Presidential debate stories accentuate the negative

Presidential debate stories accentuate the negative

Reber, Bryan H

Primary campaigns, generally, and primary debates, in particular, merit scholarly attention. Primary candidates narrow the field of contenders, presenting voters with a choice in November. This process can be particularly important when one party has a weak incumbent. In 1976 Jimmy Carter might not have been the only Democrat who could have defeated Gerald Ford. Four years later, other Republicans besides Ronald Reagan might have been capable of ousting Carter. In 1992, Bill Clinton may not have been the only Democrat who could have beaten George Bush. Thus, in some years the primaries may in effect decide the outcome of the general election by determining who gets to face and defeat the weak opponent.

Second, primary debates are more common than general debates. In 1948, Harold E. Stassen and Thomas E. Dewey participated in the first presidential (primary) debate, broadcast over radio.’ John F. Kennedy practiced for the Nixon-Kennedy debates in 1960 by debating Hubert H. Humphrey before the West Virginia primary. Scholarship indicates that every campaign from 1948 to 2000, with the sole exception of 1964, included at least one primary debate.’ Furthermore, in recent campaigns, primary debates have outnumbered general election debates. In 2000, three general election debates were held in the fall, while twenty-two primary debates informed viewers of the relative merits of the candidates for each party’s nomination. The fact that there are so many debates means that there are many more opportunities to influence voters, which may help compensate for the fact that fewer people watch each primary debate. Furthermore, primary debate performance can help candidates attract media attention.

Another reason primary messages are important is that candidates are less well-known in the primary than in the general campaign. For example, in 2000, the two nominees who competed in the general election, Al Gore and George W. Bush, were better known than Orrin Hatch or Alan Keyes. However, even Bush and Gore became better-known during the general election than they were during the primary campaign. Thus, debates have a greater opportunity for influence in the primary than in the general campaign.’

Of course, debates in the primary campaign do not attract the same number of viewers as general election debates. The electorate, in general, often does not get interested until the candidate field is narrowed. Jamieson and Birdsell noted that in November of 1987 only 15 percent of those surveyed said they were paying very close attention to news reports about Democratic presidential candidates, and 13 percent said they were paying close attention to the Republican race.’ Because relatively few voters watch individual primary debates, the media may have more impact on voters. That is, if voters don’t watch the debate themselves, they are dependent on the media to learn about the debates. Research shows that media use contributes to campaign knowledge.’

However, the media have a strong proclivity for focusing on the “horse race” aspects of campaigns: who is ahead, what states are being contested, who is campaigning where. Patterson explained, “in its coverage of a presidential campaign, the press concentrates on the strategic game played by the candidates in their pursuit of the presidency, thereby de-emphasizing the questions of national policy and leadership.” In fact, his investigation of the 1976 campaign concluded that “the election’s substance… received only half as much coverage as was accorded the game.”‘ Voters can learn about candidates from the news, but not as much as might be supposed.

Gans, also focusing on the 1976 election, did consider coverage of debates: “The debates became news and their coverage followed the daily campaign format, the news media paid major attention to candidate mistakes, and like the pollsters, treated the debates as contests, and thus part of the larger horse race.”7 More recent studies of the 1984 and 19888,19929 and 199610 presidential elections have found that, while horse race coverage may not be the majority of coverage, it is substantial. Stempel and Windhauser found that a majority of the newspaper coverage they analyzed was devoted to the horse race. Mantler and Whiteman found that 41.4 percent of coverage consisted of campaign issues rather than policy or character issues.” Domke et al. found that “concerns about the horse race aspects of the campaign were found in the most paragraphs…29.4 percent.””2 But we don’t know exactly what is contained in news stories about debates. Can voters learn more about the candidates themselves and their policy stands from coverage of particular campaign events/messages?

