Message Framing part 3: Political Attack Ads

It seems that every election cycle a series of reports and editorials pieces surface, decrying the “poisoned nature of contemporary political discourse,” usually substantiated by anecdotal references to snippets taken from negative ads being run in particular races and campaigns. Of course, accompanying these criticisms of negative ads is the expression of sentiment calling for greater civility in our political discourse, while pundits and commentators simultaneously recollect the good old days when campaigns were just about the “issues.” Such expressions of both disgust and nostalgic longing for a pristine past are hardly rooted in a historically accurate understanding of American political campaigns. Since the founding of our republic, political campaigns have often been characterized by negativity and generous portions of mud-slinging. For instance, in the 1800 presidential campaign between Thomas Jefferson and John Adams, one newspaper claimed that if Jefferson was elected president that murder, rape, and adultery would be openly practiced, while in the campaign of 1828, candidate Andrew Jackson was accused of, among other things, being a murderer and a son of a prostitute (Begley, 2008). Negative campaigning did not end there then only to and in the poisoned and polarized atmosphere of party politics in the 1980’s however; Abraham Lincoln, in the election of 1860, was described in such colorful terms as a “thief, ignoramus, buffoon and butcher” (Seiter and Gass, 2010, p.218), while Harry Truman compared Republicans to Hitler in the election of 1948, just after World War II (Knonhollz, 2008). Though it is evident from this handful of examples that so-called “attack ads” are as old as the republic, they are not however perceived to be as American as apple pie, as Underation (2009) explains, “Americans like to imagine their political discourse has a sense of moral excellence about it; in short, we like to imagine we are electing people who are morally sound and who speak to high-minded and large-hearted ideals that we are taught to embrace from our earliest years” (258). That dissonance between American civic virtue and actual political practice is a significant contributing factor to the palpable uneasiness so evident in both the public perception and discussion of political attack ads. Upon further review however, if done appropriately, it might not be the case that attack ads are all that un-American after all, since they are often rich in detail and chalk full of information which can assist the public in exercising their civic obligation to vote.

For some time, scholars have wrestled with developing a definition which is sufficiently clear and expansive. An early definition is offered by Pinkleton (1997) who explains that a negative ad is “any ad that carries discouraging words about an opponent” (p.20). A more nuanced definition is offered by Cheng and Riffe (2008) who claim that, in general, researchers seem to agree that negative political advertising directly assaults its target, citing an opponent’s broken promises, voting record, and public misstatements. In an effort to further clarify the rules of fair-play in the use of attack ads, Lau and Rovner (2009) make a distinction between ads about policy and ads which are essentially about character assassination. On the one hand, negative information about an incumbent’s policies, voting record, or failure to provide for his or her constituents provides suitable material for negative ads. Similarly, for non-incumbents, negative information such as a candidate’s experience or ideological belief that might provide clues to how the candidate may govern once in office is relevant matter to attack. On the other hand, salacious information about a candidate’s behavior in college or the candidate’s past marital problems are not matters of relevance to campaigns or to understanding of a candidates policy positions, nor are they necessarily future indicators of performance if elected. Therefore, such information is not suitable fodder for attack or analysis. A similar way of restating this clarification in more precise terms is found in the distinction, put forward by Seiter and Gass (2010), between argumentativeness and verbal aggression. They are argue that “being assertive or argumentative or being willing to advance one’s own position and refute the position of an opponent, is not the same thing as being hostile or verbally aggressive.” (219) In other words, argumentativeness focuses on policy and relevant character issues, while verbal aggression concerns ad hominem attacking. Based upon the discussion above, a negative ad is defined as an attack upon an opponent’s policy or past voting record or relevant character concerns, along with the implications which these might have for future decisions if elected to office.

Hughes (2003), distinguishes between two primary forms: attack political advertising and comparative political advertising. Attack political advertising involves an aggressive, one-sided assault designed to draw attention to an opponent’s weaknesses in either character or issue positions, while comparative political advertising identifies a competing candidate, draws comparisons, implies inferiority, and attempts to degrade prospective voter perceptions of the targeted candidate. The latter kind of ad can be either direct or implied. In the case of direct comparison ads, they use “a two-sided message to identify the targeted candidate and contrast specific aspects of the candidates’ records, experience or issue positions” while “implied comparative advertising is one-sided and does not mention the targeted candidate specifically” (164). Another form, the indirect comparative ad is employed in campaigns from time to time, but is rarely used effectively on account of it being overly subtle to serve the purpose of a negative ad.

