 

951285

(atria-"H73

This is to certify that the
dissertation entitled

MEDIA FRAMES, INTERRACIAL ATTITUDES AND
SUPPORT FOR AFFIRMATIVE ACTION

presented by
John D. Richardson

has been accepted towards fulﬁllment
of the requirements for the

 

 

 

 

Ph.D. degree in Mass Media
ij ' 7 422% .
Major Professor’s Signature
823v4
Date

MSU is an Afﬁnnative Action/Equal Opportunity Institution

.-;-c--c-o-o—o-n—-—.—-_

 

 

 

LIBRARY
Michigan State
University

 

 

 

l PLACE IN RETURN Box to remove this checkout from your record.
' TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

‘1 {DATE DUE: DATE DUE DATE DUE

 

NOV 1 o 2005

 

MM”

 

 

NOV, 0 72005

 

Emmi? idia

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DatoDuepGS—p. 1 5

MEDIA FRAMES. IN’I‘ERRACIAL ATTITUDES AND SUPPORT FOR
AFFIRMATIVE ACTION

By

John D. Richardson

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Mass Media PhD. Program

Department of Telecommunication

2004

Sill

 

COI‘

Illa

lt‘x‘t

‘iIii

(or

ABSTRACT

MEDIA FRAMES. INTERRACIAL ATTITUDES AND SUPPORT FOR
AFFIRMATIVE ACTION

By
John D. Richardson

An experiment investigated how media frames can alter the cognitive processes
by which people evaluate political issues. In particular, the joint effects of media frames
and interracial attitudes were examined in the context of Gruiler v. Bollinger (2003). a
US. Supreme Court decision upholding an afﬁrmative action policy used to admit
students to law school at The University of Michigan.

Editorials were presented in one of four randomly assigned versions, all
containing the same description of the Grutter ruling, but altering which frame(s) the
editorial presented to endorse it. The Remedial Action editorial emphasized afﬁrmative
action is necessary to redress past and present discrimination, as well as inequality in
schools in minority communities. The Diversity editorial emphasized that afﬁrmative
action is good for all students—they learn to work and get along with people different
from themselves toward the goal of becoming leaders in an increasing global
marketplace. Control (neither frame) and Combined (both frames) editorials were also
tested.

Neither editorial frame signiﬁcantly altered support for afﬁrmative action.
However. results indicated editorials presenting the Diversity frame moderated the
inﬂuence of interracial attitudes on support for afﬁrmative action. Interracial attitudes
strongly predicted support for afﬁrmative action among white participants reading

Control or Remedial Action editorials. However. among those reading the Diversity

Pit
poi
atti

Cilll

 

Iran

dClll

editorial. the relation between interracial attitudes and support for afﬁrmative action fell
sharply to an insigniﬁcant level.

Results further demonstrated the editorial frame inductions directly impacted
interracial affect. Remedial Action editorials induced white participants to score lower on
Katz and Hass‘ (1988) Pro-Black Affect Scale, compared to the other three experimental
conditions. and higher on Katz and Hass" Anti-Black Affect Scale, compared to the
control condition.

This study extends prior research demonstrating that media frames and individual
predispositions can interact. altering the cognitive processes by which citizens decide
political issues. Speciﬁcally, frames can deactivate the linkage between interracial
attitudes and political evaluations, even for an issue directly involving race. Media frames
can also alter interracial affect. An unintended consequence of the Remedial Action
frame. which traditionally has dominated much of the discourse in favor of afﬁrmative

action. could be to induce whites to view blacks less favorably.

Dedicated to the memory of my brother

Jeff

iv

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. vii

CHAPTER 1
Introduction ............................................................................................. l

CHAPTERZ
Literature Review ...................................................................................... 3
Framing... 3
Three approaches tofranie analyszs .. .....5
Frame disputes... 9
The Moderating Inﬂuence ofFrames ............lO
Aﬂirmative Action Frames... 12
Aﬂirmative action in media d1scourse ...............13
TheDiversityﬁame... . 17
Effects ofAfﬁrmative Action Frammg 21
Social Identity Theory... ...........27
Re-categorization 30
ProceduralJustice Theory 33
Proceduraljustice andsocial 1dent11y .... .....37
InterracialAttitudes... 38
Modern racism ..................................................................... 39
White guilt and belief in white privilege ........................................ 41
Racial ambivalence ............................................................... 45

CHAPTER 3
Hypotheses ........................................................................................... 47

CHAPTER 4
Method ................................................................................................ 50
Participants .................................................................................. 50
Statistical power ................................................................... 51
Procedure .................................................................................... 52
Pretest .............................................................................. 52
Experiment .......................................................................... 53
Instrumentation ............................................................................. 55
Pretest .............................................................................. 55
Experimental stimuli .............................................................. 57
Induction checks .................................................................... 59
Dependent measures .............................................................. 61

CHAPTER 5
Results ............................................................................................... 63

Participant Demographics ................................................................ 63

Interracial Attitudes ........................................................................ 63

Dependent Variables ....................................................................... 65

Tests of the Hypotheses .................................................................... 66

White participants ’ support for an afﬁrmative action policy ............... 66

The moderating inﬂuence of editorial frame .................................. 69

White participants’ affect toward blacks ........................................ 71

Additional Analyses ....................................................................... 76

CHAPTER 6

Discussion ........................................................................................... 78
APPENDICES

A. Pretest Instructions to Participants ............................................. 85

B. Experimental Instructions to Participants ..................................... 86

C-1 Experimental Stimulus for Control Framing Condition ...................... 87

C-2. Experimental Stimulus for Remedial Action Framing Condition ........... 88

C-3. Experimental Stimulus for Diversity Framing Condition .................... 89

C-4. Experimental Stimulus for Combined Framing Condition .................. 90

D. Support for an Affirmative Action Policy Scale ................................ 92

E. Intercorrelations among Variables for White Participants .................. 94

F. Intercorrelations among Variables for Black Participants. . . . . . . . . . . .95

References ........................................................................................... 96

vi

 

 

 

T1

Ta

Tal

Tat

Tab

Tabl

Tabl

Tablc’

Table 4.1

Table 5.1

Table 5.2

Table 5.3

Table 5.4

Table 5.5

Table 5.6

Table 5.7

Table 5.8

Table 5.9

Table 5.10

Table 5.11

LIST OF TABLES

Percentage of Respondents Answering Induction Check
Items in the Expected Direction as a Function of Editorial

Frame Condition ................................................................

Black and White Participants ’ Responses to Demographic Items. . . . ..

Intercorrelations and Mean Item Responses to Interracial

Attitude Scales among White Participants ..................................

Intercorrelations and Mean Item Responses to Dependent Variable

Scales among White Participants ................................ ,,,,,, .......

Mean Support for an Affirmative Action Policy among White

Participants as a Function of Editorial Frame .............................

Support for an Affirmative Action Policy among White
Participants Regressed on Scores on Interracial Attitude

Scales and Editorial Frame ...................................................

Support for an A ﬂirmative Action Policy Between Experimental
Conditions Regressed on Scores on Interracial Attitude Scales

among White Participants ......................................................

Pro-Black Affect among White Participants as a Function

of Editorial Frame ............................................................

Pro-Black A ﬂeet among White Participants Regressed on

Scores on Interracial Attitude Scales and Editorial Frame .............

Anti-Black Affect among White Participants as a Function

of Editorial Frame .............................................................

A nti—Black A jfect among White Participants Regressed on

Scores on Interracial Attitude Scales and Editorial Frame ..............

Support for an Aﬁirmative Action Policy among Black

Participants as a Function ofEditorial Frame...

vii

...60

.64

...65

....66

...67

...68

...70

....72

...73

...74

...75

77

 

 

10

Pt

 

fr

dirt

prei
(Se:
the 1
Mil
undt
(mat

the s

aUdle

CHAPTER 1

Introduction

To traverse the world men must have maps of the world.

-- Lippmann, 1922, p. 16

Media frames present preferred meanings and points of view the reader is invited
to accept—sometimes the invitation is accepted, other times not (Gamson et al., 1992).
Predicting and explaining when media frames will exert inﬂuence, and when not, is the
framing researcher’s quest. The obvious place to begin this investigation is to
systematically manipulate media frames and observe whether and to what extent they
directly induce attitudinal change. A rich body of research has followed this approach.

Prior research also suggests that media frames can interact with individual
predispositions, altering the strength or direction of their inﬂuence on ultimate attitudes
(See e.g., Domke et al., 1999; Kinder & Sanders, 1990; Nelson etal., 1997). Discovering
the cognitive processes by which media frames and individual predispositions jointly
inﬂuence evaluations of attitudinal objects holds promise toward expanded theoretical
understanding. Borrowing from Lippmann’s (1922) metaphor, different media frames
(maps) could induce audience members to follow very different routes, even if leading to
the same ultimate destination.

This study simultaneously examined the direct effects of media frames on

audience member attitudes and their moderating inﬂuence on individual predispositions.

 

 

 

 

 

th

wi

bl:

The context of the study was a controversial and important social issue: afﬁrmative
action.

The individual predispositions selected for investigation were interracial attitudes.
Race is certainly a topic about which people hold a variety of attitudes. Race is also a
topic where mass media content can be particularly inﬂuential: “The media are not only a
powerful source of ideas about race. They are also one place where these ideas are
articulated, worked on, transformed, and elaborated” (Hall, 2003, pp. 90-91). Johnson
and colleagues (1971) observed that while blacks must traverse back and forth between
the world of whites and the world of blacks, most whites have few meaningful contacts
with blacks. Accordingly, these scholars reasoned, whites’ thoughts and feelings about
blacks could be particularly susceptible to mass media inﬂuences.

Media frames examined in this study were drawn from Richardson and
Lancendorfer’s (2004) content analysis of newspaper editorial responses to Grutter v.
Bollinger (2003), a US. Supreme Court ruling on afﬁrmative action. Richardson and
Lancendorfer identiﬁed two alternate frames that editorial writers used to support the
Grutter decision: (1) the Diversity frame, emphasizing that including people from many
racial and ethnic backgrounds strengthens organizations and society, and (2) the
Remedial Action frame, emphasizing that discrimination has unfairly disadvantaged
members of certain racial/ethnic groups. Drawing on Social Identity Theory and
Procedural Justice Theory, it was predicted these frames would have differential effects

on white participants’ attitudes and cognitive processes.

CHAPTER 2
Literature Review

This chapter reviews: frame analysis; the moderating inﬂuence of frames on
individual predispositions; framing of afﬁrmative action in media discourse; the
emergence of the Diversity frame; the effects of afﬁrmative action frames; Social Identity
Theory; Procedural Justice Theory; and interracial attitudes and affect.

Framing '

Entman (1993) deﬁned framing as a process by which a communicator engages
to: “. . .select some aspects of perceived reality and make them more salient in a
communicating text, in such a way as to promote a particular problem deﬁnition, causal
interpretation, moral evaluation, and/or treatment recommendation for the problem
described” (p. 52). Kinder (1998) offered a similar deﬁnition: “Frames seek to capture
the essence of an issue. They deﬁne what the problem is and how to think about it; often
they suggest what, if anything, should be done to remedy it” (p. 172).

Goffman (1974) introduced the terms, strip and frame, to describe the process by
which people interpret complex events. A strip refers to the underlying reality, i.e., the
“raw batch” of issues and occurrences under scrutiny. However, because most situations
involve many things going on simultaneously, strips oﬁen give rise to the fundamental
question: “What is going on here?” A ﬂame is a deﬁnition of a situation (stn'p), built up
in accordance with principles of organization. Hence, frames are devices fabricated to
lend meaning to ambiguous factual realities.

Although the term framing is attributable to Goffman (1974), it was Lippmann

(1922) who introduced the underlying concept: “For the most part we do not ﬁrst see, and

then deﬁne, we deﬁne ﬁrst and then see” (p. 81). Lippmann argued that a news story is
not a “mirror of social conditions,” but merely a “stylized account” of certain aspects of
those conditions journalists regard as interpretive:

A report is the joint product of the knower and the known, in which the

role of observer is always selective and usually creative. The facts we see

depend on where we are placed, and the habits of our eyes (p. 80).

In the context of mass media, framing is concerned with which elements receive
the most attention within a particular media artifact. Framing research presumes that: (1)
issues, events and people are multi-faceted entities capable of being described and
packaged in a variety of ways, and (2) these descriptions and packages can affect
audience members’ perceptions (Price et al., 1997). The persuasiveness of frames is
inﬂuenced by many factors, including source and message attributes (Tewksbury et al.,
2000)

Some scholars argue that frames oﬁen emerge fortuitously (Entman, 1993;
Kahnemann & Tversky, 1984). Chyi and McCombs (2004) contended media outlets
intentionally change frames over time to make their coverage of ongoing stories appear
fresh and in-depth. Others emphasize that media workers and/ or outside sources often
craft frames with speciﬁc intent to inﬂuence public opinion: “Journalists themselves may
concoct their own frames, while powerful communication agents, including elected
ofﬁcials, interest groups and lobbyists, create and promote frames with hopes of planting
them in mass-media outlets” (Nelson & Oxley, 1999, p. 1041). But whether frames are
created by media workers or borrowed from outside sources, Enzenberger (1970/2003)

observed that all media content is inherently manipulated (i.e., framed, slanted, distorted):

 

 

 

m

C\

 

 

isr

Pacl
inter

elite

gTCa]

Thus every use of the media presupposes manipulation. The most

elementary processes in media production, from the choice of medium

itself to shooting, cutting, synchronization, dubbing, right up through

distribution, are all operations carried out on the raw material. There is no

such thing as unmanipulated writing, ﬁlming, or broadcasting. The

question is therefore not whether the media are manipulated, but who

manipulates them (p. 265).

Similarly, Bennett (1993) observed that even in the context of scientiﬁc surveys,
no standard or objective format exists for phrasing questions. Every survey question (like
every media artifact) is “loaded,” i.e., framed. In sum, because all media use symbols
(words and/or images) to represent issues, events and/or people, framing in media content
is ubiquitous.

Three approaches to ﬂame analysis

D’Angelo (2002) identiﬁed three major paradigms that have shaped scholarly
analysis of framing: cognitive, constructionist, and critical. According to D’Angelo, the
cognitive paradigm is characterized by a process in which journalists negotiate alternate
interpretations of an issue, in line with the norm of balanced presentation.
Constructionists focus more on the role of journalists as creators of interpretative
packages that reﬂect and add to the public discourse about a topic. Critical scholars
interpret framing primarily as the product of hegemonic ideology (i.e., the world-view of
elite and powerful social actors).

Kahnemann and Tversky’s (1984) work exploring the concept of invariance has

greatly inﬂuenced the cognitive approach to framing scholarship. Intuitively, the

preferability of two alternatives (e.g., issue positions) should depend on objective
advantages and disadvantages, without regard to the manner in which they are described.
However, a number of impressive experiments demonstrate that frame invariance does
not hold: participants’ conﬁdence in a particular choice can change dramatically
depending on the frame employed. Frames emphasizing the risk of loss create a bias for
retention of the status quo; frames emphasizing potential gain induce more willingness to
change.

Price and colleagues (1997) used the metaphor of switching trains to explain the
cognitive approach to frame analysis. Under this viewpoint, frames increase the relative
salience of certain aspects of an issue or event. Framing devices used to accomplish this
include message organization, content selection and the use of thematic structure. Story
attributes with increased salience activate pre-existing ideas within the person, priming
the receiver to regard these attributes as applicable to the cognitive task at hand: “By
prompting the activation of certain constructs at the expense of others, frames can
directly inﬂuence what enters the mind of audience members” (p. 504). In other words,
frames can serve to direct a train of thought into the recipient’s head, albeit they do not
control it.

Several scholars (See, e.g., Domke et al., 1999; Entrnan, 1993; Gamson, 1992;
Gamson & Modigliani, 1989; Kinder & Sanders, 1990; Nelson & Kinder, 1996) suggest
that media frames are analogous to cognitive schemata—both act as central organizing
principles to give coherence and meaning to an array of inputs. Some framing theorists
suggest that frames exert inﬂuence by increasing the salience (accessibility) of some

considerations relative to others (See, e. g., deVreese, 2004; Domke, 2001; Domke et al.;

Druckman, 2001a; Iyengar, 1990; Kinder, 1998; Shen, 2004). Nelson and colleagues
(Nelson & Oxley, 1999; Nelson et al., 1997) offered an alternative conceptualization,
suggesting that framing effects are more accurately viewed as altering the relative weight
(importance) that receivers assign to competing considerations.

Shen (2004) offered another way to view the inﬂuence of frames. When message
recipients encounter frames, they do not bring a blank slate. Instead, the receiver
possesses schemata acquired through experience, some of which are chronically
accessible. Under this view, the persuasive impact of a frame depends largely on the
extent to which it is congruent with pre-existing beliefs—audience members are unlikely
to accept frames incongruent with their pre-existing beliefs.

The constructionist approach to framing draws on Carey’s (1975) ritual view of
communication, which posits that media serve primarily to represent reality in
accordance with culturally shared beliefs:

We create, express, and convey our knowledge of and attitudes toward

reality through the construction of a variety of symbol systems: art,

science, journalism, religion, common sense, mythology. . ..

communication is not some pure phenomenon we can discover; there is no

such thing as communication to be revealed in nature through some

objective method free from the corruption of culture (p. 17).

Scholars applying the constructionist paradigm emphasize that because
communication is a symbolic process, the person selecting these symbols has wide
latitude to present any number of representations of reality: “Far from being stable, the

social world is a chameleon, or, to suggest a better metaphor, a kaleidoscope of potential

realities, any of which can be readily evoked by altering the ways in which observations
are framed and categorized” (Edelman, 1993, p. 232). The constructionist paradigm
emphasizes framers’ capacity to produce accounts which are seemingly factual, but are
nonetheless instilled with value-laden undertones: “This means news stories are not so
much fact as artifact; that is, they represent the active construction of the social world by
newsworkers and newsmakers. This process of news creation is a major site of
ideological production” (Gray, 1987, p. 385).

Like the constructionist approach, the critical approach to frame analysis assumes
that historical events cannot be communicated in the “raw.” Instead, media workers must
transform them into artifacts consisting of symbols (words and/or images). However,
while the constructionist approach contends that media workers have wide latitude in this
process, critical theorists assert media workers are constrained by the dominant,
hegemonic ideology: “Any society/culture tends, with varying degrees of closure, to
impose its classiﬁcations of the social and cultural and political world. These constitute a
dominant cultural order, though it is neither univocal nor uncontested” (Hall, 1980, p.
134, emphasis original). Recognizing that the professional norms of journalism proscribe
balance and objectivity, critical theorists argue these ideals are subsumed by wider socio-
cultural and political ideological assumptions which journalists, like most everyone else,
take for granted as objective truths.

F mm the critical perspective, media discourse is viewed primarily as a reﬂection
of the values of the powerful and the elite. In line with this reasoning, Schudson (2002)
observed that: (1) the day-to-day business of journalism involves the interaction of

reporters with government ofﬁcials (with the latter group wielding much greater

inﬂuence), and (2) the media are a more important forum for communication among
elites than with ordinary citizens. Schudson further indicated that: “All of this work
recognizes that news is a form of literature and that among the resources journalists work
with are the traditions of story telling, picture making, and sentence construction they
inherit from their own cultures, with vital assumptions about the world built in” (p. 262,
emphasis supplied).

Frame disputes

Framing can be particularly important when public opinion about an issue is
divided. Goffman (1974) coined the term ﬂame dispute to describe a situation in which:
“. . .parties with opposing versions of events may openly dispute with each other over
how to deﬁne what has been or is happening ” (p. 322). Frame disputes are characterized
by fundamental disagreement as to the nature, causes, implications, and preferable
courses of action surrounding an issue (Entrnan, 1993).

Swidler’s (1986) discussion of cultural change is enlightening in this regard.
According to Swidler, culture can be viewed as a toolkit of symbols and worldviews
useful toward solving problems, organizing experience, and evaluating reality. This
toolkit includes beliefs, ritual practices, art forms, ceremonies, language, stories, gossip,
and the rituals of daily life. When problems arise, members of the culture attempt to
resolve them by implementing one or more devices from the toolkit. However, in periods
of social transformation, or unsettled lives, dissatisfaction with accepted ways induces
social actors to create new strategies of action. To gain acceptance, these new strategies

of action must compete with the existing toolkit and be perceived to outperform it.

