Limits to the Indexing Hypothesis:
A Case Study of the Reykjavik Arms Control Summit
By James White
Department of Communication Studies
Submitted for presentation at the annual meeting of the Association for
Education in Journalism and Mass Communication, August 10-13, 1994, Atlanta, GA.
The paper compares presidential and Congressional rhetoric about the 1986 arms
control meeting between President Ronald Reagan and Soviet leader Mikhail
Gorbachev with news media coverage of the event. The paper provides a case study
of how elite attempts to "frame" the news are aided or impeded by journalistic
practice. Lance Bennett's "indexing hypothesis" is used as a point of departure,
but it is concluded that the theory needs modification to better account for the
complexities of press-state relations.
Limits to the Indexing Hypothesis:
A Case Study of the Reykjavik Arms Control Summit
This paper examines how the discourse of political elites is transformed into
media discourse in coverage of foreign affairs news. Specifically, it compares
administration and Congressional rhetoric about the 1986 arms control meeting
between President Ronald Reagan and Mikhail Gorbachev, the Soviet leader, with
news media coverage of the event. The paper hopes to illustrate through a
detailed case study of day-to-day discussion and coverage of the meeting how the
attempts of political elites to "frame" the news are aided or impeded by
journalistic practice. The study uses Lance Bennett's "indexing hypothesis" as a
point of departure for examining press-state relations, but concludes by
suggesting that the theory needs modification to better account for the
complexities of news coverage of political elites.
Previous research on newsgathering routines has documented the reliance of
journalists on "official sources" for the news (Hallin, 1989). This idea has
most recently been demonstrated in studies by Lance Bennett, who has argued that
journalists "tend to 'index' the range of voices and viewpoints in both news and
editorials according to the range of views expressed in mainstream government
debate about a given topic" (Bennett, 1990). The "indexing hypothesis" has been
supported in studies of news media coverage of U.S. funding for the Nicaraguan
contras (Bennett, 1990) and the debate over the Persian Gulf War (Bennett &
Manheim, 1993). In these studies, it was found that anti-administration policy
views tended to appear in news coverage only when such views were voiced by
political elites. When no anti-administration perspective existed in elite
discourse, that perspective was absent from news stories as well.
Bennett & Manheim (1993) conclude that until people become more critical readers
of news, or until the news media begin to "decide independently on what parts of
the information record to emphasize, the role of the public in the policy
process will continue to be driven by the symbolic political strategies of
Washington elites and the public relations experts they employ."
The indexing hypothesis supports the "window on the world" image of news
practice that journalists themselves use to describe their trade (Epstein,
1973). Indexing suggests that the news media serve as a transparent conduit for
the rhetoric and actions of political elites to the public. Journalists
themselves make no admonitions about defining elite rhetoric and behavior as
news, saying that it is important for the voting public to be aware of what the
president and other political leaders do and say. If we accept journalists' own
account of what they do, we would not expect anything other than "indexing" to
take place. Indeed, independently deciding upon what parts of the information
record to emphasize, as Bennett recommends, would violate the journalistic
doctrine of objectivity.
But the conclusion to which we are led by the indexing hypothesis -- that
political elites, especially the president, can "cue" the public's thinking
about policy issues with little or no interference by the news media --
contradicts many other studies of news routines. Other characteristics of news
production might at times give journalists more independent power to frame the
issues and events they cover despite a reliance on elite sources (Entman, 1989;
Bennett, 1988; Tuchman, 1979). Journalistic tendencies to focus on the simple,
dramatic and personal rather than the complex, historical and systemic aspects
of news can hinder as well as help elite attempts to control the news. Word
choice and story organization, strategic uses of sources' statements, and
getting sources to provide the material needed to fit pre-determined narratives
are all ways that journalists maintain a powerful ability to shape the news.
Also, an apparently growing tendency to report negatively on political elites
(Smoller, 1994) seems to contradict the model of powerful elite control of the
media the indexing hypothesis suggests.
