites may stem, at least in part, from
their real-life experience in threat-producing situations. Evidence further suggests
that the effects are not limited to situations that meet the conditions for stereotype
threat. A study of nearly 2,000 Black Americans found a positive relationship
between experience with racial discrimination and high blood pressure
(Krieger & Sidney, 1996). Similarly, research shows that gay men who reported
experiencing anti-gay violence or discrimination during the previous year
showed higher levels of psychological distress than gay men who did not report
such experiences (Meyer, 2003).
A second way in which experiencing prejudice and discrimination can produce
stress is through events that are nonnative and nonnonnative for certain
stigmatized groups (Allison, 1998). Examples of normative events include
identity development and school socialization; these experiences are part of the
developmental process but differ across stigmatized groups. For example, it is
typical for adolescents to struggle with their sexual identity development. Yet, fot
gay and lesbian adolescents, this nonnative experience has the added stress that
stems from knowing prejudice toward their group is widespread. Heterosexuals,
for example, do not worry about “coming out” to parents or friends about their
attraction to members of the other sex; gay lllen and lesbians know that telling
others of their attraction to sanle-sex others can result in personal rejection or
physical and verbal abuse (pilkington & D’AugeIli, 1995). Correlational data reveal
that lesbian, gay, and bisexual (LGB) youth who have disclosed their sexual orientation
to family and friends are at a greater risk for a suicide attempt than LGBs
who have not made such a disclosure (Rotheraru-Borus & Fernandez, 1995).
N onnonnative life stresses are those experienced only by menlbers of stigmatized
groups. Minority immigrants, for example, often experience prejudice and discrimination
during the acculturation process either because of language difficulties or
444 (HAPTER 11
because they violate cultural nonns, increasing their stress levels as they try to adapt
to their new environment (Allison, 1998).
Prejudice and discrimination also can produce stress through indirect means.
That is, some life events are more likely to happen to individuals from stigmatized
groups~not because of their group membership, per se, but because of situational
factors related to their group membership. Children from a low SES background,
for example, are more likely to attend poorly funded schools, live in lower quality
housing, and have poorer nutrition than students from higher SES backgrounds, all
of which affect their physical and mental health and their success in schoo!. These
effects are indirectly related to ethnicity (Arnold & Doctoroff, 2002). That is, race
in and of itself does not predict academic achievement; given similar opportunities,
both Blacks and Whites are successful in school. However, because ethnic minority
children are overrepresented in lower SES groups, they are more likely than
White children to experience academic failure that results from poverty. African
Americans are more likely than their White counterparts to experience a variety
of stressful events, including the divorce or separation of their parents, death of a
friend, birth of a sibling, or residence in a violent neighborhood (Garrison,
Schoenbach, Schluchter, & Kaplan, 1987). These stressors create a cycle of stress;
research has demonstrated, for example, that experiencing a high level of stress
makes individuals lnore vulnerable to the effects of subsequent stressors (Allison,
1998). Evidence also suggests possible indirect effects of higher SES for Blacks.
Specifically, mortality rates are higher for Blacks with more advantaged socioeconomic
status, perhaps because these individuals have more day-to-day contact
with Whites in traditionally all-White workplaces (Feagin & McKinney, 2003). As
we saw in our discussion of tokenism, when a person of a different ethnicity joins
a formerly all-majority group, the majority is more likely to think about ethnicity,
and perhaps to be uncomfortable with having to do so. This discomfort can affect
interactions betvveen minority and majority members which, in tum, can result in
stress for the minority group member that may affect both mental and physical
health. Famous minorities are not imnlune fron1 this stress, as the examples in
Box 11.3 illustrate.
Stress-Related Responses to Prejudice and Discrimination. As we noted
above, stress produces physiological changes that can affect both mental and
physical health. What effects do those stressors have on the people who experience
them? To find out, Joe Feagin and Karyn McKinney (2003) conducted
focus groups with economically successful Mrican Americans to explore their
perceptions of the effects of racism on their physical and mental health.
Respondents often noted that racially-related stress affected their immune
systems, making them more susceptible to colds and other diseases. Reports of
elevated blood pressure, headaches, insomnia, and stomach problems also were
common. Some respondents referred to “nine-to-five” headaches that would
be present while they were at the workplace, but would lift on leaving. As one
respondent put it, “I would have this headache. And it would be for eight hours
until I walked out the door and then it was like a weight was lifted off” (p. 73).
People of color remain a minority in many arenas,
including certain sports and in the television and movie
industry. Are the effects of this minority status
outweighed by success? Consider this: Arthur Ashe,
America’s first Black male tennis star, suffered much
adversity in his life, including losing his mother at a
young age and acquiring AIDS from a blood transfusion,
which led to his death. Yet when asked to
describe his most difficult challenge, he replied, “Being
Black is. No question about it. Even now it continues to
feel like an extra weight tied around me” (quoted in
Schuster, 1993, p. le).
Even respected journalists have been known to
make unfortunate racist comments concerning famous
sports figures. First, Golf Channel anchor Kelly Tilghman
stated on air that the only way for challengers to compete
against the successful golfer, Tiger Woods, would
be to “lynch him in a back alley.” This unfortunate
comment led to Golfweek magazine featuring a noose
on the cover of its January 19, 2008 issue. The ensuing
controversy resulted in the firing of the magazine’s
vice president and editor, Dave Seanor (Golfweek fires
editor, 2008). As we discussed in Chapter 1, nooses are
a strong symbol of hostility toward Blacks; the use of
such a symbol by major media outlets should be
unthinkable.
