The national news media's coverage of women and blacks at both the presidential and congressional levels continues to rely on racial and gender stereotypes, which results in distorted coverage of these officeholders. Media coverage of female and black members of Congress often reinforces stereotypical images, usually linking them to gendered or racialized issues (i.e., reproductive rights or domestic violence for female members or civil or voting rights for black members) (Carroll and Schreiber Reference Carroll, Schreiber and Norris1997; Gershon Reference Gershon2008a; Greco Larson and Andrade Reference Greco Larson and Andrade2005; Niven Reference Niven2004; Ward Reference Ward, Brown and Gershon2016a). While this is likely because of their perceived credibility on these issues (Swers Reference Swers2013), this stereotyping in effect pigeonholes minority and female legislators, which can “deny them credit for the breadth of their interests, efforts, and accomplishments” (Niven and Zilber Reference Niven and Zilber2001b, 402), reinforcing the ideas among voters that their identity is central to all of their work and that they do not have the same range of expertise and interests as their white, male colleagues, while also ignoring the latter's engagement on these issues.
Therefore, this study analyzes whether national news media coverage of House members, and presidential campaigns more broadly, continues to rely on stereotypes, but within the context of the 2008 presidential election. With prominent female and black candidates running for president and vice president that year, I analyze whether the heightened saliency of race and gender resulted in increased national news media coverage for female and/or black representatives. I argue that the 2008 election boosted the visibility of female representatives compared to 2004, but that the continued pattern of stereotypical coverage reinforced perceptions of female and black members as predominantly surrogate representatives of women and black voters. This was particularly true for black women, who are typically less likely to receive coverage than white women or minority men (Gershon Reference Gershon2012). With the national media's attention drawn toward conflict, members with perceived conflicts among their intersectional identities drew greater media attention, especially Republican and black women. The media's coverage of House members during the 2008 presidential election appeared to equate “gender” with “woman” and “race” with “black,” rather than engaging in these identities more broadly, and unnecessarily pitting black women's race and gender in opposition to one another.
GENDER, RACE, AND MEDIA COVERAGE
Female and minority candidates are often stereotyped in the media. Coverage of female and African-American presidential candidates often diminishes their viability, focuses on their race/gender and the novelty of their candidacy. For women, this often means the media focusing on personal traits, such as appearance or family, at the expense of political experience and qualifications and results in their being covered less overall (Carlin and Winfrey Reference Carlin and Winfrey2009; Falk Reference Falk2010; Heldman, Carroll and Olson Reference Heldman, Carroll and Olson2005; Kahn Reference Kahn1996; Reeves Reference Reeves1997). Specifically in 2008, coverage of Clinton was more gender-neutral in some ways, but coverage of her campaign was more negative than for Obama, sexism was ignored or downplayed by the media, and some journalists made overtly sexist remarks (Carroll Reference Carroll2009; Miller, Peake, and Boulton Reference Miller, Peake and Boulton2010). At the intersection of race and gender, Junn (Reference Junn2009) argues that 2008 had significant repercussions for women of color, as it solidified the default collective idea of woman as white and person of color as black.
Media coverage of female and black members of Congress demonstrate parallel findings; media coverage of these members tends to be less frequent and more negative than coverage of their white, male counterparts (Entman Reference Entman1994; Gershon Reference Gershon2008a, Reference Gershon2012; Kahn Reference Kahn1996; Niven Reference Niven2004; Schaffner and Gadson Reference Schaffner and Gadson2004; Ward Reference Ward, Brown and Gershon2016a, Reference Wardb; Zilber and Niven Reference Zilber and Niven2000a,Reference Zilber and Nivenb). When they draw the media's attention, coverage tends to focus disproportionately on a narrow set of stereotypical women's or minority issues, even though their self-presentation is more similar than different from their white male counterparts (Carroll and Schreiber Reference Carroll, Schreiber and Norris1997; Greco Larson and Andrade Reference Greco Larson and Andrade2005; Niven Reference Niven2004; Niven and Zilber Reference Niven and Zilber2001a; Zilber and Niven Reference Zilber and Niven2001). For example, in a follow-up to Cook's (Reference Cook1986) landmark study of members of Congress in the national news, Greco Larson and Andrade (Reference Greco Larson and Andrade2005) demonstrated that, unlike men, introducing more bills did not generate more coverage for female members, but introducing greater numbers of women's issue bills did. The media's attention more often turns to women or minority legislators if the story can be neatly packaged into a stereotypical category. This trend is amplified for minority women because of the “double barrier” of race and gender. Minority women receive less coverage than white women or minority men, and that coverage is also more negative (Gershon Reference Gershon2012, Ward Reference Ward, Brown and Gershon2016a).
