I. INTRODUCTION
Realignment and the Contribution of Negative Campaigning

Realignment, the shifting of political power from Democrats to Republicans, changed the face of Southern politics. Direct causes of secular realignment include issues of race as a result of federal intervention, the 1964 Civil Rights Act, and the implementation of the Voting Rights Act. Another factor of realignment in the South, the shift of Democrats towards a more liberal ideology, pushed many conservative Democrats into the Republican ranks during the 1980’s (Black and Black, 2002). The use of negative campaign strategies indirectly, but significantly, contributed to the transformation of political power. Negative campaigning is a valuable tool used to undermine the overwhelming support of Democratic incumbents who suffocated Republican attempts at election for decades. Without this strategy, it has been virtually impossible for Republican challengers to defeat Democrats or for incumbents to defend against attacks. The significance of negative campaigning, or “attack advertising,” is illustrated in the context of the 2002 elections in Georgia. Despite widespread criticism of negative campaigning, the 2002 Georgia elections demonstrate the increased use of negative attack advertising. The 2002 elections in Georgia also show the effectiveness of negative campaigning for undermining the support of formidable Democratic incumbents and other opponents.

Purpose

Negative campaigning changed the face of Georgia politics throughout the last twenty years. Due to the significance and increasing prevalence of attack advertising, scholars engage in research studying its effects. The purpose of this research is to examine selected 2002 Georgia elections and review specific scholarly studies in the context of real campaigns to answer the two questions: 1. Who uses negative campaigning and why? 2. Do negative advertisements win elections? 3. Is it possible to reap adverse effects using negative advertising? Upon the conclusion of my research, I offer my thesis statement: 1. Although challengers are more likely to attack first, incumbents (Democrats or Republicans) will also use negative advertising; 2. Negative advertisements are highly effective in elections and do lead to victories; 3. There is no real limit to negativity in advertising due to the phenomenon of the sleeper effect, which depletes any backlash to the source without diminishing the negative message.

Methodology

Although quantitative data such as election results enrich the research, this study is based primarily on qualitative research derived from personal interviews, and secondary research of newspaper sources, relevant scholarly literature, campaign materials, and advertisements. The examined elections include the Governor’s race between Roy Barnes and Sonny Perdue, the State Senate race between Doug Haines and Brian Kemp, the State Senate Primary race between Joyce Stevens and Renee Unterman, and the U.S. Senate race between Max Cleland and Saxby Chambliss. The campaigns are evaluated in the context of scholarly research to determine if the scholars’ results were evidenced in these real campaigns.

II. LITERATURE REVIEW

Analyzing the effects of negative attack advertising in Georgia elections requires examining public opinion, campaign rhetoric, and the political process during elections. This study focuses on contrasting the Georgia consultants’ and legislators’ definition of negative advertising with the scholarly definition; the usage of negative advertisements in Georgia campaigns; and the trends involved in “going negative” as evidenced by Georgia campaigns. For the purpose of this examination, the necessary subjects to review include: typology of negative, criticism and support of negative, rules of negative, and information processing of negative. The information cited offers a deeper understanding of negative campaigning and a context in which to conduct analysis of Georgia elections.

Typology

Defining negative campaigning requires the differentiation between negative, comparative, and positive campaigning. Positive campaigns are signified by information about a candidate’s strengths, merits, and the beneficial policies that will be pursued if he or she is elected. While positive campaigning is about the candidate, negative campaigns usually focus on the opponent and the deficient nature of his programs, accomplishments, and qualifications (Lau and Pomper, 2001). There is some discrepancy as to whether or not any contrast drawn between two candidates is considered “negative campaigning.” Mayer explains that negative campaigning is campaigning that attacks or is critical of an opposing candidate where positive campaigning dwells on the candidates own strengths and merits and the beneficial policies that will be adopted if the candidate is elected (Mayer, 1996). Assuming that a candidate, when drawing a distinction between himself and his opponent, tries to paint himself in a favorable light, it follows that the difference would be an attack on the opponent. Mayer agrees that “any serious, substantive discussion of what a candidate intends to do after the election can only be conducted by talking about the flaws and shortcomings of current policies” (Mayer, 1996). Lau and Pomper define the directional meaning of negative as statements in opposition to a person or a program and define the evaluative meaning as statements that are subjectively disliked or distrusted (Lau and Pomper, 2001).

