From prejudice to discrimination

Order Description

1.Watch this clip:

Are you surprised and/or upset when the last word comes out? Have you in some way influenced by stereotypes?

2. Watch this video clip: Wafa was attacking the extremists and their destructive acts.

And see this picture

In download

Question: Why do you think the terrorists do what they do? Is the reason similar to members of KKK lynching Blacks? You have to show your knowledge to receive full credit.

must use book in download as 1 reference

Chapter 10
*
From Prejudice to
Discrimination
Rush hour on board a bus or train …. The flood of incoming
passengers begins to solidifY …. I am sitting next to a window,
my eyes half-closed …. Dressed conservatively in a tweed
jacket and tastefully bold tie, I am an unremarkable Ulan, .. , as
unnoticed as any other COl1lllluter. Except for Oile thing: amid
the growing crush, the seat beside me remains empty. At stop
after stop, … a succession of seemingly random individual
decisions coalesces into a glaring pattern of unoccupied spaces
next to black males-inc1uding me. Soon the seats beside us are
the only ones left. Other passengers remain standing, leaving
only these seemingly quarantined seats.
-BRUCE JACOBS (1999, PP. 15-16)
Chapter Outline
What Is Discrimination?
Forms of Discrimination
Interpersonal Discrimination
The Relation between Prejudice
and Discrimination
Regressive Prejudice
Reactions to Having Acted in a
Prejudiced Manner
Discrimination in Organizations
Organizational Research
Individuals in Organizations
369
Hate Crimes
Hate Crime Offenders
Motivations for Hate Crimes
Effects on Victims
Summary
Suggested Readings
Key Terms
Questions for Review and
Discussion
370 CHAPTER 10
Prejudice, which has been the main focus up to this point in the book, is an
attitude; it deals with how people think and feel about members of other
groups. Discrimination, in contrast, is behavior; it deals with how people act to- I ward members of other groups. Discrimination consists of behaving differently
l,~oward people based solely or prirnarily on their lllembership in a social group.
‘The tenn is usually used to refer to acting in an unfair or demeaning manner, but it
can also refer to giving someone an undeserved advantage. Bruce Jacobs’s (1999)
experience cited above illustrates one aspect of the ongoing problem of discrimination
in modem fullerican society: Many White people avoid contact with members
of minority groups, even when the avoidant behaviors cause inconvenience for
themselves. Although discrimination against mel11bers of minority groups is not always
as blatant as it was prior to the civil rights l110vement of the 1960s, it still occurs.
As we discuss the research on discriminatory behavior you might be tempted to say,
“Well, it’s obvious: Prejudiced people discriminate,” but, as we will see, there is only
a moderate correlation between people’s degree of prejudice and their propensity to
engage in discriminatory behavior. That is, not all prejudiced people discriminate
every time they have an opportunity, and some nonprejudiced people discriminate
in sallIe situations. In presenting what psychologists know about discrimination,
we first look at the nature of discrimination and then focus on the two forms most
often studied by psychologists, interpersonal discrimination and organizational
discrimination. We conclude with a brief examination of the most severe fonn of
discrinlination, hate crimes.
While reading this chapter, bear in mind a point we made in Chapter 7 concerning
prejudice: some fonns, such as racial prejudice, are socially proscribed
whereas other forms, such as anti-gay prejudice and prejudice against physically
unattractive people, are more socially pernlissible. The same principle applies to
discrimination. For example, David Schneider (2004) noted the following forms
of socially acceptable discrimination: “Most of us would fight having a group
honle for convicted rapists placed next door to our honle, no matter how ‘cleaned
up’ or ‘ex-‘ the rapists claimed to be. Most church groups do not invite homeless
people to share their potluck dinners …. Many women prefer their gynecologists
to be female, and most males discriminate against males as their sexual partners.
That’s the kind of discrimination we all know and generally approve. So it is not
the fact of discrimination that is controversial, but its application to specific groups”
(p. 291). Consider, for example, college students’ ratings of the acceptability of
prejudice and discrimination directed at various groups (Crandall, Eschleman, &
O’Brien, 2002). Table 10.1 shows the groups for which prejudice and discrimination
had the highest and lowest approval ratings.
WHAT IS DISCRIMINATION?
Discrimination can manifest itself in many ways, both verbally and behaviorally,
and in many settings. For example, Jennifer Livengood and Monika Stodolsak
(2004) interviewed American Muslims about the discrimination they experienced
in post-September 11. Common experiences included hearing racist epithets,
FROM PREJUDICE TO DISCRIMINATION 371
TAB L E 10.1 Socially Approved and Disapproved Prejudices
Percent Percent
Approved Prejudices Approving Disapproved Prejudices Approving
Rapists 98 Mentally retarded people 3
Child abusers 98 Native Americans 6
Child molesters 97 Black Americans 6
Wife beaters 97 Jews 6
Terrorists 95 Catholics 6
Racists 92 Whites 7
Members of the Ku Klux Klan 91 Hispanics 7
Drunk drivers 91 Asian Americans 7
Members of the American 90 Canadians 7
Nazi Party
Pregnant women who drink 89 Ugly people 10
alcohol
Men who refuse to pay child 89 Interracial 11
support couples
Negligent parents 86 People with AIDS 11
People who cheat on their 82 Fat people 11
spouses
SOURCE: Crandall, Eshleman, and O’Brien (2002, Table 1, p. 362).
receiving hostile and disapproving looks, being subjected to obscene gestures,
avoidance by others, and, in some cases, being personally threatened, having
property vandalized, or being physically attacked. As one respondent noted,
“They scream at you and … say really mean things … like ‘you crazy terrorist,
you go back home. You rag head [referring to Muslim head coverings] and
you are so backward.'” (p. 192). Respondents reported experiencing ‘hate
stares’ as they went about their everyday activities such as shopping, driving,
or walking down the street. As a result, many changed their routines to reduce
contact with mainstream Americans and chose to remain at home whenever
possible. Many reported being fearful for their personal safety and for the
safety of their friends and loved ones. In this chapter, we being by looking at
the forms discrimination can take: blatant, subtle, or covert. We then focus on
the types of discrimination most often studied by psychologists: interpersonal
and organizational.
Forms of Discrimination
Given the many forms that discrimination can take, it is useful to have a system
for classifYing fonns of discrimination to show how they relate to one another.
Nijole Benokraitis and Joe Feagin (1995) developed one such system, based on
372 CHAPTER ,0
three forms of discrimination-blatant, subtle, and covert. Although Benokraitis
and Feagin developed their typology in the context of sexism, it applies to other
forms of discrinlination as well. In Chapter 1, we discussed levels of discrimination,
including interpersonal, organization, cultural, and institutional. Note that
blatant, subtle, and covert discrimination can be found at all four of those levels
of discrimination.
U
Blatant Discrimination. Blatant discrimination consists of “unequal and
harmful treatment … that is typically intentional, quite visible, and easily documented”
(Benokraitis & Feagin, 1995, p. 39). Extreme cases of blatant discrimination,
-such as the murder of Matthew Shepard, a gay man who was beaten and left to die
on a fence in Wyoming (Laramie Project Archives, 2002), and the murder of James
Byrd, a Black man who was chained to the back of a truck and dragged along a
road in Texas until he died (Texas NAACP, 1999), often receive national attention.
However, blatant discrimination occurs in everyday contexts as well. For example,
a Black college student interviewed by Janet Swim and her colleagues (Swim,
Hyers, Cohen, Fitzgerald, & Bylsma, 2003) told of how “a man at a party addressed
her by a racist label and ordered her to perform a menial task” (p. 52). To cite
another example, it is not unCOlumon for Black women and Inen to experience
blatant discrimination while shopping, including being followed in the store or otherwise
being treated as though they are likely to steal merchandise. One study
found that 35 percent of Blacks experienced negative treatment while shopping in
predominandy White communities, cOlupared to 10 percent reporting similar treatment
in their own community (Lee, 2000).
Some fOlTl1s of blatant discrimination against some groups are illegal and
generally condemned, such as racial discril1unation at work, in school, and in
public accommodations” However, other forms, such as discrimination against
lesbians and gay men, are often legal and are accepted as “normal” by many people.
Consider, for example, a Web site on which visitors can find a picture of
Matthew Shepard burning in hell, along with a record of how many days he
has been there. Visitors are invited to click on his picture and hear him scream
as he endures the flames of hell. However, examples of blatant discrimination
against other groups are not difficult to find, especially if they are supposedly
humorous, At the tUlle of this writing, the popular Web site You- Tube featured
of series of videos (many rated 5-star) depicting the “Anlazing Racist” discriminating
against Mexicans, Asians, and Muslims. Although these clips are categorized
as comedy, the actions depicted are based on negative stereotypes such as
the belief that all Asians eat dogs, that all Muslims have body odor, and that all
Mexicans are in United States illegally.