A complete theory of press-state relations must take into account all
journalistic routines, not just reliance on elite sources, to provide a
comprehensive explanation of the process by which elite discourse becomes news
The Reykjavik Meeting
On Oct. 1, 1986, President Reagan announced that he would meet with Mikhail
Gorbachev, the Soviet leader, in Reykjavik, Iceland to discuss arms control and
other issues. The purpose of the meeting, which took place Oct. 11 and 12, was
to set an agenda and date for a full-blown superpower summit on arms control to
be held in the United States sometime the following year. The announcement
followed the Soviet release of American journalist Nicholas S. Daniloff and the
U.S. release of Gennadi F. Zakharov, both of whom were being held on espionage
charges. The Soviets also allowed the emigration to the United States of two
The "pre-summit," as it was labeled by the administration, coincided with debate
in the House of Representatives over several defense-related issues. House
Democrats wanted to resolve a number of outstanding arms control issues with the
administration and Congressional Republicans. The issues included proposed curbs
on spending for the Strategic Defense Initiative (S.D.I.) and a requirement that
the U.S. unilaterally comply with the provisions of the 1979 Strategic Arms
Limitations Talks, which were never ratified. The administration and
Congressional Republicans wanted to postpone discussion and resolution of these
issues until after the Iceland meeting (Boyd, 9 October 1986). A compromise was
reached on the eve of the Reykjavik meeting, with the House Democratic
leadership dropping its proposal for compliance with S.A.L.T. and easing its
spending cuts for S.D.I. (Fuerbringer, 10 October 1986).
Gorbachev and Reagan opened their face-to-face talks in Iceland amongst a great
deal of public and press optimism that progress in arms control would be made.
But on the second day, the talks broke down over S.D.I., which Gorbachev wanted
included in any arms deal. Reagan's refusal to put S.D.I. "on the table" led to
the dissolution of the talks without achievement of their primary purpose, to
set a date for a future superpower summit.
The Reagan-Gorbachev meeting provides an interesting case study of the indexing
hypothesis because of the ebb and flow of elite debate surrounding the meeting.
In a relatively short period of time, political elites went from a period of
disagreement, when House Democrats and the administration wrangled over defense
legislation , to a period of agreement on the eve of the summit and back to
disagreement over the summit's breakdown. As the indexing hypothesis suggests,
news coverage of arms control and the Reykjavik meeting should mirror these
swings in elite discourse.
Data and methods
In Bennett's original test of the indexing hypothesis, coverage of U.S. funding
for the Nicaraguan contras in The New York Times was content analyzed. Coders
determined whether an opinion was voiced on administration policy and who the
source of that opinion was. The opinions were also coded as to whether they
supported, opposed or were neutral to administration policy. Bennett (1990)
found that the opinions appearing in the Times came overwhelmingly from
government officials, and that public opinion was sufficiently neglected as a
source of policy viewpoints in news stories.
The analysis provides important documentation of the power of elite sources to
manage the news. But certain conceptual and methodological limitations may have
prevented a true test of the indexing hypothesis. First, Bennett studies elite
discourse through the news text, rather than having independent measures for
elite debate. How can we know if news coverage is indeed indexed to the range of
voices and viewpoints in mainstream political debate if we have no independent
measure of that debate? Second, Bennett operationalizes the "range" of voices
and views as a dichotomy -- pro- or anti-administration policy views. This
operationalization may gloss over important differences in policy positions in
elite discourse and oversimplify the process by which elite discourse is
transformed into news discourse. Finally, Bennett's unit of analysis is the
voiced opinion in The New York Times Index. He does not examine the entire news
story and therefore may fail to account for the journalists' own "framing" of an
event, issue or policy.
The methodological shortcomings of Bennett's analyses are no doubt a trade-off
for the lengthy time-frame of his study. He examines almost three years of news
coverage. This study attempts a more detailed analysis of news coverage and
elite rhetoric by examining a particular case of foreign affairs coverage over a
comparatively short time period, about one month.
