Persuasion and "Purpleness" in Presidential Campaign Advertising
Conventional wisdom in political advertising holds that so-called "purple" regions--areas that tend to split votes evenly between Democratic and Republican candidates--tend to be more persuadable than non-purple regions. Subscribing to that theory, campaigns typically run significantly more advertisements in purple markets in hopes of swinging them toward their candidates. But is there sound rationale to support that conventional wisdom? This study uses advertising data and election data from recent presidential races to determine (1) whether campaigns actually ran more advertisements in more-purple media markets, and (2) whether campaign advertisements were statistically more effective in more-purple media markets than in less-purple markets. Specifically, this study uses election data from the 2000 and 2004 presidential elections in battleground states--states in which both candidates campaigned competitively--as well as campaign advertising data by media market in those states. Control variables include demographic information, past electoral outcomes, and economic data, all calculated at the media market level. The study then tests the same correlations for the Obama and McCain campaigns in Florida, a key battleground state in 2008, to determine whether they utilized the same targeting strategy of the 2000 and 2004 presidential campaigns. All told, the statistical regressions reveal that the presidential campaigns in 2000 and 2004 did run significantly more advertisements in purple markets, even when holding other factors constant. However, the research reveals no evidence that out-advertising an opponent in more-purple media markets actually corresponds to greater swings in vote outcome than out-advertising an opponent in less-purple markets. Further, the only type of market in which out-advertising an opponent appears to affect vote outcomes appears to be red, Republican-leaning markets. In those markets, out-advertising an opponent does correspond to more votes. In purple and blue media markets, there is no significant correlation between out-advertising an opponent and vote outcomes, holding other factors constant. In 2008, the Obama campaign appears to have subscribed to the conventional wisdom in Florida, running significantly more advertisements in purple markets than in non-purple markets; by contrast, the McCain campaign appears to have largely ignored purpleness as a factor in its Florida advertising strategy. As a result, Obama underperformed in Florida relative to his performance in other states in 2008, and purpleness again failed to correspond to persuadability.
Showing items related by title, author, creator and subject.