Frames can be viewed as devices included within the cultural toolkit (Entrnan,
1993; Gamson et al., 1992; Gamson & Modigiliani, 1989; Price et al., 1997). When issue
conﬂicts create dissatisfaction, frame disputes will occur. Frames will ebb and ﬂow in
prominence and be revised and updated to accommodate change (Gamson & Modigliani).
The media are a prime site for this contest:

The mass media are a system in which active agents with speciﬁc

purposes are constantly engaged in a process of supplying meaning.

Rather than thinking of them as a set of stimuli to which individuals

respond, we should think of them as the site of a complex symbolic

contest over which interpretation will prevail (Gamson, 1992, p. xi).
Edelman (1993) argued that frame disputes are characterized by newsmakers exploiting
the media for their own purposes: “. . .while in theory the media can inform and educate,
regimes and interest groups have learned that in practice they can use the media to place
mass audiences in invented worlds that justify the outcomes of any policies at all” (p.
242)
The Moderating Influence of Frames

A moderator is a variable affecting the direction and/or strength of the relation
between an independent or predictor variable and a dependent or criterion variable
(Baron & Kenney, 1986). Within a correlational analysis framework, a moderator is a
third variable that affects the correlation between two other variables. In analysis of
variance terms, a moderator effect occurs when there is an interaction between a focal

independent variable and a factor that speciﬁes the conditions for its operation.

10

Media framing experiments typically examine whether manipulating the content
of a media message can affect audience members’ attitudes toward the subject matter.
Such experiments are, in essence, attempts to understand how external situational
changes affect people’s attitudes. However, external situational determinants such as
frames are only part of the equation. Internal predispositions of the individual message
recipient also inﬂuence attitudes. Attitudes can be inﬂuenced by how the object is
framed, the individual predispositions of the person, or both. One-size-ﬁts-all approaches
that ignore the effects of individual differences are inherently incomplete.

This problem is neither new nor unique to media framing research. Cautioning
that applied psychology was becoming a discipline “divided against itself,” Cronbach
(195 7) argued for an approach integrating the inﬂuence of both the person and the
situation on behavior:

We are not on the right track when we conceive of adjustment or adjustive

capacity in the abstract. It is always a capacity to respond to a particular

treatment. The organism which adapts well under one condition would not
survive under another. If for each environment there is a best organism,

for every organism there is a best environment (p. 679).

More recently, Fiske (2000) observed that generations of researchers examining inter-
group attitudes have alternated between individual and contextual levels of analysis—
using either measures of individual difference or experimentally manipulated situations
as the independent variables.

A mere handful of media framing studies represent the exception to this rule,

simultaneously examining the effects of both participants’ individual differences and

11

random assignment to experimentally manipulated frame conditions (See, e.g., Domke et
al., 1999; Kinder & Sanders, 1990; Nelson et al., 1997). An advantage of this latter
approach is that it promises to better elucidate our understanding of the interplay between
frames and participants’ pre-existing values and beliefs. Given the complexity and often
emotionally-charged nature of attitudes toward afﬁrmative action, this two-pronged
approach seems particularly appropriate: “In short, progress in understanding white
opposition to racial change will require taking up more than one theory at a time”
(Kinder, 1986, p. 168).

Media frames can interact with individual differences, making it possible for
frames to have different effects on people with different predispositions (Domke, 2001;
Domke et al., 1999; Price et al., 1997; Shen, 2004; Tewksbury et al., 2000). One arena in
which media frames and individual predispositions might interact is the interface of
politics and race. Domke (2001; Domke et al., 1999) theorized a process in which media
frames could activate pre-existing racial stereotypes and inﬂuence whether message
recipients apply them in a politically meaningful way. Alternatively, Domke envisioned
the possibility that other frames could “weaken—or least temporarily hinder, or impede”
linkages between racial stereotypes and cognitions about the underlying issue.
A ﬂirmative Action Frames

Afﬁrrnative action presents a classic frame dispute—attempts to deﬁne it
inevitably lead to controversy (Crosby & Cordova, 1996). Indeed, the term aﬁirmative
action itself is not a neutral label—it is a frame suggesting the desirability of corrective
measures (Gamson, 1992). As an abstract principle, afﬁrmative action can be viewed as

an umbrella term for a number of related policies: “Afﬁrmative action is a catchall phrase

12

referring to laws, customs, and social policies intended to alleviate the types of
discrimination that limit opportunities for a variety of demographic groups in various
social institutions” (American Psychological Association, 1995). Accordingly,
afﬁrmative action means different things in different contexts and care should be taken to
focus on speciﬁc afﬁrmative action policies.

The earliest afﬁrmative action policies were initiated to redress the effects of
racial discrimination, particularly against African-Americans:

You do not wipe away the scars of centuries by saying: ‘now you are free

to go where you want, do as you desire, and choose the leaders you

please.’ You do not take a man who for years has been hobbled by chains,

liberate him, bring him to the starting of a race, saying ‘you are free to

compete with the others,’ and still justly believe you have been completely

fair (President Johnson’s address at Howard University, June 4, 1965.

Quoted in: “The Slow and Tortured Death of Afﬁrmative Action,” 2003).

Affirmative action in media discourse

Prior research provides historical perspective on how the mass media have framed
afﬁrmative action (Clawson etal., 2003; Entrnan, 1997; Gamson & Modigliani, 1987;
Richardson & Lancendorfer, 2004; Van Dijk, 1991). Different researchers have
introduced a variety of terms to describe these frames. One parsimonious way to
categorize this discourse conceptualizes one frame opposed to afﬁrmative action, No
Preferential Treatment, and two alternative frames that support afﬁrmative action,

Remedial Action and Diversity.

l3

Gamson and Modigliani (1987) analyzed media discourse on afﬁrmative action
between 1969 and 1984, in four content genres: network television news, news
magazines, political cartoons, and syndicated newspaper columns. Two frames eventually
dominated the discourse: Remedial Action (RA) and No Preferential Treatment (N PT). In
accord with President Johnson’s remarks above, the RA frame asserts that race-conscious
policies are necessary to redress continuing effects of discrimination. In sharp contrast,
the NPT frame opposes afﬁrmative action, asserting that regardless of the motive or
mechanism, race-conscious policies are unjust.

Gamson and Modigliani (1987) also coined the term package to describe a
particularized sub-theme emphasizing one aspect of a media frame. For instance, media
discourse displaying the NPT frame in 1984 most often used the reverse discrimination
package, which asserts that afﬁrmative action violates the rights of individuals who are
not members of the groups targeted for preference. Other packages they identiﬁed within
the NPT frame were unfair advantage (arguing that afﬁrmative action gives beneﬁts
without regard to individual merit), unintended harm (arguing that afﬁrmative action
could be detrimental for minorities), and SES (arguing that economic hardship, rather
than race, should be the criterion for allocating afﬁrmative action beneﬁts).

Gamson and Modigliani (1987) showed that afﬁrmative action framing varied
across time and content genre. On network television news, the RA frame dominated the
discourse in 1969, but by 1984 the NPT frame was used more often. In news magazines,
the NPT frame dominated the discourse in 1969, was brieﬂy surpassed by the RA frame
in 1979, but the NPT frame again dominated discourse in this genre in 1984.

Interestingly, a somewhat different picture emerged in their analysis of openly

14

opinionated media content. Syndicated newspaper columns were equally likely to present
the NPT and RA frames throughout the period studied. The researchers attributed this
ﬁnding to a common practice among newspapers to carry an equal number of
conservative and liberal syndicated columnists. Political cartoons were overwhelmingly
more likely to use the RA frame in every year studied.

Van Dijk (1991) performed a qualitative textual analysis of afﬁrmative action
discourse in British newspapers in 1985-86. He observed that afﬁrmative action received
very little press coverage. When afﬁrmative action was covered, the sources were almost
always white and the slant was highly negative. Although sources and journalists paid lip
service to the ideal of equal opportunity, the NPT frame dominated British media
discourse on concrete afﬁrmative action measures. Terms derogatory of afﬁrmative

,9 ‘6

action characterized the discourse: “massive blunder, race relations mumbo jumbo,”

9’ 6" 9, ‘6

“reverse discrimination,” “burden, intrusive,” “degrading to blacks, radical
philosophy,” and “controversial.”

Entman (1997) examined media discourse on afﬁrmative action during the ﬁrst
half of 1995. He reported that national news magazine stories opposed afﬁrmative action
by a 3:1 margin; network television news reports were evenly split. The NPT frame
completely dominated reports slanted against afﬁrmative action. Also consistent with
Gamson and Modigliani (1987), reverse discrimination was the NPT package displayed
most often. Overall, Entman found that news coverage exaggerated whites’ opposition to
afﬁrmative action, distorted afﬁrmative action into a black-white battleground, and

ignored the fact that afﬁrmative action policies directly beneﬁt not just blacks, but also

white women and other racial/ethnic groups.

15

Entman’s (1997) ﬁndings in pro-afﬁnnative action reports were particularly
interesting. Only 21% asserted that “discrimination remains a problem”—i.e., about one-
ﬁfth speciﬁcally invoked the RA frame. Nineteen percent asserted afﬁrmative action:
“achieves its important objectives,” and 9.6% asserted afﬁrmative action: “helps the
entire society, not just blacks, other minorities, or women.” Entman did not describe
these new “frames” in great detail. Richardson & Lancendorfer (2004) observed these
“frames” might have foreshadowed what they eventuality identiﬁed as the Diversity
frame.

Clawson and colleagues (2003) examined newspaper stories covering Adarand v.
Pena (1995), a US. Supreme Court decision overturning the consideration of race when
awarding federal highway construction contracts. Their analysis of three mainstream
newspapers (The New York Times, Los Angeles Times, and Washington Post) showed that
65% of stories invoked the NPT frame. Twenty-four percent of mainstream newspaper
stories, and 67% of stories in a sample of Black newspapers, invoked a frame the
researchers dubbed: dramatic setback. Described as an outgrowth of the RA frame, the
dramatic setback frame asserted that policies to redress racial discrimination remained
necessary and decried their reversal.

Richardson and Lancendorfer (2004) analyzed 158 newspapers’ editorial
responses to Gratz v. Bollinger (2003) and Grutter v. Bollinger (2003), two US. Supreme
Court decisions addressing afﬁrmative action in student admissions at The University of
Michigan. The underlying decisions overturned a quantitative admissions procedure
awarding extra points to minority applicants to the undergraduate school (Gratz), but

upheld a qualitative admissions procedure that treated membership in an

16

underrepresented minority group as a “plus factor” for law school applicants (Grutter).
The Grutter decision represented the ﬁrst time a majority of the Supreme Court
recognized diversity as a compelling state interest that could justify the consideration of
race——it also reversed lower court rulings (See, e. g., Hopwood v. Texas, (5th Cir., 1996)).

Richardson and Lancendorfer (2004) found the discourse surrounding The
University of Michigan cases was dominated by a frame not revealed in any prior content
analysis of news/editorial media: Diversity. More than 80% of the newspaper editorials
sampled presented the Diversity frame in a positive light, compared to 30% for the RA
frame and 17% for the NPT frame (some editorials presented two frames
simultaneously).

The Diversity frame

The liberal tradition supporting early afﬁrmative action efforts assumed race and
ethnicity are superﬁcial—group membership makes no difference (Allport, 1954; Gandy,
1998). However, a growing school of thought challenges this assumption. Under this
perspective, racial and ethnic differences are neither ﬁctitious nor pejorative; to the
contrary, inter-group differences are presumed real and valuable: “Since we are all
different, we all lose something important if the place we go to college or work fails to
reﬂect the diversity around us” (Gamson, 1999, p. 45). Jones’ (1998) diversity hypothesis
articulates this point of view:

[S]ocietal cohesion and organizational effectiveness can be increased

when racially and culturally diverse citizens are represented or employed,

provided they are represented throughout all levels of the organizational

structure; diversity exceeds some minimal ratio; the diverse citizens or

17

FR”

 

 

 

 

C)

Ill

be

Dc

workers have feelings of common purpose (we-ness); and racial and

cultural identity are valued (p. 658).

In the context of frame analysis, Richardson and Lancendorfer (2004) referred to this
perspective as the Diversity frame.

Bowen and Bok (1998) observed that educators have stressed the value of
diversity for more than 150 years. Initially, diversity was embraced in the context of
regional differences—allowing students from the North and South to associate with
dissimilar people they had not known before was believed to promote learning.
According to Foster (1993), the extension of this concept to racial and ethnic differences
is attributable to Justice Powell’s concurrence in Regents of the University of California
v. Bakke (1978). Therein, Justice Powell opined that a racially diverse student body
enhances the educational experience of all students, in part by promoting a robust
exchange of ideas in the classroom. Signiﬁcantly, Powell shifted focus away from the
backward-looking RA frame to a frame recasting afﬁrmative action as a forward-looking
policy that could also beneﬁt non-minority students and society at large (Foster).

Beginning in the mid-19803, the argument for diversity gained momentum in
corporate America. Management publications transformed afﬁrmative action from a legal
mandate into diversity management—a new way to raise proﬁts. Speciﬁcally, business
experts asserted two reasons why ﬁrms should develop diverse workforces: (a) to better
understand and serve an increasingly diverse and global market, and (b) because White
males constituted a declining percentage of the skilled labor pool—soon there would not
be enough of them to ﬁll all highly-skilled positions (Edelman et al., 2001; Kelly &

Dobbin, 1998).

18

In the context of education, Asante’s (1991) conceptualization of multiculturalism
sheds light on the rationale underpinning the Diversity frame. Asante observed that for
most of our history, a white-only perspective dominated the curriculum of American
schools. Major achievements in history, science and the humanities were attributed only
to whites, while contributions made by people of color were ignored, slighted or
misattributed. In contrast, multiculturalism acknowledges that our system of knowledge
is the product of contributions from people of many different racial and ethnic
backgrounds. Asante contended that recognizing the value of all racial and ethnic groups
would be a catalyst for the emergence of a common American culture. Moreover, he
argued that a multicultural approach to education promotes the self-esteem of minorities
and increases Whites’ awareness of the achievements of other cultures, thereby reducing
Whites’ aptness to assume a superior attitude.

Supporters of diversity in education assert that race-conscious policies provide a
number of beneﬁts. Diversity in the classroom can promote leaming—students are
exposed to different perspectives, thereby learning to work effectively and get along with
people different from themselves (Bowen & Bok, 1998; Brown, 2002; Crosby et al.,
2003; Gamson, 1999; Gurin, et al., 2002; Palmer, 2001; Terrenzini et al., 2001). Indeed,
because housing patterns and school districts in America remain highly segregated, most
Whites and minorities enter college having little prior contact with diverse peers (Gurin,
et a1.). Diversity can also prepare students for democratic citizenship (Brown; Crosby, et
al.; Palmer). Advocates further assert that diversity in education provides a stream of
well-educated minorities who can go on to serve and lead their communities (Crosby et

al.; Palmer).

19

Of course, the Diversity frame has critics. The fundamental assumption of the
Diversity frame is that members of different racial and ethnic groups are culturally
different, in part because of White domination (Merelman, 1994). Some challenge this
assumption, arguing that it is specious to assume that culture follows race/ethnicity (See
e. g., Foster, 1993; Hollinger, 1998). Diversity has also been criticized as being too
amorphous—if differences are valuable, then why not value left-handedness, obesity, red
hair, etc.? (Foster). Terrenzini and colleagues (2001) suggested the effects of diversity in
education are small and might not be linear, i.e., there might be a point where further
increasing diversity could be counter-productive.

Other critics of the Diversity frame view it as a retreat from the original goals of
afﬁrmative action, particularly parity. Although diversity requires more than a token
representation of minorities to attain its objectives, diversity does not necessarily require
proportionate racial/ethnic representation, which some view as an essential element of
equality (Grutter v. Bollinger, 2003; Jones, 1998; Palmer, 2001). By refusing to look
backward at the legacy of racism in America, the diversity frame could serve to
perpetuate the existing racial hierarchy:

[T]he liberal defenders do not expressly deny the presence of past or

current racial discrimination. Instead they claim disinterest in either,

arguing that it is more fruitful and less divisive to focus on the future than

to stir the ashes of a troubled past. But the effect is much the same. The

liberal defense is framed as if there is no structural discrimination to

remedy (Lawrence, 2001, p. 954).

20

Indeed, some advocates of the Remedial Action frame argue the Diversity frame reduces
people of color into useful tools for exploiting global markets, conquering developing
nations, or policing minority neighborhoods (Fancher, 2003; “The Slow and Tortured
Death of Afﬁrmative Action,” 2003).

The Diversity frame presents a rationale for race-conscious policies quite distinct
from the traditional RA frame. Many businesses and educational institutions have
embraced diversity for some time. The mass media and legal system have now become
sites for grappling with diversity and its implications. How the general public will react
to the Diversity frame remains to be seen. Prior research on the effects of other
afﬁrmative action frames offer some insights on this question.

Effects of Affirmative Action Frames

A number of prior studies have examined the effects of framing on attitudes
toward afﬁrmative action and/or its beneﬁciaries. Results have been mixed, in part due to
features of different research designs. Nevertheless, prior research does appear to suggest
that frames can: (a) affect attitudes toward afﬁrmative action, and/or (b) moderate the
effects of individual predispositions on attitudes toward afﬁrmative action.

Kinder and Sanders (1990) employed a survey-based experiment to examine the
effects of framing on support for afﬁrmative action in employment and college
admissions. Respondents were randomly assigned to answer questions invoking either of
two NPT packages drawn from Gamson and Modigliani (1987): reverse discrimination
(emphasizing harm to whites outside the target group) or unfair advantage (emphasizing
that beneﬁciaries did not earn the beneﬁts of the afﬁrmative action). The frame induction

did not signiﬁcantly affect support for afﬁrmative action; whites opposed it strongly in

21

either framing condition. Nevertheless, the frame induction did affect which factors
respondents took into consideration. Whites’ support for the principle of equal
opportunity was a positive predictor of support for afﬁrmative action in the unfair
advantage condition. However, in the reverse discrimination condition, whites’ support
for the principle of equal opportunity was negatively associated with support for
afﬁrmative action. Interestingly, Kinder and Sanders’ (1990) interpretation was that the
unfair advantage frame induced whites to access negative racial sentiments under which:
“Public opinion [of white people] on afﬁrmative action becomes a kind of referendum on
the moral character of black citizens” (p. 96). Unfortunately, the researchers did not
measure respondents’ attitudes or affect towards Blacks.

Fine (1992) examined data from a national survey using a split-sample approach
similar to Kinder and Sanders (1990). Again, respondents were randomly assigned to
answer survey questions invoking either of two NPT packages: reverse discrimination or
unfair advantage. This time opposition to afﬁrmative action was stronger in the reverse
discrimination framing condition, among Whites and Blacks alike. Consistent with
Kahnemann and Tversky’s (1984) work concerning the relative attractiveness of frames
emphasizing gain versus loss, Fine concluded the public appears more opposed to an
unwarranted loss than an undeserved gain.

Murrell et al. (1994) conducted a survey-based experiment involving alternative
frames for supporting afﬁrmative action. Frames were manipulated between-subj ects in
three conditions: (1) to compensate for past discrimination (the RA frame), (2) to provide
diversity within the organization (the Diversity frame), or (3) a control condition, offering

no justiﬁcation for afﬁrmative action. Results showed that compared to the control

22

condition, both the Remedial Action and Diversity frame conditions garnered more
support for afﬁrmative action. However, there was no signiﬁcant difference in support for
afﬁrmative action between respondents in the Remedial Action versus the Diversity
framing condition.

Snidennan and Piazza (1996) conducted experiments in which participants read
descriptions of ﬁctional government programs. One study involved the issue of
government assistance to a ﬁctional laid-off worker. The target beneﬁciary’s race and
dependability were manipulated between-subjects. A striking result was that politically
conservative participants were more likely to favor government assistance when the
target was described as an African-American. Race of the target made no difference for
liberal participants. Targets described as dependable garnered more support from both
groups. The researchers interpreted their result with enthusiasm:

It has long been assumed that whites are dug in on racial issues. In fact,

large numbers of whites can be dislodged from the positions they have

taken on many issues of race by calling their attention to countervailing

considerations. A large number of Americans are open to argument to

tip the scales, not on every issue of race but on the largest number of them.

