Historical trends in questioning presidents, 1953-2000.
Moreover, the culture of the press corps--in particular, its tendency toward either a deferential or adversarial posture--is a central structural contingency shaping the institution of the presidency and presidential conduct (cf. Hager and Sullivan 1994). Thus, when the partisan press of the early nineteenth century was superceded by a more independent and unwieldy commercial press, it led to the institutionalization of the presidential press conference, the press secretary, and other vehicles of presidential news management (Ponder 1998). More recently, when the news media of the 1960s was perceived as magnifying societal unrest and dissent surrounding the civil rights movement and the Vietnam War, administration officials responded with more strategic and comprehensive public relations planning, creating a new agency--the Office of Communications--devoted to this task (Maltese 1994). Accordingly, the presidency cannot be fully understood without taking account of the media landscape through which presidents must navigate in pursuit of popular support.
How hazardous is this landscape? Numerous scholars have argued that journalists since the 1950s have become increasingly independent, vigorous, and at times adversarial in their treatment of presidents, presidential candidates, and government officials generally. This idea was advanced most forcefully by Michael Robinson (1976) and has since received empirical support (Entman 2003; Hallin 1992; Hart, Smith-Howell, and Llewellyn 1990; Patterson 1993, 2000; Ragsdale 1997; Robinson 1981; Rozell 1994; Sabato 1991; Smoller 1990; see also Cohen 2004). However, much remains unknown about the magnitude and scope of this change. American journalists remain heavily dependent on government officials as sources of both information and opinion, such that negative and critical coverage tends to be contingent on the emergence of policy critics among officials themselves (Bennett 1990; Epstein 1975; Hallin 1984). Indeed, when critical content is documented in news coverage, it is often difficult to determine how much should be attributed to journalists per se as opposed to the authorities on which they are dependent, as well as the extrajournalistic reality in which both are embedded.
This complex picture has emerged from studies using traditional news stories, whether print or broadcast, as data. Overlooked are other modes of journalistic practice such as broadcast news interviews and news conferences, where journalists directly encounter public figures rather than merely writing or talking about them. Does the impetus toward greater vigorousness extend to these direct exchanges between journalists and officials? The answer is by no means obvious. Aggressiveness here would be enacted in the immediate presence of, and addressed directly to, the politician being targeted. This runs contrary to established norms of interactional politeness (Goffman 1967; Brown and Levinson 1987; Holtgraves 1992) and rituals of deference toward political leaders (Shils 1975; Schwartz 1987; Alexander 1989), which might be expected to inhibit aggressive questioning or at least temper the manner in which it is expressed. The domain of journalist-politician interactions thus represents a particularly strong test of the vigorousness of political journalism. More generally, this domain provides an appealingly direct window into the evolving relationship between journalists, politicians, and the institutions that they represent. And yet, despite its potential as a key research site, the domain of journalist-politician interactions has yet to be fully exploited by social scientists.
One partial exception is the presidential news conference. There is a long tradition of research on such conferences, including broad historical overviews in the context of president-press relations (Cornwell 1965; French 1982; Grossman and Kumar 1981; Juergens 1981; Kumar 2005; Pollard 1947; Smith 1990; Tebbel and Watts 1985) and studies of more narrowly defined topics (Cornwell 1960; Kumar 2003; Lammers 1981; Manheim 1979; Manheim and Lammers 1981; McGuire 1967). This research, while illuminating, focuses less on the substance of what actually transpires within news conferences in favor of the conditions under which they occur, such as their initial growth and institutionalization, their increasingly public character, the declining frequency with which they are held, and so on. When conduct internal to the news conference is considered, the observations tend to be broad and impressionistic in character. Efforts to analyze journalistic conduct systematically in a way that would permit quantification (McGuire 1967; Manheim 1979) are scant and underdeveloped.
The scarcity of research arises from the fact that, when interactional conduct is involved, fruitful quantification requires a thorough grasp of how journalistic vigorousness or aggressiveness is instantiated at the ground level, in actual practices of questioning and interrogation. Aggressive professionalism, to the extent that it exists, inheres in the details of linguistic and discursive behavior, details involving not only what questions are asked but also how they are asked in ways that exert varying degrees of pressure and constraint on politicians. Such micro-level practices have not been explored by scholars of journalism and political communication, and indeed scholars have been pessimistic about the prospects for measurement and quantification in this area. Michael Schudson, in a thoughtful assessment of the impact of Watergate on American journalism, observes that "civility is not something easy to measure" (Schudson 1995, 151). Focusing on presidential news conferences in particular, Carolyn Smith notes that, while adversarialness can be illustrated anecdotally, "there is little systematic evidence to demonstrate this relationship" (Smith 1990, 10-11). Finally, Samuel Kernell notes that "the adversarial aspect of presidential-press relations is an elusive quality, difficult to quantify" (Kernell 1986, 76).
Recently, however, researchers working in the tradition of conversation analysis (1) have begun to examine journalistic practices in the context of broadcast news interview talk. (2) Although news interview research is predominantly qualitative and based on contemporary data drawn from England and the United States, it has generated findings about basic forms of questioning that can be mobilized in the service of comparative research in a quantitative mode. Building on this work, Clayman and Heritage (2002b) developed a new system for quantifying the level of aggressiveness encoded in journalists' questions and applied that system in a comparative study of the news conferences of Dwight Eisenhower and Ronald Reagan. The results revealed that, for ten different features of question design, journalists were significantly more aggressive in their treatment of Reagan than they were of Eisenhower. However, with only two historical signposts examined thus far, it is not known whether these cases are exceptional or indicative of general historical trends and, if the latter, what their shape might be.
The present study begins to fill in the historical picture by refining and applying the question analysis system to a continuous sample of presidential news conferences from Eisenhower through Clinton. The objectives are both substantive and methodological. Substantively, the aim is to use the design of journalists' questions as a window into the vigorousness of the White House press corps and its evolving relationship to the presidency. Methodologically, the goal is to further refine and develop, through the question analysis system, valid and reliable measures of journalistic aggressiveness, an essential prerequisite to hypothesis testing and theorizing about how such journalism is affected by, and consequential for, the sociopolitical world in which it is embedded.
The Question Analysis System
Dimensions of Aggressive Questioning
The question analysis system, a revised version of the system developed in Clayman and Heritage (2002b), decomposes the phenomenon of aggressive questioning into four component dimensions.