One caveat to these negative findings is the context, which, if focused on race/gender, can draw the media's attention to female and minority members. For example, when major electoral issues coincide with women's or minorities' perceived expertise, such as social welfare issues, they may benefit (Dolan Reference Dolan2001; Huddy, Cassese, and Lizotte Reference Huddy, Cassese, Lizotte and Whittaker2008; Shapiro and Majahan Reference Shapiro and Majahan1986). Women newly elected in the “Year of the Woman” actually received more coverage than their male counterparts (Carroll and Schreiber Reference Carroll, Schreiber and Norris1997). For Rep. Mia Love (R-UT), media coverage highlighted the “double novelty” of her candidacy, but this emphasizes her otherness by focusing on the difference from the perceived norm of white and male (Ward Reference Ward, Brown and Gershon2016a, Reference Wardb). Therefore, the effects of gender and race on media coverage are conditional, depending on the saliency of gendered/racial issues or identities during a particular election. The 2008 presidential election was also fertile ground for media attention to race and gender identities (Carroll Reference Carroll2009; Holt Reference Holt2012; Junn Reference Junn2009). Therefore, the typical pattern of underrepresentation of women and minorities may have been reversed, and they may even be more likely to have received press attention.
Why appear on the national news? Members of Congress, especially women and minorities, can benefit from national media coverage because it raises their media profile, enabling them to be heard on national issues, helps them gain power within their caucus, and helps to advance their agenda (Ansolabehere, Behr, and Iyengar Reference Ansolabehere, Behr and Iyengar1993; Cook Reference Cook1989; Hasecke and Mycoff Reference Haseke and Mycoff2007; Hess Reference Hess1986; Sinclair Reference Sinclair1990; Swers Reference Swers2013). This study focuses on national news media appearances by members of Congress where they serve as reporters’ sources on the presidential election. Presidential elections garner significantly more national media coverage than Congress, so media interviews might raise a member's profile, and enhance standing in the party by demonstrating party loyalty, which, in turn, is linked to legislative success (Ansolabehere, Behr, and Iyengar Reference Ansolabehere, Behr and Iyengar1993; Groeling Reference Groeling2010). Members have an important stake in the outcome of the nomination process because of the president's agenda-setting power and influence on down-ballot races (Gronke, Koch, and Wilson Reference Gronke, Koch and Wilson2003; Lee Reference Lee2008; Sellers Reference Sellers and Oppenheimer2002). As prominent partisans, members of Congress can also influence the “early conversation” among activists and primary voters in search of elite cues, thereby increasing the candidate's chance of success (Cohen et al. Reference Cohen, Karol, Noel and Zaller2008). Overall there are a number of collective, individual, and electoral reasons members of Congress might be willing to be interviewed by a news show about the presidential race, even if it does not directly impact their reelection.
Typically, newsworthy sources in Congress include members of the party leadership, committee chairs, and senior members, who have greater influence, expertise, and connections and therefore greater credibility as sources (Cook Reference Cook1998; Hess Reference Hess1986; Schaffner and Sellers Reference Schaffner and Sellers2003). Since there is a limited pool of members considered newsworthy due to their institutional position, in order to be considered newsworthy, a member typically establishes himself/herself as a credible, authoritative source for a particular issue- or identity-based subconstituency (Kedrowski Reference Kedrowski1996; Loomis Reference Loomis1988), what Cook (Reference Cook1989) called “issue spokespersonship.” However, specifically in 2008, it may also be that with the heightened saliency of race and gender, female and minority members were sought after as sufficiently newsworthy by the national press and were therefore more likely to appear on the news. Journalists may seek out congresswomen as sources on women's issues or candidates, or black members as sources to speak about black candidates or issues, because they are perceived as authoritative voices with identity-based sources of credibility (Cook Reference Cook1986, 79; Kahn Reference Kahn1996; Niven Reference Niven2002; Swers Reference Swers2013; Zilber and Niven Reference Zilber and Niven2000a). This may be due in part because many black and female representatives also express a sense of surrogate representation—seeing themselves as speaking and acting for marginalized groups (Carroll Reference Carroll and Rosenthal2002; Gershon Reference Gershon2008b; Niven and Zilber Reference Niven and Zilber2001b). Black female legislators prioritize advocacy of both black and women's interests (Bratton, Haynie, and Reingold Reference Bratton, Haynie and Reingold2006). In the Senate, Democratic women often take a prominent role in debates on women's issues, either because of their own personal interest or because they are sought out by the party leadership as a credible voice on these issues (Swers Reference Swers2013). In 2008, the campaigns did anticipate media interest in race and gender by encouraging surrogates to speak out to demonstrate credibility with their surrogate subconstituencies or to respond to specific race- or gender-based attacks or issues (Sinclair-Chapman and Price Reference Sinclair-Chapman and Price2008).
Overall then, we should expect to find not only heightened interest in congressional women and blacks as sources, but also media reliance on stereotypes. In sum, in contrast to previous findings that female and minority representatives will receive less media attention, the heightened attention to raced/gendered identities, plus the potential for newsworthiness in 2008, leads to the following hypotheses:
H1a:
Female members of Congress appeared in the national news media more often than their male counterparts in 2008 but not in 2004.
