Claims, Frames, and Global Warming
The Life Course of an Environmental Issue:
Claims, Frames, and Global Warming.
By Craig Trumbo
Graduate Student
Department of Agricultural Journalism
440 Henry Mall
The University of Wisconsin
Madison WI 53706
A paper presented to the Science Interest Group
at the Annual Convention of
the Association for Education in Journalism and Mass Communication
Washington, DC
August 1995
ABSTRACT
A content analysis of a decade of coverage of global warming in five
national newspapers is presented. The empirical analysis is drawn from
a
constructionist perspective on the content of news discourse
emphasizing
claims-making and framing. The issue is also discussed in terms of
Downs' issue-attention cycle. The issue's life in the news is modeled as
exhibiting three phases that are related to the sources quoted and
the
frames presented in highest level syntactic structures.
ACKNOWLEDGMENTS
The author would like to graciously acknowledge the guidance provided on
this project by Professor Sharon Dunwoody of the School of
Journalism at
the University of Wisconsin.
INTRODUCTION: NEW KINDS OF PROBLEMS.
Of the great variety of environmental issues that have achieved social
prominence in recent times one stands out as perhaps a prime
example of
a new class of environmental problems. Global warming represents a
type
of environmental problem that is generally identified with the idea
of
global change. In fact, change is at the very root of the issue of
global warming. But perhaps a more important characteristic that issues
such as global warming, ozone depletion, loss of biodiversity, and
others share is their intangibility for the common person. We now must
acknowledge at least the potential existence of environmental
problems
that are practically invisible yet at the same time constitute
threat on
a global scale.
Solutions to such problems will defy the efforts of the physical and
biological sciences alone because the social dimension of global
change
is inescapable. An old adage holds that the first step in solving a
problem is recognizing that a problem exists D and in this adage lies
the essence of the social aspect of issues like global warming. How
society comes to recognize and define something as a problem is no
trivial question. This research is concerned with a narrowly defined
aspect of that question (news media representation) with regard to one
environmental problem (global warming).
An important aspect of global warming as a news topic is that it has
clearly and dramatically demonstrated a kind of cyclic life course
that
may be common to the treatment of this variety of long-term issue in
the
news. Previous research on the volume of news attention given to
global
warming has shown how the issue rose from virtual obscurity, became
a
competitively pursued story, and eventually fell from prominence D
nearly disappearing altogether.[1] Recent work examining public
understanding of the issue has shown that the public is generally
misinformed about global warming.[2] Together, these observations beg for
a
closer examination of the content of the news coverage of the issue.
This investigation, in an effort to understand the life course of this
environmental issue, will add to the growing literature that has
examined global warming. Using a constructivist application of the ideas
of claims-making and framing, an empirical evaluation of the content
of
major newspapers across the span of a decade will show how
significant
changes in the nature of the content of the news relate to the
definition of global warming as a problem.
Before providing a description of global warming's career in the news,
the theoretical foundation for this investigation will be
outlined.
SOCIAL PROBLEMS: CONSTRUCTED, CLAIMED, AND FRAMED.
Constructivism. Dunlap and others have shown in considerable detail how
public concern for the environment arose in the late 1960s and
has
persisted since.[3] What role did the news media play in this
phenomenon?
Mauss notes that "the growth of public concern about the
environment, as
reflected in attitude and opinion surveys, follows rather closely
the
increased attention and coverage given these issues by the media."
[4] Over
the past 20 years a wide range of studies have examined the nature
of
the mass media's coverage of the environment and the variety of
social
impacts of that coverage.[5]
Mauss notes that many studies, considered along with opinion polls
conducted over the years, show that concern about and attention to
the
environment was steadily on the rise even as actual pollution
problems
were declining in many areas. He concludes that "this finding can be
explained, at least in part, by the attention created by mass media
coverage and emphasis on pollution." [6] Munton and Bradley echo this
sentiment, observing that "article after article, book after book, and
commentator after commentator have informed the public about
environmental pollution and told them that they should be worried." [7]
The relationship between the media and the environment is a complex one
that can be fruitfully seen in the light of social problems
theory. In
recent years the social problems literature has generally embraced a
constructivist viewpoint. Constructivists argue that human perception
of
the world comes about through a process of "meaning-making," which
is
accomplished through the exchange of a variety of symbols: thus
social
reality is a constructed thing that is not defined in an absolute
sense
by the existence of any true reality.[8]
Fortunately, the constructivist viewpoint exists across a range of
abstraction. In a more moderate form, this tradition emphasizes that
evidence of social reality is present in the process of collective
definition that is embedded in various forms of social exchange, most
especially communication. This viewpoint offers some moderation in
the
constructivist argument, an argument that when taken to its opposite
extreme can present a world that is virtually unknowable.[9]
Claims-making. The 1977 work by Spector and Kitsuse that brought a new
focus to the constuctivist social problems literature was built
upon the
tradition of Mills, Gouldner, and most especially Blumer.[10] Recent
entries
within this tradition that address the environment include the work
of
Hilgartner and Bosk as well as that of Gamson and Modigliani.[11]
Spector
and Kitsuse focus on the process of the constructed definition of a
social problem, stating that this process is grounded in
"claims-making"
activities: "the activities of individuals or groups making
assertions
of grievances and claims with respect to some putative conditions.
The
emergence of a social problem is contingent upon the organization of
activities asserting the need for eradicating, ameliorating, or
otherwise changing some condition." [12]
The idea of claims-making is the conceptual component of their model
that makes it, like Blumer's, a model of reality constructed
through a
process of symbolic exchange. They point out that "claims-making is
always a form of interaction: a demand made by one party to another
that
something be done." [13] These claims-making activities may take a
variety
of forms: writing government representatives, petitioning, protest,
resolutions made by professional or other organizations, filing
lawsuits, garnering media attention or simply filling out complaint
forms. "All of those who involve themselves in these activities
participate in the process of defining social problems." [14] Participants
may take any form: individuals, groups, crusaders, officials, news
persons, professionals, or government agencies.