Subtle Discrimination. Subtle discrimination consists of “unequal and
harmful treatment .” that is typically less visible and obvious than blatant discrimination.
It is often unnoticed because people have internalized subde [discrinlinatory]
behaviors as ‘normal,’ ‘natural,’ or customary” (Benokraitis &
Feagin, 1995, p. 41). Subtle discrimination tends to be harder to document
FROM PREJUDICE TO DISCRIMINATION 373
than blatant discrimination, but it can often be done. Unlike the other fOnTIS
of discrimination, which are often intentional, subtle discrimination is often
unintentional. For example, Claude Steele (1992) related a story told by a friend
of his who
noticed over many visits [to her son’s third-grade classroom] that the extraordinary
art work of a small black boy named Jerome was ignored–or, more accurately
perhaps, its significance [as a sign of artistic talent] was ignored . .&
genuine art talent has a way of doing-even in the third grade-his stood out.
Yet the teacher seemed hardly to notice. Moreover, Jerome’s reputation, as it
was passed along from one grade to the next, included only the slightest mention
of his talent, … Had Jerome had a reading problem, which fits [American
society’s stereotypic image of black children], it might have been accepted as
characteristic of him more readily than his extraordinary art work, which contradicts
[that image] (p. 72).
Thus, subtle prejudice had the effect of directing teachers’ attention away
from Jerome’s artistic talent because, by society’s definition, Black children
do not have that particular talent. Subtle prejudice also can be manifested in
everyday speech, as described in Box 10.1.
Covert Discrimination. Covert discrimination consists of “unequal and
hannful treatment … that is hidden, pUJposeful, and, ofren, maliciously motivated ….
[It is] behavior that consciously attempts to ensure … failure, as in hiring or other
employment situations” (Benokraitis & Feagin, 1995, p. 42). Covert discrimination
tends to be very difficult to document. Examples in the employment context
include tokenism, hiring one or a few members of a group as evidence that an
organization does not discriminate; containment, restricting members of a group
to a limited number of job categories; and sabotage, arranging for members of a
group to fail, such as by assigning them low volume sales territories but setting their
sales quotas at levels similar to those of salespeople with better territories
(Benokraitis & Feagin, 1995).
Covert discrimination is comnlon outside of the workplace as well, including
the housing market. In the United States, housing discrimination is prohibited by
the Fair Housing Act of 1988; however, such unfair practices are difficult to document.
Landlords, for example, may simply tell prospective renters that an available
unit was just rented. Similarly, real estate agents can subdy encourage buyers
to look in certain areas and to avoid others. Adrian Carpusor and William Loges
(2006) studied whether landlords discriminated against prospective renters based
on their ethnicity. Via email, they contacted landlords in Los Angeles County,
California, about the availability of rental property. The emails were supposedly
sent by prospective renters whose names varied by ethnicity: Patrick McDougal
(White), Tyrell Jackson (Mrican American), and Said Al-Ralnnan (Arab). The
applicant with the White-sounding name received more replies (89 percent)
than did the applicants with Black (56 percent) or Arab (66 percent) sounding
names. Moreover, only 11 percent of the replies to the White applicant were
negative, compared to 44 percent for the African American and 34 percent for
374 CHAPTER 10
Some of the most common examples of subtle prejudice
can be found in everyday speech. Examples include:
home at Monticello, Virginia. He noted that the
tour guide always referred to Jefferson in the
active voice but always referred to the work of
Jefferson’s slaves in the passive voice. For
example, while describing a set of interior doors
with a complex operating mechanism that had
required no repair in the 166 years since
Jefferson’s slaves had built and installed them, the
guide said, “Mr. Jefferson deSigned these doors”
(p. 147). In contrast, the guide said, “These doors
were installed originally in 1809” (p. 147, italics in
original), making the enslaved carpenters whose
skilled work had created the remarkable doors
effectively disappear from history.
•
•
•
Hostile humor calls attention to the negative
stereotypes associated with outgroups, such as
low intelligence, selfishness, stinginess, and
alleged bad habits such as drunkenness (Ruscher,
2001). People are often tolerant of outgroup
disparagement that is presented as a joke because
the humor context implies that listeners should
not take the speaker’s remarks seriously (Ford &
Ferguson, 2004). Humor can therefore also
function as a justification for derogatory speech.
How often have you heard someone who has just
been confronted for making a prejudiced remark
reply with “Hey, it was only a joke”?
Patronizing speech implies that outgroups are less
competent than members of the speaker’s
ingroup. Patronizing speech can take two forms
(Ruscher, 2001). Baby talk is often directed at the
older adult, implying that they have regressed
cognitively to the level of infants (see Chapter 13).
Although baby talk is usually well intentionedspeakers
view it as comforting-recipients find it
insulting. Controlling talk is directed at members
of groups the speaker views as having lower social
status and “functions to keep low~status indivi~
duals ‘in their place'” by controlling the direction
of the conversation (Ruscher, 2001, p. 88). For
example, when talking to women, men are more
likely to interrupt and to give commands than
when talking to other men (Caplan, 1994).
Vanishing uses linguistic devices to make out~
groups disappear. A common way of doing so is to
replace the active voice of a verb (for example,
“Bill hit the ball”) with the passive voice (“The ball
was hit”). As Thomas Greenfield (1975) noted, this
rhetorical device makes “the creator or instigator
of action totally disappear from a reader’s [or
listener’s} perception” (p. 146; see also Chapter 3).
For example, Greenfield (1975) recounted his ex~
periences on a guided tour of Thomas Jefferson’s
• Abnormalization involves describing outgroup
members in ways that emphasize their lack of
compliance with ingroup norms. For example,
Maykel Verkuyten (2001) conducted focus groups
in which Dutch residents of Rotterdam discussed
life in their neighborhoods. The participants were
chosen because they lived in areas in which large
numbers of immigrants also lived. She found that
when immigrants were mentioned, participants
almost always described them in ways that
emphasized their differences from the Dutch norm
often using extreme examples to illustrate a point:
For example, one participant supported his view
that immigrants from India were ignorant and
crazy by citing a case of a family that had built a
cooking fire on the floor of their apartment.
linguistic devices such as these serve several functions.
Hostile humor and abnormalization indicate ways in
which the ingroup is superior to the outgroup and
patronizing speech reinforces the higher status and
power of the ingroup (Ruscher, 2001). Vanishing
denigrates the skills of outgroup members and denies
their contributions to society, implying that progress
comes only through the efforts of the ingroup
(Greenfield, 1975). All forms of prejudiced speech serve
to draw clear boundaries that separate the “good”
ingroup from the “bad” outgroup.
the Arab. As Carpusor and Loges note, “long before an Mrican American man
gets a chance to show what he is capable of, discrimination tilts the scales
against him in ways he may not even be able to observe. Arab Americans face
similar obstacles” (p. 948).
FROM PREJUDICE TO DISCRIMINATION 375
INTERPERSONAL DISCRIMINATION
Interpersonal discrimination refers to individual, person-to-person diSCrini.U:’~–ination:
one person treating another unfairly because of the person’s group
membership. This section addresses three aspects of interpersonal discrimination.
The first deals with the relationship between prejudice and discrimination
and the circumstances under which prejudice is more or less likely to predict
discriminatory behavior. In Chapter 5, we discussed the motivation to control
prejudice. Here, we discuss factors that can undermine that motivation and allow
discriminatory behavior to occur. The third section examines how people
react to having acted in a prejudiced manner.
The Relation between Prejudice and Discrimination
& we noted earlier, when asked what causes interpersonal discrimination, most
people would probably reply “prejudice.” If that were the case. one would expect
to find a strong correlation between people’s prejudiced attitudes and their
propensity to engage in discriminatory behavior. However, when John Dovidio
and his colleagues (Dovidio. Brigham, Johnson, & Gaertner, 1996) reviewed the
results of 23 studies of the prejudice-discrimination relationship, they found an
average of correlation of r ;:::: .32. A meta-analysis of 60 studies found a similar
correlation of r ;:::: .36, but also found a stronger relationship between prejudice
and the intention to discrintinate, r = .45 (Schutz & Six, 1996). That is, the relationship
between prejudice and what people say they will do is stronger than the
relationship between prejudice and what they actually do. Even when researchers
find fairly high correlations between prejudice and discrimination, about
30 percent of the participants in those studies exhibited behavior that was inconsistent
with their attitudes (Duckitt, 1994). Although results such as these might
seem discouraging, they are, in fact, consistent with the results of research on the
relationship between attitudes and behavior in general. For example, Stephen
Kraus (1995) found an average correlation of r = .38 between attitudes and behavior
across a wide variety of domains. As William Graziano and has colleagues
note (Graziano, Bruce, Sheese, & Tobin, 2007), it is one thing to express negative
attitudes toward a social group, but another thing to actually discriminate
against someone. Hence, “the lake of prejudice is wider and deeper than the
smaller pool of discrimination, even if the forruer feeds the latter” (p. 572). As
it turns out, prejudice is related to discrimination (see Eagly & Chaiken, 1998),
but the relationship is not a silTIple one: A number of factors influence the
strength of this attitude-behavior relationship including personal stereotypes,
attitude-behavior correspondence, and perceived social support for the attitude.
Personal Stereotypes. Recall from Chapter 3 that stereotypes can exist at two
levels. Social stereotypes are characteristics of groups that most people in a society
agree on and personal stereotypes are individuals’ beliefs about group characteristics.
Personal stereotypes usually overlap with social stereotypes, but some
of their content may be different. Similarly, people recognize the social norms
376 CHAPTER 10
about prejudice toward particular groups, but may not personally agree with
those nonns. Replicating the results shown in Table 10.1, William Graziano
and colleagues (2007) found that people recognized that there are social norms
which detelnllne whether prejudice toward particular groups is acceptable or unacceptable.
However, they also reported that people personally found it even
more acceptable to discriminate against groups such as rapists and terroriststhose
groups for which society approves prejudice.
Prejudiced people also are lllore likely to discriminate against those outgroup
lllembers who fIt their personal stereotype than against those who do not. In a
study of this process, Shawna Ramsey and her colleagues (Ramsey, Lord,
Wallace, & Pugh, 1994) assessed college students’ personal stereotypes of former
mental patients. Several weeks later, the students read about a fonner mental
patient who either closely matched their personal stereotype or who was very
different from their stereotype hut matched other students’ stereotype. For
example, some students stereotyped fonner luental patients in tenns of schizophrenic
symptonls whereas others stereotyped them in tenus of depressive symptoms.
The students then chose from a list of activities those they would be
willing to engage in with the fonner luental patient, such as showing the person
the university library or taking the person to a party. For students who read
about a person who fit their personal stereotype, Ramsey and her colleagues
found a correlation of r = .43 between attitudes toward fonner mental patients
and the nunlber of activities in which they were willing to engage. For students
who read about a person who did not fit their personal stereotype, the correlation
was only r = -.07.
[
Attitude-behavior Correspondence. The tenn attitude~behavior correspondence
refers to how well an attitude matches, as it were, the behavior it is supposed to
be associated with. A higher degree of correspondence results in a higher
attitude-behavior correlation. One type of correspondence that is important
to the prejudice-discrimination relationship is the degree to which people can
control either their responses on the attitude measure or the behavior being
measured. Dovidio and his colleagues (1996) noted that explicit paperand-
pencil measures of prejudice assess controllable responses: People can think
about how they want to respond and carefully choose their responses, so social
desirability response bias can affect their answers; however, implicit attitude
measures assess relatively uncontrollable attitudes, so there is less opportunity
for social desirability response bias to affect responses (see Chapter 2).
Similarly, some behaviors, such as the content of what a person says, are controllable
and so can be affected by a social desirability response bias, whereas
other behaviors, such as many nonverbal behaviors, are more automatic and
difficult to control and so are less likely to be influenced by a social desirability
response bias. Drawing on the correspondence principle, Dovidio and his colleagues
proposed that scores on controllable measures of prejudice should be
correlated with controllable behaviors but not with automatic behaviors and
that scores on liuplicit measures of prejudice should be correlated with automatic
behaviors but not with controllable behaviors.
FROM PREJUDICE TO DISCRIMINATION 377
John Dovidio, Kerry Kawakami, and Samuel Gaertner (2002) tested this hypothesis
in a study in which White college students completed explicit and implicit
measures of prejudice. The participants then discus,.<;ed several race-neutral
topics, such as what personal belongings were most useful to bring to college,
with a Black student confederate who played the role of another research participant.
The interactions were videotaped, and raters later coded the White students’
behaviors for the frienclliness of their (automatic) nonverbal behaviors
and for the friendliness of what they said (controllable behavior). The researchers
found that, as they had expected, implicit prejudice correlated with nonverbal
friendliness but not with verbal friendliness and explicit prejudice correlated
with verbal friendliness but not with nonverbal friendliness. Similarly, Denise
Sekaquaptewa and her colleagnes (Sekaquaptewa, Espinoza, Thompson, Vargas, &
von Hippe!, 2003) found that an implicit measure of prejudice was related to
White students’ tendency to ask a Black student stereotypic rather than nonstereotypic
questions from a list provided by the researchers. Because students were
unaware that the questions varied in how stereo typic they were, this was considered
an implicit behavioral measure.
Attitude-behavior correspondence also can be assessed by the degree to
which people’s general attitudes toward a social group nlatch behaviors toward
a specific group member. For example, Michael Olson and Russell Fazio (2007)
studied whether White students’ implicit attitudes toward Blacks in general corresponded
to their attitudes toward a specific Black person. They reasoned that
when these attitudes were similar-that is, if a White student had positive attitudes
toward Blacks, in general, and also liked a specific Black person-the
speaker’s nonverbal responses should indicate little discomfort. However, if this
same White student disliked the specific Black person, it would be evident in
nonverbal behaviors, such as self-touching and speech interruptions. Results supported
their hypothesis; regardless of whether implicit attitudes toward Blacks
were positive or negative, when reactions to a specific Black person differed
from those implicit attitudes, nonverbal “leakage” occurred. Hence, even nonprejudiced
people can appear prejudiced if they dislike a specific person.
Perceived Social Support. The term perceived social support refers to the extentl
to which people believe that others share their attitudes and opinions. Generally,J
attitudes for which people perceive more social support are more closely related
to their behavior than attitudes for which they perceive less social support. For
example, Gretchen Sechrist and Charles Stangor (2001) pretested White college
students’ level of racial prejudice and selected those with high or low scores for
participation in a study conducted several weeks later. fu part of the study, participants
learned that either 81 percent of the students at their university agreed
with their racial attitudes (high social support condition) or that 19 percent agreed
with them O-ow social support condition). The researchers then used what is
known as the waiting room ploy to assess discrimination: The research apparatus
“malfunctioned” and the experimenter asked the participant to wait in the hallway,
where seven chairs were lined up in a row, with a female African American
student seated in the chair next to the door to the laboratory. Discrimination was
378 CHAPTER 10
assessed by how lllany chairs away from the Black student the participant sat.
Not surprisingly, the students who had scored low on prejudice sat closer to
the Black student than those who had scored high, an average of 2 seats versus
3.9 seats away. In addition, perceived social support affected the behavior of prejudiced
participants, with those who thought that most of their fellow students
also were prejudiced sitting farther from the Black woman than those who
thought that most of their peers were unprejudiced, an average of 4.3 seats versus
3.4 seats. Perceived social support had no effect on seating distance of the students
low on prejudice, Moreover, the prejudice-discrimination correlation was
larger for the students high in prejudice, r = .76, than for the students low in
prejudice, r = .33.
Regressive Prejudice
As we discussed in Chapter 5, under some conditions people are motivated to
avoid appearing prejudiced. However, doing so requires a great deal of mental
work: One must recognize that prejudice might be affecting one’s behavior and
then consciously change that behavior to produce a nonprejudiced response
(Crandall & Eshleman, 2003). Although, this process might become automatic
for some people (see Devine, 2002; Plant, 1998), others have to work at it.
The laborious nature of this process is demonstrated by research showing that
people experience a sense of relief when they are allowed to express prejudices
they have been controlling (Crandall & Eshleman, 2003). Thus, it is not surptising
that even people who score low on measures of prejudice report sometimes
_making prejudiced responses toward members of other groups (Monteith &
[
Vails, 1998; Vails, Ashbum-Nardo, & Monteith, 2002). That is, even people
who see themselves as unprejudiced and who score low on measures of prejudice
sometimes find thenlselves acting in prejudiced ways. Box 10.2 provides an example
of how an everyday event can trigger prejudiced reactions even in unprej-
Cldiced people. Such expressions of prejudice by people who are otherwise low
in prejudice are labeled regressive racism or regressive prejudice (Rogers &
Prentice-Dunn, 1981). Regressive prejudice occurs because controlling prejudiced
responses requires attention and mental resources; when those resources
are not available or when the control process is mort-circuited people regress
from controlling prejudice to expressing it. This section examines some of the
factors that can produce regressive prejudice.
Control over Behavior. To avoid responding in a certain way, people must be
able to control the behavior. As we noted earlier, some behaviors, such as nonverbal
responses, are less under voluntary control than others, such as the content of
what one says. Consequendy, people may give off nonverbal cues that imply dislike
of or discomfort with a nlember of a stereotyped group even while trying to
behave in a positive manner (see, for example, Dovidio et al., 2002). In one study
of this process, Dovidio and his colleagues (2002) tested White reseatch participants
for implicit prejudice against Mrican Americans. Each participant then held a
3-minute conversation with a Black student who was working for the researchers;
Most research on stereotype arousal uses laboratory
methods, so you might wonder how well those
procedures generalize to everyday life. Consider the
following commercial that was televised in 1995.
The commercial opens with a picture of a poker~
faced, vaguely threatening and ominous Black man
with his head shaved that takes up the left two~thirds
of the screen. After a moment, the following lines appear
one at a time down the right side of the screen,
with a pause between each:
Michael Conrad.
Male. Age 28.
Trafficking.
Armed Robbery.
Assault and Battery,
Extortion,
FROM PREJUDICE TO DISCRIMINATION 379
Rape.
Murder.
Apprehended
January 1994 by
Joseph Cruthers,
Shown here. (Urban Alliance on Race Relations,
1995)
By the time “Apprehended” appears, most people
assume the picture is that of the criminal, Michael
Conrad. When they learn that the picture is that of a
policeman, they feel both surprised and upset that
they have operated on the social stereotype that Black
men are dangerous and made the racist assumption
that the pictured Black man must be a criminal (Fazio &
Hilden, 2001).
because the conversation was about a race-neutral topic, the White participants
were presumably motivated to act in a nonprejudiced manner. However, conlpared
to the nonverbal behavior of participants with lower prejudice scores, the
behavior of those with higher scores was rated as less friendly by both the participants’
Black conversational partners and White raters who saw videos of the participants’
side of the conversation. Thus, when people hold stereotyped views of
other groups, those views may “leak out” through their nonverbal behavior even
while they are trying to control the effects of those stereotypes on other behaviors
(see also Sekaquaptewa et al., 2003).
Cognitive Demands. As we saw in Chapter 4, cognitive demands, such as
trying to remember an 8-digit number or working under time pressure, make
it difficult for people to control the effects of stereotypes and prevent stereotypes
from affecting their judgments. The same processes affect the expression of prejudice:
Even for people who are motivated to control their prejudices, cognitive
demands can reduce their mental resources and so reduce control. To demonstrate
this effect, Daniel Wegner and his colleagues (reported in Wegner, 1994)
had research participants complete sentences about women such as “Women
who go with a lot of men are … ” (Wegner, 1994, p. 47). Half the participants
were instructed not to be sexist in their responses and half received no special instructions;
in addition, half the participants in each of these groups were put under
time pressure to respond and half were not. Wegner and his colleagues found that
participants who had been instructed to avoid sexist responses (and therefore were
nlotivated to do so) and were not under time pressure provided fewer sexist responses
(such as “sluts” to the sentence stem just given) than participants who
380 CHAPTER 10
received no such instructions. However, time pressure undermined the ability to
control prejudice of the participants who were motivated to be nonsexist: They
made sexist responses at the same rate as the participants who were not motivated
to control their sexism.
Disinhibitors. Because motivation to control prejudice derives from social
norms to avoid being prejudiced, factors that reduce people’s motivation to
cOlllply with social nonns can also reduce motivation to control prejudice and
disinhibit its expression. Two factors that motivate nonn compliance are the desire
to avoid the social punislunents that other people would mete out for violating
norms and the desire to receive the rewards that others provide for nonn
compliance. As a result, when people are anonymous and cannot be identified,
they are less motivated to cOlnply with nonns than when they are identifiable
(see, for example, Mym, 2008). Edward and Marcia Donnerstein (1976) tested
the effects of anonymity on the release of prejudice in a study in which White
research palticipants gave what they thought were electric shocks to a Black or
White confederate for making errors on a learning task. Aggressive discrimination
was accessed in two ways. Overt, direct aggression was measured by the
setting on a dial that indicated shock level. Covert, indirect aggression was measured
by shock duration-that is, how long participants pressed the button that
supposedly delivered the shock. Half the participants thought that their behavior
was being monitored and half thought they were anonymous.
Results showed that anonymous participants showed lllore direct aggression
toward the Black than the Wbite person, indicating that anonymity facilitated
discriminatory behavior. Moreover, in contrast to their anonymous peers, participants
who were identifiable favored the Black person by showing less direct
aggression toward him than toward the White person, perhaps reflecting a motivation
not to appear prejudiced. However, nonanonymous participants showed a
higher level of indirect aggression toward the Black person than the White person,
perhaps because covert expression of prejudice is seen as more socially
acceptable.
Strong emotions also can lead people to ignore social norms. In another
study of interracial aggression, angered and cahn White research participants
had the opportunity to administer electric shocks to a Black or White person
(Rogers & Prentice-Dunn, 1981). Results paralleled those Donnerstein and
Donnerstein (1976) found for anonymity: Angry participants gave stronger
shocks to the Black person than to the White person, whereas calm participants
gave stronger shocks to the White person. Alcohol consumption is another notorious
disinhibitor of compliance with social nonns. Not sulprisingly, then,
Laurie O’Brien and her colleagues (as reported in Crandall & Eshleman, 2003)
found a correlation of .31 between alcohol intoxication (as measured by
Breathalyzer tests of people leaving bars) and willingness to express prejudice
against racial and religious groups.
The expression of prejudice is a forbidden act, but the implied approval of
an authority figure can release normative restraints. For example, Donnerstein
and Donnerstein (1976) reported a study in which White participants observed
FROM PREJUDICE TO DISCRIMINATION 381
a peer delivering an intense electric shock to a Black person. Some of the
participants then saw the experimenter overseeing the study severely reprimand
the person for delivering such a strong shock; other participants saw the experimenter
ignore the shock level, thus implicitly giving approval to it. Participants
in a control condition saw a shock being administered but did not know its
intensity. All the participants then took part in the experiment. Compared to
the control participants, those who saw a peer reprimanded for administering a
strong shock gave weaker shocks, but those who saw no reprimand gave much
more severe shocks.
The perceived costs and rewards in a situation also can influence whether
people feel comfortable ignoring social norms, such as the nonn to help others
in need. When a motorist is stranded on a highway, for example, potential
helpers must decide whether the costs of helping, such as effort, time, or
potential risk, outweigh the rewards of helping, such feeling good about relieving
another person’s suffering. Donald Saucier, Carol Miller, and Nicole
Doucet (2005) reasoned that, in such situations, discrimination against Blacks
would be more likely if people could justify the failure to help by detennining
that the costs were too high. These researchers conducted a meta-analysis of
studies assessing whether Black and White people received help in a variety of
situations. Results showed no evidence of universal discrimination against
Black people who needed help. However, in situations where the failure to
help could be easily justified, such as when helping would take a longer time,
involved more risk, or was otherwise difficult, Black people received less help
than did White people.
Finally, other people’s behavior can disinhibit prejudice. Linda Simon and
Jeff Greenberg (1996) created a situation in which four White research participants
and a Black confederate worked in separate cubicles on a creativity task.
The participants had been earlier detennined to have positive, negative, or
ambivalent attitudes toward Mrican Americans. The participants labeled their
answers with a code letter they wrote on a yellow sticky note attached to the
answer sheet. They then evaluated what they thought were the other participants’
responses (actually standard responses created by the researchers) that the
experimenter passed around one at a time. Some participants received a list with
no comment added to the sticky note, others received a list with a handwritten
addition to the sticky note that read “1 can’t believe they stuck us with this Black
person! (please erase this)” (the ethnic criticism condition), and a third group of
participants received a list with a note that read “I can’t believe they stuck us
with this nigger! (please erase this)” (the derogatory ethnic label condition). k
shown in Figure 10.1, participants with positive attitudes toward African
Americans were unaffected by the comment nlanipulation, giving the Black
participant’s contribution a rating of about six in all conditions. Consistent with
the theory of ambivalent prejudice, racially ambivalent participants inflated their
ratings in the derogatory label condition, presumably to emphasize that they
were not prejudiced. However, anti-Black participants who had been primed
by the derogatory label felt free to express their prejudice, resulting in lower
ratings than in the other two conditions.
382 (HAPTER 10
8,—————————————,
7
6
5
No comment Ethnic crtitcism Derogatory ethnic
label
Type of comment overheard
• Pro-Black participants
Anti-Black participants
.. Ambivalent participants
FIG U R E 10.1 Priming as a Releaser of Regressive Prejudice
White research participants either saw no comment about a Black confederate or read a note with a critical comment
that induded a reference to race (“Black person”) or a critical comment that included a derogator ethnic label
(“nigger”). Ratings of the Black person made by participants who had positive attitudes toward Blacks were not
affected by the type of comment read. However, participants with negative attitudes toward Blacks made more negative
ratings when they read the derogatory ethnic labeL In contrast, the ratings made by participants characterized by
ambivalent prejudice were more positive when they read the derogatory ethnic labeL
Moral Credentials. When we, the authors, were growing up, it was not unusual
for a White person who was accused of anti-Black prejudice to defend
him- or herself by saying something along the lines of “How can you call me
Iprejudiced? Why, some of my best friends are Black!” The claim of friendship
twas used as a kind of credential to establish the person’s lack of prejudice.
More recently, Benoit Morrin and Dale Miller (2001) have suggested that complying
with the nonn to avoid prejudiced behavior can have an ironic effect: It
can increase the likelihood of behaving in a prejudiced way in the future by re- I ~UCing the motivation to appear unprejudiced. They believe that acting in a
nonprejudiced way establishes what they call moral credentials by allowing people
to show others that they are not prejudiced and by reinforcing their own beliefS
that they are not prejudiced. Having established to themselves and to others that
they are not prejudiced allows them to act in a prejudiced manner if they are
disposed to do so; if challenged, they can point to their earlier behavior as evidence
of their lack of prejudice.
Monin and Miller (2001) tested their theory in a set of studies in which
some research participants could establish their unprejudiced credentials by
either rejecting a set of stereotypical statements about women or by selecting a
well-qualified woman or African American for ajob. Other participants had no
opportunity to act in a nonprejudiced manner. All participants then rated the
extent to which they thought a given job (such as construction foreman or
FROM PREJUDICE TO DISCRIMINATION 383
police officer) was better suited for a man or woman, or a Black person or a
White person. In all three experiments, participants who earlier had had the
opportunity to establish that they were nonprejudiced rated the job in the second
part of the study as better filled by a man or White person. Monin and
Miller (2001) concluded that “the more confident people are that their past
behavior reveals a lack of prejudice, the less they will worry that their future
behavior is, or can be construed as, prejudiced …. By fostering self-image
security … the establishment of moral credentials eUlboldens the [person] to
respond honestly in circumstances in which political correctness pressure militates
against honest expression” (p. 40).
Reactions to Having Acted in a Prejudiced Manner
Given that situations can arise that can lead people to make prejudiced responses,
how do they then react to having acted in a prejudiced manner? The initial
research on this topic focused on differences between people high and low in
explicit prejudice. Patricia Devine and her colleagues (Devine, Monteith,
Zuwerink, & Elliot, 1991; Monteith, Devine, & Zuwerink, 1993; Zuwerink,
Monteith, Devine, & Cook, 1996) hypothesized that people with nonprejudiced
self-images who act or think in a prejudiced manner experience a discrepancy
between how they think they should respond (that is, in a nonprejudiced manner)
and how they do respond (that is, in a prejudiced manner). This discrepancy
then leads to feelings of discomfort and guilt. To test this hypothesis, they examined
differences in how people believed they should respond in interactions with
an African American or a gay man and how they thought they would actually
respond. Not surprisingly, Devine and her colleagues found that people low in
prejudice had more stringent personal standards for nonprejudiced behavior than
did people high in prejudice. Nonetheless, people both high and low in prejudice
felt discomfort over discrepancies between how they thought they should
act and how they thought they would act. However, whereas people low in
prejudice felt guilty about their discrepancies, people high in prejudice did not.
Rather than feeling negative toward themselves over their discrepancies, people
high in prejudice experienced negative emotions about other people, such as
feeling angry and irritated at them, perhaps because they believed that other people
expected them to be unprejudiced and would pressure them to behave in
unprejudiced ways.
In addition to discovering their own prejudiced behavior, people may be
confronted by others who point out behaviors that could indicate prejudice.
How do people respond then? Alexander Czopp and Margo Monteith (2003)
found that the response depended on two factors. The first was the type of prejudice
involved. People felt more concerned and guilty over racial prejudice than
over sexism. In fact, “the predominant evaluative sentiment resulting from confrontations
about gender-biased behavior was amusement” (Czopp & Monteith,
2003, p. 541), suggesting that many people do not take gender-based prejudice
very seriously. The second factor was the person who did the confronting.
People were nlore likely to dismiss an accusation of prejudice when it came
384 (HAPTER 10
from a menlber of the group toward which their prejudiced behavior was
directed than when it came [r0111 a member of their own group. Czopp and
Monteith suggested that this reaction occurred because people felt less threatened
when confronted by a member of their own group. Thus, it appears that the
greatest guilt and discomfort over acting in a prejudiced manner is elicited
when people have a high internal 111_otivation to control prejudice and when
they become aware of their prejudiced responses themselves or have those responses
pointed out by a member of their own group.
Researchers have found that guilt over having acted in ways that are discrepant
from one’s self-image leads to action that reaflinns that image (Steele, 1988).
Interestingly, there seems to be no recent research on whether feelings of guilt
about having acted in a prejudiced nlanner affect future behavior, but several
older studies have done so. For example, Steven Shemlan and Larry Gorkin
(1980) induced some research participants to nlake a sexist decision by leading
them to assume that a surgeon depicted in a story was male rather than female;
participants in a control condition nude no such decision. Later, all participants
took part in a seemingly unrelated study in which they acted as mock jurors in a
sex discrimination case involving a woman who alleged that she had been refused
a job because of her sex, with a man being hired instead. Shennan and
Gorkin found that, conlpared to participants in the control condition, those
who earlier had been induced to make the sexist decision were more likely to
see the hiring of the man as unjusti£ed and were more likely to decide the case
in favor of the woman.
Perhaps more telling are the results of an eXperll11ent conducted by Donald
Dutton and Robert Lake (1973). Based on a pretest, they selected 80 students
who were low on prejudice and who had also rated equality as a value that
was important to them. The participants thought they were taking part in a study
of physiological responsiveness to vatious stimuli and that high physiological
arousal indicated the presence of unconscious negative attitudes toward a stimulus.
While their physiological responses were recorded, participants watched a
series of slides that included pictures of Black people. Half the participants received
feedback that indicated unusually high responsiveness to the picture of
Black people relative to neutral stimuli (thus threatening their nonprejudiced
self-images) whereas the other participants received feedback that indicated similar
responses to the two types of stimuli (thus leaving their nonprejudiced selfimages
unthreatened). All palticipants were paid $2.00 in quarters (to ensure that
they had change for what happened next) for their participation. On leaving the
building, participants were approached by either a Black or a White panhandler
who asked “Can you spare some change for some food?” Eighty-five percent
of the participants in the self-image threat condition gave money to the Black
panhandler compared to 45 percent of those in the nonthreat condition. In
addition, the participants in the threat condition gave more money (an average
of 47 cents) than the participants in the nonthreat condition (an average of
17 cents). Threat condition did not affect donations to the White panhandler.
By the way, if 47 cents does not seenl like much money, it is equivalent to
about $2.30 today when corrected for inflation. Thus, when people who have
FROM PREJUDICE TO DISCRIMINATION 385
nonprejudiced self-images have that image called into question, they feel guilty
and that guilt motivates them to reaffiun their self-images of low prejudice by
acting in an especially nonprejudiced manner,
DISCRIMINATION IN ORGANIZATIONS
As we noted in Chapter 1, organizational discrimination occurs when “thQ
practices, rules, and policies of formal organizations, such as corporations or government
agencies” result in different outcomes for members of different groups
(Benokraitis & Feagin, 1995, p. 44). As we also saw in Chapter 1, although more
than 40 years have passed since the enactment of the landmark Civil Rights Act
of 1964, racial/ethnic and gender discrimination still exist in the workplace, We
have divided our discussion of this discrimination into two parts, The first part
examines employment discrimination at the organizational leveL The second part
examines some of the psychological variables that influence the discriminatory
behavior of individual decision makers, As you read this section, bear in mind
that although the eifects of gender and racial bias tend to be small (see, for example,
Barrett & Morris, 1993), small effects can cumulate over time to have a
large impact on career outcomes, For example, Richard Martell, David Lane,
and Cynthia Emrich (1996) conducted a computer simulation that modeled the
potential effects of gender discrimination in an organization that had eight levels
of promotion, The simulation started with an equal number of male and female
new employees. Althoughjob qualifications vatied among both male and female
employees, they were, on the average, the same for both groups; however, there
was a very small pro-male bias for promotions at each leveL Mter the simulation
had run through all eight levels of promotion, 65 percent of the top-level jobs
were filled by men, Therefore, even when discrimination has a small effect on
anyone decision, it can have larger effects in the long mn, In addition,
experiencing disctimination can have profound psychological and physical affects,
a topic we discuss in Chapter 11,
Organizational Research
Organizational researchers have conducted an enormous amount of research
to determine the forms discrimination takes in organizations and the ways in
which discrimination is related to characteristics of organizations, As we will
see, racial discrimination has been found at many points in the employment
process, including hiring, job performance evaluations, and promotions, There
appears to be less gender discrimination in hiring and performance evaluation,
but evidence still supports gender discrimination in promotion. Let us look at
the results of some of that research,
Hiring. The first step in the employment process is hiring, so quite a bit of researc has been conducted on the hiring process, most of it focusing on racial discriminaJtion.
One technique that has been used to study this topic is the employment audit.
In an employment audit, members of two groups are matched on appearance,
386 CHAPTER 10
education, and relevant experience, and then sent to apply for the same job. Thus,
one White person and one Black person, both with the same qualifications, dressed
similarly, and trained to act similarly, would apply for the same job. The dependent
variable in such studies is who is lucre likely to be hired, the Black applicant or the
White applicant? A review this research showed that for the most part, Black and
White applicants experience the same outcome: in 68 percent of the cases neither
was offered ajob and in 15 percent of the cases both were offered ajob (Heckman,
[~998). However, when discrimination did occur, it favored the White applicant:
overall, 27 percent of White applicants were offered jobs compared to 21 percent
of Black applicants. That is, a White applicant was 1.3 times more likely to get a
job than was a Black applicant.
The results of an employment audit study conducted by Marc Bendick,
Charles Jackson, and Victor Reinoso (1994) illustrate some specific forms that discrimination
can take. For example, they found that although employers accepted
applications from Black and White job seekers at about the Sa111e rate-95 percent
for Whites and 92 percent for Blacks–48 percent of the White applicants were
offered interviews cOlnpared to 40 percent of the Black applicants. When more
than one opening was available and the job was offered to both applicants, Black
applicants were offered lower starting salaries. In addition, Black applicants were
more likely to be steered toward a job at a lower level in the organization than
the one they had applied for. Bendick and his colleagues also found that discrimination
was more likely to occur when applications were made through employment
agencies rather than directly to the company. One reason for this finding
might be that the companies using employment agencies cued the agencies that
they only wanted to see White applicants. For example, evidence was presented
in one employment discrimination lawsuit that employers would use tenns such as
“all-AInerican,” “front-office appearance,” and “cOlporate image” in the job descriptions
they sent to employment agencies to indicate that they wanted to see
only White job candidates (Wolff, 1989).
r
– One shortcoming of employment audits is that it is impossible to perfectly
match a Black and a White applicant in every regard: They are different people
with different personalities, so it is always possible that any differences in the responses
they received are a result of their individual characteristics rather than of
their racial group membership. Therefore, researchers sometimes use matched
mailed-in resumes to see whether a Black or White applicant is more likely to
be called for an interview. Marianne Bertrand and Sendhil Mullainathan (2003)
conducted a large-scale study of this kind, sending out aInlOst 5,000 resumes
in response to 1,300 jobs advertised in newspapers in Chicago and Boston.
They manipulated three independent variables using infonnation in the
resumes: applicant race, applicant gender and applicants’ qualifications. They
manipulated race and gender by varying the applicant’s first name: “White”
names included Kristin, Carrie, Brad, and Jay; “Black” names included Ebony,
Latonya, Jennaine, and Leroy. The researchers chose the names based on the
frequency of their appearance in birth records and checked the racial associations
the names elicited by asking survey respondents to rate how “Black” and
“White” each name sounded.
FROM PREJUDICE TO DISCRIMINATION 387
Bertrand and Mullainathan (2003) found that, overall, only 17 percent of the
applicants were called for an interview; however, 10,1 percent of the White applicants
were called compared to 6.7 percent of the Black applicants, In other
words, White applicants received an interview offer for every 10 resumes submitted,
whereas Black applicants had to submit 15 resumes to get an interview,
In addition, although White applicants with high quality resumes were more
likely to be called than White applicants with low quality resumes, 11.3 percent
compared to 8,8 percent, resume quality had virtually no effect on the response
rate for Black applicants: 7,0 percent for high quality resumes and 6.4 percent for
low quality resumes. Note that a White applicant with a low quality resume was
26 percent more likely to be called for an interview (8.8 percent) than a Black
applicant with a high quality resume (7.0 percent). Resnlts were similar in
Boston and Chicago and for six types of jobs ranging from clerical worker to
manager across six industries, Bertrand and Mullainathan put their results in this
context: “in our sample, [having] a White name is equivalent to about 8 additional
years of [job-related] experience” in being called for an interview (p. 10).
An important part of the hiring process is the job interview, A review 31
studies of racial group differences in employment interview outcomes showed
that, on the average, Whites received higher ratings than either Blacks or
Hispanics (Huffcutt & Roth, 1998). Such racial differences in evaluations were
greater for low level jobs than for higher level jobs and discrimination increased
as the proportion of Black applicants increased. Thus, White job applicants are
more likely to get interviews than are their equally qualified Black peers and,
once interviewed, are more likely to get high ratings on suitability for the job.
Box 10.3 relates one Black man’s experience in applying for a low-level job and
a few of his experiences on that job,
Lawrence Graham (1995), a highly successful Harvard
educated lawyer, set out to uncover what life was like as
an employee at one ofthe all~White country clubs in
Greenwich, Connecticut, In his essay on his experiences,
he first describes applying for work, His goal was to be a
waiter and he applied for this job at five country clubs,
As he writes, “During each of my phone conversations, I
made sure that I spoke to the person who would make
the hiring decision, I also confirmed exactly how many
waiter positions were available, and I arranged a
personal interview within forty minutes to an hour of
the conversation, just to be sure that they could not tell
me that no such job was available” (p, 4). Upon arrival
at each of the five clubs, he was told either that there
were no openings or that he was not qualified. One
receptionist threatened to call security. Another
employer firmly insisted he could not have been the
person she talked to on the phone, No one would even
accept his application for a waiter position.
Graham did secure a job as a busboy at one of the
clubs. In that role, people made negative comments
about “Negroes” and other minority groups right in
front of him. Example comments included “My
goodness. , ‘ . That busboy had diction like an educated
White person” (p. 12). He learned that the staff
quarters were unselfconsciously called the “Monkey
House” because at one time all the workers had been
Black. At one point, he was instructed to find the
“Chinaman,” a supply clerk, and was told it was easy to
remember his location because it was right next to the
laundry. These and many similar experiences took
place in less than a one~month period. They also took
place in the 1980s, not,. as you might expect by reading
them, during the pre-dvil~rights era of the 1960s.
388 (HAPTER 10
There is less research on gender differences in hiring than on racial
differences. However, Bertrand and Mullainathan (2003) found that male and
female applicants were called for interviews at about the same rate, 17.2 percent
for women and 15.3 percent for men. Also, Heather Davison and Michael
Burke (2000) reviewed 49 studies in which participants evaluated the suitability
of equally qualifled male or female applicants for a job. Overall, they found
virtually no discrimination, although men tended to receive slightly higher
suitability ratings for male-stereotyped jobs and women tended to receive
slightly higher suitability ratings for female-stereotyped jobs. Thus, the research
indicates that currently there is little gender discrimination in the hiring process,
However, Davison and Burke noted that research reviews published in
1979 and 1988 did find evidence of discrimination, suggesring that gender discrimination
has decreased over time. For example, the 1988 review (Olian,
Schwab, & Haberfield, 1988) found that gender accounted for 4 percent of
the variance in hiring evaluations whereas the 2000 review found that it
accounted for less than 1 percent.
There is, however, one important factor that leads to gender discrimination in
lllring: pregnancy. In the United States, discriminating against pregnant women violates
the federal Pregnancy Discrimination Act, passed in 1975 as an am_endment to
the Civil Rights Act. However, the legality of this discrimination has not reduced the
number of claims filed with the Equal Opportunity Employment Commission; in
fact, number of claims has increased nearly every year since 1992 (Equal
Opportunity Employment Conmllision, 2007). Evidence that pregnancy affects hiring
decisions comes from a study of undergraduate business school students who evaluated
job applicants based on an interview. Although pregnant and nonpregnant
applicants were seen as equally qualified, pregnant women received lower hiring reconunendations
than did nonpregnant applicants (Cunningham & Macan, 2007).
Raters also believed that the pregnant applicant would be absent more often and
would be more likely to quit than the nonpregnant applicant.
[
Performance Evaluation. Once a person is on the job, career success is
strongly affected by perfonnance evaluations: Positive evaluations can lead to
valuable training opportunities, promotions, and pay raises, whereas poor evalua~
ons can lead to losing one’s job. Chieh-Chen Bowen, Janet Swim, and Rick
Jacobs (2000) reviewed 27 studies of gender differences in job perfonuance ratings
in organizations. They found that, overall, men and women received equal
job perfonnance ratings, although male evaluators tended to give somewhat
higher ratings to men than to women. Unlike Davison and Burke (2000), who
found that both men and women received slightly higher suitability ratings for
sex-stereotyped jobs, Bowen and her colleagues found that the sex typing of the
job was not related to perfonnance ratings. However, the sex stereotyping of the
items on perfOlmance evaluation measures was related to the ratings: Men received
higher ratings on measures that included mostly masculine-stereotyped items,
WOlnen received higher ratings on measures that included mostly femininestereotyped
items, and men and women received equal ratings on Ineasures that
contained mostly gender-neutral items. Bowen and colleagues also found that
FROM PREJUDICE TO DISCRIMINATION 389
supervisors who had more opportunity to directly observe employees’ job performance
gave somewhat higher performance ratings to women. They concluded
that “any overall gender discrimination that one might observe in the work force
is … not directly attributable to biases in performance appraisal” (Bowen et al.,
2000, p. 2206). However, their results do indicate that to the extent that perf ornunce
evaluations are made by Inen, especially men who do not directly observe
their employees’ perfonnance, and to the extent that an organizations’ performance
evaluations measures include masculine-stereotyped items, men will receive
higher evaluations than women.
Unlike the situation for gender, race/ethnicity does appear to affect job perfor n1.ance evaluations. Based on a review of 48 studies, Philip Roth, Allen HuffcuttJ,
and Philip Bobko (2003) found that White employees received higher evaluations
than Black employees, although there were no differences in evaluations of
Hispanic and White employees. Even when numerical ratings of job perfonnance
are the same for Black and White workers, the narrative comments that supervisors
make about employee performance might be biased. For example, Patricia Thomas
and her colleagues (Thomas, Edwards, Perry, & David, 1998) compared the perfonnance
evaluation comments made by the supervisors of 582 male U.S. Navy
officers who had received the highest nuulerical perfonnance rating. They fouud
that White officers were more likely to be described as outstanding leaders and
were more likely to be recommended for positions of command than were Black
officers. In addition, White officers were more likely to be reconnnended for early
promotion and to be described as having characteristics that other research shows
lead to early promotion (Johnson, 2001). That is, promotion boards seem to use
certain characteristics as cues when selecting officers for early promotion and evaluators
were more likely to attribute those characteristics to White officers than to
Black officers. Similarly, letters of reference for faculty positions in chemistry were
similar in length and, overall, were equally positive for male and female applicants.
However, letters supporting male candidates contained more standout words, such
as ”Inost” gifted and “rising” star than did letters supporting female candidates
(Schmader, Whitehead, & Wysocki, 2007).
Although racial group differences in job performance evaluations might reflect
racial prejudice, they nright also be accurate, reflecting actual racial group differences
in average job performance. For example, Roth and his colleagues (2003)
found that, compared to White workers, Black workers had lower average scores
on objective measures of work pelfonnance, and that these differences were about
twice as large as the differences in supervisor ratings. However, even though minority
group employees may actually perfonn at a son1.ewhat lower level than
White employees, that lower level of performance nright itself be a result of
prejudice and discrimination in the workplace (Greenhaus, Parasuramen, &
Wonnley, 1990). That is, differences in race may lead to differences work experiences
which, in tum, affect job performance. The possibility is supported by the
fmding that although both White and Black supervisors give White workers
slightly higher performance ratings than Black workers, White supervisors see larger
perfom”nce differences than do Black supervisors (Stauffer & Buckley, 2005). That
supervisor race affects perceptions of work performance suggests that subtle
390 CHAPTER 10
factors may be affecting performance ratings. Research shows, for example, that
members of higher status groups-in this case, White supervisors-are more
likely to favor the ingroup than are melnhers of low status groups-in this case,
Black supervisors (George & Chattopadhyay, 2002); if so, this would give an
edge to White workers. Organizational researchers have not paid Inuch attention
to issues such as these (Brief, 1998); however, organizational theorists have proposed
a number of workplace characteristics could adversely affect minority
group members’ job performance (Roberson & Block, 2001).
[
One such characteristic is what Daniel Ilgen and Margaret Youtz (1986) called
the lost opportunities dJect: “Differential treatment of minority and llUjOrity group
members may result in different on-the-job opportunities for these two groups. To
the extent that minority group members have fewer and less favorable opportunities,
lower performance for minorities may result” (p. 317). For example, minority group
managers report that, compared to their White peers, their supervisors appear to
view them as less competent (such as by reviewing their work more fi’equently
and more closely) and give them less support and encouragement (Blank & Slipp,
1994; Greenhaus et al., 1990; James, 2000). Prejudiced Whites also can create a
chilly climate for their Black coworkers. For example, Black managers reported
feeling less accepted than White coworkers by their White peers (Greenhaus et
al., 1990; Blank & Slipp, 1994). In addition, lower-level White workers showed
nlore work avoidance, such as absences, as the proportion of minority workers in
their work units increased and expressed stronger intentions to quit their jobs if
another employment opportunity arose (Tsui, Eagan, & O’Reilly, 1992). Such
coworker prejudice can have a direct effect on Black workers’ job performance.
For example, Black research participants who worked with a prejudiced, compared
with an unprejudiced, White participant were 30 percent less productive
(Dovidio, Kawakami, & Beach, 2001).
Ilgen and Youtz (1986) also proposed that workplace prejudice and discrimination
could have an adverse effect on minority employees’ morale, which, in
tunl, could affect job perfonnance. On a survey of the experience of workplace
mistreatment, such as being set up for failure, being treated as if they didn’t exist,
and being the target of insulting jokes or comments, Black workers reported
more mistreatment than White workers (Deitch et al., 2003). Moreover, the
experience of such mistreatment was related to reduced job satisfaction. Other
research shows that lower job satisfaction is related to lower job perfonnance
(judge, Thoreson, Bono, & Patton, 2001). For examples of some of the hassles
that Black workers face, see Box lOA.
Promotion. Promotion within an organization is based largely on job pelformance
evaluations. However, as Bowen and colleagues (2000) have noted, even
when women and members of minority groups receive the SaIne perfonnance
evaluations as men and majority group members, decisions based on those evalua- itions can still be biased. And so it seems to be with promotions. Looking first at
gender differences, researchers have found that although women and men receive,
on average, equal job perfomlance ratings, their supervisors often see them as
The editors of Harvard Business Review asked
organizational consultants Keith Caver and AnceUa
Livers (2002) to write a fictional letter from a Black
manager to a White boss describing some of the
everyday, almost certainly unintentional, incidents that
make B lack managers feel like outsiders. The events
they noted included:
• Being cast as a de facto diversity expert even
though the managers’ training gave them no
special expertise in that field: “Despite my 15
years of experience, despite my solid track record,
my new colleagues appeared to have little
interest in my business expertise. Instead, they
seemed to have assigned me some special role:
official interpreter of minority concerns for the
organization” (p. 78).
• Having one’s presence questioned when others go
unchallenged: “One weekend I went to the office
in my normal, casual weekend attire …. Before
getting into the elevator I was stopped by an
informally dressed young white man who in a
stern voice asked to see my identification …. I had
worked here for two years, but because I was out
of context, he assumed I was a thug. You might
chalk it up to an honest mistake, but I can assure
you he hadn’t challenged any of the white people
entering the building” (p. 79).
• Resentment over increased workplace diversity:
Robert, a Black manager, hired a Black woman
and promoted a Black man. Afterwards, “Robert
FROM PREJUD!CE TO D!SCR!MINAT!ON 391
began to hear whispers in the halls-suggestions
that he was building his own little ‘ghetto
fiefdom’ and having a White colleague ‘jokingly’
say to him, ‘So white people aren’t good enough
for you?'” (p.79).
• Not being trusted to do one’s job properly: In
addition to the comments of his peers, Robert’s
boss “suddenly seemed to take a greater interest
in the details of his group’s work-asking for
reports and updates he’d never needed when
Robert’s team was primarily white. Subtly, his
boss was letting him know that at some
level he expected the team’s performance to
drop” (p. 79).
• Having others assume that because you are not
White, you are a low-level employee: A Black
woman “was recently hired as a senior vice
president for a major financial institution. With
the exception of a few initial interviews and
meetings, she did not set foot in the organization
until her first day at the office. As she emerged
from the elevator, she was abruptly greeted by a
white male who directed her to a small cubicle
and asked her to quickly put her things away as
they were expecting a new senior officer to arrive
shortly” (p. 80).
The effect of experiences such as these is that for
minority workers, “race is always with us. As a friend of
mine said recently, ‘I don’t think a day goes by that I’m
not reminded I’m black'” (p. 81).
having less promotion potential than men (Landau, 1995; Shore, 1992), they mustJ
wait longer for promotion (Cannings & Montmarquette, 1991; Maume, 1999),
and they receive fewer promotions the higher they move in the organizational
structure (Lyness &Judiesch, 1999).
Even in situations in which women receive higher job performance ratingsl
than men, they may be promoted no faster (Cannings & Montmarquette, 1991; J
Shore, 1992). For example, Ted Shore (1992) studied the results of assessment
center ratings of male and female managers. When managers attend an assessment
center, they undergo a set of written and perfonnance tests designed to assess
their potential for prol11-etion. Shore found that, on the average, women and
men received the same scores on measures of mental ability and similar ratings
on interpersonal skill. In addition, women scored higher on measures of managerial
job perfonnance. Nonetheless, women received the san’le overall managerial
potential ratings as men and were promoted at the same rate.
392 CHAPTER 10