1. Initiative: the extent to which journalists are enterprising rather than passive in their questioning.
2. Directness: the extent to which journalists are blunt rather than cautious in raising issues through their questions.
3. Assertiveness: the extent to which questions invite a particular answer and are in that sense opinionated rather than neutral.
4. Adversarialness: the extent to which questions pursue an agenda in opposition to the president or his administration.
Granting the difficulty of either distinguishing or fully separating the substantive "content" of a question from its linguistic or discursive "form," the first three dimensions are predominantly concerned with matters of form, while the fourth dimension is predominantly concerned with content or what the question is "about."
Each dimension of aggressiveness is operationalized in terms of various features of question design which serve as indicators. Consider initiative--journalists exercise initiative when they (1) preface their question with statements that construct a context for the question to follow, or (2) ask more than one question within a single turn at talk, or (3) ask a follow-up question. The first two practices are illustrated below. While the first example contains a simple, unelaborated question embodying minimal initiative, the second example is much more elaborate and enterprising. The journalist begins (at the first arrow) with an extended prefatory statement drawing an ominous analogy between U.S. military operations in Lebanon and parallel operations pursued previously by the French in Vietnam. And instead of asking a single question, this journalist proceeds to ask two distinct questions (arrowed), one inviting the president to embrace the Vietnam analogy, and the other asking when U.S. troops will be redeployed.
(1) (Eisenhower 2 April 1953: question 1)
JRN: Mr. President, what is your estimation or analysis of the recent peace overtures from Russia and Communist China?
(2) (Reagan 19 October 1983: 20)
JRN: S [right arrow] Mr. President, before the United States went into Vietnam, the French suffered a devastating defeat there by putting their troops in a saucer-shaped depression with the enemy up around the sides shooting down at them.
Q [right arrow] Doesn't this appear uncomfortably similar to you to the way we are deploying our troops in Lebanon on the low ground?
Q [right arrow] And how soon can we expect that we're going to redeploy them to a spot that makes more sense?
The incorporation of multiple questions places greater demands on the president, while the statement preface--which provides grounds for the Vietnam analogy--makes them more difficult to sidestep or resist. Both features thus embody the exercise of initiative.
The third indicator of initiative, follow-up questions, is exemplified in excerpt 3. Given the conditions of speech exchange in news conferences, where large numbers of journalists must bid for the president's permission to ask each question, most journalists get only a single turn at talk. However, occasionally the same journalist who asked a particular question will, immediately after the president's response, regain the floor to press for a more substantial answer or to raise a related matter. For instance:
(3) (Reagan 16 June 1981: 19-20)
JRN: Mr. President, do you approve of conservative fundraising groups such as NCPAC making these expensive television commercials targeting liberal Democrats for defeat in the next election?
RR: I don't really know how to answer that, because the game of politics is trying to win an election ...
JRN: [right arrow] If I may follow up on that sir, is it really a sense of fair play that these groups with all their money are, in effect, ganging up on one member of Congress to make him an object lesson for other wavering Congressmen who might not see things their way?
Follow-up questions embody the exercise of initiative, in general by exceeding the one-turn-per-journalist convention, and in many cases by declining to accept as adequate a response that the president offered as adequate. However, the analysis of follow-up questions is complicated by the fact that success in getting such a question is contingent on the president, who can choose to either allow a follow-up question, or forestall it by calling on someone else. The frequency of follow-up questions is thus an imperfect indicator of journalistic initiative per se.
Directness, which involves the extent to which the journalist delivers the question bluntly rather than cautiously, is operationalized in terms of its opposite--various practices that embody an indirect or cautious stance toward the question. Journalists are conventionally indirect (Brown and Levinson 1987, 132-45) when they frame their questions with self-referencing phrases (e.g., "I wonder whether ...," "I want/would like to ask ...," "Can I/Could I/May I ask ...") that refer to the journalist's own intentions or desires before launching into the question proper. The most indirect of these frames imply a virtual request for permission to ask the ensuing question. For example:
(4) (Eisenhower 8 October 1953: 20)
JRN: Mr. President, I ask this because many of us are not well acquainted with Mr. Mitchell.
[right arrow] Could I ask you the same question, sir, in reference to him that was asked about Chief Justice Warren last week, that is, what are the qualifications that attracted Mr. Mitchell to you, as Secretary of Labor?
Cautiousness is also manifest through the use of other-referencing frames that invoke the president's ability (e.g., "Can you/Could you tell us ...") or willingness ("Will you/ Would you tell us ...") to answer the question, and hence allow for the possibility that he may be unable or unwilling to answer. The most cautious variant focuses on willingness, which licenses the possibility that the president may not answer simply as a matter of choice or preference. For example:
(5) (Eisenhower 7 April 1954: 4)
JRN: [right arrow] Mr. President, would you care to say anything to us about the loyalty and patriotism of Edward R. Murrow?
Both self- and other-referencing question frames embody indirectness and reduce the level of coercion encoded in the question. Conversely, the absence of such frames is indicative of directness in putting issues before the president. (3)
Assertiveness, which captures the extent to which the question invites a particular answer and is thus opinionated rather than neutral, is operationalized for yes/no questions only, where the phenomenon is most transparent. Yes/no questions can be designed to invite or prefer either a yes- or no-type response in two distinct ways: (1) through the question preface, or (2) through the linguistic form of the question itself. Both indicators of assertiveness are present in the following example.
(6) (Reagan 19 October 1983: 28)
JRN: 1 [right arrow] Mr. President, new figures out today show that housing starts were down pretty sharply last month, and the number of building permits went down for the second month in a row. Analysts are saying this could mean the economic recovery is going to level off, maybe kind of peter out next year. And more people are becoming concerned about high interest rates. And given the big deficits being projected by your own administration,
2 [right arrow] isn't it time for some strong action by you to get interest rates down?
Here the journalist begins with an extended preface (starting at arrow 1) detailing various bits of bad economic news portending a downturn in the business cycle, all of which favors a yes answer to the subsequent question about the need to reduce interest rates. A similar tilt toward yes is encoded in the linguistic form of the question itself, which is negatively formulated ("isn't it time ..."). As a general principle, negative interrogatives (e.g., "isn't it/doesn't it/don't you think") are heavily tilted in favor of yes-type answers (Heritage 2002b; Clayman and Heritage 2002a, Chapter 6).