Research has indicated that news coverage of campaign messages is less than optimal. Sigelman and Bullock found that issue coverage in five national newspapers dropped from 1968 to 1988 (such coverage had increased from 1888 until either 1948 or 1968, depending upon the paper).” In their study of the 1992 presidential campaign, Lichter and Noyes found that “compared with news media reports, the candidates’ daily speeches discussed far more issues, and dealt with them far more extensively, and with considerably more detail and context.” They reported that “the candidates also ran more positive campaigns than voters might have guessed, since news reports consistently emphasized their most negative rhetoric.”14 Stempel and Windhauser found that negative coverage of Republican and Democratic presidential and vice presidential candidates ranged from 25 percent to 41 percent during the 1984 and 1988 elections.” Domke et al. found that although negative coverage of Bill Clinton and Bob Dole in 1996 was not in the majority, it was substantial — 45.5 percent. 16 Thus, while citizens who have a relatively low interest in primary debates may have to rely on news reports, the question of whether they can (or should) rely on the media remains unanswered.

With this understanding of general campaign knowledge, it is important to know what messages are being conveyed during a primary campaign and what knowledge is likely being built in voters. Appallingly little research has been done on the relationship of newspaper coverage to presidential debates, either in the general or primary election. 17 Pfau and Eveland urge scholars to examine debates within the broader political environment including newspaper coverage.” Thus, this study will examine two primary debates and newspaper coverage of those campaign events. Then it will compare the debates with the coverage to determine to what extent newspaper coverage reflected the actual content of the debates.

Purpose and Hypotheses

First, we will analyze the debates themselves to determine what the newspapers could have been covering. Analysis was conducted using the functional theory, which has been applied to the 1960,1988 and 1996 presidential debates.19 Benoit and Harthcock’s analysis of the Nixon-Kennedy debates” found that positive comments (acclaims) were most common at 49 percent of the utterances, attacks constituted 39 percent of the remarks and defenses were 12 percent. Benoit and Brazeal2l reported that in the Bush-Dukakis debates, acclaims (positive remarks) were 59 percent of the utterances, attacks were 33 percent and defenses comprised 8 percent of the utterances. Benoit, Blaney and Pier’ studied the Clinton-Dole debates, indicating that 59 percent of their comments were acclaims, 33 percent were attacks and 7 percent were defenses. Thus, in each of the presidential debates, positive remarks outnumbered negative comments, with defenses the least common utterance form. Benoit and Harthcock 23 reported that in the Nixon-Kennedy debates 78 percent of their comments addressed policy matters, whereas 22 percent concerned character. Benoit and Brazeal” found that in the Bush and Dukakis campaign 66 percent of utterances were about policy and 34 percent were about character. Benoit, Blaney and Pier 21 found that Clinton and Dole devoted 72 percent of their remarks to policy and 28 percent to character. Thus, these presidential debates consistently emphasized policy over character.

Based on this previous research into presidential primary debates,26 we make these predictions about the functions and topics of primary debates (the first two hypotheses will be tested separately for each debate):

H1:

Candidates in primary debates will devote more utterances to acclaims than attacks (or defenses).

H2:

Candidates in primary debates will focus more on policy (issues) than character (personality).

Chaffee, Zhao and Leshner found that newspaper reading is a significant predictor among voters of party issue knowledge (the general differences between partisan groups), candidate issue knowledge (a specific candidate’s policy positions) and candidate personal knowledge (a specific candidate’s biography).27 Similarly, Lemert found that newspaper reading was a significant predictor of campaign knowledge and that this predictor was strengthened by exposure to a presidential debate.28 Lowden, Andersen, Dozier and Lauzen found a significant positive correlation between newspaper use and issue knowledge.29 These considerations led us to focus on newspaper (rather than television) coverage of debates.

Hart observes that “political news is reliably negative.”” Morello analyzed newspaper editorials that followed presidential election debates. He discovered that 26 of 55 editorials analyzed suggested that the principal criterion for whether a debate has been successful is whether it has established how the candidates differ. He noted these newspaper editorials judged debate outcomes in terms of unexpected behaviors, personal style and debate performance. Morello’s thesis is that while newspapers plea for substance from debates, their editorial pages look mainly for ways to continue the easiest winner/loser analysis of the campaign. 31

Shoemaker and Reese note how the widely held news values of professional journalists affect newsroom decisions.32 The ubiquity of these standards among journalists provide a uniformity to many content decisions.