The case for public policies to assist blacks can be won; and it can be lost

(p. 64).

Sniderrnan and Piazza (1996) further argued that opposition to afﬁrmative action
is driven more by program characteristics than racial sentiments. If racial sentiments were
the primary impetus, they contended, then white respondents should have been less

willing to help the target in the black beneﬁciary condition.

23

In a subsequent experiment, Sniderrnan and Piazza (1996) instructed participants
to read descriptions of ﬁctional “equal opportunity” programs. Follow-up questions
mentioning a quota system favoring Blacks induced White participants to describe Black
people as irresponsible and lazy. The researchers interpreted these ﬁndings as indicating
that afﬁrmative action causes white backlash:

The new race-conscious agenda has provoked broad outrage and

resentment. Afﬁrmative action is so intensely disliked that it has led some

whites to dislike blacks—an ironic example of a policy meant to put the

divide of race behind us in fact further widening it” (p. 62).

However, given that quotas represent the most rare and extreme form of afﬁrmative
action, it is possible that lesser forms of afﬁrmative action might have no such impact.

Kuklinski et al. (1997) conducted an elegant program of survey-based
experiments that unobtrusively examined between-groups attitudes toward afﬁrmative
action. Their results showed that respondents opposed afﬁrmative action policies much
less when framed as making “an extra effort” compared to giving “a preference.”

Entman and Rojecki (2000) conducted qualitative interviews to examine Indiana
residents’ attitudes toward afﬁrmative action. Many were vehemently opposed to
afﬁrmative action, even though they had no personal experience with it. The researchers
interpreted this ﬁnding as suggestive that mass media reports were a powerful inﬂuence.
In addition, the researchers also found that when they mentioned that special preferences
are routinely given to other groups, such as children of alumni and athletes, respondents’

opposition to afﬁrmative action softened.

24

Quinn et al. (2001) conducted an experiment in which participants read ﬁctional
newspaper editorials about afﬁrmative action. The editorials manipulated attributions of
minorities’ responsibility for causing the problem of disparate minority unemployment,
as well as their responsibility for solving it. Results showed that when editorials were
framed such that minorities were portrayed as responsible for either causing or solving
the problem, support for afﬁrmative action was higher. Frames suggesting minority
responsibility for both causing and solving the problem, or neither, induced less support
for afﬁrmative action.

Clawson and Waltenburg (2003) conducted an experiment in which participants
read ﬁctional news stories concerning the Supreme Court’s ruling in Adarand v Pena
(1995). In line with the content analysis of Clawson et al. (2003), the researchers
manipulated the story between-subj ects, presenting either the dramatic setback frame (a
regrettable retreat from remedial action), or the NPT frame. Results showed that white
participants in the NPT frame condition were more supportive of the Adarand decision,
i.e., the NPT frame induced more support for a ruling against afﬁrmative action.

Several conclusions can be drawn from this body of research. First, there is
abundant evidence that framing can affect attitudes toward afﬁrmative action (Clawson &
Waltenburg, 2003; Entman & Rojecki, 2000; Fine, 1992; Kuklinski, et a1. 1997; Murrell
et al.; 1994: Sniderrnan & Piazza, 1996; Quinn etal., 2001). The power of frame
manipulations to affect support for afﬁrmative action can depend on the race of the
respondent; whites’ support for afﬁrmative action appears more susceptible to framing
effects (Clawson & Waltenburg; Kuklinski). Framing can also moderate the effects of

individual predispositions (prior attitudes) on attitudes toward afﬁrmative action (Kinder

25

& Sanders, 1990). There is also evidence suggesting some afﬁrmative action frames
might negatively inﬂuence whites’ attitudes toward blacks (Kinder & Sanders;
Sniderrnan & Piazza).

Murrell et al. (1994) is the only published study examining the impact of the
Diversity frame. They found the Diversity frame induced more support for afﬁrmative
action compared to a control condition offering no justiﬁcation for afﬁrmative action, but
it did not induce signiﬁcantly greater support than the RA frame. One possible
explanation is that the Diversity frame and RA frame are equally better than nothing, but
neither more so than the other.

Another possibility is that Murrell and colleagues’ (1994) experimental
manipulation was too weak. A tersely worded survey question might have been
insufﬁcient to induce their respondents to give much consideration to the Diversity
frame’s implications. Respondents likely had little prior familiarity with the concept of
diversity—media discourse surrounding afﬁrmative action appears rarely to have
mentioned it (Richardson & Lancendorfer, 2004). Moreover, while Murrell and
colleagues tested the impact of the Diversity frame on support for afﬁrmative action, they
did not test its potential impact on affect toward African-Americans nor on the relative
weight respondents assign to underlying beliefs. Hence, the prior literature leaves many
questions about the Diversity frame unanswered.

Despite this gap in prior research, there is good reason to hypothesize that the
Diversity frame can affect respondents quite differently than the Remedial Action frame.
Social Identity Theory and Procedural Justice Theory suggest the Diversity frame might

sway respondents to view afﬁrmative action and its beneﬁciaries more positively.

26

Social Identity Theory

To understand the effects of media content on attitudes toward afﬁrmative action
and its beneﬁciaries, it is necessary to consider how people think and feel about social
groups. Disparities between social groups gave rise to afﬁrmative action. Social group
membership is the criterion for receiving afﬁrmative action beneﬁts. Indeed, the debate
over afﬁrmative action can be viewed as a clash of ideas concerning social groups:
“Without denying that racism remains a problem or that ideological conservatism matters
for White attitudes, afﬁrmative action is about the place racial groups should occupy in
American society” (Bobo, 1998, p. 999).

A social group can be deﬁned as two or more people who perceive themselves to
be members of the) same social category (Taj fel & Turner, 1979; Turner, 1982). Unlike
the categorization of things, every social category is either an in-group that includes the
perceiver or an out-group that excludes them (Simon, 1993). Social categories are an
important component of every person’s self-concept; they provide a frame of reference
that helps create and deﬁne a person’s place in society (Taj fel & Turner).

Social Identity Theory posits that the need for positive self-esteem motivates
people to perceive the in-group favorably and as differentiated from out-groups (Brewer,
1979; Tajfel & Turner, 1979; Turner, 1982). Identiﬁcation with a social category is also
associated with a perception that in—group members are less differentiated from the self.
People come to perceive outcomes for other in-group members as their own and to like
in-group members more. A sense of togetherness, belonging and we-ness develops:

“Thus group members do not perceive themselves as isolated individuals but as members

27

of a social entity whose properties are thereby modiﬁed” (Brown & Turner, 1981, p. 35).
All of this leads to bias favoring the in-group in evaluations and behavior.

Social Categorization Theory (Turner, 1985) holds that the self-concept includes
both a personal identity and a social identity. Personal identity is the individuated self,
i.e., the characteristics that differentiate an individual from others within a given social
context. Social identity is the sum total of signiﬁcant social categorizations in which a
person includes the self (Brewer, 1991; Turner, 1982, 1985). Hence, Social
Categorization Theory rejects the assumption that the true person lies only in their
idiosyncrasies: “Human beings are both and equally individual persons and social group
members” (Oakes & Turner, 1990, p. 114).

Tajfel and colleagues’ (1971) minimal group experiments elegantly demonstrated
that social identity and the resultant in-group bias are primarily cognitive phenomena—
the very act of social categorization, in and of itself, is sufﬁcient to generate bias.
Participants were induced to identify with ad hoc groups based on meaningless
membership criteria, e.g., aesthetic preference for Kandinsky or Klee; a supposed
inclination to over-estimate or under-estimate the number of dots in a cluster projected
brieﬂy on a screen. These “groups” had no interaction, no instrumental value and were
completely anonymous. Nevertheless, when asked to allot money to people identiﬁed
only by their “group membership,” participants with nothing at stake discriminated in
favor of in-group members, even when detrimental to the greater common good.

Race is a salient category for social identiﬁcation that can lead to exaggerated
perceptions of difference and bias favoring one’s own racial group (Jones, 1998). The

pervasive tendency to discriminate on the basis of race has been conﬁrmed in meta-

28

analyses of numerous studies (Kraiger & Ford, 1985; Mullen, et al., 1992). Social
constructions of race can have profound psychological and social meaning (Omi, 2001;
Omi & Winant, 1994; Winant, 2000).

Of course, most people identify with many social categories. Different situational
stimuli work to activate a particular social identity at a given time:

Distinctiveness of a given social identity is context-speciﬁc. It depends on

the frame of reference within which possible social entities are deﬁned at a

particular time, which can range from participants in a speciﬁc social

gathering to the entire human race (Brewer, 1993, p. 158).

Stimuli which can heighten the salience of a social identiﬁcation include: conﬂict
or confrontation, competition between identity groups, unequal distribution of resources,
group distinctiveness in a particular setting, inter-group similarities, shared fate,
interdependence, the number of in-group and/or out-group members present, and physical
proximity (Brewer, 1979; Brewer & Harasty, I996; Brewer & Miller, 1984; Brown &
Turner, 1981; Garcia-Prieto et al., 2003; Kramer & Brewer, 1984; Tajfel & Turner, 1979;
Turner, 1982, 1985). Tajfel and Turner argued that immutable social categories such as
race intensify inter-group bias and conﬂict. Among highly prejudiced people, racial
identity is more chronically salient across many situations (Brown & Turner, 1981; Byme
& McGraw, 1964).

An analogy offered by Oakes and Turner (1990) is instructive. When a person
looks at a busy street, there are several ways to conceptualize the scene: as trafﬁc; as cars,
buses, trucks and motorcycles; as Hondas, Fords, BMWs, etc. The underlying stimuli

remain the same, but the subjective momentary experience of similarity and difference

29

 

 

rel
[Cg
as

the

 

of

depends on which typology is employed. None of these typologies is more or less valid;
they neither distort nor deny the existence of one another. They simply call up different
bundles of information, making some considerations more salient. Put another way,
different frames suggest the applicability of different organizing heuristics.

Similarly, attitudes toward social issues and other people can depend on which
social identity is activated. The ability to call up different social identiﬁcations and act
accordingly, depending on external circumstances, is neither irrational nor primitive.
Rather, it is an adaptive process making possible pro-social relations such as social
cohesion, cooperation and inﬂuence (Turner, 1985).

Re-categorization

Social identity theory explains why people discriminate in favor of in-group
members. But how can people be persuaded to treat out—group members more equitably?
One approach is re-categorization (Dovidio et al., 1995; Gaertner et al., 1989, 1993,
1994), which attempts to reduce inter-group bias by inducing group members to identify
with a larger social group. The concept posits that by inducing people to identify with a
larger group, the number of in-group members can be expanded. Thus, people formerly
regarded as out-group members can be re-categorized as in-group members. Once
regarded as in-group members, bias against the former out-group members diminishes. In
a sense, re-categorization invokes the logic of the familiar maxim: “If you can’t beat
them, join them.”

Several studies suggest that inducing group members to change their perceptions
of group boundaries can invoke different social identities, thereby reducing inter-group

bias. Inducing experimental participants in two groups to re-conceive of themselves as

30

belonging to a single, super-ordinate group improves favorable affect toward former out-
group members (Dovidio etal., 1995; Gaertner et al., 1989, 1993). The perception of a
super-ordinate group identity can also cause people to act in a manner more consistent
with the common good (Kramer & Brewer, 1984). A survey of high school students
demonstrated that perceptions of the student body as “one group” and “on the same team”
were positively associated with lower affective bias toward members of different racial
and ethnic groups (Gaertner et al., 1994). Similarly, a survey of ethnically diverse union
members revealed that strong identiﬁcation with the super-ordinate group (their
employer) was positively associated with values conducive for organizational cohesion
(Huo et al., 1996). Collectively, these studies suggest that positive affect toward others is
a consequence, rather than a cause, of group cohesiveness (Brewer & Harasty, 1996).
Several scholars have suggested that social identity and re-categorization are
implicated in the afﬁrmative action debate. Chang (1996) observed that an inadvertent
effect of afﬁrmative action policies is to call attention to conﬂict between different social
groups. Gandy et al., (1997) argued blacks are more likely to favor afﬁrmative action
because they are more likely to identify with black afﬁrmative action recipients. Brewer
(1995) noted that the concept of diversity implies social identities; within an
organization, various demographic identiﬁes exist side-by-side with work subgroup
identities and the super-ordinate identity of the entire ﬁrm or institution. She suggested
that the concept of re-categorization could improve organizational performance by
reminding members of rival departments about their common super-ordinate identity.

Garcia-Prieto et al. (2003) observed that because team members have multiple, dynamic

and context-dependent social identities, diversity is best understood as a subjective

31

experience; the important social identity is not that into which a person “objectively”
falls, but the one to which they subjectively feel they belong. Perhaps most directly on
point were Murrell et al. (1994):

In the case of afﬁrmative action, individuals may be motivated to promote

their own group’s interests. Thus, resistance to afﬁrmative action policies

is likely contingent on the extent to which these policies are seen as

beneﬁting out-group members at the expense of in-group members. This

conceptualization suggests that strategies that reduce the extent to which

afﬁrmative action is seen as beneﬁting an out-group (e.g., Blacks) relative

to the in-group (e. g., Whites) should reduce opposition to afﬁrmative

action. Thus, refocusing or recategorizing social boundaries may reduce

evaluations of afﬁrmative action on the basis of racial group boundaries

(p. 83).

Re-categorization toward a common super-ordinate identity does not require
people to forfeit their subgroup identities. Two groups can be viewed as operating
independently toward a common goal, such as the offensive and defensive units of a
football team (Gaertner et al., 1994). Optimal distinctiveness theory posits that a balance
between personal and social identity is consistent with good mental health (Brewer,
1991). The self-concept involves a fundamental tension between two opposing processes:
people want to be like others, but they simultaneously want to be unique. In other words,

people are motivated to be “different and the same at the same time.”

32

Procedural Justice Theory

Social identity is clearly not the only factor that inﬂuences attitudes toward
afﬁrmative action and its beneﬁciaries. For example, many Blacks oppose afﬁrmative
action and many Whites support it (Krysan, 2000; Schumann et al., 1997; Steeh &
Krysan, 1996). This strongly indicates something more than social identity is at work.
Many scholars assert perceived procedural justice is a powerful determinant of attitudes
toward afﬁrmative action.

Tyler and colleagues (Lind et al., 1993; Lind & Tyler, 1988; Tyler, 1990; Tyler &
McGraw, 1986; Tyler, et al., 1986) challenged the long-standing presumption that
political opinions are determined primarily on the basis of economic self-interest. They
observed that many people hold opinions about issues in which they have no personal
economic stake, and that people can also take positions contrary to their economic self-
interest. Tyler and colleagues argued the explanatory power of economic self-interest was
limited and suggested an alternative explanation: Procedural Justice Theory.

Distinct from the fairness of policy outcomes, procedural justice concerns
decision-making methods and includes factors such as: consistency, predictability,
neutrality, trustworthiness and respect for all interested parties. Tyler and colleagues
(Lind et al., 1993; Lind & Tyler, 1988; Tyler, 1990; Tyler & McGraw, 1986; Tyler, et al.,
1986) posited that people use perceived procedural justice as a heuristic to evaluate
policy questions, especially when issues are complex, ambiguous or have uncertain
outcomes. This heuristic is learned in early childhood and its use is not necessarily

conscious.

33

 

Subsequent work indicates that perceived procedural justice could predict support
for afﬁrmative action. However, a complicating factor is that perceived fairness can
motivate either support or opposition for afﬁrmative action (Golden et al., 2001 ).
Perceptions of procedural justice depend on which allocation rule is judged most ﬁtting;
i.e, justice is context-dependent (Azzi, 1992; Barnes Nacoste, 1994; Brewer & Harasty,
1996; Clayton & Tangri, 1989; Esses & Seligman, 1996; Jencks, 1988; Nacoste, 1987;
Opotow, 1997; Peterson, 1994).

Thibault and Walker (1978) described one inﬂuential allocation principle, the
proportionality rule. This principle holds that beneﬁt distribution should be proportional
to the contributions made or achievements attained by those competing for the beneﬁt.
Barnes Nacoste (1994) observed that afﬁrmative action could lead to a perception that the
proportionality rule is being violated. Indeed, Gamson & Modigliani (1987) identiﬁed an
NPT package in mass media content, unfair advantage, which clearly invokes the
proportionality principle. Likewise, Chang (1996) asserted that afﬁrmative action is
procedurally unjust because it deprives candidates outside the target group of an equal
chance to compete.

A counterveiling allocation rule identiﬁed by Thibault and Walker (1978) is
equity. This principle holds that all human beings are worthy and society has an
obligation to help the disadvantaged. Jencks (1988) observed that no society fully
reconciles the conﬂict between the norms of proportionality and equity, in part because
every moment in time is both a beginning and an end. For instance, viewing admission to
college as an end suggests the proportionality principle—students should be admitted on

the basis of grades, test scores, recommendations and other past achievements. However,

34

if college admission is viewed as a beginning, i.e., as a gateway to prosperity and a share
of the American dream, then equity for all groups becomes more salient.

Prior research suggests that both of these allocation principles, equity and
proportionality, inﬂuence evaluations of afﬁrmative action policies. When participants
are speciﬁcally told that both merit and group membership factor into beneﬁt allocation
decisions, they support afﬁrmative action programs more (Major et al., 1994). “Milder”
afﬁrmative action programs that help minorities without “penalizing” Whites (e.g., job
training, mentorship) receive overwhelming support (Bobocel et al., 1998; Kravitz, 1995;
Kravitz & Platania, 1993; Krysan, 2000; Nosworthy et al., 1995; Schumann et al., 1997;
Steeh & Krysan, 1996; Summers, 1995). Racial prejudice is a strong predictor of
opposition to these types of programs (Bobocel; Nosworthy). However, as race becomes
weighted more heavily as a preferential factor, support for afﬁrmative action decreases
(Bobocel; Kravitz; Kravitz & Platania; Krysan; Nosworthy et al.; Schumann, et al.;
Skedsvold & Mann, 1996; Steeh & Krysan ). As the weight of race as a preferential
factor increases, the correlation between racial prejudice and opposition to afﬁrmative
action decreases (Bobocel; Nosworthy).

Implicit in this body of work is an assumption (by participants) that race is
irrelevant when evaluating merit (Summers, 1995). However, several scholars have
challenged this assumption (Crosby, 1994; Crosby et al., 2003; Eberhardt & Fiske, 1994;
Jackman & Muha, 1984; Moses, 1999; Winkelmann & Crosby, 1994). Under this view,
merit is a myth—an ideological construct used to perpetuate the advantaged position of
privileged groups. Merely examining grades or test scores is specious under this view,

because it ignores that members of different groups do not have equal access to resources

35

and are not treated equally by authority ﬁgures. Swain (2001) observed there is more than
one deﬁnition of merit, and that universities should be given wide latitude to deﬁne merit
in a ﬂexible way.

Schmidt and Hunter (1999) advocated a different view. They argued that
standardized tests (such as the S.A.T.) are valid and unbiased—the problem is the failure
of minorities to develop the skills and abilities measured by the tests. Schmidt and Hunter
further contended that ignoring such objective measures of merit would reduce economic
productivity and the standard of living for everyone, whites and minorities alike. Hence,
the conﬂict between equity and proportionality persists in academe as well.

Another complicating factor in deciding questions of fairness is the level of
analysis. Brickman and colleagues (1981) observed that a procedure viewed as just when
applied to society as a whole can seem very unfair when the focus shifts to a particular
case involving speciﬁc individuals. For instance, the exclusionary rule in court
proceedings (making evidence inadmissible if obtained illegally) is viewed as promoting
justice by forcing police to respect citizens’ right to freedom from unreasonable searches.
However, when it leads to release of a particular criminal clearly guilty of a heinous
crime, the exclusionary rule could seem much less fair.