Another example of how racism can affect even
the most successful comes from the media’s response
to the question of whether the prestigious Masters
Golf Tournament should open its doors to women
(Vitello, 2004). The club that hosts this tournament
only allows male members, a tradition that was
challenged publicly by Martha Burk, head of the
National Council of Women’s Organizations. Yet who
was put on the spot to respond to this challenge? Paul
Vitello (2004) asked his readers to consider which of
these possibilities was the most likely candidate: (a) the
club members themselves; (b) the Professional Golfer’s
Association; (c) any number of White male golfers who
have played in or won the tournaments and/or the
women they love; and (d) CBS, the network that
broadcasts the Masters. The answer is none of the
above; Tiger Woods, the golf phenomenon of Black
and Thai ethnicity, was the person singled out and asked
THE EXPERIENCE OF DISCRIMINATION 445
to boycott the tournament. As Vitello (2004) writes,
“somehow, because the subject was discrimination, the
attention turned to the man of color-the guy whose
ancestors were hurt most by the ugly history of
white-men-only discrimination in America” (po 253).
Never mind that the issue concerned sex discrimination
and that, by boycotting the tournament, Woods would
have given up the chance to win the Masters three years
in a row, a feat no one has accomplished. As is often the
case for tokens, Woods was put in the position of
responding not only to actions against his race, but to
all discriminatory actions.
One of the stressors these athletes were responding
to concerned being asked to serve as a spokesperson
for their race. Arthur Ashe, for example, was
referring both to his experiences with discrimination
and to his position as the first Black tennis star, noting
that this role often put him in the position of being
spokesperson for his race, a common experience for
members of underrepresented groups (Sandler & Hall,
1986). He could never be quite confident that the
attention he received was due to his success and not his
race. In response to both the decision to include the
noose on the cover of Golfweek and to the controversial
issue of Augusta National having women as members,
Woods was asked to be a spokesperson for all
underrepresented groups. In both cases, his comments
simply referred to such incidents as unfortunate.
On the surface, asking individuals to speak for their
group orto stand up for discrimination seems like a
supportive gesture on the part of the majority. After all, it
does recognize that differences in perspectives can exist.
A closer examination, however, shows the problem with
the approach. First, it is possible that the majority and
minority groups agree on an issue. Second, as we saw in
Chapter 3, it reflects a belief in outgroup homogeneitythat
outgroup members are all alike and one person can
speak for the entire group. Third, it puts the minority
group member in the spotlight, which results in her or his
actions being highly scrutinized (Fiske & Taylor, 1991).
Not surprisingly, most minority group members find this
extra attention uncomfortable and would prefer that all
perspectives representing their group’s viewpoints were
heard (Sandler & Hall, 1986).
Similarly, college students who reported having fewer positive interracial interactions
also reported having more headaches and chronic fatigue and students who anticipated
experiencing racially biased treatment scored higher on a measure of psychological
distress (Littleford & Wright, 1998).
446 CHAPTER 11
Feagin and McKinney also found that one way many Blacks coped with these
stressors was by engaging in behaviors that are more or less socially accepted but
unhealthy such as alcohol and tobacco use or excessive eating. Although such coping
mechanisms can help reduce stress in the short-run, they also can obviously
have long-teml negative health consequences. These coping mechanisms may also
affect their children’s well-being. Fredrick Gibbons and his colleagues (Gibbons,
Gerrard, Cleveland, Wills, & Brody, 2004) found that parents’ experience of racial
discrimination led to increased parental distress, which indirectly increased parental
substance use at the time and two years later. Moreover, their children were aware
of their parent’s distress and this awareness was related to the children’s future
substance use. Results also showed, however, that effective parenting reduced
the chances that the child would smoke or drink alcohol in the future.
A common, but far from universal, response to racism is frustration, anger,
and even rage (Feagin & McKinney, 2003; Swim, Aiken, Hall, & Hunter, 1995).
Although members of minority groups may be reluctant to express these feelings,
failing to do so can produce higher blood pressure and greater sleep disturbance
(E. Johnson & Greene, 1991). What is the best way to cope? Feagin and
McKinney (2003) offer several effective sttategies including internally focused or
cognitive-based coping, such as adjusting one’s own attitudes, being more
accepting of what appears to be unintentional discrimination, becoming desensitized
to discriminatory acts, or framing the acts as due to White ignorance. Also
effective are behavioral strategies, such as verbally confronting discrimination or
protesting through formal channels.
Another factor central to effective coping is knowing when one has or lacks
control of either the situation or the response to it. As one respondent from
Feagin and McKinney’s (2003) sample put it, “When I feel that I’m in control,
I never get angry …. And I am in control of every racial situation. I am not the
sick one …. I’m in control, I don’t get mad anymore” (p. 140). By recognizing
what can and cannot be controlled, members of stigmatized groups can minimize
the psychological costs they incur by reacting-that is, how their own anger can
affect their lives-and also can reduce the likelihood that they will internalize the
incident, realizing it is not them but the other person or the situation that is to
blame. If parental socialization includes a discussion about racisnl, this strategy
also appears to provide a buffer against the negative mental health consequences
of experiencing discrimination (Fischer & Shaw, 1999). These strategies focus on
how members of stigmatized groups can address prejudice and discrimination.
