Press Effects,Public Opinion,and Gender: Coverage of Sarah Palin’s Vice-Presidential Campaign

Abstract

The press had great potential to influence perceptions of Sarah Palin during the 2008 campaign, given her relative obscurity when picked to be the Republican vice-presidential nominee. Prior literature on press treatment of women running for national office suggests that Palin was likely to receive coverage that disadvantaged her due to her gender. We scrutinize press coverage of Palin’s campaign for evidence that she was treated differently and that these differences affected public opinion. Our sample of over 2,500 individual newspaper articles published during the campaign is paired with the 2008 National Annenberg Election Survey. We demonstrate that Palin’s coverage differed in significant ways from that of her male counterpart. Her gender, appearance, and family status were disproportionately mentioned in her coverage, and such mentions tended to dampen public opinion about her. In addition, the tone of Palin’s coverage was markedly negative, increasingly so over time, and was significantly related to reader opinion about her. While Palin was a unique candidate for national-level office, these findings should give pause to those concerned with equitable press treatment of women on the campaign trail.

Keywords

media effects presidential campaigns gender Sarah Palin public opinion

From early in her 2008 campaign for Vice-President of the United States, Sarah Palin accused the news media of bias against her. At the Republican National Convention, Palin proclaimed, “I’ve learned quickly . . . that if you’re not a member in good standing of the Washington elite, then some in the media consider a candidate unqualified for that reason alone.” Palin’s relationship with the media was certainly tumultuous. She routinely complained to delighted crowds who appeared to share her view that the “liberal” media wanted to see her fail.

Palin was rightly concerned about her press coverage, although for different reasons than she advanced. First, women running for national office typically receive “gendered” coverage detrimental to their campaigns. Women running for president consistently receive less press coverage than men, and it tends to be negative (Falk 2008; Heldman et al. 2005; see also Kahn 1994a, 1994b; Kahn and Goldenberg 1991). It also focuses on their appearance and perceived long-shot odds at the expense of their positions (Falk 2008; Heldman et al. 2005; Smooth 2006). While Hillary Clinton’s 2008 coverage broke some of these trends (Lawrence and Rose 2010; Miller et al. 2010), a little-known governor from Alaska might not be treated similarly.

Second, press influence over voter opinion is well documented (Jamieson and Waldman 2003; Kiousis 2005; Patterson 1993), and effects can be dramatic when women run. Experimental studies demonstrate that typical “women’s coverage” negatively influences voter sentiment toward any candidate, whether male or female (Kahn 1992, 1994a, 1996; see also Dayhoff 1983). Experiments specific to Palin further suggest that objectification of the governor—as might be found in her press coverage—depressed assessments of her competence (Heflick and Goldenberg 2009, 2011). Such effects in the real world might be especially pronounced since Palin was largely unknown, was only the second woman on a major party presidential ticket, and offered a unique gender performance on the campaign trail. The press was likely to cover Palin extensively and influence public opinion about her.

We scrutinize coverage of Palin’s campaign in seventeen leading U.S. newspapers. In keeping with prior studies, we compare Palin’s coverage to her male counterpart’s for evidence that her coverage followed typical “gendered” patterns. The evidence we find is mixed. Palin garnered voluminous coverage, but it was markedly negative. Mentions of her gender, appearance, and family status were disproportionate, but each was associated with positive, rather than negative, coverage. We also examine the temporal dynamics of Palin’s coverage and find that its tone soured over time. Finally, we demonstrate that Palin’s press coverage significantly influenced public opinion, and that mentions of her gender and appearance were detrimental.

We thus offer evidence from the real world of presidential politics, rather than the laboratory, that the differential treatment of Palin adversely affected voter opinion. While Palin’s candidacy and campaign were unique, her press coverage should give pause to those concerned with the equitable treatment of women seeking elective office.

A History of Gendered Coverage

Scholars concerned about women’s press coverage have focused on women running for president and women running for lower-level offices. Both types of studies largely assess newspaper coverage, although some work at the presidential level also examines coverage in other venues.¹

Historically, women running for president have suffered from a lack of coverage. Falk (2008) documents persistent bias in coverage afforded women running for president across three centuries. Women from Belva Lockwood to Patricia Schroeder fought for the little coverage they received. As recently as 2000, Elizabeth Dole received less coverage than men trailing her in the polls (Heldman et al. 2005). Such inequities appeared to dissipate in 2008 for Hillary Clinton. She received as much or more coverage than her male counterparts (Lawrence and Rose 2010; Miller et al. 2010; Project for Excellence in Journalism 2007).

Yet substance matters as much as quantity and presidential-level studies reveal that women suffer in terms of content. Since 1872 women presidential candidates have been consistently portrayed as unviable; press accounts have emphasized their appearance and gender while virtually ignoring their platforms (Falk 2008). Such coverage objectifies and trivializes women, treating them as novelties rather than legitimate candidates (Braden 1996; Falk 2008; Woodall and Fridkin 2007). Dole’s coverage emphasized her appearance, personal story, and gender (Aday and Devitt 2001; Heldman et al. 2005). Carol Moseley Braun was equated with Al Sharpton, who had never held elective office (Smooth 2006). While Clinton’s issue-oriented coverage was commensurate with her male counterparts’, disproportionate mentions of her gender were significantly associated with the open questioning of her electability (Miller et al. 2010). Negative coverage has also bedeviled women running for president, with Dole and Clinton as two recent examples (Heldman et al. 2005; Lawrence and Rose 2010; Miller et al. 2010).

Women running for lower-level offices have endured similar disparities in coverage amount and content—though each may be on the wane. Early studies of campaigns for U.S. House of Representatives, Senate, and governor found that women received significantly less coverage than men (Kahn 1992, 1994b, 1996; Kahn and Goldenberg 1991). More recent work suggests that such disparities may be coming into alignment. K. B. Smith (1997) found smaller differences in coverage volume in 1994 than prior studies using earlier data. Bystrom et al. (2004) found that women running for U.S. Senate and governor garnered less coverage than men in 1998 but more in 2000.

In terms of substance, coverage of women running for subpresidential offices tends to disproportionately mention their appearance, personality, and family status (Braden 1996; Kahn 1996). Their coverage has also been disproportionately negative (Kahn 1994a, 1994b; Kahn and Goldenberg 1991). As with volume, however, there are some promising recent trends in coverage content. Press focus on the gender, marital, and parental status of women candidates seems to be on the wane (Kittilson and Fridkin 2008; although see Bystrom et al. 2004). Disproportionate negativity in women’s coverage has also largely disappeared at the subpresidential level (Banwart et al. 2003; Bystrom et al. 2004; Kittilson and Fridkin 2008).

This prior work illuminates the hurdles women face when seeking press coverage of their campaigns. At the subpresidential level, some of these hurdles seem to be disappearing. Yet it would be a mistake to assume that trends toward equanimity in coverage of races for governor, for instance, would smoothly translate to coverage of Palin. She was handpicked by John McCain—landing on the presidential ticket without first campaigning in the primaries and caucuses. More importantly, Palin’s was a unique story (first-term governor, mother of five with a special-needs infant) delivered in a way that complicated the feminine and masculine.