A micro-level focus is apparent in the arguments of many afﬁrmative action
opponents. For instance, Liu (1998) argued that afﬁrmative action is procedurally unjust
because it beneﬁts individual minorities who might not personally have been victims of
discrimination, while harming non-minorities who personally committed no
discriminatory act (See also, Foster, 1993). Other opponents argue minority group

members who are personally “well off” should not be allowed to receive afﬁrmative

36

action beneﬁts (See, e.g., Esses & Seligman, 1996). Indeed, both of these sets of
arguments appeared in NPT packages identiﬁed by Gamson & Modigliani’s (1987)
content analysis of news media. J emmott & Tebbets (1980) reported afﬁrmative action
opponents were more likely to refer to micro-level information about Mr. Bakke (the
white plaintiff seeking admission to medical school in the case of Regents of the
University of California v. Bakke (1978)), while articles slanted against the Bakke ruling
(i.e., in favor of afﬁrmative action) more often referred to macro-level issues.

Scholars have suggested that shifting the focus from the micro-level to the macro-
level can soften opposition to afﬁrmative action (Clayton & Tangri, 1989; Esses &
Seligman, 1996; Opotow, 1997). Clayton and Tangri argued that afﬁrmative action
should be viewed from the macro-level because it induces people to more accurately
perceive inter-group disparities and the effects of discrimination. A micro-level focus, in
contrast, tends to perpetuate the myth that discrimination is no longer widespread.
Interestingly, Clayton and Tangri observed that a macro-level focus is counterintuitive;
people more naturally tend to view situations from a micro-level perspective.

Procedural justice and social identity

To conceptualize social identity and perceived procedural justice as entirely
separate would be inaccurate. Assessing the fairness of policies affecting groups
necessarily activates mental representations of social categories (Brewer & Harasty,
1996). Hence, in-group bias and procedural fairness can be viewed as two norms exerting
simultaneous inﬂuence (Braithwaite, et al. 1979; Brewer & Harasty; Tajfel et al., 1971).
For instance, Taj fel and colleagues’ minimal group experiments revealed participants’

tendency to discriminate in favor of in-group members was tempered by considerations

37

of fairness. Tougas and Veilleux (1989) argued that even in-group members will endorse
afﬁrmative action only if they believe the procedure is fair. Azzi’s (1992) experiments
showed that whites induced to identify with minorities were more likely to equally divide
power among groups, i.e., to invoke an equity allocation principle, rather than
proportionality. Therefore, not only is there an interplay between social identity and
perceived fairness, it also appears that personalizing the out-group can inﬂuence which
allocation rule is invoked.

Opotow (1997) described an additional consideration affecting perceptions of
fairness: the scope of justice. It is possible to believe some groups do not deserve just
procedures:

Concerns about fair procedure and distributions are irrelevant outside our

boundary of justice. Moral values and concerns about entitlements do not

apply to those viewed as nonentities, undeserving or expendable. Instead,

we justify harm that befalls those outside as inevitable or as deserved

because they are “morally excluded” from our scope of justice (Opotow,

p.237)

Thus, it is conceivable that a procedure recognized as unjust if applied to in-group
members might not be perceived as unjust if it affects out-group members only.
Interracial Attitudes

As discussed previously, the interface of politics and race is an arena in which

media frames and individual predispositions can interact. Accordingly, this study pre-

tested participants on interracial attitudes, toward the goal of attaining a richer

38

 

. h‘
l

r

i

i

1

1‘
I
r.

r.

rave. r-. u ...:-

 

understanding of their joint effects with the Diversity and Remedial Action frames. A
discussion of these interracial attitudes is provided below.

Modern racism

For centuries in America, respected members of white society openly expressed
values and beliefs that were unabashedly racist. McConahay (1981, 1986) argued that in
the second half of the twentieth century, disclosure of Nazi atrocities and the emergence
of the American civil rights movement created an environment in which it became no
longer tenable for elites to express blatantly racist beliefs. He theorized that as a result, a
more contemporary form of racial prejudice emerged: modern racism.

Modern racism is conceptualized as a combination of the Protestant work ethic
and negative affect toward Aﬁican-Americans. The principle tenets of modern racists are:
(1) discrimination is no longer widespread; (2) blacks push too hard and too fast into
areas in which they are not wanted; (3) blacks’ demands, and the tactics used to achieve
them, are unfair; (4) recent gains by blacks are undeserved; (5) racism is bad but the
foregoing beliefs are not racist, simply facts, and; (6) true racism is limited to old-
fashioned notions such as segregation and the alleged genetic inferiority of blacks.
Modern racism is believed to be acquired early in life and more difﬁcult to change than
the cognitive and policy preference components associated with it.

A key component of modern racism is a desire not to appear prejudiced.
McConahay (1981) argued that because the opinions and beliefs associated with modern
racism can be justiﬁed on other grounds (primarily the Protestant ethic), it often goes
unacknowledged. Fiske (2000) observed that modern racists hold policy positions that:

“. .. all happen to disadvantage minorities; the high-scoring individual thus has

39

ideological excuses for bias. The difference between modern racists and modern
nonracists lies in the political beliefs of the racists” (p. 301).

The concept of modern racism is controversial. Critics’ primary objection is that
modern racism is not clearly distinct from the underlying policy issues, particularly
afﬁrmative action and school busing (See, e. g., Katz et al., 1986; Sniderman & Tetlock,
1986; Wood, 1994). Under this view, race-neutral principled objections to policies are
confounded with modern racism. The result, critics charge, is that many people are
falsely labeled as racists.

Nevertheless, prior research suggests that beliefs associated with modern racism
strongly predict opposition to afﬁrmative action, above and beyond race-neutral beliefs.
McConahay’s (1986) Modern Racism Scale is one widely used measure. However,
measures of several analogous constructs also yield similar results.

For instance, Jacobson (1985) reported that a seven-item measure of new racism
strongly predicted afﬁrmative action attitudes in a national survey. Tuch and Hughes
(1996) reported the strongest predictor of afﬁrmative action attitudes was belief in
discrimination. An analysis of four-large scale surveys by Sears et al. (1997)
demonstrated that controls on political ideology, party identiﬁcation and social welfare
attitudes had weaker effects on afﬁrmative action attitudes than symbolic racism, the
effects of which were not substantially reduced by such controls. Williams et al. (1999)
found that a measure of contemporary prejudice (incorporating several items from
McConahay’s (1986) Modern Racism Scale) was by far the strongest predictor of
opposition to afﬁrmative action —controlling for contemporary prejudice reduced the

association between conservatism and afﬁrmative action opposition to marginal

4O

signiﬁcance. A survey by James et al. (2001) demonstrated that modern racism scores
were negatively associated with support for equal opportunity policies targeted to beneﬁt
African-Americans. Several commentators have suggested that a central tenet of modern
racism is also the key reason for white opposition to afﬁrmative action: the belief
discrimination is mainly a thing of the past and therefore blacks are themselves primarily
responsible for inter-group disparity (See, e.g., Gandy & Baron, 1998; Gandy et al., 1997;
Iyengar, 1989; Kluegel, 1985; Kluegel & Smith, 1983).

Interestingly, Bonilla-Silva’s (2000) in-depth interviews with white students
suggested scores on the Modern Racism Scale might actually underestimate the
pervasiveness of racial prejudice. Her qualitative study showed that even students with
low to moderate scores on the Modern Racism Scale displayed tendencies to: (I) express
racial views in a “sanitized way,” (2) claim ambivalence to avoid disclosing their true
beliefs, (3) believe that racial prejudice is a thing of the past, and (4) construct a we-they
dichotomy between whites and blacks.

White guilt and belief in white privilege

Written from his perspective as an African-American living in the Southern
United States during the 19603, Steele’s (1990) essay, White Guilt, described a sea
change in white society’s racial attitudes. After centuries of oppressing minorities, Steele
argued that whites in the 19608 came to experience a sense of collective guilt for the
harms their group had inﬂicted: “White guilt became so palpable you could see it on
people. guilt had changed the nature of the white man’s burden from the
administration of inferiors to the uplift of equals—from the obligations of dominance to

the urgencies of repentance” (pp. 497-498). Steele further argued that momentous public

41

policy changes such as the Civil Rights Act of 1964 and Great Society were directly
attributable to this collective sense of white guilt.

Steele (1990) asserted that white guilt motivated a misguided approach to
afﬁrmative action. For Steele, the goal of afﬁrmative action should be to assist African-
Americans to develop their full potential, toward achieving parity with whites.
Referencing President Johnson’s famous 1965 address at Howard University (excerpts
quoted above), Steele argued that afﬁrmative action has instead focused primarily on
alleviating white guilt:

On its surface this seems to be the most reasonable of statements, but on

closer examination one can clearly see how it deﬂects the emphasis away

from black responsibility and toward white responsibility. . .. blacks are the

passive recipients of white action. The former victimizers are challenged

now to be patrons, but where is the black challenge? This is really a

statement to and about white people, their guilt, their responsibility, and

their road to redemption. Not only does it not enunciate a black mission,

but it sees blacks only on the dimension of their victimization—‘hobbled

by chains’——and casts them once again into the role of receivers of white

beneﬁcence. Nowhere in this utterance does President Johnson show

respect for black resilience or faith in the capacity of blacks to run fast

once they get to the ‘starting line’ (p. 504).

Steele (1990) argued further that granting preferences to a handful of blacks might
be a quick and easy way to make some whites feel better, but it fails to address the core

problem of assisting blacks to accept and shoulder responsibility for uplifting themselves:

42

“Here one falls into Orwellian doublespeak, where preference equals equality” (p. 505,
emphasis original).

Tatum (1992), a professor who taught seminars on the psychology of race for two
decades, discussed white guilt in the context of a recurrent pattern she observed among
white students. Initially, typical white college students were largely ignorant about the
extent and effects of past and present racism in American society. Once better informed,
many white students expressed a deep sense of guilt for the harm their group caused.

As discussed above, modern racism is characterized by, among other things, the
belief that African-Americans and other minorities are no longer seriously disadvantaged
by racial discrimination. However, McIntosh (1989) observed that even among whites
believing that racial discrimination creates a substantial disadvantage for minorities, there
is often a failure to appreciate the corollary advantage accruing to whites.

McIntosh conceptualized white privilege as “an invisible package of assets,”
including advantages such as: (1) protection from many kinds of hostility, distress and
violence; (2) freedom to criticize the dominant culture; (3) freedom to disparage, fear,
neglect or be oblivious to anything outside the dominant culture; and (4) a sense of
belonging in mainstream society. Consistent with the critical view of a hegemonic
ideology of race, McIntosh argued the potency of white privilege is evidenced by the fact
it is rarely discussed: “The silences and denials surrounding privilege are the key political
tool here. They keep the thinking about equality or equity incomplete, protecting
unearned advantage and conferred dominance by making these taboo subjects” (p. 12).

Swim and Miller (1999) were ﬁrst to empirically investigate the concepts of white

guilt and white privilege. They developed a ﬁve-item White Guilt Scale demonstrating

43

good internal consistency and correlated only moderately (r = -.23) with the Modern
Racism Scale (McConahay, I986). Signiﬁcantly, the White Guilt Scale predicted
endorsement of statements concerning support for afﬁrmative action, over and above
controls for gender, age, political orientation, and four explicit measures of racism. The
researchers also reported that mean white guilt across their student sample was low. They
attributed this ﬁnding to the fact their student participants were not old enough to have
experienced the racial strife of the 19605. Swim and Miller (1999) also developed a ﬁve-
item Belief in White Privilege Scale. This scale predicted attitudes toward afﬁrmative
action and was only moderately correlated with the White Guilt Scale (r = .23).

Iyer et al. (2003) pointed to a limitation of the Swim and Miller (1999) studies: all
the afﬁrmative action statements they tested were couched in term of the Remedial
Action frame, i.e., as a means necessary to recompense past harms. Iyer and colleagues
asserted that because white guilt is self-focused, i.e., primarily concerned with alleviation
of dysphoria, the extent to which it predicts support for afﬁrmative action should vary
depending on how a particular afﬁrmative action policy is framed. To test this idea, they
examined the power of white guilt to predict support for two different types of
afﬁrmative action statements: compensatory (remedial action) policies, and so-called soft
afﬁrmative action programs designed to increase minority recruitment efforts. Consistent
with their hypothesis, the researchers demonstrated that white guilt was a reliable
predictor of support for compensatory afﬁrmative action, but did not predict support for
soft afﬁrmative action policies. The researchers concluded that white guilt appears to

interact with afﬁrmative action framing:

44

Not only do levels of support for an afﬁrmative action program change

based on the way it is framed, but the psychological predictors of this

support may change as well. . .. Research on support for, or opposition to,

afﬁrmative action must take account of the fact that the deﬁnitions

employed may affect the psychological basis of support. Research that

deﬁnes afﬁrmative action as compensation, for example, is more likely to

show white guilt as a predictor than research that frames the policy as

concerned with equal opportunity (p. 127).

Iyer and colleagues (2003) also found that white privilege predicted support for
compensatory afﬁrmative action (i.e., invoking the RA frame). This effect was partially
mediated by white guilt. However, they did not test the association between belief in
white privilege and support for afﬁrmative action in the context of other frames.

Racial ambivalence

Katz and Hass (I988) theorized that the thoughts and feelings of whites toward
blacks are often characterized by a conﬂict between dual perceptions. Many whites
perceive that blacks are disadvantaged by social and economic discrimination. However,
many whites also perceive blacks as deviant, lacking essential attributes of mind and
body. These ideas are believed to be orthogonal—having sympathy for blacks as targets
of discrimination does not necessarily determine how a person thinks about what blacks
can and should be doing to improve their own lot. In other word, it is entirely possible to
believe that blacks are deserving of assistance and that blacks are not doing enough to
help themselves. Katz and Hass conceptualized racial ambivalence as the tension between

dual cognitive (value-attitude) structures existing simultaneously within individuals.

45

Katz and Hass (1988) developed scales to measure these dual cognitive structures.
The Pro-Black Affect Scale consists of ten items that tap liking and sympathy for blacks
as an underdog group. High scores on the Pro-Black Affect Scale suggest greater social
acceptance of blacks. The Anti-Black Affect Scale consists of ten items reﬂecting the
view that blacks lack qualities essential to uplift themselves, such as discipline, self-
respect, and concern for one another. High scores on the Anti-Black Affect Scale suggest
disdain and hostility toward blacks.

Katz and Hass (1988) further demonstrated that experimental primes can
inﬂuence Pro-Black affect and Anti-Black affect. Speciﬁcally, participants primed with a
scale measuring humanitarianism scored higher on Pro-Black affect; participants primed

with a scale measuring Protestant ethic tended to score higher on Anti-Black affect.

46

CHAPTER 3
Hypotheses

Social Identity Theory (Brewer, 1979; Tajfel & Turner, 1978; Turner, 1982)
posits that people like in-group members more and regard outcomes for in-group
members as their own. Increasing the salience of group boundaries and inter-group
conﬂict strengthens these effects (Brewer, 1979; Brewer & Harasty, 1996; Brewer &
Miller, 1984; Brown & Turner, 1981; Garcia-Prieto et al., 2003; Kramer & Brewer, 1984;
Tajfel & Turner, 1979; Turner, 1982, 1985). Procedural Justice Theory (Lind et al., 1993;
Lind & Tyler, 1988; Tyler, 1990; Tyler & McGraw, 1986; Tyler, et al., 1986) posits that
people support policies to the extent they view them to be procedurally fair.

These theories suggest the Diversity frame could induce whites to support
afﬁrmative action more. The Diversity frame can be viewed as a re-categorization,
shifting focus away from racial conﬂict (Murrell et al., 1994). Whites and blacks are
depicted as joint members of a super-ordinate organization and/or society, with common
goals and interests. Consistent with the proportionality rule of distribution (Thibault &
Walker, 1978), the Diversity frame argues that minorities receiving afﬁrmative action
beneﬁts give something in retum—a different perspective beneﬁcial to the entire group.
Moreover, the Diversity frame suggests a macro-level focus on the common welfare of
the super-ordinate group (Brewer, 1995). Therefore,

H1: White participants reading editorials presenting the Diversity frame

will support an afﬁrmative action policy more than white participants

reading editorials without the Diversity frame.

47

 

 

These theories also suggest the Remedial Action frame could induce whites to
support afﬁrmative action less. The Remedial Action frame heightens the salience of
racial identity, depicting whites and blacks as distinct groups with conﬂicting interests.
The Remedial Action frame can also be viewed as incongruent with the proportionality
rule—blacks receive beneﬁts, but their contribution in return is not emphasized (Barnes
Nacoste, 1994; Chang, 1996; Gamson & Modigliani, 1987). Therefore,

H2: White participants reading editorials that present the Remedial

Action frame will support the afﬁrmative action policy less than white

participants reading editorials without the Remedial Action frame.

Interracial attitudes such as modern racism, white guilt and belief in white
privilege can predict attitudes toward afﬁrmative action (Iyer et al., 2003; James, 2001;
Swim & Miller, 1999; Williams et al., 1999). However, media frames can moderate the
inﬂuence of individual predispositions on attitudes toward the subject matter (Domke,
2001; Domke et al., 1999; Price et al., 1997; Shen, 2004; Tewksbury et al., 2000). In the
context of race and political cognition, prior studies suggest frames can moderate the
relation between interracial attitudes and support for a political issue (Domke; Domke et
al.; Kinder & Sanders, 1990). The Diversity frame reduces the salience of racial group
identities. Afﬁrmative action is presented not as a debt whites owe to blacks, but instead
as a means to strengthen and beneﬁt the super-ordinate group. Accordingly, the Diversity
frame could switch whites’ train of thought away from their pre-existing interracial
attitudes. Therefore,

H3: The relation between interracial attitudes and support for an

afﬁrmative action policy will be weaker among white participants

48

reading editorials that present the Diversity Frame, compared to
white participants reading editorials without the Diversity frame.

Frames that differentially affect in-group bias should also inﬂuence interracial
affect. Interracial affect is susceptible to change due to situational cues (Katz & Hass,
1988). The need for self-esteem motivates people to feel favorably toward other in-group
members (Turner, 1985). When group identities are salient, people like in-group
members more. People discriminate in favor of in-group members even when the
criterion for membership is meaningless (Taj fel et al., 1971). The Diversity frame
suggests to whites that they share a super-ordinate group identity with blacks. This should
induce more favorable affect and less disdain toward blacks. In contrast, the Remedial
Action frame emphasizes that blacks and whites are different groups with conﬂicting
interests. This frame could have the opposite effect. Therefore,

H4: Editorials presenting the Diversity frame will induce white

participants’ affect toward blacks to become more favorable compared to

whites reading editorials without the Diversity frame.

H5: Editorials presenting the Remedial Action frame will induce white

participants’ affect toward blacks to become less favorable compared to

whites reading editorials without the Remedial Action frame.

49

CHAPTER 4
Method

An experiment was conducted to test the joint effects of media frames on support
for an afﬁrmative action policy and interracial affect. This study also pre-tested
participants’ interracial attitudes to investigate whether media frames moderated the
inﬂuence of those interracial attitudes on support for an afﬁrmative action policy. This
chapter describes the participants, procedures and instrumentation involved in conducting
this study.

Participants

Participants were recruited from telecommunication and advertising classes at
Michigan State University. Students received course extra credit for participating. Data
were collected in two phases: a pretest and an experiment.

A total of 436 students completed the pretest. The pretest instrument instructed
participants to provide the last four digits of their social security number “for coding
purposes.” Two weeks thereafter, a total of 461 students from the same classes
participated in the experiment. The experimental packets also instructed participants to
provide the last four digits of their social security number “for coding purposes.”

Comparison of these four-digit responses identiﬁed a total of 293 matching data
sets. Responses to demographic items indicated that among the matched data sets there
were 186 white Non-Hispanic, 41 Asian, 40 black, ﬁve Hawaiian/Paciﬁc Islander and
four Native American participants. Of the 41 Asian participants, 23 reported they were

citizens of countries other than the United States. There were eight Hispanic/Latino

50

participants, one of whom also identiﬁed herself as African American/Black. Among the
matching data sets, racial/ethnic data was missing for the remaining ten participants

Data analysis focused mainly on the 186 white non-Hispanic participants.
Because whites constitute the majority of the US. electorate and are also the group most
likely to oppose afﬁrmative action, understanding how frames affect their opinions and
underlying cognitive processes is a pressing social concern. Given the demographics of
the sampling frame (Michigan State University students), large numbers of white subjects
were more readily available for study. For the sake of brevity, the terms white and white
non-Hispanic appear interchangeably in the remainder of this paper.