Spector and Kitsuse point out that much of the social activity
surrounding the recognition of a problem goes on within and between
social agencies such as the government, protest groups or professional
organizations. While certainly not media-centric, their model can
still
be seen to confer power to the media. This is most critical to the
legitimization stage in their model. Once a claim is recognized as
legitimate and worthy of action it will also become recognized as having
the characteristics of news.
Probably the most fruitful manner of applying Spector and Kitsuse's
model to the question of the media's role in the definition of
social
problems is to use it as a basis for recognizing the media as a
social
clearinghouse for claims. The media serve as a conduit for
communication
between social agencies and as a way for those agencies to bring
pressure to bear as they champion their claim. Claims that become news
are those that have entered one very important arena in the struggle
for
legitimacy.
Framing. The metaphor of the "frame" has been spread far and wide,
crossing disciplines to the degree that no summary definition is
possible. In a call for communications researchers to strive toward a
clarification of framing, Entman observes:
Despite its omnipresence across the social sciences and humanities,
nowhere is there a general statement of framing theory that shows
exactly how frames become embedded within and make themselves
manifest
in a text, or how framing influences thinking.[15]
In terms of looking at media content, the commonly cited roots of
framing extend to Goffman's 1974 dramaturgical perspective that
frames
are "schemata of interpretation" that people use to "locate,
perceive,
identify, and label," and subsequently to Tuchman's 1978 derivation
that
assigns frames the role of an organizing device that allows the
journalist to more efficiently net, sort, and transmit information.[16]
More recent formulations include the 1980 work of Gitlin, who writes
that media frames "are principles of selection, emphasis, and
presentation composed of little tacit theories about what exists, what
happens, and what matters."[17] Gamson and Modigliani consider
frames as
being embedded within "media packages" that can be seen to "give
meaning
to an issue. A package has an internal structure. At its core is a
central organizing idea, or frame, for making sense of relevant events,
suggesting what is at issue." [18] They identify five signifiers of
frames:
metaphors, exemplars, catchphrases, depiction, and visual images.
And
according to Dunwoody, when framing is applied to the content of
messages it is "a schema or heuristic, a knowledge structure that is
activated by some stimulus and is then employed by a journalist
throughout story construction."[19]
Two recent perspectives on framing offer significant guidance toward a
clearer conceptual definition and a useful operationalization.
Entman
offers that:
Framing essentially involves selection and salience. To frame is to
select some aspects of a 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 item described.
Frames, then, define problems D determine what a causal agent is
doing with what costs and benefits, usually measured in terms of
common
cultural values; diagnose causes D identify the forces creating
the
problem; make moral judgments D evaluate causal agents and their
effects; and suggest remedies D offer and justify treatments for the
problem and predict their likely effects. [original emphasis] [20]
As Entman brings some clarity to the concept, Pan and Kosicki provide a
basis on which to observe frames by outlining four categories of
framing
devices that may be located in news discourse.[21] Syntactical
Structures
are general organizing schemes that most obviously manifest
themselves
as the inverted pyramid. Script Structures have "distinct structure
defined by the rules that may be called story grammars. A generic
version consists of the five Ws and one H in news writing." Thematic
Structures more commonly occur in issue-related stories as opposed to
event stories. These are causal statements that can be seen as
hypothesis tests having the form of a statement and then its logical
support based on traditional modes of journalistic evidence (e.g.,
quotes, attribution). Rhetorical Devices refer back to Gamson's framing
devices: metaphors, exemplars, catchphrases, depiction, and visual
images.
There is some common ground between the work of Entman and the work of
Pan and Kosicki. These ideas will be returned to in the
operational
definitions used in this study. But first, an overview of global
warming's life in the news will be offered as background.
GLOBAL WARMING: CHARACTERIZING THE LIFE OF THE ISSUE.
At the heart of global warming is the proposition that human activities
are altering the composition of the planet's atmosphere to a
degree
sufficient to affect the natural processes that play fundamental roles
in shaping global climate. Many, perhaps even most, scientists agree
that the release of gasses such as carbon dioxide, CFCs, and methane
will have the consequence of raising the average temperature on Earth.
A
considerable amount of contention exists over issues such as when
this
might happen, how quickly it might come about, and the degree and
nature
of the consequences.
How can global warming's history in the news be summed? Figure 1 shows
the amount of news coverage given to global warming in five
major
newspapers over the past decade (details of Figure 1 are addressed
below). Perhaps it could be concluded that the issue of global warming
has simply enjoyed its day in the sun. Puns aside, a serious problem
presents itself if one considers that during the span of this decade
there was little substantive change in the science that should warrant
diminishing concern. It is easy to understand the spike of attention
associated with Dr. James Hansen's Congressional testimony (during the
drought summer of 1988) that global warming had manifested itself.
But
the overall build-up and eventual decline of news coverage presents
a
more complex problem. The existence of a cycle of attention is clear.
But in a broader sense it is important to ask what social forces
might
drive such a cycle and how these forces might exert themselves
through
the news.
Transient attention to specific issues may be typical of American public
opinion, policy, and media coverage. Downs offers his
"issue-attention
cycle" as an explanation for such coming and goings of news coverage
and
public concern. In this theory he suggests that there are typically
five
stages to the life of a given issue.[22]
1. Pre-problem. "This prevails when some highly undesirable social
condition exists but has not yet captured much public attention, even
though some experts or interest groups may already be alarmed by
it."
2. Alarmed discovery, euphoric enthusiasm. "As a result of some dramatic series
of events, the public suddenly becomes both aware of and alarmed
about
the evils of a particular problem." This is combined with a
reaction of
overconfidence, "euphoric enthusiasm," in society's ability to
discover
a solution.
3. Realizing the cost. "A gradually spreading realization that the cost of
solving the problem is very high indeed." The public and
policymakers
also realize that the problem is being caused by a condition
which is
providing benefits to society.
4. Gradual decline of interest. Three reactions occur. Some people become
discouraged. Some suppress attention out of fear. Others simply
get
bored. Often, all three reactions operate to varying degrees.
Meanwhile,
another issue is on the rise and attention shifts.