H1b:
Black members of Congress appeared in the national news media more often than their nonblack counterparts in 2008 but not in 2004.
If the media's attention is drawn to newsworthy groups, those who may be perceived to have conflicts or cross-pressures among their identities would be of particular interest to the news media. One particular subtheme of the media coverage during the 2008 race was the “race versus gender” identity debate, particularly for black women, who were framed as having to “choose” one identity over the other when deciding between Clinton (“the female candidate”) and Obama (“the black candidate”) (Junn Reference Junn2009; Sinclair-Chapman and Price Reference Sinclair-Chapman and Price2008). Electoral discourse tended to reinforce “the already prevalent notion that women equals white and race equals black” (Junn, 106). Black women did not fit easily into this framework, so the tension perceived in the competition between their identities likely drew the attention of the media.
This is also complicated by the fact that many Democratic representatives endorsed one of the primary candidates, and as superdelegates were of interest to the media as they might ultimately decide the convention outcome. In 2008, members of Congress balanced reelection goals and district concerns, institutional influence, career goals, the dynamics of the nomination contest, and policy goals in their decision of whom to endorse (Anderson Reference Anderson2013; Hasecke, Meinke, and Scott Reference Hasecke, Meinke and Scott2012). However, unlike previous years, gender also predicted endorsement of Clinton, with white women more likely to endorse Clinton than white men (Galdieri, Parsneau, and Granberg-Rademacker Reference Galdieri, Parsneau and Granberg-Rademacker2012). Reflecting broader voting trends, though many black representatives endorsed Clinton early, black voters were more cohesive in their support for Obama than female voters were for Clinton (Dolan Reference Dolan2014; Huddy and Carey Reference Huddy and Carey2009; Simien Reference Simien and Crotty2009). Therefore, we might also expect greater media attention on black members who endorsed Clinton than on white female members of Congress who endorsed Obama, because the public might expect black members of Congress to support the black candidate regardless of gender.
Another group that may be perceived as having conflicting identities is Republican women, who may be more supportive of women's issues than their male colleagues, though this trend seems to be diminishing in recent Congresses (Frederick Reference Frederick2011). While Governor Palin's choice as the vice presidential candidate was likely an attempt to reach out to female voters, the gender gap favors Democrats (Huddy, Cassesse, and Lizotte Reference Huddy, Cassese, Lizotte and Whittaker2008). On the other hand, this means that the limited number of Republican female legislators are often asked to speak out and defend the party, especially on women's issues. On some issues, Republican party leaders turned to women in the House if there wasn't sufficient interest from more moderate female Republican senators (Swers Reference Swers2013). Therefore we might also expect the media to focus its attention on Republican women. Therefore, we expect the following:
H2:
Republican female representatives will be more likely to appear in the national news than their male Republican counterparts in 2008.
H3:
Black female representatives and black representatives who endorsed Clinton will be more likely to appear in the national news in 2008 than their Democratic counterparts.
Finally, it is likely the case that the high frequency of appearances by women and blacks in 2008 will be due mainly to the increased saliency of marginalized racial and gender identities due to the presence of non-white male candidates in such a high-profile presidential campaign. Although endorsements of these candidates may partly explain this spike in appearances, with stereotypical coverage by the media, we should expect black and female representatives to also be questioned about these candidates based on their perceived raced/gendered expertise. We should also expect media interest in gendered and racialized issues and identities when women and black representatives serve as sources (Carroll and Schreiber Reference Carroll, Schreiber and Norris1997). Therefore we predict the following:
H4a:
Democratic female representatives will be more likely to discuss Clinton and gender, regardless of which primary candidate they endorsed, than Democratic men.
H4b:
Republican female representatives will be more likely to discuss gender and Palin, regardless of which primary candidate they endorsed, than Republican men.
H4c:
Black representatives will be more likely to discuss Obama and race, regardless of which primary candidate they endorsed, than non-blacks.
DATA AND METHODS
The data for this project are counts of national media appearances for a sample of members of the House of Representatives. All of the women in Congress were included, as well as similar sized samples of male representatives. Overall, the 2008 samples included 70 women (10 black women) and 95 men (22 black men), and the 2004 samples included 60 women (11 black women) and 91 men (23 black men).Footnote 1 The data were collected for two separate years, from January of 2004 or 2008 until the day after the election of that year, both recent competitive presidential election years. There are two types of national news appearances included in the sample: interviews and direct quotes. An interview would include an in-person interview on national news, while direct quotes are replayed tape of members commenting on an issue or a campaign. Direct quotes are included in the analysis because many members were not often interviewed on national news programs, but a story featured a clip of their statement. Senators are generally more visible than House members and more regularly appear on national news programs to promote their personal agendas or as potential or actual presidential candidates (Hess Reference Hess1986), so as a consequence, the sample includes only House members, whose national media appearances are more likely to be indicators of national trends, rather than personal agendas.