This paper compares elite "framing" of the Reykjavik, Iceland meeting with the
frames present in news media discourse during the same period. The data are
drawn from three sources: the Weekly Compilation of Presidential Documents, the
remarks on the Senate and House floor in The Congressional Record, and
front-page stories in the New York Times. The analysis includes all statements
from each of the three sources about arms control in general or the pre-summit
meeting in particular from September 30, when the meeting was announced, until
October 30, when the summit disappears as a continuing story from the front page
of the Times. This yielded 19 New York Times stories, 32 statements, speeches or
interviews from Weekly Compilation, and 95 speeches from the House and Senate
Finding comprehensive independent measures of elite discourse is difficult.
Admittedly, the sources of elite rhetoric here are not complete. Most, but not
all, administration communication is included in Weekly Compilation. And The
Congressional Record does not include the press releases, public statements and
press interviews given by members of Congress. But it is assumed here that these
sources provide a sufficient reflection of the entire scope of elite rhetoric to
be useful for analysis. It is reasonable to expect, for instance, that what
members of Congress tell reporters in press conferences or releases will be
echoed on the Senate or House floor.
Each of the three sources of discourse was content analyzed to determine its
dominant "frames." In popular rhetoric, frames are thought of as the particular
"spin" a news story or any other account puts on an event or issue. According to
Entman (1993), the process of framing involves selection and salience; that is,
"to frame is to select some aspects of a perceived reality and make them more
salient" through repetition or placement of certain "keywords, stock phrases,
stereotyped images, sources of information, and sentences that provide
thematically reinforcing clusters of facts or judgments." Frames function to
define problems, diagnose their causes, make moral judgments about the causal
agents, and suggest remedies (Entman, 1993).
Frames were identified and labeled for each unit of analysis in the three
sources of discourse.1 In each case, specific keywords and phrases were
identified in the text which were thought to "cue" a particular frame.
The content analysis of framing is coupled with a detailed, day-to-day
comparison of the three sources of discourse. This analysis allows us to
re-create to some degree the process by which journalists transform elite
discourse into news stories. By using independent measures of elite rhetoric,
focusing on frames rather than pro- or anti-administration opinions, and
examining the entire news text, this analysis allows a more direct and detailed
test of the indexing hypothesis.
Dominant Pre-Summit Frames in New York Times,
Congressional and Presidential Discourse
President is soft on communism
Don't tie the president's hands
Arms Control Policy
* Cells are the percentage of all units of analysis that cued a particular
frame. Most units of analysis cued more than one frame.
During the period studied, the Reagan administration seemed to be successful in
having its "framing" of the Iceland meeting and arms control issues adopted by
the news media. The themes emphasized by the president and others in the
administration were picked up and repeated, sometimes amplified, by the Times.
Table 1 shows the proportion of total units of discourse preceding the
Reagan-Gorbachev meeting in which certain frames were cued for each of the three
sources studied. The frames that appear in Table 1 are the dominant 2 frames
that appeared in either elite or news media discourse.
The "Human Rights" frame represents references to President Reagan's desire to
include human rights issues as well as arms control on the Reykjavik agenda.
Although this was a Reagan priority, this frame appeared at a higher proportion
in news media stories about the summit than it did in presidential discourse
itself. The President and administration also emphasized their "Limited
Expectations" for the pre-summit, making clear that the meeting was not designed
for the "signing of actual agreements" but rather
Dominant Post-Summit Frames in New York Times,
Congressional and Presidential Discourse
Soviets to blame for failure of talks
Arms control progress was made/Reagan
S.D.I. impeded progress/Reagan failed
Positive statements about S.D.I./Liberals weak
* Cells are the percentage of all units of analysis that cued a particular
frame. Most units of analysis cued more than one frame.
was to serve as a "base camp" for a future superpower summit. This frame also
appeared in news coverage to a greater extent than it did in presidential
discourse. With respect to these two frames, then, the news media served more as
a megaphone than a mere mouthpiece for the administration, amplifying the
But the Times made other framing choices that broke with elite discourse. The
idea that the "president is soft on communism" because he agreed to free Gennadi
F. Zakharov, a confessed Soviet spy, appeared in a much higher proportion of
news stories than it did in either source of elite rhetoric. And discussion of
specific "arms control policies", a frequent subject of Congressional
speech-making especially in the Senate, appeared infrequently in news coverage.
