Abstract
The author examines why female politicians continue to be underrepresented in the press by measuring how structural inequalities, engagement in traditional and disruptive dialogue, and gender preferences influence the amount of press coverage U.S. House representatives receive. Drawing on a data set of Tweets, press releases, and news articles and transcripts related to the 114th House of Representatives’ investigations of the Iran deal and Planned Parenthood, the author uses negative binomial regression to test the effects of gender, engagement, and interactions of the two on the press coverage received by male and female House members. The results indicate that female House members’ underrepresentation in the media mirrors their underrepresentation in public office. These findings suggest that although political discourse and gender preferences may not be keeping women out of the media when covering gendered topics, getting more women in public office is likely to be a cumbersome challenge in itself.
Explaining why gender inequality persists despite women’s advances in many different arenas is of critical importance to sociology and gender scholars (Ridgeway 2011). A growing body of literature has documented persistent inequalities in educational (England and Li 2006), economic (Budig 2002), media (Shor et al. 2015), and political (Deason, Greenlee, and Langer 2015) arenas. Theories about why these inequalities persist often engage cultural and structural barriers as two interconnected explanations.
Not surprisingly, scholars have also documented persistent inequalities in political press coverage, an arena that may be influenced by structural barriers related to women’s underrepresentation in office (Baitinger 2015) and cultural barriers related to media producers’ gender preferences (De Swert and Hooghe 2010; Ward 2016) and politicians’ behavior (Fridkin and Kenney 2010). Analyses of gender differences in political press coverage continue to find that there are fewer female politicians in the media (Baitinger 2015) and that female politicians receive less coverage than their male colleagues (Dunaway et al. 2013; Heldman, Carroll, and Olson 2005; Luhiste and Banducci 2016). When female politicians are mentioned, research suggests that media producers contribute to inequalities through stereotyped or demeaning depictions of female politicians (Dolan 2010; Hayes 2011; Kittilson and Fridkin 2008; Miller, Peake, and Boulton 2010), negative bias (Miller et al. 2010), and differences in how male and female politicians are addressed (Uscinski and Goren 2010).
Although scholars have documented how representation in public office (Baitinger 2015) and editor and journalist gender preferences (De Swert and Hooghe 2010; Ward 2016) may affect female politicians’ underrepresentation in the press, previous literature is limited in two ways. First, little work has been done to untangle how underrepresentation in public office, editor and journalist gender preferences, and women’s engagement in traditional and social media dialogues may influence gender differences in press coverage. Second, previous scholarship has focused almost exclusively on press coverage received during political campaigns, which may overestimate the press coverage female politicians receive, an issue I address in the “Data” section. In short, more work is needed to determine why female politicians continue to be less visible in the media and to contribute to answering the increasingly urgent question in contemporary gender scholarship, Why does gender inequality persist?
In this article I examine why gender inequality persists in political press coverage by analyzing the contributions of three potential factors: structural inequalities, engagement in traditional and disruptive dialogue, and editor and journalist gender preferences. Drawing on samples of Tweets and press releases written by members of the 114th House of Representatives and newspaper articles and television transcripts reporting on the House’s investigations of the Iran deal and Planned Parenthood between July 1, 2015, and November 1, 2015, I test the influence of gender on the news coverage House members receive, taking their engagement in traditional and disruptive dialogue, gender stereotyping of news topics, and other key factors into account. Through this exploration, my research contributes to sociological theory concerned with the factors that contribute to the persistence of gender inequality in political communication.
Alternative Explanations: Gender Differences in Representation
The explanations scholars have developed for understanding why gender inequalities persist in political news coverage are useful for understanding why female House members are less visible in news discussing Planned Parenthood and the Iran deal. The three explanations tested in this article are structural inequalities (Shor et al. 2015), engagement in traditional and disruptive dialogue (Brescoll 2011; Fridkin and Kenney 2010), and editor and journalist gender preferences (De Swert and Hooghe 2010; Ward 2016). These explanations are particularly important because they account for politicians’ and news producers’ actions, which may be influenced by cultural status beliefs, as well as broader structural inequalities, which may affect representation in public office. Examining each explanation provides a comprehensive framework for untangling the different ways that these explanations may contribute to gender inequalities in political press coverage.