Initially, each appearance was coded as to whether the representative addressed the presidential campaigns or candidates, or focused on an issue, or both. Also coded for 2008 was whether the member discussed one of the non-white male candidates, Obama, Clinton, or Palin. Each appearance was coded to indicate if the member of Congress specifically mentions women or women's roles, such as being a mother or trends among women voters, or issues that almost exclusively or directly affect women, and are stereotypical women's issues such as domestic violence or sexism. They were similarly coded if they specifically mentioned African-Americans or stereotypical issues related to discrimination against blacks or civil rights (Bratton, Haynie, and Reingold Reference Bratton, Haynie and Reingold2006).
The data were initially collected with the unit of analysis as each appearance but were then aggregated to sums for each member to explain which members had more appearances, and which factors best explain that behavior. The sample of national news programs includes television and radio programs, including cable and network news shows. Specifically, the sample includes the following: ABC, NBC, and CBS news (morning and evening news broadcasts), PBS shows Charlie Rose and The Newshour with Jim Lehrer, as well as news programs from MSNBC, CNBC, CNN, FOX, and NPR,Footnote 2 which was included because of its large number of national news shows. Using Lexis-Nexis, the search included each member's name as well as some variations using different titles. I searched for the representative's full name, and included possible alternatives such as “Representative Smith,” “Rep. Smith,” or “Congressman Smith.” The analysis employs negative binomial regression, which is appropriate for overdispersed count data.
For the multivariate analysis, the main independent variables of interest include whether the member is female and/or black.Footnote 3 To examine whether more influential members attract attention, three variables measure the prestige of members: leadership (member of the party leadership), chair/ranking member of a committee or subcommittee, and seniority (number of years served). However, it may also be that women and blacks are simply more interested in these types of issues. To examine this hypothesis, I also include a measure of women's issue bills, defined as the number of bills sponsored by a House member in areas traditionally associated with “care,” including education, health, women, children, and elderly issues (Greco Larson and Andrade Reference Greco Larson and Andrade2005), while racial issue bills are counts of sponsored bills that decrease racial discrimination and/or increase the socioeconomic status of minorities (Bratton and Haynie Reference Bratton and Haynie1999). Members from more liberal, urban areas may be more convenient television sources, and have constituents who pay close attention to major media outlets, so I control for the percent urban, and major media markets (districts in the top ten national media markets) (Schaffner and Gadson Reference Schaffner and Gadson2004). I measure the ideological lean of the district as the percentage the Democratic presidential candidate won in the previous election and an indicator of party (Democrats), since Democrats may have been more in demand during their longer primary season.
Although little research explores surrogate or horse-race appearances by members of Congress, a growing body of research has identified factors that explain why members endorse presidential candidates. In particular, electorally safe House members are more likely to involve themselves in the nomination process to claim credit for helping the party (Anderson Reference Anderson2013; Hasecke, Meinke, and Scott Reference Hasecke, Meinke and Scott2012). Therefore, the multivariate analysis includes a measure of the percent won by the member in the most recent election, indicating the competitiveness of the district. Members more active on legislation are also more likely to endorse a candidate (Anderson Reference Anderson2013), so I measure the total number of bills sponsored by a member. Similarly, if the member represents a state that will be contested in the general election, this may make them of interest to the national media, therefore battleground states are defined as having been reported by the Cook report as a tossup after Labor Day of 2004/2008 (Wolak Reference Wolak2006).
RESULTS
How often did members of Congress appear in the national news media in 2004 and 2008? Setting all independent variables at their modes, members of Congress appeared (in person or quoted) an average of eight times in this sample of national news shows, more in 2008 (9.9) than 2004 (7.8). However, there was a wide range: the median number of appearances is only two, and the modal category is zero, with 16% not mentioned on any of the sampled news shows in 2008, and 22% with no appearances in 2004. The speaker was the most quoted member; Nancy Pelosi (D-CA) and John Boehner (R-OH) had 200 and 198 appearances, Respectively, in the years they were speaker. Reflecting previous literature, members of Congress appear in the national media more often, even in an election year, to discuss policy issues. On average, members discussed policy issues in their appearances (5) more than the presidential campaign (3). There was a slight overall increase in the number of campaign-related appearances from 2004 (2.5) to 2008 (3.4), but the differences were not statistically significant (p = .310). Overall, congressional representatives were slightly more visible in the media's presidential coverage in 2008, but the typical member still did not appear often (if at all) in the national news media to discuss the presidential campaign.