Statistical power

The power of a statistical test is the probability that the test will detect an effect in
a sample when, in fact, a true effect exists in a population. Factors inﬂuencing the power
of a statistical test include: (1) the speciﬁc statistical test that is chosen, (2) the level of
signiﬁcance chosen, (3) the magnitude of the true effect in the population, and (4) sample
size (Cohen, 1992). Cohen suggested that .80 (an 80% chance of ﬁnding an effect in a
sample that truly exists in the population) is a good standard for the minimum power
necessary before undertaking an investigation. This standard has been widely accepted
throughout the social sciences (Aiken & West, 1991).

In calculating the number of subjects necessary to attain adequate power, a
complicating factor is that the magnitude of the true effect in the population is often
unknown a priori. Accordingly, a number of rules of thumb have emerged to guide
researchers when deciding how large a sample will sufﬁce. Green’s (1991) analysis of

these rules of thumb in the context of multiple regression analyses suggested that the

51

formula: N 2 50 + 8p, where N is the number of subjects and p is the number of predictor
variables, generally yields a sample size to ﬁnd medium-sized effects with power = .80.
The multiple correlation models tested here used up to ﬁve predictor variables,
suggesting a minimum sample of 90 participants. The sample of 186 white participants
used here would therefore appear sufﬁcient.
Procedure

Express approval for all instruments and procedures used in this study was
obtained from the University Committee on Research Involving Human Subjects at
Michigan State University. Faculty members teaching large undergraduate courses were
contacted and asked to consider offering an extra credit incentive to students for
participation in a two-phase study. Faculty members were asked not to advise students
that the studies were related, and to offer a separate extra-credit incentive for
participation in either or both phases. Data were collected during the Spring 2004 term.

Pretest

White graduate students administered the pretest in large lecture halls,
immediately at the end of regularly scheduled class time. Participants were advised the
researchers were conducting a survey to learn about students’ opinions toward a variety
of social issues. Participants were assured of the conﬁdentiality of their responses. After
obtaining their written informed consent, participants were instructed to complete the
pretest instrument and place the last four digits of their social security number on its face
“for coding purposes.” Written instructions provided to pretest participants are annexed

as Appendix A.

52

Participants completed the pencil-and-paper pretest instrument at their own pace.
Most completed the pretest in 15 minutes or less. When ﬁnished, participants placed their
completed pretest instruments in a large box, were thanked by the pretest administrator
and excused.

Experiment

Experimental data were collected approximately two weeks after the pretest from
each class. The experimental phase of the study was conducted in the same lecture halls,
immediately at the end of the same courses sampled in the pretest. A different White
graduate student conducted the experiment. Experimental packets used typeface, font
size, layout and language intentionally manipulated to appear different from the pretest
instrument.

Participants were instructed that the study required them to read a newspaper
editorial and answer questions about the issues involved. Participants were assured of the
conﬁdentiality of their responses. After obtaining their written informed consent,
participants were given an experimental packet containing one mock newspaper editorial
followed by scales to assess the dependent variables. Participants were instructed to read
the editorial carefully, paying close attention to the arguments the newspaper made to
support its editorial stance. Participants were advised the editorial was published in a
large metropolitan newspaper, the St. Louis Post-Dispatch. Participants were further
instructed to place the last four digits of their social security number on the instrument
“for coding purposes.” The written instructions provided to experimental participants are

annexed as Appendix B.

53

The St. Louis Post-Dispatch was chosen as the putative source of the
experimental stimuli for several reasons. First, it is a newspaper published in a major
market. Second, it is published in the Midwest, the same general region of the country as
the data collection site. Third, it is not published close enough to the data collection site
to present much chance that any participant had prior negative experiences with it.

Four different experimental packets were employed. Each packet included one
editorial systematically manipulated to reﬂect one editorial frame condition:
DIVERSITY, REMEDIAL ACTION, COMBINED, or CONTROL. Participants were
randomly assigned to frame conditions without regard to race. As a result, sub-samples of
participants from various racial/ethnic groups across frame conditions were unequal.
Independent sample t-tests examining white participants’ scores on the pre-tested
interracial attitude scales showed that random assignment to experimental conditions
successfully equalized the groups on these variables—none were signiﬁcantly different
from the control group.

Participants read the experimental stimuli and responded to the pencil-and-paper
dependent measures at their own pace. When ﬁnished, participants placed the completed
experimental packets in a large box and were given a document entitled: “Debrieﬁng
Form.” The debrieﬁng form advised participants that: (1) the editorial was ﬁctitious, (2)
the editorial was created by the researchers, not the St. Louis Post Dispatch, and (3) data
from the survey they participated in about two weeks earlier would be used to enhance
the researchers’ understanding of the responses provided in the experimental packet.

Participants were thanked for their participation and excused.

54

Instrumentation

Pretest

The pretest instrument included three interracial attitude scales: McConahay’s
(1986) seven-item Modern Racism Scale, Swim and Miller’s (1999, Study 1) ﬁve-item
White Guilt Scale, and a modiﬁed version of Swim and Miller’s ﬁve-item Belief in
White Privilege Scale. The original Belief in White Privilege Scale is phrased such that it
assesses whether white respondents believe that society gives them advantages because
they are white. By slightly adjusting the wording of the items, this scale was modiﬁed
such that it asked respondents from all racial/ethnic groups whether they believe that
society gives advantages to whites. In order to provide a consistent format, the interracial
attitude scales presented pretest participants with uniform response options: (1 = Strongly
Disagree, 2 = Disagree, 3 = Unsure, 4 = Agree, 5 = Strongly Agree).

The data were analyzed to assess the reliability of the interracial attitude scales. In
addition, conﬁrmatory factor analyses were conducted to test the implicit hypothesis that
the items within each scale were internally consistent (Hunter & Gerbing, 1982).
Conﬁrmatory factor analysis of McConahay’s (1986) Modern Racism Scale indicated
dropping one of the seven items (Item 2) would improve its internal consistency. The
resultant six-item scale demonstrated good reliability (Cronbach’s a = .85). Swim and
Miller’s (1999, Study 1) White Guilt Scale was internally consistent with good reliability
(Cronbach’s a = .84). Their ﬁve-item Belief in White Privilege Scale (as modiﬁed) was
internally consistency with excellent reliability (Cronbach’s a = .91).

A complicating issue was that the Modern Racism Scale (McConahay, 1986) and

White Guilt Scale (Swim and Miller, 1999, Study 1) are designed to assess the attitudes

55

of whites; when addressed to a person of color, some of the items make little sense (e.g.,
“When I learn about racism, I feel guilt due to my association with the White race”).
However, data from minority participants was desired for the other variables. Moreover,
to not provide minority students with equal access to an extra-credit opportunity would
have presented an ethical problem. In a study administered to large groups, using
different forms for minority and white participants could have created other
complications. Accordingly, it was decided to instruct respondents that: “Some questions
address issues involving race and do not apply equally to members of all groups. . . .”
Members of minority groups were advised they might wish to skip designated items (See
Appendix A for details).

Demographic information was also collected. Respondents were asked to indicate
their sex, age, race (African-American/Black, Asian, Hawaiian/Paciﬁc Islander, Native
American/Alaska Native or White/Caucasian) and ethnicity (Hispanic/Latino or Not
Hispanic/Latino) by checking the appropriate category. Political orientation was
measured using a ﬁve-point scale (1 = Strongly Conservative, 5 = Strongly Liberal).
Political party identiﬁcation was measured using a ﬁve-point scale (1 = Strong
Republican, 5 = Strong Democrat). Annual household income was measured using a 4-
category scale (1 = Under $25,000, 2 = $25,000-$50,000, 3 = $ 50,001-$75,000, 4 = Over
$75,000). Respondents were also asked whether they were: (1) US. citizens, (2)
registered to vote, and (3) voted in the 2002 General Election. Likelihood of voting in the
2004 General Election was measured using a ﬁve-point scale (1 = Very Low Likelihood,

5 = Very High Likelihood).

56

Experimental stimuli

The experimental stimuli were four mock newspaper editorials created for this
study. Each mock editorial was dated June 24, 2003 and purported to be the editorial
response of the St Louis Post-Dispatch to the US. Supreme Court’s ruling in Grutter v.
Bollinger (2003). To make the experimental stimuli appear authentic, the logo used on
newspaper’s actual Web site, St. Louis Today, was digitally copied and pasted to the top
of the stimuli. Typeface and layout were chosen to mimic the physical appearance of an
editorial printed out from the actual Web site. Also consistent with the actual St. Louis
Today Web site, no bylines were used. Copies of the experimental stimuli are annexed as
Appendices C1-C4.

Each mock editorial provided an identical headline and introductory paragraph,
advising that the US. Supreme Court ruled in favor of The University of Michigan’s
afﬁrmative action policy, treating membership in underrepresented minority groups
(African-American, Hispanic or Native American) as a “plus factor” for admission to law
school. Each mock editorial also provided an identical closing statement, expressing the
newspaper’s support for the decision on the grounds that higher education is important
and should be kept open for everyone.

The remainder of each editorial was systematically manipulated for use in one of
four frame conditions: DIVERSITY, REMEDIAL ACTION, COMBINED, or
CONTROL. In the DIVERSITY condition, the body of the editorial presented arguments
consistent with the Diversity frame. Bowen and Bok’s (1998) discussion of the case for
diversity, together with actual newspaper editorials analyzed by Richardson and

Lancendorfer (2004) were used to construct these arguments. The arguments in this

57

framing condition presented three diversity packages identiﬁed by Richardson and
Lancendorfer: LEARNING (emphasizing that afﬁrmative action enhances the
educational experience for all students), LEADERSHIP (emphasizing that afﬁrmative
action helps teach students to work and get along well with people different from
themselves), and COMPETE (emphasizing that afﬁrmative action promotes America’s
ability to compete in increasingly global and diverse markets).

In the REMEDIAL ACTION condition, the body of the editorial presented
arguments consistent with the RA frame. Gamson and Modigliani’s (1987)
conceptiralization of the RA frame, together with actual newspaper editorials analyzed by
Richardson and Lancendorfer (2004) were used to construct these arguments. The
arguments in this framing condition focused on three RA packages identiﬁed by
Richardson and Lancendorfer: HISTORY (emphasizing America’s past legacy of
racial/ethnic discrimination), CURRENT (emphasizing that discrimination is a
continuing problem), and KTWELVE (emphasizing disparities in elementary and
secondary schools located in minority communities).

The mock editorials used in the DIVERSITY and REMEDIAL ACTION
conditions were exactly the same length: 264 words. The goal was to present persuasive
arguments for afﬁrmative action emphasizing grounds consistent with each respective
frame, but without including arguments characteristic of the alternate frame.

In the COMBINED condition, the mock editorial presented all the pro-
afﬁrrnative action arguments used in both the DIVERSITY and REMEDIAL ACTION

conditions. The COMBINED editorial was 414 words long. In the CONTROL condition,

58

only the introductory and closing statements common to every other mock editorial were
presented. The CONTROL editorial was 113 words long.

Induction checks
The experimental packets included three true-false questions for use as induction checks
and instructed participants to: “. . .indicate your understanding of what the editorial
actually said (putting aside your own personal thoughts and feelings)” (Emphasis
original).

The ﬁrst induction check item asked whether the editorial asserted that without an
afﬁrmative action policy, minority enrollment at the University of Michigan law school
would drop. The introductory paragraph common to all four editorials made such a
statement. Therefore, it was anticipated that participants in all four conditions would
answer this question as true.

The second induction check item asde whether the editorial argued that racism is
a long-standing problem in America. This argument was presented in the REMEDIAL
ACTION and COMBINED experimental conditions, but was not mentioned in the
DIVERSITY or CONTROL experimental conditions. The third induction check item
asked whether the editorial argued that: “. .. a diverse student body is good for all
students.” This argument was speciﬁcally mentioned in the DIVERSITY and
COMBINED experimental conditions, but not in the REMEDIAL ACTION or
CONTROL experimental conditions.

Results from the induction check are presented in Table 4.1. On the ﬁrst two

induction check items, most respondents answered in the expected direction. However,

59

Table 4.1
Percentage of Respondents Answering Induction Check Items in the Expected

Direction as a Function of Editorial Frame Condition

 

Induction Check Item

 

 

Impact on Minority Racism Diversity Good
Experimental Condition Enrollment Longstanding for all Students
CONTROL 92.1 86.8 55.3
(n = 38)
REMEDIAL ACTION 90.6 88.7 49.1
(n = 53)
DIVERSITY 93.8 70.8 ' 97.9
(n = 48)
COMBINED 97.9 97.9 95.7
(n = 47)

 

Note: 11 = 186 white participants included in this portion of the data analysis.

about half the participants in the CONTROL and REMEDIAL ACTION conditions
answered the third induction check item contrary to expectation. In hindsight, this latter
ﬁnding probably resulted from an ambiguity in the induction check item. The closing
paragraph common to all four editorial inductions stated: “It is vital that the doors of
higher education be kept open for everyone.” The issue tapped by induction check item 3

was whether: “The editorial argued that a diverse student body is good for all students.”

60

Some participants in the CONTROL and REMEDIAL ACTION conditions might have
interpreted this closing statement of the editorial as an argument for diversity. Given the
high rate of correct responses for induction check items 1 and 2, as well as for item 3
among participants in the DIVERSITY and COMBINED conditions, the editorial
inductions were judged to have passed the checks.

Dependent measures

The dependent variable of primary interest was support for an afﬁrmative action
policy. A weakness of many prior studies is they purport to assess attitudes toward
afﬁrmative action as a generalized, abstract principle, ignoring the fact that afﬁrmative
action is an umbrella term encompassing a variety of different programs. Another
frequent shortcoming is the use of single-item dependent measures, an approach that
limits reliability.

Accordingly, a new scale was developed to assess support for the particular
afﬁrmative action policy discussed in the editorial stimuli, i.e. The University of
Michigan’s use of race as a “plus factor” when deciding law school admissions. The goal
of this scale was to measure the extent to which participants favored or opposed the
policy at The University of Michigan and all other colleges and universities throughout
Michigan—implicitly including the participants’ university, Michigan State.
Conﬁrmatory factor analysis (Hunter & Gerbing, 1982) indicated that of the original
eleven items, ten were internally consistent. The resultant scale was highly reliable
(Cronbach’s a = .95). Appendix D sets forth the ten—item Support for an Afﬁrmative

Action Policy Scale retained for analysis.

61

In addition, participants were also instructed to complete Katz and Hass’ (1988)
Pro—Black Affect Scale and Anti—Black Affect Scale. To be consistent with the remainder
of the instrument, Katz and Hass’ original response options were replaced with the
following: (1 = Strongly Disagree, 2 = Disagree, 3 = Unsure, 4 = Agree, 5 = Strongly
Agree). Conﬁrmatory factor analysis of responses to the Pro-Black Affect Scale indicated
dropping one item (Item 1) would improve its internal consistency. The resultant nine-
item scale demonstrated acceptable reliability (Cronbach’s a = .79). Internal consistency
of the Anti-Black Affect Scale was improved by dropping two items (Items 2 and 8). The
resultant eight-item scale demonstrated good reliability (Cronbach’s a = .81).

Like the pretest, a complicating issue was that the Pro-Black Affect Scale and
Anti-Black Affect Scale were designed primarily to tap whites’ affect toward blacks. A
procedure similar to that used in the pretest was proposed. However, the UCHRIS
reviewer objected to that procedure, suggesting instead that a special instruction be given.
Accordingly, it was agreed to insert the following instruction preceding these scales:

Questions 28-37 concern attitudes about race and include questions that

may be deemed inappropriate and offensive. We apologize for any

discomfort to respondents of these questions. You may choose to not

answer any of these questions. Using the scale below, write a number next

to each item that best describes your opinion. There are no right or wrong

answers. Please just give us your honest personal opinion.

62

CHAPTER 5
Results

Participant Demographics

Demographic information concerning blacks and white non-Hispanics
participating in both phases of the study is set forth in Table 5.1. Among black
participants, males were underrepresented compared to females. White participants’
mean income bracket was higher compared to black participants, t (224) = 3.217, p <
.001 (two-tailed). Black participants were more likely than white participants to identify
with the Democratic Party, t (224) = 5.949, p < .001 (two-tailed), and describe
themselves as liberal, t (220) = 2.065, p < .05 (two-tailed).
Interracial Attitudes

Table 5.2 summarizes white participants’ pretest responses on the three interracial
attitude scales. White participants’ mean scores were signiﬁcantly below the scale
midpoint (3 = “Unsure”) on the Modern Racism Scale, t (185) = -1 1.650, p < .001 (two-
tailed), White Guilt Scale, t (185) = -l3.161, p < .001 (two-tailed), and Belief in White
Privilege Scale (as modiﬁed), t (185) = -3.442, p < .001 (two-tailed). Black participants
scored higher on the Belief in White Privilege Scale (as modiﬁed) (M = 4.46, SD = .55),
compared to whites, t (224) = 11.039, p < .001 (two-tailed). Compared to the relation
between modern racism and white guilt, the relation between modern racism and belief in

white privilege was stronger, z = 3.574, p < .001.

63

Table 5.1

Black and White Participants’ Responses to Demographic Items

 

 

Whites Blacks
('1 = 186) (n = 40)

Sex

Female 49.2 % 66.7 %

Male - 50.8 % 33.3 %
Mean Age 21.8 (3.1) 22.8 (4.2)
Annual Household Income Bracket

Under $25,000 15.7 % 37.5 %

$25,001-$50,000 10.8 % 15.0 %

$50,001-$75,000 27.6 % 17.5 %

Over $75,000 45.4 % 30.0 %
Political Orientation a 2.97 (.98) 3.32 (.70)
Party Identiﬁcation b 3.02 (1.10) 4.13 (.88)
Registered to Vote 80.5 % 77.5 %
Voted in 2002 43.8 % 43.1 %
Likelihood of Voting in 2004 c 4.43 (1.07) 4.30 (1.29)

 

Note: Standard deviations in parentheses.
3 Based on ﬁve-point scale (Strong Conservative = 1, Strong Liberal = 5).
b Based on ﬁve -point scale (Strong Republican = 1, Strong Democrat = 5).

° Based on ﬁve-point scale (Very Low Likelihood = 1, Very High Likelihood = 5).

64

Table 5.2

Intercorrelations and Mean Item Responses to Interracial Attitude Scales among White

 

 

Participants

Interracial Attitude Scale 1 2 3 M SD
1. Modern Racism -- -.20 ** -.52 *** 2.33 .78
2. White Guilt -- .36 *** 2.18 .85
3. Belief in White Privilege -- 2.76 .93

 

Notes: 11 = 186. Values are mean item responses on ﬁve-point scales.

*p < .05, ** p < .01, *** p <.001(two-tailed).

Dependent Variables

Table 5.3 summarizes white participants’ responses on the dependent variable
scales. Consistent with prior research (Katz & Hass, 1988), Pro-Black affect correlated
moderately with Anti-Black affect. Compared to the relation between support for
afﬁrmative action and anti-black affect, the relation between support for afﬁrmative
action and Pro-Black affect was stronger, z = 4.225, p < .001. White participants’ mean
scores were signiﬁcantly below the midpoint (3 = “Unsure”) on the Support for an
Afﬁrmative Action Policy Scale, I (185) = -9.075, p < .001 (two-tailed), Pro-Black Affect
Scale, I (184) = -3.641, p < .001 (two-tailed) and Anti-Black Affect Scale, t (185) = -

5.165, p < .001 (two-tailed).

65

Table 5.3

Intercorrelations and Mean Item Responses to Dependent Variable Scales among White

 

 

Participants

Dependent Variable Scale 1 2 3 M SD

1. Support for Afﬁrmative Action -- .63 *** -.26 *** 2.42 .87
2. Pro-Black Affect -- -.29 *** 2.85 .57 >
3. Anti-Black Affect -- 2.73 .72

 

Notes: 11 = 186. Values are mean item responses on ﬁve-point scales.

* p < .05, ** p < .01, *** p <.001(two-tailed).

Tests of the Hypotheses

The following sections describe data analyses used to test the hypotheses. No
speciﬁc hypothesis predicted the relation between participant sex and the dependent
variables. Post hoc multiple regression analyses adding this variable to the equations
demonstrated participant sex was not signiﬁcant predictor of any dependent variable
discussed below.