Palin rejected the proverbial pantsuit of women candidates, styled her hair in a feminine fashion, talked openly about her children, and showcased them at the convention (McGinley 2009). Gibson and Heyse’s explication of her convention speech reveals numerous markers of a maternal persona including a personal tone, incorporation of anecdotes, and a number of folksy colloquialisms. Yet Palin undermined this feminine/maternal style by lauding McCain as a masculine warrior while emasculating Obama for being a mere community organizer (Gibson and Heyse 2010).

Palin continued to present a complicated gender performance post-convention. She wore high heels, tailored jackets, and short skirts to campaign events, yet the campaign circulated images of Palin hunting and shouldering firearms. These brought to mind a horseback-riding Ronald Reagan or a woodchopping George W. Bush (Beail and Longworth 2013). Even the campaign’s portrayal of Palin as a “hockey mom” wed the feminine/maternal with stereotypically masculine traits like toughness and aggressiveness (Beail and Longworth 2013). Palin’s “pit bulls with lipstick” joke further fused the masculine and feminine. Rather than carefully negotiate the “double bind” women face—who risk appearing either too feminine or too masculine (Jamieson 1995)—Palin seemed to reject the bind altogether. She openly touted hyperfemininity and hypermasculinity.

As anyone who followed the 2008 campaign knows, Palin received voluminous coverage. Understandable, as she held each marker Ulbig (2010) associates with heightened press coverage of vice-presidential nominees: newness to the national scene, a distinctive characteristic, a unique shortcoming, and scandal.² The content of Palin’s coverage is somewhat more difficult to succinctly summarize. We thus test several hypotheses that spring from the literature on women’s presidential and subpresidential campaigns. In keeping with this literature, we compare Palin’s coverage to that of her male counterpart, Joe Biden.

Hypothesis 1: Palin’s coverage was more negative than Biden’s.

Hypothesis 2: Palin’s coverage was more likely than Biden’s to emphasize gender.

Hypothesis 3: Palin’s coverage was more likely than Biden’s to emphasize appearance.

Hypothesis 4: Palin’s coverage was more likely than Biden’s to emphasize family status.

The Potential for Press Influence

Not only do we expect Palin’s coverage to differ in predictable ways from Biden’s, we also expect it to be influential. Media coverage across a variety of outlets influences voter perceptions of candidates through the process of framing. Framing “involves selecting a few aspects of a perceived reality and connecting them together in a narrative that promotes a particular interpretation” (Entman 2010: 391).³ Framing can affect how individuals view issues and political actors, and can shape public opinion about candidates during an election (Scheufele and Tewksbury 2007). Studies across presidential campaigns suggest that how candidates are framed in the press influences public opinion about them (Jamieson and Waldman 2003; Johnston et al. 2004).⁴

The media’s role in framing candidate evaluations is likely to increase in low-information contexts. Voters tend to rely on information in media accounts when little is known about a candidate (Mutz 1998). Coverage of a little-known candidate for national office can thus have pronounced effects (Sears and Kosterman 1994; see also Ulbig 2010). McCain’s pick of Palin triggered this sort of low-information scenario. In a Gallup poll conducted the day after Palin’s selection, a majority of Americans did not know enough about her to offer an opinion (Newport 2008). Yet the public was eager to learn about Palin. Her speech at the GOP Convention drew 37.2 million viewers (U.S. News and World Report 2008) and 69.9 million watched her debate Biden, making theirs the “most widely viewed vice presidential debate in history” (Benoit and Henson 2009: 48).

The degree to which a candidate is framed positively versus negatively has further potential to influence voter evaluations (Entman 1989; Johnston et al. 2004; Shaw 1999). As Entman (2010: 392) argues, “One-sided framing emphasizes some elements and suppresses others in ways that encourage recipients to give attention and weight to evaluative attributes.” When negative information about a candidate accumulates, one-sided negative framing becomes more likely. Such negative “slant” is especially likely to influence public opinion when a consensus develops within the media (Entman 2010; see also Chong and Druckman 2007). Palin’s coverage appears to have played out in this fashion. It soured over time, as we show below.

The McCain campaign did not help matters. Its initial strategy of shielding Palin from the media “fueled press speculation about her competence” (Kenski et al. 2010). Subsequent interviews with ABC’s Charles Gibson and CBS’s Katie Couric further stoked suspicions, given her uneven and rambling performances (Kenski et al. 2010). A negative frame focusing on Palin’s perceived incompetence gained momentum. This focus on Palin’s alleged incompetence likely fed gender-based stereotypes—that women are less competent on “presidential” issues like defense and foreign policy, and less strong and decisive than men (e.g., Lawless 2004; Sapiro 1981–1982; Woodall and Fridkin 2007).

In part because gender was already salient in the 2008 campaign, conditions were ripe for gender stereotyping. The press had a long history of disproportionately mentioning the gender, appearance, and family status of women candidates. Cueing voters to think about Palin as an attractive woman, mother, and wife were likely to dampen opinion about her. The presidency, after all, is an “office predicated on masculinity” (Duerst-Lahti 2008: 733). Voters tend to associate stereotypically masculine traits like toughness and aggressiveness with the presidency—not stereotypically feminine traits like warmth and compassion (Heldman 2007; Huddy and Terkildsen 1993; Lawless 2004). The fact that Palin combined large doses of each blurs expectations about how such gendered coverage might influence public opinion. As Entman (2010: 398) puts it, “Palin . . . hearkened to the cultural archetype of the plucky, self-reliant pioneer woman who could raise a passel of kids—and handle a rifle to protect them.” She was overtly feminine and overtly masculine. We nevertheless anticipate that press emphasis on Palin’s gender, appearance, and family status depressed public opinion about her, in keeping with the prior literature (see Braden 1996; Falk 2008; Woodall and Fridkin 2007).

In sum, press coverage of candidates may prime particular frames of reference and stereotypes among voters that can influence their attitudes. Such effects were likely pronounced for Palin. We thus hypothesize:

Hypothesis 5: The tone of Palin’s coverage had a direct influence on reader opinion.

Hypothesis 6: Press mentions of Palin’s gender, appearance, and family status were inversely related to reader opinion about her.

Methodology

To test each hypothesis, we content-coded articles and editorials covering the 2008 presidential campaign from the top-circulating newspapers in each of thirteen battleground states. States were chosen based on preelection polling and/or closeness of recent presidential elections. Four leading newspapers of record were also included, for a total of seventeen newspapers (see Appendix A).⁵

We examine newspapers, rather than a cross section of media outlets, in part due to their availability and ease of comparisons across outlets.⁶ Newspapers also lend themselves readily to content analysis. Most importantly, we merged our press coverage data with the National Annenberg Election Survey (NAES) to estimate press effects, and all seventeen newspapers in our sample were among those reported as sources of campaign information by NAES respondents.

Focusing on newspapers limits our analysis somewhat, given the growing importance of the Internet, blogs, and cable television in intermedia agenda setting (e.g., Meraz 2011). Notwithstanding these concerns, in 2008 newspapers were a significant part of the media landscape with a sizable audience compared with other news outlets,⁷ suggesting their importance even in a period of declining circulation.⁸ Research further demonstrates that the slant of newspaper coverage can affect voter decisions and favorability toward U.S. Senate candidates (Druckman and Parkin 2005; Kahn and Kenney 2002) and president (e.g., Jamieson and Waldman 2003; Johnston et al. 2004). Moreover, Druckman (2005) suggests that newspaper coverage plays a greater role in informing the electorate than television news.