Underrepresentation in Public Office and the Media
One explanation for why gender inequality persists in political press coverage is that men outweigh women in politics, and news coverage mirrors societal-level inequalities (Baitinger 2015; Shor et al. 2015). In other words, if there are more men serving in public office than women, there will be more male politicians in the media than female politicians. Historical and contemporary measures of women’s representation in American politics confirm that the underrepresentation of women in public office is an ongoing problem (Deason et al. 2015; Dolan 2010; Kunovich and Paxton 2005). In the United States, women make up only 20 percent of the House of Representatives and 20 percent of the Senate. Although women’s representation has increased over time, it falls short compared with nations that have boosted women’s representation over recent years (Clayton and Zetterberg 2018). Baitinger (2015) supported the argument that female politicians are less prominent in television news coverage because there are fewer of them in office and in leadership positions. However, because newspaper and television are fundamentally different, with television offering a smaller news hole, additional research is necessary for determining if a similar trend is seen in newspaper coverage. The explanation that female politicians are underrepresented in television and newspaper coverage because there are fewer of them in public office is the basis for hypothesis 1.
Hypothesis 1: Male and female politicians’ representation in traditional media is proportional to their representation in public office.
Gendered Participation Disparities
Another explanation for why gender inequality persists in political press coverage is that journalists rely on political dialogue for newsworthy information, and gender stereotypes reduce female politicians’ engagement in the dialogue (Brescoll 2011; Fridkin and Kenney 2010). Research shows that journalists from national and local newspapers rely heavily on press releases when writing articles (Blyskal and Blyskal 1985; Walters and Walters 1992) and increasingly on social media platforms for news content (Hayes and Lawless 2016; Peterson 2012). Paradoxically, although bypassing traditional media may benefit marginalized groups that face inequities in traditional media (Evans and Hayes Clark 2016; Gershon 2012; Loiseau and Nowacka 2015; Vergeer 2015), women may limit their online communication because of gendered communication expectations related to how much women talk and what they talk about (Fridkin and Kenney 2010).
The persistence of gender inequality in how men and women are expected to communicate is well documented. Scholars have found that although men gain power from talking more, talking more than others negatively influences both men’s and women’s perceptions of women’s competence and suitability for leadership positions (Brescoll 2011). Similarly, men are rewarded for being aggressive, and women are criticized when their enthusiasm is misinterpreted as emotionality (Heath and Flynn 2015). As a result, women may strategically alter or limit their communication by monitoring the number and type of press releases and social media posts they publish (Vochocová 2018) and reducing their willingness to engage in self-promotion (Deason et al. 2015; Lawless and Fox 2010). The explanation that female politicians are underrepresented in television and newspaper coverage because they share fewer press releases and social media posts is the basis for hypothesis 2.
Hypothesis 2: There is a positive relationship between the number of press releases and social media posts male and female politicians write and their representation in traditional media on average.
Gender Preferences in Traditional Media
Another factor that may contribute to gender inequality in press coverage is news producers’ editorial and journalistic gender preferences (Shor et al. 2015). Scholars have found evidence of gender discrimination in the quantity and quality of coverage editors and journalists dedicate to female politicians, especially minority women (De Swert and Hooghe 2010; Heldman et al. 2005; Ward 2016). Women and feminine topics may receive less news coverage than their male counterparts (Carlin and Winfrey 2009; Dunaway et al. 2013; Luhiste and Banducci 2016), and depictions of women in traditional media may be stereotyped or demeaning (Dolan 2010; Hayes 2011; Huddy and Terkildsen 1993; Kittilson and Fridkin 2008; Miller et al. 2010). Scholars suggest that gender stereotypes may also influence journalists’ tendency to favor male candidates because they perceive men as possessing the appropriate characteristics and skills for serving in public office (Alexander and Anderson 1993; Lawless 2004), although several analyses of mayoral (Atkeson and Krebs 2008) and congressional (Hayes and Lawless 2015) elections suggest that gender bias does not influence journalists’ coverage of male and female political candidates.