Criticism and Support of Negative Advertising

Is the media an effective source of information or a powerful negative force depressing voter turnout? Researchers differ in their opinions on this controversial topic. Early opinions are highly critical of the media. Robinson (1975) finds that the media increases distrust of the national government and leaders because of its perceived high credibility, mass appeal, and focus on conflict. Becker and Whitney (1980) conclude that reliance on television decreases political knowledge and trust in the government. Likewise, O’Keefe, Mendelsohn, and Liu (1976) claimed that media reliance increases cynicism. More recently, Cappella and Jamieson (1997) concluded that exposure to news about campaign strategies rather than issues increases public cynicism. In support of the media, Pinkleton, Austin, and Fortman (1998) found that media use increases the likelihood of voting. Also, O’Keefe’s research (1980) shows that reliance on newspapers increases positive public attitude while reliance on television does not cause an increase in cynicism. Perhaps the greatest criticism is not that negative advertising in the media causes political disaffection and cynicism, but that it causes political inefficacy by suppressing voter turnout. A decrease in voter turnout may be the most distressing potential consequence of negative advertising. Stephen Ansolabehere and Shanto Iyengar claim “negative tactics not only discourage turnout but suppress dissent…indirectly influence policy by limiting the terms and parameters of policy-making between elections” (Elving, 1996). Ansolabehere and Iyengar, using a set of California races under a variety of controlled experiments, studied the turnout of the 1992 Senate elections (Freedman and Goldstein, 1999). Their research concludes that subjects who view a single attack advertisement during the local news broadcast are about five percent less likely to vote in the upcoming election than subjects who do not view the attack (Lau and Pomper, 2001). Ansolabehere and Iyengar are the first social scientists to suggest the demobilizing effect of negative campaigning. However, their research generated interest and some controversy among analysts. Other social scientists such as Finkel and Geer (1998), Freedman and Goldstein (1999), Lau and Pomper (2001), and Tinkham and Weaver Lariscy (1993) report findings that refute Ansolabehere and Iyengar’s research on demobilization.

Conditions to “Go Negative”

Despite criticisms of negative campaigning, scholars widely agree that one of the most effective ways to reach voters is to appeal to their emotions through negative advertisements. Lau and Pomper outline the broadest spectrum of characteristics of candidates likely to employ negative campaign tactics. Using virtually all U.S. Senate contests from 1988 to 1998, Lau and Pomper test their hypotheses based on a core group of candidates who chose campaign strategies with the goal of electoral success (Lau and Pomper, 2001). The seven hypotheses outline characteristics of candidates, shown by previous research, who are most likely to engage in negative campaigning: (1) Candidates who are behind in a race or expect to lose; (2) Candidates in close elections; (3) Challengers with no office to lose; (4) Candidates with fewer resources than their opponents; (5) Republican candidates more than Democratic candidates; (6) Men more than women; (7) Candidates once attacked (Lau and Pomper, 2001). After extensive research, Lau and Pomper conclude that challengers or those in open-seat contests, candidates in close races, males, Republicans, and candidates previously attacked are all likely to run negative campaigns.

However, incumbents, candidates in races with easily predicted outcomes, and well-financed candidates are less likely to have negative campaigns (Lau and Pomper, 2001). These results complement studies by Tinkham and Weaver Lariscy. In their repeat sample time-series design, Tinkham and Weaver Lariscy find strong similarities between U.S. Congressional elections in 1982 and 1990 that support the hypothesis that the perception of incumbency advantage determines a candidate’s choice of strategy (Tinkham and Weaver Lariscy, 1995). Their results in 1992 replicate those of the 1982 patterns. That is, challengers are more negative than incumbents and a candidate’s perception of the utility of incumbency advantage is predictive of the decision to “go negative” (Tinkham and Weaver Lariscy, 1995). From their results, they conclude that a challenger’s best chance to unseat an opponent is a direct attack and incumbents should not wait to be attacked before attacking. Using data from the same Congressional study, Tinkham and Weaver Lariscy report findings that the percentage of the vote obtained is negatively related to campaign tactics focused on opponents rather than those focused on the sponsor (Tinkham and Weaver Lariscy, 1996).