The data collection window spanned from August 29, 2008 (the day Palin was placed on the GOP ticket) through November 4, 2008 (Election Day).⁹ Articles and editorials had to be at least five sentences long and address the presidential race. To save on subscription costs and minimize coder burden, only weekday coverage was analyzed. All front-page articles were selected for the sample (n = 753). The remaining 1,839 articles and editorials were selected at random, bringing the total sample to 2,592. This represents 56 percent of the total population of articles published that met the selection criteria in these seventeen newspapers. For each paper, the proportion of articles in the sample mirrors the proportion of articles in the population.

Content coding was conducted by the authors, who were joined by a team of graduate and undergraduate coders.¹⁰ Specific information was coded about each candidate mentioned in each article: whether or not they were mentioned in the headline, its tone toward them (negative/neutral/positive), how many sentences were devoted to them, and whether their gender, clothing/appearance, parental and marital status were mentioned. The article’s overall tone toward each candidate (negative/neutral/positive) was also assessed. Tests of intercoder reliability suggest it was moderate to high.¹¹

The NAES (2008) is merged with our coverage data to determine whether press coverage influenced reader opinion about Palin. The NAES used random digit dialing and a national, rolling cross-section design. A total of 57,967 U.S. adult residents were interviewed daily between December 17, 2007 and November 3, 2008. The response rate was 19 percent.¹² We include only those 1,931 respondents interviewed on or after September 1 who reported reading one of the seventeen sampled newspapers listed in Appendix A.¹³

We model reader favorability toward Palin and assessments of her readiness to be president. Both are measured from 0 to 10; higher scores indicate greater favorability/readiness.¹⁴ Descriptive statistics for each appear in Appendix B.

Coverage Amount, Tone, and Visibility

Table 1 compares the volume and tone of Palin’s coverage to Biden’s. Palin’s coverage was clearly more extensive than Biden’s, in terms of headline and article mentions. Reader opinions are unlikely to be influenced if coverage is minimal, so Palin’s coverage satisfied a condition necessary for press effects. On balance, Palin’s coverage was also more negative than Biden’s. Forty-three percent of headlines mentioning Palin were negative, compared with 36 percent of Biden’s. Similarly, 35 percent of articles mentioning Palin reflected negatively on her, versus 15 percent of Biden’s.

Table 1.

Volume and Tone of Coverage—Palin versus Biden.

	Palin % (Except Difference)	Biden % (Except Difference)
Headline
Candidate mentioned^a	42.2***	23.2
Tone toward candidate^b
Positive	27.8	28.9
Negative	43.2^†	35.6
Neutral	29.0^†	35.6
Difference (positive − negative)	−15.4	−6.7
Article
Candidate mentioned^c	51.6***	23.8
Tone toward candidate^a
Positive	31.1	33.2
Negative	34.9***	15.2
Neutral	33.9***	51.6
Difference (positive − negative)	−3.8	+18.0

Percentages based on total articles in which candidate mentioned (Palin n = 1,109, Biden n = 585).

Percentages based on total headlines in which candidate mentioned (Palin n = 472, Biden n = 135).

Percentages based on total articles in sample beginning on date candidate was placed on ticket (Palin n = 2,148, Biden n = 2,449).

†

p < .10 (one-tailed) for difference between Palin and Biden. ***p < .001 (one-tailed) for difference between Palin and Biden.

The tone of a candidate’s coverage matters insofar as it is visible to readers. By this standard, Biden’s positive coverage likely mattered little, while Palin’s negative coverage mattered a lot. Figure 1 displays positive and negative visibility indexes for each candidate. Each index is the candidate’s number of positive/negative headlines divided by the total number of headlines. We focus on headlines because they are visible to even casual readers. In theory, scores can range from 0 (no positive/negative headlines for that candidate) to 1 (uniformly positive/negative headlines in every headline in the sample).

Figure 1.

Candidate visibility: Positive/negative headlines divided by total headlines.

While Palin’s press exposure was more visibly negative than Biden’s, it was also more visibly positive. Her proportion of positive headlines (28 percent) was slightly smaller than his (29 percent), but there were far more of them, so Palin’s positive visibility outstripped Biden’s. On balance, however, Table 1 and Figure 1 make clear that the press presented a relatively visible, negative portrait of Palin.

While Hypothesis 1 is supported, the question arises whether Palin’s negative coverage simply reflected the struggling state of the GOP ticket and/or negativity toward McCain. To check the first possibility, we examined the relationship between coverage tone and horse-race mentions. While McCain’s headline and article tone were strongly correlated with simultaneous mentions of the Republican ticket’s status in the polls, this was not true for Palin. Her negative coverage was not easily explained by the horse race.¹⁵ To check the second possibility, we compared the tone of Palin’s coverage with McCain’s, limiting the analysis to articles primarily about each (see Table 2).¹⁶ In terms of headline and overall article tone, articles primarily about McCain tended to reflect much less negatively on him than was true for Palin. The notion that Palin’s negative coverage reflected McCain’s is not supported.

Table 2.

Tone of Coverage—Palin versus McCain: Limited to Articles Primarily about Each Candidate.

	Positive (%)	Negative (%)	Neutral (%)	Difference (% Positive − % Negative)	n
Headline^a
Palin	28.8***	43.1*	28.1	−14.3	288
McCain	41.3	35.1	23.6	6.2	271
Article^b
Palin	36.2**	47.8***	16.0	−11.6	312
McCain	46.9	34.1	19.0	12.8	311

Headline percentages calculated based on total articles mentioning each candidate that were primarily about them.

Article percentages calculated based on total articles that were primarily about the candidate.

p < .05 (one-tailed) for difference between Palin and McCain. **p < .01 (one-tailed) for difference between Palin and McCain. ***p < .001 (one-tailed) for difference between Palin and McCain.

The Temporal Dynamics of Tone

Coverage tone is not static; it responds to events on the campaign trail. Figure 2 displays Palin’s article tone differential by week and clearly suggests some periodicity.¹⁷ Palin’s coverage skewed positive during the convention stage before trending neutral. It took a severe negative turn following Palin’s poor performance in a network television interview with Couric and a vice5 presidential debate in which she merely exceeded “basement-level expectations” (Kenski et al. 2010: 157).

Figure 2.

Article tone differential by week: % positive articles toward Palin – % negative articles toward Palin.

The data displayed in Figure 2 resemble that in Entman’s (2010) analysis of network evening news. According to Entman, news reports moved from focusing on Palin’s novelty and contributions to the GOP ticket to emphasizing her inexperience and lack of skill in the aftermath of shaky televised interviews with Gibson and Couric. Kenski et al. (2010) analyze NAES survey data and document a corresponding slide in voter assessments of Palin. Yet neither Entman nor Kenski et al. test the empirical relationship between coverage of Palin and public opinion. After examining specific, gendered aspects of Palin’s coverage, we do just that.