Baitinger’s (2015) inclusion of leadership positions in her analysis suggests that gender inequality may also be tied to how established women are in public office. Serving in office for a longer period of time or holding formal leadership and committee positions may contribute to journalists’ decision to cover one House member over another (Baitinger 2015; Ridgeway and Walker 1995). Because women’s engagement in public office has risen recently and is still relatively low, men may have more of the status characteristics news producers are looking for and subsequently receive more coverage (Herman and Chomsky 1988). From this perspective, female politicians’ underrepresentation in the media may be because they are underrepresented in strategic political positions.
On the basis of research that documents news producers’ preferences related to gender and other status characteristics, it appears as though female politicians may be underrepresented in the media because of journalistic preferences that are captured by controls, committee membership and years served, and the residual, where discrimination may also be present. Although the remaining effects in the gender difference, residual, cannot be isolated, it will include other, unmeasured status-related characteristics as well as gender discrimination. The explanation that female politicians are underrepresented in television and newspaper coverage because of unmeasured status characteristics and gender discrimination is the basis for hypothesis 3.
Hypothesis 3: Male politicians receive greater news coverage in traditional media than female politicians after accounting for their representation in public office, online engagement, and other factors on average.
However, some research suggests that perceived gendered issue expertise influences editors’ and journalists’ decisions to mention male or female politicians (Major and Coleman 2008). Scholars studying gendered personality traits have found that women are expected to possess communal qualities, while men are expected to possess agentic qualities (Dolan 2014; Eagly 1987; Lawless 2004; Paul and Smith 2008). These gender-congruent personality traits contribute to differences in perceived status and gendered areas of expertise (Rudman, Glick, and Phelan 2008). Women’s communal nature is perceived as being well suited for social issues such as health, education, and childcare, while men’s rational nature is perceived as being well suited for high-stakes issues such as finance and security (Alexander and Andersen 1993; Deason et al. 2015; Dolan and Lynch 2014; Meeks 2012).
Because women are expected to be knowledgeable about social issues and men are expected to understand financial and security issues, traditional media may include and favor engagement by female politicians in media covering feminine issues and exclude female politicians and favor engagement by male politicians in media covering masculine issues.
Hypothesis 4: Female politicians receive greater news coverage in traditional media discussing “feminine issues” than male politicians after accounting for their representation in public office, online engagement, other factors, and an additional effect from writing press releases and Tweets about “feminine issues.”
Hypothesis 5: Male politicians receive greater news coverage in traditional media discussing “masculine issues” than female politicians after accounting for their representation in public office, online engagement, other factors, and an additional effect from writing press releases and Tweets about “masculine issues.”
Data
Data were collected on politicians, their engagement in traditional and disruptive media, and press coverage they received in traditional media to test the proposed hypotheses. Before beginning the data collection process, I selected two topics, Planned Parenthood and the Iran deal, that capture how politicians discuss and news producers cover a stereotypical “feminine” issue and a stereotypical “masculine” issue.
These specific topics were selected for two reasons: they took place during the same time and they generated a similar amount of public interest on Twitter. Defunding Planned Parenthood, a controversial and highly publicized women’s health issue, was a Republican initiative that began when the House of Representatives began investigating secretly recorded videos that allegedly contained evidence of the selling of fetal tissue and House members proposed the defunding of Planned Parenthood (Carmon 2016; U.S. House of Representatives 2015a). The Iran deal, a controversial and highly publicized security issue, was a Democratic initiative that began when the Obama administration proposed an agreement with Iran that lifted nuclear-related sanctions to prevent Iran from constructing a nuclear weapon (Korte 2016; U.S. House of Representatives 2015b).