Of course, the burden of addressing these beliefS and behaviors does not rest
solely with them. In Chapter 14, we discuss in detail strategies privileged group
members can use to reduce prejudice and discrimination.
Threats to Self-Esteem
Fifty years ago, most social scientists would have said that minority group members
have low self-esteem. Theoretically, this situation was expected because it
was assumed that these individuals would have accepted and internalized the
dominant group’s stigmatizing beliefS about them. This viewpoint was consistent
THE EXPERIENCE OF DISCRIMINATION 447
with important theories of the time, such as the concept of the “looking glass
self” proposed by Charles Horton Cooley (1902). According to this and similar
theories, our self-images are, in part, formed by imagining how we look to others
and how others judge us. When those imaginings are negative, self-worth suffers.
As we have seen throughout this book, stigmatized group members have a multitude
of experiences, both historical and personal, that suggest such negative
evaluations exist. But do these experiences affect their self-esteem?
It turns out that this question cannot be answered with a simple yes or no.
We focus here on the results of a major review of this literature, conducted by
Jean Twenge and Jennifer Crocker (2002), who examined data from 712 studies
that encompassed over 375,000 participants. Results showed that important differences
in self-esteem emerged across ethnic groups. One major finding is that
Blacks have higher self-esteem than any other ethnic group studied, including
Whites. This pattern of results is particularly interesting because studies of
Americans’ general attitudes toward social groups find Blacks to be more devalued
than Whites, Asians, or Hispanics (Wilson, 1996a). Clearly, Black’s self-concepts
are not simply reflecting societal attitudes toward their group. Interestingly, the
evidence also suggests that Blacks’ higher self-esteem is linked to their racial identity.
For example, by looking at studies across time, Twenge and Crocker showed
that there were larger differences between Blacks’ and Whites’ self-esteem around
and after the time of the civil rights and Black power movements in the United
States (see also Gray-Little & Hafdahl, 2000), presumably because such actions
raised awareness of and confidence in their group identity. Moreover, self-esteem
was higher aruong college-age Blacks, who may be leaming more about their
culture and heritage (Twenge & Crocker, 2002).
However, the different patterns observed across ethnic groups suggest that a
single theory is unlikely to account for the data. In contrast to Blacks, Asians and
Hispanics had lower self-esteem than Whites, so not all ethnic groups form a
positive social identity relative to the dominant group. In addition, Twenge
and Crocker (2002) found little evidence that racial identity improved selfesteem
for Asians and Hispanics. Instead, the data for these groups appear to be
most consistent with the idea that there are cultural differences in how people
evaluate themselves. For example, cultures differ in their endorsement of individualism
or collectivism. Individualism refers to the idea that people are independent
of one another and that individuals should focus on their personal goals, personal
uniqueness, and personal controls. Collectivism refers to the idea that group members
are bound together and are obligated to one another (Oyserman, Coon, &
Kemmelmeier, 2002). People with an individualist perspective tend to see the
self as stable and transcending relationships and situations. That is, they emphasize
the individual over the group. In contrast, people with a collectivist perspective
tend to believe the self is flexible and varies with the situational context and,
accordingly, de-emphasize the importance of the self relative to the group.
Maintaining and enhancing self-esteem is associated with individualism; it is
acceptable in an individualist culture to stand out from and be superior to others.
Collectivist cultures, in contrast, emphasize self-criticism both because it is
believed this leads to self-improvement and because it promotes harmony with
448 CHAPTER 11
Statistically, Asian Americans are an underrepresented
group in the United States. Yet when people think
about minorities, particularly those who are
stigmatized, this group does not readily come to mind.
You may have noticed, for example, that Asian
Americans are rarely a subject of social psychological
research on prejudice and discrimination, especially
when compared to Blacks and women. One reason
Asian Americans are often overlooked may stem from
the perception that they are seen as the “model
minority.” That is, as a group, Asian Americans are
well-integrated into the culture of the United States
and the characteristics associated with them-high
achievement and economic success-are the same
characteristics associated with Americans in general
(Lee, 1996). As Daphna Oyserman and Izumi Sakamoto
(1997) point out, however, the blurred boundaries
between “Asian” and “American” is a mixed bag. It is
a good thing to be seen as a model, but viewing Asian
Americans in this light is also a way to marginalize
the group.
Oyserman and Sakamoto (1997) studied Asian
Americans’ perceptions of the stereotypes held about
their group and their reaction to those stereotypes,
Results showed that some respondents believed that
non-Asians perceive them as high achieving and highly
motivated-in short, a model minority. Those who
made this observation also believed this to be a
positive perception that held a kernel of truth, Other
respondents, however, viewed the “model minority”
label negatively and believed it overlooked their
personal role in their success. That is, they thought
their success was being attributed to their group
membership, rather than their own abilities and
efforts. They also believed that the label kept them out
of the societal mainstream. Oyserman and Sakamoto
(1997) also found that Asian Americans believe
non-Asians hold negative stereotypes about their
physical appearance and mannerisms, stereotyping them
as short, nearsighted, and having poor English-speaking
ability. Asian Americans also believed others perceived
them as exclusionist, keeping with their own race and
holding condescending views about other races,
These perceptions also are not unfounded.