Conflict and controversy lend news value to issues and events. “Conflict is inherently more interesting than harmony,” write Shoemaker and Reese.”

Conflict is inherent to the political process and perhaps to primary and general election debates.

News values are audience-centered, according to Shoemaker and Reese, and editors must determine in assessing news value what is most desirable to the audience.34

Sears and Chaffee suggest that in 1976, “The debates themselves were heavily issue-oriented, but the subsequent coverage of them decidedly less so.”35 Lichter and Noyes, looking at speeches rather than debates, found that the news media devoted more attention to character than did campaign speeches.

The final set of hypothesis is based on these observations that actual message content has been found to differ dramatically from journalistic coverage of that content and that journalistic tenets value conflict over harmony. Therefore, we make the following predictions about coverage of these debates (again, these two hypotheses will be tested separately for coverage of each debate):

H3:

Newspaper coverage of primary debates will focus more on attacks than acclaims (or defenses).

H4:

Newspaper coverage of primary debates will focus more on policy (issues) than character (personality).

Next, we will compare the debates with newspaper coverage of those debates to provide an indication of the accuracy of coverage. Because campaign coverage has been found to be substantially negative?31 we predict that:

H5:

The number of attacks covered in newspaper stories about primary debates will be higher than the number of attacks in the debates; the number of acclaims in newspapers will be lower than the actual number of acclaims.

Finally, given the importance of people or personalities in coverage, our final hypothesis predicts that:

H6:

The number of character comments in newspaper stories about primary debates will be higher than the number of character utterances in the debates; the number of policy remarks in newspapers will be lower than the actual number of policy comments.

We will test each set of hypotheses on two debates (one Democratic, one Republican) from the 2000 primary campaign. Together, these hypotheses will provide insight in the nature of the debates, the nature of newspaper coverage and the relationship of the two in coverage of particular campaign events.

Method

Benoit, Blaney and Pier argue that voting is an inherently comparative process.-18 The functional theory of political campaign discourse is based, in part, on this notion that candidates must differentiate among themselves for the electorate. Therefore, they argue that “all political campaign discourse has three potential functions,” acclaims (self-praise), attacks (negative comments or criticism) or defenses (refutations of attacks)?31 Acclaims give voters reasons to vote for the candidate, attacks give reasons to vote against the opponent and defenses respond to attacks from an opponent. Together, they serve as an informal form of cost-benefit analysis: acclaims stress a candidate’s own benefits, attacks identify opponents’ costs and defenses attempt to refute alleged costs.

The functional theory of political campaign discourse is ideally suited as a method for this study. It employs content analysis of political messages, including primary debates, which will allow quantitative comparisons of the debates and newspaper coverage of those debates. The major categories of this theory functions (acclaims, attacks, defenses) and topic (policy, character) also permit a clear operationalization of the variables in this analysis.

The analytic procedure used in this study consisted of three steps. First, the statements in the debates and stories were unitized into themes or utterances that address a coherent idea (in our discussion, we use the terms “utterances,” “comments,” and “remarks” synonymously with “themes”).” Because discourse is inherently enthymematic (arguments frequently contain suppressed premises that viewers must supply), themes vary in length from a phrase to several sentences. Second, each theme was classified as an acclaim, attack, or defense:

Themes that portray the candidate in a favorable light are acclaims. Themes that portray the opposing candidate in an unfavorable light are attacks.

Themes that attempt to repair the candidate’s reputation (from attacks by the opposition) are defenses.

Third, a judgment was made about whether the theme concerned policy or character:

Themes that concern governmental action (past, current or future) and problems amenable to governmental action are policy utterances.

Themes that address characteristics, traits, abilities or attributes of the candidates are character utterances.