Among assertive prefaces, the coding system further distinguishes between those that are innocuous in that the answer being sought is not unfavorable toward the president, versus those that are substantively unfavorable. Excerpt 6 exemplifies the latter. The preface not only favors a yes answer, but this answer would require President Reagan to back down from an avowed position on fiscal policy (i.e., the theory that tax cuts would be self-regenerating and would not increase the federal deficit), and would also confirm the premise that the economy has begun to slow under his watch. In effect, the preface tilts the question against the president.
Adversarialness, which captures the extent to which the question pursues an agenda in opposition to the president or his administration, is concerned mainly with question content. An oppositional stance can be encoded (1) in the preface to the question only, or (2) in the design of the question as a whole. For an illustration of the former, consider example 7, where the preface reports an accusation made by Adlai Stevenson to the effect that Eisenhower is taking credit for civil rights progress achieved by Democrats. Here the preface is overtly critical of the president, but the subsequent question merely invites Eisenhower to respond to the accusation and so is not in itself critical.
(7) (Eisenhower 5 October 1956: 1)
JRN: S [right arrow] Mr. President, Adlai Stevenson said in a civil rights speech in Harlem that you were trying to run on the Democratic record, that the Democrats started desegregation of the Armed Forces, and that the Republicans have made a brazen attempt to take credit for civil rights progress.
Q [right arrow] Would you care to comment on that, Sir?
Among adversarial prefaces, the coding system further distinguishes between varieties in terms of how they are treated by the subsequent question. An adversarial preface can be the focus of the subsequent question, as in example 7, which gives the president a direct opportunity to counter or refute the prefatory information and is thus relatively mild in character. In the more aggressive variant, an adversarial preface is presupposed by the subsequent question, as in example 8 below. Here the preface sets up a damaging contrast between President Reagan's promises regarding tax cuts and his actual performance, such that the contrast portrays Reagan as twice caving in to special interests. Notice that in this case the subsequent question does not invite Reagan to comment on the prefatory criticism. Instead, the question assumes that the preface is true, and draws out an inference about Reagan's general susceptibility to pressure from special interests. It is this inference that Reagan is asked to respond to.
(8) (Reagan 16 June 1981: 14)
JRN: S [right arrow] Mister President, for months you said you wouldn't modify your tax cut plan and then you did. And when the business community vociferously complained, you changed your plan again.
Q [right arrow] I just wondered whether Congress and other special interest groups might get the message that if they yelled and screamed loud enough, you might modify your tax plan again.
This type of question, by presupposing the truth of the preface and denying the president a direct opportunity to counter it, is plainly more adversarial.
In addition to the coding of preface adversarialness, which can only apply to statement-prefaced questions, questions were also coded for global adversarialness, which applies to all questions and is based on an assessment of the gestalt of the question as a whole. Questions were coded as globally adversarial if, however brief or elaborate they might be, it was determined that an oppositional or critical posture ran through the question in its entirety (e.g., example 8).
Beyond preface and global adversarialness is a third and somewhat different manifestation of the same underlying dimension: accountability questions, which ask the president to explain why he's adopted a policy or taken some course of action. Because such questions require the president to provide a rationale for and in effect justify his actions, they embody at least an implicitly critical stance toward official policy. However, the degree of adversarialness depends on the form of such questions--Why did you-type questions are formally neutral regarding the justifiability of the president's actions, whereas How could you-type questions imply an attitude of doubt or skepticism regarding the president's capacity to adequately defend his actions.
Table 1 summarizes the various dimensions of aggressiveness, the design features that serve as indicators of each dimension, and the values or levels of each indicator. For a more thorough discussion, see Clayman and Heritage (2002b).
Reliability and Validity
Although both question "content" and "form" are captured by the coding system, the majority of indicators are concerned with relatively formal aspects of question design. This has various advantages. First, it facilitates a focus on journalistic culture. Coding categories based on the thematic or topical content of questions (e.g., questions about the president's conflicts with Congress, his private life, etc.) are less likely to tap into the culture of journalism per se, so much as the extra-journalistic reality of a particular administration. Thus, the present coding system focuses on relatively formal aspects of question design, and incorporates question content in ways that are general enough to be comparable across presidents.
The emphasis on formal design features has additional advantages for both the reliability and validity of the coding system. Thematic content categories tend to be highly interpretive, and their application requires considerable judgment (Krippendorff 1980, 62-63); formal design features are relatively concrete and hence more reliably codable. Coding was performed by a team of fourteen coders who worked in pairs and whose decisions required consensus. Problem cases were resolved in weekly meetings involving the authors and the entire coding team. Reliability was assessed by recoding a subsample of ten conferences, using Cohen's kappa to evaluate the level of agreement. Of the ten indicators, four exceeded the .90 threshold, three others exceeded .75, and three exceeded .65. Note that the reliability of dimensions, each composed of two to three indicators, tends to be higher than that of discrete indicators (Table 1, far right column), with three measures at or above the .80 threshold and one just shy of that level at .78. Because the statistical analysis is based mainly on composite measures of the four basic dimensions (discussed below), these composite kappa scores more accurately capture the reliability of the coding system. Given that kappa scores at or above .75 are believed to indicate an adjusted agreement level of 90 percent or higher (although this general rule may underestimate the level of agreement when a coding category has few component codes; see Bakeman et al. 1997), the coding system is indeed highly reliable.
As for validity, formal features of question design have been the subject of substantial prior research, both on journalistic questioning per se and on questioning practices in interaction generally. (4) This research demonstrates that specific design features are indeed understood and treated by interactants as embodying aggressiveness in various forms. Consequently, the validity of these design features as indicators of aggressiveness has for the most part been established.
Question Design and Social Norms
The forms of conduct captured by the coding system can be understood in terms of how they relate to both widespread norms of interaction and norms specific to the professional context of journalism. Consider, first, general interactional norms. It has long been recognized that social conduct is shaped by certain fundamental human desires that persons are obliged to respect. Erving Goffman (1967) conceptualized these basic desires in terms of "face," and the mutual obligations in terms of "face-work." Goffman's analysis was further developed and systematized by Brown and Levinson (1987), who also drew inspiration from Durkheim's (1915) distinction between negative and positive rites. Brown and Levinson thus distinguished between negative face (the desire to be free from imposition) and positive face (the desire for approval or validation), and they operationalized face-work in terms of a wide range of linguistic strategies that attend to these different aspects of face. Brown and Levinson also proposed that the selection of linguistic strategies is correlated with social variables such as the participants' social distance and relative power vis-a-vis one another, although the underlying desires--to be unimpeded and to be favorably regarded--are putatively universal features of human conduct.