A Gendered Portrayal

Hypotheses 2, 3, and 4 posit disproportionate press attention to Palin’s gender, appearance, and family status. Each hypothesis is supported. Palin’s gender was mentioned six times more often than Biden’s (20.9 vs. 3.4 percent). Her clothing/appearance and parental status were mentioned four times as often (6.7 vs. 1.7 percent and 18.1 vs. 4.6 percent, respectively). The disparity was smallest, but still statistically significant, for mentions of marital status. Palin’s was mentioned twice as often (14.9 vs. 7.0 percent).¹⁸

Heightened focus on Palin’s clothing/appearance and parental status may have stemmed from scandals surrounding each. Yet the finding for clothing/appearance remains statistically significant when analysis is limited to the period prior to October 20, when the wardrobe scandal broke. Likewise, the finding for parental status remains statistically significant when analysis is limited to the period between September 5 and October 20—one week after news of her daughter’s pregnancy broke and just prior to the wardrobe scandal.

Press mentions of a female candidate’s gender, clothing/appearance, and family status have traditionally been thought to delegitimize her by treating her as novel, objectifying her, and identifying her with the private rather than the public sphere.¹⁹ It is thus puzzling that such mentions were associated with more positive coverage of Palin. Figure 3 demonstrates that articles mentioning Palin’s gender were significantly more likely to be positive in tone and suggest she strengthened the GOP ticket. Similar results were found for mentions of Palin’s clothing/appearance and parental status. The findings differ somewhat for mentions of Palin’s marital status, where effects were smaller and significant for article tone only.

Figure 3.

Positive article tone and strengthens ticket by gendering of Palin’s coverage.

We did not expect gendered coverage to be associated with positivity. Granted, we do not yet account for how such coverage was processed by readers. Well-intentioned mentions of the historic nature of Palin’s nomination, for instance, may have inadvertently suggested that Palin was different, novel, and thus less legitimate. It is to such possibilities we now turn.

Coverage Effects on Reader Opinion

To assess the impact of Palin’s press coverage on reader opinion about her, NAES respondents were assigned measures of coverage from the paper they reported reading most often for information about the campaign. Not all NAES respondents reported reading the paper daily, and we cannot be certain they read it on the day of their NAES interview. We thus use cumulative coverage measures up to the day each respondent was interviewed. The first is a generalized measure of tone: the cumulative positivity in headline visibility minus the cumulative negativity in headline visibility in the respondent’s favored newspaper (hereafter cumulative relative positivity in headline visibility). Three additional coverage measures record the cumulative percent of gender, clothing/appearance, and parental status mentions in the respondent’s favored newspaper.²⁰

We use the cumulative approach given variability in respondent reading habits. Using cumulative scores also makes sense theoretically, since voters are not “blank-slates,” wiping clean everything they have heard about a candidate on a daily or weekly basis. New information is filtered through predispositions, such as partisanship, but may also add incrementally to prior information (Chong and Druckman 2007; Graber 1988). The cumulative tone measure also allows us to account for the changing nature of Palin’s tone reported above.

Cumulative relative positivity in headline visibility and cumulative mentions of Palin’s gender, clothing/appearance, and parental status are used to predict favorability toward Palin and assessments of her readiness to be president (scored from low to high) in the models below. Alternative specifications of coverage tone do not change the findings.²¹ Each model includes respondent sex, race, age, education, partisanship, ideology, degree to which followed the campaign,²² and favorability toward McCain. The latter helps ensure that effects documented for Palin’s coverage are independent of favorability toward McCain. The “week” variable records when the respondent was interviewed (e.g., Week 1, 2, 3, etc.) to account for events unfolding on the campaign trail. Our approach allows for variance across newspapers, as well as variance within newspapers over time,²³ since the changing nature of the campaign and its coverage must be taken into account.

Endogeneity is an important concern since individuals may select a newspaper based on the expectation of favorable coverage toward a particular candidate or party. Controlling for factors such as partisanship and ideology mitigates the problem somewhat. Favorability toward McCain is controlled for the same reason. Given these controls and the temporal nature of the coverage variables, it is reasonable to conclude that if a significant, positive relationship exists between coverage and Palin’s ratings, then the coverage independently contributed to the ratings.

Effect of Coverage on Assessments of Palin

Table 3 displays ordinary least squares (OLS) regression models of favorability toward Palin and assessments of her readiness to be president using respondent attributes and coverage indicators from the respondent’s favored newspaper as predictors.²⁴

Table 3.

Predicting Attitudes Toward Palin Using Respondent Characteristics, Headline Visibility, and Gendering.

	Favorability Toward . . . (0 to 10)		Readiness to Be President . . . (0 to 10)
	b	SE	b	SE
Respondent attributes
Female	−0.101	0.095	0.100	0.092
African-American or black	0.283^†	0.190	0.178	0.185
Other nonwhite race	−0.163	0.184	0.188	0.180
Age	0.007*	0.003	0.012***	0.003
Education	−0.140***	0.023	−0.181***	0.023
Party ID (strong Democrat → strong Republican)	0.388***	0.033	0.310***	0.032
Ideology (liberal → conservative)	0.355***	0.054	0.321***	0.053
Following campaign (low → high)	−0.130^†	0.088	−0.213**	0.086
Favorability toward McCain (low → high)	0.597***	0.021	0.440***	0.021
Newspaper read by respondent
Headline visibility (cumulative relative positivity)	1.728**	0.556	1.358**	0.543
Palin’s gender mentioned (cumulative %)	−0.015**	0.006	−0.007	0.006
Palin’s clothing/appearance mentioned (cumulative %)	−0.009	0.012	−0.018^†	0.012
Palin’s parental status mentioned (cumulative %)	−0.001	0.008	−0.011^†	0.008
Week	−0.109***	0.023	−0.082***	0.023
Constant	1.017*	0.516	1.172*	0.506
Number of cases	1,732		1,719
Adjusted R²	.67		.57
Model F-statistic	247.606***		162.029***

Note. Dependent variables range from 0 (low) to 10 (high). Entries are slope coefficients (b) and standard errors (SEs) derived from ordinary least squares regressions. Authors’ newspaper data merged with National Annenberg Election Survey (2008) for purposes of this table.

†

p < .10. *p < .05. **p < .01. ***p < .001 (one-tailed).

Beginning with the findings for respondent attributes, it is noteworthy that respondent gender is insignificant in both models. While this is consistent with prior research (Carroll 2010; Dolan 2008), it clearly rebuts a key strategy of the McCain campaign: to attract women voters by placing a woman on the ticket.

Results for other respondent attributes are straightforward. Age is significant in both models, with older respondents rating Palin higher. Education has the opposite effect; better-educated respondents rated Palin less highly. Partisanship and ideology perform as expected, with Republicans and conservatives tending to rate Palin more highly. Respondents who followed the campaign more closely tended to rate Palin less highly, especially in terms of readiness. Favorability toward McCain is significant and positive in each model.

The variable “week” is negative and significant in both models, indicating that readers soured on Palin as the campaign progressed. In terms of effect size, Palin’s average favorability and readiness ratings were reduced by roughly 1 point on the 10-point favorability and readiness scales by the end of the tenth week.