White participants ’ support for an affirmative action policy

Hypothesis I predicted white participants reading editorials that presented the
Diversity frame would support the afﬁrmative action policy more compared to those
reading editorials without the Diversity frame. Hypothesis 2 predicted white participants

reading editorials that presented the Remedial Action frame would support the

66

Table 5.4

Mean Support for an Affirmative Action Policy among White Participants as a Function

 

 

 

of Editorial Frame
DIVERSITY Frame
REMEDIAL ACTION Frame Absent Present
2.29 (.80) 2.59 (.90) 2.43 (.86)
Present
11 = 53 n = 47 n = 100
2.42 (.89) 2.41 (.88) 2.42 (.88)
Absent
n = 38 n = 48 n = 86
2.34 (.84) 2.50 (.89) 2.42 (.87)
n = 91 n = 95 n = 186

 

 

 

Notes: Mean-item responses to a Support for an Afﬁrmative Action Policy Scale

(Average response on a ﬁve-point scale across ten items retained for analysis). Higher

scores indicate more support for the law school admissions policy upheld in Grutter v

Bollinger (2003). Standard deviations in parentheses.

afﬁrmative action policy less compared to those reading editorials without the Remedial

Action frame.

Table 5.4 sets forth white participants’ mean item responses to the Support for an

Afﬁrmative Action Policy Scale as a function of editorial frame. Multiple comparison

67

Table 5.5

Support for an A ﬁirmative Action Policy among White Participants Regressed on Scores

on Interracial Attitude Scales and Editorial Frame

 

mu
Modern Racism -.29 ***
White Guilt .24 ***
Belief in White Privilege .17 *
Remedial Action Frame
Diversity Frame
Adjusted R2 .26
A F 23.73 ***

mid;
-.28 ***
.24 ***
.18 *
.04
.03
.26

.29

 

Notes: n = 186 white participants. Numbers in the table are standardized beta coefﬁcients.

*p < .05, ** p < .01, *** p < .001 (two-tailed).

and contrast testing indicated no signiﬁcant between-groups differences in support for the

afﬁrmative action policy.

Support for afﬁrmative action was also examined in a multiple regression

analysis. Model 1 regressed this dependent variable on White participants’ pretest scores

on the interracial attitude scales. Model 2 added as independent variables the editorial

frame inductions, dummy coded for the presence (coded 1) or absence (coded O) of the

Remedial Action and Diversity frames. Table 5.5 summarizes the results of these

analyses.

68

Examination of these data reveals that scores on the pretest interracial attitude
scales explained 26% of the variance in white participants’ support for the afﬁrmative
action policy. The corresponding F ~Change value indicates these variables were
signiﬁcant predictors of support for the afﬁrmative action policy.

In Model 2, the standardized beta coefﬁcients indicate neither editorial frame
induction signiﬁcantly inﬂuenced support for the afﬁrmative action policy. Adding the
editorial frame inductions to the original model did not increase the amount of variance
explained. The corresponding F -Change value indicates the addition of the editorial
frame inductions had no signiﬁcant effect on the original model’s capacity to predict
support for the afﬁrmative action policy.

The data were not consistent with Hypotheses l and 2. Neither the Diversity nor
the Remedial Action editorial frame inductions signiﬁcant affected white participants’
support for the afﬁrmative action policy.

The moderating inﬂuence of the Diversity frame

Hypothesis 3 predicted support for the afﬁrmative action policy would be less
strongly associated with interracial attitudes among white participants reading editorials
presenting the Diversity frame, compared to those reading editorials without the Diversity
frame. Table 5.6 summarizes the results of analyses of support for an afﬁrmative action
policy regressed on participants’ scores on interracial attitude scales between
experimental groups.

Examination of these data reveal that white participants’ pretest interracial
attitude scores were powerful predictors of their support for the afﬁrmative action policy

in the CONTROL and REMEDIAL ACTION conditions, explaining roughly half the

69

Table 5.6
Support for an Affirmative Action Policy Between Experimental Conditions Regressed on

Scores on Interracial Attitude Scales among White Participants

 

Editorial Frame Condition

 

Control Remedial Action Diversity Combined
n=38 n=53 n=48 n=47
Modern Racism -.09 -.45 *** -.18 -.36 *
White Guilt .28 .28 ** .21 .16
Belief in White Privilege .51 ** .23 -.01 .06
Adjusted 1t2 .54 .49 .03 .14
A F 15.67 *** 17.65 *** 1.52 3.43 *

 

variance. The corresponding F -Change values demonstrate the signiﬁcance of these
relations.

However, among white participants in the DIVERSITY condition, i.e., when
participants read editorials presenting the Remedial Action frame but not the Diversity
frame, interracial attitude scale scores explained only three percent of the variance in
support for the afﬁrmative action policy. Put another way, compared to white participants
in the DIVERSITY condition, the relation between interracial attitudes and support for
the afﬁrmative action policy was 18 times stronger in the CONTROL condition and more
than 16 times stronger in the REMEDIAL ACTION condition. The corresponding F -

Change value demonstrates that among white participants in the DIVERSITY condition,

70

interracial attitude scores were not signiﬁcant predictors of support for the afﬁrmative
action policy.

The data were fully consistent with Hypothesis 3. Editorials presenting the
diversity frame moderated the relation between white participants’ interracial attitudes
and their level of support for the afﬁrmative action policy. When white participants read
editorials presenting the Diversity frame, the relation between their interracial attitudes
and their support for the afﬁrmative action policy was weaker compared to participants in
the other experimental groups.

White participants’ affect toward blacks

Hypothesis 4 predicted that editorials presenting the Diversity frame would
induce white participants’ affect toward blacks to be more favorable compared to those
reading editorials without the Diversity frame. Hypothesis 5 predicted editorials
presenting the Remedial Action frame would induce white participants’ affect toward
blacks to be less favorable compared to those reading editorials without the Remedial
Action frame. Table 5.7 sets forth white participants’ mean-item responses to the Pro-
Black Affect Scale as a function of editorial frame.

Examination of these data reveals that compared to the other three experimental
conditions, Pro-Black affect was lower among white participants in the REMEDIAL
ACTION frame condition, i.e., when participants read editorials presenting the Remedial
Action frame but not the Diversity frame. Contrast testing indicated Pro-Black affect
among this group was signiﬁcantly lower, t (181) = -3.870, p < .001 (two-tailed),

compared to the other three experimental groups.

71

Table 5.7

Pro-Black Aﬂect among White Participants as a Function of Editorial Frame

 

 

 

DIVERSITY Frame
REMEDIAL ACTION Frame Absent Present
2.59 (.63) 2.94 (.47) 2.76 (.58)
Present
n = 53 n = 47 n = 100
2.84 (.58) 3.04 (.51) 2.95 (.55)
Absent
r1 = 38 n = 47 a n = 85
2.70 (.62) 2.99 (.49) 2.85 (.54)
n = 91 n = 94 n = 185

 

 

 

Notes: Mean-item responses to Pro-Black Affect Scale (Average response on a ﬁve-point
scale across nine items retained for analysis). Higher scores indicate more positive affect
toward blacks. Standard deviations in parentheses.

“ Data on this variable were missing for one white participant.

Multiple regression analyses were also conducted. Model I examined the relation
between White participants’ pretest scores on the interracial attitude scales and their
scores on the Pro-Black Affect scale. Model 2 added editorial frame inductions as

predictor variables, dummy coded for the presence (coded 1) or absence (coded O) of the

72

 

Table 5.8
Pro-Black Affect among White Participants Regressed on Scores on Interracial Attitude

Scales and Editorial Frame

 

Mum M2
Modern Racism -_41 Hair -.38 slur-six
White Guilt 29 *** .25 *ilulr
Belief in White Privilege .22 *** .25 am
Remedial Action Frame -.12 It:
Diversity Frame .17 He
Adjusted R2 .48 .53
A F 58.43 m 8.90 ***

 

Notes: n = 185 white participants (Data missing on this variable from one white
participant). Numbers in the table are standardized beta coefﬁcients.

*p < .05, I""‘p < .01, ***p < .001 (two-tailed).

Diversity and Remedial Action frames. Table 5.8 summarizes the results of these
analyses.

Examination of these data reveals that the pretest interracial attitude scores used
as predictors in Model I explained 48% of the variance in pro-black affect. By adding the
editorial frame inductions to the model, explained variance rose to 53%, a signiﬁcant

improvement as demonstrated by the corresponding F -Change value.

73

Table 5.9

Anti-Black Affect among White Participants as a Function of Editorial Frame

 

 

 

DIVERSITY Frame
REMEDIAL ACTION Frame Absent Present
2.89 (.78) 2.68 (.81) 2.80 (.80)
Present
11 = 53 n = 47 n = 100
2.58 (.48) 2.70 (.71) 2.65 (.62)
Absent
n = 38 n = 48 n = 86
2.76 (.69) 2.69 (.76) 2.73 (.72)
n = 91 n = 95 n = 186

 

 

 

Notes: Mean-item responses to Anti-Black Affect Scale (Average response on a ﬁve-

point scale across eight items retained for analysis). Higher scores indicate more negative

affect toward blacks. Standard deviations in parentheses.

The joint effects of editorial frame and interracial attitudes on anti-black affect

were also examined. Table 5.9 sets forth white participants’ mean-item responses to the

Anti-Black Affect Scale as a function of editorial frame.

Examination of these data reveals that anti-black affect was somewhat higher

among white participants in the REMEDIAL ACTION frame condition, i.e., when

participants read editorials presenting the Remedial Action frame but not the Diversity

frame. Multiple comparison testing demonstrated that when the Diversity frame was

74

Table 5.10

A nti-Black Affect among White Participants Regressed on Scores on Interracial Attitude

Scales and Editorial Frame

 

DAQQELI
Modern Racism .45***
White Guilt -. 1 3
Belief in White Privilege .07
Remedial Action Frame
Diversity Frame
Adjusted R2 .20
A F 16.37***

bdcdelz
.47***
al3
.08
11
.05
.20

1.42

 

Notes: 11 = 186 white participants. Numbers in the table are standardized beta coefﬁcients.

*p < .05, ”p < .01, ***p < .001 (two-tailed).

absent, the Remedial Action frame had a signiﬁcant simple effect on Anti-Black affect, t

(89) = -2.l97,p < .05.

Multiple regression analyses were also conducted. Model I examined the relation

between White participants’ pretest scores on the interracial attitude scales and their

scores on the Anti-Black Affect scale. Model 2 added as predictor variables the editorial

frame inductions, dummy coded for the presence (coded l) or absence (coded O) of the

Remedial Action and Diversity frames. Table 5.10 summarizes the results of these

analyses.

75

Examination of these data reveals the pretest interracial attitude scores included in
Model I explained 20% of the variance in anti-black affect. The corresponding F -Change
value indicates these variables were signiﬁcant predictors of anti-black affect.

In Model 2, the standardized beta coefﬁcients indicate neither editorial frame
induction had a signiﬁcant main effect on anti-black affect. Adding the editorial frame
inductions to the original model did not increase the amount of variance explained. The
corresponding F -Change value was also insigniﬁcant.

The data were partially consistent with Hypotheses 4 and 5. When white
participants read editorials presenting the Remedial Action frame but not the Diversity
frame, i.e., the REMEDIAL ACTION condition, pro-black affect was signiﬁcantly lower
compared to all other experimental groups. Compared to the CONTROL group, white
participants in the REMEDIAL ACTION condition also scored higher on anti-black
affect.

Additional Analyses

No speciﬁc hypothesis was made concerning the effect of editorial frame on black
participants’ support for the afﬁrmative action policy. Nevertheless, Table 5.11 sets forth
black participants’ mean-item responses to the Support for an Afﬁrmative Action Policy
Scale as a function of editorial framing condition.

Support for an afﬁrmative action policy was signiﬁcantly higher among black
participants compared to whites, t (222) = 11.053, p < .001. Interestingly, blacks in the
REMEDIAL ACTION frame condition supported afﬁrmative action most. Unfortunately,
ﬁrm conclusions cannot be drawn from these data due to the small number of black

participants. However, future studies should examine whether this frame is more

76

Table 5.11

Support for an A ﬂirmative Action Policy among Black Participants as a Function of

 

 

 

Editorial Frame
DIVERSITY Frame
REMEDIAL ACTION Frame Absent Present
4.26 (.63) 3.97 (.92) 4.13 (.75)
Present
11 = 9 n = 7 n = 16
4.01 (.82) 4.12 (.65) 4.06 (.74)
Absent
n = 13 n = 9 n = 22
4.11 (.74) 4.06 (.75) 4.09 (.74)
n = 22 n = 16 n = 38

 

 

 

Notes: Mean-item responses to a Support for an Afﬁrmative Action Policy Scale

(Average score on a 5-point scale across 10 items retained for analysis). Higher scores

indicate more support for the law school admissions policy upheld in Grutter v Bollinger

(2003). Standard deviations in parentheses.

effective among black message receivers, and if so, the causes thereof.

Appendices E and F set forth the correlations among the demographic variables,

interracial attitudes, and the dependent variables for white and black participants,

respectively. Among white participants, males scored higher on the Modern Racism

Scale (r = .26, p < .001) and Anti-Black Affect Scale (r = .24, p < .001).

77

CHAPTER 6
Discussion

Race and ethnicity are issues regarded as something of a hot potato by many
media scholars. Studies have expressed concern, with some justiﬁcation, that media
content mentioning or even hinting at race or ethnicity will induce audience members to
respond in a prejudicial manner. Domke (2001), for example, argued that media
references to race-associated cues activate racial stereotypes within the minds of the
audience, increasing the extent to which these stereotypes are cognitively embedded and
politically enmeshed.

Certainly there might be situations when the best course for the media would be to
refrain from referencing race and ethnicity. At the same time, it is important to recognize
that media references to race and ethnicity should not be made taboo. Real problems of
inequality, prejudice and discrimination still persist—failure to discuss race and ethnicity
would only serve to perpetuate, not alleviate them. Some issues, such as afﬁrmative
action and bilingual education, cannot be understood without reference to race and
ethnicity. Perhaps even more fundamentally, eliminating racial and ethnic cues from the
media threatens to deprive our society of many rich contributions that people of color
have to offer.

Keeping quiet about race and ethnicity, in the hope negative stereotypes might
remain dormant, is not a solution. We can tiptoe around the problem, but that does not
make the problem go away. Fear of activating negative stereotypes should not induce us

to treat race and ethnicity as stains to be cleansed from our media artifacts.

78

People of color have good reason to take pride in their racial/ethnic identities.
The best way to counter the negative pictures in audience members’ heads is to provide a
steady stream of positive pictures as replacements. The focus should be to investigate
ways race and ethnicity might be reframed so as to weaken and/or break down negative
stereotypes, and improve interracial ideations.

This study was one small step in that direction. Results indicated that editorials
presenting the Diversity frame weakened the linkage between interracial attitudes and
support for an afﬁrmative action policy. The Diversity frame editorial represented a
media artifact laided with racial cues that squarely addressed a racially charged political
issue. Nevertheless, it actually deactivated the inﬂuence of white participants’
preconceptions about race.

Consistent with Katz and Hass’s (1988) conceptualization of racial ambivalence,
the evidence presented here highlights another important consideration: the interracial
attitudes of a given individual can include ideations that are positive, negative or both.
Media frames that moderate the inﬂuence of interracial beliefs on political cognition can
reduce the impact of negative interracial attitudes, but they can also reduce the impact of
pro-social interracial attitudes, such as white guilt and belief in white privilege. Reducing
the inﬂuence of racial preconceptions upon the political cognition equation might not be
advisable in all cases.

The long-terrn consequence of reducing the linkage between support for
afﬁrmative action and pro-social interracial attitudes, i.e., white guilt and belief in white
privilege, bears consideration. Could weakening these linkages eventually undermine

support for afﬁrmative action? Future research should address this question. In the

79

meantime, a two-pronged argument combining the Remedial Action and Diversity frames
might be a reasonable compromise.

The results of this study also suggest that media frames can alter interracial affect.
White participants reading editorials that presented the Remedial Action frame but not
the Diversity frame scored lower on the Pro-Black Affect Scale, compared to the other
three experimental conditions, and scored higher on the Anti-Black Affect Scale,
compared to the control group. For afﬁrmative action advocates, this ﬁnding has
important and practical social implications. Presented alone, the Remedial Action frame
appears counter-productive for white audience members. It did not induce any signiﬁcant
change in support for afﬁrmative action, but it did induce whites’ affect toward blacks to
become less favorable. Consistent with the ﬁndings reported by Sniderrnan and Piazza
(1996), these data reveal that ironically, a frame intended to improve interracial harmony
induced the opposite effect.

The ﬁndings reported here are consistent with Social Identity Theory (Brewer,
1979; Taj fel & Turner, 1979; Turner, 1982). Because the Diversity frame recasts
afﬁrmative action as a means to beneﬁt the entire super-ordinate group, it might serve to
reduce the salience of racial identity. The weaker linkage between interracial ideations
and support for afﬁrmative action among whites exposed to the Diversity frame editorial
seems to nicely ﬁt this theoretical expectation. Because the Remedial Action frame
emphasizes a conﬂict of interest between whites and members of underrepresented
minority groups, Social Identity Theory would predict that this frame would increase
whites’ bias against blacks. The ﬁndings of less favorable affect toward blacks among

whites reading the Remedial Action editorial are consistent with this expectation.

80

However, further research to elucidate the connection between Social Identity
Theory and these afﬁrmative action frames is needed. For example, does the Diversity
frame increase whites’ aptness to regard blacks as members of the in-group? Does the
Remedial Action frame induce whites to identify more closely with white candidates not
included in the beneﬁciary group? Future studies could employ dependent variables such
as these to probe this connection more directly.

Contrary to expectations, the frame inductions did not signiﬁcantly inﬂuence
support for afﬁrmative action. Similarly, Kinder and Sanders (1990) found no signiﬁcant
change in support for equal opportunity policies between groups exposed to either of two
different No Preferential Treatment frames. They argued such a result was inevitable
because people hold strong opinions toward afﬁrmative action and because their frame
inductions were on the same side of the issue, i.e., they tested two anti-afﬁrmative action
frames. Given that both frames in this experiment supported afﬁrmative action, Kinder
and Sanders might have predicted no between-groups difference here.

But there are many other possible explanations as well. The experimental
inductions used here might simply have been too weak. A longer, more poignant message
about the beneﬁts of diversity might have induced much stronger effects. The assertion
that diversity is beneﬁcial might not ring true for many people. If people do not view
diversity as a better justiﬁcation than remedial action, Procedural Justice Theory would
predict no difference in support for afﬁrmative action. Future research should explore the
connection between Procedural Justice Theory and afﬁrmative action frames more
directly. For instance, whites exposed to the Diversity frame believe that afﬁrmative

action is more consistent with the proportionality rule, in that minority beneﬁciaries give

81

something back in the form of a different perspective? Does the diversity frame increase
the likelihood that people will view afﬁrmative action from a macro-level perspective?
Are whites exposed to the diversity frame more likely to believe that afﬁrmative action
can beneﬁt not only minorities, but whites as well?

The data reported here showed that among whites exposed to the Diversity frame,
interracial attitudes had no signiﬁcant inﬂuence on their support for an afﬁrmative action
policy. The question then becomes, what factors do predict support for afﬁrmative action
among participants exposed to this frame? Open-mindedness, one of the Big Five
personality traits, is one possibility that would seem to merit exploration. Another is
novelty, i.e., future studies might examine the extent to which prior familiarity with the
Diversity frame interacts with its persuasive effect on support for afﬁrmative action.

Limitations of this study include the artiﬁciality of the experimental setting.
Results from student participants instructed in college classrooms to pay close attention
to mock editorials might not generalize to newspaper readers in the real world. However,
real world newspaper readers are exposed to media repeatedly, not just once. Here, one-
shot exposure to 264-word essays induced signiﬁcant effects.

Another limitation was use of a convenience sample of students. Student samples
have restricted ranges of age, intelligence, education and work experience. Moreover,
students do not reﬂect the demographic or geographic diversity of the general population.
Prior commentators have observed an alarming trend in research about race: whites are
asked to report their attitudes toward minorities and issues affecting minorities, while
minorities are treated as passive targets whose attitudes don’t seem to matter (See e.g.,

Shelton, 2000). The study reported here leaves an important gap—the joint effects of

82

media frames and individual predispositions on minorities’ political attitudes. Data from
geographic areas outside the State of Michigan is also needed.