More will be said about these two frames later. Finally, the debate between the
administration and Congressional Republicans, and House Democrats over whether
arms control provisions being discussed in Congress would "tie the president's
hands" at the negotiating table, appeared to a much lesser degree in the media
than they did in elite discourse.
Discourse following the dissolution of the pre-summit meeting in Iceland showed
a similar pattern. According to Table 2, the administration's top two frames --
that the "Soviets were to blame for the failure" of the talks, and that despite
the breakdown in negotiations, "progress was made" -- were also the most
frequent frames appearing in Times stories. But discrepancies between elite
discourse and news coverage again appear when we consider the other dominant
frames. A small number of House Democrats made speeches attempting to frame the
Iceland meeting as a "Reagan failure" in which the president allowed the
Strategic Defense Initiative to "get in the way" of real arms control progress.
This frame was amplified greatly by the Times. In fact, it appeared more
frequently than the other dominant administration frame that described S.D.I. in
positive terms and liberals who opposed it as "weak on defense."
This analysis offers mixed support for the indexing hypothesis. While some
frames being promoted by elites in both the administration and Congress were
picked up and relayed by the news media, others were not. More interesting is
the disproportionate relationship between media framing choices and the frames
present in elite discourse. While the evidence here suggests that the media must
rely on elite sources for news, it also suggests that other news practices may
impede the ability of elites to provide cues directly to the public through the
news media. In the next section, a closer analysis of news coverage of the
Iceland meeting reveals some of the other news practices and how they influence
the transformation of elite discourse into media discourse.
Keys to press power: creating conflict, creating frames, simplification and
Upon closer inspection, it becomes apparent that despite a frequent reliance on
elite sources for the news, journalists retain a great deal of power to
"reconstruct" elite discourse in ways that meet the demands of other news norms.
In covering the Reykjavik meeting, journalists "created conflict" when little or
none existed in elite discourse; they created news frames of their own when
elites were reluctant to provide information; they simplified rather than
"indexed" the range of frames in elite discourse; and they employed "critical
distance" to cast doubt on elite comments and reveal the political motivation
behind elite rhetoric.
The indexing hypothesis suggests that when "official debate is intense and
extended, information that is critical of existing or proposed administration
policies is more likely to become emphasized in mainstream news coverage"
(Bennett, 1993). Conversely, when official debate is restricted, critical
information is expected to be limited in the press as well. But in coverage of
the Reykjavik meeting, conflicting points of view appeared in news texts without
direct attribution to an elite source and when no conflicting point of view was
offered in the sources of administration or Congressional rhetoric studied.
The Oct. 1 New York Times story framed the announcement of the pre-summit
meeting as part of a deal that included freeing an admitted Soviet spy, despite
Reagan's repeated comments denying any such deal. This discrepancy between media
and elite framing can be seen clearly in Table 1 above by comparing the degree
to which the "Reagan is soft on communism" frame is cued in each of the three
sources of discourse.
No statements conflicting with the Reagan line appeared that day in
Congressional discourse. Rather, the story apparently drew from the reporter's
own observations and suspicions to contest Reagan's account. Without
attribution, the story said that the "simultaneous announcement" of the
pre-summit meeting by Reagan and the Soviets "underscored the appearance that
the United States had agreed to the meeting as part of a resolution of the
Daniloff-Zakharov cases." To whom, other than the journalist, this "appearance"
presented itself is unclear.
