How Voters Construct Images of Political Candidates: The Role of Political Advertising and Televised News

Abstract

The voting literature has belittled the impact of campaigns on electoral outcomes, focusing instead on the greater contribution of partisan alignment [Lazarsfeld, Berelson and Gaudet, 1948] and retrospective judgments of governmental performance [Fiorina, 1981] as key predictors of the vote. But some recent evidence (Hershey, in Pamper, 1989; Pfau and Kensky, 1990] and much popular press are devoted to the power of campaigns to sway the electorate. If there is agreement between these two schools of thought it centers on the impact of the campaign on marginal voters and on the role of the campaign in influencing candidate evaluations and highlighting resonant issues.

The extent to which public discourse during a campaign centers on issues favoring one candidate as opposed to another is considered a good predictor of recent election outcomes. A case in point is the success of the 1988 Bush presidential campaign in directing discourse to patriotism (the Pledge of Allegiance) and crime (prison furloughs, Willie Horton). The best kind of candidate issue, as illustrated by the "1988 campaign, is one that not only casts the sponsor in a good light ("Bush is tough on crime") but at the same time casts the opponent in a bad light ("Dukakis is soft on crime"). Emphasizing agenda variables over partisan predisposition or economic preordination, points to the active role of voters in campaign communication. issues are only important in the campaign agenda when they resonate with voters and shape their assessments of the candidate.

This research was designed to investigate the role of news and advertising in stimulating discourse about campaign issues and in the formation of candidate images. The vehicle for the study is a series of focus groups exposed to news and advertising broadcast during the 1990 Senate race in North Carolina.

This analysis of issue resonance and candidate image formation utilizes concepts drawn from schema theories of learning and memory, which increasingly have been tested experimentally [Crocker, Fiske and Taylor, 1984; Crockett, 1988; Fiske and Taylor, 1984; Graber, 1988; Reeves, Chaffee and Tims, 1982; Wicks, 1986 and 1990; Wicks and Drew, 1991 j. A schema is a "cognitive structure that represents organized knowledge about a given concept or type of stimulus" which enables individuals to integrate new bits of information [Fiske and Taylor, 1984, p. 139). So, for example, individuals who know that Jesse Helms is a Republican and Harvey Gantt is a Democrat, can employ party schemas in constructing mental images of the two candidates. They will expect the candidates to hold typical partisan views, and they will revise their candidate schemas as they receive new information which confirms or contradicts the images or the candidates they have provisionally constructed from party identification.

Our purpose here is not so much to show how schema aid in information acquisition, as to show the dynamics of schema construction. This should prove a useful enterprise from a theoretical standpoint, as schema theory is often criticized for taking the schemas as given and offering no explanation for their origin [Kuklinski et al, 1991; Wicks, in press]. In this exercise we take two groups of individuals who are essentially unfamiliar with the 1990 North Carolina Senate race and show them news and political advertising-from the climatic weeks of the campaign. We observe how individuals employ their values, beliefs, prior information and experiences (i.e. schemas) in elaborating new mini-schemas for Jesse Helms and Harvey Gantt.