Most importantly, differences in coverage are directly related to reader assessments of Palin. Table 3 indicates a significant, positive relationship between Palin’s cumulative relative positivity in headline visibility and ratings of her favorability and readiness to be president. Readers of newspapers that covered Palin more positively tended to rate her more highly. In terms of effect size, a one standard deviation increase in relative positivity yields a 0.15-point increase in favorability toward Palin and a 0.12-point increase in assessments of her readiness. The effects are small but statistically significant.²⁵ The results thus support Hypothesis 5.

The findings for mentions of Palin’s gender, clothing/appearance, and parental status are mixed. The more readers were exposed to press accounts mentioning Palin’s gender, the less favorably they felt toward her.²⁶ Gender mentions were not, however, related to assessments of Palin’s readiness. Contrarily, press mentions of Palin’s clothing/appearance and parental status did not affect favorability assessments, but they approached statistical significance in the readiness model. In both cases, the sign on the coefficients is negative, suggesting that opinion dampened when readers encountered press coverage that mentioned Palin’s clothing/appearance and status as a mother. We find mixed support for Hypothesis 6.

Effects over Time

As shown above, Palin’s coverage during the final weeks of the campaign was more negative than during the first weeks. Given the greater significance readers attribute to negative information (Baum and Groeling 2010; Cobb and Kuklinksi 1997; Lau 1985), as well as the cascading negativity in Palin’s press coverage, press effects may be more pronounced during the latter stage of the campaign. To test for this, we split respondents into those interviewed during Weeks 1 to 4 versus 5 to 10. We expect the effects of coverage tone to be more pronounced in the latter period. We are agnostic about expectations for the other coverage variables; the literature asserts that press mentions of gender, appearance, and family status are detrimental without regard to campaign stage (see Braden 1996; Falk 2008; Woodall and Fridkin 2007).

Table 4 displays OLS regressions on respondent favorability toward Palin and assessments of her readiness to be president, with separate models for Weeks 1 to 4 and 5 to 10. The independent predictors are identical to those in Table 3.

Table 4.

Predicting Palin Ratings Pre- and Post-Televised Interviews Using Respondent Characteristics, Headline Visibility, and Gendering.

	Favorability Toward . . .				Readiness to Be president . . .
	Weeks 1–4		Weeks 5–10		Weeks1–4		Weeks 5–10
	b	SE	b	SE	b	SE	b	SE
Respondent attributes
Female	−0.145	0.143	−0.084	0.127	0.152	0.136	0.040	0.126
African-American or black	0.161	0.276	0.387^†	0.265	0.292	0.264	0.044	0.263
Other nonwhite race	−0.274	0.291	−0.049	0.240	0.201	0.280	0.195	0.239
Age	0.008^†	0.005	0.006^†	0.004	0.010*	0.005	0.014***	0.004
Education	−0.142***	0.035	−0.128***	0.032	−0.156***	0.034	−0.195***	0.032
Party ID (strong Democrat → strong Republican)	0.432***	0.049	0.341***	0.046	0.313***	0.047	0.302***	0.046
Ideology (liberal → conservative)	0.287***	0.080	0.420***	0.074	0.302***	0.076	0.350***	0.074
Following campaign (low → high)	−0.110	0.132	−0.142	0.120	−0.309**	0.126	−0.120	0.119
Favorability toward McCain (low → high)	0.605***	0.032	0.591***	0.029	0.440***	0.031	0.439***	0.029
Newspaper read by respondent
Headline visibility toward candidate (cumulative relative positivity)	0.938^†	0.650	3.688**	1.249	1.083*	0.621	1.806^†	1.243
Palin’s gender mentioned (cumulative %)	−0.009	0.008	−0.022*	0.010	−0.005	0.007	−0.009	0.010
Palin’s clothing/appearance mentioned (cumulative %)	0.013	0.015	−0.053*	0.023	−0.003	0.014	−0.044*	0.023
Palin’s parental status mentioned (cumulative %)	−0.007	0.009	0.023	0.020	−0.014^†	0.009	−0.001	0.020
Week	−0.168**	0.068	−0.106*	0.048	−0.136*	0.065	−0.042	0.048
Constant	1.055^†	0.786	1.087	0.876	1.675*	0.756	0.534	0.875
Number of cases	794		937		784		934
Adjusted R²	.67		.66		.57		.56
Model F-statistic	118.039***		128.351***		76.412***		84.594***

Note. Dependent variable is R’s favorability toward Palin, ranging from 0 (low) to 10 (high). Entries are slope coefficients (b) and standard errors (SEs) derived from ordinary least squares regressions. Weeks 1–4 begin 9/1, 9/8, 9/15, and 9/22; Weeks 5–10 begin 9/29, 10/6, 10/13, 10/20, 10/27, and 11/3. Authors’ newspaper data merged with National Annenberg Election Survey (2008) for purposes of this table.

†

p < .10. *p < .05. **p < .01. ***p < .001 (one-tailed).

In terms of favorability toward Palin, the control variables perform similarly during both time periods. This is not the case for the coverage variables. Coverage tone merely approaches significance in the first four weeks, but becomes highly significant—with a larger effect size—in the latter six weeks when Palin’s coverage was more consistently negative. Substantively, the effect is substantial; a one standard deviation decrease in cumulative relative positivity in headline visibility leads to an estimated 0.41-point decrease in favorability toward Palin. Similarly, press mentions of Palin’s gender and clothing/appearance are not significant in predicting favorability toward her in the first stage but are significant (and negative) in the latter stage. A one standard deviation increase in gender and clothing/appearance mentions leads to corresponding decreases in favorability of 0.14 and 0.18 points, respectively. Mentions of Palin’s parental status are not significant in either stage.

In the readiness model, the findings for the control variables are basically consistent across time periods. Findings for the coverage variables are mixed and less consistent with expectations. The tone measure is predictive in the expected direction during the first four weeks, but becomes attenuated in the latter period. Mentions of Palin’s gender have no apparent effect on opinion related to her readiness during either time period. Mentions of clothing/appearance are negatively associated with assessments of Palin’s readiness to be president during the latter period only. Mentions of Palin’s parental status depress assessments of her readiness to be president during the early stage, but not the latter.²⁷

Discussion

Sarah Palin’s 2008 campaign offered a rare opportunity to assess how the press covered a woman running for national office, whether her coverage differed from that of her male counterpart, and how it affected public opinion. Drawing on gender, framing, and media effects theories, we hypothesized that Palin’s coverage would be disproportionately negative and would disproportionately emphasize her gender and specific aspects of it to her detriment in the public arena. We confirm and extend existing theory on press treatment of women and its effects on voter opinion. While we acknowledge the unique aspects of Palin and her campaign, our results should give pause to those concerned with the equitable treatment of women running for elective office.

Our findings extend those of Falk (2008), Heldman et al. (2005), Lawrence and Rose (2010), and Miller et al. (2010), who report that coverage of women running for president is disproportionately negative. Though Palin was running for vice-president, she was not immune from this trend. We tested two possible explanations for Palin’s disproportionately negative coverage, but neither found support. Her negative coverage was not easily explained by the faltering status of the GOP ticket. Nor did her negative coverage simply reflect McCain’s. It appears to be a further example of the disproportionately negative treatment women on the presidential trail receive from the press.