Monica Lin and her colleagues (Lin, Kwan, Cheung, &
Fiske, 2005) showed that Whites believe Asian
others. Because the commonly used self-esteem measures contain items consistent
with the individualist perspective, self-esteem should be higher in groups
that COlne from those cultures rather than from collectivist cultures.
Twenge and Crocker’s (2002) results are consistent with this perspective, The
self-esteem of individuals from collectivist cultures, such as Asian Americans,
Hispanics, and Native Americans was lower than the self-esteem of individuals
from individualist cultures, such as Whites and Blacks in the United States, It might
surprise you to learn that Asian Americans experience lower self-esteem than do
Blacks or Whites, Many people view this group as a so-called model minority
and, as such, expect them to be unaffected by prejudice and discrimination. See
Box 11,4 for more about this stereotypic perception.
The question of whether and why stereotyping and prejudice affect seIfesteem
is far from settled. We saw in our discussion of attributional ambiguity
that attributing discrimination to prejudice can actually buffer self-esteem.
However, this buffering effect is far from universal, particularly when viewed
outside of the context of a particular instance of discrimination. Women who
perceive high levels of discrimination, for example, also report higher levels of
depression than women who perceive less discrimination (Kobrynowicz &
Branscombe, 1997). In addition, asking women to think specifically about the
ways they have been discriminated against leads to reductions in self-esteem
Americans have poor social skills. Indeed, a variety of
negative stereotypes about Asians exist in concert
with the “model minority” label. Whites believe
Asians are ambitious, hardworking, and intelligent,
but they also see them as pushy, selfish, deceitful,
and nerdy (Ho & Jackson, 2001). Moreover, negative
attitudes and emotions are particularly strong when
White people feel threatened by Asians Americans’
academic success (Maddux, Galinsky, Cuddy, &
Polinfroni, 2008).
The belief that Asian Americans are highly
competent workers appears to coexist with the belief
that they are unsociable. Research suggests this latter
belief is used to justify discrimination against Asian
Americans (Un et aL, 2005). That is, Asian Americans are
characterized as working too hard and unfairly
succeeding at the cost of positive social relations. The
“model minority,” then, pays a price for being
perceived as competent. This price is evident in Asian
American’s reports about their experiences. Oyserman
and Sakamoto (1997), for example, found that about
half of their Asian American sample reported having a
set of coping strategies to deal with these negative
perceptions. Many of these experiences are similar to
those of other stigmatized group members, including
the experience of being singled out, being stared at, not
THE EXPERIENCE OF DISCRIMINATION 449
having their groups’ voice represented in the media, or,
relatedly, having people make assumptions about their
perspective based solely on their group membership.
Moreover, model minority status does not
appear to ameliorate workplace discrimination; Asian
Americans report levels of workplace discrimination
that are similar to Hispanics, and significantly less
than Whites (although Blacks report higher levels of
such discrimination; Bell, Harrison, & McLaughlin,
1997). Another downside to model minority status is
that needed help is sometimes not offered. Asian
Americans who are poor at math (and so violate the
stereotype that all Asian Americans are mathematically
talented), for example, might not receive
mentoring or other help (Goto, 1999). Research
shows that successful Asians Americans are less likely
to have a mentor than are successful managers from
other minority groups and they report being less
satisfied with the mentoring experiences they do
have (Thomas, 1991). It may be that the “model
minority” stereotype is preventing Asian Americans
from getting effective mentoring, perhaps because
of the very success that led to this label (Goto, 1999).
Many Asian Americans have succeeded in spite of
this, but that does not mean it has not hindered
others’ progress.
(Branscombe, 1998). It seems clear that there is no one answer to the question of
whether one’s group membership, or experiencing prejudice and discrimination
related to that membership, affects self-esteem. The outcomes depend on when
and how the question is asked, the cultural context within which a stigmatized
group lives, and whether the individuals themselves readily perceive discrimination
in their lives. Research questions related to this issue will no doubt continue
to attract the attention of social science researchers.
COPING WITH DISCRIMINATION
As we have seen throughout this chapter, living with prejudice and discrimination
creates a threatening situation that can be difficult to deal with and individuals
who are in this situation use a variety of coping strategies to do so. Some of
these strategies were discussed in the section on stress-related responses to prejudice
and discrimination. We next consider two additional coping strategies that
have been studied by social justice researchers: psychological disengagement and
behavioral compensation.
450 CHAPTER 11
Psychological Disengagement and Disidentification
One coping strategy employed by stigmatized group members is psychological
disengagement, “a defensive detachment of self-esteem from outcomes in a
particular domain, such that feelings of self-worth are not dependent on successes or
failures in that domain” (M’ior, Spencer, Schmader, Wolfe, & Crocker, 1998, p. 35).