Against this backdrop, the first three dimensions of the coding system--initiative, directness, and assertiveness--impose varying forms of pressure and constraint on presidents' responses, thereby impinging primarily on presidents' negative face. In contrast, the fourth dimension--adversarialness--introduces information that disagrees with or is critical of the president, and is thereby primarily (if not exclusively) threatening to presidents' positive face. Correspondingly, variations in the frequency of these forms of conduct may be indicative of changing levels of social distance and/or asymmetries of power between journalists and the president.
Regarding the more specialized norms of journalism, it seems clear that the dimensions of aggressiveness vary in their professional salience. Initiative, assertiveness, and adversarialness are each closely bound up with journalistic norms and ideals concerning independence, objectivity, and the watchdog role of the press. Correspondingly, many of the specific practices that serve as indicators of these dimensions (especially multiple questions, follow-up questions, and prefaced questions including both assertive and adversarial prefaces) are specialized for the environment of broadcast journalism and are found infrequently in ordinary conversation. By contrast, the dimension of directness stands out in the degree to which it is unrelated to journalistic norms per se, but is tied to more general norms of politeness and civility. Correspondingly, the specific indicators of indirectness--self- and other-referencing frames--are by no means restricted to journalism, but are found across a wide range of interactions including ordinary conversation (Brown and Levinson 1987, 132-45).
Applying the System
Historical Time Frame and Sampling Procedure
This study encompasses the administrations of Eisenhower through Clinton, a time frame that roughly spans the era of the public news conference. Earlier news conferences were essentially private encounters between presidents and journalists, with strict rules governing the use of quotations and the manner in which they could be attributed to the president (Cornwell 1965; Smith 1990). For much of the first half of the twentieth century, the default assumption was that the president's remarks were "off the record" unless otherwise indicated, so that journalists could not quote the president without receiving explicit permission to do so. These rules were relaxed significantly during the Truman administration, but it was not until Eisenhower that the news conference became fully public and "for the record," with most content available for verbatim quotation, full attribution, and subsequent broadcast.
Using transcripts reprinted in Public Papers of the Presidents, four conferences were sampled per year from 1953 to 2000. The conferences were staggered quarterly over the course of each year--using February 1, April 1, July 1, and October 1 as starting dates, the first conference held after each date was selected. A temporally stratified sample was chosen on the grounds that president-press relations are believed to be cyclical in nature, with an initial honeymoon period followed by more aggressive treatment (Manheim 1979; Smoller 1990). Conferences held beyond White House grounds, and those involving other officials in addition to the president, were excluded from the data sample. Because no conferences were held in 28 quarters, this sampling procedure yielded a database of 164 conferences and 4,608 distinct questions.
For each dimension of aggressiveness, discrete indicators were combined to form a single composite measure of that dimension (as outlined in Tables 2-5). In general, higher values of each measure are indicated when the more aggressive practices are used, or when aggressive practices are used in combination. With the exception of initiative (a binary measure), the composite measures have multiple levels and were treated as ordinal variables, not assuming interval scale properties or a normal distribution. To determine whether a single underlying construct is being measured ordinally throughout each measure, we predicted each outcome from the time at which the press conference was held in an ordinal logistic regression model and examined the test of the assumption of proportional odds. The tests revealed only a single violation of the assumption of proportional odds (p < 0.05), which was corrected by collapsing adjacent levels of the scale.
The only original outcome variable that for conceptual reasons could not be assimilated into a composite measure is the variable for accountability questions. Accordingly, this variable is analyzed separately.
Ordinal logistic regression models were run, fitted with continuous linear and (centered) quadratic time variables as predictors. When the quadratic term was not significant (at p < 0.05), it was removed from the final model.
All five dimensions of questioning showed substantial historical trends toward greater journalistic aggressiveness over the course of the sampling period. Moreover, ordinal logistic regression models (summarized in Table 6) demonstrate that the trends are all statistically significant (p < 0.05). Initiative, assertiveness, and adversarialness show significant upward trends; for directness and accountability the upward trend is curvilinear, with each trend leveling off over time. All trends thus point in the same general direction, indicating a long-term decline in journalistic deference to the president and a rise in more aggressive forms of questioning.
However, as graphs of the trends demonstrate, this generalization glosses over some important issues concerning both the aggregate level of aggressiveness and the shape of the trendlines. Figures 1-5 show the percentage of questions per four-year term embodying each form of aggressiveness. The lines on each multiline graph are "stacked" or
cumulative, with each line showing the proportion of questions embodying either that amount of aggressiveness or lesser amounts.
[FIGURES 1-5 OMITTED]
Concerning the aggregate level of aggressiveness, it bears emphasis that most questions are not aggressive even during the most contentious of times. With the exception of directness, which characterizes a majority of questions, all other dimensions peak at much lower levels (initiative at 35 percent, assertiveness at 15 percent, adversarialness at 18 percent, and accountability at 6 percent). Thus, even when the press corps has been at its most vigorous, most of the questions put to the president are not aggressive.
Concerning the shape of the trendlines, the pattern for directness (Figures 2A and 2B) stands out as more gradual and unidirectional than all other dimensions. Over most of the sampling period, direct questions (Figure 2A) have monotonically increased, while all indirect forms (Figure 2B) have monotonically decreased. Thus, where journalists in the 1950s were exceedingly indirect in their questioning (often asking questions in the form "Would you care to tell us...," "Can I ask whether...," etc.), they have steadily become more straightforward in putting issues before the president. The only exceptions to this pattern are at the tail ends of the sampling period--questions became less direct rather than more so over the course of the Eisenhower and Clinton administrations, and they remained flat during the first Bush administration. But across the intervening administrations, each president received more direct questions than did his predecessor.