Research on political messages often analyzes this discourse using two dimensions: functions (positive and negative or attack ads) and topics (issues or policy and image or character).

Sample

Democratic and Republican debates held in New Hampshire on January 5 and 6, 2000 were selected for analysis. This selection was based on the importance of New Hampshire’s first-in-the-nation primary, on the fact that frontrunner status was not completely established and because the debates fell backto-back and therefore might have taken on additional comparative importance within the media. Debate transcripts were located via online computer search using Lexis-Nexis Academic Universe.

Newspaper articles were identified via online computer search using LexisNexis Academic Universe. The search was conducted through the news category of Lexis-Nexis and by searching articles that contained candidates’ names and the words “debate” and “New Hampshire” (i.e., Gore + Bradley + debate + New Hampshire). The search was limited to northeastern newspapers as preselected by Lexis-Nexis and to the week of the debates, Sunday, January 2, to Saturday, January 8,2000. The purpose of selecting northeastern newspapers was to focus on those newspapers that would most directly impact prospective primary voters who may or may not have watched the New Hampshire debates. This mode of identifying a sample of articles yielded 48 articles that met the criteria for the Republican debate and 56 articles that met the criteria for the Democratic debate. News articles were then visually scanned for their appropriateness to be included in the final pool. Only articles that focused on the debate – in which the debate was the subject of more than half the article – were included. The final sample of articles consisted of those printed on January 5, 6 and 7 for the Democratic debate and on January 6,7 and 8 for the Republican debate. Twenty articles were coded for the Republican debate; 31 for the Democratic debate. Print stories came from the Associated Press, Bergen County Record, The Boston Globe, Boston Herald, New York Daily News, New York Post, The New York Times, Providence journal, The Patriot Ledger, The Union Leader and Telegram & Gazette.

Cohen’s kappa, which controls for agreement by chance, was used to calculate intercoder reliability on 20 percent of each of the two debates, 20 percent of the articles on the Republican debate and 20 percent of the articles on the Democratic debate.41 For the debates, the intercoder reliability was 1.0 for functions and .94 for topics. For the newspaper articles, intercoder reliability was 1.0 for functions and .95 for topics. Fleiss explains that “values [of kappa] greater than .75 or so may be taken to represent excellent agreement beyond chance, values below .40 or so may be taken to represent poor agreement beyond chance and values between .40 and .75 may be taken to represent fair to good agreement beyond chance.”‘ Thus, these values indicate excellent reliability for our analysis.

Results

New Hampshire Democratic Debate

The Democratic debate occurred on January 5,2000, in Durham, N. H., and included the two candidates in this party, Bill Bradley and Al Gore. The first hypothesis was confirmed: there were more acclaims (58 percent) than attacks (23 percent) or defenses (19 percent) (X2[df=21=22.9, p

For example, Bradley acclaimed his stance on abortion: “I am for freedom of choice for women.” This statement functioned to increase his attractiveness to pro-choice adherents. Gore attacked Bradley’s proposals as inadequate: “The plan that he put out… doesn’t save a penny for Medicare. It does not contain a comprehensive education reform agenda.” Bradley used defense when confronted with his acceptance of campaign donations from the pharmaceutical industry: “Less than one percent of the money that I ever raised when I was running in all my Senate campaigns came from anybody connected to a pharmaceutical company. Less than three-tenths of a percent in my presidential campaign. So, from my standpoint, that’s not a problem.” These excerpts illustrate how the candidates enacted the three functions in this debate.

The second hypothesis was also confirmed in the Democratic debate: Significantly more utterances addressed policy (72 percent) than character (28 percent) (X2[df=11=35.8, p

For example, Bradley argued that “When it comes to gun control it’s common sense that we have a registration and licensing system for guns, just like we have for automobiles.” Gun control is obviously a question of governmental policy. Gore discussed leadership ability, a character trait: “We both have leadership. We both have experience. The question is: Who has the leadership to get these big things done for the country?” Clearly the answer to his rhetorical question is that he is better qualified than Bradley to lead our country. Thus, both candidates discussed policy and character issues, with more emphasis on the former.

