A functional analysis of 2008 U.S. presidential primary debates.

• Author: Benoit, William L.
• Position: Report

Presidential primary debates have been employed to help inform voters in the United States since 1948, when Dewey and Stassen participated in a debate on radio during the Oregon Republican primary campaign. In recent years, primary debates have been much more common than debates in the general election campaign. In 2004, for instance, there were 21 primary, three presidential, and one vice presidential debates in the campaign (Benoit et al., 2007). Various studies have found that voters can be influenced by watching presidential primary debates (Benoit, McKinney, & Stephenson, 2002; Benoit & Stephenson, 2004; Lemert, Elliot, Nesvoldet, & Rarick, 1987; Pfau, 1984, 1987, 1988; Wall, Golden, & James, 1988; Yawn, Ellsworth, Beatty, & Kahn, 1998). Meta-analysis has established that primary debates can increase issue knowledge, affect perceptions of candidate character, and change vote choice (Benoit, Hansen, & Verser, 2003); these effects are even larger in primary than general debates, probably because voters have less knowledge, fewer character perceptions, and weaker commitment to vote choice earlier than later in the campaign. Data from the 2008 campaign indicates that 24 of the Democratic and Republican primary debates attracted 90 million viewers (Kurtz, 2008; Memmott & Carnia, 2007; Page, 2008). Clearly, presidential primary debates merit scholarly attention.

The 2008 U.S. presidential campaign is particularly noteworthy for two reasons. First, 2008 was the only open campaign in recent memory. Not since 1952, when General Dwight Eisenhower faced Governor Adlai Stevenson, had a U.S. presidential campaign not included either a sitting president or vice president. However, in 2008, President George W. Bush was completing his second and final term and Vice President Dick Cheney decided not to run for the top slot. Although some recent campaigns have seen challenges to renomination of the incumbent (e.g., in 1992 Pat Buchanan contested President George H. W. Bush for the Republican nomination and in 2000 Bill Bradley ran against Vice President A1 Gore for the Democratic nomination), the lack of an incumbent makes both primary races competitive and in all likelihood makes campaign messages even more important for voters who have less information about the candidates compared to typical election campaigns.

Furthermore, although the primary campaign commenced earlier than ever before, the Democratic nominee was not decided until much later than usual, with Senator Barack Obama wresting the nomination away from Senator Hillary Clinton in June. This is the first rime a nominee for one of the two major U.S. political parties was not a white male. When Senator John McCain selected Governor Sarah Palin as his running mate, it assured that for the first rime in U.S. history the elected President or Vice President would not be a white male. Thus, the debates that led up to this historic election merit scholarly attention.

LITERATURE REVIEW

Relatively few studies have content analyzed U.S. presidential primary debates (for other kinds of research on this message form, see Berquist, 1960; Best & Hubbard, 2000; Blankenship, Fine, & Davis, 1983; Hellweg & Phillips, 1981; Kane, 1987; Ray, 1961; Stelzner, 1971). A study of U.S. presidential primary debates from 1948-2000 offers some insights into the content of these messages (Benoit, Pier, et al., 2002). Acclaims were the most common function of these primary debates (63%), followed by attacks (32%) and defenses (4%). When presidential candidates did attack in primary debates, they were more likely to attack members of their own political party (47%) than candidates in the opposing party (30%) or to attack the status quo (criticisms that include members of both major political parties, 24%). The candidates in primary debates discussed policy (63%) more frequently than character (37%). More of these policy utterances concerned general goals (40%) or past deeds (37%) than future plans (24%). When the candidates discussed character in primary debates, they discussed ideals (45%) and personal qualifies (36%) more than leadership ability (19%). Benoit et al. (2007) reported similar patterns: The Democratic primary debates of 2004 employed acclaims (63%) more than attacks (32%) or defenses (4%). About three-fourths of their statements concerned policy (and the remainder character). One contrasting finding was that these Democratic candidates were more likely to direct attacks toward Republican President Bush (65%) than to one another (21%) or the establishment generally (15%). We extend this line of analysis, deploying Functional Theory to analyze 36 primary debates from Democrats and Republicans in the 2008 U.S. presidential campaign.