A third possibility merits consideration: Might Palin’s negative coverage reflect concern about her qualifications? While 40 percent of the articles that reflected negatively on Palin mentioned her qualifications (or lack thereof), a full 60 percent did not.²⁸ At best, Palin’s qualifications provide a partial explanation for the negativity in her coverage. We concur with Carlin and Winfrey (2009: 339) who assert that “Palin was inexperienced and naive much as Dan Quayle was but no one made sexist comments about him and related his lack of qualifications to sex role stereotypes.”

Palin’s gender, clothing/appearance, and family status were disproportionately mentioned in her coverage, as has been true of women seeking presidential office since the nineteenth century. But for Palin, these press mentions were associated with positive, rather than negative, coverage. One explanation is that reporters mentioned, say, Palin’s status as the “first woman on the GOP presidential ticket” in the context of a positive article about her.²⁹ Article tone would be positive, but the gender-based stereotypes that concern Carlin and Winfrey might nevertheless be triggered. We tested this possibility by pairing our press coverage data with NAES survey data.

Press mentions of Palin’s gender dampened favorability toward her and assessments of her readiness to be president. Moreover, the harmful effect on favorability became more pronounced as the campaign progressed. Press mentions of Palin’s clothing/appearance were also negatively associated with reader opinion about her; this effect was especially pronounced in the latter weeks of the campaign. Findings were less significant for mentions of Palin’s parental status, but were directionally consistent with expectations. Prior work asserting the harmful effects of media emphasis on female candidates’ gender, appearance, and family status is thus corroborated and extended here (see Braden 1996; Falk 2008; Woodall and Fridkin 2007).

Our finding on the harmful effects of press attention to clothing/appearance is especially notable, as it corroborates prior experimental work. Research demonstrates that female candidates who are physically attractive are rated lower by study subjects (Bowman 1985; Chiao et al. 2008). Thus, attractive female candidates are thought to be adversely affected when their appearance is emphasized by reporters. In 2008, Heflick and Goldenberg (2009, 2011) found that experimental subjects primed to think about Palin’s appearance rated her as significantly less competent and expressed less willingness to vote for the GOP ticket. Our pairing of press coverage with actual public opinion data brings this experimental result into the real world of presidential politics.

Similarly, our results on the temporal nature of Palin’s coverage tone—and its effects on reader opinion—corroborate and extend earlier work. The direct relationship we find between tone and reader opinion is consistent with extensive literature on media effects (e.g., Jamieson and Waldman 2003; Shaw 1999). But it also speaks to specific theory about the manner in which “one-sided framing” can cascade in the media, leading to more pronounced effects on public opinion (Entman 2010). Palin’s press coverage not only became more negative over time, it also became more influential. By pairing press coverage with public opinion data, we thus extend the work of Elis et al. (2010), Entman (2010), and Kenski et al. (2010).

While our study is limited to newspapers, analyses of other mediums corroborate our results. Studies of Palin’s coverage in a variety of media obtained similar results for tone (see Center for Media and Public Affairs 2009; Entman 2010; Project for Excellence in Journalism 2008a, 2008b).³⁰ And by studying newspapers, we may actually understate media effects. The newspapers we coded were doubtless subject to tighter editorial control than, say, talk radio or blogs. Heldman et al.’s (2009) comparison of Palin’s coverage in top-circulation newspapers to political blogs revealed more pronounced negativity and overt sexism on blogs than in newspapers. Likewise, analysis of Clinton’s presidential campaign coverage revealed offensive gendering among a variety of media, but especially online sources (Lawrence and Rose 2010). The fact that we document any effects on public opinion for Palin’s newspaper coverage may represent the proverbial tip of the iceberg.

Palin was a unique candidate on the campaign trail. She was only the second woman on a major party presidential ticket, a first-term governor, the mother of a special-needs infant, with a persona that uniquely combined the feminine and masculine. Not only that, but the public knew little about her and was eager for information. A perfect storm of conditions may have been ripe for press effects. Given these factors, the generalizability of the findings to other candidates and campaign contexts is an important consideration.

Finally, campaign agency should be taken into account. The McCain campaign launched Palin’s candidacy by showcasing her family and emphasizing her gender. This may have triggered some of the gendered coverage we document. We are unable to disentangle the exact source of the press effects we identify. Emphasis on Palin’s gender and family at the Republican National Convention likely structured how the press covered Palin, and may have triggered harmful gender stereotypes. Kenski et al. (2010) go further, asserting that the press selectively applied a negative, incompetence frame to Palin but not Biden. Such selective framing could have triggered the harmful gender-based stereotype that women are less fit to serve as president and commander-in-chief.³¹ Negative stereotyping, fed in part by Palin’s poor performance in network television interviews, “undermined Palin’s image as a knowledgeable politician prepared to be president if needed” (Graber 2010: 201). It also appears to have dampened the tone of Palin’s press coverage and public opinion toward her.

Our findings fit neatly into long-observed trends in press coverage of women running for national office, as well as those on media effects. We corroborate and extend theory on the effects of gendered coverage on women seeking elective office. Disproportionate negativity and emphasis on gender and related characteristics has harmful effects on public opinion. These effects can manifest even when some coverage appears positive, focused on a woman’s historic bid for high office. They can also accelerate over time as negative press attention accumulates. For women running for office for the first time—or running in constituencies where they are not well known—these results should therefore give pause. While Sarah Palin was a unique candidate in unique circumstances, she was neither the first nor the last fashionably dressed mother with young children to introduce herself to an electorate largely unfamiliar with her.

Footnotes

Appendix A

Acknowledgements

The authors wish to thank Kim Fridkin for comments on an early version of this manuscript, as well as three anonymous reviewers who provided very helpful comments about the submission. The authors also wish to acknowledge the students at Bowling Green State University who assisted with the coding and keying of data. Their contribution is especially appreciated.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a 2009 American Political Science Association Small Research Grant.

Notes

Author Biographies

Melissa K. Miller is an Associate Professor of Political Science at Bowling Green State University. Her research focuses on women in politics, voter behavior, and political participation in the United States. Her work appears in Public Opinion Quarterly, Political Research Quarterly, Politics & Gender, and PS: Political Science and Politics, among others.

Jeffrey S. Peake is a Professor of Political Science at Clemson University. His research focuses on the presidency, congress and the media and he is coauthor of Breaking through the Noise: Presidential Leadership, Public Opinion and the News Media (Stanford, 2011).

References

Aday

Sean

Devitt

James

. 2001. “Style Over Substance: Newspaper Coverage of Elizabeth Dole’s Presidential Bid.” International Journal of Press/Politics 6 (2): 52–73.

Banwart

Mary Christine

Bystrom

Dianne G.

Robertson

Terry A.

2003. “From the Primary to the General Election: A Comparative Analysis of Media Coverage of Candidates in Mixed-Gender 2000 Races for Governor and U.S. Senate.” American Behavioral Scientist 46 (5): 658–76.

Baum

Matthew A.