The vote to defund Planned Parenthood took place on September 18, 2015, and the vote on the Iran deal took place on September 10, 2015. Looking at the Twitter conversation discussing Planned Parenthood and the Iran deal, it is evident that the conversations took place during the same time period, with conversations beginning and ending within days of each other, and were roughly similar in volume: 5.2 million Tweets were written about Planned Parenthood and 4.6 million Tweets about the Iran deal. After selecting Planned Parenthood and the Iran deal as my topics and July 1, 2015, to November, 1 2015, as my time period, I selected my sample of politicians, members of the 114th U.S. House of Representatives (2015–2017).
I selected the House of Representatives as my sample for two key reasons. First, the House of Representatives was directly involved in both of these issues. With regard to Planned Parenthood, several members of the House were directly involved in sponsoring a bill to defund Planned Parenthood, and this bill was then debated and voted on by the House (U.S. House of Representatives 2015a). Similarly, one House member was directly involved in sponsoring the Iran deal resolution, and this resolution was then debated and voted on by the House (U.S. House of Representatives 2015b). Thus, there was incentive for journalists to cover House members rather than other politicians when covering these two issues.
Second, the vast majority of studies on gender differences in press coverage analyze political campaigns (Atkeson and Krebs 2008; Carlin and Winfrey 2009; Gershon 2012; Hayes and Lawless 2015, 2016; Luhiste and Banducci 2016). Studying politicians when they are not running for office is important because press coverage of political campaigns, especially presidential campaigns, may be more likely to include female candidates than during nonelection periods because elections are newsworthy, and female candidates are particularly newsworthy because they are breaking norms by running for office (Meeks 2012). As a result, studies of campaign coverage may be overreporting the amount of press coverage female politicians normally receive.
Members serving in the 114th House of Representatives were selected because they were in office when the House was investigating Planned Parenthood and the Iran deal. Data were manually collected from the Biographical Directory to identify all members and collect demographic characteristics including sex, race, age, religion, party, congressional district and state, years served in the House, and committee memberships. Although there are many different committees in the House, three (Energy and Commerce, Judiciary, and Oversight) investigated Planned Parenthood (House Republicans 2016), and five (Judiciary, Oversight, Foreign Affairs, Financial Services, and Ways and Means) investigated the Iran deal (U.S. House of Representatives 2015b). My sample (n = 435) does not include the five delegates who represent the District of Columbia, the Virgin Islands, Guam, American Samoa, the Commonwealth of the Northern Mariana Islands, and Puerto Rico.
Politician-level descriptive statistics are provided in Table 1. Approximately 19 percent of House members are women, and 18 percent are people of color. House members’ age ranges from 31 to 86, with a mean of 58 years. While the number of years served ranges from 1 to 51, with a mean of 9.7. Democrats make up 43 percent of the sample, and House members identify with a number of different religious affiliations: Christian (56 percent), Catholic (32 percent), Jewish (5 percent), and other religions (6 percent). House members also serve on a variety of committees, including Energy and Commerce (12 percent), Judiciary (9 percent), Oversight (9 percent), Foreign Affairs (10 percent), Financial Services (14 percent), and Ways and Means (9 percent).
Sociodemographic Descriptive Statistics (n = 435).
After House members’ demographic information was collected, all of the press releases written by House members about Planned Parenthood and the Iran deal were collected to capture their engagement in traditional political dialogue. Press releases that were published between July 1, 2015, and November 1, 2015, and discussed Planned Parenthood and/or the Iran deal were collected from the media section of each House member’s official congressional website. A total of 1,254 press releases were downloaded and saved for analysis, 467 discussing Planned Parenthood and 787 discussing the Iran deal. On average, House members wrote approximately 1.1 press releases about Planned Parenthood and 1.8 press releases about the Iran deal.
