Campaign Dynamics: The Race for Governor

Chapter 2. Campaigns and Candidates: What Should We Expect?

CHAPTER 2 Campaigns and Candidates: What Should We Expect? More than anything else, an election campaign is a learning process for both candidates and voters.Voters receive information about the candidates from the media, neighbors and coworkers, and directly from the candidates themselves. Voters also learn about the electorate as a whole through the publication of public-opinion polls. Candidates gather information about potential supporters through polls, focus groups, and cof fee-shop discussions as they hit the campaign trail. Candidates also learn more about each other during the campaign. Central to the unfolding of a campaign is the struggle between candidates to provide new information to voters. Candidates shape the informational context within which voters make their decisions by battling to influence what i salient to voters when they cast their ballots. Through this process, the content of campaigns influences voting behavio. To reach this conclusion, I blend into a single theoretical framework the motivations of voters, party activists, and candidates, the basic institutional structure of the electoral process, and a contextual view of voting behavior. Gubernatorial candidates are goal-seeking individuals. Like other political actors, they may be driven by the commitment to enact specific policies, th desire to obtain power or prestige, or the ambition to use an elective of fice as stepping-stone along a particular career path. Regardless of which is paramount , the achievement of any of these goals by elected officials depends fir and foremost on getting elected. Thus, this study begins with the same assumption with which Downs (1957) began: that candidates’ primary goal is winning elections. Voters are also goal oriented. Specificall , voters in a two-candidate race are assumed to vote for their more preferred candidate. To understand voting behavior, those factors that shape the preferences of voters and thus their voting behavior must be discovered. I am not concerned here with the psychological process by which voters synthesize information, predispositions, and past behavior into preferences. Instead, I focus on how the content of a gubernato8 rial campaign influences the relative importance of factors linked with a vote ’s preference for one candidate over another. Candidates and voters interact under these basic assumptions.The bulk of that interaction process takes place during the campaign, and it culminates on election day.1 Voters know that candidates want to win, candidates know that voters want to vote for their preferred candidate, and candidates know that their opponents also want to win. Thus, each participant has a basic understanding of the motivations of everyone involved in the process. In a world of complete and costless information, candidates and voters would reveal their preferences honestly. Candidates would choose which policy positions to advocate, and casting ballots would be free of costs. Political science has produced simple yet elegant theories based on such a model, but alas, gubernatorial elections do not take place in such a world. This situation makes theory building more complicated and analysis more cumbersome, but it also makes both processes more interesting and more fun. First, the assumption of complete and cost-free information is untenable. When pressed, it is difficult to define exactly what would constitute comple information, but both logic and observation suggest that it is not necessary to do so. Voters do not have complete information about candidates or other voters , and candidates, try as they may , never seem to have enough information about the electorate or each other . The simple fact of incomplete information alters the campaign process markedly, particularly for candidates. Incomplete information leads in part to uncertainty , and uncertainty encourages candidates to behave strategically during the campaign. Under complete information, candidates may attempt to misrepresent their strategy to their opponents or their policy stances to voters during the course of the campaign. Thus, some strategic behavior may take place. However , if voters have full information, they realize that candidates engage in such behavior , and voters will wait until candidates make their “last moves” before making a decision (Ordeshook 1992). If the policy choices open to a candidate fall on a single di mension (an assumption addressed later), and there are only two candidates, such strategic plays by the candidates will not alter the outcome of the election if voters have complete information: the median voter theorem articulated by Downs (1957) holds. In response to uncertainty, candidates test various campaign themes and alter their approaches to accommodate new information.They react to changes in the behavior of their opponents as well as to shifts in public opinion...