The use of one editorial exemplar per condition was another limitation. Careful
effort was made to design experimental stimuli that would exemplify a persuasive case
prototypical of each respective frame. Nevertheless, every message of any length is in
some sense idiosyncratic—this limits the ability to generalize from a speciﬁc message to
a targeted message population (Morley, 1988). Future studies should test the effects of
these frames with newly developed stimuli. The wording of messages should be varied.
Future studies should also test these frames across a variety of different media (e. g.,
television, Internet) and content formats (e.g., straight news reports, editorial content,
statements by news sources).

Afﬁrmative action is an umbrella term for a wide range of policies and programs.
This study focused on support for a speciﬁc afﬁrmative action policy employed at The
University of Michigan and afﬁrmed in Grutter v. Bollinger (2003). Future research
should examine the inﬂuence of frames on a variety of other afﬁrmative action policies,
and other political issues as well. The University of Michigan afﬁrmative action cases
received extensive national and local media attention. The administration of Michigan
State University provided its student body with access to a variety of information on this
controversy. Participants sampled from Michigan State University might have entered the
experiment with entrenched positions particularly resistant to frame effects. A sample
from a more distant pool, and/or an induction concerning a policy that was less well

known, could potentially have yielded larger effects.

83

Lippmann (1922) observed that: “To traverse the world men must have maps of
the world” (p. 16). For too many citizens, internal maps used to evaluate political issues
and make day-to-day choices are ﬂawed by negative racial and ethnic stereotypes.
Improving this misguided state of affairs requires newer, more accurate maps; maps that
simultaneously respect the unique value of all racial and ethnic groups, and that
appreciate our connectedness as one human family. Discovering how these new maps can
be drawn most effectively, and how citizens might be induced to put their old maps away,

merits further study.

84

Appendix A
Instructions to Pretest Participants
STUDFY INSTRUCTIONS

Thank you for agreeing to participate in this study. We are interested in learning
about students’ feelings toward a variety of social issues, and their news media habits.

Please read the questions and answer them in the order presented. There are no
right or wrong answers. Please be as candid, forthright, and honest as possible.

Some questions concern race and ethnicity and do not apply to members of all
groups. African-American/Black people might wish to skip questions 42 to 53. Members
of other racial and/or ethnic minority groups might wish to skip questions 49 to 53. We
apologize in advance if any question seems offensive.

Your responses to this questionnaire will be kept conﬁdential to the maximum
extent provided by law. For coding purposes, please place the last four numbers of your

social security number below.

Last Four Digits of SS#:

 

85

Appendix B

Instructions to Experiment Participants
W

This study requires you to read an editorial published in a large metropolitan
newspaper, the St. Louis Post-Dispatch. Please read the editorial carefully,
paying close attention to the arguments the newspaper made to support its
editorial stance. After carefully reading the editorial, please answer each

question in the order presented.
Your responses to this questionnaire will be kept conﬁdential to the

maximum extent provided by law. For coding purposes, please place the last

four numbers of your social security number below.

Last Four Digits of SS#: _

86

Appendix C-l
Experimental Stimulus for Control Framing Condition

JAR,
l.3I.llIII_lliIa;pé.

M 3! (00's WEBSITE

EDITORIAL: Affirming afﬁrmative action
06/24/2003

In a 5-4 ruling on Monday, the US. Supreme Court decided in favor of the afﬁrmative
action program at the University of Michigan law school. Under the program, the school
considers race and ethnicity a “plus factor” for applicants who are African American,
Native American or Hispanic. An expert testiﬁed that without afﬁrmative action,
minority enrollment at the school would drop from 14.5 to 4 percent.

The ruling essentially saved afﬁrmative action as an effective means to promote minority

enrollment on campus. Graduate schools are important gateways to power, inﬂuence and
success in life. It is vital that the doors of higher education be kept open for everyone.

87

Appendix 02

Experimental Stimulus for Remedial Action Framing Condition

.”‘\,
tilliltllﬂj.

#1 SI (OMS WEB SITE

EDITORIAL: Afﬁrming afﬁrmative action
06/24/2003

In a 5-4 ruling on Monday, the US. Supreme Court decided in favor of the afﬁrmative
action program at the University of Michigan law school. Under the program, the school
considers race and ethnicity a “plus factor” for applicants who are Aﬁican American,
Native American or Hispanic. An expert testiﬁed that without afﬁrmative action,
minority enrollment at the school would drop from 14.5 to 4 percent.

The court recognized that the effects of slavery and prejudice marring this nation’s
history still remain. Racism has infected this country for hundreds of years. Afﬁrmative
action exists because prejudice does damage over generations. The children of a man
denied his full potential are themselves hurt by bias, kept from opportunities they would
otherwise have and certainly deserve. This means that, even if one could magically make
all prejudice disappear, there would still be those who suffer from the ugliness of an
earlier time. Without some kind of acknowledgement of this legacy of hate, society
cannot hope to overcome it.

Even today, children in minority communities attend schools that are less well funded
and more dangerous than schools in afﬂuent, predominately White communities.
Afﬁrmative action means just what it says—taking afﬁrmative steps to address the fact
that opportunities have been denied to people because of institutionalized racism, both
past and present.

The ruling essentially saved afﬁrmative action as an effective remedy for past and present

discrimination. Graduate schools are important gateways to power, inﬂuence and success
in life. It is vital that the doors of higher education be kept open for everyone.

88

Appendix C-3

Experimental Stimulus for Diversity Framing Condition

.~’~m\,

a,

1‘ SII tIIIlIlII- -
L-..-._-. lawn,“ .5. ‘ ..

#151 (0018 WEB SITE

EDITORIAL: Afﬁrming afﬁrmative action
06/24/2003

In a 5-4 ruling on Monday, the US. Supreme Court decided in favor of the afﬁrmative
action program at the University of Michigan law school. Under the program, the school
considers race and ethnicity a “plus factor” for applicants who are African American,
Native American or Hispanic. An expert testiﬁed that without afﬁrmative action,
minority enrollment at the school would drop from 14.5 to 4 percent.

Afﬁrmative action improves the educational experience for everyone, not just minorities.
A racially diverse student body brings different perspectives to the classroom,
dormitories and campus. It promotes interracial understanding and breaks down
stereotypes. College students learn from books and professors, but they also learn from
each other. College is a time when students encounter a wide variety of people on an
intimate, day-to-day basis. Indeed, a big reason why young people go away to college is
the idea there is something they can learn from new people and surroundings.

Global companies seek qualiﬁed employees who can work with a diverse workforce and
serve diverse customers. Learning to work effectively and get along with people from
different backgrounds is an essential job skill. The court Monday upheld a sensible
principle by allowing university ofﬁcials the discretion to build a diverse community of
students that can learn important lessons for life from each other.

The ruling essentially saved afﬁrmative action as an effective means to promote diversity

on campus. Graduate schools are important gateways to power, inﬂuence and success in
life. It is vital that the doors of higher education be kept open for everyone.

89

Appendix C-4

Experimental Stimulus for Combined Framing Condition

{‘5
1310111111 -
i~a--3...- L i .11}.

... .14-...“.8

#15! [MRS WEB SITE

EDITORIAL: Afﬁrming afﬁrmative action
06/24/2003

In a 5-4 ruling on Monday, the US. Supreme Court decided in favor of the afﬁrmative
action program at the University of Michigan law school. Under the program, the school
considers race and ethnicity a “plus factor” for applicants who are African American,
Native American or Hispanic. An expert testiﬁed that without afﬁrmative action,
minority enrollment at the school would drop from 14.5 to 4 percent.

Afﬁrmative action improves the educational experience for everyone, not just minorities.
A racially diverse student body brings different perspectives to the classroom,
dormitories and campus. It promotes interracial understanding and breaks down
stereotypes. College students learn from books and professors, but they also learn from
each other. College is a time when students encounter a wide variety of people on an
intimate, day-to-day basis. Indeed, a big reason why young people go away to college is
the idea there is something they can learn from new people and surroundings.

Global companies seek qualiﬁed employees who can work with a diverse workforce and
serve diverse customers. Learning to work effectively and get along with people from
different backgrounds is an essential job skill. The court Monday upheld a sensible
principle by allowing university ofﬁcials the discretion to build a diverse community of
students that can learn important lessons for life from each other.

The court also recognized that the effects of slavery and prejudice marring this nation’s
history still remain. Racism has infected this country for hundreds of years. Prejudice
does damage over generations. The children of a man denied his full potential are
themselves hurt by bias, kept from opportunities they would otherwise have and certainly
deserve. This means that, even if one could magically make all prejudice disappear, there
would still be those who suffer from the ugliness of an earlier time. Without some kind of
acknowledgement of this legacy of hate, society cannot hope to overcome it.

Even today, children in minority communities attend schools that are less well funded
and more dangerous than schools in afﬂuent, predominately White communities.
Afﬁrmative action means just what it says—taking afﬁrmative steps to address the fact
that opportunities have been denied to people because of institutionalized racism, both
past and present.

90

The ruling essentially saved afﬁrmative action as an effective means to promote diversity
and redress discrimination. Graduate schools are important gateways to power, inﬂuence
and success in life. It is vital that the doors of higher education be kept open for
everyone.

91

Appendix D
Support for an Affirmative Action Policy Scale
Questions 1-10 ask for your personal opinion about UM law school’s policy for admitting
students. Using the scale below, write a number next to each item that best describes your
opinion. There are no right or wrong answers. Please just give us your honest personal
opinion.
Strongly Disagree Disagree Unsure Agree Strongly Agree
1 2 3 4 5

1. All colleges and universities should have an afﬁrmative action program like the
one at Michigan law school

2. Michigan law should be changed to make afﬁrmative action illegal. (Reverse-
Coded)
__ 3. It is reasonable for Michigan colleges and universities to give special
consideration for admission to Minorities.

4. Afﬁrmative action at Michigan colleges and universities is good for society as a
whole.

5. I believe the Supreme Court was wise to uphold afﬁrmative action at Michigan
law school.

6. I would be willing to attend a student rally to show my support for afﬁrmative
action at Michigan colleges and universities.

7. I would be willing to donate one dollar ($1.00) to promote afﬁrmative action in

Michigan colleges and universities.

92

__ 8. I would be willing to vote for a candidate for Governor of Michigan who
promised to end afﬁrmative action at Michigan colleges and universities. (Reverse-
Coded)

__ 9. Overall, Michigan law school’s afﬁrmative action policy is just a bad idea.
(Reverse-Coded)

10. I support the use of afﬁrmative action at Michigan colleges and universities.

93

one coma-m :3 :5. v e..-...... .5. v at. .8. v a... .3 1 2a: a n oaaob 881885.. .2: n c aeoz

 

poo-ca. xosmrzﬁx .:

 

atom.- -- one? sea-ea .2
...-8... .18. .- o=oa 5. ea ac 8&3 .e
...-...3. ...-an- ...-3..- -- sauna .582 a
2.8.- :S. ......cm. .- 3- -- :30 223 .e
ta.- ..ram. :- :.. ...-mn- rsrcm. -- amaze; 223 52.3 .c
L _ .- a _. :8. ...-x.- 8 .. _ N. -- cocoa-Eco: and .m
2.. 2. ...2.- :om.- no. ...-em. ...-...3. 5:855 .aoéoa .e
3.- ..2.- ar 8. ...3- co. 8.- -- 2.85 .m
3.- 3. ...2. _ T 8.- 8. 2. 8. -- ow< .N
tam. 2.- 2.- ...-cm. :.- 8.- so. ...3- :.- -- . am a
: 2 o a a c m m N _ 28m

 

SERGEQR 2.2%: Lo\§35.:§ Meets 3233x3225

m 55...?-

94

Appendix F

Intercorrelations among Variables for Black Participants

 

 

Variable 1 2 3 4 5 6 7
1. Sex 3 -- .04 .00 -.26 -.63*** .04 -.31
2. Age -- -.41** .02 .02 -.02 .00
3. Income -- -.08 -.08 -.32* .18
4. Political Orientation -- .09 .07 .27
5. Party Identiﬁcation -- .34* .23
6. Belief in White Privilege -- .12

7. Support for an AA Policy

 

Notes: 11 = 40.
a Dummy-coded (Female = 1, Male = 2).

*p < .05, ”p < .01, ***p < .001 (All 2-tailed tests).

95

References

Adarand Contractors, Inc. v. Pena, 515 US. 200 (1995).
Allport, G.W. (1954). The nature of prejudice. Reading, MA: Addison-Wesley.

American Psychological Association. (1995). Affirmative action: Who beneﬁts?
Retrieved October 3, 2003 from the World Wide Web:
http://www.apa.org/pubinfo/afﬁrmaction.html?CFID=245 1 377&CFTOKEN=609
94088

Asante, MK. (1991). Multiculturalism: An exchange. The American Scholar, 60, 267-
276.

Azzi, AB. (1992). Procedural justice and the allocation of power in intergroup relations:
Studies in the United States and South Africa. Personality and Social Psychology
Bulletin, 18, 736-747.

Barnes Nacoste, R. (1990). Sources of stigma: Analyzing the psychology of afﬁrmative
action. Law & Policy, 12, 175-195.

Barnes Nacoste, R. (1994). If empowerment is the goal. . .: Afﬁrmative action and social
interaction. Basic and Applied Social Psychology, 15(1 & 2), 87-112.

Baron, R.M., & Kenny, DA. (1986). The moderator-mediator variable distinction in
social psychological research: Conceptual, strategic and statistical considerations.
Journal of Personality and Social Psychology,51, 1 173-1 182.

Bennett, W.L. (1993). Constructing publics and their opinions. Political Communication,
10, 101-120.

Bobo, L. (1998). Race, interests and beliefs about afﬁrmative action: Unanswered
questions and new directions. American Behavioral Scientist, 41, 985-1003.

Bobocel, D.R., Son Hing, L.S., Davey, L.M., Stanley, D.J., & Zanna, MP. (1998).
J ustice-based opposition to social policies: Is it genuine? Journal of Personality
and Social Psychology, 75, 653-669.

Bonilla-Silva, E., & Forman, TA. (2000). “I am not a racist but. . .”: Mapping White
college students’ racial ideology in the USA. Discourse & Society, 11, 50-85.

Bowen, W.G., & Bok, D. (1998). The shape of the river: Long-term consequences of

considering race in college and university admissions. Princeton, NJ: Princeton
University Press.

96

Braithwaite, A., Doyle, S., & Lightbown, N. (1979). The balance between fairness and
discrimination. European Journal of Social Psychology, 9, 149-163.

Brewer, MB. (1979). In-group bias in the minimal intergroup situation: A cognitive-
motivational analysis. Psychological Bulletin, 86, 307-324.

Brewer, MB. (1991). The social self: On being the same and different at the same time.
Personality and Social Psychology Bulletin, 17, 475-482.

Brewer, MB. (1993). Social identity, distinctiveness, and in-group homogeneity. Social
Cognition, 11, 150-164.

Brewer, MB. (1995). Managing diversity: The role of social identities. In S.E. Jackson &
M.N. Ruderman (Eds), Diversity in work teams: Research paradigms for a
changing work place (pp. 47-68). Washington, DC: American Psychological
Association.

Brewer, M.B., & Harasty, AS. (1996). Seeing groups as entities: The role of perceiver
motivation. In R.M. Sorrentino & E.T. Higgins (Eds), Handbook of motivation
and cognition, .' The interpersonal context (Vol. 3, pp. 347-370). New York:
Guilford.

Brewer, M.B., & Miller, N. (1984). Beyond the contact hypothesis: Theoretical
perspectives on desegregation. In N. Miller & M.B. Brewer (Eds), Groups in
contact: The psychology of desegregation (pp. 281- 302). Orlando, FL: Academic.

Brickman, P., Folger, R., Goode, E., & Schul, Y. (1981). Microjustice and macrojustice.
In M.J. Lerner & S.C. Lerner (Eds), The justice motive in social behavior:
Adapting to times of scarcity and change (pp. 173-208). New York: Plenum
Press.

Brown, RH. (2002). Overcoming educational exclusion: Is diversity an appropriate
model for democratic higher education? American Behavioral Scientist, 45, 1061 -
1087.

Brown, R.J., & Turner, J .C. (1981). Interpersonal and intergroup behaviour. In J .C.
Turner & H. Giles (Eds), Intergroup behaviour, pp. 33-65. Oxford, England:
Basil Blackwell.

Byme, D., & McGraw, C. (1964). Interpersonal attraction toward Negroes. Human
Relations, 1 7, 201-213.

Carey, J .W. (1975). A cultural approach to communication. Communication, 2, 1-22.

97

Chang, WC. (1996). Toward equal opportunities: Fairness, values, and afﬁrmative action
programs in the US. Journal of Social Issues, 52, 39-97.

Chyi, H.I., & McCombs, M. (2004). Media salience and the process of framing:
Coverage of the Columbine school shootings. Journalism & Mass
Communication Quarterly, 81, 22-35.

Clawson, R.A., Strine, H.C., IV, & Waltenburg, EN. (2003). Framing Supreme Court
decisions: The mainstream versus the black press. Journal of Black Studies, 33,
784-800.

Clawson, R.A., & Waltenburg, EN. (2003). Support for a Supreme Court afﬁrmative
action decision: A story in black and white. American Politics Research, 31, 251-
279.

Clayton, S.D., & Tangri, SS. (1989). The justice of afﬁrmative action. In F.A. Blanchard
& F.J. Crosby (Eds), Affirmative action in perspective (pp. 177-192). New York:
Springer-Verlag.

Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159.

Cronbach, L]. (1957). The two disciplines of scientiﬁc psychology. American
Psychologist, 12, 671-684.

Crosby, F.J. (1994). Understanding afﬁrmative action. Basic and Applied Social
Psychology, 15, 13-41.

Crosby, F .J ., & Cordova, D1. (1996). Words worth of wisdom: Toward an understanding
of afﬁrmative action. Journal of Social Issues, 52, 33-49.

Crosby, F.J., Iyer, A., Clayton, S., & Downing, RA. (2003). Afﬁrmative action:
Psychological data and the policy debates. American Psychologist, 58, 93-115.

D’Angelo, P. (2002). News framing as a multiparadigmatic research program: A response
to Entman. Journal of Communication, 52, 870-888.

deVreese, CH. (2004). The effects of frames in political television news on issue
interpretation and frame salience. Journalism & Mass Communication Quarterly,
81, 36-52.

Domke, D. (2001). Racial cues and political ideology: An examination of associative
priming. Communication Research, 28, 772-801.

Domke, D., McCoy, K., & Torres, M. (1999). News media, racial perceptions, and
political cognition. Communication Research, 26, 570-607.

98

Dovidio, J .F., Gaertner, S.L., Isen, A.M., & Lowrance, R. (1995). Group representations
and intergroup bias: Positive affect, similarity and group size. Personality and
Social Psychology Bulletin, 21, 856-865.

Druckman, J .N. (2001). The implications of framing effects for citizen competence.
Political Behavior, 23, 225-256.

Eberhardt, J .L., & Fiske, S.T. (1994). Afﬁrmative action in theory and practice: Issues of
power, ambiguity, and gender versus race. Basic and Applied Social Psychology,
15, 201-220.

Edelman, M. (1993). Contestable categories and public opinion. Political
Communication, 10, 231-242.

Edelman, L.B., Fuller, S.R., & Mara-Drita, I. (2001). Diversity rhetoric and the
managerialization of law. American Journal of Sociology, 106, 1589-1641.

Entman, RM. (1993). Framing: Toward clariﬁcation of a fractured paradigm. Journal of
Communication, 43 (4), 51-58.

Entman, RM. (1997). Manufacturing discord: Media in the afﬁrmative action debate.
Harvard International Journal of Press/Politics 2(4), 39-59.

Entman, R.M., & Rojecki, A. (2000). The black image in the white mind: Media and race
in America. Chicago: University of Chicago Press.

Enzenberger, HM. (2003). Constituents of a theory of the media. In N Wardrip-Fruin &
N. Montfort (Eds), The new media reader, pp. 261-275. Cambridge, MA: The
MIT Press. (Reprinted from New Left Review, 64, 13-36, Nov/Dec 1970).