That is, when individuals disengage, they produce a psychological separation from
themselves and the arena in which they might fail, thereby protecting their
self-esteem. A person who fears poor perfOlTI1al1Ce in an achievelnent-related area,
for exanlple, might psychologically prepare for failure by deemphasizing the impor-
, tanee of success in that area. One way to manage this is by disidentification or deli
valuing the domain (Schmader, Major, & Gramzow, 2001; C. Steele, 1997). When
l people disidentifY with a domain, they define or redefine their self-concept so that the
domain is no longer an area of self-identification. Women who believe they might
‘be unsuccessful at math, then, might disidenti±y with a career in luathematics, and
instead associate their self-worth with literature. For these waHlen, then, failure at
math would not affect their self-esteem. A second way to separate stereotypic
expectations of failure from individual self-esteem is by discounting the feedback
as inaccurate or invalid, particularly because its source was a prejudiced other
(Crocker et aI., 1991). We discussed this process under the section of this chapter
on attributional ambiguity.
It is important to note that psychological disengagement refers to the distancing
of the self from areas in which one’s group is negatively stereotyped
and, therefore, expected to fail. In our ex~unple, a woman’s belief that she m.ight
fail at math is tied to the cultural stereotype that women, in general, are not
mathematically inclined. Tills process also can operate on an individual levelthat
is, for reasons not linked to group stereotypes. Sabotaging one’s performance
out of fear of failure, called self-handicapping, can have similar self-protective
outcomes (Berglas & Jones, 1978). Our focus, here, however, is on disengagement
that is related to group membership, Theoretically, the disengagement
emerges in response to systemic injustice, or the belief that discrimination has
produced differences between social groups that cannot be overcome at a personal
level, no matter how motivated or competent an individual member of
that gmup is (Schmader et al., 2001).
Research suggests that Black smdents are particularly likely to disengage their
self-esteem from their performance in intellectual or academic domains, especially
when compared to Wh.ites. A study of a large, nationally representative sample of
middle and high school-aged children found that Blacks’ achievement was lower
than Whites’ in three of the four content areas studied (Osbourne, 1995). Despite
this, Blacks reported higher self-esteem than did Whites. Moreover, the relationship
between grade point average and self-esteem weakened for Blacks as the children
reached more advanced grade levels, suggesting that disengagement is more likely
after luore academic feedback and experience. A similar pattern emerged for Black
males, but not Black females, when the relationship between academic acillevement
and self-esteem was examined, No such weakening occurred for White students.
Moreover, based on both empirical data and a recent review of this literature,
THE EXPERIENCE OF DISCRIMINATION 451
Toni Schmader and her colleagues (2001) concluded that this disengagement of the
self from academic domains occurs not because ethnic minorities devalue education
but rather because they discount the academic feedback they receive from White
evaluators.
Much of the evidence on ethnic differences in engagement is based on correlational
data that cannot definitely confinn the cause. To test the role of engagement
experimentally, Brenda Major and her colleagues (1998, Study One)
had Blacks and Whites take either an easy or difficult test, so that participants
experienced success or failure in an academic domain. If Blacks are disengaging
from academic perfonnance in general, their self-esteem should be less affected
by negative feedback in such situations than should the self-esteem of Whites.
Results supported this prediction; White’s perfonnance-related self-esteem was
lower when they took the difficult test and experienced failure than when they
took the easy test and succeeded. In contrast, Blacks’ perfonnance-related selfesteem
was not affected by which test they took.
Major and her colleagues (1998) conducted a follow-up study that examined
whether test failure would be more likely to affect individuals who were chronically
disengaged intellectually, compared with those who chronically engaged on
this factor. Level of intellectual engagement was assessed in advance on a measure
designed for that purpose. Procedures were otherwise similar to Study One,
except that all participants in Study Two believed they performed poorly on
the test. Results showed that Blacks who were chronically disengaged with intellectual
tests tended to have higher sel£-esteem following failure than Blacks
who were not so disengaged. Whites’ self-esteem was unrelated to their level
of intellectual engagement. Taken together, research on psychological disengagement
suggests that Blacks can protect their self-esteem by disengaging themselves
from academic or achievement-related domains. Yet, doing so has costs; disconnecting
from academic achievement may result in poor perfonnance in school,
which is linked to higher dropout rates, lower college acceptance rates, and the
receipt of fewer scholarships to support higher education. It is also linked to
fewer opportunities for job success (C. Steele, 1997).
Behavioral Compensation
As we have seen throughout this chapter, for members of stigmatized groups the
experience of prejudice and discrimination is not a one-time or unusual event.
Because of this, individuals develop strategies that help them cope with the
experience. A strategy labeled behavioral compensation has recendy receiveD
attention and concerns how people behave when they expect to be discriminated
against. Carol Miller and her colleagues (Kaiser & Miller, 2001b; Miller & Myers,
1998) have proposed that, in such situations, people sometimes compensate for
potential discrimination by changing their behavior in ways that disconfirm the
stereotype.
According to this perspective, individuals develop a set of skills to help them
achieve desired outcomes. In the case of potential discrimination, these skills are
above and beyond what is needed to succeed in a typical social interaction. This is
452 CHAPTER 11
because when prejudice is a possibility, the individual must overcome an added
burden to be successful in an interaction. Heavyweight people, for example,
know that they may face discrimination because of their weight; this discrinllnation
can take the fann of overhearing unflattering comments about their size, being
avoided or excluded, job discrimination, and even physical violence (Miller &
Myers, 1998). To compensate for possible discrimination, heavyweight individuals
might use humor in a social interaction to increase the chances that they will be
liked. According to Miller and Myers (1998), when prejudice is particularly severe,
higher levels of compensation are required to overcome it. Increased prejudice also
reduces the chances that the compensation will be successful.