5. Post-problem. "A prolonged limbo D a twilight realm of lesser attention
or spasmodic recurrence of interest."[23]
Looking at the issue of global warming in terms of Downs'
issue-attention cycle is a useful way to present a brief history of the
issue, and will present a useful tool for approaching the problems
and
results of this investigation. The time prior to 1988, when global
warming was primarily the concern of scientists and top policy-makers,
can easily be characterized as the pre-problem stage. During this
time
there was considerable scientific activity, extending back to the
1750s
in fact. The policy attention that the early science produced caused
some mild controversy that served to earn the issue prominent display
in
the news on a couple of occasions. But public awareness of the issue
remained low in the absence of any sustained media attention.
An important aspect of the pre-problem stage involves the preparation of
the issue for its alarmed discovery. For global warming, much of
this
preparation was in the form of the generally rising level of
environmental concern in society[24] and the linkage of global warming to
the related atmospheric problem of ozone depletion.[25] This linkage
gave
global warming added legitimacy and plausibility.
This preparation is also political. Between 1985 and 1988 a number of
influential Congresspersons adopted climate change as an
important
concern. The scientists who were becoming increasingly concerned about
global warming therefore had excellent access to an important public
arena as conditions became favorable for the alarmed discovery of
warming.
It takes little imagination to see the alarmed discovery heralded by
Hansen's mid-drought Congressional testimony, set against the
backdrop
of the Yellowstone fires. It is interesting to speculate on how the
issue might have behaved if there had not been a circumstantial heat
wave that summer and if Yellowstone had not become so engulfed.
But Downs' second stage has an inherent dualism as it is also
characterized by a euphoric optimism over solutions. This contrast was
abundantly clear in the headlines of late 1988: Calculating the
consequences of a warmer planet earth; Major greenhouse impact is
unavoidable, experts say; Scientists dream up bold remedies for ailing
atmosphere; Fighting the greenhouse effect. But perhaps nothing
captured
this dualism better than the contrast between Hansen's testimony and
President Bush's pledge to counter the greenhouse effect with "the
White
House effect."[26]
The third stage, a realization of the true cost, gradually replaced this
optimism and alarm. This change came about primarily through
actions in
the political sphere. The science bashing carried out by then Chief
of
Staff John Sununu served to promote the idea that solving the
problem of
global warming would bear an enormous price tag, even though many
experts disagreed. It was a fear of the economics of preventing climate
change that motivated Bush to non-action and prompted Sununu to
order
that Hansen's written testimony be watered down.
But it was more than just politics. By 1992 the true complexity of the
problem was becoming evident as nations of the world began
contemplating
a treaty to slow the release of greenhouse gases. Downs points out
that
it is during this part of the cycle that society becomes aware that
the
problem at hand is related to things that are held dear, things that
provide benefit.
A May 25, 1992, New York Times business page article tells how "the
price of driving a car has never been lower" because of the
plummeting
cost of oil.[27] It goes on to say that, world-wide, automobiles are
reproducing faster than people and this poses dire consequences for
global warming. Then of course, there's the complex interactions of
population and economics. Another Times article, on the same day's
front
page, reveals how China's "contribution to global warming may be
rising
more quickly than that of any other country."[28] This is due to the
fact
that one fifth of earth's population is entering a period of economic
growth fueled by coal. Once again, the consequences are related to
global warming.
On a more fundamental level what is becoming increasingly evident is
global warming's utter complexity, both scientific and social.
Writing
from the Earth Summit in Rio, The New York Times' William K. Stevens
contrasts a 1972 conference on the environment with the 1992 Earth
Summit:
In those palmy save-the-whales years, full of hope and idealism, the
delegates to the United Nations Environment Conference in Sweden
asserted confidently that "the capability of man to improve the
environment increases with each passing day."
Here, the optimism of 1972 has been replaced by a hard realism. The
delegates in Rio have discovered how hard it is for nations to
unite on
fundamental environmental problems facing them.[29]
The article goes on to emphasize the difficulties associated with
simultaneously negotiating the reduction of both deforestation in
developing nations and industrial emissions in developed ones. Two
aspects of the same problem, in more ways than one. The article's
headline summarizes that the "Earth Summit finds the years of optimism
are a fading memory."
It's unlikely that the stages of Downs cycle operate independently or in
any strict linear sense. There should be considerable overlap
between
the grim realizations of phase three and the declining attention of
phase four. There should also be other demands being made on the
public's attention. The public arenas and ecologies of news perspectives
tells us that there are only so many issues that can be supported at
a
given time and that those issues must compete with each other in a
sort
of zero-sum game.[30] As the difficult nature of global warming
became news
D and the volume of media attention began to decline D the nation
was
also sliding into increasingly difficult economic conditions, was
captivated by Operation Desert Storm, and began anticipating the most
unusual presidential election in recent memory. As Downs suggests,
there
were new issues to attend to.
And what of the fifth stage in the cycle? Did the issue of global
warming enter "a prolonged limbo D a twilight realm of lesser
attention?" News coverage of global warming had a brief comeback in
mid-1992 thanks to the Earth Summit. However, the amount of media
attention to global warming during the first six months of 1993 is
similar in volume to that of the first half of 1988.[31] Clinton's
announcement of his "Climate Change Action Plan" in late 1993 received
cursory coverage and no follow-up. Coverage in 1994 was scant. And
while
outside of the bounds of this analysis, another recurrence of
attention
occurred in early 1995 as Antarctic pack ice went to sea while the
Earth
Summit treaty was being reviewed in Berlin.
But Downs holds that an issue in the fifth stage of the cycle is not a
simple return to its earlier state. In the wake of its rise and
fall,
global warming has entered the popular lexicon D even being featured
in
television and film drama D and has created significant
international
agreements. Downs points out that such factors "almost always persist
and often have some impact even after public attention has shifted
elsewhere."[32] So may be the case with global warming.
While examining the volume of media attention and the nature of the news
story can inform many questions about the life course of this
issue,
much of the knowledge to be gained about global warming's history as a
socially defined problem lies embedded in the content of the news.
