Record numbers of negative television ads aired in 2018, no doubt reflecting the increase in the number of competitive congressional races across the country. Digital advertising, as the Wesleyan Media Project noted, was also on the rise, but campaigns and their surrogates had not by any means abandoned the 30-second television spot. Nearly half of those ads that aired were purely negative, and all told, 7 of 10 ads contained an attack. Such negativity is not new nor is scholarly interest in the effects of attack advertising. But to date, political scientists have puzzled over mixed findings. Are negative ads more effective, persuading Americans to vote for one candidate over another? Does such negativity demobilize the electorate and debase political discourse? Kim Fridkin and Patrick Kenney, outstanding scholars of modern campaigns, offer a nuanced take on these questions, putting forth a theoretical framework that improves our understanding of how negative campaigning affects the beliefs and behaviors of Americans.
In Fridkin and Kenney’s framework, which they coin the “tolerance and tactics theory of negativity,” both voters and the messages matter. Individuals vary in their tolerance for negativity. Some are comfortable with strident and even personal attacks on the political opposition; others find negative attacks off-putting or unpleasant. Tactics differ too, with substantial variation in the tone and content of negative messages. Negative ads might include ad hominem attacks or focus on important policy matters. Fridkin and Kenney argue that we can better understand the effects of negative campaigning by jointly considering voters’ tolerance for negativity and campaigns’ tactics for attacking the opposition. They hypothesize that the most persuasive negative ads are harsh in tone (uncivil) and deliver politically meaningful content about the risk of supporting a candidate (relevant). They also posit that negativity can, under some conditions, damage the body politic, with uncivil, irrelevant messages turning Americans off from politics and away from the polls. Crucially, they argue that both sets of effects will be concentrated among those least tolerant of negativity in politics.
Fridkin and Kenney employ a diverse array of research designs to test each aspect of their theory. Individuals’ tolerance for negativity is examined using survey data from the Cooperative Congressional Election Study (CCES). A thorough content analysis of ads aired in the 2014 US Senate elections reveals how negative ads vary in tone and substance. Perceptions of the civility and relevance in negative advertising are assessed with the help of focus groups and survey experiments; facial recognition software is used to look at people’s emotional reactions to negative ads. Finally, they match CCES respondents to their actual campaign environments to test their hypotheses about how negativity relates to people’s assessments of candidates and decisions to participate.
They find that individuals’ tolerance of negativity varies systematically with political and demographic characteristics. Women, older Americans, and the less politically interested tend to express much less tolerance. And though their earlier work suggests that Republicans are more tolerant of negativity than Democrats, they do not find a consistent relationship between tolerance and ideology or strength of partisanship. Interestingly, although there is variation in the civility and relevance of negative ads, those sponsored by outside groups in 2014 did not look markedly different from those sponsored by candidates or parties.
Their findings, however, are most compelling when they bring both pieces of their framework together. Using facial recognition software, for example, Fridkin and Kenney test individuals’ emotional reactions to campaign messages. The results are consistent with their theory: uncivil and relevant messages register the strongest emotional reactions, but only among subjects who scored low in tolerance for negativity. In their matched CCES analysis, they show that uncivil and relevant negative ads tend to lower evaluations of the attacked candidate, but only among Americans with a low tolerance for negativity; whether negative ads are uncivil, relevant, or both bears no statistically significant relationship to candidate evaluations among high-tolerance Americans.
Fridkin and Kenney’s central argument is that individuals low in tolerance for negativity will be most sensitive to the tone and substance of negative ads. Their turnout models are in line with that expectation: those who do not think that negative attacks are fair game tend to be less likely to go to the polls when they see lots of uncivil and irrelevant attack ads. But their findings on the persuasiveness of negative ads bear some additional consideration. If voters are not tolerant of these kinds of messages, why are they more (not less) likely to be persuaded by them? Alternatively, it could be that low-tolerance voters are more likely to punish sponsors of negative ads, thereby reducing the net effectiveness of negative campaigning. In previous work, Fridkin and Kenney, along with several others, have shown that negative ads can generate a backlash against the candidate who sponsors or is helped by the attack. Although low-tolerance Americans appear to give greater weight to shrill attacks about a candidate’s policy stands or legislative votes, Fridkin and Kenney do not test whether these potential voters also punish the candidate behind the attack.
If I have one critique of the empirical analyses in the book, it is that they have not taken full advantage of their experimental design (chap. 6) to assess how people react to ads that vary in civility. In an experiment embedded in the CCES, Fridkin and Kenney manipulated the storyboards of real negative ads. As expected, ads manipulated to be uncivil were generally rated as less civil (though they find no treatment effect on perceptions of an ad’s relevance). What remains untested, however, is whether their civility treatment has any effect on the outcome of interest—that is, individuals’ impressions of the candidates—or whether their treatment interacts with an individual’s level of tolerance for negativity, as their theory suggests. It is possible that their experimental treatment, which does not always pass the manipulation check, is too weak to detect any such effects.
This book provides a comprehensive examination of negative advertising, and its theoretical framework will likely be taken up by many in the field. In the end, Fridkin and Kenney concur that “America’s campaigns…do not fulfill the democratic ideal of generating thoughtful, reasoned, balanced, reflective, and prospective debate and discussion” (p. 160). Reviewing their results and the tenor of the 2016 presidential election in the conclusion, they suggest that we might be at a tipping point, at which negative advertising in future campaigns will do more harm than good. It is not clear that all readers will share that interpretation. If there is a silver lining here, it is that few negative ads can be characterized as pure mudslinging; indeed, their coders rated roughly 90% of negative ads aired in the 2014 Senate elections as relevant to voters’ worries or interests. But Americans generally disfavor negativity in politics, and it should be deeply disconcerting both that the less politically interested report even greater intolerance for negativity and that campaign rhetoric appears to do little to engage these potential voters in democratic life. Fridkin and Kenney have brought fresh insight and a wealth of data into debates about the consequences of negative campaigning for the quality of representative democracy. This book will be a great addition to graduate seminars and upper-division undergraduate courses on campaigns, elections, and research methods.
