Comunication and Democracy - McCombs
Comunication and Democracy - McCombs
Edited by
Maxwell McCombs
University of Texas at Austin
Donald L. Shaw
University of North Carolina at Chapel Hill
David Weaver
Indiana University
Routledge
Taylor &. Francis Group
NEW YORK AND LONDON
First Published by
Lawrence Erlbaum Associates, Inc., Publishers
10 Industrial Avenue
Mahwah, New Jersey 07430
Publisher’s Note
The publisher has gone to great lengths to ensure the quality of this reprint
but points out that some imperfections in the original may be apparent.
Contents
v
vi_________________________________________________________ CONTENTS
References 231
Contributors 255
emphasis was on the agenda of issues found in the news media and among the
general public. Although a few investigators probed other kinds of agendas and
agenda items—such as Weaver, Graber, McCombs, and Eyal’s (1981) investi-
gations of the salience of political interest and candidate images on personal
agendas during the 1976 U.S. presidential campaign—the vast majority fol-
lowed the traditions of public opinion research and tracked issues.
In the abstract, most of these studies, whether they examined issues, political
interest, the slate of contenders during the presidential primaries, or whatever,
focused on an agenda of objects. Explicitly stating the focus of the theory in
this way, an agenda of objects, immediately suggests a second level of attention.
Objects have attributes. When the news media report on public issues, political
candidates, the presidential campaign, or any other object, they describe that
object. In these descriptions some attributes are very prominent and frequently
mentioned, some are given passing notice, and others are omitted. In short, news
reports also define an agenda of attributes that vary considerably in their
salience. Similarly, when people talk about and think about these objects—pub-
lic issues, presidential candidates, or whatever—the attributes ascribed to these
objects also vary considerably in their salience. These agendas of attributes are
the second level of agenda setting. The core theoretical idea is the same for
agendas of attributes as it is for agendas of objects: The salience of elements,
objects or attributes, on the media agenda influences the salience of those
elements on the public agenda (McCombs, 1992; McCombs & Evatt, 1995).
By extension, of course, we can talk about the transfer of salience from any
agenda to another, the president of the United States to the news media, the New
York Times to other newspapers, and so forth.
At the first level of agenda setting are agendas of objects. This is the
traditional domain of agenda-setting research, represented by an accumulation
of hundreds of studies over the past quarter-century. At the second level of
agenda setting are agendas of attributes. This second level is one of the new
theoretical frontiers whose aspects are discussed in detail in the opening
chapters of Communication and Democracy. Other chapters offer maps of yet
other theoretical frontiers. Among the frontiers mapped by these chapters are
political advertising agendas and their impact on behavior, the framing of
various agendas in the mass media and the differential impact of print and TV,
the theoretical role of individual differences in the agenda-setting influence of
the news media on the public agenda, methodological advances for determining
cause and effect roles in agenda setting, and the application of agenda-setting
theory to historical analysis.
There is an overview in Fig. P.l of these frontier territories that are mapped
by the chapters of Communication and Democracy. At the top of the diagram
are the familiar elements of the agenda-setting idea, the media agenda, and the
public agenda. This diagram has been expanded to call explicit attention to the
two levels of agenda setting, the influence of object salience in the news media
on object salience among the public (Level 1) and the influence of attribute
salience in the news media on attribute salience among the public (Level 2).
PROLOGUE: "THE GAME IS AFOOT1 xi
OBJECT
Newspaper vs. TV
Individual differences
International comparisons
Agendas and democracy
Establishing causality
Historical analysis
ATTRIBUTES ATTRIBUTES
Centered under the first level are summary labels for a variety of theoretical
maps presented in the book. Of course, these research expeditions also can be
pursued at the second level of agenda setting. Listed first under the second level
is a summary label for the opening chapters of Communication and Democracy.
Also at the second level are additional theoretical topics listed under either the
media agenda or the public agenda. It is a rich array of topics.