STATEMENT OF THE PROBLEM AND OPERATIONAL CONCEPTS.
Because of their presumed influence, the media become, to quote
Gurevitch and Levy (1985), "a site on which various social groups,
institutions, and ideologies struggle over the definition and
construction of social reality." The media, in this view, provide a
series of arenas in which symbolic contests are carried out among
competing sponsors of meaning.
Participants in symbolic contests read their success or failure by how
well their preferred meanings and interpretations are doing in
various
media arenas. Prominence in these arenas is taken as an outcome
measure
in its own right, independent of evidence on the degree to which
the
messages are being read by the public. Essentially, sponsors of
different frames monitor media discourse to see how well it tells the
story they want told, and they measure their success or failure
accordingly.[33]
The ultimate rationale for this study is well captured by the above
passage from Gamson, Croteau, Hoynes, and Sasson. In this light it
is
taken that media content will capture important aspects of an
evolving
struggle to define a problem.
Two ideas require operationalization: claims-makers and frames. Both are
closely related and are being cast under the constructionist
rubric. The
idea of the claims-maker is being operationally defined as a
function of
the attributed source. While journalists bring a great deal more to
a
story than a collection of sources D things like background and
emphasis
D it is in the source that the broader authority of the story
resides.
Attribution is the first lesson in journalism.
But sources are used for a wide variety of reasons, including their past
history with both individual journalists and the media in
general,
prominence in their field, availability, and their ability to provide
useful material such as interesting quotes. Nonetheless, any party
wishing to place a claim in the media arena has a keen interest in
becoming a source or to be represented by a source. And as was pointed
out above, prominence in the news may be taken by the claims-maker
as
success in its own right. Perhaps the best indicator of that variety
of
success is the quote. While journalists quote for as many different
reasons as they choose sources, from the viewpoint of the claims-maker
nothing signifies successful access to the media arena quite as well
as
a direct quote. This study will define the claims-maker as the
quoted
source. The following section addressing measurement details how
claims-maker categories are developed and identified.
Framing is being operationalized in concert with claims-making. In fact,
for this study it will be held that the frame is the claim being
made by
the media and that this claim is manifest in the macro level meeting
of
Pan and Kosicki's syntactical and thematic structures. In other
words,
the frame is the claim presented by the media in the headline and
the
lead paragraph (lead is addressed below). Frames of this form are
held
to have the four functions specified by Entman: to define problems,
diagnose causes, make moral judgments, and to suggest remedies.
Locating the frame at the top of the inverted pyramid draws from
Entman's emphasis of salience in his definition of framing. Within
that
emphasis, salience "means making a piece of information more
noticeable,
meaningful, or memorable to audiences. An increase in salience
enhances
the probability that receivers will perceive the information,
discern
meaning, and thus process it, and store it in memory." [34]
Restricting the operationalization of the frame to only the headline and
lead places this analysis firmly on the macro level. Tradition
holds
that the headline and lead should be written to inform the reader as
to
what is most important about the story. While styles vary and there
are
exceptions to the inverted pyramid, journalists and their editors
are
aware of the fact they are competing for the reader's attention and
that
the top of the story is generally the point of the readers' first
contact with the information content of the story.
The specification of claims-makers and frames continues below in the
section on measurement. At this point the research questions being
addressed may be presented:
RQ1: Can Downs' issue-attention cycle serve as a model in which the
amount of coverage of this issue may be seen in terms of phases?
RQ2: How are frames and claims-makers distributed in media coverage of
global warming? Do these distributions change through time?
RQ3: Are there associations among the frames and claims-makers?
METHODS.
Sample. The newspaper story is the unit of analysis, based on a set of
newspapers chosen to represent national level media. Selecting a
set of
newspapers to represent the national media is always a somewhat
debatable matter. This study follows the lead of others[35] in selecting
The
New York Times, The Washington Post, The Los Angeles Times, The
Christian Science Monitor, and The Wall Street Journal.
Further support for this selection comes from a qualitative review which
seeks to define the world's 20 most important newspapers.
Merrill
states:
In the vast global wasteland of crass and mass journalistic mediocrity
is a small coterie of serious and thoughtful internationally
oriented
newspapers that offers a select group of readers an in-depth,
rational
alternative. . . . They are well-informed, articulate papers
that
thoughtful people the world over take seriously.[36]
Merrill cites the five papers above plus The Miami Herald as the best in
the United States (in no assigned order). Two other factors
support this
selection. Each of the five papers selected is generating its own
coverage of the issue at hand through the employment of its own science
writers. Therefore, each story selected from these newspapers is
original and unique. Many of these stories go on to live a second life
in other papers across the nation via the Associated Press and other
wire services. Finally, this set of five newspapers is represented in
a
single consistent reference index: the National Newspaper Index.
[37]
Only news stories are used in this study. News stories are defined as
content containing references to global warming or the greenhouse
effect, excluding editorials, opinion columns, letters to the editor and
advertisements. The selection of stories was done using a
computerized
version of the National Newspaper Index.
Across the period of the study approximately 500 items on global warming
appeared in these newspapers. Only about half that number was
determined
to be necessary to empower statistical analysis. However, subsequent
analytic needs involving comparison of distinct phases of coverage
necessitated over-sampling of the early and late phases of coverage
(determination of the Phases is discussed shortly). Therefore, a random
half of the stories that fell in Phase 2 were selected and all of
the
stories that fell in Phases 1 and 3 were selected. The final sample
yielded a total of 252 stories entered into the analysis.
Measurement: Claims-makers. As discussed above, claims-makers are being
operationalized as quoted sources. Examination of the full content
of
the Washington Post coverage revealed the following categories of
individuals quoted: university scientists, government scientists (NASA,
NOAA, etc.), other scientists (including foreign), Congresspersons,
Presidential administrations, foreign officials, environmental interest
groups, business and industry interest groups. All but 6, or 98% of
all
quoted sources, fell into these categories. For analysis, categories
were collapsed to scientists, politicians, and interest groups.
The number of quotes for each category were summed by story for an
interval level measure. Distributions were found to be highly
skewed, so
a nominal level measure was also computed as the presence or absence
of
each source category in a story.