By contrast, the other dimensions--initiative (Figure 1), assertiveness (Figure 3), adversarialness (Figure 4), and accountability (Figure 5)--are more historically volatile, rising in a more concentrated manner in certain historical periods and falling in others. Moreover, these patterns of growth and decline are highly correlated across the four dimensions, such that four phases in the development of aggressive questioning may be distinguished.
The first phase spans the administrations of Eisenhower through Johnson (1953-1968). During this time, all four dimensions of aggressiveness remained at a relatively low level, albeit with some fluctuations.
The second phase spans Nixon through Reagan's first term (1969-1984). Over the course of this sixteen-year period, the dimensions of aggressiveness rose almost continuously. The only partial exception to this continuous upward trend occurred during the Carter administration--questions addressed to Carter were less likely to embody adversarialness, although his questions were more aggressive in most other ways.
The third phase begins with Reagan's second term and continues through Bush (1985-1992). During this time, aggressive questioning was generally on the decline, although it would not fall as far as pre-Nixon levels.
The fourth and final phase spans the two Clinton terms (1993-2000), during which time aggressiveness was again on the rise, and one dimension (adversarialness) grew to levels exceeding the previous peak in Reagan's first term.
The divergent historical trends documented here strongly suggest that different causal processes are at work for different forms of aggressiveness. Consider, first, the trend in directness. Because the growth of direct questioning has been gradual and more or less constant across eight administrations spanning three decades, increasing directness appears to be a deeply ingrained secular trend, one that is not particularly sensitive to discrete historical events or local sociopolitical conditions. Accordingly, this change represents a slow "tectonic" shift in the culture of the White House press corps and its relationship to the office of the presidency. Indeed, the steady and virtually inexorable character of this change suggests that it may not be a journalistic trend per se, so much as one manifestation of broader cultural changes involving the decline of formality in American life and the coarsening of public discourse (Ferris 2002; Tannen 1998; see also Maynard 2003, 55). This is consistent with the initial observation that directness differs from the other three dimensions of aggressiveness in being least tied to journalistic norms, and most tied to highly general interactional practices.
Regardless of how widespread and general this trend may be, that it is manifest in presidential news conferences may be consequential for perceptions of both the presidency and the press. As a fundamental property of interaction, directness has been theorized (by Brown and Levinson 1987) to be inversely associated with perceptions of the social distance separating interactants from one another, and directly associated with perceptions of the relative power of speakers versus hearers. Accordingly, its growth in news conferences may have a leveling effect, reducing perceptions of the social distance separating White House journalists from the president, and increasing perceptions of journalistic power vis-a-vis the president.
The other measures, which exhibit more volatile trends, suggest that these aspects of aggressiveness are more contextually sensitive. The sharp rise in aggressive questioning from Nixon through Reagan's first term may result from a series of historical events and conditions that prompted journalists to exercise their watchdog role much more vigorously from the late 1960s through the early 1980s.
The most proximate factor is declining journalistic trust in the president that followed in the wake of the Vietnam War and the Watergate affair (Broder 1987, 167-68; Cannon 1977, 289-92) and the more systematic news management practices initiated by the Nixon administration (Maltese 1994). The latter included holding White House news conferences much less frequently than previous presidents (Ragsdale 1998, 171). Lou Cannon of the Washington Post cites these events as having a transformative impact on how reporters view administration officials: "An attitude of basic trust that was tinged with skepticism was replaced with an attitude of suspicion in which trust occasionally intervened" (Cannon 1977, 291). Watergate, in particular, had a singular impact on American journalism. Although journalists played only a marginal role in the trajectory of that scandal (Epstein 1975), the idea that two young reporters were primarily responsible for "bringing down the President" has nonetheless become something of a professional myth among American journalists (Schudson 1992, Chapter 6). In the aftermath of Watergate, Bob Woodward and Carl Bernstein became celebrities and role models, and the long-dormant traditions of muckraking and investigative journalism were revived. Such journalism traditionally requires extensive behind-the-scenes research, but an atmosphere favorable to muckraking might also have encouraged members of the White House press corps, in the public forum of the news conference, to demonstrate their independence and professional skepticism by questioning the president aggressively. As David Broder (1987, 167) has observed, even meetings with the president's press secretary were affected: "The style of questioning at White House briefings became, after Watergate, almost more prosecutorial than inquisitive."
A second factor is the decline of political consensus that characterized this period. The events of 1968--in particular the Tet offensive and Johnson's subsequent decision not to seek a second term--stimulated substantial elite and public opposition to the war (Hallin 1986, 167-74). Correspondingly, Nixon's election launched an extended period of divided government, with different parties controlling the presidency and Congress. It has been demonstrated that journalists are sensitive to societal conflict and dissent, both within Congress (Bennett 1990; Hallin 1984) and in the public at large (Entman 2003). Given that such conditions tend to yield more independent and adversarial news stories, they might also influence how journalists conduct themselves when asking questions of political leaders.
Perhaps the least obvious factor, but one that is also potentially relevant, has to do with practical economic conditions. Nixon and Reagan span a period of time when the long post-World War II economic expansion came to an end. It may be that high levels of both unemployment and inflation soured the public mood and, perhaps, the public's journalistic surrogates in the White House press corps.
The subsequent decline of aggressive questioning from Reagan's second term through Bush may be due to a countervailing set of factors. Economic conditions steadily improved following the recession of the early 1980s. Reagan's persistent popularity after that point, his landslide reelection, and the fact that he weathered the Iran-Contra scandal may have suggested to White House reporters the limitations of the Watergate model of adversarial journalism. Moreover, during this period, journalism came under increasing criticism for being excessively negative and overly concerned with strategy and scandal, and for fostering public apathy and cynicism. This would in turn stimulate a reform movement within journalism, the so-called civic journalism or public journalism movement spearheaded by Jay Rosen and Davis Merritt. The latter development did not get off the ground until the middle of Bush's term in office--indeed, the movement was triggered in part by what some journalists regarded as poor coverage of the 1988 Bush-Dukakis presidential campaign (Fallows 1996, 247-54). Nevertheless, it could have further contributed to trends already in progress, trends that show journalists to be reining in their aggressiveness during this period.