In this study, the effects of incumbent entrenchment are expanded using the outcome of the 1990 election. An examination of the strategies chosen by incumbents, challengers, and candidates in open races support the conclusions of Lau and Pomper among others. This predictive analysis of message strategy and descriptive analysis of differences between incumbents, challengers, and candidates in open races shows that challengers and candidates in open races “go negative” much more than incumbents (Tinkham and Weaver Lariscy, 1996). Other supportive results for Lau and Pomper’s assertions concern the conclusion that Republicans use negative advertising more than Democrats. Theilmann and White, using a series of thought experiments with political consultants, ask the question: “Do consultants who work for Democratic candidates employ attacks differently than their Republican counterparts?” and test the applicability of two theoretical models (Theilmann and White, 1998). Their findings agree with those of Lau and Pomper—Republican consultants appear more willing to employ attack strategies. In summary and for the purposes of this study, the characteristics of those candidates most likely to “go negative” include: challengers, candidates lacking sufficient funds, Republicans, candidates once attacked, and candidates in open races. Because of the descriptive, non-causal nature of several of the studies these characteristics are based on, there are possible flaws in the research. One obvious problem is that self-reports, as used in Tinkham and Weaver Lariscy’s research after the 1982 and 1990 elections, may not be as reliable as actual voting reports. Despite the descriptive nature of the research, this taxonomy of candidate characteristics is an excellent source of examination for real campaigns.

Guidelines for Negative Advertising

Several effects may result when employing negative campaign techniques. The effects are dependent on actions taken by the sponsor and how negative advertising is supported with the facts. Houston and Doan argue that candidates must provide evidence for their negative claims. In their 1999 study, participants evaluate two U.S. Senate candidates, one “shared-ideology” and one “opposing-ideology,” in which one candidate provides evidence to support a claim and the other does not (Houston and Doan, 1999). Results show that shared ideology candidates are significantly hurt by the failure to support the character assertions while the opposing-ideology candidates are not harmed by their failure to provide evidence. In fact, voter evaluations of the opposing-ideology candidates are not affected by the presence or the absence of supportive evidence (Houston and Doan, 1999). Houston and Doan assert that candidates should provide thorough documentation to prevent negative consequences during campaigning. Other than failing to prove their credibility by backing up negative advertisements, candidates must also be wary to avoid going “too negative.” Where a candidate fails to document his charges, the negative advertising may produce “dual effects” also known as the “boomerang effect.” The boomerang effect occurs when the attack harms the sponsor more than it harms the opponent. Jasperson and Fan explain that the intended effect of a negative advertisement is to shape voter perception of the opponent or target. However, if the public believes an advertisement is too harsh, too negative, or too personal the result could be a backlash harming the sponsor rather than the target (Jasperson and Fan, 2002). Jasperson and Fan’s results show that negative information could boomerang against a candidate outside of the laboratory in a real-world context. In summary, the claim by Houston and Doan holds that candidates who choose an attacking campaign strategy must provide evidence for claims made during attacks to maintain credibility. Jasperson and Fan also find that candidates should be conscious of the level of negativity in their advertisements. If the public perceives a message as too negative then a boomerang effect may occur and harm the sponsor’s campaign rather than the opponent’s.

Information Processing of Negative Advertising

The reception of negative advertising by the public determines the outcome of elections. Austin and Pinkleton define public opinion in the context of several specific variables: cynicism, apathy, negativism, and the third-person effect (Austin and Pinkleton, 1995). Cynicism is the belief that the government does not deserve respect or support characterized by mistrust. Whereas cynicism is a lack of confidence in the political system, apathy is a general lack of caring about the political process and a low desire to invest oneself in the political process. Negativism is the perception that political advertising is “unfair, uninformative, unethical, and deceptive” (Austin and Pinkleton, 1995). The third-person effect occurs when an individual believes that he is immune to political advertising while others lack this immunity. Austin and Pinkleton find that there is a correlation between positive and negative public mood and political cynicism, but Leshner and Thorson’s results from their 2000 study claim that negative attitude increases actual voting by motivating the apathetic voters. Also, more positive public mood leads to an increase in self-reported voting but a decrease in actual voting (Leshner and Thorson, 2000). Other studies on public opinion offer similar findings. O’Cass’ studies of political advertising believability and the perceived value of information sources, voter involvement, confidence, and emotion show that negative campaigns run by opposition are just as believable as positive campaigns run by the incumbent. The factors affecting the believability of positive campaigns are voter involvement, satisfaction, and emotion, while only involvement and satisfaction affect the believability of negative campaigns. Thus, incumbent advertising is believable if the voters are satisfied with the party’s performance, and the voters are involved in politics (O’Cass, 2000). O’Cass concludes that the more emotional the public becomes, the more they tend to believe negative campaigns.