New Hampshire Republican Debate

The Republican debate was January 6, 2000, in Manchester, N. H. Debate participants were candidates Gary Bauer, George W. Bush, Steve Forbes, Orrin Hatch, Alan Keyes and John McCain. The first hypothesis was confirmed: candidates devoted most of their remarks to acclaims (57 percent), then to attacks (24 percent) and least to defenses (15 percent) (X2[df=21=89.1, P

To illustrate their use of acclaims, Bush often acclaimed past deeds as Governor of Texas: “As Governor of Texas, I fought for and signed the two largest tax cuts in my state’s history. I fought for charter schools and public school choice in our public schools. I reformed welfare by insisting on work. I fought for tort reform. I’ve got a record of accomplishment.” McCain attacked Bush on the issue of campaign finance: “I don’t think you have an idea of how important campaign finance reform is to restore the confidence of young Americans in their government.” When McCain charged that Bush’s plan gives the entire surplus back to the taxpayers, Bush quite simply denied this accusation: “No it doesn’t.” Thus, the three functions were all present in this debate, although not in equal proportions.

The second hypothesis was again confirmed: Republicans devoted significantly more of their remarks to policy (64 percent) than character (36 percent) (X2[df=2]=18.4, p

Bauer attacked Bush on several policy issues at once: “You’ve rejected fundamental tax reform. You won’t agree to a pro-life running mate. You won’t agree to put pro-life judges on the court. And your China policy, just like Clinton’s, puts trade ahead of national security and human rights.” Forbes stressed his integrity, a character trait: “I am independent. The special interests and lobbyists have no hooks into me.” This aspect of his character could make him seem more attractive to voters. Similarly, McCain explained that he believed one should “get the truth. Get honesty. Get honor.” Thus, these candidates discussed both policy and character, focusing more on policy. News Coverage of the Debates

As predicted in the third hypothesis, newspaper coverage did not devote an equal amount of space to each function. In the Democratic debate articles, attacks (45 percent) were discussed more frequently than acclaims (40 percent) and defenses the least (14 percent) (X2[df=21=47.1, p

The fifth hypothesis was supported: Newspaper coverage of the Democratic debate was significantly more negative than the debate itself (X2[df=21=22.9, p

This hypothesis was also supported for the Republican debate LZ[df=21=19.2, p

The final hypothesis was not supported for the Democratic debates (X?[df=21=0.08, ns). Policy accounted for 72 percent of debate themes and 73 percent of newspaper comments, while character comprised 28 percent of debate utterances and 17 percent of article themes. This hypothesis was not supported for the Republican debate either (X2[df=21=1.0, ns). Policy accounted for 64 percent of remarks in the debate and 69 percent of the comments in the newspaper coverage; character accounted for 36 percent of debate themes and 31 percent of newspaper utterances.

Implications and Conclusion

This research provides a much-needed window into how newspapers cover presidential primary election debates. We found that these debates are more positive than negative overall (58 percent acclaims, 31 percent attacks, 12 percent defenses). This is very consistent with the research conducted on general election presidential debates.43

In contrast, the newspaper coverage accentuated the negative (overall, 45 percent attacks, 40 percent acclaims, 16 percent defenses). If the general public has the impression that debates are mostly negative, it could well be the skewed news coverage, rather than the debates themselves, that created this erroneous impression.

These results are consistent with the observation advanced by Lichter and Noyes: “The candidates also ran more positive campaigns than voters might have guessed, since news reports consistently emphasized their most negative rhetoric.”44 The findings confirm earlier studies of campaign coverage;45 negative coverage was not the majority of coverage. However, coverage of the debates was significantly more negative than the debates themselves. We now have data demonstrating the magnitude of the discrepancy between debates and newspaper coverage. This bias in coverage — accentuating the negative, focusing on attacks and conflict – is readily understandable.