‘ Ironically, the male advantage in promotion may be greater in femaledominated
occupations, such as clerical and sales, than in male-dominated occupations,
such as nunufacturing (Maume, 1999), an outcome called the glass
escalator effect (Williams, 1992). For example, David Maume (1999) found that
—fnen were 17 percent more likely than WOlnen to be promoted in femaledominated
occupations. Being in a female-dominated job not only increased
men’s odds of promotion, it decreased women’s. Similarly, Christine Williams
(1992) found that male nurses, librarians, elementary teachers and social workers
benefited from being in a female-dominated profession both at the hiring
stage and on the job. However, these men often encounter negative stereotypes,
such as the perception that they are passive or felninine and they are
sometimes viewed -with suspicion-so much so that they “alter their work behavior
to guard against sexual abuse charges, particularly those in specialties
requiring intimate contact with women and children” (Williams, 1992, pp.
261-262). These negative stereotypes sometimes lead to negative self-esteem
and, ultimately, career change. Interestingly, negative stereotypes also may
contribute to the glass escalator effect: To avoid outsiders’ criticisms and suspicions,
supervisors may promote the men to higher level jobs where they have
less contact with the public (Williams, 1992).
[
In regard to race, Black workers are less likely to be promoted than their
White counterparts (Roth et aI., 2003) and must wait longer for promotion
(Maume, 1999). As with women and men, even when Black workers receive
the same job perfOlmance ratings as their White peers, they are likely to receive
lower ratings on promotion potential Ganles, 2000; Landau, 1995; Thomas et al.,
1998). Even after controlling for differences in education and experience, Black
assistant football coaches are less likely to be promoted to head coach than are
their white counterparts (Sagas & Cunningham, 2005). In addition, White employees
are paid more than Black enlployees in jobs with equivalent levels of
authority (Dreher & Cox, 1996; Smith, 1997). Ryan Smith (1997) also found
that although Black employees’ absolute salaty levels increased with promotion,
the difference in Black and White workers’ salaries increased with higher levels
of authority: Black workers with no authority were paid 86 percent of what their
White peers received, 81 percent at low levels of authority, and 77 percent at
middle levels of authority. (There were too few Black managers at high level
of authority for Smith to make meaningful comparisons.)
The problem begins at the lowest level of promotion. Using the results of
separate surveys, David Maume (1999) and Ryan Smith and James Elliott (2002)
both found that Mrican Americans in nomnanagerial jobs were 50 percent less
likely to be promoted to Inanagerialjobs than were their White peers. Smith and
Elliott further noted that when Mrican Americans did hold first-level managerial
positions, they were nl0re likely to supervise Black workers than White workers.
Because most Black workers are in low-level jobs, this ethnic matching of supervisors
and employees means that most Black managers are found in low-level
positions of authority. Smith and Elliott (2002) refer to this ethnic matching phenomenon
as the “sticky floor” effect: “The relative position of one’s ethnic
group within an organization constitutes the ‘sticky floor’-one to which
FROM PREJUDICE TO DISCRIMINATION 393
individual opportunity for authority ‘adheres.’ If one’s ethnic group dominates
only entry-level jobs within an organization, then one’s authority chances will
be restricted largely to supervising entry-level workers. If one’s ethnic group
dominates higher-level positions, then one’s authority chances will increase accordingly”
(p. 274; see also James, 2000).
At least two other organizational factors are related to Black managers’
slower promotion rates relative to their White peers. First, Black employees
tend to be “tracked” into certain job categories, such as affinnative action officer.
Black coaches, for example, often fill positions such as recruiter or minority affairs
officer that derail them from advancelnent to a head coach position (Sagas &
Cunningham, 2005). Similarly, Sharon Collins (1997) interviewed Black executives
in large White-owned corporations and found that 63 percent had been
career-tracked into jobs designed to “mediate the social pressures related to black
protest for civil rights, [such as] affinnative action or urban affairs manager”
(p. 327). Often, the only “qualification” these executives had was their race:
Professionals with advanced degrees in fields such as accounting, engineering,
and chemistry were put into these positions despite the oct that they had no
training or previous experience in the field. These managers were moved from
their chosen career fields based solely on a stereotype-that members of minority
groups make better diversity managers simply because of their group membership.
Jobs in these categories tend to be “dead-end” positions or have slower
promotion rates and “top out” at lower levels of authority than other jobs in
other categories, such as sales and operations management, regardless of the race
of the people holding the jobs (James, 2000).
Second, one’s mentor can have important effects on career progression because
of the fonnal and informal training and social support mentors provide, and
because of the connections they have with upper-level management (Roberson &
Block, 2001). Because White men predominate in upper-level management positions,
they can be especially helpful as menton;. For example, George Dreher and
Taylor Cox (1996) found that having had a White male mentor added more than
$22,000 per year to manager’s salaries after 10 years of experience compared to
managers who had had female or non-White mentors. They also found that
women, African American men, and Hispanic men were less likely than White
managers to have had a White male mentor.
Although it is tempting to attribute the differential outcomes for female and
Black workers to intentional discrimination, Smith and Elliott (2000) see the situation
differendy: “We believe that something more subtle and profound occurs
in the process of doing ‘business as usual’-mere maintenance of the status quo is
more than enough to perpetuate racial stratification” (p. 274); the same principle
applies to gender stratification.
Individuals in Organizations
Although people very often talk about organizational discrimination and institutional
discrimination, if you think about it, organizations and institutions do not
discriminate: Individuals in organizations and institutions discriminate. That is,
394 CHAPTER 10
individuals set and enforce discriminatory policies, make discriminatory hiring
decisions, and give discriminatory perfom1allCe evaluations. Even when decisions
are made by committees, individuals have input into those joint decisions. Given
this situation, surprisingly little research has been conducted on how individuallevel
psychological processes influence discriminatory outcomes in organizations
(Brief, 1998; Roberson & Block, 2001). This section discusses some of those processes:
stereotype fit, intergroup respect, shifting standards, contem-porary prejudice,
and confom-rity to perceived organizational norms.
i-Stereotype Fit. Madeline Heilman (1983; 2001) developed the stereotype fit
hypothesis to explain why WOlllen hold few managerial or executive positions
in organizations relative to men. The hypothesis postulates that the characteristics
associated with effective managers are very similar to the cultural stereotypes
_9f men and very different from the cultural stereotypes of Wonlen. Therefore,
nlen are perceived as fItting into the managerial role but women are not; as a
result, women are less likely to be hired for managerial positions and, once hired,
less likely to be promoted into higher positions (see also Eagly & Karau, 2002).
More generally, people see men as better suited for “masculine” jobs such
as business manager and construction worker and women as better suited for
“feminine” jobs such as nurse and secretary (Biernat & Kobrynowicz, 1997;
Cejka & Eagly, 1999).
Evidence for the manager-as-rnale stereotype has come from studies in
which experienced male corporate managers rated the target groups male manager
and female manager on traits that characterize effective managers (Dodge, Gilroy, &
Fenzel, 1995; Heilman, Block, & Martell, 1995; Mat~ell, Parker, Emrich, &
Crawford, 1998). These studies have found that male managers as a group received
higher ratings on the traits than did female managers as a group. However,
the results of the studies were less consistent when the research participants rated
succesiful male managers and succesifulfemale managers: two studies found a pro-male
bias (Dodge et a!., 1995; Heilman et a!., 1995) but one found no difference in
the ratings of successful male and feuule managers (Martell et a!., 1998).
Interestingly, the one study that included female managers as participants
(Dodge et al., 1995) found that they expressed a pro-female bias when rating
male and female managers in general, although not when rating successful male
and fenule managers. These results suggest that an egocentric bias might be at
work when rating managers in general: Because managers are making the
ratings, they see managers as having characteristics similar to their own-malestereotypical
when men are making the ratings and female-stereotypical when
wonlen are making the ratings. However, given that men predominate in
positions of power in organizations, nlen’s belief systems are more likely to influence
organizational decisions than are women’s.
Further evidence that evaluators may adjust the criteria to fit their expectations
comes from a study of potential applicants for the job of police chief
(Uhhnann & Cohen, 2005). Undergraduates evaluated a male or a female applicant
who was described as either streetwise (tough and had experience in rough
neighborhoods) but not fonnally educated or educated but not streetwise. Raters
FROM PREJUDICE TO DISCRIMINATION 395
adjusted the criteria they believed were important for the job such that, regardless
of which set of credenrials the male applicant had, he was considered better
qualified than the female applicant. When the male applicant was described as
streetwise, for example, those characteristics were seen as most important for a
police chief and not his education. Raters, then, did not stereotype the applicants,
but rather adjusted “their notion of ‘what it takes’ to do the job well in
a manner tailored to the idiosyncratic credentials of the person they wanted to
hire” (Uhhnann & Cohen, 2005, p. 478)-in this case by choosing the male
applicant.
An example of stereotype fit in operation comes from a study of the evaluations
and decisions that employment interviewers made about male and female
applicants for managerial jobs (Van Vianen & Willemsen, 1992). Consistent with
the stereotype fit hypothesis, interviewers believed that the ideal job applicant
would have more masculine traits than feminine traits. Moreover, although interviewers
regarded the male and female applicants as being equally qualified in
terms of education and experience, they were more likely to recommend that
male applicants be hired. Finally, interviewers attributed more masculine traits
to successful applicants than to unsuccessful applicants, indicating that masculinity
played an important role in their decisions. In short, because the interviewers saw
female applicants as less masculine than male applicants and viewed the jobs as
requiring masculine traits, they were less likely to recommend that female applicants
be hired even though they had the same objective qualifications as the male
applicants.
Stereotype fit (or lack of fit) can also influence performance evaluations.
Jennifer Boldry, Wendy Wood, and Deborah Kashy (2001) examined this influence
in a study of ratings male and female ROTC cadets nlade of one another.
They found that although male and female cadets scored equally well on objective
measures of military performance, female cadets received lower ratings on
motivation and leadership from their fellow cadets. Thus, equal performance,
when flltered through gender stereotypes, can lead to different evaluations. Also
recall the research discussed earlier that found that women are less likely to be
promoted despite equal job performance evaluations (Landau, 1995; Lyness &
Judiesch, 1999; Maume, 1999); similar processes might be operating in those
situations.
Although Heilman (1983, 2001) developed the stereotype fit hypothesis to explain
gender differences in organizational outcomes, it can also explain racial and
ethnic group differences in outcomes; however, less research has been conducted in
that context. One study that has addressed this question examined experienced
White managers’ ratings of managers, African Americans, and Whites on a set of
traits characteristic of good managers (Tomkiewicz, Brenner, & Adeyemi-Bello,
1998). Results showed a correlation of r = .54 between the ratings of managers
and the ratings of Whites but a correlation of on1y r = .17 between the ratings of
managers and the ratings of African Anlericans, indicating that Whites were seen as
fitting the managerial role better than were Mrican Americans. Similarly, White
managers in the hospitality industry saw a correspondence between the category
“successful middle manager” and the categories “Caucasian American middle
396 CHAPTER 10
manager” (r ~ .53) and “Asian American middle manager” (r ~ .56; ChungHerrera
& Lankau, 2005). They also saw a relationship between the categories
“successful middle manager” and “Mrican American manager,” but this correlation,
r::::: .24, was significantly lower than the relationships found between successful
and White or successful and Asian managers. The relationship between
“successful middle manager” and “Hispanic American manager,” r = .13, was not
significant.
Similar processes operate for lower level jobs as well. For example, Jaleen
Kirschenman and Kathryn Neckerman (1990) interviewed 185 Chicago-area
employers about their impressions of workers from various racial and ethnic
groups who fill low-level jobs such as sales clerk, typist, restaurant worker, and
assembly worker. They found that employers generally perceived Blacks and
Hispanics as having none of the characteristics of good workers, seeing them as
unskilled, illiterate, dishonest, unmotivated, involved with drugs and gangs,
lacking a work ethic, and having few interpersonal skills. The employers contrasted
Blacks and Hispanics with recent immigrants from central Europe,
whom they saw as having all the characteristics of desirable workers. These
stereotypes were reflected in em-ployee recruitment strategies. For example,
Kirschenman and Neckerman (1990) noted that “one company advertised for
skilled workers in Polish- and German-language newspapers, but hired all its
unskilled workers, 97 percent of whom were Hispanic,” through employee
referrals (p. 210).
Employers did have subcategories within their stereotypes, however. For
example, they differentiated between “desirable” Black workers-those who had
middle- or working-class backgrounds-from “undesirable” Black workersthose
who resided in urban ghettos. However, employers tended to use race as
a marker for ghetto resident, assuming that Black job applicants were ghetto residents
(and therefore would not be good workers) unless the applicants provided
evidence of being “desirable.” Such evidence included speaking standard English,
dressing appropriately for the job interview, being able to demonstrate job-related
skills, having a history of steady employment, and providing a non-ghetto address.
In sum, Kirschenman and Neckerman (1990) found that employers
assumed that Blacks and Hispanics were unqualified for even low level jobs
unless the applicant proved otherwise, but applied less stringent criteria to
members of other groups.
Finally, although stereotype fit can result in a disadvantage for some groups,
it can confer an advantage on other groups. For example, the Asian stereotype
includes being talented at math and computers whereas the female stereotype includes
having little talent in those domains. Jennifer Steele and Nalini Ambady
(2004) hypothesized that an Asian American woman’s perceived qualification for
a computer technician job would depend on whether her Asian identity or her
female identity was salient to the decision maker. College student participants
read the description of a student computer technician job, reviewed a completed
application for the job, and then interviewed the applicant, an Asian American
woman. The application form was designed to emphasize either the Asian or
female aspect of the applicant’s identity. When the applicant’s Asian identity,
FROM PREJUDICE TO DISCRIMINATION 397
rather than her female identity was salient, participants gave her a stronger hiring
reconunendation and recommended a 12 percent higher starting salary.
Intergroup Respect. One explanation for organizational diSCriminationll
then, is stereotype fit: Because the stereotype of White lllen provides a better
fit to the managerial stereotype than do the stereotypes of minority group
members or women, White men are selected over equally, or even better,
qualified members of other groups. Lynne Jackson, Victoria Esses, and
Christopher Burris (2001) have proposed a different explanation. They have
hypothesized that it is not a decision maker’s group stereotypes that primarily
affect discriminatory decisions, but rather it is the amount of respect the decision
maker has for the group that affects decisions. They define respect as “feelings
of esteem for another that manifest in both valuing the person’s feelings,
thoughts, and behaviors as well as willingness to be influenced by that person”
(pp. 48-49). Jackson and her colleagues propose that people hold different degrees
of respect for various groups based on the amount of power the groups
have in society, For example, “differential respect for men and women may
have its origins in the social structure. In most contemporary cultures, men
continue to hold higher social status than women” .. Relatedly, men generally
still have more power than women in society (e.g., in politics and economics),
in the workplace, and in interpersonal relationships … Because many positions
occupied by men are high in status, people are likely to more frequently act
deferentially to men than to women” (Jackson et al., 2001, p. 49). Living in
such a social context, they argue, leads people to develop more respect for men
as a group than for women as a group (although, of course, individual women
can be respected more than individual men).
Jackson and her colleagues (2001) conducted three studies to test the hypothesis
that respect outweighs stereotypes in affecting hiring decisions. In the
first two studies, research participants read a description of a gender-neutral job
and an application for the job from a man or a woman. Participants rated the
applicant on the degree of respect they felt for him or her, the extent to which
male and female stereotypes described the applicant, and how suitable the applicant
was for the job. The researchers found that although both respect and
stereotypes had positive correlations with job suitability ratings, respect had a
much stronger relationship. In the third study, the researchers manipulated both
how respected the applicant was by a previous employer and the applicant’s male
and female stereotypic characteristics using a letter of recommendation in the
applicant’s file. They also manipulated whether the applicant was applying for a
high or low status job. They found that greater respect led to stronger hiring
recommendations, especially for the high status job, and that having more masculine
stereo typic traits led to stronger hiring recommendations for both types of
job; feminine stereotypic traits were not related to hiring recommendations. As
in the first two studies, respect had a stronger effect than stereotypes. Taken together,
these findings indicate that the effect group membership has on hiring
goes beyond how decision makers think about groups (stereotypes) to include
how they feel about groups (respect). This distinction is important because it
398 (HAPTER 10
suggests that eliminating reliance on stereotypes may not eliminate discrimination
that is rooted in differences in respect for different social groups.
Shifting Standards. Think back to two of the sets of research results on gender
discrimination in organizations that we presented earlier. First, on the average,
women and ll1en receive equal job performance evaluations. Second, on the
average, women are less likely to get promoted than men. Given that rewards
such as promotions are supposed to be based on performance, these two sets of
findings appear to contradict one another: given equal performance, women and
men should be promoted at equal rates. One explanation for this apparent contradiction
lies in Monica Biemat’s (2003) shifting standards model of evaluation
that we introduced iu Chapters 3 and 4. Recall that the model proposes that
negative stereotypes lead people to hold lower performance expectations for
women and members of minority groups. When evaluators use subjective criteria
to rate performance, a person is rated relative to the expectations the evaluator
has for the person’s group. Because most workplace perfonnance evaluation
measures use subjective rating scales (see, for example, Murphy & Cleveland,
1995), job perfonnance ratings are vulnerable to this bias. For example, a woman
-manager’s perfonnance would be rated relative to the evaluator’s expectations for
woman managers as a group. The top section of Figure 10.2 illustrates this process.
Janet and Jason work for the same manager and perform at the same level.
However, because their manager has lower expectations for female employees
(that is, the female standards are shifted to the left of the male standards in this
example), the same level of performance results in a higher rating for Janet than
for Jason.
The bottom section of Figure 10.2 provides a hypothetical example of how
shifting standards can also influence the interpretation of performance ratings. In
this example, Jamal, who is Black, and Jerry, who is White, work for the same
manager, who gives them both a (very good) rating of 4 on a 5-point scale. But
their manager has, probably unconsciously, rated Jamal and Jerry relative to the
expectations he has for the performance of Black and White employees. That is,
he saw Jamal’s performance as “very good for a Black employee” and Jerry’s as
“very good for a White employee.” However, because of the different standards
used for rating Black and White performance, Jamal actually scores lower on an
objective common scale that takes both race-based scales into account. Biernat
(2003) notes that it is the objective comnlon scale) not the race-based subjective
scales, that determines the distribution of organizational rewards such as promotions
aud pay raises. Thus, although both Jamal and Jerry received ratings of 4 on
their annual performance evaluations, Jerry is more likely to get a promotion or
pay raise because his 4 translates to a 7 (very good) on the objective scale whereas
Jamal’s 4 translates to a 4 (average).
Recall, for example, the study conducted by Monica Biernat and Theresa
Vescio (2002) that we desctibed in Chapter 4, in which participants took the
role of the manager of a coed softball teanl. Even when male and female players
were evaluated as equal in athletic ability, men were more likely to be chosen
as starting players and to be placed in the top of the batting order. Thus, men
FROM PREJUDICE TO DISCRIMINATION 399
Objective performance:
0 2 3
“Female” scale:
0 2 3
“Male” scale:
0
“Black” scale:
0 2
“White” scale:
Underlying common scale:
0 2
FIGURE 10.2 Shifting Standards of Evaluation
Janet
Jason
4
4
Jamal
3 4
0
3 4
5
5
2
5
2
5
6 7 8
3 4 5
Jerry
3 5
6 7 8
Top Section: Janet and Jerry perform at the same objective level, but because the evaluator’s expectations for female
performance are lower than his expectations for male performance, Janet gets a higher performance rating. Bottom
Section: Jamal gets a high rating of 4, but it is relative to the low expectations the rater has of Black employees. Jerry
gets a high rating of 4, but it is relative to the higher expectations that the rater has for White employees. As a result,
when the two ratings are transformed to a common scale, such as if the rater had to rank employees, Jerry comes out
ahead of Jamal.
SOURCE: Adapt-ed from Biernat (2003, Figure 1. p. 1021).
received more rewards than women, presumably because shifting standards put
them higher on the objective scale the participants used to make their choices.
These kinds of differential decisions are made not only by evaluators but also by
other people who use the evaluations to make decisions. For example, Biernat
(2003) reported the results of a study in which participants read a letter of recommendation;
half the participants thought it was written on behalf of a Black
job applicant and half that it was written on behalf of a White job applicant. The
participants interpreted the letter to mean that, in objective terms, the Black
400 CHAPTER 10
applicant had perfonned less well on the job described in the letter than had the
White applicant. Participants apparently assumed that the letter writer had used
lower standards to evaluate the Black applicant’s performance.
Contemporary Prejudice. In Chapter 6 we noted that, because of contemporary
social norms that condemn prejudice, prejudice and discruuination tend to
manifest themselves in subtle ways and in situations in which prejudiced behavior
can be attributed to other causes, Thus, for example, a prejudiced person could
claim that she voted against a Black political candidate not because he was Black,
but because his platform was too liberal. In the em_ployment context, a prejudiced
employer may use applicant characteristics as a reason to reject a Black
job applicant even while hiring a White applicant with the same characteristic.
For example, Gordon Hodson, John Dovidio, and Samuel Gaertner (2002) asked
White college students who had been classified as being high or low on racial
prejudice to make college admission decisions about Black and White applicants.
The researchers created two mixed qualification applicant conditions, a clearly
high-qualification applicant condition, and a clearly low-qualification applicant
condition. In one mixed-qualification condition, the applicant had high
Scholastic Assessment Test (SAT) scores but a low high school grade point average
(GPA); in the other mixed qualification condition, the applicant had a
high GPA but low SAT scores. The higb qualification applicant had both high
SAT scores and a high GPA; the low qualification applicant had both low SAT
scores and a low GPA.
Both high- and low-prejudice participants accepted the highly qualified
Black and White applicants at the same rate (100 percent) and rejected the
poorly qualified Black and White applicants at the same rate (69 percent).
However, high-prejudice participants were more likely to accept the mixedqualification
White applicant (74 percent) than the mixed-qualification Black applicant
(44 percent); the reverse was true for low-prejudice participants although
the difference was not as large: 86 percent accepted the Black applicant and
64 percent accepted the White applicant. Participants also rated how much the
applicant’s SAT scores and GPA influenced their decisions. The high-prejudice
participants who had evaluated the mixed qualification Black applicant rated
whichever piece of negative infonnation they had seen-SAT scores in the one
condition and GPA in the other condition-as most influential. Low-prejudice
participants showed the opposite pattenl, focusing on whichever piece of information
was more positive. Thus, both high- and low-prejudice participants
seized on the infonnation that was consistent with their racial attitudes-positive
for those low in prejudice and negative for those high in prejudice-and used
that infonnation to justify- their decisions.
Employers may also use business-related factors to justifY discrimination, the
most common being maintaining workplace hannony and placating customers
(Brief, 1998; Kirschenman & Neckennan, 1990). Kirschenman and Neckennan
(1990) found that many of the employers they interviewed were reluctant to
hire minority workers because they thought doing so would upset their White
employees, leading to morale and productivity problems. This reluctance was
FROM PREJUDICE TO DISCRIMINATION 401
reflected in hiring practices: Employers who said they valued teamwork highly
were twice as likely to have racially homogeneous workforces than those who
thought that teamwork was less important. Unfortunately, these employers’ concerns
may have a basis in reality: lower-level White workers, for example, show
more work avoidance as the proportion of minority workers in their work units
increases (Tsui et al., 1992). Kirschenman and Neckerman (1990) also noted that
some employers believed that they would lose customers if they hired minority
workers. They quoted one restaurant owner as saying, “I have all white waitresses
for a very basic reason. My clientele is 95 percent white. I simply wouldn’t
last very long if I had some black waitresses out there” (p. 220). This concern
may also be based in reality: A White suburban restaurant owner who had hired
Black wait staff because he could not find enough White workers received comments
from his White customers such as “Why do you have those people out
here?” (p. 220, emphasis in original). Customer demand can also work in the
other direction: Harry Holzer (1996) found that employers with a predominandy
Black customer base were more likely to hire Black workers than White workers.
At colleges and universities, important “customers” include alumni, particular
those who donate to the university. Michael Sagas and George Cunningham
(2005) suggest that the possibility of alienating White alumni boosters may be
one reason why Blacks are less likely to be hired as head college football coaches.
Conformity to Perceived Norms. Finally, individuals might make diSCrimi~J
natOlY decisions because they believe those decisions are consistent with company
nOrms. For example, using a methodology developed by Arthur Brief and
his colleagues (Brief, Dietz, Cohen, Pugh, & Vaslow, 2000), Lars-Eric Petersen
and Joerg Dietz (2005) asked German students to imagine they were a department
head for a Gennan fast-food chain. Their task was to select three candidates
for a job interview, based on a pool that contained both foreign and
German applicants. Half of the participants read a memo from the company
president that commented on the number of foreign applicants and noted the
importance of nlaintaining a homogeneous workforce. The memo also noted
that the company employed almost exclusively Germans. The remaining participants
read no such memo. In addition, participants were categorized as either
high on a measure of subtle prejudice, high on a measure of blatant prejudice,
or nonprejudiced. Individuals high on subde prejudice were less likely to select
foreign applicants when they saw the memo than when they did not. Ratings of
nonprejudiced and blatandy prejudiced participants were unaffected by the
memo. These findings suggest that individuals who are blatandy prejudiced do
not need a justification to discriminate: these students were less likely to select
foreign applicants regardless of whether they saw the memo. Consistent with
findings from previous research (see Brief et al., 2000), those who were subdy
prejudice made discriminatory decisions when they believed that doing so would
please their boss. By the way, if pressure from a company president to discriminate
seems far-fetched, see Box 10.5, which describes the event that led to Brief
and colleagues’ research.
402 CHAPTER 10
In 1992 the restaurant chain Shoney’s Incorporated
paid $132.5 million to settle an employment
discrimination lawsuit. The evidence in the case
revealed a long-standing policy of minimizing the
number of Black employees in the company, especially
in customer-contact jobs. For example, 75 percent of
Shoney’s Black restaurant workers held minimum-wage
jobs such as dishwasher, cook, and breakfast-bar
attendant (Watkins, 1993). These employment policies
were a direct reflection of Chief Executive Officer Ray
Danner’s views. For example, one former Shoney’s vice
president described what he called Danner’s Laws:
“Blacks were not qualified to run a store. Blacks were
not qualified to run a kitchen of a store. Blacks should
not be employed in any position where they would be
seen by customers” (Watkins, 1993, p.427).
Danner’s justification for his policies was that
White customers did not want to see Black employees
and would not patronize restaurants that employed
Black customer service staff. In pretrial testimony
Danner said, “In looking for anything to identify why
[a restaurant] is under~performjng in some cases, I
would probably have said that this is a neighborhood
of predominantly white neighbors, and we have a
considerable amount of black employees and this
might be a problem” (Watkins, 1993, p. 427). Steve
Watkins (1993) reported that “the smoking gun in the
case came in the form of a letter Danner wrote
complaining about the performance of [one
restaurant] and comparing the racial makeup of the
store, which had several black employees-some of
whom were later fired-to the all~white, or nearly
all~white, composition of other fast~food restaurants
Danner visited in the area” (p. 427).
When executives from company headquarters
visited restaurants, they would tell managers whom
they thought had hired too many Black workers “to
‘lighten up’ their store-a company euphemism for
reducing the number of black workers-and hire
‘attractive white g’lrls’ instead” (Watkins, 1993, p. 424).
In another instance, “two black Shoney’s employees said
they were ordered by their manager to hide in a
restroom because some company executives had shown
up for a surprise visit and there were ‘too many’ blacks at
work that day” (Watkins, 1993, p. 426). These anti~Black
policies did not affect only Black employees: Watkins
reported that White restaurant managers who
disobeyed orders to “Iighten up” their staff were fired.
HATE CRIMES
Hate crimes (also called bias criines) are, perhaps, the most severe fonn of dis- !Crimination. Hate crimes are criminal offenses in which there is evidence that
the victims were chosen because of their race, ethnicity, national origin, religion,
disability, or sexual orientation (U.S. Federal Bureau of Investigation
-[FBI], 1999). Although whether a crime is caused by bias rather than some other
motive (such as personal animosity unrelated to prejudice) is sometimes a matter
of judgment, in many cases the evidence is fairly obvious. For example, a survey
of gay, lesbian, and bisexual hate crime victin1S found that in 53 percent of the
cases the offender made an explicit statement about the victim’s sexual orientation
(Herek, Cogan, & Gillis, 2002). In addition, hate crimes usually have no
motivation other than attacking a member of a particular group. “There appear
to be no gains for the assailant: There is no attempt to take money or personal
items and there is no prior relationship between the victim and offender” that
could provide a personal motive, such as revenge (McDevitt, Levin, & Bennett,
2002, p. 304).
Table 10.2 presents some of the characteristics of the hate crimes reported
to the FBI in 2006. The FBI received 7,720 reports of hate crimes that year
FROM PREJUDICE TO DISCRIMINATION 403
TAB L E 10.2 Characteristics of Hate Crimes
Reported in 2006
Victim Group
Racelethnicity (65.6% of all hate crimes)
Black
White
Hispanic
Asian
American Indian
Multiple races, group
Religion (18.1 of all hate crimes)
Jewish
Islamic
Catholic
Protestant
Other
Sexual orientation (15.3% of all hate crimes)
Male homosexual
Female homosexual
Heterosexual
Other
Disability (1 % of alJ hate crimes)
Mental
Physical
Race of Offender
White
Black
Unknown
Type of Offense
Against Persons (56.5% of all hate crimes)
Aggravated assault
Simple assault
Intimidation
Other violent crime
Property (43.2% of all hate crimes)
Vandalism
Burglary or theft
Arson
Other
SOURCE: u.s. Federal Bureau of Investigation (2006).
52.7%
16.7%
12.9%
3.8%
1.2%
12.7%
65.4%
11.9%
4.9%
3.7%
14.2%
62.0%
13.7%
2.0%
29.9%
77.8%
22.1%
58.6%
20.6%
12.9%
21.6%
31.9%
46.0%
0.3%
80.4%
16.9%
1.2%
2.7%
404 CHAPTER 10
involving 9652 victims, which is probably fewer than the true number given that
most crimes of all kinds go unreported (Strom, 2001). fu Table 10.2 shows, most
hate crimes were based on the victims’ race or ethnidty. However, one does not
have to be a member of a minority group to be a victim of a hate crime: in 2006,
16.7 percent of the victims were White. The majority of the crimes (56.5 percent)
were against persons, such as assault or intimidation; 43.2 percent were
crimes against property, such as vandalism, burglary, or a motor vehicle theft. A
study of hate crimes reported from 1997-1999 showed that most victims were
male (62 percent) and young (65 percent were under 35 years of age), most of
the crimes were violent (60 percent), and 18 percent involved the use of a deadly
weapon (Stonn, 2001). Jack Levin and Jack McDevitt (2002) have further noted
that hate crimes are excessively brutal compared to crune in general: 58 percent
of hate crimes involve assaults compared to 7 percent of all crimes and 30 percent
of hate crime assault victims receive physical injuries compared to 7 percent
of victims for assaults in general.
In this section, we discuss three aspects of hate crimes. First, we look at some
of the characteristics of hate crime offenders. We then examine the motivations
that offenders have for taking part in hate crimes. We conclude with a brief discussion
of the effects of hate crimes on the victims.
Hate Crime Offenders
Who conunits hate crimes? There are two ways of looking for answers to this
question. One approach is to examine victim descriptions of offenders, such as
those contained in the reports of hate crimes collected by the FBI. Kevin
Strom’s (2001) analysis of hate crimes reported from 1997~1999 showed that,
like most offenders, hate crime perpetrators are disproportionately male (84 percent)
and young: 62 percent are under 24 years of age and 79 percent are under
35 years of age. Vicky Kielinger and Susan Paterson (2007) examined London
police records of hate crimes based on racial or sexual orientation bias; their analysis
showed that two thirds of the suspects were known by the victim, either as
neighbors, colleagues, or school mates. As we will discuss, this is consistent with
research showing that hate crimes are generally not comrnitted by members of
hate groups, but instead tend to be opportunistic crimes.
Another approach to detennining offender characteristics is to conduct
surveys and examine the characteristics of people who admit to having participated
in hate crimes. Karen Franklin (2000) investigated anti-gay behavior and
found that 10 percent of her sanlple of 489 community college students admitted
having assaulted a lesbian or gay man (or someone the respondent thought was
lesbian or gay) and that an additional 24 percent admitted engaging in verbal
abuse. Like Strom, Franklin found that offenders were disproportionately male.
(She did not report results by age of respondent.) Franklin also found that men
were increasingly likely to be offenders as the violence of the behavior increased:
Men were the perpetrators in 64 percent of name-calling incidents, 79 percent of
intimidation incidents, and 92 percent of physical attacks. In addition, Franklin
found that, relative to their representation in her sample, Mrican Americans were
FROM PREJUDICE TO DISCRIMINATION 405
more likely than members of other racial groups to admit to having engaged in
anti-gay behaviors: African Americans constituted 19 percent of offenders versus
12 percent of the sample. African Americans were also more likely to admit to