In their 1999 collaboration, Weaver Lariscy and Tinkham found that negative attack advertising has both a strong initial impact and a sustained impact. They determine that negative messages are only risky for sponsors in the short term because the public cannot remember the author of the message. Thus, any initial damage done to the attacker will not last (Weaver Lariscy and Tinkham, 1999). Contrary to early thought on the effects of negative messages, the sleeper effect explains how the impact of negative political advertisements persists and even increases over time. “One of the things we can say is that although many voters indicate they don’t like negative advertising, it still impacts them,” explains Weaver Lariscy. “It may be that three weeks later in the voting booth, you can’t remember anything about where you heard that horrible information, but that piece of information is still there. We call that the sleeper effect. The conventional wisdom fifteen years ago was that negative ads created a boomerang effect. In other words, the person who did the attack would be the one who would suffer the greatest consequences. A lot of candidates chose to take the high road” (Schanche, 2002). Weaver Lariscy, in collaboration with Tinkham, has studied the effects of negative messages and the sleeper effect since Michael Dukakis’ loss against President Bush in 1988. During the late 1970’s and 1980’s, negative campaigning was less prevalent than in current elections When attacked by Bush, Dukakis chooses to ‘take the high road’ and does not dignify the accusations with a response. Dukakis’ loss illustrates that this approach is not effective. Another approach used by defensive candidates is to inoculate themselves early against a negative attack. However, this approach could unnecessarily raise damaging issues. Indirect defenses, such as positive ads and endorsements, prove even less effective because they rarely have the power to suppress the impact of the attack (Weaver Lariscy and Tinkham, 1999). In their study, Weaver Lariscy and Tinkham examine patterns of voter response to attack advertising using attack and defensive messages. Participants report their vote immediately after exposure to the message and reported again in a delayed telephone callback. The sleeper effect is based on the “negativity bias” which states that, because they are less powerful, positive messages do not offset negative messages. In the context of a political campaign, the negative attack ad, or base message, is more likely to have a stronger initial impact because the attacker initiates an attitude change. Essentially, the sleeper effect has a delayed impact unlike typical message effects which erode over time. To recognize a sleeper effect, there must be an increased impact at some point in time after the stimulus occurs (Weaver Lariscy and Tinkham, 1999).

Explanations of the sleeper effect are offered to describe the occurrence of the phenomenon. Inherent to discussion of the sleeper effect, the discounting cue holds an important role in the perception of the negative message. Discounting cues designate sources of the negative message by providing a direct reference to the source. In television ads, the discounting cue is typically placed at the bottom of the screen stating the source of the ad. Placement of the discounting cue, before or after the message is arguably critical for a sleeper effect to occur. One explanation, the Discounting Cue Hypothesis (also known as the Dissociation of Memory routes), follows the impact by beginning with an untrustworthy source that reduces the impact of the negative message initially. The hypothesis explains that the negative message is better learned due to its critical examination. Eventually, the source is forgotten and the impact of the negative message increases significantly.

Another hypothesis, Differential Decay, describes the sleeper effect by explaining that the impact of the discounting cue decays faster than the impact of the negative message. Therefore, the source is forgotten more quickly than the message (Weaver Lariscy and Tinkham, 1999). Weaver Lariscy and Tinkham examine “political strategy as an applied context for the sleeper effect phenomenon” using the parameters of the absolute sleeper effect as opposed to the relative sleeper effect (Weaver Lariscy and Tinkham, 1999). The absolute sleeper effect is statistically significant between the first exposure of the negative message and the delayed increase in impact. For an absolute sleeper effect to occur, the discounting cue must appear after the stimulus. Whereas the absolute sleeper effect requires a delayed impact increase, the relative sleeper effect shows that the impact of the initial message might decrease. Without a discounting cue, there would be a loss of impact for sources of high and low credibility with no delayed increase in impact. The nature of the political campaign creates the “prototypical context for the examination of the sleeper effect” with constant back and forth attack and defense messages. Through their collaboration, Weaver Lariscy and Tinkham conclude that, “If an attack stands alone, unanswered, it is effective. If an attack is refuted, the attack will, ultimately, still be effective. If an attacker damages his or her own credibility, the attack message itself will still have a positive impact over time. If attacked, therefore, the best defense seems a strong, swift offense” (Weaver Lariscy and Tinkham, 1999).

This study concludes that the harmful boomerang effect of negative advertising in the media dissipates after time. The message to candidates is to “attack early and attack often” due to the results that the lasting impact essentially harms the opponent, not the attacker. The psychology of public opinion is a complex element in the field of political campaigning. Research shows that candidates who appeal to voter’s emotions are more likely to run successful negative campaigns (O’Cass, 2000). In addition to emotion, public opinion is comprised of cynicism, negativism, apathy, and the third-person effect. Also, attacks are effective even after their sponsor is forgotten (Weaver Lariscy and Tinkham, 1999). However, more research is needed in this area to explain voter’s actions. Extensive research was useful to create more effective negative attack advertisements and stronger defenses against such attacks.

[ Abstract - 1 - 2 - 3 - 4 - Bibliography]