Experimental evidence suggests stigmatized people do compensate for
potential discrimination. In one relevant experiment, Cheryl Kaiser and Carol
Miller (2001b) asked women to complete a test of their future career success.
This test required then”l to write an essay about what their lives would be like
in 10 years. The women also were told, either before or after completing the
essay, that the panel of n”len who would be evaluating their results was composed
entirely of prejudiced men or that either 50 percent or none of the panelists
were prejudiced men. Independent evaluators rated the essays on the extent to
which they confonned to gender stereotypes and gave their overall in”lpression of
the essays. Results showed the content of the essays varied depending on who
the participants thought would evaluate them: The essays of those forewarned
about prejudice included fewer references to stereotypically feminine topics,
such as the importance of family and niceness, compared to essays written by
those who believed none of the panelists were prejudiced or who were informed
of possible discrimination after the fact. The authors believe that these differences
were due to the women’s desire to distance themselves from felnininity as a way
to compensate in advance for the judges’ possible sexism. An interesting additional
finding was that those who wrote the essays that distanced themselves
from femininity created a more negative impression overall; Kaiser and Miller
speculate that this outcome was due to overcompensation-that is, these women
inadvertently created the impression that they were unfeminine and strident.
To test the possibility that behavioral compensation varies by the demands of
the situation, Carol Miller and her colleagues (Miller, Rothblum, Brand, &
Felicio, 1995) studied the impressions overweight and nonnal-weight women
made in a telephone conversation. In SOllle cases, their conversational partner
could see them, in others the partners could not be seen. This manipulation
was designed to increase the demands of the situation for those who could be
seen; presumably, the overweight women felt an extra burden because they
expected discrimination based on their weight. The authors also varied whether
the overweight women believed or did not believe their partner could see them.
In all cases, the interaction was videotaped, so a visual record was created. Of
interest were the partners’ ratings of the overweight women’s social skills after
the conversation. Results showed that overweight WOlllen received more negative
evaluations than nonnal-weight women when their partner could see them
and they were not aware that they could be seen. When they were aware that
they could be seen, overweight women were rated similarly to nonnal-weight
THE EXPERIENCE OF DISCRIMINATION 453
women. That is, they were able to successfully compensate for potential prejudice
by using their social skills more effectively.
People’s ability to compensate for prejudice depends on a number off actors. First,
the demands of the prejudice-related situation must not be so high that the person
cannot overcome them. Second, the person must acquire and effectively use the skills
needed for compensation and there are probably individual differences in the ability
to do so (see Miller & Myers, 1998 for a review). Finally, there may be unintended
consequences to behavioral compensation. People who expect to compensate for the
effects of prejudice may “slack off” in situations where prejudice is not a factor and, in
doing so, fail to use the appropriate level of effort required in that social setting. They
also may overcompensate, as we saw in the Kaiser and Miller (2001b) study, by trying
too hard, talking too much, or coming on too strong. In short, stigmatized individuals
may misjudge the requirements of the social interaction and either do too
much or too litde. Certainly, this is a burden not faced by members of nonstigmatized
groups.
SUMMARY
This chapter reviewed the research on stereotyping and prejudice from the perspective
of stigmatized group members. Five factors distinguish whether a stigma
is benign or harmful: course, concealability, aesthetic qualities, origin, and peril.
Stigmas can be acquired by association: people associate negative characteristics
with dominant group members who socialize with or support stigmatized others.
Tokens are individuals who are a minority in a majority group; these individuals
stand out from the group and often have negative experiences because of it.
Rosabeth Kanter (1977) emphasized the perceptual tendencies of visibility, contrast,
and assimilation that produce these negative experiences. More recent work
describes the chilly climate that can accompany tokenism and its effect on the
individuals who experience it.
Social psychologists have proposed several theories about how people respond
to prejudice and discrimination. Work on attributional ambiguity shows that
members of stigmatized groups know that dominant group members have both
positive and negative reactions to them and that these reactions can lead to both
favorable and unfavorable evaluations. What is important is that, in both cases, the
stigmatized person believes these evaluations may be due to their group m_embership,
not their ability. Ironically, this makes it difficult for members of stigmatized
groups to know which actions and behaviors are sincere and which stem from
prejudice or from a desire to appear unbiased. Attributing negative feedback to
prejudice can provide a buffer to self-esteem, but this buffer may come at a cost.
One potential cost can be seen in research on the personal! group discrirnination
discrepancy (pGDD)-the tendency to believe that one’s group is more likely to
experience discrimination than one is as an individual group member. The PGDD
has been demonstrated in a number of contexts and may be explained by cognitive
factors, such as differences in how infonnation about individuals and groups is
454 CHAPTER 11
processed, or by lTlotivational factors, such as the desire to deny personal discrimination
as a justification for not taking action against it.