physical violence than members of other groups: 29 percent versus 4 percent to
11 percent for other groups.
Although one might think that hate crime offenders hold extremely negative
intergroup attitudes or are unusually aggressive people, that is not always the
case, as we will see shortly, Also, as we saw in Chapter 9, very few hate group
members commit hate crimes. What factors, then, motivate people to commit
hate crimes? The next section addresses that question.
Motivations for Hate Crimes
Why do people commit hate crimes? Based on an analysis of 169 hate crimes
reported to the Boston police in 1991 and 1992, Jack Levin and Jack McDevitt
(2002) proposed four motivations for engaging in hate ctimes: thrill seeking, defending
one’s “territory” from outsiders, retaliation for a perceived offense committed
by an outgroup member, and a conunitment to a racist, anti-gay, or other
bigoted ideology. Franklin (2000) has identified a fifth motivation, peer group
dynamics, that Levin and McDevitt had included as part of thrill seeking
(McDevitt et al., 2002). Bryan Byers and Benjamin Crider (2002) have suggested
that seeing hate crime as a nonnal behavior may facilitate, if not motivate, such
behavior. Franklin (2000) also examined people’s reasons for not catrying out
anti-gay behaviors. Let us look at these motives.
Thrill Seeking. Thrill seeking as a motivation for a hate crime entails cOurmit-J~
ting the crime out of a desire for excitement, as an antidote for boredom. For ,
example, McDevitt and his colleagues (2002) noted that young people who had
been arrested for hate crimes “often told police that they were just bored and
looking for some fun …. The attack in these thrill-motivated cases was triggered
by an immature desire to display power and to experience a rush at the expense
of someone else…. Several of these young offenders revealed that their only
benefit from the attack was some vague senses of their own importance: a sadistic
high as well as bragging rights with their friends” (pp. 307-308).
Thrill seekers generally have little commitment to bias and often express little
animosity toward the group whose members they have attacked (Byers, Crider, &
Biggers, 1999; Franklin, 2000; McDevitt et al., 2002). Rather, they are bored and
see violence as a way of alleviating their boredom. For exanlple, Byers and Crider
(2002; Byers et al., 1999) interviewed young men who, as teenagers, had participated
in criminal and harassing behavior against the Amish residents of the rural
county in which they and their Amish victims lived. These offenses are so common,
and so commonly accepted by the non-Amish, that there is a local word for
them, daping. The role played by boredom in hate crimes is illustrated by the comments
of one of Byers and colleagues’ (1999) respondents, who offered this explanation
for his behavior: “It was what our friends were doing at the time to pass the
sununer months away or whatever. That was what we were doing on Friday
406 CHAPTER 10
nights” (p. 85). The results of Byers and colleagues’ (1999; Byers & Crider 2002)
interviews suggest that a lack of respect rather than animosity might be the emotional
facilitator of thrill seekers’ behavior. One of their respondents told them, “I
just had the mentahty that they are just Amish …. It is like, we can pick on them
because they are so different” (Byers et al., 1999, p. 87) and another said, “It is
because I still have some feeling that they ahnost ‘deserved it’ for some unknown
reason because they are different” (p. 88).
Thrill seekers tend to choose as targets people they see as providing easy and
safe opportunities for violence. For example, an offender who, along with a
friend, targeted gay men as robbery victims said, “It wasn’t because we had
sonlething against gays, but because we could get SOBle money and have some
fun. It was a rush. A serious rush. Massive rush … , It was nothing at all against
gays. They’re just an easy target. Gays have a reputation that they can’t fight
[back]” (Franklin, 1998, p. 14, emphasis in original). Similarly, one of Byers
and Crider’s (2002) respondents said of the Amish, “They are an easy target.
They offer an easy target because they ‘turn the other cheek’ and don’t fight
back” (p. 131). Another characteristic that makes a group an easy target for thrill
seekers is an unwillingness or inability to report the crime. For exanlple, one of
Byers and Crider’s (2002) interviewees noted that the Amish “can’t call the cops
[because of their rejection of modern technology, such as telephones] and don’t
believe in suing” (p. 135). Similarly, lesbians, gay men, and bisexuals may be
seen as easy targets because they are reluctant to report hate crilnes due to concern
over police harassment or public disclosure of their sexual orientation
(Herek et aI., 2002).
Thrill seekers often justify their actions by minimizing the crime’s impact on
the victims and by portraying their actions as hannless fun (Byers et al., 1999;
Franklin, 1998). One of Byers and colleagues’ (1999) respondents said, “It was
all, I always thought clean fun …. We always looked at it as there are a lot worse
things that we could be doing” (p. 85). Byers and his colleagues found that denying
that they had hurt anyone was a common justification offenders gave for
their behavior. One of their respondents said about destroying an outhouse, “No
one ever really got hurt, and it wasn’t really that much property damage. It was
pretty much just a mess to clean up” (p. 85). Besides, one respondent explained,
claping causes no real injury because the Amish should expect to be harassed:
“Stuff like that happens to them. It happens to them all the time. They are
used to it I think” (Byers et al., 1999, p. 86).
[
Territorial Defense. In defensive hate crimes, the perpetrators see themselves
as protecting their own territory fronl invasion by outsiders. The purpose of this
~type of hate crime is to coerce the outsiders to go away and to send a more
general message that members of the victim’s group are not wanted in the offenders’
neighborhood. For example, Donald Green, Dara Strolovitch, and Janelle
Wong (1998) found that the incidence of hate crimes was higher in all-White
neighborhoods into which minority group members were moving compared to
similar neighborhoods which remained all White. Defensive hate crimes constituted
25 percent of Levin and McDevitt’s (2002) sample of cases.
FROM PREJUDICE TO DISCRIMINATION 407
Retaliation. In retaliatory hate crimes, the offenders are seeking revenge for 0-real
or rumDred attack on a member of their ingroup, McDevitt and his colleagues
(2002) note that “whether the [attack] actually occurred is often irrelevan .
Sometimes a rumor of an incident may cause a group of offenders to take vengeance,
only to learn later that their original infonnation was merely unfounded
hearsay” (p. 309). Although retaliatory attackers cite revenge as the reason for
their actions, they usually do not seek out the person they believe committed
the offense against their group, but target any available member of the group.
This is consistent with the vicarious retribution model, discussed in Chapter 9, which
explains how groups decide when to retaliate against an outgroup (Lickel, Miller,
Stenstrom, Denson, & Schmader, 2006).
This kind of generalized retaliation may be especially likely to occur when the
real target of the offenders’ anger is out of their reach, As Levin and McDevitt
(2002) noted, “After [the September 11, 2001, terrorist attacks], what made it especially
tempting to target college students who spoke with an accent and had a dark
complexion was the ambiguity in identifYing the real enemy. President Bush
blamed Osama bin Laden, a shadowy figure who resided in a far-off land and had
been seen only a few times on videotape, For most Americans, bin Laden ‘” was
an abstraction, little more than a caricature. It was therefore far more satisfying
psychologically to target flesh-and-blood human beings [who were close at hand]
-international students” (p, 5). Retaliatory hate crimes constituted 8 percent of
Levin and McDevitt’s (2002) sample of cases.
Mission. Mission-motivated hate crimes are carried out because of a commit”::-l
ment to a bigoted ideology. In these kinds of crimes, “the perpetrator seeks toJ
rid the world of evi!” (McDevitt et al., 2002, p. 309). Some mission-oriented offenders
are members of hate groups, although they may be acting without the
knowledge or support of the group’s leadership. As we saw in Chapter 9, the leadership
of many hate groups publicly oppose violence, seeing it as a threat to their
recruitment efforts. Other mission-oriented offenders act on their own, seeing
themselves as victims of conspiracies by groups against whom they seek revenge
(Levin & McDevitt, 2002). Mission hate crimes are extremely rare; they constituted
less than 1 percent of Levin and McDevitt’s (2002) sample of cases.
Peer Group Dynanrics. Many hate crimes, especially thrill- and defense-J-motivated
crimes, are committed by groups of offenders, almost always young
men who know one another (Levin & McDevitt, 2002), so peer group dynamics
can play an important part in motivating participation in these crimes (Byers
et aI., 1999; Franklin, 1998, 2000; McDevitt et al., 2002). As Kathleen Blee
(2007) notes, to outsiders the violence can appear “pointless, irration~ or the
product of -immaturity or personality disorders” but to group members may convey
“a sense of strength, inviolabiliry, purpose, and agency” (pp. 263-264) that promotes
group afflliation and solidarity. Offenders motivated by peer group concerns
act out of “the desire to feel closer to friends, to live up to friends’
expectations, and to prove toughness and [in the case of anti-gay crimes] heterosexuality
to friends” (Franklin, 2000, p. 347). As this emphasis on toughness and
408 CHAPTER 10
heterosexual masculinity suggests, Franklin (1998) noted that men are more
strongly tllotivated by this factor than are women, at least in the anti-gay context
(see also Chapter 12). The group-centered nature of some hate crnnes is illustrated
in the interviews Byers and his colleagues (1999) conducted with perpetrators
of hate crimes against the Amish: “When asked if a person were to clare
alone, subjects responded that the person would have to be ‘sick’ to do such a
thing” (p. 84). Another said, “It was a kind of male bonding …. It kind of drew
us all closer because we went out and did something” (p. 89).
Like thrill seekers, peer-motivated hate crime offenders exhibit little animosity
toward their victims’ groups, but also exhibit little respect for them (Byers et aI.,
1999; Franklin, 1998). However, unlike thrill seekers, they sometimes do acknowledge
that the victim was hanned, but tend to minimize their personal responsibility.
Instead, they portray themselves as having had little choice in the matter (Byers et al.,
1999; Franklin, 1998). For example, some of Byers and colleagues’ (1999) respondents
blamed their behavior on peer pressure or local nonns, giving explanations
such as “The harassment was ahnost conunon nature” and “[It] is because of the
way I was raised” (p. 92).
McDevitt and his colleagues (2002) note that many offenders who act on
the basis of peer group concerns may be unwilling participants in the crime:
They do not approve of violence (or, perhaps, even prejudice), but go along
with the group because they feel that if they do not they will lose the approval of
their friends. In many instances, unwilling participants do not actively take part in
the crime, but also do nothing to prevent it or stop it once it has begun and are
unwilling to provide infonnation to authorities afterwards (see also Byers et al.,
1999). Thrill-motivated hate crimes were the ones most commonly found in
Levin and McDevitt’s (2002) Boston sample, constituting 66 percent of tlie cases.
Normalization. Based on their interviews with participants in anti-Amish hate
crimes, Byers and Ctider (2002) liave suggested that one factor tliat facilitates, if
[
not motivates, such crimes is community acceptance of such behavior. That is,
the community in which the offenders live views such actions as nonnal behavior
and so do not strongly condenm it, try to prevent it, or punish it. Byers and
— Crider (2002) noted that “victimization of the Amish was sometimes considered
a rite of passage [in the local non-Amish community] … ‘ If people believe that
claping is hannless, there is a lower likelihood of intervention from parents,
teachers, or criminal justice officials” (pp. 133-134). This attitude was reflected
in comments made by Byers and Crider’s (2002) interviewees. For exalnple,
[Claping] is socially acceptable here …. [Interviewer: In this community?]
Extremely, in this community. They just pretty much shrug and it is “boys
growing up.” If I lived here and I had kids and they were 16 or 17, then it
would not surprise me if I just said it is just “kids growing up.” … I think that
most people would see it as something as I will turn my back to it and pretend
that it is not there. Like when people are watching t.v. and you see the starving
kids, everyone just turns the channel. It is the same thing …. If something bad
really did happen like say somebody got killed or whatever I would say that the
Although the exact number of lynchings of Blacks that
occurred in the United States in the 19th and
20th centuries remains unknown, estimates range
from 2,500 to over 3,400 (see Leader, Mullen, &
Abrams, 2007). As we have seen throughout this text,
history is replete with examples of such violence
against outgroups and, at one point in U.S. history,
lynchings were, if not sanctioned, certainly
overlooked. Consider the now well-known lynchings
of two Black teenagers that occurred in Marion,
Indiana in 1930. These young men were accused of
raping a White woman and killing her boyfriend.
Although the teens had confessed and were awaiting
trial, rumors spread that they would be let off easy;
in response, family members and their supporters
stormed the jail. The teens were, one at a time,
dragged to the courthouse square and lynched.
Evidence that such actions were normalized at that
time in U.S. history comes from James Madison’s
(2001) analysis of the crowd’s response. He writes:
[Police] reported a scene of peace and remarkable
good humor. One of the mob who had helped
with the rope went with his young wife to a
nearby restaurant for a late dinner. But the crowds
did not disperse. People milled around through the
night. A woman nursed her baby, Fathers held up
older children to see the two bodies, .. Newcomers
kept arriving, including the youth group from
Antioch Methodist Church”. Cars [were] parked at
all angles, jammed together like matchsticks
around the Courthouse Square, . ,Souvenir collectors
cut pieces of clothing from the two bodies and
bark from the lynching tree (p. 10).
These actions continued through the night.
In fact, although the second lynching occurred at
FROM PREJUDICE TO DISCRIMINATION 409
10:30 p.m., it was 5:45 the next morning before the
crowd would allow the bodies to be cut down,
What conditions lead people to participate in such
savage cruelty? To examine this question, Tirza Leader,
Brian Mullen, and Dominic Abrams (2007) studied
reports of 300 lynchings that occurred in the United
States between 1899 and 1946. Results showed that as
crowd size increased, lynchings became more savage,
perhaps because bystanders felt anonymous in the
larger groups, /ls the crowd size increased, norms
seemed to shift, allowing the display of antisocial
behavior. Leader and her colleagues also found that
atrocities were greater when the victim’s alleged crime
was more severe, perhaps providing further
justification for the group’s cruelty, The researchers
also considered the possibility that increases in the
Black population fed White people’s desire to protect
their territory; however, they found no evidence
that the severity of the lynchings increased in those
areas with higher proportions of Blacks, suggesting
that people were not responding to population
changes.
As Leader and colleagues (2007) note, lynching in
the United States has come to define a certain time
in the past and a culture that was less sophisticated
and less educated” (p. 1340). Even at the time, not
everyone accepted lynching as normal behavior. As
Madison (2001) notes in his report on the Marion
lynching, the teens’ jailors tried desperately to prevent
the lynchings and, although the details are not
certain, a voice of reason, shouting from the crowd,
was able to stop the lynching of a third teen.
Moreover, some people in the crowd were horrified
by what they witnessed and some were physically ill.
Then, as now, acceptance of violent group behavior is
far from inevitable.
community would be, “That is a shame and all that,” but then the entire town
would be like “Oh well.” I really think it would be that way (pp. 136-137).
Claping is so acceptable in the community where Byers and Crider conducted
their interviews that one of the researchers was invited out on a claping
expedition (he declined the invitation). Box 10.6 describes the normalization of a
more severe hate crime, the lynching blacks in 19th and 20th centuries.
Inhibiting Factors. It is often as informative to understand why people do n~
engage in a behavior as why they do engage in it. A unique aspect of Franklin’~
(2000) study is that she asked the nonoffenders in her sample why they “have
410 CHAPTER 10
never harassed or beaten up a homosexual” (p. 353). She identified a number of
reasons, although she did not report the number of people giving each type of
reason. Some people’s responses implied that they refrained from anti-gay behavior
because of a lack of opportunity, saying they avoid homosexuals or never see
homosexuals. Others reported a fear of negative consequences, such as getting
into trouble with authorities or peer disapproval. Some respondents cited a belief
in nonviolence or other religious or moral beliefS. A final reason was knowing
someone who was homosexual. Franklin also noted that these restraining beliefs
may be fairly weak and so may easily break down under pressure. For example,
38 percent of her respondents who had never engaged in anti-gay behavior
“reported some likelihood to verbally or physically assault a homosexual who
flirted with or propositioned them” (Franklin, 2000, p. 353).
Effects on Victims
Surveys conducted in different parts of the United States over a period of 10 years
have provided an unusually consistent set of results regarding the effects hate
crimes have on their victims: Hate crune victims suffer more severe psychological
consequences from their victimization and these negative effects last longer
compared to victims of similar crimes that were not motivated by bias (Ehrlich,
1999; Herek, Gillis, & Cogan, 1999; McDevitt, Balboni, Garcia, & Gu, 2001).
Moreover, hate crime victims who experienced severe violence, such as sexual assault,
reported greater psychological distress than those who were threatened but
did not experience physical violence (Rose & Mechanic, 2002). Table 10.3 lists
SOille of the outcomes that hate Crlll1e victims experience to a greater extent than
victims of other crinles. In addition, cOlllpared to victims of nonbias crimes, hate
crune victims report feeling less control over their lives. One factor that helps
crime victims deal psychologically with their victimization is the feeling that they
can control what happens to them and, as a result, do things that will prevent
TAB L E 10.3 Effects of Hate Crimes on Victims
Compared to victims of similar crimes that were not motivated by bias, hate crime
victims experience more:
• Nervousness, anxiety, depression, and stress
• Intrusive thoughts about the crime
• Trouble concentrating or working
• Anger and a desire to retaliate
• Feelings of being exhausted and weak for no reason
• Fear of future trouble in life
• Distrust of people
• Fear of crime and feelings of personal vulnerability
• Difficulty coping with the effects of victimization
• Difficulty in relationship with spouse or significant other
Ehrlich, Larcom, and Purvis (1995); Herek, Gillis, and Cogan (1999); and McDevitt, Balboni, Garcia, and Gu (2001).
FROM PREJUDICE TO DISCRIMINATION 411
them from being victimized again (Davis, Taylor, & Titus, 1997). However, hate
crime victims tend to be chosen at random and so believe that there is nothing they
can do to avoid becoming a victim again: “Victims are aware that their overt actions
did nothing to precipitate their victimization; being the ‘wrong person,’ at the wrong
time and place, qualifies the bias victim [to become a victim]. Therefore, if the impetus
for victimization is something that is outside the bias victim’s control before the
incident, it is reasonable that there would be little the victim would do differently
subsequent to the incident” (McDevitt et al., 2001, p. 711). These feelings oflack
of control exacerbate the negative psychological consequences of having been a
crime victim.
Perhaps because of this, hate crime victims do not always immediately label
their experience as a hate crime, even if authorities have made that determination.
For example, Kathleen Blee (2007) interviewed members of an Islamic community
whose mosque had been destroyed by arson. Despite clear evidence that the
arson was not a random act of violence, including a note left at the arson site explicitly
stating “we hate you and we want you to leave” (p. 265), members resisted
that interpretation, at least initially. Interestingly, those with higher status positions
or who had been in the community a longer period of time were particularly
likely to dismiss the possibility that the crime was directed at Muslims. Similarly,
Blee interviewed members of a Jewish community which experienced the firebombing
of its Holocaust museum, Some initially interpreted the event as due to
anti-Semitic violence but later changed their minds; others initially looked for
other interpretations but, over time, came to believe it was a hate crime. Hence,
“[w]hether and how victims interpret violent acts as hate violence are the products
of collective and individual processes of interpretation” (Blee, 2007, p. 266).
A special characteristic of hate crimes is what McDevitt and his COlleague]
(2001; see also Ehrlich, 1999) call secondary victimization: A hate crime has
psychological effects not only on the victim but also on members of the victim’s
group. These secondary victims experience, at least temporarily, heightened anxiety
over the possibility of becoming victims themselves. Secondary victimization
is a major goal of defense-motivated hate crimes and is often a secondary goal of
others (McDevitt et al., 2001). fu McDevitt and his colleagues (2001) note,
“a cross burning not only affects the immediate family [that was victimized],
but any African American who beconles aware of the incident” (p. 698). As
the Chinese military strategist Sun T’zu said over 2,000 years ago: “Kill one,
frighten 10,000.” There are few data on the extent of secondary victimization
in hate crimes, but Howard Ehrlich (1999) reported that surveys of college students
following on-campus hate crimes have found that about two-thirds of
other members of the victim’s group experience fear of becoming victims themselves.
Paul Iganski (2007) interviewed people who, although not victims of hate
crimes themselves, saw hate-related violence in their jobs as district attorneys or
police officers. These individuals reported that hate crimes had many consequences
for the communities in which they took place, including increased anxiety,
the potential for more crime due to retaliation, and ripple effects that led
some group members to respond as if they had themselves been victimized.
Thus, hate crimes victimize not just individuals but entire social groups.
412 CHAPTER 10
SUMMARY
Discrimination consists of treating people differently, and usually unfairly, based
solely or priluarily on their membership in a social group. Discrimination is therefore
a matter of behavior (including verbal and nonverbal behavior), whereas prejudice
is an attitude that can motivate discriminatory behavior. Discrimination can
take any of three forms. Blatant discrimination is intentional and obvious. Subtle
discrimination is less visible and obvious than blatant discrimination, is often unintentional,
and derives from people having internalized discriminatory customs and
social norms. Covert discrimination is hidden but intentional, and often motivated
by malice.
Although prejudice can motivate discrimination, not all prejudiced people
discriminate when they have the opportunity and nonprejudiced people can
discriminate without intending to. A number of factors influence the relationship
between prejudice and discrimination. Prejudice is more likely to manifest
itself in discrimination when the target of the discrimination matches the prejudiced
individual’s personal stereotype of the outgroup. Implicit prejudice is
most likely to result in automatic, uncontrollable behaviors, whereas explicit
prejudice is most likely to affect controllable behaviors. Finally, people are
more likely to act on their prejudices when they believe that other people
agree with them.
Because of the egalitarian nonn that exists in modern society, most people
are motivated to control any prejudice they feel and to avoid discriminatory
behavior. However, regressive prejudice occurs when people lose control over
their prejudiced responses and act in a discriminatory manner. Thus, people can
exhibit nonverbal indicators of prejudice, which are usually not under voluntaty
control, while trying to appear unprejudiced through their controllable behaviors.
Just as high cognitive demands can lead people to apply stereotypes,
such demands can let prejudiced behaviors “leak out” by undermining control.
Alcohol consumption and strong emotions, such as anger, can also reduce control
over behavior. Because people are less motivated to conlply with social
norms when other people cannot identifY them, anonymity facilitates discriminatory
behavior, as does the actual or implied approval of authority figures.
When people can otherwise justifY their actions, such as when they decide
the cost of helping another person is too high, discrimination is more likely.
Seeing other people act in a prejudiced manner can also disinhibit prejudice.
Finally, if individuals believe that they have established their credentials as unprejudiced
people, they may let their control lapse and act in a discriminatory
manner.
Research on discrimination in organizations indicates that Black job applicants
who submit resumes are less likely to be called for interviews than equally
qualified White applicants and receive lower ratings on interview performance.
As a result, they are less likely to be hired. In contrast, there currently seems to
be little gender discrimination in hiring—-unless the applicant is pregnant.
Once on the job, there seems to be little gender bias in performance evaluations;
FROM PREJUDICE TO DISCRIMINATION 413
however, Black workers receive lower perfonnance evaluations than White
workers, especially from White supervisors. Even when numerical ratings for
Black and White employees are identical, White employees tend to get more
positive narrative comments. Black employees perfonn less well on objective
measures of job perfonnance, so evaluations might simply be reflecting that difference,
However, the lower objective perfonnance might itself be a result of
prejudice, reflecting lost opportunities, such as for additional training, and lower
morale caused by prejudice and discrimination.
Even when women and members of minority groups receive the same performance
evaluations as men and Whites, they are less likely to be promoted,
are promoted more slowly, and are more likely to end their careers at a lower
organizational level. Also, ethnic minorities can be “tracked” into certain job
categories, such as affirmative action officer, that offer fewer opportunities for
advancement. In addition, the pay differential between Black and White workers
increases as level of authority increases. Black workers may experience slower
promotions because of the sticky floor effect-Black managers tend to supervise
Black workers who are disproportionately found at lower organizational levels,
are tracked into jobs with little promotion potential, and have fewer influential
mentors to help them in their careers.
A number of individual-level processes contribute to discrimination in
organizations. The stereotype fit hypothesis holds that women and members
of minority groups are underrepresented in managerial positions relative to
White men because the White male stereotype matches the stereotype of the
effective manager whereas the female and minority stereotypes do not, As a
result, women and minority group members are perceived as less qualified despite
their objective qualifications. Similar processes also can operate for lower
level jobs: The generally negative stereotypes of minority groups contradict the
“good worker” stereotype. Wonlen and members of minority groups also may
be excluded from prestigious jobs because, as groups, they garner less respect
than White men. The flnding that women and members of minority groups
are less likely to be promoted even when they receive the same perfonnance
evaluations as men and Whites may be a result of the shifting standards effect:
Because evaluators have lower expectations for WOlnen’s and minority groups’
perfonnance relative to men’s and Whites’, the same subjective rating translates
into a lower rating on an objective common scale that takes the race-based evaluations
into account. Rewards such as promotions are based on the common
scale.
Contemporary prejudice can lead prejudiced decision makers to put more
weight on the negative aspects of a minority group member’s qualifications
when both positive and negative infonnation is available, leading to an adverse
decision that can be justified by the negative infonnation. Employers also may
use business-related justifications, such as maintaining work group harmony and
customers’ prejudices, as justifications for not hiring minority workers, Finally,
people tend to comply with what they perceive to be the requirements of organizational
nonns and authority figures. Thus, if they perceive the organizational
414 CHAPTER 10
nonn as calling for discrimination or perceive that authority figures prefer to
have as few minority workers as possible, even low prejudiced people may discriminate
to comply with those demands.
Hate crimes are criminal offenses in which evidence shows the victims were
chosen because of their group membership. Hate crime offenders are primarily
young men. Several motives exist for hate crime. The most common seems to
be thrill seeking: People are bored and see picking on or assaulting a member of
an outgroup as a way of getting some excitement. They often have no strong
animosity toward their victlins’ groups; they choose as targets members of groups
they believe are unlikely to fight back or to report the crime. Tliey often justify
their actions by minimizing their impact on the victim or portraying their actions
as harmless fun. Defensive hate crimes are designed to drive outgroup members
from ingroup “territory” and to send a general message to other melnbers of the
victim’s group to stay away. Retaliation crimes are carried out in response to an
actual or rumored hate crune against a member of the offender’s group. Mission
hate crimes occur because of a commitment to a bigoted ideology and to rid the
world of a perceived evil. Peer group dynamics contribute to hate crimes because
offenders are often trying to impress members of the peer group, are going along
with what they see as the group nonn, or have succumbed to group pressure to
participate. Community norms also can facilitate hate crimes by viewing them as
normal behavior and refraining from disapproving of or punishing them. Factors
that can inhibit people from conunitting hate crimes include lack of opportunity,
fear of negative consequences, religious or moral beliefs, or knowing a member of
the targeted group.
Hate crimes generally have more severe psychological consequences for their
victinlS than do nonbias-motivated crimes, and those effects last longer. The
effects may be more severe and longer lasting because hate crime victims feel
that they cannot do anything to avoid being victimized in the future. However,
not all hate crime victims see the experience in the same way and how they view
the experience can change over time. Hate crimes also result in secondary victimization:
A hate crime has negative psychological effects not only on the victim, but
also on other members of the victim’s group, who experience heightened anxiety
over the possibility of becoming victims themselves.
SUGGESTED READINGS
The Relation of Prejudice to Discrimination
Crandall, C. S., Eshleman, A., & O’Brien, L. (2002). Social norms and the expression
and suppression of prejudice: The struggle for internalization. Journal 4 Personality
and Social Psychology, 82, 359-378.
Crandall and his colleagues present a set of studies on the role of social nonns in
prejudice and discrimination.
FROM PREJUDICE TO DISCRIMINATION 415
Regressive Prejudice
Crandall, C. S., & Eshleman, A. (2003). Ajustification-suppression model of the expression
and experience of prejudice. Psychological Bulletin, 129,414-446.
This article includes a comprehensive review of factors that act as releasers of
regressive prejudice, which Crandall and Eshleman refer to as justifications.
Reactions to Having Acted Prejudiced
Czopp, A. M., & Monteith, M.]. (2003). Confronting prejudice (literally): Reactions to
confrontations of racial or gender bias, Personality aHd Social Psychology Bulletin, 29,
532-544.
Devine, P. G., Monteith, M. J., Zuwerink, J. R., & Elliot, A. J. (1991). Prejudice with
and without compunction, Journal of Personality and Social Psychology, 60,817-830.
Devine and colleagues’ article describes some of the initial theory and research
on how people with unprejudiced self-concepts react to having acted in
a prejudiced manner. Czopp and Monteith examine the issue in the context
of responses to having acted in a prejudiced manner toward different groups.
Organizational Discrimination
Roberson, L., & Block, C. J. (2001). Racioethnicity and job perfonnance: A review and
critique of theoretical perspectives on the causes of group differences. Research in
Otgal1izational Behavior, 23, 247-325.
Roberson and Block discuss four models of factors that affect minority group members’
work perfonnance. Although the discussion focuses on work perfonnance, the
principles they discuss generalize to other processes, such as hiring and promotion.
Hate Crimes
Blee, K. (2007). The microdynamics of hate violence: interpretive analysis and implications
for responses. American Behavioral Scie11tist, 51, 258-270.
Ehrlich, H. J. (1999). Campus ethnoviolence. In F. L. Pincus & H. J. Ehrlich (Eds.), Race
and ethnic conflict: Contending views 011 prejudice, discrimination, and ethl1ovioience (2nd ed.,
pp. 277-290). Boulder, CO: Westview.
Levin, J., & McDevitt, J. (2002). Hate cdmes revisited: Amen’ca’s war ON those who are different.
Boulder, CO: Westview.
McDevitt,]., Levin,]., & Bennett, S. (2002). Hate crime offenders: An expanded typology.
Journal of Social Issues, 58,303-317.
Blee interviewed victims of hate crimes and discusses the factors the influence
their interpretation of the event. McDevitt and his colleagues (2002) present an
overview of Levin and McDevitt’s· (2002) typology of hate crime offenders.
Levin and McDevitt’s (2002) discuss the typology in more detail and provides a
number of case descriptions for each type. They also discuss issues involved in
policing, public policy, and prevention, among others. Ehrlich (1999) discusses
the problem of hate crime on college campuses, usually considered to be
bastions of tolerance (see also Levin and McDevitt’s Chapter 9).
416 CHAPTER 10
blatant discrimination
covert discrimination
discrimination
hate crimes
interpersonal
discrimination
KEY TERMS
organizational
discrimination
regressive racism
(prejudice)
secondary victimization
shifting standards model
social noons
stereotype fit hypothesis
subtle discrimination
QUESTIONS FOR REVIEW AND DISCUSSION
1. Define discrimination. How does discrimination differ from prejudice? How
are the two concepts similar?
2. Define the three fonus discrimination can take and give an example of each.
Review the types of contemporary prejudice we discussed in Chapter 6.
What forms of discrimination do you think those types of prejudice likely
result in?
3. Describe the factors that influence the relationship between prejudice and
discrimination. That is, under what conditions is prejudice most likely to
result in discrimination?
4. What are social nonns? How are they related to prejudice and discrimination?
What experiences have you had with social norms and prejudice and
discrimination?
5. What is regressive prejudice? Describe the factors that can precipitate it.
Have you observed any instances of regressive prejudice? If so, describe them
and explain what factors led to the release of discriminatory behavior in
those cases.
6. What is an employment audit? Do you think that employment audits are
effective tools for studying discrimination in hiring? Why or why not?
7. What has research discovered about race and gender discrimination in
hiring? What has research discovered about race and gender discrimination
in perfonnance evaluation?
8. Researchers have found that Black workers usually get lower scores on
objective measures of job perfonnance than do White workers. What is the
relevance of this finding for interpreting race differences in supervisor
evaluations, which generally have a strong subjective element?
9. What has research discovered about race and gender discrimination in
promotions? What organizational factors might contribute to these
differences?
10. Describe the costs and benefits for men who are in a female-dominated
profession.
FROM PREJUDICE TO DISCRIMINATION 417
11. What is the stereotype fit hypothesis? How does it explain race and gender
differences in hiring, perfonnance evaluation, and promotion?
12. Describe how differences in the amount of respect that different social
groups receive are related to organizational discrimination.
13. What is the shifting standards effect? How does it explain race and gender
differences in hiring, perfonnance evaluation, and promotion?
14. Explain the role contemporary prejudice plays in orgartizational
discrimination.
15. Explain the role confonnity to perceived nouns plays in organizational
discrimination.
16. If an employer believes that his White customers do not want to interact
with people of other ethnicities, does that justify his decision not to hire
non-White workers? Explain your reasoning.
17. Rather than attributing the differential outcomes miuority and female
workers experience in organizations to intentional discrimination, Smith and
Elliott (2002) wrote that ‘We believe that something more subtle and
profound occurs in the process of doing ‘business as usual’-lnere
maintenance of the status quo is more than enough to perpetuate ‘”
stratification” (p. 274). Do you agree or disagree? Explain the reasons for
your position.
18. What are hate crones?
19. Describe the characteristics of hate crime offenders.
20. Explain how thrill seeking can motivate hate crimes. Who do thrill seekers
choose as victims? How do they justify their behavior?
21. Some researchers believe that thrill-seeking hate crime offenders feel little
animosity toward their victims or their groups. Do you agree or disagree?
Explain the reasons for your position.
22. Explain territorial defense as a motivation for hate crimes.
23. Explain retaliation as a motivation for hate crimes.
24. What are mission-motivated hate crimes? Why do you think they are so
rare?
25. Explain the role that peer group dynamics play in hate crimes.
26. Explain how community attitudes can affect the occurrence of hate crimes.
27. In what ways do the psychological consequences differ for the victims of
hate crimes and those of crimes not motivated by bias? What causes these
differences?
28. Explain the concept of secondary victimization.