Race, Gender, and Party in Media Visibility
Did the increased salience of race and gender identities in 2008 shape the number of media appearances by female and black members of Congress? Table 1 presents negative binomial regressions predicting the number of appearances discussing the presidential campaigns in 2004 (Model 1) and 2008 (Model 2). Comparing the first two columns, female representatives made nearly three times as many appearances on national news shows as male representatives in 2008 but were not significantly different from men in 2004. In contrast, black members in both 2004 and 2008 were significantly more likely than nonblack members to discuss the presidential race in the national news.Footnote 4 One potential reason the national news media's interest in speaking to black members about the presidential election was not limited to 2008, as it was for women, could be the salience of race in both presidential races, with black turnout a concern for Kerry in 2004, as the parties expanded their base mobilization efforts even compared to 2000 (Panagopoulos and Wielhouwer Reference Panagopoulos and Wielhouwer2008). Overall, there is support for the hypothesis that women appeared more in 2008 (H1a), but black members were visible on the media in both 2004 and 2008, so Hypothesis 1b is not completely supported. Both models also show that members of the party leadership teams were ten times more likely to appear on the national news to comment on the presidential race. Congressional party leadership messaging has become increasingly centralized and nationalized (Sellers Reference Sellers2010). These results suggest that messaging by members of the congressional party leadership extends beyond the institution to the party's broader electoral message as well, which includes the candidates at the top of the ticket.
Table 1. Negative binomial regressions predicting number of appearances discussing presidential campaign in 2004 and 2008
Notes: ***p < .001, **p < .01, *p < .05, two-tailed test. Standard errors in parentheses.
If women were more likely to appear in the news media in 2008, was this especially the case for black women and Republican women? The media's interest in conflict and drama might lead to an interest in intersectional identities seemingly in conflict, which suggests greater attention might be paid to black women and Republican women. To test this claim, the first column of Table 2 presents the predicted number of campaign appearances in 2008, supporting the claim that Republican women ( H 2 ) and black women ( H 3 ) appeared more often than their partisan and nonblack colleagues. First, it is important to note that the impact of the increased salience of gender was bipartisan; women from both parties had more predicted appearances than their male partisan counterparts. Overall, Democratic women had three more predicted appearances than Democratic men. However, Republican women had the highest number of predicted appearances; ten more than Republican men, a much larger gap than among Democrats.
Table 2. Predicted number of appearances discussing presidential campaign in 2008
Notes: ***=p < .001, **=p < .01, *=p < .05, +=.10, two-tailed test. Standard errors in parentheses. Effects calculated using margins command in Stata.
Second, the combined effects of race and gender identities for black women also led to more predicted appearances than any of their Democratic counterparts, including black men, although the prediction is only significant at the p < .10 level. Black women were likely in demand by both media outlets because they could speak from personal experience regarding any tensions between their racial and gender identities. For example, in one interview, Representatives Elijah Cummings (D-MD) and Laura Richardson (D-CA), both African-Americans, were asked about the role that race played in their endorsement decision. When the host said to Rep. Richardson, “you've got both race and gender to consider” and asked “so how important was gender for you,” she responded, “Not only is it gender, it's race. I come from a bi-racial family … So if there is a poster person who you would think would be supporting Senator Obama, it would be me. But, again, we are talking about the next president of the United States.”Footnote 5
Both have an identity-based credibility due to their race, but for Rep. Richardson, the interviewer acknowledged her intersectional identity, echoing interest in the “race vs. sex” loyalty test ostensibly faced by black women who would have to choose between Clinton (sex/woman) and Obama (race/black) (Junn Reference Junn2009). It is also apparent she is trying to turn the focus to Clinton's qualifications, often lacking in coverage of female candidates (Falk Reference Falk2010). Another example of this was Rep. Sheila Jackson Lee (D-TX), who made several appearances on national news shows in support of Clinton's candidacy. She noted, “What I see in Hillary is history. And, yes, I'm an African-American woman and I frankly celebrate Senator Obama and Michelle Obama,” but she felt Clinton was more qualified.Footnote 6 Often, the format would be her paired with an African-American advocate for Obama, where Lee was symbolically able to counter the credibility of male African-American experts by embodying both marginalized categories — female and African American.Footnote 7 At the same time, her presence problematized this race vs. gender trope used to describe the presidential candidates because she embodied intersectionality that did not fit the default category of white woman or black male (Junn Reference Junn2009). Rep. Jackson Lee was called upon by the media for her embodied gender expertise but also at times confounded the media's race-vs.-gender frame, particularly by acknowledging this historic moment, but framing her argument around qualifications, not just race or gender.