Groeling

Tim J.

2010. War Stories: The Causes and Consequences of Public Views of the War. Princeton: Princeton University Press.

Beail

Linda

Longworth

Rhonda Kinney

. 2013. Framing Sarah Palin: Pit Bulls, Puritans and Politics. New York: Routledge.

Benoit

William L.

Henson

Jayne R.

2009. “A Functional Analysis of the 2008 Vice Presidential Debate: Biden versus Palin.” Argumentation and Advocacy 46 (1): 39–50.

Bowman

Ann

. 1985. “Physical Attractiveness and Electability.” Women & Politics 4 (4): 55–65.

Braden

Maria

. 1996. Women Politicians and the Media. Lexington: University of Kentucky Press.

Bystrom

Dianne G.

Banwart

Mary Christine

Kaid

Lynda Lee

Robertson

Terry A.

2004. Gender and Candidate Communication: VideoStyle, WebStyle, NewsStyle. New York: Routledge.

Carlin

Diana B.

Winfrey

Kelly L.

2009. “Have You Come a Long Way, Baby? Hillary Clinton, Sarah Palin, and Sexism in 2008 Campaign Coverage.” Communication Studies 60 (4): 326–43.

10.

Carroll

Susan J.

2010. “Voting Choices: The Politics of the Gender Gap.” In Gender and Elections: Shaping the Future of American Politics, ed. Carroll

Susan J.

Fox

Richard L.

2nd Edition, 117–43. Cambridge: Cambridge University Press.

11.

Center for Media and Public Affairs. 2009. “Election Watch: Campaign 2008 Final.” Media Monitor 23 (1): 1–9. http://cmpa.com/pdf/media_monitor_jan_2009.pdf (accessed July 1, 2013).

12.

Center for the People and the Press. 2010. “Americans Spending More Time Following the News.” http://www.people-press.org/2010/09/12/section-1-watching-reading-and-listening-to-the-news/ (accessed July 1, 2013).

13.

Chiao

Joan Y.

Bowman

Nicholas E.

Gill

Harleen

. 2008. “The Political Gender Gap: Gender Bias in Facial Inferences that Predict Voting Behavior.” PLoS ONE 3:e3666. doi:10.1371/journal.pone.0003666

14.

Chong

Dennis

Druckman

James N.

2007. “Framing Theory.” Annual Review of Political Science 10:103–26.

15.

Cobb

Michael D.

Kuklinksi

James H.

1997. “Changing Minds: Political Arguments and Political Persuasion.” American Journal of Political Science 41 (1): 88–121.

16.

Dayhoff

Signe

. 1983. “Sexist Language and Person Perception: Evaluation of Candidates from Newspaper Articles.” Sex Roles 9 (4): 527–39.

17.

Dolan

Kathleen

. 2008. “Is There a ‘Gender Affinity Effect’ in American Politics? Information, Affect and Candidate Sex in U.S. House Elections.” Political Research Quarterly 61 (1): 79–89.

18.

Druckman

James N.

2005. “Media Matter: How Newspapers and Television News Cover Campaigns and Influence Voters.” Political Communication 22 (4): 463–81.

19.

Druckman

James N.

Parkin

Michael

. 2005. “How Editorial Slant Affects Voters.” The Journal of Politics 67 (4): 1030–49.

20.

Duerst-Lahti

Georgia

. 2008. “‘Seeing What Has Always Been’: Opening Study of the Presidency.” PS: Political Science & Politics 41 (4): 733–37.

21.

Elis

Roy D.

Hillygus

Sunshine

Nie

Norman

. 2010. “The Dynamics of Candidate Evaluations and Vote Choice in 2008: Looking to the Past or Future?” Electoral Studies 29 (4): 582–93.

22.

Entman

Robert M.

1989. “How the Media Affect What People Think: An Information Processing Approach.” The Journal of Politics 51 (2): 247–70.

23.

Entman

Robert M.

2010. “Media Framing Biases and Political Power: Explaining the Slant in News of Campaign 2008.” Journalism 11 (4): 389–408.

24.

Falk

Erika

. 2008. Women for President: Media Bias in Eight Campaigns. Chicago: University of Illinois Press.

25.

Gibson

Katie L.

Heyse

Amy L.

2010. “‘The Difference between a Hockey Mom and a Pit Bull’: Sarah Palin’s Faux Maternal Persona and Performance of Hegemonic Masculinity at the 2008 Republican National Convention.” Communication Quarterly 58 (3): 235–56.

26.

Graber

Doris A.

1988. Processing the News: How People Tame the Information Tide. 2nd Edition. New York: Longman.

27.

Graber

Doris A.

2010. Mass Media and American Politics. 8th Edition. Washington, D.C.: CQ Press.

28.

Hayes

Andrew F.

Krippendorff

Klaus

. 2007. “Answering the Call for a Standard Reliability Measure for Coding Data.” Communication Methods and Measures 1 (1): 77–89.

29.

Heflick

Nathan A.

Goldenberg

Jamie L.

2009. “Objectifying Sarah Palin: Evidence that Objectification Causes Women to Be Perceived As Less Fully Human.” Journal of Experimental Social Psychology 45 (3): 598–601.

30.

Heflick

Nathan A.

Goldenberg

Jamie L.

2011. “Sarah Palin, A Nation Object(fie)s: The Role of Appearance Focus in the 2008 U.S. Presidential Election.” Sex Roles 65 (3–4): 149–55.

31.

Heldman

Caroline

. 2007. “Cultural Barriers to a Female President in the United States.” In Rethinking Madam President: Are We Ready for a Woman in the White House? ed. Han

Lori Cox

Heldman

Caroline

, 17–42. Boulder: Lynne Rienner.

32.

Heldman

Caroline

Carrol

Susan J.

Olson

Stephanie

. 2005. “She Brought Only a Skirt.” Political Communication 22 (3): 315–35.

33.

Heldman

Caroline

Oliver

Sarah

Conroy

Meredith

. 2009. “From Ferraro to Palin: Sexism in Media Coverage of Vice Presidential Candidates.” Paper presented at the Annual Meeting of the American Political Science Association, Ontario, Canada.

34.

Huddy

Leonie

Terkildsen

Nayda

. 1993. “The Consequences of Gender Stereotypes for Women Candidates at Different Levels and Types of Office.” Political Research Quarterly 46 (3): 503–25.

35.

Iyengar

Shanto

. 1987. “Television News and Citizens’ Explanations of National Affairs.” American Political Science Review 81 (3): 815–32.

36.

Jamieson

Kathleen Hall

. 1995. Beyond the Double Bind: Women and Leadership. New York: Oxford University Press.

37.

Jamieson

Kathleen Hall

Waldman

Paul

. 2003. The Press Effect: Politicians, Journalists, and the Stories that Shape the Political World. New York: Oxford University Press.

38.

Johnston

Richard

Hagen

Michael G.

Jamieson

Kathleen Hall

. 2004. The 2000 Presidential Election and the Foundations of Party Politics. New York: Cambridge University Press.

39.

Kahn

Kim Fridkin

. 1992. “Does Being Male Help? An Investigation of the Effects of Candidate Gender and Campaign Coverage on Evaluations of U.S. Senate Candidates.” The Journal of Politics 54 (2): 497–517.