Drawing further on the Sherlock Holmes tradition, Arthur Conan Doyle
described life at 22IB Baker Street as consisting of periods of intense excite-
ment when a case was afoot that were interspersed among lengthy periods of
intellectual torpor, punctuated by pistol practice on the living room wall or other
efforts at arousal. This may be overdone as an analogue of life in social science
research, but in the intellectual life of agenda-setting research most certainly
there have been periods of intense excitement over the exploration of new ideas
interspersed among more quiescent, conservative periods of careful extension
and explication. At times, scholars have been explorers like Lewis and Clark,
whose journals sketched the outlines of a vast new territory. At other times, the
role of scholars has been more like those anonymous surveyors who painstak-
ingly mapped the routes for the transcontinental railroads. We are now entering
a new period of creative exploration in agenda-setting research. The chapters
of this book have been prepared by early explorers for the benefit of other
scholars who would join them in mapping new theoretical territories over the
next few years.
AN INTERNATIONAL EXPEDITION
The editors of this book—Maxwell McCombs, Donald Shaw, and David
Weaver—represent the established generation of agenda-setting scholars.
McCombs and Shaw, of course, carried out the original Chapel Hill study.
xii PROLOGUE: "THE GAME IS AFOOT"
Weaver joined in the very next project, a panel study of the 1972 U.S.
presidential election reported in The Emergence o f American Political Issues
(Shaw & McCombs, 1977) and was the lead author of Media Agenda-Setting
in a Presidential Election (Weaver et al., 1981), a year-long panel study of the
1976 U.S. presidential election. All three have continued to contribute to
agenda-setting research, but a new generation of young scholars also has made
important contributions to the evolution of agenda-setting theory. So in the fall
of 1994, the three editors invited 14 young scholars to prepare theoretical essays
outlining the priority areas that should be pursued in agenda-setting research in
the coming years. For purposes of this invitation, “young scholars” were
defined as persons who had finished graduate school in the last 5 or 6 years and
who had published agenda-setting research.
Nearly all of these persons have some direct personal link with one of the
three editors, most commonly as a graduate student who studied with one of
the editors for his or her PhD, but sometimes the role is faculty colleague or
research collaborator. This network of scholars constitutes a loosely linked
invisible college, too loosely linked in the view of the editors because with few
exceptions the links are to a single senior editor without direct connections
among the young scholars themselves. The international scatter of this col-
lege—from Europe across the United States to Asia—is an inadequate expla-
nation for the dispersion of this college in a time of e-mail, fax, and frequent
international conferences.
In view of this situation as agenda-setting theory moves into a new era, the
goals of this book are threefold:
More than two decades have passed since the original agenda-setting hypothesis
was stated by McCombs and Shaw (1972), and scholars have published more
than 200 articles weaving agenda-setting research into a rich theory (Rogers,
Dearing, & Bregman, 1993). Over these years this research has detailed the
patterns in the transfer of issue salience from the media to the public, the
contingent conditions for agenda setting, and influences on the media agenda.
The underlying assumption for all three areas is that what is covered in the
media affects what the public thinks about.
Agenda setting is now detailing a second level of effects that examines how
media coverage affects both what the public thinks about and how the public
thinks about it. This second level of agenda setting deals with the specific
attributes of a topic and how this agenda of attributes also influences public
opinion (McCombs & Evatt, 1995).
In the abstract, every agenda consists of a set of objects. In turn, each of these
objects possesses a set of attributes. For example, the agenda of issues examined
in the original Chapel Hill study—and in numerous subsequent studies—is, in
the abstract, a set of objects. In contrast, Benton and Frazier’s (1976) examina-
tion of one object on the issue agenda, the economy, probed two sets of
attributes: the specific problems, causes, and proposed solutions associated with
this general issue; and the pro and con rationales for economic policies. Agenda
setting is about more than issue or object salience.
This shift in emphasis does not negate the basic agenda-setting hypothesis,
but rather builds on what already exists. It is one highway linking up with another
3
4 GHANEM
major thoroughfare. The first level of agenda setting deals with the transfer of
object salience from the media to the public agenda, whereas the second level
of agenda setting involves two major hypotheses about attribute salience:
1. The way an issue or other object is covered in the media (the attributes
emphasized in the news) affects the way the public thinks about that
object.