By focusing on race and gender in member's endorsement decisions, the media gives the impression that they are primarily surrogate representatives choosing a candidate based on their race or gender. For black women gender identity was often subtly disregarded and racial identity perceived as the default by the media. For example, Katie Couric claimed, “That's why she's [Jackson Lee's] defying many of her African-American constituents [by supporting Clinton].”Footnote 8 It was often left to black representatives supporting Clinton to explain why they were not mirroring the black community in supporting Obama. In another example, in an interview with Rep. Emanuel Cleaver (D-MO) and Rep. Barbara Lee (D-CA), who supported Clinton and Obama, respectively, the interviewer asked Rep. Cleaver, “Your district went for Senator Obama. You are supporting Hillary Clinton. Do you feel comfortable with that?” Cleaver responded, “I have not heard anybody question Ted Kennedy or John Kerry about why they haven't switched from Senator Obama to Senator Clinton” even though Massachusetts voted for Clinton. He went on, “There are many white superdelegates who were in states that went for Senator Clinton. And, for some reason … all of the questions about responding to your district seem to be coming to the African-Americans.”Footnote 9 Both male and female black House members were asked by the media to defend what the media implies should make them uncomfortable—choosing the nonblack candidate. In this case, the assumption is that black representatives are primarily surrogate representatives of the black community. This demonstrates the extent to which the media assumes that in these interviews black and/or female members speak as surrogates for black and/or female voters, further minimizing the audience's perception of their range of interests, as well as the reasons these members might endorse a black or female candidate.
Role of Endorsements
In the previous examples, Representatives Jackson Lee and Richardson were likely media sources because they were black women supporting Hillary Clinton. This raises the issue of the role of endorsements in these trends. Is it simply that these representatives endorsed candidates and therefore took to the airwaves to support them? In addition to interest in black representatives supporting Clinton, white female members of Congress were early and strong supporters of her candidacy, so it is possible that Clinton endorsers were driving these appearances simply because they were acting as campaign surrogates for the more beleaguered candidate over the course of a long primary season (Galdieri, Parsneau, and Granberg-Rademacker Reference Galdieri, Parsneau and Granberg-Rademacker2012). To test this alternative, Model 3 in Table 1 includes an indicator of whether the member of Congress originally endorsed Clinton, compared to those who endorsed Obama, Edwards, or made no endorsement. Those who endorsed Clinton had nearly twice as many appearances. It is important to note however, that endorsements do not explain all of the increased visibility for women and blacks, since both variables are still significant and have similar incidence rate ratios to Model 2. Therefore, although endorsing Clinton translated into more national news media appearances, women and blacks were still more likely to appear regardless of candidate choice, so endorsements do not undermine support for the hypotheses presented in the previous section. Breaking down the predicted number of appearances by cohort in the final column of Table 2, in each case those who endorsed Clinton had more appearances than similar members who endorsed other or no candidates. However, women and blacks still had more appearances than their male/nonblack counterparts. Women who endorsed Clinton had 5.7 more appearances than men who endorsed Clinton, and blacks who endorsed Clinton appeared 2.9 more times than nonblacks who endorsed Clinton (p > .001). The results show the additive effects for black female members.
Role of Presidential Candidates
The most likely explanation for the greater national news media visibility by female and black representatives in 2008 is that they were asked to discuss the non-white male candidates in the race (Clinton, Obama, Palin). On the other hand, perhaps it was simply that black/female members were discussing other issues of interest, rather than the media targeting black and female members to comment on the nontraditional candidates. Table 3 presents negative binomial regressions for appearances where members discussed Obama, Clinton and Palin, and the results support H 4a and H 4b as women discussed the female candidates who shared their party identification to a greater extent than their fellow partisans. Model 1 demonstrates Democratic women and Clinton endorsers were particularly active in discussing the Clinton campaign. However, it is also important to note that female representatives discussed Clinton four times as often as men, even when controlling for endorsement. This suggests that women were asked about Clinton regardless of their particular support for her candidacy. For example, when Rep. Betty McCollum (D-MN), an Obama supporter, discussed the presidential race with Tucker Carlson, he began by noting “Obama's appeal to women, weaker than we thought.” Although McCollum defended Obama's record on women's issues, the interview turned to Clinton's campaign strategy on gender.Footnote 10 Even though it appears that Rep. McCollum was not seeking to highlight “the gender question” or Clinton, the host pursued the topic.
Table 3. Negative binomial regressions predicting number of appearances discussing Clinton, Palin, and Obama and pairing representatives with minorities and/or women
Notes: ***p < .001, **p < .01, *p < .05, two-tailed test. Standard errors in parentheses.
Model 2 presents the results for discussing Palin, revealing that only Republican women were more likely to discuss Palin. Being a member of the leadership is the only other variable that is significant. In fact, Republican women were the most active cohort among House members from the announcement of Palin as the vice presidential nominee to Election Day (results not shown). We might expect more hesitancy by Republican women, since in recent decades the Republican Party has been less sympathetic to gender-based claims (Wolbrecht Reference Wolbrecht2000). When Obama made a comment about putting “lipstick on a pig,” though, Palin supporters came to her defense. Rep. Mary Fallin (R-OK) noted on Hardball, “And by the way, you know, I've run for office for 18 years. I know, I can feel it in my spirit, I feel it in my heart, when someone's taking a dig at me as a woman.”Footnote 11 Rep. Candice Miller (R-MI) claimed, “The Obama campaign has decided that the way to get at Sarah Palin is through personal attacks and sexist insults. Senator Obama might find such jokes funny but women will only find them insulting.”Footnote 12 Of note, in many of these cases, these women were introduced as part of a “Palin Truth Squad.” This is similar to the way Senate Republican women are often called upon to defend the party's agenda or nominee from “anti-woman” charges from Democrats (Swers Reference Swers2013). While among Democrats the primary was an internal party battle and therefore there was no party mobilization of surrogates, it appears Republicans mobilized surrogates to support Palin.