40.

Kahn

Kim Fridkin

. 1994a. “Does Gender Make a Difference? An Experimental Examination of Sex Stereotypes and Press Patterns in Statewide Campaigns.” American Journal of Political Science 38 (1): 162–95.

41.

Kahn

Kim Fridkin

. 1994b. “The Distorted Mirror: Press Coverage of Women Candidates for Statewide Office.” The Journal of Politics 56 (1): 154–73.

42.

Kahn

Kim Fridkin

. 1996. The Political Consequences of Being a Woman. New York: Columbia University Press.

43.

Kahn

Kim Fridkin

Goldenberg

Edie N.

1991. “Women Candidates in the News: An Examination of Gender Differences in U.S. Senate Campaign Coverage.” Public Opinion Quarterly 55 (2): 180–99.

44.

Kahn

Kim Fridkin

Kenney

Patrick J.

2002. “The Slant of the News: How Editorial Endorsements Influence Campaign Coverage and Citizens’ Views of Candidates.” American Political Science Review 96 (2): 381–94.

45.

Kenski

Kate

Hardy

Bruce W.

Jamieson

Kathleen Hall

. 2010. The Obama Victory: How Media, Money, and Message Shaped the 2008 Election. Oxford: Oxford University Press.

46.

Kiousis

Spiro

. 2005. “Compelling Arguments and Attitude Strength: Exploring the Impact of Second-level Agenda Setting on Public Opinion of Presidential Candidate Images.” International Journal of Press/Politics 10 (2): 3–27.

47.

Kittilson

Miki Caul

Fridkin

Kim

. 2008. “Gender, Candidate Portrayals and Election Campaigns: A Comparative Perspective.” Politics & Gender 4 (3): 371–92.

48.

Lau

Richard R.

1985. “Two Explanations for Negativity Effects in Political Behavior.” American Journal of Political Science 29 (1): 119–38.

49.

Lawless

Jennifer

. 2004. “Women, War, and Winning Elections: Gender Stereotyping in the Post-September 11th Era.” Political Research Quarterly 57 (3): 479–90.

50.

Lawrence

Regina G.

Rose

Melody

. 2010. Hillary Clinton’s Race for the White House: Gender Politics & the Media on the Campaign Trail. Boulder: Lynne Rienner.

51.

McGinley

Ann C.

2009. “Hillary Clinton, Sarah Palin, and Michelle Obama: Performing Gender, Race, and Class on the Campaign Trail.” Denver University Law Review 86 (Special Election Issue): 709–25.

52.

Meraz

Sharon

. 2011. “Using Time Series Analysis to Measure Intermedia Agenda-Setting Influence in Traditional Media and Political Blog Networks.” Journalism & Mass Communication Quarterly 88 (1): 176–94.

53.

Miller

Melissa K.

Peake

Jeffrey S.

Boulton

Brittany Anne

. 2010. “Testing the Saturday Night Live Hypothesis: Fairness and Bias in Newspaper Coverage of Hillary Clinton’s Presidential Campaign.” Politics & Gender 6 (2): 169–98.

54.

Mondak

Jeffrey J.

1995. Nothing to Read: Newspapers and Elections in a Social Experiment. Ann Arbor: University of Michigan Press.

55.

Mutz

Diana C.

1998. Impersonal Influence: How Perceptions of Mass Collectives Affect Political Attitudes. New York: Cambridge University Press.

56.

National Annenberg Election Survey. 2008. Phone Edition (NAES08-Phone). Philadelphia: Annenberg Public Policy Center, University of Pennsylvania.

57.

Newport

Frank

. 2008. Palin Unknown to Most Americans: Immediate Reaction on Par with Reaction to Biden. Princeton: Gallup. http://www.gallup.com/poll/109951/palin-unknown-most-americans.aspx (accessed July 1, 2013).

58.

Patterson

Thomas E.

1993. Out of Order. New York: Alfred A. Kopf.

59.

Prior

Markus.

2013. “Media and Political Polarization.” Annual Review of Political Science 16: 101-27.

60.

Project for Excellence in Journalism. 2007. “The Invisible Primary—Invisible No Longer: A First Look at the Coverage of the 2008 Presidential Campaign.” October 29. http://www.journalism.org/node/8187 (accessed July 1, 2013).

61.

Project for Excellence in Journalism. 2008a. “The Color of News: How Different Media Covered the General Election.” October 29. http://www.journalism.org/node/13436 (accessed July 1, 2013).

62.

Project for Excellence in Journalism. 2008b. “Winning the Media Campaign: How the Press Reported the 2008 Campaign.” October 22. http://www.journalism.org/node/13310 (accessed July 1, 2013).

63.

Project for Excellence in Journalism. 2012. “The State of the News Media.” March 19. http://stateofthemedia.org/ (accessed July 1, 2013).

64.

Sapiro

Virginia

. 1981–1982. “If US Senator Baker Were a Woman: An Experimental Study of Candidate Images.” Political Psychology 3 (1/2): 61–83.

65.

Scheufele

Dietram A.

Tewksbury

David

. 2007. “Framing, Agenda Setting, and Priming: The Evolution of Three Media Effects Models.” Journal of Communication 57 (1): 9–20.

66.

Sears

David O.

Kosterman

Rick

. 1994. “Mass Media and Political Persuasion.” In Persuasion: Psychological Insights and Perspectives, ed. Shavitt

Sharon

Brock

Timothy C.

(pp.251-78). New York: Prentice Hall.

67.

Shaw

Darren R.

1999. “The Impact of News Media Favorability and Candidate Events in Presidential Campaigns.” Political Communication 16 (2): 183–202.

68.

Smith

Aaron

. 2009. “The Internet’s Role in Campaign 2008.” PEW Internet & American Live Project. http://web.pewinternet.org/~/media/Files/Reports/2009/The_Internets_Role_in_Campaign_2008.pdf (accessed July 1, 2013).

69.

Smith

Kevin B.

1997. “When All’s Fair: Signs of Parity in Media Coverage of Female Candidates.” Political Communication 14 (1): 71–82.

70.

Smooth

Wendy

. 2006. “African American Women and Electoral Politics: Journeying from the Shadows to the Spotlight.” In Gender and Elections: Shaping the Future of American Politics, ed. Carroll

Susan J.

Fox

Richard L.

, 117–42. Cambridge: Cambridge University Press.

71.

Ulbig

Stacy G.

2010. “The Appeal of Second Bananas: The Impact of Vice Presidential Candidates on Presidential Vote Choice, Yesterday and Today.” American Politics Research 38 (2): 330–55.

72.

U.S. News and World Report. 2008. “Palin Changing Course of Race?” September 5, 2008. http://www.usnews.com/usnews/politics/bulletin/bulletin_080905.htm (accessed July 1, 2013).

73.

Woodall

Gina Serignese

Fridkin

Kim L.

2007. “Shaping Women’s Chances: Stereotypes and the Media.” In Rethinking Madam President: Are We Ready for a Woman in the White House? ed. Han

Lori Cox

Heldman

Caroline

, 69–86. Boulder: Lynne Rienner.