2. The way an issue or other object is covered in the media (the attributes
emphasized in the news) affects the salience of that object on the public
agenda.
Figure 1.1 shows the difference between the two levels of agenda setting and
also illustrates the two hypotheses about second-level agenda-setting effects.
For both the first and second levels of agenda setting, the independent and
dependent variables are the same: There is a media agenda and a public agenda.
The theoretical difference between the two levels is in the details of the way
that the variables are conceptualized and operationalized.
For first-level agenda setting, the independent variable is considered in terms
of objects, the topics or issues discussed on the media agenda. For the second
level, the media agenda (the same independent variable as at the first level) is
considered in terms of attributes or perspectives. The dependent variable for
both levels of agenda setting still remains the public agenda. However, in the
case of the first level, the public agenda is operationalized in terms of issue or
topic salience, whereas at the second level the salience of the attributes of the
issue or topic are measured.
These two levels of agenda-setting effects are represented by the two
horizontal arrows in Fig. 1.1. The top arrow illustrates the influence of issue or
topic salience. This is the original agenda-setting hypothesis. The bottom arrow
illlustrates the influence of attribute salience on how people think about an
object on the agenda. This is the initial second-level agenda-setting hypothesis
stated earlier. The diagonal arrow represents the second hypothesis stated
earlier, the influence of particular attributes or perspectives in news coverage
on the salience of an issue or topic on the public agenda. Some of the attributes
of an object presented in the media can have striking influence on the salience
of that object on the public agenda. This influence of attributes or frames from
the second level of the media agenda on the salience of objects on the first level
of the public agenda defines “compelling arguments” in the media message
(McCombs, 1996).
Second Level
FIG. 1.1. Two levels of agenda setting and three hypothesized effects.
1. FILLING IN THE TAPESTRY 5
The agenda of objects and the agenda of attributes can be looked at as two
concentric circles with the agenda of issues being the outer circle and the agenda
of attributes imbedded within that circle. Kosicki (1993) referred to agenda
setting as the “shell of the topic.” The shell of the topic can be compared to the
issues or other objects examined, whereas the attributes are an exploration of
what is inside the shell. Noelle-Neumann and Mathes (1987) suggested that
media content can be examined at three levels: agenda setting, focusing, and
evaluation. Agenda setting deals with the importance of issues and problems;
focusing deals with the definition of issues; and problems and evaluation deals
with the creation of a climate of opinion. We can easily replace their term agenda
setting with the first level of agenda setting and their second and third levels by
the second level of agenda setting.
The link between the two levels of agenda setting takes us back to
Lippmann’s (1922) idea of pictures in our heads. By examining the attributes
of an issue, we get a more detailed examination of that picture. The examination
of attributes is similar to the examination of an issue or other object under a
magnifying lens, whereas the agenda of issues deals with examination by the
naked eye. The attributes of an object are the set of perspectives or frames that
journalists and the public employ to think about each object. How news frames
impact the public agenda is the emerging second level of agenda setting.
A recent study on the coverage of crime in the news media (Ghanem, 1996)
examined both levels of agenda setting and introduced the concept of compel-
ling arguments, which is represented by Hypothesis 2 earlier and by the
diagonal arrow in Fig. 1.1. Replicating first-level agenda-setting effects, sali-
ence of crime on the media agenda correlated positively (.73) with salience of
crime on the public agenda over a period of several years. Supporting Hypothe-
sis 2 about compelling arguments, 13 attributes of this coverage that were
examined also correlated positively with the salience of crime on the public
agenda. In this particular study, the salience of crime on the public agenda was
not affected by actual crime rates, which had been decreasing during the years
examined. Hypothesis 1 about second-level effects was not tested in this study.
Several studies did explore this second level of agenda setting even before
explicit theorizing had begun. During the 1976 presidential election, Weaver,
Graber, McCombs, and Eyal (1981) conducted a nine-wave panel study that
looked at the agenda of attributes in the descriptions of presidential candidates
in the news and the agenda of attributes in voters’ descriptions of the candidates.
