In her book, Michele Claibourn links presidential campaigning and governing by arguing that citizens, by means of public opinion polls, hold newly elected presidents accountable for keeping or breaking promises made on the campaign trail. From the perspective of campaign scholars, this is a welcome argument because it offers another response to the question of why campaigns matter. It is also intriguing normatively because it suggests that the most important criterion for evaluating campaigns is whether or not the candidates are making their issue priorities clear.
In pursuing this argument, Claibourn makes a number of interesting moves. First, she offers a new theory of agenda accountability, arguing that most citizens do not remember the issue positions offered by candidates during campaigns but, instead, develop a sense of the issues about which candidates care the most. These “issue agendas” matter because they convey to voters the issues that the new candidate will tackle first upon entering office. She reminds us that in our system of “separate institutions sharing power,” the president's power to shape the legislative agenda is far greater than his ability to shape the specifics of legislation (p. 207). As a result, the author argues, understanding a candidate's issue agenda is more important than understanding a candidate's specific issue positions.
Next, Claibourn challenges the manner in which scholars of media effects view priming as a short-term factor in shaping candidate evaluations. She introduces the work of social psychologists who argue that priming can have long-term effects if an issue is frequently associated with a candidate and the issue is considered to be applicable—for example, if the issue is owned by the candidate's party (pp. 74–75). The author explains that people organize information around a “person node” and that our approach to political candidates is no different. When a candidate repeatedly mentions an issue during the campaign, we create a mental link between the candidate and the issue that is activated whenever we think about the candidate. Claibourn contends that these long-term priming effects carry over into a newly elected president's first year and enable citizens to hold him accountable for the campaign promises he made.
A strength of the book is that the author draws on an impressive array of data in her empirical analyses. To establish that candidates do indeed develop campaign issue agendas, she uses television advertising data to show that major party candidates in 1992, 2000, and 2008 did offer distinct issue agendas. She then establishes that citizens registered the candidates' issue agendas by using a national rolling cross-sectional survey administered by the University of Wisconsin Survey Center in 2000 (unfortunately, the National Annenberg Election Study did not include the measures she needed) in order to demonstrate that in media markets where candidate ads emphasized certain issues, citizens noticed. She uses advertising data again to make the case that long-term priming effects do indeed happen. Finally, in Chapters 6 and 7, she uses presidential remarks and legislative histories to determine which campaign issues presidents Bill Clinton, George W. Bush, and Barack Obama addressed during their first year in office and which ones they ignored. Finally, she turns to public opinion polls to show that the issues emphasized during their campaigns carried the greatest weight in presidential favorability ratings.
In general, Claibourn employs appropriate methods to analyze her data, but I found myself frustrated at times by the lack of robustness checks. For example, in Chapter 5 where she tries to show that long-term issue priming occurs, she uses the Wisconsin Advertising Project's coding of issue mentions in Gore and Bush ads during the 2000 campaign to create two measures of ad issue exposure: one that counts the total number of ads mentioning an issue in a respondent's media market up to the day before her interview and one that counts the number of ads mentioning the issue in the three days prior to the interview (p. 81). The author hypothesizes that the cumulative measure will have a greater effect on evaluations of candidates than the recent ad measure indicating that long-term priming does in fact occur. In her models, she does find the cumulative measure to have a stronger effect more often than not. The problem is that she seems to be unaware of the recent spate of studies showing that the effects of ads rapidly decay, sometimes in as little as two or three days (Alan S. Gerber, James G. Gimpel, Donald P. Green, and Daron R. Shaw, “The Influence of Television and Radio Advertising on Candidate Evaluations,” Unpublished. 2007; Seth J. Hill, James Lo, Lynn Vavreck, and John Zaller, “The Duration of Advertising Effects in the 2000 Presidential Campaign,” Paper prepared for the Annual Meeting of the American Political Science Association, 2008; Hill et al., “The Duration of Advertising Effects in Political Campaigns,” Paper prepared for the Annual Meeting of the American Political Science Association, 2007). Why then does her cumulative ad measure perform better than the recent ad measure? The reason may be that her models do not control for any other measure of campaign activity. She does not control for the number of days until the election or even the total number of ads in a respondent's media market, and so her measure of “tax” ads or “education” ads may be picking up the effects of other covariates. It is disappointing that she did not put in the extra legwork to really nail down her findings, especially given that they contradict some very solid recent work in the advertising-effects literature.
I was also unconvinced by Claibourn's key analyses showing that the issues emphasized during a president's candidacy are important in explaining presidential approval in the first year of office. In Chapter 6, she shows that voter approval of Bush on the four issues he emphasized during his campaign (education, social security, health care, and taxes) were strong predictors of presidential approval in his first year (at least before 9/11). The problem is that these issues are strong predictors of presidential approval in general, and so she has to make the case that their strong effect is at least in part a result of priming during the campaign. She tries to address this issue in Table 6.3 (p. 111) by showing that Bush's campaign issues had a stronger effect on presidential approval than other issues. In the table, we see that the economy and foreign affairs have a larger effect than the issues emphasized during the campaign, but Claibourn convincingly argues that this is because the economy is always a strong predictor of presidential approval (and reports on the economic slowdown did not come out until November 2001) and because Bush had successfully resolved his first foreign policy crisis, the Hainan Island incident. Yet we also see that voters' sense of how the president is doing on the environment, energy, the federal budget, and the defense budget all have an effect on overall presidential approval that is equal to or greater than some of the issues emphasized during the campaign. Claibourn tries to explain this pattern, but it struck me as involving more than a little hand waving.
Ultimately, the author's theory is seductive, but it seems to involve too many qualifications and caveats: Voters hold presidents accountable for their campaign issue agendas only in the first year of their presidency (those are the only years she examined), only when the issues are applicable to the candidate, only if they live in battleground states and are exposed to ads from the candidate, only if more recent events do not drown out the long-term priming effects, and so on. I do believe that Claibourn should be commended for her effort to bring the disparate fields of campaign effects and presidential studies together. More scholars should give attention to the relationship between campaigning and governing. Perhaps one of the most important lessons to emerge from the book is that if we are to study this connection, we need surveys that bridge the general election campaign and the first year of governing. Part of the reason that Claibourn's analysis is not as convincing as it could have been is that such data do not exist.