Experiencing discrimination has a number of personal consequences for the
target. For exam_pIe, research on stereotype threat suggests that stigmatized group
melubers are aware that they are stereotyped and that, especially in achievement
settings, they fear confirming those stereotypes. This fear itself can then undermine
academic achievement. The general features of stereotype threat include
the importance of situational factors in producing the threat and the generality
of its effects. Stereotype threat also can change the way people process information.
However, researchers have shown that, under certain conditions, stereotype
threat can be reduced. Finally, nonstereotyped group members sometimes experience
stereotype lift, a gain that elnerges from the same situations that produce
stereotype threat for the stereotyped group.
Throughout this chapter, the stressful bypro ducts that result from experiencing
prejudice were emphasized. These byproducts affect physical health, producing
hypertension, headaches, and other ailments, and mental health, such as
depression and coping. Although strategies are available that reduce these effects,
some are unhealthy and many put the burden on the stigmatized group member
rather than the prejudiced actor. One negative outcome of discrimination is low
self-esteem, which appears to affect Asians and Hispanics more than Blacks and
Whites. Differences between individualist and collectivist cultures may explain
these effects, particularly because the fOImer emphasizes the self and the latter
emphasizes the group.
Individuals can cope with discrimination by psychologically disengaging or
putting a psychological separation between themselves and the arena in which
they might fail. This can be accomplished by disidentification, or devaluing the
domain, or by discounting the feedback. Doing so often has the unfortunate
effect of lowering academic achievement. Individuals may also use behavior
con”lpensation as a way to cope with prejudice and discrimination. That is, they
develop a set of skills that allow theln to compensate for potential discrimination
by changing their behavior in ways that disconfirm the stereotype. Doing so has
both an upside and a downside.
This chapter includes a nUlnber of personal stories of the effects of
experiencing prejudice and discrilnination. It is important that dominant group
members listen to those stories and understand the cumulative effect of even
seemingly small incidents of discrimination. It is this cUll”lulative impact that is
often most detrilnental to those who experience discrimination because of their
group membership.
SUGGESTED READINGS
The Target’s Perspective
Oysemlan, D. & Swim, J. K. (Ed.) (2001). Stigma: An insider’s perspective [Special
Issue]. Journal if Sodal Issues, 57(1).
THE EXPERIENCE OF DISCRIMINATION 455
Swim, J. K. & Stangor, C. (Eds.) (1998). Prejudice: The target’s perspective. San Diego:
Academic Press.
Both resources have a number of articles that are relevant to the issues in this
chapter. Both include general discussions of theories and data as well as articles
devoted to specific stigmatized groups, such as the overvveight, women, or
specific racial groups.
Social Stigma
Jones, E. E., Farina, A., Hastorf, A. H., Markus, H., Miller, T., & Scott, R. (1984). Social
stigma: The psychology of marked relationships. New York: Freeman.
Although a great deal of research has addressed stigma since the publication of
this book, it remains one of the best resources on this topic because of its clear
explanations.
Weiner, B., Perry, R. P., & Magnusson, J. (1988). An attributional analysis of reactions
to stigmas. Journal if Persol1ality and Social Psychology, 55, 738-748.
An accessible research article that examines people’s perceptions of and
reactions to stigmatized individuals.
Personal Experiences as Tokens or Members
of Stigmatized Groups
Dews, C. L. B. (Ed.) (1995). This fine place so far from home. Philadelphia: Temple
University Press.
Tokarczyk, M. M. & Fay, E, A. (Eds,) (1993), Working-class women 111 the academy: Laborers
il2 the knowledgeJactory. Amherst, MA: The University of Massachusetts Press.
Both books contain collections of essays from women and men from lower
class andlor ethnic minority backgrounds who are currendy working in
academia. Many essays explore their experiences, including their feelings of
isolation and the ways in which the subtle message that they are “different” is
conveyed. Their experiences will resonate with many students, especially
students of color and first generation college students,
Graham, L. 0, (1995). Member oJtl1e club: Riflectiol1s on life il1 a racially polarized world, New
York: Harper Collins.
This highly readable book contains a series of essays that address racism in the
United States, Graham is a highly successful Harvard-trained lavvyer who writes
about his difficulty in finding acceptance in either the White professional or the
Black community. One essay, for example, describes his undercover job as a
busboy in an all-White Connecticut country club. Others address topics from
interracial marriage to Black men’s dining experiences in top New York
restaurants.
Herek, G, M, & Berrin, K, T. (Eds.) (1992). Hate crimes: Confronting violence agai/1st lesbians
at1d gay men. Newbury Park, CA: Sage,
This excellent volume contains a number of powerful “Survivor Stories” of
victims of anti-gay and lesbian violence, Many were based on testimony at the
1996 anti-gay violence hearing before the Subcommittee on Criminal Justice of
456 CHAPTER 11
the Conun.ittee on the Judiciary, House of Representatives. The stories are
brief and memorable and put a human face on the problem of violence against
gays and lesbians.
Responses to Negative Behaviors
Crocker, J. & Major, B. (2003). The self-protective properties of stigma: Evolution of a
modern classic. Psychological Inquiry, 14,232-237.
Crocker and Major’s work on attributional ambiguity was recognized as a
modem classic; in this article, they discussed the source of their ideas and how
they developed them. They also discuss how their research has influenced other
theorists.
Crosby, F. J. (1984). The denial of personal discrimination. American Behavioral Scientist,
27, 371-386.
