This book begins with a series of quotes from scholars and media pundits arguing that negative political advertisements are uncivil, mean-spirited, emotion-laden distortions of the truth, which not only smear their target but also mislead voters and undermine the political process itself; in short, they are the scourge of democracy. Or so the conventional wisdom goes. But John Geer asks us to reject the starting assumption that attack ads are bad. In fact, he argues that democracy requires candidates to attack each other. His argument is simple but powerful: All advertisements, by their very nature, exaggerate the truth. But this point applies equally to negative and positive ads. We cannot expect political candidates to fairly discuss their weaknesses along with their strengths. That is the role of the opposition, and it can only be done by offering criticisms of the opposing candidate or party. The real questions, then, are 1) whether negative ads perform their role effectively, or instead distort and confuse the vote choices citizens must make; and 2) whether attack ads, in performing a role that seems to be required of them, nonetheless undermine the democratic political system of which they are a crucial part.

In Defense of Negativity is organized into seven chapters. After an introduction that previews many of the surprising points to be made in the book, the second chapter describes the extraordinary data collection that forms the basis of the analysis: a content analysis of the appeals included in virtually every (795 in total) televised ad from the major presidential candidates between 1960 and 2000. The strength of this data set is the incredible detail with which every ad is coded. The major unit of analysis is not the ad itself but the themes or appeals that are made in the ad—and there is an average of just over 12 made in every ad. The only important weakness of the data is also discussed—that we do not know how often the different ads were aired during the campaign—forcing Geer to assume that all ads were shown with equal frequency. While it is surely not true, I cannot imagine this assumption introducing any serious bias into the results. And as Geer argues, his focus is on the content of the appeals within the ads—on the messages the candidates want to send—rather than on estimating the effects of the ads on voters.

The author's theory—and preliminary evidence for it—is presented in the third chapter, which should be required reading for both scholars who study political campaigns and the journalists who try to cover them. Geer first establishes three criteria for judging whether ads inform the public: 1) Do they involve issues or policy? 2) Are claims backed by evidence? And 3) do they expose the ground on which the candidates disagree? He then argues that there is a fundamental asymmetry between positive and negative ads because candidates face different constraints when promoting themselves and attacking their opponent. In a nutshell, the argument is that while candidates are rarely challenged about claims they make about themselves, they feel compelled to provide evidence whenever they attack their opponent. This asymmetry leads to a difference in the content of the appeals typically made in positive and negative ads. It is very difficult to provide evidence that supports or disputes a claim that some candidate holds a particular value, but policy claims can often be supported (or challenged) by past behaviors (e.g., votes in Congress) or clear records of performance or accomplishment. This leads Geer to predict many more value appeals in positive ads but a preponderance of evidence-backed policy appeals in negative ads, a prediction that is clearly borne out by the evidence. These facts make negative ads generally superior to positive ads on all three criteria for informativeness, and seriously undermines the claims underlying any fears about negative ads weakening democracy.

Geer also argues that all candidates who make it to the presidential level are of high quality, leaving relatively few avenues for person-based attacks. This fact is not true for candidates for lower offices, however, leading him to predict many more character-based attacks in campaigns for lower offices than for the presidency. This chapter led me to think of several related hypotheses of my own. While neither one of us has any evidence for these predictions, this is just what good theories do: They inspire additional hypotheses, and they can be extended beyond the domain in which they were originally presented.

The next two chapters look specifically at the nature of character- and issue-based appeals in presidential ads. Geer shows that relatively few attacks at the presidential level are character-based, and that they have barely increased over time. Thus, the overall rise in negativity that has occurred since 1960 is due primarily to an increase in policy-based attacks backed by verifiable evidence—a trend which, now that the evidence is before us, should alarm few observers of the American political scene. Chapter 5 on issue-based appeals is another particularly strong one, including numerous theoretically driven hypothesis tests I do not have the space to describe, but which again led me to generate several additional hypotheses of my own, exactly what good theories should do.

The penultimate chapter takes a closer look at the 1988 presidential campaign, widely viewed as the most notoriously negative in recent history. Geer dispels several points of accepted truth about that campaign (which I have repeated regularly to my students), including the purported “facts” that it was unusually negative to begin with, and composed mostly of attacks from the Republicans. The final chapter is devoted to an attempt to find explanations for the observed rise in negativity, which he places firmly on the increasing polarization of the parties, who now clearly have much more to fight about than they did 40 years ago.

All told, this book has a great deal to recommend it for undergraduate and graduate students alike. This is what high-quality social science is all about. It should become required reading for all journalists and political pundits before the next round of federal elections.