Asymmetric Political Image Effects and the Logic of Negative Campaigning

Political marketing campaigns frequently use negative campaigning approaches, in which a candidate’s advertising criticizes an opponent, but the practice is controversial. Negative advertising in the 2012 US Presidential election was widely chastised in t

  • PDF / 181,559 Bytes
  • 4 Pages / 595.276 x 790.866 pts Page_size
  • 42 Downloads / 215 Views

DOWNLOAD

REPORT


ION The impact of the issues on voting intentions is shown in Table 1. (The image variables were set up with a preference for Romney coded as -0.5, a non-committal answer as 0 and a preference for Obama as +0.5, to give the difference between a preference for the two candidates a unit value in the model). Eight of the 13 image attributes had a significant independent explanatory effect, but some were stronger than others. Which of the two candidates scores highest on the item “Can be effective in Washington” is the most powerful discriminator, with around three times as strong a relationship as any other variable in the model, the odds ratio being 49.7 (meaning when you compare two otherwise identical voters, who differ only on which candidate would be more effective in Washington, the odds against the one who prefers Romney on this issue being an Obama voter are almost fifty times as high as for the one who picks Obama). At the second level of importance to the voting decision are two variables, which candidate is most “Presidential” and which “Has the right values”. On the third level, around half as powerful as these, is “Understands people like me”. Less powerful, in order of importance, are “Likeable”, “Would be fun to meet in person”, “Smart enough for the job” and “Will protect American jobs”. The remaining five image attributes have no extra explanatory effect on voting intention. Table 2 shows how strongly the same list of attributes relates to the voters’ favorability towards Obama and Romney, deriving a separate linear regression model for each. (Coefficients for Romney are negative since a negative score on each of the independent variables indicates a preference for Romney.) Of the 13 attributes, eight are significantly related to the voters’ opinions of Obama and five to Romney, with three common to both models and three finding no place in either. Which of the two candidates is seen as most “Presidential” is a powerful predictor of their views of both (it is the single best predictor of favorability/unfavorability towards Obama). Which of them best “Represents America” and which can best be described as “A man of faith” is also related to opinions of both candidates, although the former is more influential for Obama and the latter for Romney. Six other issues are tied to views of Obama, while having no significant relationship with views of Romney, and a further two are relevant to views of Romney but not of Obama, including “Will protect American jobs” which is the best single predictor of overall attitudes to Romney. Finally, we consider which candidate had a lead (and how large) among the potential swing voters on each of the 13 image attributes. These are shown in Table 3 with a summary of the findings from the two previous tables, giving an overview of the tactical situation for each candidate on each of the 13 issues. Table 3 illustrates that the campaigning situation creates some very clear incentives for negative campaigning. For Romney, the issues of effectiveness in Washington and having the