A strong cross-lagged correlation was found between the media agenda and the
public agenda. Becker and McCombs (1978) looked at the 1976 presidential
primaries and found considerable correspondence between the agenda of
attributes in Newsweek and the agenda of attributes in New York Democrats’
descriptions of the contenders.
Outside the setting of presidential elections, Benton and Frazier (1976)
examined the issue of the economy at three levels of information holding. Level
1 included general issue names, the economy in this case; Level 2 consisted of
subissues, including problems, causes, and solutions; and Level 3 contained
6 GHANEM
more specific information about the subissues. Levels 2 and 3 are similar to the
focusing level mentioned by Noelle-Neumann and Mathes (1987). Once again
significant correspondence between the media and the public agenda was
found. Iyengar and Simon’s (1993) research on the coverage of the Persian Gulf
crisis provides another example that illustrates the difference between the two
levels. When respondents state that the Gulf crisis is the most important problem
facing the nation, we are dealing with the first level. When respondents describe
the crisis in terms of military or diplomatic options, we are dealing with the
second level.
Many researchers, including Iyengar and Simon (1993), have distinguished
between the first and second levels of agenda setting by labeling the attributes
frames. No discussion of the second level of agenda setting is complete without
an extensive examination of media frames. The principal difference between
the research literature on frames and on the second level of agenda setting is
that the latter examines the impact of news frames on the public agenda,
whereas many framing studies have focused solely on the frames themselves.
DEFINITIONS OF FRAMES
exclusion and elaboration” (p. 3). A story angle or story line “which transforms
an occurrence into a news event, and that, in turn, into a news report, is a frame”
(Mendelsohn, 1993, p. 150). The essence of news judgment, according to Tiffen
(1989), are story frames or angles. Framing is defined in cognitive psychology
“as the function that specifies the relations that hold among the arguments
comprising a particular conceptual bundle at a particular level of abstraction”
(Friedman, 1979, p. 321).
How options are framed affects decision making. Studies have shown that
the wording of examples leads people to choose differently (Machina, 1990).
If a situation is presented to a person in terms of losses, the decision is very
different than if it is presented to that person in terms of gains (Elster, 1990).
The perception of problems and evaluation of probabilities produce predictable
shifts of preference depending on how the problem is framed (Tversky &
Kahneman, 1981, 1990).
Social scientists have long discussed the importance of word choice in
survey questionnaires. For example, Fine (1992) conducted a study on the
impact of issue framing on public opinion toward affirmative action programs.
He concluded that the way questions are framed affects how people perceive
issues. The same logic can be applied to the coverage of issues in the media:
Depending on how an issue is presented or framed in the media, the public will
think about that issue in a particular way. Media coverage of issues in the media
has been linked to public opinion where the salience of issues in the media
agenda leads to the salience of issues in the public agenda. How the media
covers an issue, argue researchers, also can have a cognitive influence on how
the public thinks about the issue. This transference of the salience of attributes
is the core of the second level of agenda setting.
international stories and observed that some types of international story frames
have a stronger agenda-setting impact on the public than other frames. Patterson
(1993a) also made the distinction between agenda setting and framing and said
that every news story has a theme or a frame that functions as the central
organizing idea, but that the topic is the summary label of the domain of social
experiences covered by a story.
Another way to look at the relationship between agenda setting and framing
is that traditional agenda setting posits that the media tell us what to think about,
and framing deals with the issue of the media telling us how to think about an
issue. Framing “shares with agenda-setting research a focus on the public policy
issues in the news and in voters’ minds. However, it expands beyond what
people talk or think about by examining how they think and talk ” (Patterson,
1993b, p. 70).
Basically, the first level of agenda setting deals with the selection of issues
by the news media and its impact on the public agenda. Framing deals with the
selection of elements within a particular issue. The second level of agenda
setting deals with the influence of the particular elements of an issue on the
public’s agenda of attributes.