From prejudice to discrimination

Order Description

1.Watch this clip:

Are you surprised and/or upset when the last word comes out? Have you in some way influenced by stereotypes?

2. Watch this video clip: Wafa was attacking the extremists and their destructive acts.

And see this picture

In download

Question: Why do you think the terrorists do what they do? Is the reason similar to members of KKK lynching Blacks? You have to show your knowledge to receive full credit.

must use book in download as 1 reference

Chapter 10
*
From Prejudice to
Discrimination
Rush hour on board a bus or train …. The flood of incoming
passengers begins to solidifY …. I am sitting next to a window,
my eyes half-closed …. Dressed conservatively in a tweed
jacket and tastefully bold tie, I am an unremarkable Ulan, .. , as
unnoticed as any other COl1lllluter. Except for Oile thing: amid
the growing crush, the seat beside me remains empty. At stop
after stop, … a succession of seemingly random individual
decisions coalesces into a glaring pattern of unoccupied spaces
next to black males-inc1uding me. Soon the seats beside us are
the only ones left. Other passengers remain standing, leaving
only these seemingly quarantined seats.
-BRUCE JACOBS (1999, PP. 15-16)
Chapter Outline
What Is Discrimination?
Forms of Discrimination
Interpersonal Discrimination
The Relation between Prejudice
and Discrimination
Regressive Prejudice
Reactions to Having Acted in a
Prejudiced Manner
Discrimination in Organizations
Organizational Research
Individuals in Organizations
369
Hate Crimes
Hate Crime Offenders
Motivations for Hate Crimes
Effects on Victims
Summary
Suggested Readings
Key Terms
Questions for Review and
Discussion
370 CHAPTER 10
Prejudice, which has been the main focus up to this point in the book, is an
attitude; it deals with how people think and feel about members of other
groups. Discrimination, in contrast, is behavior; it deals with how people act to- I ward members of other groups. Discrimination consists of behaving differently
l,~oward people based solely or prirnarily on their lllembership in a social group.
‘The tenn is usually used to refer to acting in an unfair or demeaning manner, but it
can also refer to giving someone an undeserved advantage. Bruce Jacobs’s (1999)
experience cited above illustrates one aspect of the ongoing problem of discrimination
in modem fullerican society: Many White people avoid contact with members
of minority groups, even when the avoidant behaviors cause inconvenience for
themselves. Although discrimination against mel11bers of minority groups is not always
as blatant as it was prior to the civil rights l110vement of the 1960s, it still occurs.
As we discuss the research on discriminatory behavior you might be tempted to say,
“Well, it’s obvious: Prejudiced people discriminate,” but, as we will see, there is only
a moderate correlation between people’s degree of prejudice and their propensity to
engage in discriminatory behavior. That is, not all prejudiced people discriminate
every time they have an opportunity, and some nonprejudiced people discriminate
in sallIe situations. In presenting what psychologists know about discrimination,
we first look at the nature of discrimination and then focus on the two forms most
often studied by psychologists, interpersonal discrimination and organizational
discrimination. We conclude with a brief examination of the most severe fonn of
discrinlination, hate crimes.
While reading this chapter, bear in mind a point we made in Chapter 7 concerning
prejudice: some fonns, such as racial prejudice, are socially proscribed
whereas other forms, such as anti-gay prejudice and prejudice against physically
unattractive people, are more socially pernlissible. The same principle applies to
discrimination. For example, David Schneider (2004) noted the following forms
of socially acceptable discrimination: “Most of us would fight having a group
honle for convicted rapists placed next door to our honle, no matter how ‘cleaned
up’ or ‘ex-‘ the rapists claimed to be. Most church groups do not invite homeless
people to share their potluck dinners …. Many women prefer their gynecologists
to be female, and most males discriminate against males as their sexual partners.
That’s the kind of discrimination we all know and generally approve. So it is not
the fact of discrimination that is controversial, but its application to specific groups”
(p. 291). Consider, for example, college students’ ratings of the acceptability of
prejudice and discrimination directed at various groups (Crandall, Eschleman, &
O’Brien, 2002). Table 10.1 shows the groups for which prejudice and discrimination
had the highest and lowest approval ratings.
WHAT IS DISCRIMINATION?
Discrimination can manifest itself in many ways, both verbally and behaviorally,
and in many settings. For example, Jennifer Livengood and Monika Stodolsak
(2004) interviewed American Muslims about the discrimination they experienced
in post-September 11. Common experiences included hearing racist epithets,
FROM PREJUDICE TO DISCRIMINATION 371
TAB L E 10.1 Socially Approved and Disapproved Prejudices
Percent Percent
Approved Prejudices Approving Disapproved Prejudices Approving
Rapists 98 Mentally retarded people 3
Child abusers 98 Native Americans 6
Child molesters 97 Black Americans 6
Wife beaters 97 Jews 6
Terrorists 95 Catholics 6
Racists 92 Whites 7
Members of the Ku Klux Klan 91 Hispanics 7
Drunk drivers 91 Asian Americans 7
Members of the American 90 Canadians 7
Nazi Party
Pregnant women who drink 89 Ugly people 10
alcohol
Men who refuse to pay child 89 Interracial 11
support couples
Negligent parents 86 People with AIDS 11
People who cheat on their 82 Fat people 11
spouses
SOURCE: Crandall, Eshleman, and O’Brien (2002, Table 1, p. 362).
receiving hostile and disapproving looks, being subjected to obscene gestures,
avoidance by others, and, in some cases, being personally threatened, having
property vandalized, or being physically attacked. As one respondent noted,
“They scream at you and … say really mean things … like ‘you crazy terrorist,
you go back home. You rag head [referring to Muslim head coverings] and
you are so backward.'” (p. 192). Respondents reported experiencing ‘hate
stares’ as they went about their everyday activities such as shopping, driving,
or walking down the street. As a result, many changed their routines to reduce
contact with mainstream Americans and chose to remain at home whenever
possible. Many reported being fearful for their personal safety and for the
safety of their friends and loved ones. In this chapter, we being by looking at
the forms discrimination can take: blatant, subtle, or covert. We then focus on
the types of discrimination most often studied by psychologists: interpersonal
and organizational.
Forms of Discrimination
Given the many forms that discrimination can take, it is useful to have a system
for classifYing fonns of discrimination to show how they relate to one another.
Nijole Benokraitis and Joe Feagin (1995) developed one such system, based on
372 CHAPTER ,0
three forms of discrimination-blatant, subtle, and covert. Although Benokraitis
and Feagin developed their typology in the context of sexism, it applies to other
forms of discrinlination as well. In Chapter 1, we discussed levels of discrimination,
including interpersonal, organization, cultural, and institutional. Note that
blatant, subtle, and covert discrimination can be found at all four of those levels
of discrimination.
U
Blatant Discrimination. Blatant discrimination consists of “unequal and
harmful treatment … that is typically intentional, quite visible, and easily documented”
(Benokraitis & Feagin, 1995, p. 39). Extreme cases of blatant discrimination,
-such as the murder of Matthew Shepard, a gay man who was beaten and left to die
on a fence in Wyoming (Laramie Project Archives, 2002), and the murder of James
Byrd, a Black man who was chained to the back of a truck and dragged along a
road in Texas until he died (Texas NAACP, 1999), often receive national attention.
However, blatant discrimination occurs in everyday contexts as well. For example,
a Black college student interviewed by Janet Swim and her colleagues (Swim,
Hyers, Cohen, Fitzgerald, & Bylsma, 2003) told of how “a man at a party addressed
her by a racist label and ordered her to perform a menial task” (p. 52). To cite
another example, it is not unCOlumon for Black women and Inen to experience
blatant discrimination while shopping, including being followed in the store or otherwise
being treated as though they are likely to steal merchandise. One study
found that 35 percent of Blacks experienced negative treatment while shopping in
predominandy White communities, cOlupared to 10 percent reporting similar treatment
in their own community (Lee, 2000).
Some fOlTl1s of blatant discrimination against some groups are illegal and
generally condemned, such as racial discril1unation at work, in school, and in
public accommodations” However, other forms, such as discrimination against
lesbians and gay men, are often legal and are accepted as “normal” by many people.
Consider, for example, a Web site on which visitors can find a picture of
Matthew Shepard burning in hell, along with a record of how many days he
has been there. Visitors are invited to click on his picture and hear him scream
as he endures the flames of hell. However, examples of blatant discrimination
against other groups are not difficult to find, especially if they are supposedly
humorous, At the tUlle of this writing, the popular Web site You- Tube featured
of series of videos (many rated 5-star) depicting the “Anlazing Racist” discriminating
against Mexicans, Asians, and Muslims. Although these clips are categorized
as comedy, the actions depicted are based on negative stereotypes such as
the belief that all Asians eat dogs, that all Muslims have body odor, and that all
Mexicans are in United States illegally.