Attacks may be, in general, more interesting than acclaims. Attacks are certainly more dramatic. They may also highlight differences between the candidates, although candidates can certainly differ in their acclaims as well. Defenses highlight the clashes of the candidates, and the newspapers did not slight this function. Defenses constituted 15 percent of the debates and 16 percent of the newspaper coverage. Although it is understandable, these newspapers did not accurately represent the functions that occurred in these two debates. While this result may not be surprising, we can not only confirm our suspicions but also quantify them.

We found that debates focus significantly more on policy (issues) than character (image).

This finding is also consistent with research on general presidential debates.46 It is also consistent with research by Zhu, Milavsky and Biswas.47 But they also argued that televised debates may have the negative effect of focusing prospective voters’ attention on candidate image rather than issue knowledge. Clearly, the debates themselves are not focusing on candidates rather than issues.

Furthermore, we have to give the coverage high marks for not characterizing these debates as merely a clash of personalities. Even the Republicans, who might have been expected to dwell on character in the aftermath of President Clinton’s trial and impeachment, devoted only about one-third of their utterances to character. Both the debates (68 percent policy overall, 32 percent character) and the newspaper articles (72 percent policy overall, 28 percent character) stress policy substantially more than character. This finding, unlike the bias toward attacks, is not consistent with previous analyses. Lichter and Noyes argued that “the problem was not that the candidates avoided substance during the campaign; the problem was that the available substance lacked the dramatic packaging necessary for it to become ‘news.”‘ However, these articles did not slight substance (policy) in their coverage of these two primary debates.

These candidates – and these journalists – may be responding to voter preferences. In October of 1999, Princeton Survey Research Associates reported that 90 percent of respondents reported that their presidential vote was more influenced by policy, while only 8 percent indicated that the candidates’ character was a more important determinant of their vote.49

Of course, one important limitation of this study is that it only analyzes news coverage in newspapers. Television, radio and, increasingly, the Internet are other sources of news. Lichter and Noyes argue that voters learn more from newspaper reports than television reports.” Still, future research could profit from contrasting messages with both print and electronic news coverage.

Notes

1. Thomas E. Dewey and Harold E. Stassen, “Should the Communist Party in the United States

be Outlawed? Affirmative Harold E. Stassen; Negative Thomas E. Dewey,” Vital Speeches of the Day 14, no. 16, 1 June 1948,482-489.

2. Samuel J. Best and Clark Hubbard, “The Role of Televised Debates in the Presidential Nominating Process,” in In Pursuit of the White House 2000: How We Choose Our Presidential Nominees, William G. Mayer, ed. (New York: Chatham House, 2000), 255-284; John W. Davis, U.S. Presidential Primaries and the Caucus-Convention System: A Sourcebook (Westport, Conn.: Greenwood Press, 1997); Kathleen E. Kendall, Communication in the Presidential Primaries: Candidates and the Media, 1912-2000. (Westport, Conn.: Praeger Publishers, 2000).

3. Best and Hubbard, “The Role of Televised Debates;” David J. Lanoue and Peter R. Schrott, “The Effects of Primary Season Debates on Public Opinion,” Political Behavior 11, no. 3 (September 1989): 289-306; Michael Pfau, “Intra-Party Political Debates and Issue Learning,” Journal of Applied Communication Research 16, no. 2 (fall 1988): 99-112; Mike Yawn et al., “How a Presidential Primary Debate Changed Attitudes of Audience Members,” Political Behavior 20, no. 2 (June 1998): 155-181.

4. Kathleen H. Jamieson and David S. Birdsell, Presidential Debates: The Challenge of Creating an Informed Electorate (New York: Oxford University Press, 1988), 122.