Teasing out these explanatory factors will have to await further analysis. In the meantime, these historical trends are themselves illuminating, helping to resolve certain persistent controversies regarding White House journalism. One such controversy concerns the historical trajectory of president-press relations. Some have painted a broadly unidirectional picture of increasing aggressiveness by the White House press corps (Hallin 1992; Patterson 1993, 2000) and by Washington journalists more generally (Rozell 1994; Robinson 1981). Others suggest that journalistic aggressiveness peaked under Nixon, after which relations grew more harmonious (Smith 1990). The present study suggests that, at least in the news conference environment, both views underestimate the dynamism of president-press relations. The Nixon era marks the beginning of an extended period of increasingly vigorous questioning, rather than the pinnacle as some have suggested. Nixon's immediate successors were for the most part treated more aggressively than was Nixon himself. Furthermore, for most dimensions of aggressiveness, the trend since Nixon has been neither linear nor unidirectional. After an extended period of rising aggressiveness, aggressiveness declined from Reagan's second term through Bush, and then rose again over the course of the Clinton presidency.
Another controversy concerns the significance of partisan bias in journalism. While most academic studies have failed to find evidence for systematic partisanship (D'Alessio and Allen 2000; Niven 2002), numerous popular commentators continue to propose that the news media systematically favor either the Republican or more often the Democratic side of the American political spectrum. The data presented here, while not controlling for other variables, offer little support for arguments about partisanship in either direction. Although it is true that the more volatile dimensions steadily rose during mostly Republican administrations (Nixon through first-term Reagan), the subsequent period of decline also occurred under Republicans (second-term Reagan through Bush), and the more recent increase occurred when a Democrat (Clinton) occupied the White House. The volatile dimensions thus do not correlate with the president's party affiliation. Correspondingly, the trend for directness is even more transparently nonpartisan. Questions have steadily become less cautious and more blunt from Eisenhower through Reagan, a period encompassing four Republican and three Democratic presidents. Plainly these long-term historical trends overwhelm president-level characteristics such as party affiliation as predictors of journalistic aggressiveness.
Finally, this study has demonstrated that journalistic aggressiveness is indeed measurable in the context of direct encounters with presidents and other public figures. The question analysis system developed here represents a viable new method for evaluating the tenor of president-press relationships and for tracking the evolving journalistic environment in which presidents operate. This, in turn, is a prerequisite to any systematic investigation of either the causes or consequences of a more vigorous White House press corps.
AUTHORS' NOTE: This research was supported by grants from the National Science Foundation (grant #SES-0112221) and the LeRoy Neiman Center for the Study of American Culture at UCLA. Additional support was provided by an outstanding team of research assistants: Hanna Abner, Tamara Black, Carleen Curley, Amira Day, Hedy Javahery, Marian Katz, Wendy Klein, Min Lee, Seung-Hee Lee, Parisa Leviadin, Lisa McConnell, Shawna Miller, Julie Peggar, Danielle Pillet-Shore, Shannon Seibert, Sachiko Takita, and Nancy Yuen. We are grateful for feedback given by the anonymous reviewers and by colleagues who heard various components of the paper presented in 2004 at conferences of the American Political Science Association, the American Sociological Association, and the National Communication Association.
Alexander, Jeffrey. 1989. Culture and political crisis: "Watergate" and Durkheimian sociology. In J. Alexander, Structure and meaning: Relinking classical sociology, 174-216. New York: Columbia University Press.
Bakeman, R., V. Quera, D. McArthur, and B. E Robinson. 1997. Detecting sequential patterns and determining their reliability with fallible observers. Psychological Methods 2: 357-70.
Baum, Matthew A., and Samuel Kernell. 1999. Has cable ended the golden age of presidential television? American Political Science Review 93:99-114.
Bennett, W. Lance. 1990. Toward a theory of press-state relations in the United States. Journal of Communication 40: 103-25.
Blum-Kulka, Shoshana. 1987. Indirectness and politeness in requests: Same or different? Journal of Pragmatics 11: 131-46.
Boden, D., and D. H. Zimmerman, eds. 1991. Talk and social structure. Cambridge, UK: Polity Press.
Broder, David S. 1987. Behind the frontpage: A candid look at how news is made. New York: Simon and Schuster.
Brown, Penelope, and Stephen Levinson. 1987. Politeness: Some universals in language usage. Cambridge, UK: Cambridge University Press.
Cannon, Lou. 1977. Reporting: An inside view. Sacramento: California Journal Press.
Clark, Herbert H., and Dale H. Schunk. 1980. Polite responses to polite requests. Cognition 8:111-43.
Clayman, Steven E. 1988. Displaying neutrality in television news interviews. Social Problems 35: 474-92.
--. 1992. Footing in the achievement of neutrality: The case of news interview discourse. In Talk at work, edited by E Drew and J. Heritage, 163-98. Cambridge, UK: Cambridge University Press.
--. 2001. Answers and evasions. Language in Society 30: 403-42. Clayman, Steven, and John Heritage. 2002a. The news interview: Journalists and public figures on the air. Cambridge, UK: Cambridge University Press.
--. 2002b. Questioning presidents: Journalistic deference and adversarialness in the press conferences of U.S. Presidents Eisenhower and Reagan. Journal of Communication 52:749-75.
Cohen, Jeffrey E. 2004. If the news is so bad, why are presidential poll numbers so high? Presidents, the news media, and the mass public in an era of new media. Presidential Studies Quarterly 34: 493- 515.
--. 2005. Presidential going public in an age of new media. Paper presented at the American Political Science Association annual meeting, September, Washington DC.
Cornwell, Elmer. 1960. The presidential press conference: A study in institutionalization. Midwest Journal of Political Science 4: 370-89.
--. 1965. Presidential leadership of public opinion. Bloomington: Indiana University Press. D'Alessio, Dave, and Mike Allen. 2000. Media bias in presidential elections: A meta-analysis. Journal of Communication 50: 133-56.
Drew, Paul, and John Heritage. 1992. Analyzing talk at work: An introduction. In Talk at work, edited by P. Drew and J. Heritage, 3-65. Cambridge, UK: Cambridge University Press.
Durkheim, Emile. 1915. The elementary forms of the religious life. New York: Free Press.
Entman, Robert M. 2003. Projections of power: Framing news, public opinion, and U.S. foreign policy. Chicago: University of Chicago Press.
Epstein, Edward. 1975. Did the press uncover Watergate? In Between fact and fiction: The problem of journalism, edited by E. Epstein, 19-33. New York: Vintage.