Next, all of the Tweets House members wrote about Planned Parenthood and the Iran deal were collected to capture engagement in disruptive dialogue using ForSight, a commercial social media analytics platform constructed by Crimson Hexagon. This platform greatly expedites the data collection process and prevents errors that may occur during manual data collection. Tweets were collected if they contained one or more of my keywords, “Planned Parenthood,” “#plannedparenthood,” “Iran deal,” and “#IranDeal”; were written between July 1, 2015, and November 1, 2015; and were written by a House member’s official, verified Twitter account.
To ensure that all Tweets were written by the members of the 114th House of Representatives, Tweets were collected only from House members with verified Twitter accounts. Profiles designated as “verified accounts” ensure that they are run by the politician and/or his or her staff. Removing the Twitter accounts that could not be verified improved my findings because it prevented the inclusion of fake data and affected only a small percentage number of House members (3.7 percent).
Although 428 House members had Twitter accounts, only data from the 412 verified accounts were included in my sample. Additionally, if House members had two verified accounts, only data from the official account were included in the sample. In these cases (n = 25), House members had one official account that was identified by the presence of “rep” or “representative” in the Twitter handle. Removing the duplicate Twitter accounts improved my findings because it removed personal and, more frequently, campaign accounts that were inactive or were used for sending personal messages. As with nonverified accounts, the removal of unofficial verified accounts affected only a small percentage of House members with Twitter accounts (3.7 percent). Thus, the sample contains only House members’ verified official accounts (n = 412). A total of 3,894 Tweets were collected, 1,503 on Planned Parenthood and 2,410 on the Iran deal. House members wrote approximately 3.4 Tweets about Planned Parenthood and 5.6 Tweets about the Iran deal on average.
LexisNexis and NewsBank were used to collect newspaper articles and television transcripts that appeared between July 1, 2015, and November 1, 2015, and discussed Planned Parenthood and/or the Iran deal. To create a balanced sample of national media coverage of Planned Parenthood and the Iran deal and prevent bias that may result from relying on one media source, such as the New York Times, I collected data from six national media sources. Following the work of Bail (2012), I collected newspaper articles from liberal (the New York Times), centrist (USA Today), and conservative (The Washington Times) newspapers and transcripts from liberal (CBS), centrist (CNN), and conservative (Fox News) television programs. A total of 784 newspaper articles and 103 transcripts were collected; 381 articles and 33 transcripts discussing Planned Parenthood and 403 articles and 70 transcripts discussing the Iran deal. On average, House members received roughly .51 mentions in news discussing Planned Parenthood and .52 mentions in news discussing the Iran deal.
Methods
Content Analysis
Content analysis was used to analyze how gender, engagement, and politician characteristics influence press coverage in media covering Planned Parenthood and the Iran deal. All coding was completed manually by the researcher and a research assistant. Intercoder reliability was achieved by coding separately and reviewing small sections of coding completed by the researcher and research assistant. To ensure the accuracy of the coding procedure, automated searches were used to identify topics, keywords, and names to confirm that topic and mention had been coded accurately.
During the first stage of this process, every Tweet, press release, newspaper article, and television transcript in the data set was read and coded as “Planned Parenthood,” “Iran deal,” or “both,” depending on which topic was being discussed in each piece of content. After receiving coding instructions, the researcher and assistant coded 100 randomly selected Tweets independently, and after confirming 95 percent reliability, another 100 randomly selected Tweets were coded independently to reconfirm that the researcher and assistant reached 98 percent reliability. The assistant was responsible for coding the remaining Tweets. I used automated searches for related keywords to confirm that the Tweets had been accurately coded as “Planned Parenthood,” “Iran deal,” or “both.”