Measurement: Frame. Frame is being conceptualized as the claim presented in
the highest syntactical structures, the headline and the lead. A
qualitative reading of all headlines and lead paragraphs in the
Washington Post revealed four prominent categories that agreed well
with Entman's four purposes of frames:
1. Define Problems: impacts of global warming. These stories deal with
what will happen as a consequence of this phenomenon. Impacts
may be
negative (coastal flooding), positive (improved regional agriculture),
or debated.
2. Diagnose Causes: evidence as to the realty of global warming as a
problem. These are typically presentations of scientific findings
that
support the idea that there is a problem (evidence of rising sea
level),
refute the idea that there is a problem (evidence that changes are
within limits of natural variance), or present the argument that the
nature of the problem is unknown.
3. Make moral judgments: action statements. These stories present
general statements calling for action or reporting action taken (U.S.
should sign a treaty, did sign a treaty), arguing against action or
reporting action blocked (emission standards not needed, scientific
testimony altered), or present the argument that a course of action is
not clear.
4. Suggest remedies: provide specific information about how solutions
should be implemented. These stories report specific solutions
that have
been proposed or implemented (tougher emission standards), solutions
that have been rejected or deemed inadequate (voluntary programs), or
present a debate about a specific solution or solutions. Note that
the
specificity of the solution D a statement of exactly how the
solution
should be carried out D is an important distinction between an action
statement and a solution statement.
All but 17, or 93% of all stories fell into one of these four
categories. Most of the stories could be classified by reading only the
headline. When headlines were ambiguous (often because they were too
short) the first paragraph was read. Most stories were classified by
this point. In a few cases it was necessary to read into the story by
an
additional paragraph or two (typically when the story begins with an
anecdote). The goal of the classification is to identify the most
immediately identifiable characteristic of the story with respect to the
four categories.[38]
DISCUSSION OF THE RESULTS.
Figure 1 presents the distribution of the sample through time and Figure
2 presents the distribution of stories through the 5 newspapers
(without
over-sampling). Inspection of this distribution supported the idea
that
the attention paid to this issue might be divided into distinct
phases
for analysis.
Using the ideas of Downs, three distinct phases were identified. The
overall distribution fits Downs' 5 stage model fairly well. Downs
proposes that attention to an issue will remain low until a dramatic
discovery brings a sudden increase in salience. While the
issue-attention cycle does not offer specific predictions about
changing
salience during the middle three stages it does suggest general
aspects
of the content of these stages and also suggests that salience
during
these stages should be at its highest before feathering into a
decline.
Finally, Downs directly predicts that the final phase will involve a
lowering of the salience of the issue, but not a lowering to the
levels
seen in the first stage.
An examination of the time distribution of stories clearly suggests 2
important points in the series: mid-1988 when Hansen testifies
before
Congress and mid-1992 when the Earth Summit concludes. A fifth order
polynomial was found to fit the data so that the important points in
the
issue fell near the curve's inflections. The curve clearly suggests
Downs' overall propositions as they would be applied to the volume of
media attention. Dividing the series into these three segments and
fitting linear functions to each segment shows that the means and the
slopes vary between the phases (analysis of variance significant at p
<
.001).
Stories were thus coded as being in Phase 1, 2, or 3. Because of the
content of the news, the three phases are being labeled as
pre-controversy, controversy, and post-controversy.
Regarding the relationship between frames and claims-makers, it is first
necessary to report the obvious: political and special
interests are
strongly associated with the judgment frame while scientists are
strongly associated with the causes frame. This is true across the full
span of the decade, as shown in Figure 3.
More interesting results are found by examining changes across phases in
prevalence of each claims-maker and each frame. Figure 4 shows
that
there was a significant decline in scientists as a percentage of all
claims-makers quoted across the decade.[39] Quotes of political and
special
interest claims-makers both increased slightly, but not
significantly.
Previous research on global warming by Miller and others has shown
that
scientific sources were being crowded out by political sources
during
the late 1980s.[40] This analysis of a full decade supports their
observation and shows that it is part of a longer-term trend.
It is an open question as to why scientists declined as quoted sources
so dramatically. Of course it must be recognized that story
space is a
finite resource. Although not statistically significant, special
interests made the second largest gain across the decade. Mazur and
others have noted that as the issue of global warming matured the cold
war as simultaneously coming to an end. This allowed a number of
scientifically-oriented special interest groups (for example the Union
of Concerned Scientists) to shift their attention from nuclear
weapons
issues to environmental issues such as global warming.
An equally interesting question looks at how the prevalence of frames
shifted as the issue evolved. Figure 5 shows that the framing of
the
issue moved away from defining problems and diagnosing causes and
toward
making judgments and suggesting remedies. These results can again be
seen as a situation in which a set of perspectives must compete for
finite space in a limited number of stories. The relationship that
exists between the results in Figure 3 and the results in Figure 4 can
be seen in terms of the impacts reported in Figure 5. As politicians
and
interest groups were increasingly successful in making their claims
(Figure 4) they brought along their associated frames (Figure 3) in a
process that influenced the make-up of the content of the news
(Figure
5).
These results beg the interesting question of the relative role of
claims-making and framing in the changing attention that the media
give
to the issue. How do claims-makers and frames compare in their
ability
to predict the amount of coverage given to the issue? To address
this
question a secondary dataset was extracted as a time series. The unit
of
analysis was set as 2 month periods (n = 60) and interval level
variables were created as: number of stories per unit, number of quotes
for each claims-maker category per unit, and number of stories for
each
frame in each unit. Each unit was assigned to one of the three
phases.
First, what relationships exist within the variables representing
claims-makers and frames? It must be noted that these variables are
highly correlated with one another because each is a product of the
number of stories in a given 2 month period (average correlation .38
with upper range of .89). Measures of association involving
claims-makers, frames, and the number of stories must be used with
caution because they contain a strong spurious element. However, it is
reasonable to see if the set of claims-maker and frame variables
form
interpretable factors that could describe their relationship and that
might be useful in further analysis.