A particularly interesting line of investigation into attack ads pertains to an examination of the situations in which they are utilized. In this connection a few separate studies are worthy of some consideration. First, Damore (2002), surveyed TV ads run by major party presidential candidates competing in general elections from 1976 to 1996 and came up with the following generalizations about the use of negative ads: negative campaign tactics are a function of candidate poll standings, of the behavior of their opponents at prior points in campaigns, of characteristics of issues in the campaign agenda, and of proximity to election day (679). Second, in a study of U.S. Senate campaigns between 1988 and 2002, Lau and Pomper (2001, 2004) found that negative ads were generally used by candidates who are behind, are in close elections, are challengers, are candidates with relatively few campaign resources, are Republicans, are males, and are candidates whose opponents are attacking them. Third, Airne and Benoit (2005) examined television spots from campaigns in the 2000 election reaching some significantly different results from these previous studies. Prior research had attributed negative ads to poll standing, to challenger status, and to Republican candidates, but Airne and Benoit found that “incumbents attacked at about the same level (35%) as challengers,” and that Democrats and Republicans attack at about the same rate (p.481). Other findings of interest in this study were that winners and losers attacked and acclaimed at about the same rate -- winners: 68%; losers: 66%)—however, winners discussed policy more than losers in their negative ads (64% to 60%), whereas losers focused more on character (40%, 36%) (p. 483). Clearly, this data exposes the fact that negative campaigning is neither a challenger issue or a party issue, rather, it is a widely, if not universally used media tool candidates use to disseminate their concerns to the voting public.

Now, moving on to address the question of undoubtedly the most interest, “do they work?” The answer is, “yes,” depending on who you talk to. For instance, ScienceDaily (2008) reported on a study by Phillips, Urbany, and Reynolds (2008) who conducted a field study of registered voters aged 18-23 using real advertisements from the 2004 presidential election. This study found that 14% of the participants who viewed an ad that attacked their preferred candidate – were influenced by the ad’s content and were moved in the direction of the advertising candidate while viewing positive ads did not lead to significant change in voter attitude. The article went on to note the significance of the study results commenting that, “77 percent of college-educated 18-24 year olds who were registered cast a vote in the 2004 presidential election, compared to 64 percent of registered voters as a whole” (“Young Voters,” 2008). A separate study, conducted by Tinkham, Lariscy, and Avery (2009), concluded that voters in the 60 and above age group are “the group that is most likely to yield behaviorally to the sponsor of a political ad” (117). This is truly a significant find given that “citizens over 60 both register to vote and do vote in higher percentages than any other age group” (105). Without getting into the game of citing study for study, it is notable that although there is disagreement among researchers, there is at the same time persistent belief in the effectiveness of negative ads as Seiter and Gass (2010) explain, “Campaign managers and political consultants are unequivocal in their belief that attack ads are highly effective” (227).

Since there is an apparent disconnect between public sentiments expressed about attack ads and the actual effectiveness of attack ads, it is worth taking a moment to offer a suggestion about why they work. The answer proposed here is rooted in message framing which has its roots in Prospect Theory (for an overview of PT, see the first article in this series). One component of the rhetorical strategy of message framing, which helps account for the effectiveness of attack ads, is the certainty effect. The certainty effect says that people underweight outcomes that are merely probable in comparison with certain outcomes, unless they are persuaded that their situation is perilous enough that they must take a risk in order to enhance it. That means, that most voters, even if they aren’t terribly thrilled with an incumbent are still less likely to vote for the challenger even if they perceive that he might do a better job. In that case, using only acclaim (positive) ads wont tend to secure the vote for the challenger since the voter is more likely to settle with what they know as safe, though not thoroughly desirable. Negative ads (loss framed ads) are necessary to get the voter to see that voting for the incumbent has the potential of causing serious economic or security harms to the voter.

The last point which remains to be addressed is what it is that makes an attack ad effective. Morris (2008) argues that the most effective attack ads share three characteristics: accuracy, believability, and thematic consistency. When ads do not meet these criteria, they backfire. A further component of effective ads is discussed by Knonholz (2008), who contends that effective negative ads are ads which are targeted to address concerns that voters already have about candidates or policies. In other words, if attack ads are to strike a responsive chord in voter perceptions they must be relevant to the context of the campaign and the concerns of voters (Seiter and Gass, 2010). Negative political ads that make an impact usually reflect and amplify bits of public thought and feeling according to Uderation (2009).

In conclusion, negative political campaigning is not new, and, like it or not, it is here to stay. Prior research has done some good work in clarifying certain positive aspects of negative ads while refuting false claims about its supposedly harmful effects. Much work remains however, in order to better understand why these ads are persuasive, what kinds of negative messages are actually effective, and under what circumstances they are most successfully employed.