FUNCTIONAL THEORY

The Functional Theory of Political Campaign Discourse was developed by Benoit and his associates through a series of studies (see, e.g., Benoit, 1999, 2007; Benoit & Brazeal, 2002; Benoit & Harthcock, 1999; Benoit et al., 1998, 2003; Benoit, Stein, et al., 2007; Benoit, Pier, et al., 2002). Functional Theory posits that citizens vote for the candidate who appears preferable on the criteria that are considered most important to each voter (Benoit, 2007). Candidates can demonstrate their desirability in three ways. First, the candidate can engage in acclaiming or self-praise. The greater the perceived benefits or advantages of one candidate, the more likely that person will appear preferable to voters, compared to opponents. Second, candidates can attack or criticize opponents; as voters become more aware of alleged costs or disadvantages of opponents, those competitors should appear less desirable to voters (of course, it is possible that the source of these attacks can experience a backlash from voters who dislike mudslinging). Finally, candidates who have been the target of attack can defend against (refute) those attacks. The fewer and smaller the costs or disadvantages, the more likely that a candidate will appear preferable to opponents. These three options can be seen as roughly similar to cost-benefit analysis, providing information that can help persuade the voter to prefer one candidate (we do not claim that voters systematically quantify the impact of acclaims, attacks, or defenses or perform mathematical calculations to decide their vote choice; acclaims tend to increase one's benefits, attacks may increase an opponent's costs, and defenses can reduce one's costs). For example, research on German presidential debates has confirmed that viewers react differently to acclaims and attacks (Reinemann & Maurer, 2005). However, it is important to keep in mind that all voters do not react in the same way to a given utterance. For example, Jarman (2005) found that those who watched the second general election presidential debate in 2004 evaluated the comments from the candidate representing their own party more favorably than viewers who affiliate with the opposing party.

The three functions (acclaims, attacks, and defenses) can be employed on two distinct topics: policy (issues) or character (personality). Policy utterances address governmental action or problems that are amenable to governmental action. Character comments are about the candidates as individuals (personality, leadership experience, and values). Of course, the relative importance of these two general topics of discourse can vary from one voter to another. Functional Theory also subdivides the two topics into three forms of policy (past deeds, future plans, general goals) and three forms of character (personal qualities, leadership ability, and ideals) (sec, e.g., Benoit, 2007 for examples of the forms of policy and character).

Functional Theory (Benoit, 2007) anticipates that acclaims will be more common than attacks. Acclaims have no drawbacks, but because many voters report that they dislike mudslinging (Merritt, 1984; Stewart, 1975) there is some incentive to moderate attacks. Defenses are expected to be rare for three reasons: Most attacks occur where a candidate is weak, so responding to an attack will usually take the candidate off-message; one must identify an attack to refute it and that identification may inform or remind voters of a potential weakness; and attacks may create the undesirable impression that the candidate is reactive rather than proactive.

H1. U.S. presidential primary debate utterances in the 2008 campaign will include more acclaims than attacks and more attacks than defenses.

This prediction is consistent with past research on presidential primary debates (Benoit, 2007).

Lau and Pomper (2004), relying on newspaper accounts of campaigns, observed that Republicans attacked more than Democrats in U.S. Senate campaigns from 1992-2002. Airne and Benoit's (2005) analysis of 2000 non-presidential advertisements found mixed results: Democrats attacked more than Republicans in Senate and gubernatorial advertisements but less than Republicans in House advertisements. Cooper and Knotts's (2004) study of gubernatorial ads in 2000 reported that Republicans attacked more (36% to 25%) and acclaimed less (42% to 54%) than Democrats. Because the literature on the functions of campaign discourse from the two major political parties is mixed, we pose a research question:

RQ1. Is there a difference in frequency of the functions from Democrats and Republicans in the 2008 U.S. presidential primary debates?

Functional Theory indicates that candidates competing in primary campaigns can direct their attacks to three different targets: candidates from their own political party, candidates from the opposing party, or the status quo (e.g., "the Washington establishment," including both Democrats and Republicans). In earlier presidential campaigns (1948-2000), candidates in presidential primary debates were prone to attack members of their own political party, who were their immediate opponents (Benoit et al., 1998, 2002). However, Democratic primary...