5. See, e.g., Stephen H. Chaffee, Xinshu Zhao, and Glenn Leshner, “Political Knowledge and the Campaign Media of 1992,” Communication Research 21, no. 3 (June 1994): 305-324; James B. Lemert, “Do Televised Presidential Debates Help Inform Voters?” Journal of Broadcasting and Electronic Media 37, no. 1 (winter 1993): 83-94; Jian-Hua Zhu, Ronald J. Milavsky, and Rahul Biswas, “Do Televised Debates Affect Image Perception More than Issue Knowledge? A Study of the First 1992 Presidential Debate,” Human Communication Research 20, no. 3 (March 1994): 302-333.

6. Thomas E. Patterson, The Mass Media Election: How Americans Choose their President (New York: Praeger, 1980), 21, 24.

7. Herbert J. Gans, “Lessons 1976 Can Offer 1980,” Columbia Journalism Review 15, no. 5 (January/February 1977): 25-28, 25.

8. Guido H. Stempel III and John Windhauser, “Newspaper Coverage of the 1984 and 1988 Campaigns,” in The Media in the 1984 and 1988 Presidential Campaigns, Guido H. Stempel III and John Windhauser, eds. (Westport, Conn.: Greenwood Press, 1991),13-66.

9. Gordon Mantler and David Whiteman, “Attention to Candidates and Issues in Newspaper Coverage of 1992 Presidential Campaign,” Newspaper Research Journal 16, no. 3 (summer 1995):1428.

10. David Domke et al., “News Media, Candidates and Issues, and Public Opinion in the 1996 Presidential Campaign,” Journalism and Mass Communication Quarterly 74, no. 4 (winter 1997): 718737.

11. Mantler and Whiteman, “Attention to Candidates and Issues,” 23. 12. Domke et al., “News Media, Candidates and Issues,” 727.

13. Lee Sigelman and David Bullock, “Candidates, Issues, Horse Races, and Hoopla: Presidential Campaign Coverage, 1888-1988,” American Politics Quarterly 19, no. IJanuary 1991): 5-32.

14. Robert S. Lichter and Robert E. Noyes, Good Intentions Make lead News: Why Americans bate Campaign Journalism (Lanham, Md.: Rowman& Littlefield, 1995),124, xvii.

15. Stempel and Windhauser, “Newspaper Coverage,” found in 1984 that 33 percent of newspaper coverage of Democratic presidential and vice presidential candidates and 41 percent of coverage of Republican presidential and vice presidential candidates were negative. In 1988, they found that 25 percent of the coverage of Democratic candidates was negative; 41 percent of the coverage of Republican candidates was negative.

16. Domke et al., “News Media, Candidates and Issues,” 728.

17. For example, see Wayne Jacques et al., “Some Aspects of Major Newspaper Coverage of the 1992 Presidential Debates,” American Behavioral Scientist 37, no. 2 (November 1993): 252-256; John T. Morello, “‘Who won?’: A Critical Examination of Newspaper Editorials Evaluating Nationally Televised Presidential Debates,” Argumentation and Advocacy 27, no. 3 (winter 1991): 114-125.

18. Michael Pfau and William P. Eveland, “Debates Versus other Communication Sources: The Pattern of Information and Influence,” in The 1992 Presidential Debates in Focus, Diana B. Carlin and Mitchell S. McKinney, eds. (Westport, Conn.: Praeger Publishers, 1992),155-173.

19. William L. Benoit and William T. Wells, Candidates in Conflict: Persuasive Attack and Defense in the 1992 Presidential Debates (Tuscaloosa, Ala.: University of Alabama Press, 1996).

20. William L. Benoit and Allison Harthcock, “Functions of the Great Debates: Acclaims, Attacks, and Defenses in the 1960 Presidential Debates,” Communication Monographs 66, no. 4 (December 1999): 341-357.

21. William Benoit and Leann M. Brazeal, “A Functional Analysis of the 1988 Bush-Dukakis Presidential Debates,” Argumentation and Advocacy (forthcoming).

22. William L. Benoit, Joseph R. Blaney, and P. M. Pier, Campaign ’96: A Functional Analysis of Acclaiming, Attacking, and Defending (Westport, Conn.: Praeger Publishers, 1998).