Next, all of the newspaper articles and television transcripts were read, and all mentions of House members were documented. For those articles in which multiple House members were mentioned, each member was documented separately. After receiving coding instructions, the researcher and assistant coded 20 randomly selected articles and 5 randomly selected transcripts independently, and after confirming 90 percent reliability, 30 randomly selected articles and 10 randomly selected transcripts were coded independently to reconfirm that the researcher and assistant maintained 95 percent reliability. The assistant was responsible for coding the remaining articles and transcripts. I used automated searches to confirm that all mentions of House members had been captured in the articles and transcripts. This step was not taken for Tweets and press releases, because the author information had already been collected during the data collection process.
Negative Binomial Analysis
After identifying the topics and mentions in my data, I used negative binomial regression to analyze the relationship between the number of mentions House members received in traditional media discussing Planned Parenthood or the Iran deal and representation, participation, and preferences. Negative binomial regression was used because of evidence of overdispersion (α > 1 and a likelihood ratio test with a p value < .001). My two primary dependent variables are the number of total mentions House members receive in traditional media discussing Planned Parenthood and the Iran deal. My independent variables include categorical variables for female Republicans, female Democrats, male Republicans, and male Democrats (the reference category), race (person of color = 1), age, years served in the House, and membership on relevant congressional committees (Energy and Commerce, Judiciary, Oversight, Foreign Affairs, Financial Services, and Ways and Means). I also include interaction terms that capture the additional effects House members may experience when they write Tweets and/or press releases.
Findings
Building on previous research, this analysis untangles three explanations for why gender inequality persists in the amount of press coverage male and female House members receive in news covering two specific issues, Planned Parenthood and the Iran deal. The three explanations tested are as follows: the media mirrors underrepresentation in public office, traditional and disruptive dialogue contributes to news coverage and women limit their engagement, and media-level gender preferences exist for male politicians, especially when covering masculine topics. The findings presented in the following sections suggest that female House members’ underrepresentation is due to underrepresentation in public office rather than gender differences in engagement or news producers’ gender preferences. Explaining why gender inequality persists in political media representation will aid efforts to combat persistent inequality in press coverage and contribute to sociological theory by isolating the factors that contribute to the persistence of gender inequality.
Evidence of Persistent Gender Inequality
Evidence for hypothesis 1, that women’s representation in the media is proportional to women’s representation in public office, appears when looking at the proportion of mentions male and female House members receive in the media collected for this analysis. A total of 784 newspaper articles and 103 transcripts were collected, and House members were mentioned in a total of 188 newspaper articles (96 on Planned Parenthood and 92 on the Iran deal) and 16 transcripts (9 on Planned Parenthood and 7 on the Iran deal). When these mentions are broken down by gender, there is a significant difference in the number of articles that mention male House members and articles that mention female House members, which can be seen in Table 2. Although male House members are mentioned in 167 articles and 15 television transcripts, female House members are mentioned in only 73 articles and 3 television transcripts. Interestingly, when these mentions are broken down by topic, the gender differences are relatively similar in media covering Planned Parenthood and the Iran deal.
Mentions Received by Male and Female House Members.
The difference in the number of articles and transcripts that mention House members provided support for hypothesis 1 because only 19 percent of the 114th House of Representatives was made up of women. Thus, the fact that women are present in 38.8 percent of the news articles covering Planned Parenthood and the Iran deal may actually be higher than expected if it is assumed that women will be present in 19 percent of articles covering these issues.
Looking specifically at the number of mentions each House member received, the proportion of mentions female House members receive is roughly similar to their presence in the House. Of the 427 newspaper mentions and 23 TV transcript mentions of individual House members, female House members received 22.5 percent of newspaper mentions and 17.4 percent of television transcript mentions, while male House members received 77.5 percent of newspaper mentions and 82.6 percent of TV transcript mentions. The gender inequality in press coverage supports hypothesis 1, that women’s representation in the media, 22 percent, is roughly proportional to women’s representation in public office, 19 percent. This observation is formally tested in the following sections, in which Planned Parenthood and Iran deal mentions are examined separately and two additional explanations for why there is gender inequality in media coverage are tested.