Model 3 predicts appearances discussing Obama. Black members had three times the number of appearances compared to white members, although the predicted number of appearances for blacks just misses the conventional level of significance (p = .08). Overall, the results from Table 3 suggest that the media asked black representatives more about the black presidential candidate and female representatives about the female candidates of their party regardless of who they endorsed.
Role of Media Interest in Presidential Candidate Race and Gender
How then did this affect the substance of the news coverage? While female and black representatives might have been asked about the non-white male candidates more often, was there a link to race and gender in those discussions? It is clear from the examples above, women asked about Clinton or Palin were often asked specifically about gendered topics. Table 4 presents analyses of gender- and race-related appearances specifically about Obama and Clinton. The dependent variables here are only those appearances where Obama and race or Clinton and gender were discussed together. The results from the first column support the notion that race, gender, and party interacted in divergent ways in 2008. Female representatives were more likely to discuss gender and Clinton, as were members who endorsed Clinton. Also, being a member of the leadership is predictive of discussing Clinton and gender. However, this is likely driven by female House members in the leadership (i.e., Speaker Pelosi, Rep. Wasserman Schultz, Rep. McMorris Rodgers). Women in leadership had an average of 5.4 gender-based appearances, while men in leadership had only 1.6 (p = .001).
Table 4. Negative binomial regressions predicting appearances discussing gender, race
Notes: ***p < .001, **p < .01, *p < .05, two-tailed test. Standard errors in parentheses.
In the Obama-race model, black members were once again more likely to discuss both the candidate and race. However, in contrast to the Clinton model, those who endorsed Obama were less likely to do so. This may be partly due to the individual presidential campaign strategies, as Obama ran a more deracialized campaign, whereas Clinton eventually saw some advantages to emphasizing gender (Clayton Reference Clayton2010; Ifill Reference Ifill2009; Lawrence and Rose Reference Lawrence and Rose2010). Those who endorsed her may have felt less trepidation speaking about her gender in their support for her candidacy. However, the findings from Table 4 may also suggest a differential treatment of the media of race and gender in the campaign. Obama received significantly more positive coverage than Clinton throughout the primary (Miller, Peake, and Boulton Reference Miller, Peake and Boulton2010) and more favorable coverage than McCain and other recent presidential candidates (Farnsworth and Lichter Reference Farnsworth and Lichter2011). Indeed, in coding House members who discussed the Rev. Jeremiah Wright comments, only 7% of members did so, though more black than nonblack members (16% versus 5%). While this was a major story, it is surprising that so few members were asked about it. Overall, this suggests the media was asking black representatives, and not Obama supporters more generally, about the connection between the candidate and race. Meanwhile sexism directed at Clinton in the media, has been well documented (Carroll Reference Carroll2009). The media were less hesitant to pursue questions of Clinton's gender among all her supporters.
Table 4 does not include mentions of Palin's gender because of the small number of cases in the sample. However, 87% of Republican women's gender-related appearances mentioned Palin, compared to 78% of Democratic women's gender-related appearances mentioning Clinton (p < .001). Of the 33 appearances discussing Palin's gender in this sample, 64% were by Republican women, while nearly all of the remainder were by members of the party leadership. This affirms the importance of female Republican members speaking about Palin's gender, particularly her role as a mother. For example, Bill O'Reilly wanted to have Rep. Michele Bachmann (R-MN) on his show because she “ha[s] five children, and 23 foster children at one time or another, many of whom were special needs.”Footnote 13 Similarly, Rep. Cathy McMorris Rodgers, another mother of a special needs child, appeared on the CBS Early Show to answer the question, “Can a mother of five, including an infant with Down syndrome, be an effective vice president?”Footnote 14 Interviewers called upon their roles as mothers to connect with Palin's family experiences and juggling family and politics.