23. Benoit and Harthcock, “Functions of the Great Debates.” 24. Benoit and Brazeal, “Functional Analysis.”

25. Benoit, Blaney, and Pier, Campaign ’96. 26. Ibid.

27. Chaffee, Zhao, and Leshner, “Political Knowledge.” 28. Lemert, “Televised Presidential Debates.”

29. Nancy B. Lowden et al., “Media Use in the Primary Election: A Secondary Medium Model,” Communication Research 21, no. 3 (June 1994): 293-304.

30. Roderick P. Hart, Campaign Talk: Why Elections are Good for Us (Princeton, N.J.: Princeton

University Press, 2000), 173.

31. J. T. Morello, “‘Who won?”‘ 114-125.

32. Pamela J. Shoemaker and Stephen D. Reese, Mediating the Message: Theories of Influences on Mass Media Content (White Plains, N.Y.: Longman Publishing Group, 1991), 90.

33. Shoemaker and Reese, Mediating the Message, 91. 34. Shoemaker and Reese, Mediating the Message, 88.

35. D. O. Sears and S. H. Chaffee, “Uses and Effects of the 1976 Debates: An Overview of Empirical Studies,” in The Great Debates: Carter vs. Ford, 1976, S. Kraus, ed. (Bloomington, Ind.: Indiana University Press), 228.

36. Lichter and Noyes, Good Intentions Make Bad News.

37. Domke et al., “News Media, Candidates and Issues;” RobertS. Lichter, Robert E. Noyes, and Lynda Lee Kaid, “No News or Negative News: How the Networks Nixed the ’96 campaign,” in The Electronic Election: Perspectives on the 1996 Campaign Communication, Lynda Lee Kaid and Dianne G. Bystrom, eds. (Mahwah, N.J.: Lawrence Erlbaum, 1999), 3-13; Mantler and Whiteman, “Attention to Candidates and Issues;” Stempel and Windhauser, “Newspaper Coverage.”

38. Benoit, Blaney and Pier, Campaign ’96, 3.

39. Benoit, Blaney and Pier, Campaign ’96, 13-14.

40. William L. Benoit and Leann M. Brazeal, “A Functional Analysis of the 1988 Bush-Dukakis

Presidential Debates” Argumentation and Advocacy (in press).

41. Jacob Cohen, “A Coefficient of Agreement for Nominal Scales,” Educational and Psychological Measurement 20, no. 1 (spring 1960): 37-46.

42. Joseph L. Fleiss, Statistical Methods for Ratios and Proportions (New York: Wiley, 1981).

43. Benoit and Brazeal, “Functional Analysis;” Benoit, Blaney, and Pier, Campaign ’96; Benoit and Harthcock, “Functions of the Great Debates.”

44. Lichter and Noyes, Good Intentions Make Bad News, xvii.

45. Domke et al., “News Media, Candidates and Issues;” Mantler and Whiteman, “Attention to Candidates and Issues;” and Stempel and Windhauser, “Newspaper Coverage.”

46. Benoit and Brazeal, “Functional Analysis;” Benoit, Blaney, and Pier, Campaign ’96; Benoit and Harthcock, “Functions of the Great Debates.”

47. Zhu, Milavasky, and Biswas, “Televised Debates Affect Image Perception,” 327. 48. Lichter and Noyes, Good Intentions Make Bad News, 112.

49. NBC News/Wall Street Journal poll, “Roper Center at University of Connecticut,” Public Opinion Online, 12 November 1999, (24 August 2001).

50. Lichter and Noyes, Good Intentions Make Bad News, 101.

by Bryan H. Reber and William L. Benoit

Reber is a doctoral candidate in the School of Journalism at the University of Missouri-Columbia, and Benoit is a professor in the Department of Communication at the University of Missouri-Columbia.

Copyright Newspaper Research Journal, Department of Journalism, University of Memphis Summer 2001

Provided by ProQuest Information and Learning Company. All rights Reserved