Subtle Discrimination. Subtle discrimination consists of “unequal and
harmful treatment .” that is typically less visible and obvious than blatant discrimination.
It is often unnoticed because people have internalized subde [discrinlinatory]
behaviors as ‘normal,’ ‘natural,’ or customary” (Benokraitis &
Feagin, 1995, p. 41). Subtle discrimination tends to be harder to document
FROM PREJUDICE TO DISCRIMINATION 373
than blatant discrimination, but it can often be done. Unlike the other fOnTIS
of discrimination, which are often intentional, subtle discrimination is often
unintentional. For example, Claude Steele (1992) related a story told by a friend
of his who
noticed over many visits [to her son’s third-grade classroom] that the extraordinary
art work of a small black boy named Jerome was ignored–or, more accurately
perhaps, its significance [as a sign of artistic talent] was ignored . .&
genuine art talent has a way of doing-even in the third grade-his stood out.
Yet the teacher seemed hardly to notice. Moreover, Jerome’s reputation, as it
was passed along from one grade to the next, included only the slightest mention
of his talent, … Had Jerome had a reading problem, which fits [American
society’s stereotypic image of black children], it might have been accepted as
characteristic of him more readily than his extraordinary art work, which contradicts
[that image] (p. 72).
Thus, subtle prejudice had the effect of directing teachers’ attention away
from Jerome’s artistic talent because, by society’s definition, Black children
do not have that particular talent. Subtle prejudice also can be manifested in
everyday speech, as described in Box 10.1.
Covert Discrimination. Covert discrimination consists of “unequal and
hannful treatment … that is hidden, pUJposeful, and, ofren, maliciously motivated ….
[It is] behavior that consciously attempts to ensure … failure, as in hiring or other
employment situations” (Benokraitis & Feagin, 1995, p. 42). Covert discrimination
tends to be very difficult to document. Examples in the employment context
include tokenism, hiring one or a few members of a group as evidence that an
organization does not discriminate; containment, restricting members of a group
to a limited number of job categories; and sabotage, arranging for members of a
group to fail, such as by assigning them low volume sales territories but setting their
sales quotas at levels similar to those of salespeople with better territories
(Benokraitis & Feagin, 1995).
Covert discrimination is comnlon outside of the workplace as well, including
the housing market. In the United States, housing discrimination is prohibited by
the Fair Housing Act of 1988; however, such unfair practices are difficult to document.
Landlords, for example, may simply tell prospective renters that an available
unit was just rented. Similarly, real estate agents can subdy encourage buyers
to look in certain areas and to avoid others. Adrian Carpusor and William Loges
(2006) studied whether landlords discriminated against prospective renters based
on their ethnicity. Via email, they contacted landlords in Los Angeles County,
California, about the availability of rental property. The emails were supposedly
sent by prospective renters whose names varied by ethnicity: Patrick McDougal
(White), Tyrell Jackson (Mrican American), and Said Al-Ralnnan (Arab). The
applicant with the White-sounding name received more replies (89 percent)
than did the applicants with Black (56 percent) or Arab (66 percent) sounding
names. Moreover, only 11 percent of the replies to the White applicant were
negative, compared to 44 percent for the African American and 34 percent for
374 CHAPTER 10
Some of the most common examples of subtle prejudice
can be found in everyday speech. Examples include:
home at Monticello, Virginia. He noted that the
tour guide always referred to Jefferson in the
active voice but always referred to the work of
Jefferson’s slaves in the passive voice. For
example, while describing a set of interior doors
with a complex operating mechanism that had
required no repair in the 166 years since
Jefferson’s slaves had built and installed them, the
guide said, “Mr. Jefferson deSigned these doors”
(p. 147). In contrast, the guide said, “These doors
were installed originally in 1809” (p. 147, italics in
original), making the enslaved carpenters whose
skilled work had created the remarkable doors
effectively disappear from history.
•
•
•
Hostile humor calls attention to the negative
stereotypes associated with outgroups, such as
low intelligence, selfishness, stinginess, and
alleged bad habits such as drunkenness (Ruscher,
2001). People are often tolerant of outgroup
disparagement that is presented as a joke because
the humor context implies that listeners should
not take the speaker’s remarks seriously (Ford &
Ferguson, 2004). Humor can therefore also
function as a justification for derogatory speech.
How often have you heard someone who has just
been confronted for making a prejudiced remark
reply with “Hey, it was only a joke”?
Patronizing speech implies that outgroups are less
competent than members of the speaker’s
ingroup. Patronizing speech can take two forms
(Ruscher, 2001). Baby talk is often directed at the
older adult, implying that they have regressed
cognitively to the level of infants (see Chapter 13).
Although baby talk is usually well intentionedspeakers
view it as comforting-recipients find it
insulting. Controlling talk is directed at members
of groups the speaker views as having lower social
status and “functions to keep low~status indivi~
duals ‘in their place'” by controlling the direction
of the conversation (Ruscher, 2001, p. 88). For
example, when talking to women, men are more
likely to interrupt and to give commands than
when talking to other men (Caplan, 1994).
Vanishing uses linguistic devices to make out~
groups disappear. A common way of doing so is to
replace the active voice of a verb (for example,
“Bill hit the ball”) with the passive voice (“The ball
was hit”). As Thomas Greenfield (1975) noted, this
rhetorical device makes “the creator or instigator
of action totally disappear from a reader’s [or
listener’s} perception” (p. 146; see also Chapter 3).
For example, Greenfield (1975) recounted his ex~
periences on a guided tour of Thomas Jefferson’s
• Abnormalization involves describing outgroup
members in ways that emphasize their lack of
compliance with ingroup norms. For example,
Maykel Verkuyten (2001) conducted focus groups
in which Dutch residents of Rotterdam discussed
life in their neighborhoods. The participants were
chosen because they lived in areas in which large
numbers of immigrants also lived. She found that
when immigrants were mentioned, participants
almost always described them in ways that
emphasized their differences from the Dutch norm
often using extreme examples to illustrate a point:
For example, one participant supported his view
that immigrants from India were ignorant and
crazy by citing a case of a family that had built a
cooking fire on the floor of their apartment.
linguistic devices such as these serve several functions.
Hostile humor and abnormalization indicate ways in
which the ingroup is superior to the outgroup and
patronizing speech reinforces the higher status and
power of the ingroup (Ruscher, 2001). Vanishing
denigrates the skills of outgroup members and denies
their contributions to society, implying that progress
comes only through the efforts of the ingroup
(Greenfield, 1975). All forms of prejudiced speech serve
to draw clear boundaries that separate the “good”
ingroup from the “bad” outgroup.
the Arab. As Carpusor and Loges note, “long before an Mrican American man
gets a chance to show what he is capable of, discrimination tilts the scales
against him in ways he may not even be able to observe. Arab Americans face
similar obstacles” (p. 948).
FROM PREJUDICE TO DISCRIMINATION 375
INTERPERSONAL DISCRIMINATION
Interpersonal discrimination refers to individual, person-to-person diSCrini.U:’~–ination:
one person treating another unfairly because of the person’s group
membership. This section addresses three aspects of interpersonal discrimination.
The first deals with the relationship between prejudice and discrimination
and the circumstances under which prejudice is more or less likely to predict
discriminatory behavior. In Chapter 5, we discussed the motivation to control
prejudice. Here, we discuss factors that can undermine that motivation and allow
discriminatory behavior to occur. The third section examines how people
react to having acted in a prejudiced manner.
The Relation between Prejudice and Discrimination
& we noted earlier, when asked what causes interpersonal discrimination, most
people would probably reply “prejudice.” If that were the case. one would expect
to find a strong correlation between people’s prejudiced attitudes and their
propensity to engage in discriminatory behavior. However, when John Dovidio
and his colleagues (Dovidio. Brigham, Johnson, & Gaertner, 1996) reviewed the
results of 23 studies of the prejudice-discrimination relationship, they found an
average of correlation of r ;:::: .32. A meta-analysis of 60 studies found a similar
correlation of r ;:::: .36, but also found a stronger relationship between prejudice
and the intention to discrintinate, r = .45 (Schutz & Six, 1996). That is, the relationship
between prejudice and what people say they will do is stronger than the
relationship between prejudice and what they actually do. Even when researchers
find fairly high correlations between prejudice and discrimination, about
30 percent of the participants in those studies exhibited behavior that was inconsistent
with their attitudes (Duckitt, 1994). Although results such as these might
seem discouraging, they are, in fact, consistent with the results of research on the
relationship between attitudes and behavior in general. For example, Stephen
Kraus (1995) found an average correlation of r = .38 between attitudes and behavior
across a wide variety of domains. As William Graziano and has colleagues
note (Graziano, Bruce, Sheese, & Tobin, 2007), it is one thing to express negative
attitudes toward a social group, but another thing to actually discriminate
against someone. Hence, “the lake of prejudice is wider and deeper than the
smaller pool of discrimination, even if the forruer feeds the latter” (p. 572). As
it turns out, prejudice is related to discrimination (see Eagly & Chaiken, 1998),
but the relationship is not a silTIple one: A number of factors influence the
strength of this attitude-behavior relationship including personal stereotypes,
attitude-behavior correspondence, and perceived social support for the attitude.
Personal Stereotypes. Recall from Chapter 3 that stereotypes can exist at two
levels. Social stereotypes are characteristics of groups that most people in a society
agree on and personal stereotypes are individuals’ beliefs about group characteristics.
Personal stereotypes usually overlap with social stereotypes, but some
of their content may be different. Similarly, people recognize the social norms
376 CHAPTER 10
about prejudice toward particular groups, but may not personally agree with
those nonns. Replicating the results shown in Table 10.1, William Graziano
and colleagues (2007) found that people recognized that there are social norms
which detelnllne whether prejudice toward particular groups is acceptable or unacceptable.
However, they also reported that people personally found it even
more acceptable to discriminate against groups such as rapists and terroriststhose
groups for which society approves prejudice.
Prejudiced people also are lllore likely to discriminate against those outgroup
lllembers who fIt their personal stereotype than against those who do not. In a
study of this process, Shawna Ramsey and her colleagues (Ramsey, Lord,
Wallace, & Pugh, 1994) assessed college students’ personal stereotypes of former
mental patients. Several weeks later, the students read about a fonner mental
patient who either closely matched their personal stereotype or who was very
different from their stereotype hut matched other students’ stereotype. For
example, some students stereotyped fonner luental patients in tenns of schizophrenic
symptonls whereas others stereotyped them in tenus of depressive symptoms.
The students then chose from a list of activities those they would be
willing to engage in with the fonner luental patient, such as showing the person
the university library or taking the person to a party. For students who read
about a person who fit their personal stereotype, Ramsey and her colleagues
found a correlation of r = .43 between attitudes toward fonner mental patients
and the nunlber of activities in which they were willing to engage. For students
who read about a person who did not fit their personal stereotype, the correlation
was only r = -.07.
[
Attitude-behavior Correspondence. The tenn attitude~behavior correspondence
refers to how well an attitude matches, as it were, the behavior it is supposed to
be associated with. A higher degree of correspondence results in a higher
attitude-behavior correlation. One type of correspondence that is important
to the prejudice-discrimination relationship is the degree to which people can
control either their responses on the attitude measure or the behavior being
measured. Dovidio and his colleagues (1996) noted that explicit paperand-
pencil measures of prejudice assess controllable responses: People can think
about how they want to respond and carefully choose their responses, so social
desirability response bias can affect their answers; however, implicit attitude
measures assess relatively uncontrollable attitudes, so there is less opportunity
for social desirability response bias to affect responses (see Chapter 2).
Similarly, some behaviors, such as the content of what a person says, are controllable
and so can be affected by a social desirability response bias, whereas
other behaviors, such as many nonverbal behaviors, are more automatic and
difficult to control and so are less likely to be influenced by a social desirability
response bias. Drawing on the correspondence principle, Dovidio and his colleagues
proposed that scores on controllable measures of prejudice should be
correlated with controllable behaviors but not with automatic behaviors and
that scores on liuplicit measures of prejudice should be correlated with automatic
behaviors but not with controllable behaviors.
FROM PREJUDICE TO DISCRIMINATION 377
John Dovidio, Kerry Kawakami, and Samuel Gaertner (2002) tested this hypothesis
in a study in which White college students completed explicit and implicit
measures of prejudice. The participants then discus,.<;ed several race-neutral
topics, such as what personal belongings were most useful to bring to college,
with a Black student confederate who played the role of another research participant.
The interactions were videotaped, and raters later coded the White students’
behaviors for the frienclliness of their (automatic) nonverbal behaviors
and for the friendliness of what they said (controllable behavior). The researchers
found that, as they had expected, implicit prejudice correlated with nonverbal
friendliness but not with verbal friendliness and explicit prejudice correlated
with verbal friendliness but not with nonverbal friendliness. Similarly, Denise
Sekaquaptewa and her colleagnes (Sekaquaptewa, Espinoza, Thompson, Vargas, &
von Hippe!, 2003) found that an implicit measure of prejudice was related to
White students’ tendency to ask a Black student stereotypic rather than nonstereotypic
questions from a list provided by the researchers. Because students were
unaware that the questions varied in how stereo typic they were, this was considered
an implicit behavioral measure.
Attitude-behavior correspondence also can be assessed by the degree to
which people’s general attitudes toward a social group nlatch behaviors toward
a specific group member. For example, Michael Olson and Russell Fazio (2007)
studied whether White students’ implicit attitudes toward Blacks in general corresponded
to their attitudes toward a specific Black person. They reasoned that
when these attitudes were similar-that is, if a White student had positive attitudes
toward Blacks, in general, and also liked a specific Black person-the
speaker’s nonverbal responses should indicate little discomfort. However, if this
same White student disliked the specific Black person, it would be evident in
nonverbal behaviors, such as self-touching and speech interruptions. Results supported
their hypothesis; regardless of whether implicit attitudes toward Blacks
were positive or negative, when reactions to a specific Black person differed
from those implicit attitudes, nonverbal “leakage” occurred. Hence, even nonprejudiced
people can appear prejudiced if they dislike a specific person.
Perceived Social Support. The term perceived social support refers to the extentl
to which people believe that others share their attitudes and opinions. Generally,J
attitudes for which people perceive more social support are more closely related
to their behavior than attitudes for which they perceive less social support. For
example, Gretchen Sechrist and Charles Stangor (2001) pretested White college
students’ level of racial prejudice and selected those with high or low scores for
participation in a study conducted several weeks later. fu part of the study, participants
learned that either 81 percent of the students at their university agreed
with their racial attitudes (high social support condition) or that 19 percent agreed
with them O-ow social support condition). The researchers then used what is
known as the waiting room ploy to assess discrimination: The research apparatus
“malfunctioned” and the experimenter asked the participant to wait in the hallway,
where seven chairs were lined up in a row, with a female African American
student seated in the chair next to the door to the laboratory. Discrimination was
378 CHAPTER 10
assessed by how lllany chairs away from the Black student the participant sat.
Not surprisingly, the students who had scored low on prejudice sat closer to
the Black student than those who had scored high, an average of 2 seats versus
3.9 seats away. In addition, perceived social support affected the behavior of prejudiced
participants, with those who thought that most of their fellow students
also were prejudiced sitting farther from the Black woman than those who
thought that most of their peers were unprejudiced, an average of 4.3 seats versus
3.4 seats. Perceived social support had no effect on seating distance of the students
low on prejudice, Moreover, the prejudice-discrimination correlation was
larger for the students high in prejudice, r = .76, than for the students low in
prejudice, r = .33.
Regressive Prejudice
As we discussed in Chapter 5, under some conditions people are motivated to
avoid appearing prejudiced. However, doing so requires a great deal of mental
work: One must recognize that prejudice might be affecting one’s behavior and
then consciously change that behavior to produce a nonprejudiced response
(Crandall & Eshleman, 2003). Although, this process might become automatic
for some people (see Devine, 2002; Plant, 1998), others have to work at it.
The laborious nature of this process is demonstrated by research showing that
people experience a sense of relief when they are allowed to express prejudices
they have been controlling (Crandall & Eshleman, 2003). Thus, it is not surptising
that even people who score low on measures of prejudice report sometimes
_making prejudiced responses toward members of other groups (Monteith &
[
Vails, 1998; Vails, Ashbum-Nardo, & Monteith, 2002). That is, even people
who see themselves as unprejudiced and who score low on measures of prejudice
sometimes find thenlselves acting in prejudiced ways. Box 10.2 provides an example
of how an everyday event can trigger prejudiced reactions even in unprej-
Cldiced people. Such expressions of prejudice by people who are otherwise low
in prejudice are labeled regressive racism or regressive prejudice (Rogers &
Prentice-Dunn, 1981). Regressive prejudice occurs because controlling prejudiced
responses requires attention and mental resources; when those resources
are not available or when the control process is mort-circuited people regress
from controlling prejudice to expressing it. This section examines some of the
factors that can produce regressive prejudice.
Control over Behavior. To avoid responding in a certain way, people must be
able to control the behavior. As we noted earlier, some behaviors, such as nonverbal
responses, are less under voluntary control than others, such as the content of
what one says. Consequendy, people may give off nonverbal cues that imply dislike
of or discomfort with a nlember of a stereotyped group even while trying to
behave in a positive manner (see, for example, Dovidio et al., 2002). In one study
of this process, Dovidio and his colleagues (2002) tested White reseatch participants
for implicit prejudice against Mrican Americans. Each participant then held a
3-minute conversation with a Black student who was working for the researchers;
Most research on stereotype arousal uses laboratory
methods, so you might wonder how well those
procedures generalize to everyday life. Consider the
following commercial that was televised in 1995.
The commercial opens with a picture of a poker~
faced, vaguely threatening and ominous Black man
with his head shaved that takes up the left two~thirds
of the screen. After a moment, the following lines appear
one at a time down the right side of the screen,
with a pause between each:
Michael Conrad.
Male. Age 28.
Trafficking.
Armed Robbery.
Assault and Battery,
Extortion,
FROM PREJUDICE TO DISCRIMINATION 379
Rape.
Murder.
Apprehended
January 1994 by
Joseph Cruthers,
Shown here. (Urban Alliance on Race Relations,
1995)
By the time “Apprehended” appears, most people
assume the picture is that of the criminal, Michael
Conrad. When they learn that the picture is that of a
policeman, they feel both surprised and upset that
they have operated on the social stereotype that Black
men are dangerous and made the racist assumption
that the pictured Black man must be a criminal (Fazio &
Hilden, 2001).
because the conversation was about a race-neutral topic, the White participants
were presumably motivated to act in a nonprejudiced manner. However, conlpared
to the nonverbal behavior of participants with lower prejudice scores, the
behavior of those with higher scores was rated as less friendly by both the participants’
Black conversational partners and White raters who saw videos of the participants’
side of the conversation. Thus, when people hold stereotyped views of
other groups, those views may “leak out” through their nonverbal behavior even
while they are trying to control the effects of those stereotypes on other behaviors
(see also Sekaquaptewa et al., 2003).
Cognitive Demands. As we saw in Chapter 4, cognitive demands, such as
trying to remember an 8-digit number or working under time pressure, make
it difficult for people to control the effects of stereotypes and prevent stereotypes
from affecting their judgments. The same processes affect the expression of prejudice:
Even for people who are motivated to control their prejudices, cognitive
demands can reduce their mental resources and so reduce control. To demonstrate
this effect, Daniel Wegner and his colleagues (reported in Wegner, 1994)
had research participants complete sentences about women such as “Women
who go with a lot of men are … ” (Wegner, 1994, p. 47). Half the participants
were instructed not to be sexist in their responses and half received no special instructions;
in addition, half the participants in each of these groups were put under
time pressure to respond and half were not. Wegner and his colleagues found that
participants who had been instructed to avoid sexist responses (and therefore were
nlotivated to do so) and were not under time pressure provided fewer sexist responses
(such as “sluts” to the sentence stem just given) than participants who
380 CHAPTER 10
received no such instructions. However, time pressure undermined the ability to
control prejudice of the participants who were motivated to be nonsexist: They
made sexist responses at the same rate as the participants who were not motivated
to control their sexism.
Disinhibitors. Because motivation to control prejudice derives from social
norms to avoid being prejudiced, factors that reduce people’s motivation to
cOlllply with social nonns can also reduce motivation to control prejudice and
disinhibit its expression. Two factors that motivate nonn compliance are the desire
to avoid the social punislunents that other people would mete out for violating
norms and the desire to receive the rewards that others provide for nonn
compliance. As a result, when people are anonymous and cannot be identified,
they are less motivated to cOlnply with nonns than when they are identifiable
(see, for example, Mym, 2008). Edward and Marcia Donnerstein (1976) tested
the effects of anonymity on the release of prejudice in a study in which White
research palticipants gave what they thought were electric shocks to a Black or
White confederate for making errors on a learning task. Aggressive discrimination
was accessed in two ways. Overt, direct aggression was measured by the
setting on a dial that indicated shock level. Covert, indirect aggression was measured
by shock duration-that is, how long participants pressed the button that
supposedly delivered the shock. Half the participants thought that their behavior
was being monitored and half thought they were anonymous.
Results showed that anonymous participants showed lllore direct aggression
toward the Black than the Wbite person, indicating that anonymity facilitated
discriminatory behavior. Moreover, in contrast to their anonymous peers, participants
who were identifiable favored the Black person by showing less direct
aggression toward him than toward the White person, perhaps reflecting a motivation
not to appear prejudiced. However, nonanonymous participants showed a
higher level of indirect aggression toward the Black person than the White person,
perhaps because covert expression of prejudice is seen as more socially
acceptable.
Strong emotions also can lead people to ignore social norms. In another
study of interracial aggression, angered and cahn White research participants
had the opportunity to administer electric shocks to a Black or White person
(Rogers & Prentice-Dunn, 1981). Results paralleled those Donnerstein and
Donnerstein (1976) found for anonymity: Angry participants gave stronger
shocks to the Black person than to the White person, whereas calm participants
gave stronger shocks to the White person. Alcohol consumption is another notorious
disinhibitor of compliance with social nonns. Not sulprisingly, then,
Laurie O’Brien and her colleagues (as reported in Crandall & Eshleman, 2003)
found a correlation of .31 between alcohol intoxication (as measured by
Breathalyzer tests of people leaving bars) and willingness to express prejudice
against racial and religious groups.
The expression of prejudice is a forbidden act, but the implied approval of
an authority figure can release normative restraints. For example, Donnerstein
and Donnerstein (1976) reported a study in which White participants observed
FROM PREJUDICE TO DISCRIMINATION 381
a peer delivering an intense electric shock to a Black person. Some of the
participants then saw the experimenter overseeing the study severely reprimand
the person for delivering such a strong shock; other participants saw the experimenter
ignore the shock level, thus implicitly giving approval to it. Participants
in a control condition saw a shock being administered but did not know its
intensity. All the participants then took part in the experiment. Compared to
the control participants, those who saw a peer reprimanded for administering a
strong shock gave weaker shocks, but those who saw no reprimand gave much
more severe shocks.
The perceived costs and rewards in a situation also can influence whether
people feel comfortable ignoring social norms, such as the nonn to help others
in need. When a motorist is stranded on a highway, for example, potential
helpers must decide whether the costs of helping, such as effort, time, or
potential risk, outweigh the rewards of helping, such feeling good about relieving
another person’s suffering. Donald Saucier, Carol Miller, and Nicole
Doucet (2005) reasoned that, in such situations, discrimination against Blacks
would be more likely if people could justify the failure to help by detennining
that the costs were too high. These researchers conducted a meta-analysis of
studies assessing whether Black and White people received help in a variety of
situations. Results showed no evidence of universal discrimination against
Black people who needed help. However, in situations where the failure to
help could be easily justified, such as when helping would take a longer time,
involved more risk, or was otherwise difficult, Black people received less help
than did White people.
Finally, other people’s behavior can disinhibit prejudice. Linda Simon and
Jeff Greenberg (1996) created a situation in which four White research participants
and a Black confederate worked in separate cubicles on a creativity task.
The participants had been earlier detennined to have positive, negative, or
ambivalent attitudes toward Mrican Americans. The participants labeled their
answers with a code letter they wrote on a yellow sticky note attached to the
answer sheet. They then evaluated what they thought were the other participants’
responses (actually standard responses created by the researchers) that the
experimenter passed around one at a time. Some participants received a list with
no comment added to the sticky note, others received a list with a handwritten
addition to the sticky note that read “1 can’t believe they stuck us with this Black
person! (please erase this)” (the ethnic criticism condition), and a third group of
participants received a list with a note that read “I can’t believe they stuck us
with this nigger! (please erase this)” (the derogatory ethnic label condition). k
shown in Figure 10.1, participants with positive attitudes toward African
Americans were unaffected by the comment nlanipulation, giving the Black
participant’s contribution a rating of about six in all conditions. Consistent with
the theory of ambivalent prejudice, racially ambivalent participants inflated their
ratings in the derogatory label condition, presumably to emphasize that they
were not prejudiced. However, anti-Black participants who had been primed
by the derogatory label felt free to express their prejudice, resulting in lower
ratings than in the other two conditions.
382 (HAPTER 10
8,—————————————,
7
6
5
No comment Ethnic crtitcism Derogatory ethnic
label
Type of comment overheard
• Pro-Black participants
Anti-Black participants
.. Ambivalent participants
FIG U R E 10.1 Priming as a Releaser of Regressive Prejudice
White research participants either saw no comment about a Black confederate or read a note with a critical comment
that induded a reference to race (“Black person”) or a critical comment that included a derogator ethnic label
(“nigger”). Ratings of the Black person made by participants who had positive attitudes toward Blacks were not
affected by the type of comment read. However, participants with negative attitudes toward Blacks made more negative
ratings when they read the derogatory ethnic labeL In contrast, the ratings made by participants characterized by
ambivalent prejudice were more positive when they read the derogatory ethnic labeL
Moral Credentials. When we, the authors, were growing up, it was not unusual
for a White person who was accused of anti-Black prejudice to defend
him- or herself by saying something along the lines of “How can you call me
Iprejudiced? Why, some of my best friends are Black!” The claim of friendship
twas used as a kind of credential to establish the person’s lack of prejudice.
More recently, Benoit Morrin and Dale Miller (2001) have suggested that complying
with the nonn to avoid prejudiced behavior can have an ironic effect: It
can increase the likelihood of behaving in a prejudiced way in the future by re- I ~UCing the motivation to appear unprejudiced. They believe that acting in a
nonprejudiced way establishes what they call moral credentials by allowing people
to show others that they are not prejudiced and by reinforcing their own beliefS
that they are not prejudiced. Having established to themselves and to others that
they are not prejudiced allows them to act in a prejudiced manner if they are
disposed to do so; if challenged, they can point to their earlier behavior as evidence
of their lack of prejudice.
Monin and Miller (2001) tested their theory in a set of studies in which
some research participants could establish their unprejudiced credentials by
either rejecting a set of stereotypical statements about women or by selecting a
well-qualified woman or African American for ajob. Other participants had no
opportunity to act in a nonprejudiced manner. All participants then rated the
extent to which they thought a given job (such as construction foreman or
FROM PREJUDICE TO DISCRIMINATION 383
police officer) was better suited for a man or woman, or a Black person or a
White person. In all three experiments, participants who earlier had had the
opportunity to establish that they were nonprejudiced rated the job in the second
part of the study as better filled by a man or White person. Monin and
Miller (2001) concluded that “the more confident people are that their past
behavior reveals a lack of prejudice, the less they will worry that their future
behavior is, or can be construed as, prejudiced …. By fostering self-image
security … the establishment of moral credentials eUlboldens the [person] to
respond honestly in circumstances in which political correctness pressure militates
against honest expression” (p. 40).
Reactions to Having Acted in a Prejudiced Manner
Given that situations can arise that can lead people to make prejudiced responses,
how do they then react to having acted in a prejudiced manner? The initial
research on this topic focused on differences between people high and low in
explicit prejudice. Patricia Devine and her colleagues (Devine, Monteith,
Zuwerink, & Elliot, 1991; Monteith, Devine, & Zuwerink, 1993; Zuwerink,
Monteith, Devine, & Cook, 1996) hypothesized that people with nonprejudiced
self-images who act or think in a prejudiced manner experience a discrepancy
between how they think they should respond (that is, in a nonprejudiced manner)
and how they do respond (that is, in a prejudiced manner). This discrepancy
then leads to feelings of discomfort and guilt. To test this hypothesis, they examined
differences in how people believed they should respond in interactions with
an African American or a gay man and how they thought they would actually
respond. Not surprisingly, Devine and her colleagues found that people low in
prejudice had more stringent personal standards for nonprejudiced behavior than
did people high in prejudice. Nonetheless, people both high and low in prejudice
felt discomfort over discrepancies between how they thought they should
act and how they thought they would act. However, whereas people low in
prejudice felt guilty about their discrepancies, people high in prejudice did not.
Rather than feeling negative toward themselves over their discrepancies, people
high in prejudice experienced negative emotions about other people, such as
feeling angry and irritated at them, perhaps because they believed that other people
expected them to be unprejudiced and would pressure them to behave in
unprejudiced ways.
In addition to discovering their own prejudiced behavior, people may be
confronted by others who point out behaviors that could indicate prejudice.
How do people respond then? Alexander Czopp and Margo Monteith (2003)
found that the response depended on two factors. The first was the type of prejudice
involved. People felt more concerned and guilty over racial prejudice than
over sexism. In fact, “the predominant evaluative sentiment resulting from confrontations
about gender-biased behavior was amusement” (Czopp & Monteith,
2003, p. 541), suggesting that many people do not take gender-based prejudice
very seriously. The second factor was the person who did the confronting.
People were nlore likely to dismiss an accusation of prejudice when it came
384 (HAPTER 10
from a menlber of the group toward which their prejudiced behavior was
directed than when it came [r0111 a member of their own group. Czopp and
Monteith suggested that this reaction occurred because people felt less threatened
when confronted by a member of their own group. Thus, it appears that the
greatest guilt and discomfort over acting in a prejudiced manner is elicited
when people have a high internal 111_otivation to control prejudice and when
they become aware of their prejudiced responses themselves or have those responses
pointed out by a member of their own group.
Researchers have found that guilt over having acted in ways that are discrepant
from one’s self-image leads to action that reaflinns that image (Steele, 1988).
Interestingly, there seems to be no recent research on whether feelings of guilt
about having acted in a prejudiced nlanner affect future behavior, but several
older studies have done so. For example, Steven Shemlan and Larry Gorkin
(1980) induced some research participants to nlake a sexist decision by leading
them to assume that a surgeon depicted in a story was male rather than female;
participants in a control condition nude no such decision. Later, all participants
took part in a seemingly unrelated study in which they acted as mock jurors in a
sex discrimination case involving a woman who alleged that she had been refused
a job because of her sex, with a man being hired instead. Shennan and
Gorkin found that, conlpared to participants in the control condition, those
who earlier had been induced to make the sexist decision were more likely to
see the hiring of the man as unjusti£ed and were more likely to decide the case
in favor of the woman.
Perhaps more telling are the results of an eXperll11ent conducted by Donald
Dutton and Robert Lake (1973). Based on a pretest, they selected 80 students
who were low on prejudice and who had also rated equality as a value that
was important to them. The participants thought they were taking part in a study
of physiological responsiveness to vatious stimuli and that high physiological
arousal indicated the presence of unconscious negative attitudes toward a stimulus.
While their physiological responses were recorded, participants watched a
series of slides that included pictures of Black people. Half the participants received
feedback that indicated unusually high responsiveness to the picture of
Black people relative to neutral stimuli (thus threatening their nonprejudiced
self-images) whereas the other participants received feedback that indicated similar
responses to the two types of stimuli (thus leaving their nonprejudiced selfimages
unthreatened). All palticipants were paid $2.00 in quarters (to ensure that
they had change for what happened next) for their participation. On leaving the
building, participants were approached by either a Black or a White panhandler
who asked “Can you spare some change for some food?” Eighty-five percent
of the participants in the self-image threat condition gave money to the Black
panhandler compared to 45 percent of those in the nonthreat condition. In
addition, the participants in the threat condition gave more money (an average
of 47 cents) than the participants in the nonthreat condition (an average of
17 cents). Threat condition did not affect donations to the White panhandler.
By the way, if 47 cents does not seenl like much money, it is equivalent to
about $2.30 today when corrected for inflation. Thus, when people who have
FROM PREJUDICE TO DISCRIMINATION 385
nonprejudiced self-images have that image called into question, they feel guilty
and that guilt motivates them to reaffiun their self-images of low prejudice by
acting in an especially nonprejudiced manner,
DISCRIMINATION IN ORGANIZATIONS
As we noted in Chapter 1, organizational discrimination occurs when “thQ
practices, rules, and policies of formal organizations, such as corporations or government
agencies” result in different outcomes for members of different groups
(Benokraitis & Feagin, 1995, p. 44). As we also saw in Chapter 1, although more
than 40 years have passed since the enactment of the landmark Civil Rights Act
of 1964, racial/ethnic and gender discrimination still exist in the workplace, We
have divided our discussion of this discrimination into two parts, The first part
examines employment discrimination at the organizational leveL The second part
examines some of the psychological variables that influence the discriminatory
behavior of individual decision makers, As you read this section, bear in mind
that although the eifects of gender and racial bias tend to be small (see, for example,
Barrett & Morris, 1993), small effects can cumulate over time to have a
large impact on career outcomes, For example, Richard Martell, David Lane,
and Cynthia Emrich (1996) conducted a computer simulation that modeled the
potential effects of gender discrimination in an organization that had eight levels
of promotion, The simulation started with an equal number of male and female
new employees. Althoughjob qualifications vatied among both male and female
employees, they were, on the average, the same for both groups; however, there
was a very small pro-male bias for promotions at each leveL Mter the simulation
had run through all eight levels of promotion, 65 percent of the top-level jobs
were filled by men, Therefore, even when discrimination has a small effect on
anyone decision, it can have larger effects in the long mn, In addition,
experiencing disctimination can have profound psychological and physical affects,
a topic we discuss in Chapter 11,
Organizational Research
Organizational researchers have conducted an enormous amount of research
to determine the forms discrimination takes in organizations and the ways in
which discrimination is related to characteristics of organizations, As we will
see, racial discrimination has been found at many points in the employment
process, including hiring, job performance evaluations, and promotions, There
appears to be less gender discrimination in hiring and performance evaluation,
but evidence still supports gender discrimination in promotion. Let us look at
the results of some of that research,
Hiring. The first step in the employment process is hiring, so quite a bit of researc has been conducted on the hiring process, most of it focusing on racial discriminaJtion.
One technique that has been used to study this topic is the employment audit.
In an employment audit, members of two groups are matched on appearance,
386 CHAPTER 10
education, and relevant experience, and then sent to apply for the same job. Thus,
one White person and one Black person, both with the same qualifications, dressed
similarly, and trained to act similarly, would apply for the same job. The dependent
variable in such studies is who is lucre likely to be hired, the Black applicant or the
White applicant? A review this research showed that for the most part, Black and
White applicants experience the same outcome: in 68 percent of the cases neither
was offered ajob and in 15 percent of the cases both were offered ajob (Heckman,
[~998). However, when discrimination did occur, it favored the White applicant:
overall, 27 percent of White applicants were offered jobs compared to 21 percent
of Black applicants. That is, a White applicant was 1.3 times more likely to get a
job than was a Black applicant.
The results of an employment audit study conducted by Marc Bendick,
Charles Jackson, and Victor Reinoso (1994) illustrate some specific forms that discrimination
can take. For example, they found that although employers accepted
applications from Black and White job seekers at about the Sa111e rate-95 percent
for Whites and 92 percent for Blacks–48 percent of the White applicants were
offered interviews cOlnpared to 40 percent of the Black applicants. When more
than one opening was available and the job was offered to both applicants, Black
applicants were offered lower starting salaries. In addition, Black applicants were
more likely to be steered toward a job at a lower level in the organization than
the one they had applied for. Bendick and his colleagues also found that discrimination
was more likely to occur when applications were made through employment
agencies rather than directly to the company. One reason for this finding
might be that the companies using employment agencies cued the agencies that
they only wanted to see White applicants. For example, evidence was presented
in one employment discrimination lawsuit that employers would use tenns such as
“all-AInerican,” “front-office appearance,” and “cOlporate image” in the job descriptions
they sent to employment agencies to indicate that they wanted to see
only White job candidates (Wolff, 1989).
r
– One shortcoming of employment audits is that it is impossible to perfectly
match a Black and a White applicant in every regard: They are different people
with different personalities, so it is always possible that any differences in the responses
they received are a result of their individual characteristics rather than of
their racial group membership. Therefore, researchers sometimes use matched
mailed-in resumes to see whether a Black or White applicant is more likely to
be called for an interview. Marianne Bertrand and Sendhil Mullainathan (2003)
conducted a large-scale study of this kind, sending out aInlOst 5,000 resumes
in response to 1,300 jobs advertised in newspapers in Chicago and Boston.
They manipulated three independent variables using infonnation in the
resumes: applicant race, applicant gender and applicants’ qualifications. They
manipulated race and gender by varying the applicant’s first name: “White”
names included Kristin, Carrie, Brad, and Jay; “Black” names included Ebony,
Latonya, Jennaine, and Leroy. The researchers chose the names based on the
frequency of their appearance in birth records and checked the racial associations
the names elicited by asking survey respondents to rate how “Black” and
“White” each name sounded.
FROM PREJUDICE TO DISCRIMINATION 387
Bertrand and Mullainathan (2003) found that, overall, only 17 percent of the
applicants were called for an interview; however, 10,1 percent of the White applicants
were called compared to 6.7 percent of the Black applicants, In other
words, White applicants received an interview offer for every 10 resumes submitted,
whereas Black applicants had to submit 15 resumes to get an interview,
In addition, although White applicants with high quality resumes were more
likely to be called than White applicants with low quality resumes, 11.3 percent
compared to 8,8 percent, resume quality had virtually no effect on the response
rate for Black applicants: 7,0 percent for high quality resumes and 6.4 percent for
low quality resumes. Note that a White applicant with a low quality resume was
26 percent more likely to be called for an interview (8.8 percent) than a Black
applicant with a high quality resume (7.0 percent). Resnlts were similar in
Boston and Chicago and for six types of jobs ranging from clerical worker to
manager across six industries, Bertrand and Mullainathan put their results in this
context: “in our sample, [having] a White name is equivalent to about 8 additional
years of [job-related] experience” in being called for an interview (p. 10).
An important part of the hiring process is the job interview, A review 31
studies of racial group differences in employment interview outcomes showed
that, on the average, Whites received higher ratings than either Blacks or
Hispanics (Huffcutt & Roth, 1998). Such racial differences in evaluations were
greater for low level jobs than for higher level jobs and discrimination increased
as the proportion of Black applicants increased. Thus, White job applicants are
more likely to get interviews than are their equally qualified Black peers and,
once interviewed, are more likely to get high ratings on suitability for the job.
Box 10.3 relates one Black man’s experience in applying for a low-level job and
a few of his experiences on that job,
Lawrence Graham (1995), a highly successful Harvard
educated lawyer, set out to uncover what life was like as
an employee at one ofthe all~White country clubs in
Greenwich, Connecticut, In his essay on his experiences,
he first describes applying for work, His goal was to be a
waiter and he applied for this job at five country clubs,
As he writes, “During each of my phone conversations, I
made sure that I spoke to the person who would make
the hiring decision, I also confirmed exactly how many
waiter positions were available, and I arranged a
personal interview within forty minutes to an hour of
the conversation, just to be sure that they could not tell
me that no such job was available” (p, 4). Upon arrival
at each of the five clubs, he was told either that there
were no openings or that he was not qualified. One
receptionist threatened to call security. Another
employer firmly insisted he could not have been the
person she talked to on the phone, No one would even
accept his application for a waiter position.
Graham did secure a job as a busboy at one of the
clubs. In that role, people made negative comments
about “Negroes” and other minority groups right in
front of him. Example comments included “My
goodness. , ‘ . That busboy had diction like an educated
White person” (p. 12). He learned that the staff
quarters were unselfconsciously called the “Monkey
House” because at one time all the workers had been
Black. At one point, he was instructed to find the
“Chinaman,” a supply clerk, and was told it was easy to
remember his location because it was right next to the
laundry. These and many similar experiences took
place in less than a one~month period. They also took
place in the 1980s, not,. as you might expect by reading
them, during the pre-dvil~rights era of the 1960s.
388 (HAPTER 10
There is less research on gender differences in hiring than on racial
differences. However, Bertrand and Mullainathan (2003) found that male and
female applicants were called for interviews at about the same rate, 17.2 percent
for women and 15.3 percent for men. Also, Heather Davison and Michael
Burke (2000) reviewed 49 studies in which participants evaluated the suitability
of equally qualifled male or female applicants for a job. Overall, they found
virtually no discrimination, although men tended to receive slightly higher
suitability ratings for male-stereotyped jobs and women tended to receive
slightly higher suitability ratings for female-stereotyped jobs. Thus, the research
indicates that currently there is little gender discrimination in the hiring process,
However, Davison and Burke noted that research reviews published in
1979 and 1988 did find evidence of discrimination, suggesring that gender discrimination
has decreased over time. For example, the 1988 review (Olian,
Schwab, & Haberfield, 1988) found that gender accounted for 4 percent of
the variance in hiring evaluations whereas the 2000 review found that it
accounted for less than 1 percent.
There is, however, one important factor that leads to gender discrimination in
lllring: pregnancy. In the United States, discriminating against pregnant women violates
the federal Pregnancy Discrimination Act, passed in 1975 as an am_endment to
the Civil Rights Act. However, the legality of this discrimination has not reduced the
number of claims filed with the Equal Opportunity Employment Commission; in
fact, number of claims has increased nearly every year since 1992 (Equal
Opportunity Employment Conmllision, 2007). Evidence that pregnancy affects hiring
decisions comes from a study of undergraduate business school students who evaluated
job applicants based on an interview. Although pregnant and nonpregnant
applicants were seen as equally qualified, pregnant women received lower hiring reconunendations
than did nonpregnant applicants (Cunningham & Macan, 2007).
Raters also believed that the pregnant applicant would be absent more often and
would be more likely to quit than the nonpregnant applicant.
[
Performance Evaluation. Once a person is on the job, career success is
strongly affected by perfonnance evaluations: Positive evaluations can lead to
valuable training opportunities, promotions, and pay raises, whereas poor evalua~
ons can lead to losing one’s job. Chieh-Chen Bowen, Janet Swim, and Rick
Jacobs (2000) reviewed 27 studies of gender differences in job perfonuance ratings
in organizations. They found that, overall, men and women received equal
job perfonnance ratings, although male evaluators tended to give somewhat
higher ratings to men than to women. Unlike Davison and Burke (2000), who
found that both men and women received slightly higher suitability ratings for
sex-stereotyped jobs, Bowen and her colleagues found that the sex typing of the
job was not related to perfonnance ratings. However, the sex stereotyping of the
items on perfOlmance evaluation measures was related to the ratings: Men received
higher ratings on measures that included mostly masculine-stereotyped items,
WOlnen received higher ratings on measures that included mostly femininestereotyped
items, and men and women received equal ratings on Ineasures that
contained mostly gender-neutral items. Bowen and colleagues also found that
FROM PREJUDICE TO DISCRIMINATION 389
supervisors who had more opportunity to directly observe employees’ job performance
gave somewhat higher performance ratings to women. They concluded
that “any overall gender discrimination that one might observe in the work force
is … not directly attributable to biases in performance appraisal” (Bowen et al.,
2000, p. 2206). However, their results do indicate that to the extent that perf ornunce
evaluations are made by Inen, especially men who do not directly observe
their employees’ perfonnance, and to the extent that an organizations’ performance
evaluations measures include masculine-stereotyped items, men will receive
higher evaluations than women.
Unlike the situation for gender, race/ethnicity does appear to affect job perfor n1.ance evaluations. Based on a review of 48 studies, Philip Roth, Allen HuffcuttJ,
and Philip Bobko (2003) found that White employees received higher evaluations
than Black employees, although there were no differences in evaluations of
Hispanic and White employees. Even when numerical ratings of job perfonnance
are the same for Black and White workers, the narrative comments that supervisors
make about employee performance might be biased. For example, Patricia Thomas
and her colleagues (Thomas, Edwards, Perry, & David, 1998) compared the perfonnance
evaluation comments made by the supervisors of 582 male U.S. Navy
officers who had received the highest nuulerical perfonnance rating. They fouud
that White officers were more likely to be described as outstanding leaders and
were more likely to be recommended for positions of command than were Black
officers. In addition, White officers were more likely to be reconnnended for early
promotion and to be described as having characteristics that other research shows
lead to early promotion (Johnson, 2001). That is, promotion boards seem to use
certain characteristics as cues when selecting officers for early promotion and evaluators
were more likely to attribute those characteristics to White officers than to
Black officers. Similarly, letters of reference for faculty positions in chemistry were
similar in length and, overall, were equally positive for male and female applicants.
However, letters supporting male candidates contained more standout words, such
as ”Inost” gifted and “rising” star than did letters supporting female candidates
(Schmader, Whitehead, & Wysocki, 2007).
Although racial group differences in job performance evaluations might reflect
racial prejudice, they nright also be accurate, reflecting actual racial group differences
in average job performance. For example, Roth and his colleagues (2003)
found that, compared to White workers, Black workers had lower average scores
on objective measures of work pelfonnance, and that these differences were about
twice as large as the differences in supervisor ratings. However, even though minority
group employees may actually perfonn at a son1.ewhat lower level than
White employees, that lower level of performance nright itself be a result of
prejudice and discrimination in the workplace (Greenhaus, Parasuramen, &
Wonnley, 1990). That is, differences in race may lead to differences work experiences
which, in tum, affect job performance. The possibility is supported by the
fmding that although both White and Black supervisors give White workers
slightly higher performance ratings than Black workers, White supervisors see larger
perfom”nce differences than do Black supervisors (Stauffer & Buckley, 2005). That
supervisor race affects perceptions of work performance suggests that subtle
390 CHAPTER 10
factors may be affecting performance ratings. Research shows, for example, that
members of higher status groups-in this case, White supervisors-are more
likely to favor the ingroup than are melnhers of low status groups-in this case,
Black supervisors (George & Chattopadhyay, 2002); if so, this would give an
edge to White workers. Organizational researchers have not paid Inuch attention
to issues such as these (Brief, 1998); however, organizational theorists have proposed
a number of workplace characteristics could adversely affect minority
group members’ job performance (Roberson & Block, 2001).
[
One such characteristic is what Daniel Ilgen and Margaret Youtz (1986) called
the lost opportunities dJect: “Differential treatment of minority and llUjOrity group
members may result in different on-the-job opportunities for these two groups. To
the extent that minority group members have fewer and less favorable opportunities,
lower performance for minorities may result” (p. 317). For example, minority group
managers report that, compared to their White peers, their supervisors appear to
view them as less competent (such as by reviewing their work more fi’equently
and more closely) and give them less support and encouragement (Blank & Slipp,
1994; Greenhaus et al., 1990; James, 2000). Prejudiced Whites also can create a
chilly climate for their Black coworkers. For example, Black managers reported
feeling less accepted than White coworkers by their White peers (Greenhaus et
al., 1990; Blank & Slipp, 1994). In addition, lower-level White workers showed
nlore work avoidance, such as absences, as the proportion of minority workers in
their work units increased and expressed stronger intentions to quit their jobs if
another employment opportunity arose (Tsui, Eagan, & O’Reilly, 1992). Such
coworker prejudice can have a direct effect on Black workers’ job performance.
For example, Black research participants who worked with a prejudiced, compared
with an unprejudiced, White participant were 30 percent less productive
(Dovidio, Kawakami, & Beach, 2001).
Ilgen and Youtz (1986) also proposed that workplace prejudice and discrimination
could have an adverse effect on minority employees’ morale, which, in
tunl, could affect job perfonnance. On a survey of the experience of workplace
mistreatment, such as being set up for failure, being treated as if they didn’t exist,
and being the target of insulting jokes or comments, Black workers reported
more mistreatment than White workers (Deitch et al., 2003). Moreover, the
experience of such mistreatment was related to reduced job satisfaction. Other
research shows that lower job satisfaction is related to lower job perfonnance
(judge, Thoreson, Bono, & Patton, 2001). For examples of some of the hassles
that Black workers face, see Box lOA.
Promotion. Promotion within an organization is based largely on job pelformance
evaluations. However, as Bowen and colleagues (2000) have noted, even
when women and members of minority groups receive the SaIne perfonnance
evaluations as men and majority group members, decisions based on those evalua- itions can still be biased. And so it seems to be with promotions. Looking first at
gender differences, researchers have found that although women and men receive,
on average, equal job perfomlance ratings, their supervisors often see them as
The editors of Harvard Business Review asked
organizational consultants Keith Caver and AnceUa
Livers (2002) to write a fictional letter from a Black
manager to a White boss describing some of the
everyday, almost certainly unintentional, incidents that
make B lack managers feel like outsiders. The events
they noted included:
• Being cast as a de facto diversity expert even
though the managers’ training gave them no
special expertise in that field: “Despite my 15
years of experience, despite my solid track record,
my new colleagues appeared to have little
interest in my business expertise. Instead, they
seemed to have assigned me some special role:
official interpreter of minority concerns for the
organization” (p. 78).
• Having one’s presence questioned when others go
unchallenged: “One weekend I went to the office
in my normal, casual weekend attire …. Before
getting into the elevator I was stopped by an
informally dressed young white man who in a
stern voice asked to see my identification …. I had
worked here for two years, but because I was out
of context, he assumed I was a thug. You might
chalk it up to an honest mistake, but I can assure
you he hadn’t challenged any of the white people
entering the building” (p. 79).
• Resentment over increased workplace diversity:
Robert, a Black manager, hired a Black woman
and promoted a Black man. Afterwards, “Robert
FROM PREJUD!CE TO D!SCR!MINAT!ON 391
began to hear whispers in the halls-suggestions
that he was building his own little ‘ghetto
fiefdom’ and having a White colleague ‘jokingly’
say to him, ‘So white people aren’t good enough
for you?'” (p.79).
• Not being trusted to do one’s job properly: In
addition to the comments of his peers, Robert’s
boss “suddenly seemed to take a greater interest
in the details of his group’s work-asking for
reports and updates he’d never needed when
Robert’s team was primarily white. Subtly, his
boss was letting him know that at some
level he expected the team’s performance to
drop” (p. 79).
• Having others assume that because you are not
White, you are a low-level employee: A Black
woman “was recently hired as a senior vice
president for a major financial institution. With
the exception of a few initial interviews and
meetings, she did not set foot in the organization
until her first day at the office. As she emerged
from the elevator, she was abruptly greeted by a
white male who directed her to a small cubicle
and asked her to quickly put her things away as
they were expecting a new senior officer to arrive
shortly” (p. 80).
The effect of experiences such as these is that for
minority workers, “race is always with us. As a friend of
mine said recently, ‘I don’t think a day goes by that I’m
not reminded I’m black'” (p. 81).
having less promotion potential than men (Landau, 1995; Shore, 1992), they mustJ
wait longer for promotion (Cannings & Montmarquette, 1991; Maume, 1999),
and they receive fewer promotions the higher they move in the organizational
structure (Lyness &Judiesch, 1999).
Even in situations in which women receive higher job performance ratingsl
than men, they may be promoted no faster (Cannings & Montmarquette, 1991; J
Shore, 1992). For example, Ted Shore (1992) studied the results of assessment
center ratings of male and female managers. When managers attend an assessment
center, they undergo a set of written and perfonnance tests designed to assess
their potential for prol11-etion. Shore found that, on the average, women and
men received the same scores on measures of mental ability and similar ratings
on interpersonal skill. In addition, women scored higher on measures of managerial
job perfonnance. Nonetheless, women received the san’le overall managerial
potential ratings as men and were promoted at the same rate.
392 CHAPTER 10