Table 1 presents a factor analysis of the frame and claims-maker
variables that provides a satisfactory solution. The associations
found
in the crosstabulations are upheld here as it is shown that
political
and special interest claims-makers group together with their
associated
frames to form one factor while scientists and their associated
frames
group to form another. An alternative result would have found frames
grouping together and claims-makers grouping together. This strongly
suggests that the two concepts of frames and claims-makers are part of
a
larger single concept relating to the content of news discourse.
How well do these factors perform in predicting the most salient
characteristic of the issue D the dramatically changing amount of news
coverage? While, as noted above, the absolute strength of an
association
with the number of stories per unit of time involves a spurious
element,
it is reasonable to pit the two factors against one another in a
relative evaluation. Table 2 present the results of an analysis in which
the number of stories per unit is regressed on dependent variables
made
up of factor scores (the political-interest factor is simply being
called political for the remaining analysis). The political factor
proved to be only a slightly stronger predictor of the number of
stories, in fact there is not a significant difference in the two
equations.
Discriminant analysis was used to more thoroughly evaluate the
relationship between the two derived factors and to use them to judge
the validity of the three proposed phases. Table 3 reports the
results
of three discriminant analyses in which each factor was first
evaluated
by itself before the two were combined in a single analysis.
Taken individually, it is seen that the scientific factor does a
somewhat better job than the political factor in correctly classifying
cases into the three phases, presenting classification rates of 50%
and
45% respectively. Both rates are moderately successful as compared
to
the 33% rate that would be expected by chance. An examination of the
relative success rates for each phase (on the diagonal in each matrix)
shows that the political factor did an excellent job in correctly
classifying Phase 1 but did a poor job of distinguishing between Phases
2 and 3. On the other hand, the scientific factor did an equally
good
job of distinguishing between Phases 1 and 2 but a poor job
classifying
Phase 3. This can be explained by the changing prevalence of
claims-ma
kers and frames in each of the phases.
If both factors are entered into the analysis together they combine to
do a quite respectable job of correctly classifying the matrix,
hitting
the mark 63.3% of the time, a marked improvement over the 33% chance
rate. This clearly indicates that the two factors overlap considerably
in their relationship within the three phase model. Further, it is
again
seen that the classification rates on the diagonal decline across
the
phases with Phase 1 being perfectly classified while Phase 3 is less
well classified.
CONCLUSION.
This project does not hold as an express purpose the operationalization
of Downs' issue-attention cycle. But the model can be used as a
more
general basis for a division of the decade's media coverage of global
warming into three distinct phases. Inferring from Downs' model to
what
might be expected to be seen regarding the amount of media attention
to
an issue does allow the first research question to be affirmatively
resolved.
It must be emphasized that Downs' issue-attention cycle is a social
process model and is not specifically designed to evaluate news
media
attention to an issue. Nonetheless, elements of the issue-attention
cycle do seem to fit a qualitative reading of the news coverage of
global warming. This, combined with the good fit between the observed
quantity of news attention and the expectations of the Downs model,
suggests that it might be reasonable to interpret the three phases used
in this study as a partial expression of the issue-attention cycle.
The results of the second research question show that scientists become
less dominant sources as the issue matures. At the same time the
emphasis of the news coverage shifts away from a presentation of the
issue in terms of its causes and problematic nature and toward a
presentation more grounded in political debate and the proposal of
solutions. These observations seem to dovetail most closely with the
first three stages in the issue-attention cycle. The progression from
the pre-problem stage, to alarmed discovery, and then to a
realization
of the costs strongly suggests that there should be a politicization
of
the issue, an increase in its level of controversy, and a shift
toward
judgments and solutions. That progression was observed in this
study.
Overall, these results suggest that the most appropriate way to relate
Downs' model to the changes observed in media coverage of global
warming
is to argue that what has been observed across this decade is just
the
first three stages of the cycle. This would predict that the years
following 1994 should present a continued decline of media attention to
the issue punctuated only by a "spasmodic recurrence of interest,"
as
Downs' puts it. A casual observation of the issue during 1995
suggests
that this is in fact what is happening. A follow-up study may in
future
years provide evidence of this.
The results of the third research question apply less to the
issue-attention cycle and more to the theoretical basis of this
investigation as it relates to news media coverage of environmental
controversies. To answer the third research question: yes, strong
associations do exist between the claims-makers and frames observed in
this study. What might these associations tell us about journalistic
coverage of environmental controversy?
The results show that there is greater independence between phases and
frames than there is between phases and claims-makers. Thus,
changes
that occurred in the life course of the issue perhaps involved shifts
linked to who was getting their message into the media rather than
how
the media was choosing to present the information. It may be
reasonable,
at least in this case, to argue that a good deal of the journalistic
discretion that goes into shaping media coverage of an environmental
controversy occurs by way of deciding which sources to use and how
much
overall attention to give the issue. These decisions seem to hold
more
sway over the life of an issue compared to the decisions that allow
a
point of view D a frame D to dominate a story. In essence, this
supports
a model of transmission rather than processing: reporting over
interpretation.
The more alarming aspect of the results of this study is unfortunately
also the least surprising: that scientists left the debate as it
heated
up. In fact, scientists found themselves sharing a shrinking portion
of
a growing media pie during an important part of the public debate
over
global warming. Whether they were squeezed out by other sources or
chose
to become distanced from an increasingly political debate is an open
question.
[--- Pict Graphic Goes Here ---]
[--- Pict Graphic Goes Here ---]
Table 1.
Factor Analysis: Claim and Frame Variables
Factor 1 Factor 2
Variable Mean SD Political Scientific
Debate Proclamation
Claimsmaker: Special Interest Sources 2.4 3.5 .85 * .12
Frame: Remedies for Problem 0.5 1.1 .84 * -.33
Claimsmaker: Political Sources 3.2 4.8 .76 * .48
Frame: Moral Judgments about Problem 1.5 2.2 .70 * .55
Claimsmaker: Scientist Sources 5.5 5.4 .22 .87 *
Frame: Causes of Problem 1.3 1.3 .19 .86 *
Frame: Problem Definition 0.7 0.9 -.09 .53 *
Percentage of total variance explained 48.5 23.3
Principle components analysis with varimax rotation. Two factors explain 71.8
percent of total
variance, n = 60.