Fallows, James. 1996. Breaking the news. New York: Vintage.
Ferris, Abbott L. 2002. Studying and measuring civility: A framework, trends, and scale. Sociological Inquiry 72: 376-92.
French, Blaire Atherton. 1982. The presidential press conference: Its history and role in the American political system. Washington, DC: University Press of America.
Goffman, Erving. 1967. Interaction ritual: Essays in face to face behavior. Garden City, NY: Doubleday.
Grossman, Michael Baruch, and Martha Kumar. 1981. Portraying the president: The White House and the news media. Baltimore: Johns Hopkins University Press.
Hager, Gregory L., and Terry Sullivan. 1994. President-centered and presidency-centered explanations of presidential public activity. American Journal of Political Science 38: 1079-103.
Hallin, Daniel C. 1984. The media, the war in Vietnam, and political support: A critique of the thesis of an oppositional media. Journal of Politics 46: 2-24.
--. 1986. The "uncensored war": The media and Vietnam. New York: Oxford University Press.
--. 1992. Sound bite news. Journal of Communication 42: 5-24.
Hart, Roderick P., Deborah Smith-Howell, and John Llewellyn. 1990. Evolution of presidential news coverage. Political Communication and Persuasion 7: 213-30.
Heritage, John. 1984. Garfinkel and ethnomethodology. Cambridge, UK: Polity Press.
--. 2002a. Designing questions and setting agendas in the news interview. In Unearthing the taken-for-granted: Studies in language and social interaction, edited by J. Mandelbaum, P. Glenn, and C. LeBaron, 57-90. Mahwah, NJ: Erlbaum.
--. 2002b. The limits of questioning: Negative interrogatives and hostile question content. Journal of Pragmatics 34: 1427-46.
Heritage, John, and David Greatbatch. 1991. On the institutional character of institutional talk: The case of news interviews. In Talk and social structure, edited by D. Boden and D. H. Zimmerman, 93-137. Berkeley: University of California Press.
Heritage, John, and Andrew Roth. 1995. Grammar and institution: Questions and questioning in the broadcast news interview. Research on Language and Social Interaction 28: 1-60.
Holtgraves, Thomas. 1992. The linguistic realization of face management: Implications for language production and comprehension, person perception, and cross-cultural communication. Social Psychology Quarterly 55: 141-59.
Juergens, George. 1981. News from the White House: The presidential-press relationship in the progressive era. Chicago: University of Chicago Press.
Kernell, Samuel. 1986. Going public: New strategies of presidential leadership. Washington, DC: Congressional Quarterly.
Krippendorff, Klaus. 1980. Content analysis: An introduction to its methodology. Newbury Park, CA: Sage.
Kumar, Martha Joynt. 2003. Does this constitute a press conference? Defining and tabulating modern presidential press conferences. Presidential Studies Quarterly 33: 221-37.
--. 2005. Presidential news conferences: The importance and evolution of an enduring forum. Presidential Studies Quarterly 35: 166-92.
Lammers, William W. 1981. Presidential press-conference schedules: Who hides, and when? Political Science Quarterly 96: 261-78.
Macaulay, Marcia. 1996. Asking to ask: The strategic function of indirect requests for information in news interviews. Pragmatics 6: 491-509.
Maltese, John Anthony. 1994. Spin control: The White House Office of Communications and the management of presidential news, 2d ed. Chapel Hill: University of North Carolina Press.
Manheim, Jarol B. 1979. The honeymoon's over: The news conference and the development of presidential style. Journal of Politics 41: 55-74.
Manheim, Jarol B., and William W. Lammers. 1981. The news conference and presidential leadership of public opinion: Does the tail wag the dog? Presidential Studies Quarterly 11: 177-88.
Maynard, Douglas W. 2003. Bad news, good news: Conversational order in everyday talk and clinical settings. Chicago: University of Chicago Press.
McGuire, Delbert. 1967. Democracy's confrontation: The presidential press conference I. Journalism Quarterly 44: 638-44.
Niven, David. 2002. Tilt? The search for news bias. Westport, CT: Praeger.
Patterson, Thomas E. 1993. Out of order. New York: Vintage.
--. 2000. Doing well and doing good: How soft news and critical journalism are shrinking the news audience and weakening democracy--And what news outlets can do about it. Joan Shorenstein Center for Press, Politics, and Public Policy, John E Kennedy School of Government, Harvard University.
Pollard, James E. 1947. The presidents and the press. New York: MacMillan.
Pomerantz, Anita. 1988. Offering a candidate answer: An information seeking strategy. Communication Monographs 55: 360-73.
Ponder, Stephen. 1998. Managing the press: Origins of the media presidency 1897-1933. New York: St. Martin's Press.
Ragsdale, Lyn. 1997. Disconnected politics: Public opinion and presidents. In Understanding public opinion, edited by B. Norrander and C. Wilcox, 229-51. Washington, DC: Congressional Quarterly Press.
--. 1998. Vital statistics on the presidency, rev. ed. Washington, DC: Congressional Quarterly Press. Raymond, Geoffrey. 2003. Grammar and social organization: Yes/no interrogatives and the structure of responding. American Sociological Review 68: 939-67.
Robinson, Michael. 1976. Public affairs television and the growth of political malaise. American Political Science Review 70:409-32.
--. 1981. Three faces of congressional media. In The new Congress, edited by Thomas E. Mann and Norman J. Ornstein, 55-96. Washington, DC: American Enterprise Institute.
Roth, Andrew. 1998. Who makes news: Descriptions of television news interviewers' public personae. Media, Culture & Society 20: 79-107.
--. 2005. "Pop quizzes" on the campaign trail: Journalists, candidates, and the limits of questioning. Harvard International Journal of Press/Politics 10: 28-46.
Rozell, Mark J. 1994. Press coverage of Congress 1946-1992. In Congress, the press, and the public, edited by Thomas E. Mann and Norman J. Ornstein, 59-129. Washington, DC: American Enterprise Institute and Brookings Institution.
Sabato, Larry J. 1991. Feeding frenzy: How attack journalism has transformed American politics. New York: Free Press.
Schudson, Michael. 1992. Watergate in American memory. New York: Basic Books. --. 1995. Watergate and the press. In The power of news, edited by M. Schudson, Chapter 7. Cambridge, MA: Harvard University Press.