‘ Ironically, the male advantage in promotion may be greater in femaledominated
occupations, such as clerical and sales, than in male-dominated occupations,
such as nunufacturing (Maume, 1999), an outcome called the glass
escalator effect (Williams, 1992). For example, David Maume (1999) found that
—fnen were 17 percent more likely than WOlnen to be promoted in femaledominated
occupations. Being in a female-dominated job not only increased
men’s odds of promotion, it decreased women’s. Similarly, Christine Williams
(1992) found that male nurses, librarians, elementary teachers and social workers
benefited from being in a female-dominated profession both at the hiring
stage and on the job. However, these men often encounter negative stereotypes,
such as the perception that they are passive or felninine and they are
sometimes viewed -with suspicion-so much so that they “alter their work behavior
to guard against sexual abuse charges, particularly those in specialties
requiring intimate contact with women and children” (Williams, 1992, pp.
261-262). These negative stereotypes sometimes lead to negative self-esteem
and, ultimately, career change. Interestingly, negative stereotypes also may
contribute to the glass escalator effect: To avoid outsiders’ criticisms and suspicions,
supervisors may promote the men to higher level jobs where they have
less contact with the public (Williams, 1992).
[
In regard to race, Black workers are less likely to be promoted than their
White counterparts (Roth et aI., 2003) and must wait longer for promotion
(Maume, 1999). As with women and men, even when Black workers receive
the same job perfOlmance ratings as their White peers, they are likely to receive
lower ratings on promotion potential Ganles, 2000; Landau, 1995; Thomas et al.,
1998). Even after controlling for differences in education and experience, Black
assistant football coaches are less likely to be promoted to head coach than are
their white counterparts (Sagas & Cunningham, 2005). In addition, White employees
are paid more than Black enlployees in jobs with equivalent levels of
authority (Dreher & Cox, 1996; Smith, 1997). Ryan Smith (1997) also found
that although Black employees’ absolute salaty levels increased with promotion,
the difference in Black and White workers’ salaries increased with higher levels
of authority: Black workers with no authority were paid 86 percent of what their
White peers received, 81 percent at low levels of authority, and 77 percent at
middle levels of authority. (There were too few Black managers at high level
of authority for Smith to make meaningful comparisons.)
The problem begins at the lowest level of promotion. Using the results of
separate surveys, David Maume (1999) and Ryan Smith and James Elliott (2002)
both found that Mrican Americans in nomnanagerial jobs were 50 percent less
likely to be promoted to Inanagerialjobs than were their White peers. Smith and
Elliott further noted that when Mrican Americans did hold first-level managerial
positions, they were nl0re likely to supervise Black workers than White workers.
Because most Black workers are in low-level jobs, this ethnic matching of supervisors
and employees means that most Black managers are found in low-level
positions of authority. Smith and Elliott (2002) refer to this ethnic matching phenomenon
as the “sticky floor” effect: “The relative position of one’s ethnic
group within an organization constitutes the ‘sticky floor’-one to which
FROM PREJUDICE TO DISCRIMINATION 393
individual opportunity for authority ‘adheres.’ If one’s ethnic group dominates
only entry-level jobs within an organization, then one’s authority chances will
be restricted largely to supervising entry-level workers. If one’s ethnic group
dominates higher-level positions, then one’s authority chances will increase accordingly”
(p. 274; see also James, 2000).
At least two other organizational factors are related to Black managers’
slower promotion rates relative to their White peers. First, Black employees
tend to be “tracked” into certain job categories, such as affinnative action officer.
Black coaches, for example, often fill positions such as recruiter or minority affairs
officer that derail them from advancelnent to a head coach position (Sagas &
Cunningham, 2005). Similarly, Sharon Collins (1997) interviewed Black executives
in large White-owned corporations and found that 63 percent had been
career-tracked into jobs designed to “mediate the social pressures related to black
protest for civil rights, [such as] affinnative action or urban affairs manager”
(p. 327). Often, the only “qualification” these executives had was their race:
Professionals with advanced degrees in fields such as accounting, engineering,
and chemistry were put into these positions despite the oct that they had no
training or previous experience in the field. These managers were moved from
their chosen career fields based solely on a stereotype-that members of minority
groups make better diversity managers simply because of their group membership.
Jobs in these categories tend to be “dead-end” positions or have slower
promotion rates and “top out” at lower levels of authority than other jobs in
other categories, such as sales and operations management, regardless of the race
of the people holding the jobs (James, 2000).
Second, one’s mentor can have important effects on career progression because
of the fonnal and informal training and social support mentors provide, and
because of the connections they have with upper-level management (Roberson &
Block, 2001). Because White men predominate in upper-level management positions,
they can be especially helpful as menton;. For example, George Dreher and
Taylor Cox (1996) found that having had a White male mentor added more than
$22,000 per year to manager’s salaries after 10 years of experience compared to
managers who had had female or non-White mentors. They also found that
women, African American men, and Hispanic men were less likely than White
managers to have had a White male mentor.
Although it is tempting to attribute the differential outcomes for female and
Black workers to intentional discrimination, Smith and Elliott (2000) see the situation
differendy: “We believe that something more subtle and profound occurs
in the process of doing ‘business as usual’-mere maintenance of the status quo is
more than enough to perpetuate racial stratification” (p. 274); the same principle
applies to gender stratification.
Individuals in Organizations
Although people very often talk about organizational discrimination and institutional
discrimination, if you think about it, organizations and institutions do not
discriminate: Individuals in organizations and institutions discriminate. That is,
394 CHAPTER 10
individuals set and enforce discriminatory policies, make discriminatory hiring
decisions, and give discriminatory perfom1allCe evaluations. Even when decisions
are made by committees, individuals have input into those joint decisions. Given
this situation, surprisingly little research has been conducted on how individuallevel
psychological processes influence discriminatory outcomes in organizations
(Brief, 1998; Roberson & Block, 2001). This section discusses some of those processes:
stereotype fit, intergroup respect, shifting standards, contem-porary prejudice,
and confom-rity to perceived organizational norms.
i-Stereotype Fit. Madeline Heilman (1983; 2001) developed the stereotype fit
hypothesis to explain why WOlllen hold few managerial or executive positions
in organizations relative to men. The hypothesis postulates that the characteristics
associated with effective managers are very similar to the cultural stereotypes
_9f men and very different from the cultural stereotypes of Wonlen. Therefore,
nlen are perceived as fItting into the managerial role but women are not; as a
result, women are less likely to be hired for managerial positions and, once hired,
less likely to be promoted into higher positions (see also Eagly & Karau, 2002).
More generally, people see men as better suited for “masculine” jobs such
as business manager and construction worker and women as better suited for
“feminine” jobs such as nurse and secretary (Biernat & Kobrynowicz, 1997;
Cejka & Eagly, 1999).
Evidence for the manager-as-rnale stereotype has come from studies in
which experienced male corporate managers rated the target groups male manager
and female manager on traits that characterize effective managers (Dodge, Gilroy, &
Fenzel, 1995; Heilman, Block, & Martell, 1995; Mat~ell, Parker, Emrich, &
Crawford, 1998). These studies have found that male managers as a group received
higher ratings on the traits than did female managers as a group. However,
the results of the studies were less consistent when the research participants rated
succesiful male managers and succesifulfemale managers: two studies found a pro-male
bias (Dodge et a!., 1995; Heilman et a!., 1995) but one found no difference in
the ratings of successful male and feuule managers (Martell et a!., 1998).
Interestingly, the one study that included female managers as participants
(Dodge et al., 1995) found that they expressed a pro-female bias when rating
male and female managers in general, although not when rating successful male
and fenule managers. These results suggest that an egocentric bias might be at
work when rating managers in general: Because managers are making the
ratings, they see managers as having characteristics similar to their own-malestereotypical
when men are making the ratings and female-stereotypical when
wonlen are making the ratings. However, given that men predominate in
positions of power in organizations, nlen’s belief systems are more likely to influence
organizational decisions than are women’s.
Further evidence that evaluators may adjust the criteria to fit their expectations
comes from a study of potential applicants for the job of police chief
(Uhhnann & Cohen, 2005). Undergraduates evaluated a male or a female applicant
who was described as either streetwise (tough and had experience in rough
neighborhoods) but not fonnally educated or educated but not streetwise. Raters
FROM PREJUDICE TO DISCRIMINATION 395
adjusted the criteria they believed were important for the job such that, regardless
of which set of credenrials the male applicant had, he was considered better
qualified than the female applicant. When the male applicant was described as
streetwise, for example, those characteristics were seen as most important for a
police chief and not his education. Raters, then, did not stereotype the applicants,
but rather adjusted “their notion of ‘what it takes’ to do the job well in
a manner tailored to the idiosyncratic credentials of the person they wanted to
hire” (Uhhnann & Cohen, 2005, p. 478)-in this case by choosing the male
applicant.
An example of stereotype fit in operation comes from a study of the evaluations
and decisions that employment interviewers made about male and female
applicants for managerial jobs (Van Vianen & Willemsen, 1992). Consistent with
the stereotype fit hypothesis, interviewers believed that the ideal job applicant
would have more masculine traits than feminine traits. Moreover, although interviewers
regarded the male and female applicants as being equally qualified in
terms of education and experience, they were more likely to recommend that
male applicants be hired. Finally, interviewers attributed more masculine traits
to successful applicants than to unsuccessful applicants, indicating that masculinity
played an important role in their decisions. In short, because the interviewers saw
female applicants as less masculine than male applicants and viewed the jobs as
requiring masculine traits, they were less likely to recommend that female applicants
be hired even though they had the same objective qualifications as the male
applicants.
Stereotype fit (or lack of fit) can also influence performance evaluations.
Jennifer Boldry, Wendy Wood, and Deborah Kashy (2001) examined this influence
in a study of ratings male and female ROTC cadets nlade of one another.
They found that although male and female cadets scored equally well on objective
measures of military performance, female cadets received lower ratings on
motivation and leadership from their fellow cadets. Thus, equal performance,
when flltered through gender stereotypes, can lead to different evaluations. Also
recall the research discussed earlier that found that women are less likely to be
promoted despite equal job performance evaluations (Landau, 1995; Lyness &
Judiesch, 1999; Maume, 1999); similar processes might be operating in those
situations.
Although Heilman (1983, 2001) developed the stereotype fit hypothesis to explain
gender differences in organizational outcomes, it can also explain racial and
ethnic group differences in outcomes; however, less research has been conducted in
that context. One study that has addressed this question examined experienced
White managers’ ratings of managers, African Americans, and Whites on a set of
traits characteristic of good managers (Tomkiewicz, Brenner, & Adeyemi-Bello,
1998). Results showed a correlation of r = .54 between the ratings of managers
and the ratings of Whites but a correlation of on1y r = .17 between the ratings of
managers and the ratings of African Anlericans, indicating that Whites were seen as
fitting the managerial role better than were Mrican Americans. Similarly, White
managers in the hospitality industry saw a correspondence between the category
“successful middle manager” and the categories “Caucasian American middle
396 CHAPTER 10
manager” (r ~ .53) and “Asian American middle manager” (r ~ .56; ChungHerrera
& Lankau, 2005). They also saw a relationship between the categories
“successful middle manager” and “Mrican American manager,” but this correlation,
r::::: .24, was significantly lower than the relationships found between successful
and White or successful and Asian managers. The relationship between
“successful middle manager” and “Hispanic American manager,” r = .13, was not
significant.
Similar processes operate for lower level jobs as well. For example, Jaleen
Kirschenman and Kathryn Neckerman (1990) interviewed 185 Chicago-area
employers about their impressions of workers from various racial and ethnic
groups who fill low-level jobs such as sales clerk, typist, restaurant worker, and
assembly worker. They found that employers generally perceived Blacks and
Hispanics as having none of the characteristics of good workers, seeing them as
unskilled, illiterate, dishonest, unmotivated, involved with drugs and gangs,
lacking a work ethic, and having few interpersonal skills. The employers contrasted
Blacks and Hispanics with recent immigrants from central Europe,
whom they saw as having all the characteristics of desirable workers. These
stereotypes were reflected in em-ployee recruitment strategies. For example,
Kirschenman and Neckerman (1990) noted that “one company advertised for
skilled workers in Polish- and German-language newspapers, but hired all its
unskilled workers, 97 percent of whom were Hispanic,” through employee
referrals (p. 210).
Employers did have subcategories within their stereotypes, however. For
example, they differentiated between “desirable” Black workers-those who had
middle- or working-class backgrounds-from “undesirable” Black workersthose
who resided in urban ghettos. However, employers tended to use race as
a marker for ghetto resident, assuming that Black job applicants were ghetto residents
(and therefore would not be good workers) unless the applicants provided
evidence of being “desirable.” Such evidence included speaking standard English,
dressing appropriately for the job interview, being able to demonstrate job-related
skills, having a history of steady employment, and providing a non-ghetto address.
In sum, Kirschenman and Neckerman (1990) found that employers
assumed that Blacks and Hispanics were unqualified for even low level jobs
unless the applicant proved otherwise, but applied less stringent criteria to
members of other groups.
Finally, although stereotype fit can result in a disadvantage for some groups,
it can confer an advantage on other groups. For example, the Asian stereotype
includes being talented at math and computers whereas the female stereotype includes
having little talent in those domains. Jennifer Steele and Nalini Ambady
(2004) hypothesized that an Asian American woman’s perceived qualification for
a computer technician job would depend on whether her Asian identity or her
female identity was salient to the decision maker. College student participants
read the description of a student computer technician job, reviewed a completed
application for the job, and then interviewed the applicant, an Asian American
woman. The application form was designed to emphasize either the Asian or
female aspect of the applicant’s identity. When the applicant’s Asian identity,
FROM PREJUDICE TO DISCRIMINATION 397
rather than her female identity was salient, participants gave her a stronger hiring
reconunendation and recommended a 12 percent higher starting salary.
Intergroup Respect. One explanation for organizational diSCriminationll
then, is stereotype fit: Because the stereotype of White lllen provides a better
fit to the managerial stereotype than do the stereotypes of minority group
members or women, White men are selected over equally, or even better,
qualified members of other groups. Lynne Jackson, Victoria Esses, and
Christopher Burris (2001) have proposed a different explanation. They have
hypothesized that it is not a decision maker’s group stereotypes that primarily
affect discriminatory decisions, but rather it is the amount of respect the decision
maker has for the group that affects decisions. They define respect as “feelings
of esteem for another that manifest in both valuing the person’s feelings,
thoughts, and behaviors as well as willingness to be influenced by that person”
(pp. 48-49). Jackson and her colleagues propose that people hold different degrees
of respect for various groups based on the amount of power the groups
have in society, For example, “differential respect for men and women may
have its origins in the social structure. In most contemporary cultures, men
continue to hold higher social status than women” .. Relatedly, men generally
still have more power than women in society (e.g., in politics and economics),
in the workplace, and in interpersonal relationships … Because many positions
occupied by men are high in status, people are likely to more frequently act
deferentially to men than to women” (Jackson et al., 2001, p. 49). Living in
such a social context, they argue, leads people to develop more respect for men
as a group than for women as a group (although, of course, individual women
can be respected more than individual men).
Jackson and her colleagues (2001) conducted three studies to test the hypothesis
that respect outweighs stereotypes in affecting hiring decisions. In the
first two studies, research participants read a description of a gender-neutral job
and an application for the job from a man or a woman. Participants rated the
applicant on the degree of respect they felt for him or her, the extent to which
male and female stereotypes described the applicant, and how suitable the applicant
was for the job. The researchers found that although both respect and
stereotypes had positive correlations with job suitability ratings, respect had a
much stronger relationship. In the third study, the researchers manipulated both
how respected the applicant was by a previous employer and the applicant’s male
and female stereotypic characteristics using a letter of recommendation in the
applicant’s file. They also manipulated whether the applicant was applying for a
high or low status job. They found that greater respect led to stronger hiring
recommendations, especially for the high status job, and that having more masculine
stereo typic traits led to stronger hiring recommendations for both types of
job; feminine stereotypic traits were not related to hiring recommendations. As
in the first two studies, respect had a stronger effect than stereotypes. Taken together,
these findings indicate that the effect group membership has on hiring
goes beyond how decision makers think about groups (stereotypes) to include
how they feel about groups (respect). This distinction is important because it
398 (HAPTER 10
suggests that eliminating reliance on stereotypes may not eliminate discrimination
that is rooted in differences in respect for different social groups.
Shifting Standards. Think back to two of the sets of research results on gender
discrimination in organizations that we presented earlier. First, on the average,
women and ll1en receive equal job performance evaluations. Second, on the
average, women are less likely to get promoted than men. Given that rewards
such as promotions are supposed to be based on performance, these two sets of
findings appear to contradict one another: given equal performance, women and
men should be promoted at equal rates. One explanation for this apparent contradiction
lies in Monica Biemat’s (2003) shifting standards model of evaluation
that we introduced iu Chapters 3 and 4. Recall that the model proposes that
negative stereotypes lead people to hold lower performance expectations for
women and members of minority groups. When evaluators use subjective criteria
to rate performance, a person is rated relative to the expectations the evaluator
has for the person’s group. Because most workplace perfonnance evaluation
measures use subjective rating scales (see, for example, Murphy & Cleveland,
1995), job perfonnance ratings are vulnerable to this bias. For example, a woman
-manager’s perfonnance would be rated relative to the evaluator’s expectations for
woman managers as a group. The top section of Figure 10.2 illustrates this process.
Janet and Jason work for the same manager and perform at the same level.
However, because their manager has lower expectations for female employees
(that is, the female standards are shifted to the left of the male standards in this
example), the same level of performance results in a higher rating for Janet than
for Jason.
The bottom section of Figure 10.2 provides a hypothetical example of how
shifting standards can also influence the interpretation of performance ratings. In
this example, Jamal, who is Black, and Jerry, who is White, work for the same
manager, who gives them both a (very good) rating of 4 on a 5-point scale. But
their manager has, probably unconsciously, rated Jamal and Jerry relative to the
expectations he has for the performance of Black and White employees. That is,
he saw Jamal’s performance as “very good for a Black employee” and Jerry’s as
“very good for a White employee.” However, because of the different standards
used for rating Black and White performance, Jamal actually scores lower on an
objective common scale that takes both race-based scales into account. Biernat
(2003) notes that it is the objective comnlon scale) not the race-based subjective
scales, that determines the distribution of organizational rewards such as promotions
aud pay raises. Thus, although both Jamal and Jerry received ratings of 4 on
their annual performance evaluations, Jerry is more likely to get a promotion or
pay raise because his 4 translates to a 7 (very good) on the objective scale whereas
Jamal’s 4 translates to a 4 (average).
Recall, for example, the study conducted by Monica Biernat and Theresa
Vescio (2002) that we desctibed in Chapter 4, in which participants took the
role of the manager of a coed softball teanl. Even when male and female players
were evaluated as equal in athletic ability, men were more likely to be chosen
as starting players and to be placed in the top of the batting order. Thus, men
FROM PREJUDICE TO DISCRIMINATION 399
Objective performance:
0 2 3
“Female” scale:
0 2 3
“Male” scale:
0
“Black” scale:
0 2
“White” scale:
Underlying common scale:
0 2
FIGURE 10.2 Shifting Standards of Evaluation
Janet
Jason
4
4
Jamal
3 4
0
3 4
5
5
2
5
2
5
6 7 8
3 4 5
Jerry
3 5
6 7 8
Top Section: Janet and Jerry perform at the same objective level, but because the evaluator’s expectations for female
performance are lower than his expectations for male performance, Janet gets a higher performance rating. Bottom
Section: Jamal gets a high rating of 4, but it is relative to the low expectations the rater has of Black employees. Jerry
gets a high rating of 4, but it is relative to the higher expectations that the rater has for White employees. As a result,
when the two ratings are transformed to a common scale, such as if the rater had to rank employees, Jerry comes out
ahead of Jamal.
SOURCE: Adapt-ed from Biernat (2003, Figure 1. p. 1021).
received more rewards than women, presumably because shifting standards put
them higher on the objective scale the participants used to make their choices.
These kinds of differential decisions are made not only by evaluators but also by
other people who use the evaluations to make decisions. For example, Biernat
(2003) reported the results of a study in which participants read a letter of recommendation;
half the participants thought it was written on behalf of a Black
job applicant and half that it was written on behalf of a White job applicant. The
participants interpreted the letter to mean that, in objective terms, the Black
400 CHAPTER 10
applicant had perfonned less well on the job described in the letter than had the
White applicant. Participants apparently assumed that the letter writer had used
lower standards to evaluate the Black applicant’s performance.
Contemporary Prejudice. In Chapter 6 we noted that, because of contemporary
social norms that condemn prejudice, prejudice and discruuination tend to
manifest themselves in subtle ways and in situations in which prejudiced behavior
can be attributed to other causes, Thus, for example, a prejudiced person could
claim that she voted against a Black political candidate not because he was Black,
but because his platform was too liberal. In the em_ployment context, a prejudiced
employer may use applicant characteristics as a reason to reject a Black
job applicant even while hiring a White applicant with the same characteristic.
For example, Gordon Hodson, John Dovidio, and Samuel Gaertner (2002) asked
White college students who had been classified as being high or low on racial
prejudice to make college admission decisions about Black and White applicants.
The researchers created two mixed qualification applicant conditions, a clearly
high-qualification applicant condition, and a clearly low-qualification applicant
condition. In one mixed-qualification condition, the applicant had high
Scholastic Assessment Test (SAT) scores but a low high school grade point average
(GPA); in the other mixed qualification condition, the applicant had a
high GPA but low SAT scores. The higb qualification applicant had both high
SAT scores and a high GPA; the low qualification applicant had both low SAT
scores and a low GPA.
Both high- and low-prejudice participants accepted the highly qualified
Black and White applicants at the same rate (100 percent) and rejected the
poorly qualified Black and White applicants at the same rate (69 percent).
However, high-prejudice participants were more likely to accept the mixedqualification
White applicant (74 percent) than the mixed-qualification Black applicant
(44 percent); the reverse was true for low-prejudice participants although
the difference was not as large: 86 percent accepted the Black applicant and
64 percent accepted the White applicant. Participants also rated how much the
applicant’s SAT scores and GPA influenced their decisions. The high-prejudice
participants who had evaluated the mixed qualification Black applicant rated
whichever piece of negative infonnation they had seen-SAT scores in the one
condition and GPA in the other condition-as most influential. Low-prejudice
participants showed the opposite pattenl, focusing on whichever piece of information
was more positive. Thus, both high- and low-prejudice participants
seized on the infonnation that was consistent with their racial attitudes-positive
for those low in prejudice and negative for those high in prejudice-and used
that infonnation to justify- their decisions.
Employers may also use business-related factors to justifY discrimination, the
most common being maintaining workplace hannony and placating customers
(Brief, 1998; Kirschenman & Neckennan, 1990). Kirschenman and Neckennan
(1990) found that many of the employers they interviewed were reluctant to
hire minority workers because they thought doing so would upset their White
employees, leading to morale and productivity problems. This reluctance was
FROM PREJUDICE TO DISCRIMINATION 401
reflected in hiring practices: Employers who said they valued teamwork highly
were twice as likely to have racially homogeneous workforces than those who
thought that teamwork was less important. Unfortunately, these employers’ concerns
may have a basis in reality: lower-level White workers, for example, show
more work avoidance as the proportion of minority workers in their work units
increases (Tsui et al., 1992). Kirschenman and Neckerman (1990) also noted that
some employers believed that they would lose customers if they hired minority
workers. They quoted one restaurant owner as saying, “I have all white waitresses
for a very basic reason. My clientele is 95 percent white. I simply wouldn’t
last very long if I had some black waitresses out there” (p. 220). This concern
may also be based in reality: A White suburban restaurant owner who had hired
Black wait staff because he could not find enough White workers received comments
from his White customers such as “Why do you have those people out
here?” (p. 220, emphasis in original). Customer demand can also work in the
other direction: Harry Holzer (1996) found that employers with a predominandy
Black customer base were more likely to hire Black workers than White workers.
At colleges and universities, important “customers” include alumni, particular
those who donate to the university. Michael Sagas and George Cunningham
(2005) suggest that the possibility of alienating White alumni boosters may be
one reason why Blacks are less likely to be hired as head college football coaches.
Conformity to Perceived Norms. Finally, individuals might make diSCrimi~J
natOlY decisions because they believe those decisions are consistent with company
nOrms. For example, using a methodology developed by Arthur Brief and
his colleagues (Brief, Dietz, Cohen, Pugh, & Vaslow, 2000), Lars-Eric Petersen
and Joerg Dietz (2005) asked German students to imagine they were a department
head for a Gennan fast-food chain. Their task was to select three candidates
for a job interview, based on a pool that contained both foreign and
German applicants. Half of the participants read a memo from the company
president that commented on the number of foreign applicants and noted the
importance of nlaintaining a homogeneous workforce. The memo also noted
that the company employed almost exclusively Germans. The remaining participants
read no such memo. In addition, participants were categorized as either
high on a measure of subtle prejudice, high on a measure of blatant prejudice,
or nonprejudiced. Individuals high on subde prejudice were less likely to select
foreign applicants when they saw the memo than when they did not. Ratings of
nonprejudiced and blatandy prejudiced participants were unaffected by the
memo. These findings suggest that individuals who are blatandy prejudiced do
not need a justification to discriminate: these students were less likely to select
foreign applicants regardless of whether they saw the memo. Consistent with
findings from previous research (see Brief et al., 2000), those who were subdy
prejudice made discriminatory decisions when they believed that doing so would
please their boss. By the way, if pressure from a company president to discriminate
seems far-fetched, see Box 10.5, which describes the event that led to Brief
and colleagues’ research.
402 CHAPTER 10
In 1992 the restaurant chain Shoney’s Incorporated
paid $132.5 million to settle an employment
discrimination lawsuit. The evidence in the case
revealed a long-standing policy of minimizing the
number of Black employees in the company, especially
in customer-contact jobs. For example, 75 percent of
Shoney’s Black restaurant workers held minimum-wage
jobs such as dishwasher, cook, and breakfast-bar
attendant (Watkins, 1993). These employment policies
were a direct reflection of Chief Executive Officer Ray
Danner’s views. For example, one former Shoney’s vice
president described what he called Danner’s Laws:
“Blacks were not qualified to run a store. Blacks were
not qualified to run a kitchen of a store. Blacks should
not be employed in any position where they would be
seen by customers” (Watkins, 1993, p.427).
Danner’s justification for his policies was that
White customers did not want to see Black employees
and would not patronize restaurants that employed
Black customer service staff. In pretrial testimony
Danner said, “In looking for anything to identify why
[a restaurant] is under~performjng in some cases, I
would probably have said that this is a neighborhood
of predominantly white neighbors, and we have a
considerable amount of black employees and this
might be a problem” (Watkins, 1993, p. 427). Steve
Watkins (1993) reported that “the smoking gun in the
case came in the form of a letter Danner wrote
complaining about the performance of [one
restaurant] and comparing the racial makeup of the
store, which had several black employees-some of
whom were later fired-to the all~white, or nearly
all~white, composition of other fast~food restaurants
Danner visited in the area” (p. 427).
When executives from company headquarters
visited restaurants, they would tell managers whom
they thought had hired too many Black workers “to
‘lighten up’ their store-a company euphemism for
reducing the number of black workers-and hire
‘attractive white g’lrls’ instead” (Watkins, 1993, p. 424).
In another instance, “two black Shoney’s employees said
they were ordered by their manager to hide in a
restroom because some company executives had shown
up for a surprise visit and there were ‘too many’ blacks at
work that day” (Watkins, 1993, p. 426). These anti~Black
policies did not affect only Black employees: Watkins
reported that White restaurant managers who
disobeyed orders to “Iighten up” their staff were fired.
HATE CRIMES
Hate crimes (also called bias criines) are, perhaps, the most severe fonn of dis- !Crimination. Hate crimes are criminal offenses in which there is evidence that
the victims were chosen because of their race, ethnicity, national origin, religion,
disability, or sexual orientation (U.S. Federal Bureau of Investigation
-[FBI], 1999). Although whether a crime is caused by bias rather than some other
motive (such as personal animosity unrelated to prejudice) is sometimes a matter
of judgment, in many cases the evidence is fairly obvious. For example, a survey
of gay, lesbian, and bisexual hate crime victin1S found that in 53 percent of the
cases the offender made an explicit statement about the victim’s sexual orientation
(Herek, Cogan, & Gillis, 2002). In addition, hate crimes usually have no
motivation other than attacking a member of a particular group. “There appear
to be no gains for the assailant: There is no attempt to take money or personal
items and there is no prior relationship between the victim and offender” that
could provide a personal motive, such as revenge (McDevitt, Levin, & Bennett,
2002, p. 304).
Table 10.2 presents some of the characteristics of the hate crimes reported
to the FBI in 2006. The FBI received 7,720 reports of hate crimes that year
FROM PREJUDICE TO DISCRIMINATION 403
TAB L E 10.2 Characteristics of Hate Crimes
Reported in 2006
Victim Group
Racelethnicity (65.6% of all hate crimes)
Black
White
Hispanic
Asian
American Indian
Multiple races, group
Religion (18.1 of all hate crimes)
Jewish
Islamic
Catholic
Protestant
Other
Sexual orientation (15.3% of all hate crimes)
Male homosexual
Female homosexual
Heterosexual
Other
Disability (1 % of alJ hate crimes)
Mental
Physical
Race of Offender
White
Black
Unknown
Type of Offense
Against Persons (56.5% of all hate crimes)
Aggravated assault
Simple assault
Intimidation
Other violent crime
Property (43.2% of all hate crimes)
Vandalism
Burglary or theft
Arson
Other
SOURCE: u.s. Federal Bureau of Investigation (2006).
52.7%
16.7%
12.9%
3.8%
1.2%
12.7%
65.4%
11.9%
4.9%
3.7%
14.2%
62.0%
13.7%
2.0%
29.9%
77.8%
22.1%
58.6%
20.6%
12.9%
21.6%
31.9%
46.0%
0.3%
80.4%
16.9%
1.2%
2.7%
404 CHAPTER 10
involving 9652 victims, which is probably fewer than the true number given that
most crimes of all kinds go unreported (Strom, 2001). fu Table 10.2 shows, most
hate crimes were based on the victims’ race or ethnidty. However, one does not
have to be a member of a minority group to be a victim of a hate crime: in 2006,
16.7 percent of the victims were White. The majority of the crimes (56.5 percent)
were against persons, such as assault or intimidation; 43.2 percent were
crimes against property, such as vandalism, burglary, or a motor vehicle theft. A
study of hate crimes reported from 1997-1999 showed that most victims were
male (62 percent) and young (65 percent were under 35 years of age), most of
the crimes were violent (60 percent), and 18 percent involved the use of a deadly
weapon (Stonn, 2001). Jack Levin and Jack McDevitt (2002) have further noted
that hate crimes are excessively brutal compared to crune in general: 58 percent
of hate crimes involve assaults compared to 7 percent of all crimes and 30 percent
of hate crime assault victims receive physical injuries compared to 7 percent
of victims for assaults in general.
In this section, we discuss three aspects of hate crimes. First, we look at some
of the characteristics of hate crime offenders. We then examine the motivations
that offenders have for taking part in hate crimes. We conclude with a brief discussion
of the effects of hate crimes on the victims.
Hate Crime Offenders
Who conunits hate crimes? There are two ways of looking for answers to this
question. One approach is to examine victim descriptions of offenders, such as
those contained in the reports of hate crimes collected by the FBI. Kevin
Strom’s (2001) analysis of hate crimes reported from 1997~1999 showed that,
like most offenders, hate crime perpetrators are disproportionately male (84 percent)
and young: 62 percent are under 24 years of age and 79 percent are under
35 years of age. Vicky Kielinger and Susan Paterson (2007) examined London
police records of hate crimes based on racial or sexual orientation bias; their analysis
showed that two thirds of the suspects were known by the victim, either as
neighbors, colleagues, or school mates. As we will discuss, this is consistent with
research showing that hate crimes are generally not comrnitted by members of
hate groups, but instead tend to be opportunistic crimes.
Another approach to detennining offender characteristics is to conduct
surveys and examine the characteristics of people who admit to having participated
in hate crimes. Karen Franklin (2000) investigated anti-gay behavior and
found that 10 percent of her sanlple of 489 community college students admitted
having assaulted a lesbian or gay man (or someone the respondent thought was
lesbian or gay) and that an additional 24 percent admitted engaging in verbal
abuse. Like Strom, Franklin found that offenders were disproportionately male.
(She did not report results by age of respondent.) Franklin also found that men
were increasingly likely to be offenders as the violence of the behavior increased:
Men were the perpetrators in 64 percent of name-calling incidents, 79 percent of
intimidation incidents, and 92 percent of physical attacks. In addition, Franklin
found that, relative to their representation in her sample, Mrican Americans were
FROM PREJUDICE TO DISCRIMINATION 405
more likely than members of other racial groups to admit to having engaged in
anti-gay behaviors: African Americans constituted 19 percent of offenders versus
12 percent of the sample. African Americans were also more likely to admit to