Table 2.
Regression Analysis: Comparison of Factor Scores.
Dependent variable = number of stories in 2 month period, n = 60.
Independent Variable
R2
Std. Beta
t value
p
Fact Score: Political
.50
.71
7.6
<.001
Fact Score: Scientific
.45
.67
6.9
<.001
difference between equations not significant
Table 3.
Discriminant Analysis: Comparison of Political and Scientific Factors.
For the case of a 3 group analysis, n = 60.
Ind. Variable: Factor Score Political
Predicted Phase
Actual Phase
N
Phase 1
Phase 2
Phase 3
Phase 1: Pre-controversy
20
100.0% (20)
0.0% (0)
0.0% (0)
Phase 2: Controversy
25
44.0% (11)
12.0% (3)
44.0% (11)
Phase 3: Post-controversy
15
46.7% (7)
26.7% (4)
26.7% (4)
Prior Probability = 33% Correct Classification = 45.0%
Canonical Correlation = .39 Wilks' Lambda = .85 p < .01
Ind. Variable: Factor Score Scientific
Predicted Phase
Actual Phase
N
Phase 1
Phase 2
Phase 3
Phase 1: Pre-controversy
20
55.0% (11)
5.0% (1)
40.0% (8)
Phase 2: Controversy
25
20.0% (5)
60.0% (15)
20.0% (5)
Phase 3: Post-controversy
15
46.7% (7)
26.7% (4)
26.7% (4)
Prior Probability = 33% Correct Classification = 50.0%
Canonical Correlation = .46 Wilks' Lambda = .79 p < .01
Ind. Variables: Scientific and Political
Predicted Phase
Actual Phase
N
Phase 1
Phase 2
Phase 3
Phase 1: Pre-controversy
20
95.0% (19)
5.0% (1)
0.0% (0)
Phase 2: Controversy
25
36.0% (9)
48.0% (12)
16.0% (4)
Phase 3: Post-controversy
15
33.3% (5)
20.0% (3)
46.7% (7)
Prior Probability = 33% Correct Classification = 63.3%
Scientific: Canonical Correlation = .27 Wilks' Lambda = .66 p < .001
Political: Canonical Correlation = .54 Wilks' Lambda = .79 p < .01
NOTES.
[1] Ungar, S (1992). The rise and (relative) decline of global wa
rming as a social
problem. The Sociological Quarterly
33(4)483-501; Mazur, A. & Lee, J. (1993). Sounding
t
he global alarm: Environmental issues in the US national news. Social Stu
dies of
Science 23:681-720; Trumbo, C. (1995) Longi
tudinal modeling of public issues with the
agenda-se
tting process: The case of global warming. Journalism and Mass Communicat
ion
Monographs 152.
[2] Kempton, W. (1991). Lay p
erspectives on global climate change. Global
Environme
ntal Change. Human and Policy Dimensions 1(3):183-208; Bostrom, A., Morga
n, M.
G., Fischoff, B., & Read, D. (1994). What do people know about gl
obal climate change?
(1) Mental models and (2) Surve
y studies of educated laypeople. Risk Analysis 14(6):
959-982.
[3] Dunlap, R. E., & Mertig, A. G. (1992). The evolution of
the U.S. environmental
movement from 1970 to 1990: An
overview. In R. E. Dunlap & A. G. Mertig (Eds.) American
Environmentali
sm: The U.S. Environmental Movement 1970-1990 (pp. 1-9). Philadelphia:
Taylor and Francis.
[4] Mauss, A. (1975). Social Pro
blems as Social Movements. Philadelphia: J. B.
Lippinc
ott.
[5] A quasi-representative sample: Schoenfeld, A. C., Meier, R. F.
, & Griffin, R. J.
(1979). Constructing a social probl
em: The press and the environment. Social Problems,
27 (1), pp. 38-61, (p. 38); Howenstine, E. (1987). Environmental reportin
g: Shift from
1979 to 1982. Journalism Quarterly, 64
(4), pp. 842-846; Schoenfeld, A. C. (1980).
Newsper
sons and the environment today. Journalism Quarterly, 57 (3), pp. 456-462
;
Strodthoff, G. C., Hawkins, R. P., & Schoenfeld, A
. C. (1985). Media roles in a social
movement: A mod
el of ideology diffusion. Journal of Communication, 35 (2), pp. 134-53;
Atwater, T., Salwen, M. B., & Anderson, R. B. (1985).
Media agenda-setting with
environmental issues. Jou
rnalism Quarterly, 62 (2), pp. 393-397; Protess, D. L., et al.
(1987).
The impact of investigative reporting on public opinion and policy-making
:
targeting toxic waste. Public Opinion Quarterly, 5
1 (2), pp. 166-185; Mazur, A. and
Lee, J. (1993). So
unding the Global Alarm: Environmental Issues in the U.S. National
News. Social Studies of Science 23:681-720; Medler, J. F.
, & Medler, M. J. (1993).
Media images as environmen
tal policy. In R. J. Spitzer (Ed.), Media and public policy
(pp. 121-132). Westport, CT: Praeger.
[6] Mauss, A. (1975) p. 5
88.
[7] Munton, D., & Bradley, L. (1970). American Public Opinion and E
nvironmental
Pollution. The Behavioral Science Researc
h Report, Ohio State University (p. 23).
[8] Constructionism is an extr
act of the symbolic interactionist and phenomenological
traditions in sociology. For an overview see: Ritzer, G. (1992). Contem
porary
Sociological Theory . New York: McGraw-Hill.
For an application of this perspective to
issues of
science see: Sismondo, S. (1993). Some social constructions. Social Studi
es
of Science, 23, pp. 515-553.
[9] Best, J. (1989
). Afterword: Extending the constructionist perspective: A
conclusion and an introduction. In J. Best (Ed.), Images of Issues:
Typifying
Contemporary Social Problems (pp. 243-250)
, New York: Aldine De Gruyter.