Schwartz, Barry. 1987. George Washington: The making of an American symbol. New York: Free Press.
Shils, Edward. 1975. Center and periphery: Essays in macrosociology. Chicago: University of Chicago Press.
Smith, Carolyn. 1990. Presidential press conferences: A critical approach. New York: Praeger.
Smoller, Frederic T. 1990. The six o'clock presidency. New York: Praeger.
Tannen, Deborah. 1998. The argument culture. New York: Random House.
Tebbel, John, and Watts, Sarah Miles. 1985. The press and the presidency. Oxford, UK: Oxford University Press.
Ten Have, Paul. 1999. Doing conversation analysis. London: Sage.
Van der Wijst, Per. 1995. The perception of politeness in Dutch and French indirect requests. Text 15: 477-501.
Wattenberg, Martin P. 2004. The changing presidential media environment. Presidential Studies Quarterly 34: 557-72.
(1.) For an introductory overview of conversation analysis, see Heritage (1984, Chapter 8). For a thorough discussion of conversation analytic methods, see Ten Have (1999). For discussions of the use of conversation analysis to study institutional forms of talk, see Boden and Zimmerman (1991) and Drew and Heritage (1992).
(2.) For a sampling of research relevant to the present study of question design, see Clayman (1988, 1992, 2001), Heritage (2002a, 2002b), Heritage and Greatbatch (1991), Heritage and Roth (1995), Macaulay (1996), Roth (1998, 2005), and Clayman and Heritage (2002a, Chapter 6; 2002b).
(3.) For a more detailed discussion of question frames and the levels of deference that they embody, see Clayman and Heritage (2002b).
(4.) For research on question design in news interviews, see the references cited in note 2. For relevant research bearing on question design in interaction generally, see Pomerantz (1988), Raymond (2003), and the extensive line of research concerning conventional indirectness (e.g., Blum-Kulka 1987; Brown and Levinson 1987; Clark and Schunk 1980; Van der Wijst 1995). For further arguments pertaining to the validity of the present coding system, see Clayman and Heritage (2002b).
STEVEN E. CLAYMAN
University of California, Los Angeles
MARC N. ELLIOTT
University of California, Los Angeles
LAURIE L. MCDONALD
Steven E. Clayman is professor of sociology and communication studies at UCLA, and is coauthor (with John Heritage) of The News Interview: Journalists and Public Figures on the Air.
Marc N. Elliott is senior statistician for the RAND Corporation.
John Heritage is professor of sociology at UCLA.
Laurie L. McDonald is senior statistical research programmer for the RAND Corporation.
TABLE 1 The Question Analysis System Dimension Indicator Description Initiative Statement prefaces Q preceded by statement(s) Multiple questions 2+ Qs in a single turn at talk Follow-up questions Subsequent Q by the same journalist Directness Absence of other- Frame refers to referencing frames president's ability or willingness to answer Absence of self- Frame refers to referencing frames journalist's own intention or desire to ask Assertiveness Preface tilt Preface favors either yes or no Negative questions Isn't it ...? Couldn't you ...? Adversarialness Preface adversarialness Q preface is oppositional Global adversarialness Overall Q is oppositional Accountability questions Q seeks explanation for administration policy Dimension Values Kappa Initiative 0 No preface 0.93 0.93 1 Preface 0 Single Q 0.99 1 Multiple Qs 0 Not a follow-up Q 0.71 1 Follow-up Q Directness 0 No frame 0.88 0.87 1 Can you/Could you ... 2 Will you/Would you ... 0 No frame 0.91 1 I wonder 2 I'd like to ask ... 3 Can/May I ask... Assertiveness 0 No tilt 0.67 0.80 1 Innocuous tilt 2 Unfavorable tilt 0 Not a negative Q 0.94 1 Negative Q Adversarialness 0 Nonadversarial preface 0.79 0.78 1 Oppositional preface focus of Q 2 Oppositional preface presupposed 0 Not oppositional 0.66 overall 1 Oppositional overall 0 Not an 0.76 accountability Q 1 Why did you ... 2 How could you ... TABLE 2 Initiative Initiative Statement Multiple Follow-Up N Preface Qs 0 0 0 0 869 0 1 0 535 0 0 1 550 1 0 0 1435 1 0 1 1 41 1 1 0 986 1 0 1 156 1 1 1 36 Initiative % Derived N Derived 0 19 3389 74 12 12 31 1 1 1219 26 21 3 1 TABLE 3 Directness Directness Other-Referencing Self-Referencing N 0 2 3 2 0 3 30 2 2 5 2 1 45 2 0 134 1 1 2 2 0 2 65 1 1 116 2 0 1 235 1 0 480 3 0 0 3494 Directness % Derived N Derived 0 0 216 5 1 0 1 3 1 0 183 4 1 3 2 5 715 15 10 3 76 3494 76 TABLE 4 Assertiveness Assertiveness Preface Negative N Tilt Questions 0 0 0 1806 1 0 376 1 0 1 101 1 1 25 2 0 165 2 2 1 46 Assertiveness % Derived N Derived 0 72 2182 87 15 1 4 291 11 0 7 2 2 46 2 TABLE 5 Adversarialness Adversarialness Preface Global N Adversarial Adversarial 0 0 0 4077 1 1 0 186 2 2 0 82 3 0 1 135 1 1 20 4 2 1 108 Adversarialness % Derived N Derived 0 88 4077 88 1 4 186 4 2 2 82 2 3 3 155 3 0 4 2 108 2 TABLE 6 Results Outcome Time Coefficient p Value Initiative 0.140 <0.001 Directness 0.194 <0.001 Assertiveness 0.030 0.021 Adversarialness 0.125 <0.001 Accountability 0.167 <0.001 Outcome Time Squared, Centered p Value Coefficient Initiative Directness -0.020 <0.001 Assertiveness Adversarialness Accountability -0.028 <0.001
|Printer friendly Cite/link Email Feedback|
|Author:||Clayman, Steven E.; Elliott, Marc N.; Heritage, John; McDonald, Laurie L.|
|Publication:||Presidential Studies Quarterly|
|Date:||Dec 1, 2006|
|Previous Article:||When Trumpets Call: Theodore Roosevelt after the White House.|
|Next Article:||Polls, elite opinion, and the president: how information and issue saliency affect approval.|