Travis Ridout and Michael Franz's The Persuasive Power of Campaign Advertising presents a detailed look at the complex effects of political advertising. It is a comprehensive work that analyzes the effects of campaign advertising in Senate and Presidential races across several elections between 2000 and 2008. At the same time, Ridout and Franz do not profess to examine the many possible effects of political advertising—an impossible task. Rather, this book focuses on how political ads affect vote choice and candidate favorability.

The book is divided into eight chapters. The first three introduce the book, offer a concise overview of previous literature on the effects of political advertising, and provide details about the employed methods. The next three present mixed findings regarding how electoral context, advertising content, and receiver characteristics moderate the influence of political ads. The seventh chapter discusses advertisement exposure via the news media and the Internet and the final chapter summarizes the findings of the book and describes future challenges for political advertising researchers.

The authors take a survey-based approach, a fitting complement to existing experimental work on the effects of political advertisements. In doing so, the book joins a handful of others, most notably Richard Johnston's, Michael G. Hagen's, and Kathleen Hall Jamieson's The 2000 Presidential Election and the Foundations of Party Politics (2004), in analyzing how ads affect election outcomes by pairing market and survey data. The authors combine ad-tracking data from the Wisconsin Advertising Project with data from a host of panel surveys, including the National Annenberg Election Survey and the American National Election Studies, to investigate the influence of political ads. Using data from both sources provides the authors with a measure of their key independent variable, exposure to political advertising. For this measure, the authors multiply the number of ads shown in a media market with survey respondents' reported media exposure.

The book's main analyses are contained in Chapters 4 through 6. Each chapter tackles a different set of variables that may condition the effects of political ads on vote choice and candidate favorability. Chapter 4 looks at the campaign context, hypothesizing that the effects of political advertising will vary based on whether: (a) the race is for the presidency or a Senate seat, (b) there is an incumbent, (c) the campaign is competitive, and (d) it is early in the campaign season or closer to the election. The findings are mixed. Sometimes ad exposure affects vote choice and candidate favorability, and sometimes not. This holds true for all of the contextual variables (a-d). There are, however, some contextual variables that are more consistently associated with the effects of ad exposure. Based on the patterns of significance across elections and dependent variables, Ridout and Franz conclude that ads are more strongly related to vote choice and favorability when the candidates are less familiar to the electorate.

Chapter 5 shifts to ad content. In addition to the more traditional approach of examining ad tone (positive, negative, and contrast ads), Ridout and Franz also evaluate the emotions targeted by ads (fear, anger, and enthusiasm). Each measure of tone and emotion had a significant effect on vote choice in at least one instance. In 2000, for example, the probability of voting for Gore increased with exposure to positive Democratic ads. The result did not appear for Kerry in 2004, however. Regardless of the ad's tone and desired emotional response, the bulk of the significant findings are in the direction intended by the ad sponsor, with the ad increasing votes for the candidate promoted by the ad. In only a few instances do the authors find evidence of a backlash, whereby exposure to one party's ads sparks more votes/greater favorability for the opposing candidate. The results reveal that tone and emotions can matter—but the task of explaining why the various tones and emotions matter at different times requires additional research.

Chapter 6 moves to analyzing two individual attributes that should moderate the effects of political advertising: political knowledge and partisanship. Contrary to expectation, the politically knowledgeable and those less so are affected by political advertising. The results, however, varied. In some instances, only the politically knowledgeable were affected. In others, effects were confined to those with less political knowledge. Looking across elections and dependent variables, there were more significant relationships between ad exposure and vote choice/favorability for those low in political knowledge compared to those more equipped with political facts.

Turning to partisanship, the results for likeminded partisans and independents were as anticipated. Partisans agreeing with a candidate and those without partisan predispositions were more favorably inclined toward the candidate the more they were exposed to his/her advertising. The relationships were not, however, consistently significant. Democrats' exposure to Democratic Senate ads in 2004 increased the probability of voting for a Democrat, for example, but the same was not true in 2000. More intriguing is what happens with opposing partisans. Although Ridout and Franz suspected that ad exposure would not affect those opposed to the candidate promoted in an ad, several results are in the opposite direction. In fact, ad exposure was positively related to vote choice and candidate favorability among opposing partisans in several instances. The effects aren't huge—ads aren't converting Democrats to Republicans or vice versa—but several are significant.

More speculative than the previous three chapters, Chapter 7 turns to alternative means by which citizens may be exposed to political advertising. In addition to seeing a political ad aired during a commercial break, citizens can encounter political ads during news broadcasts and online. After documenting that political advertising was a frequent subject of news coverage, Ridout and Franz present an analysis showing that exposure to news coverage of political ads was unrelated to Senate vote choice in 2006. Appropriately cautious about the results, the authors note that they look only at the amount of ad-based news coverage, not the tone of the coverage. With respect to those searching for political ads online, the book suggests that selective exposure, where partisans are inclined to seek more likeminded ads than ads sponsored by the political opposition, may mute the effects of ads. The effects of “nano-targeting” (p. 141), where campaigns use sophisticated targeting algorithms to reach potential voters, seem ripe for additional analysis.

From a methods perspective, the book has several commendable features. Ridout and Franz should be applauded for analyzing data from several elections. Further, the combination of panel survey and ad-tracking data to determine exposure to ad content offers many benefits. Although the measure is unable to pinpoint precisely which ads a respondent may have seen, it seems to be a stronger approach than using market data or self-reported media exposure alone. The use of panel survey data complements experimental work on political advertising effects. Although internal validity is stronger for panel surveys than for cross-sectional surveys, some will quibble with drawing causal conclusions from these data. The mixed findings across elections, however, could be used as a starting point for experimentalists to examine the causal effects of political ads.

The Persuasive Power of Campaign Advertising makes a strong case that political advertising is related to vote choice and candidate favorability. Even more, the book demonstrates that campaign context, advertising content, and receiver characteristics play a role in explaining when political advertising works. Yet the book highlights the complexity of the effects of campaign advertising. Yes, advertising has an effect. But predictors of when the effect occurs, what is responsible for the effect, and who is affected are inconsistent. What works in one election may not in the next. As such, this book is a persuasive cautionary note against generalizing advertising effects found in one election to another and a productive starting point for those embarking on the difficult task of explaining why variability occurs.