A paragraph later, the story continues to provide statements that conflict with
Reagan's explanation of the impetus for the Iceland meeting.
The agreement on the Reagan-Gorbachev meeting in Iceland was apparently a key
element in the negotiations that led to the release of an American journalist
Nicholas S. Daniloff, and of Mr. Zakharov, both of whom were held on espionage
charges. . . . President Reagan and other officials insisted that the Iceland
meeting was not part of a deal. But in private briefings and in some public
statements, officials made it clear that the meeting would not have taken place
as long as Mr. Daniloff remained in Soviet custody (my emphasis).
Reagan is then quoted as saying, "The release of Daniloff made the meeting
possible . . . I could not have accepted and held that meeting if he was still
being held." What does not appear is another statement from Reagan apparently
from the same news conference, the transcript of which appears in Weekly
Compilation. In that news conference, Reagan responded to the following question
from a reporter:
Q. Well, didn't you agree to this meeting in Iceland as part of the effort to
get Daniloff free?
Reagan: No. I just said that there wouldn't be any meeting until he was free.
By omitting this statement, which would have further clarified the Reagan
position, and using words such as "insisted" to describe the Reagan position,
and "but" to counterpose what is framed as contrary evidence, the journalist
constructs a frame for which there apparently was no quotable support in elite
Drama and conflict are important components of news narratives even when such
conflict isn't directly provided by the sources of news. While elite
disagreement may ebb and flow, the narrative demand for conflict in news stories
remains constant. As this story shows, in instances when conflict is not readily
available among elite sources, journalists must rely on word choice, careful
selection of quoted material and unattributed descriptions to build conflict
into the narrative.
Journalists not only have to construct conflict at times but also have to
construct dominant frames for stories when sources do not reveal frames of their
own. In coverage of the Reykjavik meeting, these choices had important
implications for future news coverage. The Times story about the first day of
meetings between Reagan and Gorbachev drew on glimpses of the two leaders,
unnamed sources and brief passing comments of administration officials to
construct a story that framed the meeting in very positive and hopeful terms.
The story began by saying that "prospects were high" for arms control progress;
that the mood of the U.S. and Soviet leaders was "upbeat"; and that they
"expressed optimism" and met in "good-natured friendliness." The story went onto
say that Gorbachev wanted to "sign something" in Iceland "that would mark a
substantial step forward in Soviet relations with the United States." While most
of these comments were attributed to "administration sources" none of them
appeared in quotation marks, suggesting they were the reporter's interpretation
of administration sentiment.
In fact, the only direct quote appearing in the story from a named source that
describes the mood of the talks came from Vice Adm. John Poindexter, who, when
passing reporters in a hallway, described the meetings as "businesslike." Later
in the story we discover that Poindexter was apparently severely reprimanded by
White House press spokesman Larry Speakes for letting even that much information
The optimistic and hopeful framing of this story appeared despite the apparent
success with which Reagan got across his "Limited Expectations" frame throughout
the week preceding the Iceland meeting. In this sense, the journalist could be
seen as again constructing a conflicting account of the event in the ongoing
story of Reykjavik. But this story also illustrates the ability of journalists
to develop frames on their own, drawing from very little elite discourse. The
frame that was chosen for this story becomes important in future coverage of the
Reagan-Gorbachev meeting because it functions to intensify disappointment when
the meeting breaks down. This is, no doubt, the effect Reagan was trying to
avoid by promoting the "Limited Expectations" frame throughout the preceding
To satisfy the production cycle of news and the narrative constraints of writing
news stories, journalists often simplify events and issues. The process of
simplification was at work in Times coverage of the Iceland meeting. Rather than
indexing to the range of elite discourse, oftentimes journalists would simplify
this range, opting to gloss over the differences in political arguments and
instead fit arguments into a simple Republican v. Democrat dichotomy.