Frames are not necessarily positions for or against some policy measures
(Gamson, 1988), but rather the criteria by which the policy measures might be
evaluated. One could argue that the selection process or adopted frames should
not be labeled bias because of the pejorative nature of the term. According to
Hackett (1984), bias is not a useful construct because language is not neutral
and cannot be removed from values or viewpoints.
Indexing. The indexing hypothesis states that the media tend to index
the range of viewpoints expressed in government debate (Bennett, 1990). Both
framing and indexing deal with the parameters of an issue or of a debate. The
difference between the two is that indexing links the range of options available
in the media to official sources (Althaus, Edy, Entman, & Phalen, 1995),
whereas the range of meaning in framing is not confined by a particular source,
but is left wide open.
Because of this distortion, people who watch more television are more likely
to view the world as mean and are more like to perceive themselves as potential
victims of violence than those who watch less television. Cultivation studies
have focused on the entertainment media. Framing, on the other hand, deals
with the idea that the news media may also be presenting a worldview construed
in a particular way that does not necessarily mesh with reality. Because of the
framing that takes place in the news, the question becomes: What are the effects
of these frames on the public? One way to conceptualize framing is to look at
it as an example of cultivation theory that deals with the news media rather than
the entertainment media.
also reflect McCombs’ (1992) assertion that news messages are both cognitive
and affective.
These four dimensions (see Fig. 1.2) can help in the development of frames
for analyzing a variety of issues and topics. One of the weaknesses of most
framing studies is that the attributes of the issue or topic are not generalizable
across issues, so even though the specific subdimensions are not generalizable,
these four larger dimensions could be the basis of comparisons across many
different agendas.
Subtopics
Attributes, which are the independent variable at the second level of agenda
setting, are the subtopics within a particular issue. Brosius and Eps (1994)
examined the coverage of attacks against foreigners and asylum seekers in
Germany and came up with subtopics such as trials, foreigners, assaults and
political action. Examining the agenda-setting process for AIDS, Rogers,
Dearing, and Chang (1991) identified 13 distinct subissues ranging from
government policy to ethics. Takeshita and Mikami (1995) divided their
agenda-setting study into general-issue salience and subissue salience. They
found evidence of second-level agenda-setting effects for both television and
newspapers.
In terms of theory building, the weakness of studies that just focus on
subtopics is that the researcher usually comes up with a list of topic-specific
frames based on his or her perceptions and then proceeds to content analyze the
material on hand. The problem then becomes the lack of distinction between
content analysis in general and the examination of frames. Coming up with a
list of frames for each issue is similar to— and shares the weaknesses of—the
dictionary approach of computerized content analysis where separate diction-
aries are developed for specific discourse (Krippendorff, 1980).
Framing Mechanisms
The second thing that needs to be considered when examining media frames
deals with the emphasis given to topics in the media, such as placement and
size as well as other elements that influence the prominence of a news item.
Second Level
1. Subtopics
2. Framing Mechanisms
3. Affective Elements
4. Cognitive Elements
FIG. 1.2. Describing agenda attributes: four dimensions of framing.
12 GHANEM
Photographs, pull quotes, subheads, and so on, all serve to give a story in a
newspaper more prominence. This aspect of salience needs to be examined
when we are looking at the relationship between the salience of items on the
media agenda and the salience of those items on the public agenda. Tankard et
al. (1991) referred to these focal points of news presentations and labeled them
“framing mechanisms” (p. 15).
Agenda-setting studies have focused on how frequently an issue is mentioned
in the media. The frequency with which a topic is mentioned probably has a more
powerful influence than any particular framing mechanism, but framing mecha-
nisms could serve as catalysts to frequency in terms of agenda setting.