Esses, V.M., & Seligman, C. (1996). The individual-group distinction in assessments of
strategies to reduce prejudice and discrimination. In R.M. Sorrentino & E.T.
Higgins (Eds), Handbook of motivation and cognition: The interpersonal context
(Vol. 3, pp. 570-590). New York: Guilford.

Fancher, MP. (2003, July 17). Afﬁrmative action as a tool of imperialist expansion and
aggression. The Black Commentator, 50. Retrieved September 4, 2003 from the
World Wide Web:
http://www.blackcommentator.com/50/50_afﬁrmative_action_pf.html

Fine, TS. (1992). The impact of issue framing on public opinion: Toward afﬁrmative
action programs. The Social Science Journal, 29, 323-334.

F iske, S.T. (2000). Stereotyping, prejudice and discrimination at the seam between the

centuries: Evolution, culture, mind and brain. European Journal of Social
Psychology, 30, 299-322.

99

Foster, S. (1993). Difference and equality: A critical assessment of the concept of
“diversity.” Wisconsin Law Review, 105-161.

Gaertner, S.L., Dovidio, J .F., Anastasio, P.A., Bachman, B.A., & Rust, MC. (1993). The

common ingroup identity model: Recategorization and the reduction of intergroup
bias. In W. Stroebe & M. Hewstone (Eds), European review of social

psychology, (V 01. 4, pp. 1- 26).

Gaertner, S.L., Mann, J ., Murrell, A., & Dovidio, J .F. (1989). Reducing intergroup bias:
The beneﬁts of recategorization. Journal of Personality and Social Psychology,
5 7, 239-249.

Gaertner, S.L., Rust, M.C., Dovidio, J .F., Bachman, B.A., & Anastasio, PA. (1994). The
contact hypothesis: The role of a common ingroup identity on reducing intergroup
bias. Small Group Research, 25, 224-249. New York: John Wiley & Sons.

Gamson, WA. (1992). Talking politics. New York: Cambridge University Press.
Gamson, WA. (1999). Ambivalence about afﬁrmative action. Society, 9, 41-45.

Gamson, W.A., Croteau, D., Haynes, W., & Sasson, T. (1992). Media images and the
social construction of reality. Annual Review of Sociology, 18, 373-393.

Gamson, W.A., & Modigliani, A. (1987). The changing culture of afﬁrmative action. In
R.G. Braungart & M.M. Braungart (Eds), Research in political sociology
(Volume 3, pp. 137-177. Greenwich, CT: JAI Press.

Gamson, W.A., & Modigliani, A. (1989). Media discourse and public opinion on nuclear
power: A constructionist approach. American Journal of Sociology, 95, 1-3 7.

Gandy, O.H., Jr. (1998). Communication and race: A structural perspective. New York:
Oxford University Press.

Gandy, O.H., Jr., & Baron, J. (1998). Inequality: It’s all in the way you look at it.
Communication Research, 25, 505-527.

Gandy, O.H., Jr., Kopp, K., Hands, T., Frazer, K., & Phillips, D. (1997). Race and risk:
Factors affecting the framing of stories about inequality, discrimination, and just
plain bad luck. Public Opinion Quarterly, 61, 158-182.

Garcia-Prieto, P., Bellard, E., & Schneider, SC. (2003). Experiencing diversity, conﬂict,

and emotions in teams. Applied Psychology: An International Review, 52, 413-
440.

100

Goffman, E. (1974). Frame analysis: An essay on the organization of experience.
Cambridge, MA: Harvard University Press.

Golden, H., Hinkle, S., & Crosby, F. (2001). Reactions to afﬁrmative action: Substance
and semantics. Journal of Applied Social Psychology, 31, 73-88.

Gratz v. Bollinger, 123 S. Ct. 2411 (2003).
Gray, H. (1987). Race relations as news. American Behavioral Scientist, 30, 381-396.

Green, SB. (1991) How many subjects does it take to do a regression analysis?
Multivariate Behavioral Research, 26, 499-510.

Grutter v. Bollinger, 123 S. Ct. 2325 (2003).

Gurin, P., Dey, E.L., Hurtado, S., & Gurin, G. (2002). Diversity and higher education:
Theory and impact on educational outcomes. Harvard Educational Review, 72,
330-366.

Hall, S. (1980). Encoding/decoding. In S. Hall, A. Lowe & P. Willis (Eds), C ulture.
media, language (pp. 128-139). London: Hutchinson.

Hall, S. (2003). The whites of their eyes: Racist ideologies and the media. In G. Dines &
J .M. Humez (Eds), Gender, race and class in media (2“‘1 ed., pp. 89-93).
Thousand Oaks, CA: Sage.

Hollinger, DA. (1998). Group preferences, cultural diversity, and social democracy:
Notes toward a theory of afﬁrmative action. In R. Post & M. Rogin (Eds), Race
and representation: Affirmative action, pp. 97-109. New York: Zone Books.

Hopwood v. Texas, 95 F.3d 53 (5th Cir. 1996).

Huo, Y.J., Smith, H.J., Tyler, T.R., & Lind, EA. (1996). Superordinate identiﬁcation,
subgroup identiﬁcation, and justice concerns: 18 separatism the problem; is
assimilation the answer? Psychological Science, 7, 40-45.

Hunter, J .E., & Gerbing, D.W. (1982). Unidimensional measurement, second order factor
analysis, and causal models. In B.M. Straw & L.L. Cummings (Eds), Research in
organizational behavior (V 01. 4 ) (pp. 267-320). Greenwich, CT: JAI Press.

' Iyengar, S. (1989). How citizens think about national issues: A matter of responsibility.
American Journal of Political Science, 33, 878-900.

Iyengar, S. (1990). Framing responsibility for political issues: The case of poverty.
Political Behavior, 12, 1 9-40.

101

Iyer, A., Leach, C.W., & Crosby, F.J. (2003). White guilt and racial compensation: The
beneﬁts and limits of self-focus. Personality and Social Psychology Bulletin, 29,
117-129.

Jacobson, CK. (1985). Resistance to afﬁrmative action: Self-interest or racism? Journal
of Conﬂict Resolution, 29, 306-329.

Jackman, M.R., & Muha, M.J. (1984). Education and intergroup attitudes: Moral
enlightenment, superﬁcial democratic commitment, or ideological reﬁnement.
American Sociological Review, 49, 751-769.

James, E.H., Brief, A.P., Dietz, J ., & Cohen, RR. (2001). Prejudice matters:
Understanding the reactions of Whites to afﬁrmative action programs targeted to
beneﬁt Blacks. Journal of Applied Psychology, 86, 1120-1128.

Jemmott, J .B. 111, & Tebbets, R. (1980). Applying social cognition: A content analysis of
the Bakke case. Personality and Social Psychology Bulletin, 6, 30-36.

Jones, J ..M. (1998). Psychological knowledge and the new American dilemna of race.
Journal of Social Issues, 54, 641-662.

Kahnemann, D., & Tversky, A. (1984). Choices, values and frames. American
Psychologist, 39, 34 l -3 50.

Katz, I. & Hass, R.G. (1988). Racial ambivalence and American value conﬂict:
Correlational and priming studies of dual cognitive structures. Journal of

Personality and Social Psychology, 55, 893-905.

Katz, 1., Wackenhut, J ., & Hass, R.G. (1986). Racial ambivalence, value duality and
behavior. In Prejudice, Discrimination and Racism (pp. 35-59).

Kelly, B, & Dobbin, F. (1998). How afﬁrmative action became diversity management:
Employer response to anti-discrimination law: 1961 to 1996. American
Behavioral Scientist, 41, 960-984.

Keppel, G. (1991). Design and analysis: A researcher ’s handbook (3rd Ed.) Englewood
Cliffs, NJ: Prentice-Hall.

Kinder, DR. (1986). The continuing American dilemna: White resistance to racial
change 40 years after Myrdal. Journal of Social Issues, 42, 151-171.

Kinder, DR. (1998). Communication and opinion. Annual Review of Political Science, I,
167-197.

102

Kinder, D.R., & Sanders, L.M. (1990). Mimicking political debate with survey questions:

The case of white opinion on afﬁrmative action for blacks. Social Cognition, 8,
73-103.

Kluegel, J .R. (1985). If there isn’t a problem, you don’t need a solution: The bases of
contemporary afﬁrmative action attitudes. American Behavioral Scientist, 28,
761-784.

Kluegel, J .R. & Smith, ER. (1983). Affrrrnative action attitudes: Effects of self-interest,
racial affect and stratiﬁcation beliefs on Whites’ views. Social Forces, 61, 797-
824.

Kramer, R.M., & Brewer, MB. (1984). Effects of group identity on resource use in a
simulated commons dilemma. Journal of Personality and Social Psychology, 46,
1044-1057.

Kraiger, K., & Ford, J .K. (1985). A meta-analysis of rate race effects in performance
ratings. Journal of Applied Psychology, 70, 56-65.

Kravitz, DA. (1995). Attitudes toward afﬁrmative action plans directed at blacks: Effects
of plan and individual differences. Journal of Applied Social Psychology, 25 ,
2192-2220.

Kravitz, D.A., & Platania, J. (1993). Attitudes and beliefs about afﬁrmative action:
Effects of target and of respondent sex and ethnicity. Journal of Applied
Psychology, 78, 928-938.

Krysan, M. (2000). Prejudice, politics, and public opinion: Understanding the sources of
racial policy attitudes. Annual Review of Sociology, 26, 135-168.

Kuklinski, J .H., Sniderman, P-M., Knight, K., Piazza, T., Tetlock, P.E., Lawrence, G.R.,
& Mellers, B. (1997). Racial prejudice and attitudes toward afﬁrmative action.
American Journal of Political Science, 41 , 402-419.

Lawrence, C .R., III. (2001). Two views of the river: A critique of the liberal defense of
afﬁrmative action. Columbia Law Review, 101, 928-975.

Lind, B.A., Kulik, C.T., Ambrose, M., & Park, M.V.D. (1993). Individual and corporate
dispute resolution: Using procedural fairness as a decision heuristic.

Administrative Science Quarterly, 38, 224-251.

Lind, B.A., & Tyler, TR. (1988). The social psychology of procedural justice. New
York: Plenum Press.

Lippmann, W. (1922). Public opinion. New York: Harcourt, Brace.

103

Liu, G. (1998). Afﬁrmative action in higher education: The diversity rationale and the
compelling interest test. Harvard Civil Rights-Civil Liberties Law Review. 33,
381-442.

Major, B., Feinstein, J ., & Crocker, J. (1994). Attributional ambiguity of afﬁrmative
action. Basic and Applied Social Psychology, 15(1 &2), 1 13-141.

McConahay, J .B. (1981). Has racism declined in America? It depends on who is asking
and what is being asked. Journal of Conﬂict Resolution, 25 , 563-579.

McConahay, J .B. (1986). Modern racism, ambivalence and the Modern Racism Scale. In
J .F . Dovidio & S.L. Gaertner (Eds), Prejudice, discrimination and racism, pp.
91-125. Orlando, FL: Academic Press.

McIntosh, P. (1989, July/August). White privilege: Unpacking the invisible knapsack.
Peace and Freedom, 10-12.

Merelman, RM. (1994). Racial conﬂict and cultural politics in the United States. Journal
of Politics, 56, 1-20.

Morley, DD. (1988). Meta-analytic techniques: When generalizing to message
populations is not possible. Human Communication Research, 15, 112-126.

Moses, Y.T. (1999). Race, higher education and American society. Journal of
Anthropolical Research, 55, 265-277.

Mullen, B., Brown, R., & Smith, C. (1992). Intergroup bias as a function of salience,
relevance and status: An integration. European Journal of Social Psychology, 22,
103-122.

Murrell, A.J., Dietz-Uhler, B.L., Dovidio, J .F., Gaertner, S.L., & Drout, C. (1994).
Aversive racism and resistance to afﬁrmative action: Perceptions of justice are not

necessarily color blind. Basic and Applied Social Psychology, 15, 71-86.

Nacoste, R.W. (1987). Social psychology and afﬁrmative action: The importance of
process in policy analysis. Journal of Social Issues, 43, 127-132.

Nelson, T.E., & Kinder, DR. (1996). Issue frames and group-centrism in American
public opinion. The Journal of Politics, 58, 1055-1078.

Nelson, T.E., & Oxley, Z.M. (1999). Issue framing effects on belief importance and
opinion. The Journal of Politics, 61 , 1040-67.

Nelson, T.E., Oxley, Z.M., & Clawson, RA. (1997). Toward a psychology of framing
effects. Political Behavior, 19, 221-246.

104

Nosworthy, G.J., Lea, J .A., & Lindsay, R.C.L. (1995). Opposition to afﬁrmative action:
Racial affect and traditional value predictors across four programs. Journal of
Applied Social Psychology, 25, 314-337.

Oakes, P.J., & Turner, J .C. (1990). Is limited information processing capacity the cause
of social stereotyping? In W. Strobe & M. Hewstone (Eds), European review of
social psychology (Volume 1, pp.11 1-135). New York: John Wiley & Sons.

Omi, MA. (2001). The changing meaning of race. In N.J. Smelser, W.J. Wilson, & F.
Mitchell (Eds), America becoming: Racial trends and their consequences, pp.
243-263. Washington, DC: National Academy Press.

Omi, M.A., & Winant, H. (1994). Racial formation in the United States: From the 19608
to the I990s. New York: Routledge.

Opotow, S. (1997). What’s fair? Justice issues in the afﬁrmative action debate. American
Behavioral Scientist, 41 , 23 2-245.

Palmer, SR. (2001). A policy framework for reconceptualizing the legal debate
concerning afﬁrmative action in higher education. In G. Orﬁeld (Ed.), Diversity
challenged: Evidence on the impact of aﬁirmative action (pp. 49-80). Cambridge,
MA: Harvard Education Publishing Group.

Peterson, RS. (1994). The role of values in predicting fairness judgments and support of
afﬁrmative action. Journal of Social Issues, 50, 95-115.

Price, V., Tewksbury, D., & Powers, E. (1997). Switching trains of thought: The impact
of news frames on readers’ cognitive responses. Communication Research, 24,
481-506.

Quinn, K.A., Ross, E.M., & Esses, V.M. (2001). Attributions of responsibility and
reactions to afﬁrmative action: Afﬁrmative action as help. Personality and Social
Psychology Bulletin, 27, 321-331.

Regents of the Univ. of Cal. v. Bake, 438 US. 265 (1978).

Richardson, J .D., & Lancendorfer, KM. (2004, May). Newspaper editorial responses to
the University of Michigan aﬂirmative action decisions. Paper presented at the
annual meeting of the International Communication Association, New Orleans,

LA.

Schudson, M. (2002). The news media as political institutions. Annual Review of
Political Science, 5, 249-269.

Schuman, H., Steeh, C., Bobo, L., & Krysan, M. (1997). Racial attitudes in America.
Cambridge, MA, Harvard University Press.

105

Sears, D.O., Van Laar, C., Carillo, M., & Kosterman, R. (1997). Is it really racism? The
origins of White Americans’ opposition to race-targeted policies. Public Opinion
Quarterly, 61 , 16-53.

Shelton, J .N. (2000). A reconceptualization of how we study issues of racial prejudice.
Personality and Social Psychology Review, 4, 374-390.

Shen, F. (2004). Chronic accessibility and individual cognitions: Examining the effects of
message frames in political advertisements. Journal of Communication, 54(1),
123-137.

Simon, B. (1993). On the asymmetry in the cognitive construal of ingroup and outgroup:
A model of egocentric social categorization. European Journal of Social
Psychology, 23, 131-147.

Sniderman, P.M., & Tetlock, PE. (1986). Symbolic racism: Problems of motive
attribution in political analysis. Journal of Social Issues, 42, 129-150.

Sniderman, P.M., & Piazza, T. (1996). The scar of race. In J. Arthur & A. Shapiro (Eds),
Color, class, identity: The new politics of race, pp. 44-64. Boulder, CO: Westview
Press.

Steeh, C., & Krysan, M. (1996). Afﬁrmative action and the public, 1970-1995. Public
Opinion Quarterly, 60, 128-158.

Steele, S. (1990). White guilt. The American Scholar, 59, 497-506.

Summers, R.J. (1995). Attitudes towards different methods of afﬁrmative action. Journal
of Applied Social Psychology, 25 , 1090-1 104.

Swain, CM. (2001). Afﬁrmative action: Legislative history, judicial interpretations,
public consensus. In N.J. Smelser, W.J. Wilson, & F. Mitchell (Eds), America
becoming: Racial trends and their consequences, pp. 318-347. Washington, DC:
National Academy Press.

Swidler, A. (1986). Culture in action: Symbols and strategies. American Sociological
Review, 51, 273-286.

Swim, J .K., & Miller, D. L. (1999). White guilt: Its antecedents and consequences for
attitudes toward afﬁrmative action. Personality and Social Psychology Bulletin,

25, 500-514.

Tajfel, H., Billig, M.G., Bundy, R.P., & Flament, C. (1971). Social categorization and
intergroup behavior. European Journal of Social Psychology, 1(2), 149-178.

106

Taj fel, H., & Turner, J. (1979). An integrative theory of intergroup conﬂict. In W.G.
Austin & S. Worchel, The social psychology of intergroup relations (pp. 33-47).
Monterey, CA: Brooks/Cole.

Tatum, ED. (1992). Talking about race, learning about racism: The application of racial

identity development theory into the classroom. Harvard Educational Review, 62,
1-24.

Terenzini, P.T., Cabrera, A.F., Colbeck, C.L., Bjorklund, S.A., & Parente, J .M. (2001).
Racial and economic diversity in the classroom: Does it promote student
learning? The Journal of Higher Education, 72, 509-531.

Tewksbury, D., Jones, J ., Peske, M.W., Raymond, A., & Vig, W. (2000). The interaction
of news and advocate frames: Manipulating audience perceptions of a local public
policy issue. Journalism & Mass Communication Quarterly, 77, 804-829.

The slow and tortured death of afﬁrmative action. (2003, July 3). The Black
Commentator, 49. Retrieved September 4, 2003 from the World Wide Web:
http://www.blackcommentator.com/49/49_cover_pr.html

Thibault, J .W., & Walker, L. (1978). A theory of procedure. California Law Review, 66,
541-566.

Tougas, F ., & Veilleux, F. (1989). Who likes afﬁrmative action: Attitudinal processes
among men and women. In F.A. Blanchard & F.J. Crosby (Eds), Affirmative
action in perspective (pp. 111-124). New York: Springer-Verlag.

Tuch, S.A., & Hughes, M. (1996). Whites’ racial policy attitudes. Social Science
Quarterly, 77, 723-745.

Turner, J.C. (1982). Towards a cognitive redeﬁnition of the social group. In H. Tajfel
(Ed.), Social identity and intergroup relations, pp. 15-40. New York: Cambridge
University Press.

Turner, J.C. (1985). Social categorization and the self-concept: A social cognitive theory
of group behavior. In E.J. Lawler (Ed.), Advances in group processes (Vol. 2, pp.
77-122). Greenwich, CT: JAI Press.

Tyler, TR. (1990). Why people obey the law. New Haven, CT: Yale University Press.

Tyler, T.R., & McGraw, KM. (1986). Ideology and the interpretation of personal
experience: Procedural justice and political quiescence. Journal of Social Issues,

42, 115-128.

Tyler, T.R., Rasinski, K.A., & Grifﬁn, E. (1986). Alternative images of the citizen:
Implications for public policy. American Psychologist, 41 , 970-978.

107

Van Dijk, TA. (1991). Racism and the press. New York: Routledge.
Williams, D.R., Jackson, J .S., Brown, T.N., Forman, T.A., & Brown, K. (1999).
Traditional and contemporary prejudice and urban whites’ support for afﬁrmative

action and government help. Social Problems, 46, 503-516.

Winkelman, C.S., & Crosby, F .J . (1994). Afﬁrmative action: Setting the record straight.
Social Justice Research, 7, 309-328.

Winant, H. (2000). Race and race theory. Annual Review of Sociology, 26, 169-185.

Wood, J. (1994). Is “symbolic” racism really racism? A review informed by intergroup
behavior. Political Psychology, 15, 673-686.

108

MICHIGANS STATE UNIVERSITVL I:

111113111111111111311111 11111 11113111 1111111