[10] Spector, M. & Kitsuse, J. (1977). C
onstructing Social Problems. Menlo Park, CA:
Cummings;
Gouldner, A. (1970). The Coming Crisis in Western Sociology. New York:
Avon;
Blumer, H. (1971). Social problems as collective behavior. Social
Problems
18:298-306.
[11] Hilgartner, S., & Bosk
, C. L. (1988). The rise and fall of social problems: A
public arenas model. American Journal of Sociology 94(1):53-78; Gamson
, W. A., &
Modigliani, A. (1989). Media discourse an
d public opinion on nuclear power: A
constructionist
approach. American Journal of Sociology 95(1): 1-37.
[12] Spector & Ki
tsuse (1977) p. 75.
[13] Spector & Kitsuse (1977) p. 78.
[14] Spector
& Kitsuse (1977) p. 79.
[15] Entman, R. M. (1993). Framing: Toward cla
rification of a fractured paradigm.
Journal of Communi
cation 43(4): 51-58.
[16] Goffman, E. (1974). Frame Analysis: An Essay
on the Organization of Experience.
New York: Harper &
Row; Tuchman, G. (1978). Making News: A Study in the Construction of
Re
ality. New York: Free Press.
[17] Gitlin, T. (1980). The Whole World i
s Watching. Berkeley, CA: University of
California Pre
ss (p. 6).
[18] Gamson, W. A., & Modigliani, A. (1989). Media discourse
and public opinion on
nuclear power: A constructionis
t approach. American Journal of Sociology 95(1):1-37
(p. 3).
[19] Dunwoody, S. (1992). The media ad public perceptions of
risk: How journalists
frame risk stories. In D. W. Bro
mley & K. Segerson (Eds.), The Social Response to
En
vironmental Risk (pp. 75-99). Boston: Kluwer Academic. (p. 78).
[20] En
tman, R. M. (1993).
[21] Pan, Z., & Kosicki, G. M. (1993). Framing anal
ysis: An approach to news
discourse. Political Communi
cation 10:55-75 (pp. 59-62).
[22] Downs, A. (1972). Up and down with e
cology: The issue-attention cycle. The Public
Interest
, 28 , 38-50.
[23] Downs (1972).
[24] Dunlap, R. E. (1992). Trends i
n public opinion toward environmental issues:
1965-199
2. In, R. E. Dunlap & A. G. Mertig (Eds.), American Environmentalism. The
U.S.
Environmental Movement 1970-1990, (pp. 89-115
). Philadelphia: Taylor and Francis.
[25] Mazur & Lee (1993).
[26] Bu
sh's campaign comments are reviewed on the ABC News, 12/27/89.
[27] Wal
d, M. L. (1992, May 25). Cheap fuel hurts goal for climate. The New York
Times, pp. A-35, A-38.
[28] WuDunn, S. L. (1992, May
25). Difficult algebra for China: Coal = growth =
pol
lution. The New York Times, pp. A-1, A-5.
[29] Stevens, W. K. (1992, Ju
ne 9). Earth summit finds the years of optimism are a
fading memory. The New York Times, p. C-4.
[30] Hilgartner, S., & Bosk,
C. L. (1988). The rise and fall of social problems: A
public arenas model. American Journal of Sociology, 94 (1), pp. 53-78; M
olotch, H. L.,
Protess, D. L., & Gordon, M. T. (1987
). The media-policy connection: Ecologies of news.
In D. Paletz (Ed.) (
1988). Political Communication: Theories, Cases and Assessments
(pp. 26-48). Norwood, NJ: Ablex.
[31] During the first 6 mo
nths of 1993 there are approximately 350 column inches of
copy devoted to global warming in the five newspapers, during the fir
st 6 months of
1988 there were approximately 400 col
umn inches. For comparison, during the peak in
atten
tion in the last 6 months of 1989, the five papers devoted approximately
1200
column inches.
[32] Downs (1972) p. 41.
[33
] Gamson, W. A., Croteau, D., Hoynes, W., & Sasson, T. (1992). Media ima
ges and the
social construction of reality. Annual Review of Sociology
18:373-93.
[34] Entman (1993) p. 53.
[35] Reese, S, & Danielian, L.
(1989). Intermedia influence and the drug issue:
Conve
rging on cocaine. In P. Shoemaker (Ed.), Communication campaigns about dr
ugs (pp.
29-46). Hillsdale, NJ: Erlbaum.
[36] Merr
ill, J. (1990). Global elite: A newspaper community of reason. Gannett Ce
nter
Journal 4(4):91-102.
[37] It must be recognized that a number of
other metro newspapers contribute
significantly to th
e population of high visibility science stories produced in this
country (e.g., The Dallas Morning News, The Boston Globe, Th
e Seattle
Post-Intelligencer ). An effort was made t
o integrate a selection of such papers into
the stud
y. However, the only reasonable method of searching a wider variety of pa
pers,
and of obtaining their stories, is via the Nex
is database. Unfortunately, none of the
major papers
that would have been interesting to add to this study are represented in
Nexis prior to about 1990-1991. Many were added in
1991-1992. Since this study involves
looking at the issue longitudinall
y, newspaper selection was thus constrained.
[38] Coding for the other
variables involved manifest content and required very little
(if any) j
udgment, so reliability was not evaluated. Coding for the frame variable
does involve judgment so an inter-coder reliability te
st was executed. A value for
Scott's pi of .78 was
achieved and deemed acceptable.
[39] Note that because the claims-maker
categories are not mutually exclusive within
the unit
of analysis the crosstabulations for Figure 4 were executed exclusive to
each
claims-maker category, rather than by each pha
se as Figure 4 presents.
[40] Lichter, R. S. & Lichter, L. S. (1992). T
he Great Greenhouse Debate. Media
Coverage and Expert
Opinion on Global Warming. Media Monitor VI(10):1-6; Miller, M.
M., J. Boone, and D. Fowler, (1990). "The Emergence of Green
house Effect on the Issue
Agenda: A News Stream Anal
ysis," News Computing Journal 7(4):25-38.
|