Taylor, D. M., Wright, S. C, & Porter, L. E. (1994). Dimensions of perceived
discrimination: The personal/group discrimination discrepancy. In M. P. Zanna &
]. M. Olson (Eds.), The psychology of prejudice: TIle Ontario symposium (Vol 7,
pp. 233-255). Hillsdale, NJ: Lawrence Erlbaum.
Crosby’s paper provides a highly readable discussion of the personal group
discrimination discrepancy. Taylor’s review discusses newer theories about the
causes of the PGDD.
Stereotype Threat
Steele, C. (1997). A threat in the air: How stereotypes shape intellectual identity and
perfonnance. American Psychologist, 52, 613-629.
Steele, C., & Aronson, J. (1995). Stereotype threat and the intellectual test performance
of Mrican Americans. Journal qf PersonalitJ’ an.d Social PSj’c!wlogy, 69, 797-811.
These widely cited articles provide both a theoretical overview of stereotype
threat and a description of research demonstrating the basic effects. The authors
also discuss the applied implications of their findings and offer suggestions for
overcoming stereotype threat. Although many more recent articles exist,
including more current review articles, these remain the standards for those
new to the literature.
Coping with Discrimination
Feagin,]. R., & McKinney, K. D. (2003). TIle many costs qfracism. Lanham, MD:
Rowman & Littlefield.
Reviews the cost of White racism from the perspective of African Americans,
covering physical and mental health costs and family and community costs. The
book has many engaging examples and also focuses on strategies for overcoming
raClsnl.
Kaiser, C. R., & Miller, C. T. (2001). Reacting to impending discrimination:
Compensation for prejudice and attributions to discrimination. Personality and
Social Psychology Bulletin, 27, 1357-1367.
THE EXPERIENCE OF DISCRIMINATION 457
Relatively little research has addressed how people’s behavior changes when
they expect to be discriminated against. This clever study gets at both the
behavior of the person expecting discrimination and independent assessments
about how those behavioral changes might affect the interaction.
Major, B., Spencer, S. J., Schmader, T., Wolfe, C., & Crocker, J. (1998). Coping with
negative stereotypes about intellectual perfonnance: The role of psychological
disengagement. Personality and Social Psychology Bulletin, 24, 34-50.
This groundbreaking paper provides experimental evidence for psychological
disengagement.
attributional ambiguity
behavioral
compensation
disidentification
KEY TERMS
personal/group
discrimination
discrepaucy (PGDD)
psychological
disengagement
stereotype threat
stigmatized
QUESTIONS FOR REVIEW AND DISCUSSION
1. Explain the concept of stigma and describe the five factors that distinguish
between harmful and benign stigmas.
2. Do you think the basis of their stigma matters from the point of view of
marked, or stigmatized, individuals? Explain your answer.
3. Give examples of groups in the modem world who are numerically a
majority but are nonetheless stigmatized.
4. Explain why researchers often consider women to be a stigmatized group.
5. Define stigma by association. How are your own interactions affected by the
possibility of this stigma?
6. Give examples of token groups outside the corporate setting studied by
Kanter (1977). Explain how the concepts of visibility, assimilation, and
contrast relate to these groups.
7. If you were doing research on the effects of tokenism, how would you
detennine whether a particular person’s experiences were unique to that
person or part of an overall pattern of discrimination toward her or his social
group?
8. Explain the concept of attributional ambiguity. Why would a stigmatized
group member experience attributional ambiguity?
9. What reactions might a member of a stigmatized group have to positive
feedback given by a maj ority group member? Think of situations where
each reaction might be lllOre likely.
458 CHAPTER 11
10. What is the pe”onal/group discrimination discrepancy? Outline the available
SUppOlt for cognitive and motivational explanations for the personal/group
discrimination discrepancy. Which explanation do you think is more accurate
and why?
11. Explain the concept of stereotype threat. Outline the keys to understanding
how stereotype threat operates.
12. If you were an elementary school teacher, how would you prepare your
students for standardized tests so that the effects of stereotype threat would
be minimized?
13. Describe the stressors that result from experiencing discrimination. How
could these stressors be minimized?
14, Is the stress associated with experiencing discrimination the same or different
from other types of stress? Explain your reasoning.
15. Who is more responsible for reducing the mental health consequences of
experiencing discrimination, the stigmatized or the majority group member?
Explain your reasoning.
16. Consider the cunent literature addressing how experiencing discrimination
affects self-esteenl. What are the lllost important questions that renuin
unanswered? If you were planning to conduct research on this issue, what
would be your focus’ Why?
17. Explain the concept of psychological disengagement.
18. Distinguish between disidentification and discounting.
19. How might psychological disengagement affect the school performance of
Hispanics in the United States?
20. Explain the concept of behavioral compensation. Explain how members of
stigmatized groups use behavioral compensation in situations where they
might be discriminated against. Does behavioral compensation do more
harm than good? Explain your reasoning.
21. The work cited in this chapter examines social psychological theory, but also
emphasizes individual experience. How should researchers balance these two
perspectives?
22. Consider the quotations that opened this chapter. Do you believe dominant
group members can ever understand what it is like to experience discriminatory
behaviors? Why or why not? Are there factors that will make this
understanding more likely?