These trends can therefore be attributed to the media's interest in pursuing newsworthy stories about the race and/or gender of these presidential and vice presidential candidates. Was the media driving these trends? Stories were coded for whether the subject of the candidates’ race or gender was initiated by the interviewer, measured (for each member) as the percent of cases where the interviewer asked a question about race/gender, rather than the guest. Interestingly, questions of race/gender of the candidates were initiated by the media 60% of the time. So it was more often the case that the media was focusing on the presidential candidates’ race or gender, such as when female representatives were asked about sexism in the campaign. Wolf Blitzer asked Rep. Debbie Wasserman Schultz (D-FL) whether Clinton was a “victim of sexism in this campaign.” She replied, “I have gotten questions about how I can be a good mom and a good member of Congress. And no male candidate for Congress gets asked questions like that … and it's something that Hillary faced as well.”Footnote 15 Similarly, Rep. Heather Wilson (R-NM) pushed back against this line of inquiry to Gwen Ifill's question, “What advice do you give [Palin] on how to balance this? She's got five kids ….” Her reply: “That bothers me. No one ever asked John Kennedy whether he could be president and be a dad.”Footnote 16
However, the percentage of questions about the race or gender of the presidential candidates initiated by the media is significantly higher for black men (72%) and black women (79%), than it was for white women (49%) and white men (56%) (p = .043). This indicates that discussions of the presidential candidates race/gender were driven by the media, but even more so for black members. When race or gender came up in an interview, it was more likely to be initiated by the media when they were interviewing black representatives and, significantly, not simply with those who were Obama supporters. This supports the earlier point that for black representatives, and especially black women, the media was fascinated by their intersectional identity and novelty of what were portrayed as conflicting identities. On the other hand, the interest among white female representatives in invoking their role as a surrogate for women is also significant. If discussions of gender and race of the presidential candidates were least likely to be initiated by the media for white women; conversely, they were more likely to voluntarily discuss the candidate's ability to connect with women, to give voice to their concerns, or to bring up the candidates’ gendered experiences, such as being a mother.
CONCLUSION
The media continued its reliance on racial and gender stereotypes in its coverage of the 2008 presidential campaign (Carroll and Schreiber Reference Carroll, Schreiber and Norris1997; Greco Larson and Andrade 1995; Zilber and Niven Reference Zilber and Niven2000a). Data show that female representatives were more likely to speak about both gender and Clinton in 2008. Unlike white women, black members were asked more about race than their nonblack colleagues in both 2004 and 2008 and were therefore more likely than nonblack representatives to discuss race. Although this may typically marginalize female and black members of Congress, the media's interest in nonwhite male candidates at the presidential level boosted their visibility, especially for female representatives. Inversely, without a female or minority at the top of the ticket, female and minority members of Congress (and likely other surrogates) may become less visible. This was particularly true for black women, who are typically less likely to receive coverage than white women or minority men (Gershon Reference Gershon2012). In 2008, this double disadvantage was turned into an advantage in the amount of media coverage, as their “double novelty” (Ward Reference Ward2016b) resulted in black women having the highest predicted number of appearances of all cohorts.
In addition, when the media call on women or blacks because of their perceived race- or gender-based expertise to interpret the meaning of gender in presidential politics, this still reinforces seeing black and female members of Congress mainly as surrogate representatives of blacks and women, ignoring their range of interests. For black women, it also overemphasized the conflict between their racial and gendered identities. Another group with perceived conflicting intersectional identities, Republican women, was more likely to discuss Palin and those appearances included nearly all of the gender-based appearances by Republican women. Although Republican women may not typically have an advantage in discussing gender because of their party's lack of issue ownership on women's issues (Swers Reference Swers2013), in 2008 they were mobilized to defend the party's vice presidential nominee, defending Palin from sexism in the media and brandishing their credentials as mothers and experienced campaigners who had faced the same kinds of sexism and challenges. One potential implication of this could be that coverage may influence voter perceptions of the prominence of subgroups, like Republican women, in Congress, which, in turn, correlates with support for increasing women's representation (Sanbonmatsu Reference Sanbonmatsu2003).
This study has broader implications for our understanding of the ways the media employs stereotypes in coverage of the presidential campaigns. The 2008 election was certainly unique in many ways, and Sinclair-Chapman and Price (Reference Sinclair-Chapman and Price2008) showed that the campaigns strategically deployed surrogates to address the issues of race and gender that emerged, anticipating the media's interest, and these findings also suggest that Obama supporters spoke less about race than Clinton's about gender. However, the media's questions about Obama's race were focused on black members, while their questions on gender were not limited to female representatives, suggesting greater comfort with gender over race by interviewers.
More centrally, this study shows that discussions of race and gender were mainly aimed at female and black members. This trend contributes to the same kind of pigeonholing of race/gender issues in the presidential campaign, giving viewers the sense that only members of these groups can speak on these issues or are concerned with claims of sexism or racism. By linking women sources, female presidential candidates, and gendered issues, media coverage reinforces the novelty of female candidates for president and vice president and the idea that their expertise lies in these gendered areas and emerges from these identities. It also assumes their support for these presidential candidates is based on race or gender, not their qualifications. Finally, it ignores the engagement of others, like white men, in engaging with gendered and racialized issues. Journalists linking representatives and source credibility to these raced/gendered issues and identities continue to pigeonhole female and black presidential candidates, and women and black legislators into roles that diminish their broader agenda and expertise and potentially make them more susceptible to stereotyping by media and voters, which could have important consequences for their time in office should they be elected.