For instance, several very lengthy and sophisticated discussions of particular
arms control policy positions appear in Congressional discourse, especially in
the Senate. But none of these policy positions are reported. Mentions of policy
were restricted in the Times to descriptions of the administration agenda in
Also, throughout the period surrounding the Iceland meeting, lengthy and highly
nuanced political speeches by Democratic Sen. William Proxmire broke from the
standard pro- and anti-administration characterization of the debate. Yet these
statements were not included in the Time's discussion of Congressional debate
about the summit aftermath. Following the meeting's dissolution, for example,
Sen. Proxmire gave a long and detailed speech on the Senate floor critical of
both the Reagan arms agenda at Reykjavik and the arms reduction proposals of
many of his fellow Democrats.
But in the only story which mentioned Congressional reaction to the breakdown of
the meeting, the Times reported that "Congressional reaction generally was
mixed and along partisan lines with Republicans supporting Mr. Reagan's refusal
to restrict "Star Wars" development and Democrats criticizing the President for
missing what they said was an historic opportunity to eliminate offensive
The journalistic tendency to simplify the news, then, may function to limit the
range of voices and views in elite discourse even when that range is fairly
When elite disagreement did occur, the frames were seldom transmitted unhindered
through news texts. Rather, journalists assumed a critical stance that revealed
the strategies and political gamesmanship behind elite rhetoric and made
explicit the journalist's awareness that elites were attempting to "spin" or
"manipulate" the news. This critical distance also at times resembled the "horse
race" coverage of political election campaigns. In horse race coverage,
journalists favor discussion of winners, losers and political strategizing over
substantive discussion of policy proposals.
Following the meeting, the efforts of Reagan, Gorbachev, and other elites to
define what happened in Reykjavik were described in a way that exposed the
motivation behind the rhetoric and functioned to cast doubt on the validity of
the statements. For instance, after the talks broke down, an Oct. 14 story about
Reagan "defending his stand" said, "Each (meaning the U.S. and Soviet Union) in
the coming days will undoubtedly seek to portray the meeting as a success for
itself and a defeat for the other." The same story mentioned that Reagan was
"trying to put the best face on the deadlock over the Star Wars plan." The
following day, the paper's lead story said that the Russians had engaged the
U.S. "in a public competition to promote their own interpretations of the
outcome of the meeting in Iceland."
Another characteristic of this frame is to subtlety discredit statements by
changing the verbs of attributed statements from said to "claims," "insists,"
asserts," or "contends." These words do not necessarily suggest that the
statement is false, but rather recognize that the words are uttered in the
context of a debate in which the truth is not known.
These words are often included in stories where journalists openly recognize
that the information they are reporting comes from official "campaigns" of news
manipulation. For instance, one post-summit passage read: "The White House also
began an effort today to salvage something from the Iceland meeting by asserting
that the issue of intermediate range missiles had been effectively resolved in
Reykjavik despite the absence of a formal agreement." And elsewhere, "The White
House has begun an extensive campaign to blame Moscow for the summit results."
Critical distance is achieved here by placing the administration frames --
"progress was made" and "the Soviets are to blame" -- in the context of a public
relations "effort" or "campaign" to shape news coverage and public opinion.
While the administration frames are conveyed, they are presented in a context
that does not necessarily reflect favorably on them.
Critical distance also is achieved by revealing the political motivation behind
elite rhetoric. In a story about the White House debate with House Democrats
preceding the Reykjavik meeting, the Times noted that "both parties were
maneuvering for possible advantage in the Congressional elections." Similarly,
the Oct. 11 story about the administration compromise with Congressional leaders
mentioned that Democratic "leaders have been worried this week about the
political damage in this election year from appearing to tie the President's
hands in his negotiations with Moscow."