physical violence than members of other groups: 29 percent versus 4 percent to
11 percent for other groups.
Although one might think that hate crime offenders hold extremely negative
intergroup attitudes or are unusually aggressive people, that is not always the
case, as we will see shortly, Also, as we saw in Chapter 9, very few hate group
members commit hate crimes. What factors, then, motivate people to commit
hate crimes? The next section addresses that question.
Motivations for Hate Crimes
Why do people commit hate crimes? Based on an analysis of 169 hate crimes
reported to the Boston police in 1991 and 1992, Jack Levin and Jack McDevitt
(2002) proposed four motivations for engaging in hate ctimes: thrill seeking, defending
one’s “territory” from outsiders, retaliation for a perceived offense committed
by an outgroup member, and a conunitment to a racist, anti-gay, or other
bigoted ideology. Franklin (2000) has identified a fifth motivation, peer group
dynamics, that Levin and McDevitt had included as part of thrill seeking
(McDevitt et al., 2002). Bryan Byers and Benjamin Crider (2002) have suggested
that seeing hate crime as a nonnal behavior may facilitate, if not motivate, such
behavior. Franklin (2000) also examined people’s reasons for not catrying out
anti-gay behaviors. Let us look at these motives.
Thrill Seeking. Thrill seeking as a motivation for a hate crime entails cOurmit-J~
ting the crime out of a desire for excitement, as an antidote for boredom. For ,
example, McDevitt and his colleagues (2002) noted that young people who had
been arrested for hate crimes “often told police that they were just bored and
looking for some fun …. The attack in these thrill-motivated cases was triggered
by an immature desire to display power and to experience a rush at the expense
of someone else…. Several of these young offenders revealed that their only
benefit from the attack was some vague senses of their own importance: a sadistic
high as well as bragging rights with their friends” (pp. 307-308).
Thrill seekers generally have little commitment to bias and often express little
animosity toward the group whose members they have attacked (Byers, Crider, &
Biggers, 1999; Franklin, 2000; McDevitt et al., 2002). Rather, they are bored and
see violence as a way of alleviating their boredom. For exanlple, Byers and Crider
(2002; Byers et al., 1999) interviewed young men who, as teenagers, had participated
in criminal and harassing behavior against the Amish residents of the rural
county in which they and their Amish victims lived. These offenses are so common,
and so commonly accepted by the non-Amish, that there is a local word for
them, daping. The role played by boredom in hate crimes is illustrated by the comments
of one of Byers and colleagues’ (1999) respondents, who offered this explanation
for his behavior: “It was what our friends were doing at the time to pass the
sununer months away or whatever. That was what we were doing on Friday
406 CHAPTER 10
nights” (p. 85). The results of Byers and colleagues’ (1999; Byers & Crider 2002)
interviews suggest that a lack of respect rather than animosity might be the emotional
facilitator of thrill seekers’ behavior. One of their respondents told them, “I
just had the mentahty that they are just Amish …. It is like, we can pick on them
because they are so different” (Byers et al., 1999, p. 87) and another said, “It is
because I still have some feeling that they ahnost ‘deserved it’ for some unknown
reason because they are different” (p. 88).
Thrill seekers tend to choose as targets people they see as providing easy and
safe opportunities for violence. For example, an offender who, along with a
friend, targeted gay men as robbery victims said, “It wasn’t because we had
sonlething against gays, but because we could get SOBle money and have some
fun. It was a rush. A serious rush. Massive rush … , It was nothing at all against
gays. They’re just an easy target. Gays have a reputation that they can’t fight
[back]” (Franklin, 1998, p. 14, emphasis in original). Similarly, one of Byers
and Crider’s (2002) respondents said of the Amish, “They are an easy target.
They offer an easy target because they ‘turn the other cheek’ and don’t fight
back” (p. 131). Another characteristic that makes a group an easy target for thrill
seekers is an unwillingness or inability to report the crime. For exanlple, one of
Byers and Crider’s (2002) interviewees noted that the Amish “can’t call the cops
[because of their rejection of modern technology, such as telephones] and don’t
believe in suing” (p. 135). Similarly, lesbians, gay men, and bisexuals may be
seen as easy targets because they are reluctant to report hate crilnes due to concern
over police harassment or public disclosure of their sexual orientation
(Herek et aI., 2002).
Thrill seekers often justify their actions by minimizing the crime’s impact on
the victims and by portraying their actions as hannless fun (Byers et al., 1999;
Franklin, 1998). One of Byers and colleagues’ (1999) respondents said, “It was
all, I always thought clean fun …. We always looked at it as there are a lot worse
things that we could be doing” (p. 85). Byers and his colleagues found that denying
that they had hurt anyone was a common justification offenders gave for
their behavior. One of their respondents said about destroying an outhouse, “No
one ever really got hurt, and it wasn’t really that much property damage. It was
pretty much just a mess to clean up” (p. 85). Besides, one respondent explained,
claping causes no real injury because the Amish should expect to be harassed:
“Stuff like that happens to them. It happens to them all the time. They are
used to it I think” (Byers et al., 1999, p. 86).
[
Territorial Defense. In defensive hate crimes, the perpetrators see themselves
as protecting their own territory fronl invasion by outsiders. The purpose of this
~type of hate crime is to coerce the outsiders to go away and to send a more
general message that members of the victim’s group are not wanted in the offenders’
neighborhood. For example, Donald Green, Dara Strolovitch, and Janelle
Wong (1998) found that the incidence of hate crimes was higher in all-White
neighborhoods into which minority group members were moving compared to
similar neighborhoods which remained all White. Defensive hate crimes constituted
25 percent of Levin and McDevitt’s (2002) sample of cases.
FROM PREJUDICE TO DISCRIMINATION 407
Retaliation. In retaliatory hate crimes, the offenders are seeking revenge for 0-real
or rumDred attack on a member of their ingroup, McDevitt and his colleagues
(2002) note that “whether the [attack] actually occurred is often irrelevan .
Sometimes a rumor of an incident may cause a group of offenders to take vengeance,
only to learn later that their original infonnation was merely unfounded
hearsay” (p. 309). Although retaliatory attackers cite revenge as the reason for
their actions, they usually do not seek out the person they believe committed
the offense against their group, but target any available member of the group.
This is consistent with the vicarious retribution model, discussed in Chapter 9, which
explains how groups decide when to retaliate against an outgroup (Lickel, Miller,
Stenstrom, Denson, & Schmader, 2006).
This kind of generalized retaliation may be especially likely to occur when the
real target of the offenders’ anger is out of their reach, As Levin and McDevitt
(2002) noted, “After [the September 11, 2001, terrorist attacks], what made it especially
tempting to target college students who spoke with an accent and had a dark
complexion was the ambiguity in identifYing the real enemy. President Bush
blamed Osama bin Laden, a shadowy figure who resided in a far-off land and had
been seen only a few times on videotape, For most Americans, bin Laden ‘” was
an abstraction, little more than a caricature. It was therefore far more satisfying
psychologically to target flesh-and-blood human beings [who were close at hand]
-international students” (p, 5). Retaliatory hate crimes constituted 8 percent of
Levin and McDevitt’s (2002) sample of cases.
Mission. Mission-motivated hate crimes are carried out because of a commit”::-l
ment to a bigoted ideology. In these kinds of crimes, “the perpetrator seeks toJ
rid the world of evi!” (McDevitt et al., 2002, p. 309). Some mission-oriented offenders
are members of hate groups, although they may be acting without the
knowledge or support of the group’s leadership. As we saw in Chapter 9, the leadership
of many hate groups publicly oppose violence, seeing it as a threat to their
recruitment efforts. Other mission-oriented offenders act on their own, seeing
themselves as victims of conspiracies by groups against whom they seek revenge
(Levin & McDevitt, 2002). Mission hate crimes are extremely rare; they constituted
less than 1 percent of Levin and McDevitt’s (2002) sample of cases.
Peer Group Dynanrics. Many hate crimes, especially thrill- and defense-J-motivated
crimes, are committed by groups of offenders, almost always young
men who know one another (Levin & McDevitt, 2002), so peer group dynamics
can play an important part in motivating participation in these crimes (Byers
et aI., 1999; Franklin, 1998, 2000; McDevitt et al., 2002). As Kathleen Blee
(2007) notes, to outsiders the violence can appear “pointless, irration~ or the
product of -immaturity or personality disorders” but to group members may convey
“a sense of strength, inviolabiliry, purpose, and agency” (pp. 263-264) that promotes
group afflliation and solidarity. Offenders motivated by peer group concerns
act out of “the desire to feel closer to friends, to live up to friends’
expectations, and to prove toughness and [in the case of anti-gay crimes] heterosexuality
to friends” (Franklin, 2000, p. 347). As this emphasis on toughness and
408 CHAPTER 10
heterosexual masculinity suggests, Franklin (1998) noted that men are more
strongly tllotivated by this factor than are women, at least in the anti-gay context
(see also Chapter 12). The group-centered nature of some hate crnnes is illustrated
in the interviews Byers and his colleagues (1999) conducted with perpetrators
of hate crimes against the Amish: “When asked if a person were to clare
alone, subjects responded that the person would have to be ‘sick’ to do such a
thing” (p. 84). Another said, “It was a kind of male bonding …. It kind of drew
us all closer because we went out and did something” (p. 89).
Like thrill seekers, peer-motivated hate crime offenders exhibit little animosity
toward their victims’ groups, but also exhibit little respect for them (Byers et aI.,
1999; Franklin, 1998). However, unlike thrill seekers, they sometimes do acknowledge
that the victim was hanned, but tend to minimize their personal responsibility.
Instead, they portray themselves as having had little choice in the matter (Byers et al.,
1999; Franklin, 1998). For example, some of Byers and colleagues’ (1999) respondents
blamed their behavior on peer pressure or local nonns, giving explanations
such as “The harassment was ahnost conunon nature” and “[It] is because of the
way I was raised” (p. 92).
McDevitt and his colleagues (2002) note that many offenders who act on
the basis of peer group concerns may be unwilling participants in the crime:
They do not approve of violence (or, perhaps, even prejudice), but go along
with the group because they feel that if they do not they will lose the approval of
their friends. In many instances, unwilling participants do not actively take part in
the crime, but also do nothing to prevent it or stop it once it has begun and are
unwilling to provide infonnation to authorities afterwards (see also Byers et al.,
1999). Thrill-motivated hate crimes were the ones most commonly found in
Levin and McDevitt’s (2002) Boston sample, constituting 66 percent of tlie cases.
Normalization. Based on their interviews with participants in anti-Amish hate
crimes, Byers and Ctider (2002) liave suggested that one factor tliat facilitates, if
[
not motivates, such crimes is community acceptance of such behavior. That is,
the community in which the offenders live views such actions as nonnal behavior
and so do not strongly condenm it, try to prevent it, or punish it. Byers and
— Crider (2002) noted that “victimization of the Amish was sometimes considered
a rite of passage [in the local non-Amish community] … ‘ If people believe that
claping is hannless, there is a lower likelihood of intervention from parents,
teachers, or criminal justice officials” (pp. 133-134). This attitude was reflected
in comments made by Byers and Crider’s (2002) interviewees. For exalnple,
[Claping] is socially acceptable here …. [Interviewer: In this community?]
Extremely, in this community. They just pretty much shrug and it is “boys
growing up.” If I lived here and I had kids and they were 16 or 17, then it
would not surprise me if I just said it is just “kids growing up.” … I think that
most people would see it as something as I will turn my back to it and pretend
that it is not there. Like when people are watching t.v. and you see the starving
kids, everyone just turns the channel. It is the same thing …. If something bad
really did happen like say somebody got killed or whatever I would say that the
Although the exact number of lynchings of Blacks that
occurred in the United States in the 19th and
20th centuries remains unknown, estimates range
from 2,500 to over 3,400 (see Leader, Mullen, &
Abrams, 2007). As we have seen throughout this text,
history is replete with examples of such violence
against outgroups and, at one point in U.S. history,
lynchings were, if not sanctioned, certainly
overlooked. Consider the now well-known lynchings
of two Black teenagers that occurred in Marion,
Indiana in 1930. These young men were accused of
raping a White woman and killing her boyfriend.
Although the teens had confessed and were awaiting
trial, rumors spread that they would be let off easy;
in response, family members and their supporters
stormed the jail. The teens were, one at a time,
dragged to the courthouse square and lynched.
Evidence that such actions were normalized at that
time in U.S. history comes from James Madison’s
(2001) analysis of the crowd’s response. He writes:
[Police] reported a scene of peace and remarkable
good humor. One of the mob who had helped
with the rope went with his young wife to a
nearby restaurant for a late dinner. But the crowds
did not disperse. People milled around through the
night. A woman nursed her baby, Fathers held up
older children to see the two bodies, .. Newcomers
kept arriving, including the youth group from
Antioch Methodist Church”. Cars [were] parked at
all angles, jammed together like matchsticks
around the Courthouse Square, . ,Souvenir collectors
cut pieces of clothing from the two bodies and
bark from the lynching tree (p. 10).
These actions continued through the night.
In fact, although the second lynching occurred at
FROM PREJUDICE TO DISCRIMINATION 409
10:30 p.m., it was 5:45 the next morning before the
crowd would allow the bodies to be cut down,
What conditions lead people to participate in such
savage cruelty? To examine this question, Tirza Leader,
Brian Mullen, and Dominic Abrams (2007) studied
reports of 300 lynchings that occurred in the United
States between 1899 and 1946. Results showed that as
crowd size increased, lynchings became more savage,
perhaps because bystanders felt anonymous in the
larger groups, /ls the crowd size increased, norms
seemed to shift, allowing the display of antisocial
behavior. Leader and her colleagues also found that
atrocities were greater when the victim’s alleged crime
was more severe, perhaps providing further
justification for the group’s cruelty, The researchers
also considered the possibility that increases in the
Black population fed White people’s desire to protect
their territory; however, they found no evidence
that the severity of the lynchings increased in those
areas with higher proportions of Blacks, suggesting
that people were not responding to population
changes.
As Leader and colleagues (2007) note, lynching in
the United States has come to define a certain time
in the past and a culture that was less sophisticated
and less educated” (p. 1340). Even at the time, not
everyone accepted lynching as normal behavior. As
Madison (2001) notes in his report on the Marion
lynching, the teens’ jailors tried desperately to prevent
the lynchings and, although the details are not
certain, a voice of reason, shouting from the crowd,
was able to stop the lynching of a third teen.
Moreover, some people in the crowd were horrified
by what they witnessed and some were physically ill.
Then, as now, acceptance of violent group behavior is
far from inevitable.
community would be, “That is a shame and all that,” but then the entire town
would be like “Oh well.” I really think it would be that way (pp. 136-137).
Claping is so acceptable in the community where Byers and Crider conducted
their interviews that one of the researchers was invited out on a claping
expedition (he declined the invitation). Box 10.6 describes the normalization of a
more severe hate crime, the lynching blacks in 19th and 20th centuries.
Inhibiting Factors. It is often as informative to understand why people do n~
engage in a behavior as why they do engage in it. A unique aspect of Franklin’~
(2000) study is that she asked the nonoffenders in her sample why they “have
410 CHAPTER 10
never harassed or beaten up a homosexual” (p. 353). She identified a number of
reasons, although she did not report the number of people giving each type of
reason. Some people’s responses implied that they refrained from anti-gay behavior
because of a lack of opportunity, saying they avoid homosexuals or never see
homosexuals. Others reported a fear of negative consequences, such as getting
into trouble with authorities or peer disapproval. Some respondents cited a belief
in nonviolence or other religious or moral beliefS. A final reason was knowing
someone who was homosexual. Franklin also noted that these restraining beliefs
may be fairly weak and so may easily break down under pressure. For example,
38 percent of her respondents who had never engaged in anti-gay behavior
“reported some likelihood to verbally or physically assault a homosexual who
flirted with or propositioned them” (Franklin, 2000, p. 353).
Effects on Victims
Surveys conducted in different parts of the United States over a period of 10 years
have provided an unusually consistent set of results regarding the effects hate
crimes have on their victims: Hate crune victims suffer more severe psychological
consequences from their victimization and these negative effects last longer
compared to victims of similar crimes that were not motivated by bias (Ehrlich,
1999; Herek, Gillis, & Cogan, 1999; McDevitt, Balboni, Garcia, & Gu, 2001).
Moreover, hate crime victims who experienced severe violence, such as sexual assault,
reported greater psychological distress than those who were threatened but
did not experience physical violence (Rose & Mechanic, 2002). Table 10.3 lists
SOille of the outcomes that hate Crlll1e victims experience to a greater extent than
victims of other crinles. In addition, cOlllpared to victims of nonbias crimes, hate
crune victims report feeling less control over their lives. One factor that helps
crime victims deal psychologically with their victimization is the feeling that they
can control what happens to them and, as a result, do things that will prevent
TAB L E 10.3 Effects of Hate Crimes on Victims
Compared to victims of similar crimes that were not motivated by bias, hate crime
victims experience more:
• Nervousness, anxiety, depression, and stress
• Intrusive thoughts about the crime
• Trouble concentrating or working
• Anger and a desire to retaliate
• Feelings of being exhausted and weak for no reason
• Fear of future trouble in life
• Distrust of people
• Fear of crime and feelings of personal vulnerability
• Difficulty coping with the effects of victimization
• Difficulty in relationship with spouse or significant other
Ehrlich, Larcom, and Purvis (1995); Herek, Gillis, and Cogan (1999); and McDevitt, Balboni, Garcia, and Gu (2001).
FROM PREJUDICE TO DISCRIMINATION 411
them from being victimized again (Davis, Taylor, & Titus, 1997). However, hate
crime victims tend to be chosen at random and so believe that there is nothing they
can do to avoid becoming a victim again: “Victims are aware that their overt actions
did nothing to precipitate their victimization; being the ‘wrong person,’ at the wrong
time and place, qualifies the bias victim [to become a victim]. Therefore, if the impetus
for victimization is something that is outside the bias victim’s control before the
incident, it is reasonable that there would be little the victim would do differently
subsequent to the incident” (McDevitt et al., 2001, p. 711). These feelings oflack
of control exacerbate the negative psychological consequences of having been a
crime victim.
Perhaps because of this, hate crime victims do not always immediately label
their experience as a hate crime, even if authorities have made that determination.
For example, Kathleen Blee (2007) interviewed members of an Islamic community
whose mosque had been destroyed by arson. Despite clear evidence that the
arson was not a random act of violence, including a note left at the arson site explicitly
stating “we hate you and we want you to leave” (p. 265), members resisted
that interpretation, at least initially. Interestingly, those with higher status positions
or who had been in the community a longer period of time were particularly
likely to dismiss the possibility that the crime was directed at Muslims. Similarly,
Blee interviewed members of a Jewish community which experienced the firebombing
of its Holocaust museum, Some initially interpreted the event as due to
anti-Semitic violence but later changed their minds; others initially looked for
other interpretations but, over time, came to believe it was a hate crime. Hence,
“[w]hether and how victims interpret violent acts as hate violence are the products
of collective and individual processes of interpretation” (Blee, 2007, p. 266).
A special characteristic of hate crimes is what McDevitt and his COlleague]
(2001; see also Ehrlich, 1999) call secondary victimization: A hate crime has
psychological effects not only on the victim but also on members of the victim’s
group. These secondary victims experience, at least temporarily, heightened anxiety
over the possibility of becoming victims themselves. Secondary victimization
is a major goal of defense-motivated hate crimes and is often a secondary goal of
others (McDevitt et al., 2001). fu McDevitt and his colleagues (2001) note,
“a cross burning not only affects the immediate family [that was victimized],
but any African American who beconles aware of the incident” (p. 698). As
the Chinese military strategist Sun T’zu said over 2,000 years ago: “Kill one,
frighten 10,000.” There are few data on the extent of secondary victimization
in hate crimes, but Howard Ehrlich (1999) reported that surveys of college students
following on-campus hate crimes have found that about two-thirds of
other members of the victim’s group experience fear of becoming victims themselves.
Paul Iganski (2007) interviewed people who, although not victims of hate
crimes themselves, saw hate-related violence in their jobs as district attorneys or
police officers. These individuals reported that hate crimes had many consequences
for the communities in which they took place, including increased anxiety,
the potential for more crime due to retaliation, and ripple effects that led
some group members to respond as if they had themselves been victimized.
Thus, hate crimes victimize not just individuals but entire social groups.
412 CHAPTER 10
SUMMARY
Discrimination consists of treating people differently, and usually unfairly, based
solely or priluarily on their membership in a social group. Discrimination is therefore
a matter of behavior (including verbal and nonverbal behavior), whereas prejudice
is an attitude that can motivate discriminatory behavior. Discrimination can
take any of three forms. Blatant discrimination is intentional and obvious. Subtle
discrimination is less visible and obvious than blatant discrimination, is often unintentional,
and derives from people having internalized discriminatory customs and
social norms. Covert discrimination is hidden but intentional, and often motivated
by malice.
Although prejudice can motivate discrimination, not all prejudiced people
discriminate when they have the opportunity and nonprejudiced people can
discriminate without intending to. A number of factors influence the relationship
between prejudice and discrimination. Prejudice is more likely to manifest
itself in discrimination when the target of the discrimination matches the prejudiced
individual’s personal stereotype of the outgroup. Implicit prejudice is
most likely to result in automatic, uncontrollable behaviors, whereas explicit
prejudice is most likely to affect controllable behaviors. Finally, people are
more likely to act on their prejudices when they believe that other people
agree with them.
Because of the egalitarian nonn that exists in modern society, most people
are motivated to control any prejudice they feel and to avoid discriminatory
behavior. However, regressive prejudice occurs when people lose control over
their prejudiced responses and act in a discriminatory manner. Thus, people can
exhibit nonverbal indicators of prejudice, which are usually not under voluntaty
control, while trying to appear unprejudiced through their controllable behaviors.
Just as high cognitive demands can lead people to apply stereotypes,
such demands can let prejudiced behaviors “leak out” by undermining control.
Alcohol consumption and strong emotions, such as anger, can also reduce control
over behavior. Because people are less motivated to conlply with social
norms when other people cannot identifY them, anonymity facilitates discriminatory
behavior, as does the actual or implied approval of authority figures.
When people can otherwise justifY their actions, such as when they decide
the cost of helping another person is too high, discrimination is more likely.
Seeing other people act in a prejudiced manner can also disinhibit prejudice.
Finally, if individuals believe that they have established their credentials as unprejudiced
people, they may let their control lapse and act in a discriminatory
manner.
Research on discrimination in organizations indicates that Black job applicants
who submit resumes are less likely to be called for interviews than equally
qualified White applicants and receive lower ratings on interview performance.
As a result, they are less likely to be hired. In contrast, there currently seems to
be little gender discrimination in hiring—-unless the applicant is pregnant.
Once on the job, there seems to be little gender bias in performance evaluations;
FROM PREJUDICE TO DISCRIMINATION 413
however, Black workers receive lower perfonnance evaluations than White
workers, especially from White supervisors. Even when numerical ratings for
Black and White employees are identical, White employees tend to get more
positive narrative comments. Black employees perfonn less well on objective
measures of job perfonnance, so evaluations might simply be reflecting that difference,
However, the lower objective perfonnance might itself be a result of
prejudice, reflecting lost opportunities, such as for additional training, and lower
morale caused by prejudice and discrimination.
Even when women and members of minority groups receive the same performance
evaluations as men and Whites, they are less likely to be promoted,
are promoted more slowly, and are more likely to end their careers at a lower
organizational level. Also, ethnic minorities can be “tracked” into certain job
categories, such as affirmative action officer, that offer fewer opportunities for
advancement. In addition, the pay differential between Black and White workers
increases as level of authority increases. Black workers may experience slower
promotions because of the sticky floor effect-Black managers tend to supervise
Black workers who are disproportionately found at lower organizational levels,
are tracked into jobs with little promotion potential, and have fewer influential
mentors to help them in their careers.
A number of individual-level processes contribute to discrimination in
organizations. The stereotype fit hypothesis holds that women and members
of minority groups are underrepresented in managerial positions relative to
White men because the White male stereotype matches the stereotype of the
effective manager whereas the female and minority stereotypes do not, As a
result, women and minority group members are perceived as less qualified despite
their objective qualifications. Similar processes also can operate for lower
level jobs: The generally negative stereotypes of minority groups contradict the
“good worker” stereotype. Wonlen and members of minority groups also may
be excluded from prestigious jobs because, as groups, they garner less respect
than White men. The flnding that women and members of minority groups
are less likely to be promoted even when they receive the same perfonnance
evaluations as men and Whites may be a result of the shifting standards effect:
Because evaluators have lower expectations for WOlnen’s and minority groups’
perfonnance relative to men’s and Whites’, the same subjective rating translates
into a lower rating on an objective common scale that takes the race-based evaluations
into account. Rewards such as promotions are based on the common
scale.
Contemporary prejudice can lead prejudiced decision makers to put more
weight on the negative aspects of a minority group member’s qualifications
when both positive and negative infonnation is available, leading to an adverse
decision that can be justified by the negative infonnation. Employers also may
use business-related justifications, such as maintaining work group harmony and
customers’ prejudices, as justifications for not hiring minority workers, Finally,
people tend to comply with what they perceive to be the requirements of organizational
nonns and authority figures. Thus, if they perceive the organizational
414 CHAPTER 10
nonn as calling for discrimination or perceive that authority figures prefer to
have as few minority workers as possible, even low prejudiced people may discriminate
to comply with those demands.
Hate crimes are criminal offenses in which evidence shows the victims were
chosen because of their group membership. Hate crime offenders are primarily
young men. Several motives exist for hate crime. The most common seems to
be thrill seeking: People are bored and see picking on or assaulting a member of
an outgroup as a way of getting some excitement. They often have no strong
animosity toward their victlins’ groups; they choose as targets members of groups
they believe are unlikely to fight back or to report the crime. Tliey often justify
their actions by minimizing their impact on the victim or portraying their actions
as harmless fun. Defensive hate crimes are designed to drive outgroup members
from ingroup “territory” and to send a general message to other melnbers of the
victim’s group to stay away. Retaliation crimes are carried out in response to an
actual or rumored hate crune against a member of the offender’s group. Mission
hate crimes occur because of a commitment to a bigoted ideology and to rid the
world of a perceived evil. Peer group dynamics contribute to hate crimes because
offenders are often trying to impress members of the peer group, are going along
with what they see as the group nonn, or have succumbed to group pressure to
participate. Community norms also can facilitate hate crimes by viewing them as
normal behavior and refraining from disapproving of or punishing them. Factors
that can inhibit people from conunitting hate crimes include lack of opportunity,
fear of negative consequences, religious or moral beliefs, or knowing a member of
the targeted group.
Hate crimes generally have more severe psychological consequences for their
victinlS than do nonbias-motivated crimes, and those effects last longer. The
effects may be more severe and longer lasting because hate crime victims feel
that they cannot do anything to avoid being victimized in the future. However,
not all hate crime victims see the experience in the same way and how they view
the experience can change over time. Hate crimes also result in secondary victimization:
A hate crime has negative psychological effects not only on the victim, but
also on other members of the victim’s group, who experience heightened anxiety
over the possibility of becoming victims themselves.
SUGGESTED READINGS
The Relation of Prejudice to Discrimination
Crandall, C. S., Eshleman, A., & O’Brien, L. (2002). Social norms and the expression
and suppression of prejudice: The struggle for internalization. Journal 4 Personality
and Social Psychology, 82, 359-378.
Crandall and his colleagues present a set of studies on the role of social nonns in
prejudice and discrimination.
FROM PREJUDICE TO DISCRIMINATION 415
Regressive Prejudice
Crandall, C. S., & Eshleman, A. (2003). Ajustification-suppression model of the expression
and experience of prejudice. Psychological Bulletin, 129,414-446.
This article includes a comprehensive review of factors that act as releasers of
regressive prejudice, which Crandall and Eshleman refer to as justifications.
Reactions to Having Acted Prejudiced
Czopp, A. M., & Monteith, M.]. (2003). Confronting prejudice (literally): Reactions to
confrontations of racial or gender bias, Personality aHd Social Psychology Bulletin, 29,
532-544.
Devine, P. G., Monteith, M. J., Zuwerink, J. R., & Elliot, A. J. (1991). Prejudice with
and without compunction, Journal of Personality and Social Psychology, 60,817-830.
Devine and colleagues’ article describes some of the initial theory and research
on how people with unprejudiced self-concepts react to having acted in
a prejudiced manner. Czopp and Monteith examine the issue in the context
of responses to having acted in a prejudiced manner toward different groups.
Organizational Discrimination
Roberson, L., & Block, C. J. (2001). Racioethnicity and job perfonnance: A review and
critique of theoretical perspectives on the causes of group differences. Research in
Otgal1izational Behavior, 23, 247-325.
Roberson and Block discuss four models of factors that affect minority group members’
work perfonnance. Although the discussion focuses on work perfonnance, the
principles they discuss generalize to other processes, such as hiring and promotion.
Hate Crimes
Blee, K. (2007). The microdynamics of hate violence: interpretive analysis and implications
for responses. American Behavioral Scie11tist, 51, 258-270.
Ehrlich, H. J. (1999). Campus ethnoviolence. In F. L. Pincus & H. J. Ehrlich (Eds.), Race
and ethnic conflict: Contending views 011 prejudice, discrimination, and ethl1ovioience (2nd ed.,
pp. 277-290). Boulder, CO: Westview.
Levin, J., & McDevitt, J. (2002). Hate cdmes revisited: Amen’ca’s war ON those who are different.
Boulder, CO: Westview.
McDevitt,]., Levin,]., & Bennett, S. (2002). Hate crime offenders: An expanded typology.
Journal of Social Issues, 58,303-317.
Blee interviewed victims of hate crimes and discusses the factors the influence
their interpretation of the event. McDevitt and his colleagues (2002) present an
overview of Levin and McDevitt’s· (2002) typology of hate crime offenders.
Levin and McDevitt’s (2002) discuss the typology in more detail and provides a
number of case descriptions for each type. They also discuss issues involved in
policing, public policy, and prevention, among others. Ehrlich (1999) discusses
the problem of hate crime on college campuses, usually considered to be
bastions of tolerance (see also Levin and McDevitt’s Chapter 9).
416 CHAPTER 10
blatant discrimination
covert discrimination
discrimination
hate crimes
interpersonal
discrimination
KEY TERMS
organizational
discrimination
regressive racism
(prejudice)
secondary victimization
shifting standards model
social noons
stereotype fit hypothesis
subtle discrimination
QUESTIONS FOR REVIEW AND DISCUSSION
1. Define discrimination. How does discrimination differ from prejudice? How
are the two concepts similar?
2. Define the three fonus discrimination can take and give an example of each.
Review the types of contemporary prejudice we discussed in Chapter 6.
What forms of discrimination do you think those types of prejudice likely
result in?
3. Describe the factors that influence the relationship between prejudice and
discrimination. That is, under what conditions is prejudice most likely to
result in discrimination?
4. What are social nonns? How are they related to prejudice and discrimination?
What experiences have you had with social norms and prejudice and
discrimination?
5. What is regressive prejudice? Describe the factors that can precipitate it.
Have you observed any instances of regressive prejudice? If so, describe them
and explain what factors led to the release of discriminatory behavior in
those cases.
6. What is an employment audit? Do you think that employment audits are
effective tools for studying discrimination in hiring? Why or why not?
7. What has research discovered about race and gender discrimination in
hiring? What has research discovered about race and gender discrimination
in perfonnance evaluation?
8. Researchers have found that Black workers usually get lower scores on
objective measures of job perfonnance than do White workers. What is the
relevance of this finding for interpreting race differences in supervisor
evaluations, which generally have a strong subjective element?
9. What has research discovered about race and gender discrimination in
promotions? What organizational factors might contribute to these
differences?
10. Describe the costs and benefits for men who are in a female-dominated
profession.
FROM PREJUDICE TO DISCRIMINATION 417
11. What is the stereotype fit hypothesis? How does it explain race and gender
differences in hiring, perfonnance evaluation, and promotion?
12. Describe how differences in the amount of respect that different social
groups receive are related to organizational discrimination.
13. What is the shifting standards effect? How does it explain race and gender
differences in hiring, perfonnance evaluation, and promotion?
14. Explain the role contemporary prejudice plays in orgartizational
discrimination.
15. Explain the role confonnity to perceived nouns plays in organizational
discrimination.
16. If an employer believes that his White customers do not want to interact
with people of other ethnicities, does that justify his decision not to hire
non-White workers? Explain your reasoning.
17. Rather than attributing the differential outcomes miuority and female
workers experience in organizations to intentional discrimination, Smith and
Elliott (2002) wrote that ‘We believe that something more subtle and
profound occurs in the process of doing ‘business as usual’-lnere
maintenance of the status quo is more than enough to perpetuate ‘”
stratification” (p. 274). Do you agree or disagree? Explain the reasons for
your position.
18. What are hate crones?
19. Describe the characteristics of hate crime offenders.
20. Explain how thrill seeking can motivate hate crimes. Who do thrill seekers
choose as victims? How do they justify their behavior?
21. Some researchers believe that thrill-seeking hate crime offenders feel little
animosity toward their victims or their groups. Do you agree or disagree?
Explain the reasons for your position.
22. Explain territorial defense as a motivation for hate crimes.
23. Explain retaliation as a motivation for hate crimes.
24. What are mission-motivated hate crimes? Why do you think they are so
rare?
25. Explain the role that peer group dynamics play in hate crimes.
26. Explain how community attitudes can affect the occurrence of hate crimes.
27. In what ways do the psychological consequences differ for the victims of
hate crimes and those of crimes not motivated by bias? What causes these
differences?
28. Explain the concept of secondary victimization.