Affective Dimension
Bearing in mind that the news communicates much more than facts (McCombs,
1992; Patterson, 1993b), this dimension examines the affective aspect of the
news. The affective dimension deals with the public’s emotional response that
may result from media coverage. One of the ways that the media exerts this
affective response is through the narrative structure of the news. Koch (1990)
went as far as to equate framing with the narrative itself. The way a news story
is structured focuses and thus limits the causes and outcomes of the issue
(Schulman, 1990). Schudson (1982) argued that the power of the media lies in
the forms in which declarations appear. The narrative is the link among the
components of who, what, where, why, how, and when (Bennett & Edelman,
1985) that form the content of the message. Here we are looking at the journalist
as story teller (Barkin, 1984).
Researchers distinguish between two forms of narrative, the chronicle and the
story. The chronicle relies more on an inverted pyramid style that provides a
record that something noteworthy has happened. The story, on the other hand,
engages the reader because of its style. Donohew (1983), for example, suggested
that narrative or chronologically ordered stories tend to produce significantly
greater physiological arousal for readers than the traditional summary style.
Two other elements that may cause concern among the public are based on
news values. Proximity and human interest are possible news values that might
make a reader or a viewer identify more closely with what is being reported.
Such news values are a cultural as well as an institutional product (Hall, 1981).
Price and Tewksbury (1995) argued that news values help determine which
angles to take in writing the news and which details need to be emphasized.
According to Elliott (1988), in a study of the hijacking of an airline, political
causes of the hijackers are too complex to cover and thus the media focus on
human elements of the story by focusing on friends and families of hostages.
Not only are the political causes too complex, but the human element can easily
become the focus of a story with a beginning, middle, and end. The human
element in much of the coverage is akin to the tendency of the media to
personalize stories in the press: “Happenings (or fantasies) involving individual
1. FILLING IN THE TAPESTRY 13
persons stand a high chance of becoming news stories” (Fowler, 1991, p. 91).
According to Hall (1981), the personalization of the news isolates the person
from relevant social and institutional context. Bringing a story to such a
personal level might help the reader identify with the happenings in the story
and thus feel more concern for what is going on.
Cognitive Dimension
This dimension deals with general cognitive categories that might shed light on
whether the media and the audience are thinking about the problem in the same
way. Cognitive categories could move us from topical categories by identifying
meaning in topics regardless of what the topic is. Edelstein, Ito, and Kepplinger
(1989) came up with the problematic situation, which could address the issue
of generalizable frames. Although they did not use the term framing, their
concept is very much applicable to framing.
Edelstein (1993) argued that the problematic situation sheds light on mean-
ing equivalence between the media and the audience. For example, if the media
present news about a condition of conflict and the audience also perceives
conflict, then both the media and the audience are in accord. Edelstein et al.
(1989) conducted a comparative cultural study in the United States, Germany,
Japan, and Hong Kong. Using the problematic situation as a conceptual tool for
their comparison, they focused on two major elements of their theory, which
are the problem and the steps taken to address the problem.
Hendrickson (1995) also used general categories and examined the coverage of
child maltreatment using an ecological framework. This framework consists of five
dimensions: the individual, the microsystem, the mesosystem, the exosystem, and
the macrosystem. The similarity between the ecological framework and the
problematic situation is the distinction the researchers make on whether the
problem is identified from the individual or the social perspective. The ecological
framework is more detailed than the problematic situation.
Iyengar (1991) used the general categories of thematic and episodic frames
to examine news coverage. The episodic news frame focuses on specific events
or cases, whereas the thematic frame places the issue in some general context.
Similar to the thematic and episodic frames are Yagade and Dozier’s (1990)
division of coverage into abstract or concrete categories. They equated concrete
issues with visual and easy to understand issues.
Focusing on causes and solutions, Klandermans and Sidney (1988) exam-
ined social movements and argued that a social movement has an ideology that
contains a diagnosis (causes and agents responsible), a prognosis (what must
be done), and a rationale (who must do the job and arguments that action needs
to be taken). According to Rucinski (1992), the social construction of reality
deals with attributions of causes and solutions. Maher (1995) looked at the
causes of a pollution problem in Austin, Texas. He examined the causes
mentioned in the local newspaper and found a perfect correspondence with the
perception of the causes held by the public.
14 GHANEM
CONCLUSION