Previous research has demonstrated that in low-information elections voters compensate for a lack of information by taking informational shortcuts, based on candidate cues, to make voting decisions. To date, this research has focused on candidate party identification and incumbency cues. This article argues that candidate demographic cues, specifically race and gender, also play an influential role. Unlike past psychological research that focuses on potential voter bias against women or black candidates, this article examines the informational content of voters' stereotypes about women and black candidates, and how these stereotypes affect voting behavior. I use quasi-experimental data from the Los Angeles Times Poll to demonstrate that candidate gender and candidate race signal voters in two different ways. Voters stereotype candidates ideologically: women and black candidates are stereotyped as more liberal than the average white male. Voters also stereotype candidates on issues: black candidates are seen as more concerned with minority rights than whites; while women candidates are viewed as more dedicated to honest government. As a result, voters choose candidates for office based on how much they agree or disagree with the ideological and issue positions they attribute, through stereotyping, to candidates.
In the post-World War II era, there have been two significant changes in American elections which, combined, have created a void within scholarly research on American elections and voting behavior. The first change is that elections in the United States have become increasingly candidate-centered (Jacobson 1992; Wattenberg 1991). A decline in partisan identification has fostered a corresponding decline in strict party-line voting, resulting in increased importance of the candidate as a person, rather than as an agent of a political party, in elections. At the same time, an increasingly diverse candidate pool at all levels of electoral office has literally changed the face of elections. In what was once an arena dominated almost completely by the white male candidate, candidates of different backgrounds, races, and gender have started to make concerted and productive runs for political office.
These two trends have changed the nature of electoral politics. While candidates are more central to elections, aspects of the candidates themselves have changed. This suggests that while candidates have increasingly become the focus of elections, their changing demographic characteristics have also increasingly become the focus of voters' attention. For this reason, understanding how candidate demographic characteristics impact voting behavior has become important to understanding contemporary elections in the United States.
In the past decade and a half, research on candidate demographics has proliferated. Almost without exception, however, this research has focused on whether voters are reluctant to vote for specific types of candidates, namely blacks, women, and elderly candidates (among others: Eckstrand and Eckert 1981; Leeper 1991; Rosenwasser and Seale 1988; Sigelman and Sigelman 1984; Terkildsen 1993). Failing to find voter bias against such candidates, many of these studies conclude that candidate demographics do not influence voting behavior.
Past research on candidate personality characteristics (Miller, Wattenberg and Malanchuk 1986), however, suggests that the bias approach may be overly simplified. This research argues that voters judge candidates for political office the same way they judge people they meet every day While bias or prejudice is certainly a part of this, it is only a small part. As Lippmann (1922) originally reasoned, the most important method of judging the world around us, and the people in it, is through the process of stereotyping. Rather than being merely emotive, as bias is, stereotypes are content-based.
Research on informational shortcuts and low-information voting has adopted a stereotype approach with respect to partisan affiliation. …