But even in stories where critical distance was present, journalists sometimes
would still uncritically relay elite frames. In a story that in one place
described Reagan as using the Iceland summit "as a partisan issue in an election
campaign," the journalist later without attribution, relayed Reagan's frame of
blaming the Soviets for the dissolution of talks:
In the meeting last weekend, Mr. Reagan and Mr. Gorbachev reached tentative
understandings on a wide range of arms issues, but the talks broke down over
Soviet insistence that the American missile defense program be limited to
research and testing within the laboratory for 10 years.
Finally, Poll results, often part of a horse race frame in campaign coverage,
also appeared to gauge whether the President's attempts to manipulate public
opinion were working politically. A poll favorable to Reagan was cited in one
post-summit story followed by the statement, "But there was no sign that the
support had translated into a Republican advantage in the November elections."
This study presents both evidence of the indexing hypothesis and evidence that
suggests refinement. An analysis of the dominant frames present in elite and
media discourse illustrates that political elites, especially the president,
have a great deal of power to frame news coverage of an issue or event. But it
is also clear that in this case study, news routines other than reliance on
elite sources impacted upon media frames. These included the needs for drama and
conflict in news narratives; the development of frames when not provided
explicitly by elites; the simplification of elite rhetoric; and the application
of critical distance to reporting about elite debate.
This study also suggests that "indexing" is not proportional. That is, while the
frames that appear in the news are often those that appear in elite discourse,
they do not always appear with the same intensity. For instance, while the
debate over House Democrats was present in a large amount of Congressional and
administration rhetoric, this intensity did not carry over into news coverage
(See Table 1). In other instances, the media exaggerated the intensity of elite
opinions. Though the proportion of members of Congress promoting the "Reagan
failed" frame following the summit was minimal, this viewpoint was prevalent in
This disproportionality could be explained by the journalistic norm of "balance"
which requires representing both sides of a debate. Balance does not, however,
mean proportional balance. Rather, it functions to equalize sides of a debate
that may in other discursive realms be unevenly represented.
The indexing hypothesis is a useful starting point from which to analyze the
relationship between the news media and political elites. But news coverage of
the Reykjavik meeting suggests it is not a complete explanation of press-state
relations. Rather, other newsgathering routines complicate the tendency of the
news media to "index" to elite sources. These routines sometimes help and
sometimes hinder elite attempts to manage the news. Future research is needed to
explore indexing and other news practices to determine in what specific
instances certain news routines influence news production more than others.
Bennett, W. Lance & Manheim, Jarol B. (1993). Information, cuing, and the
Democratic process in the Gulf conflict. Political Communication, Vol. 10, No.
Bennett, W. Lance (1990). Toward a theory of press-state relations in the United
States. Journal of Communication, 40(2), Spring, 103-125.
Bennett, W. Lance (1988). News: The politics of illusion. New York: Longman.
Boyd, Gerald (1986, October 9). Reagan rules out House compromise on arms
control. The New York Times, p. A1.
Entman, Robert M. (1993). Framing: Toward clarification of a fractured paradigm.
Journal of Communication, 43(3), Autumn, 51-58.
Entman, Robert M. (1989). Democracy without citizens: Media and the decay of
American politics. New York: Oxford University Press.
Epstein, E. J. (1973). News from nowhere: Television and the news. New York:
Fuerbringer, Jonathan (1986, October 10). The Iceland summit; and a compromise
on Capital Hill. The New York Times, p. A1.
Hallin, Daniel C. (1989). The uncensored war: The media and Vietnam. Berkeley:
University of California Press.
Smoller, Fred T. (1994). The six o'clock presidency: Patterns of news coverage
of the president. In Graber, D.A. (Ed.), Media power in politics (pp. 227-238).
Washington DC: CQ Press.
Tuchman, Gaye (1979). Making news by doing work: Routinizing the unexpected.
American Journal of Sociology, Vol. 79, No. 1, 110-131.
1 The unit of analysis for Weekly Compilation was the speech, letter, press
release or interview; for Congressional Quarterly, the speech; and for the New
York Times, the news story.
2 Dominant frames are those that appeared in at least 30% of the discourse for
at least one source.