Press Coverage of Mayoral Candidates: The Role of Gender in News Reporting and Campaign Issue Speech
Atkeson, Lonna Rae, Krebs, Timothy B., Political Research Quarterly
Some research on gender bias in news coverage of political campaigns indicates that the media portray male and female candidates differently. Research to date, however, has focused only on elections to national or statewide offices, where confounding variables such as party, incumbency, and competitiveness are present. The authors resolve this problem by focusing their analysis of media campaign coverage on nonpartisan, open-seat, and competitive mayoral races. The authors' content analysis of press coverage in six mayoral elections suggests that press coverage is not biased in favor of male candidates. The authors, however, find that the presence of a woman on the ballot expands the range of issue coverage in local campaigns in ways favorable to perceived strengths of female candidates.
Keywords: gender; elections; candidates; urban politics; political psychology; stereotypes; mayors; media; campaigns
Because media coverage of electoral contests is the primary mechanism for informing citizens about political candidates and issues, the type of coverage provided candidates may play an important role in shaping voters' perceptions (Kahn 1994b). Indeed, research suggests that gender cues and stereotypes affect voting behavior (Paolino 1995; Kahn 1996; Dolan 1998; Sanbonmatsu 2002), especially in low-information contests (McDermott 1997; Plutzer and Zipp 1996; Dolan 1998). Women candidates are perceived as having more typical feminine traits and fewer masculine traits; being more liberal, Democratic, and feminist; and better able to handle so-called compassion issues such as health care, education, and women's rights issues. In contrast, male candidates are perceived as having more typical masculine traits and fewer feminine traits, being more conservative, and better able to handle issues such as foreign policy and crime (Rosenwasser and Seale 1988; Leeper 1991; Kahn 1994a; Koch 2000; McDermott 1998; Sanbonmatsu 2002). Male candidates are also seen as more electable (Leeper 1991; Sapiro 1981-1982).
If journalists carry the same prejudices as voters, such stereotyping may influence their reporting, leading to news analysis that favors male candidates over female candidates, thus affecting outcomes in elections involving male and female candidates (Kahn and Goldenberg 1991). More specifically, if journalists present candidates through a gendered lens, coverage may be biased in favor of male candidates who are stereotypically seen as having the appropriate traits, policy priorities, and leadership skills for public office (Koch 1999; Alexander and Anderson 1993; Huddy and Terkildsen 1993; Lawless 2004). Because of the apparent link between news coverage and electoral outcomes, it is important to examine whether bias exists in the press's treatment of male and female candidates.
Studies of gender media bias have focused on national and state gubernatorial contests (Kahn 1994a, 1994b, 1996; Kropf and Boiney 2001; Devitt 2002; Bystrom, Robertson, and Banwart, 2001; Smith 1997; Aday and Devitt 2001; Heith 2001; Heldman, Carroll, and Olson 2005). One of the difficulties in studying media bias in these elections is sorting out the influence of factors such as incumbency, party, and the competitiveness of the race. For example, if women receive less coverage than men, is it because they are running as challengers as opposed to incumbents, who are likely to receive more coverage because of their status? If coverage of women focuses on soft issues such as health care and education, as opposed to hard issues such as national defense and taxes, is it because of their gender or because the Democratic Party, the label under which most women run, is perceived to be stronger on these issues? Furthermore, if women receive coverage that makes them appear less electable, is it because they are women, or is it because they have too few resources to create a competitive contest? These factors are important confounding variables that may help produce bias when not adequately controlled. …