Governance: The Political Science Journal at UNLV

Governance: The Political Science Journal at UNLV


Campaign Advertisements; Campaign Ads; Attack Ads; Contrast Ads

File Format


File Size

273 Kb


This research is intended to find out how Senate candidates’ vote shares are affected by advertisements with a negative tone. Prior literature is largely inconsistent in determining the impact of advertisement tone, but separating negative advertisements into two subcategories, advertisements that attack an opponent and advertisements that contrast between oneself and one’s opponent, clearer correlations become evident. Literature suggests that attack advertisements are more effective at generating voter responsiveness and mobilizations compared to contrast advertisements, they also come with a backlash effect that often damages the attacker more than the target. Thus, candidates with a higher percentage of attack advertisements should have a lower vote share, and contrast advertisements should generate mixed results. Using data from the Wesleyan Media Project and the MIT Election Lab, it was clear based on a linear regression test that contrast advertisements instead of attack advertisements showed a more significant relationship between percentage and vote share. However, while the relationship between contrast percentage and vote share was statistically significant, the relationship between attack percentage and vote share was not.