Planned Parenthood: Gender Inequality When Discussing a Feminine Topic
To uncover how Tweets and press releases may influence news coverage, I turn to the mentions of House members in media covering the investigation of Planned Parenthood. Male and female House members wrote roughly the same number of Tweets and press releases about Planned Parenthood. Male House members wrote 3.31 Tweets and 1.01 press releases on average, while female House members wrote 4.01 Tweets and 1.35 press releases on average. Although House members are engaging in both traditional and disruptive media, they are creating more content on Twitter on average.
The analysis in Table 3 provides partial support for hypothesis 2, that engaging in the political dialogue is positively related to mentions. There is a positive and statistically significant relationship between Tweets and mentions in models 1 and 2. In model 1, moving from zero to one Tweet is expected to increase mentions by .01, when holding all other variables at their means. Interestingly, there is no relationship between press releases and mentions. This finding supports research that has documented the increasingly important role that social media plays in political media and the proposition that social media may actually be on its way to replacing traditional communication techniques, like press releases, in the future (Hayes and Lawless 2016; Peterson 2012). That being said, press releases do play an interesting role in model 2, which is discussed further below.
Effect of Gender and Participation on Planned Parenthood Mentions (n = 435).
Note: Values in parentheses are standard errors.
p < .10. *p < .05. **p < .01. ***p < .001.
Additionally, the analysis provides no support for hypothesis 3, that men receive more coverage when controls are included in the model. Neither female Republicans nor female Democrats receive fewer mentions than the reference group, male Republicans, at a statistically significant level. This suggests that men are not benefiting from news producers’ gender preferences. However, years served and being a member of the House Oversight Committee both have a statistically significant positive effect on press coverage.
Although male House members may not be benefiting from news producers’ gender preferences, many may benefit from the preference for covering strategic actors because male House members are generally more established than female House members, many of whom have been elected to the House recently and may struggle to be nominated to serve on committees and in leadership positions. As a result, news producers’ preference for mentioning strategic actors may not be as benign as it first appears, because male House members reap greater rewards because of the ongoing structural imbalance of male and female representation in office.
Finally, hypothesis 4, that women receive more press coverage in media covering stereotypical women’s issues, can also be rejected. There are no statistically significant differences between the mentions received by female Democrats, female Republicans, or male Republicans, the reference group. The only difference is that male Democrats are expected to receive .20 fewer mentions than male Republicans, holding all other variables at their means. Additionally, there are no statistically significant results found when looking at the interactions between female House members and press releases and Tweets. Neither female Republican nor female Democrats experience an additional boost from writing Planned Parenthood Tweets in comparison with Republican men.
Interestingly, press releases written by male Democrats do have a positive effect on mentions, although it does not outweigh the negative effect showing that male Democrats are less likely to receive mentions in news discussing Planned Parenthood in comparison with male Republicans on average. Male Democrats are the only group that receives a boost from writing press releases about Planned Parenthood, a stereotypical feminine issue. This boost may be because journalists search for more traditional input from groups that are not at the center of the discussion. However, further research is needed to make sense of this unexpected finding.
The fact that female House members do not experience an additional effect for engaging in the dialogue surrounding Planned Parenthood suggests that journalists and editors do not favor content created by female House members when discussing feminine issues, an unexpected result. These models suggest that women’s underrepresentation in the media corresponds to their underrepresentation in public office rather than differences in political dialogue or gender preferences.
The Iran Deal: Explanations of Difference
To uncover how Tweets and press releases may influence news coverage of a stereotypical masculine issue, I turn to the mentions of House members in media covering the Iran deal. Men and women wrote roughly the same number of Tweets and press releases about the Iran deal. Male House members wrote 6.22 Tweets and 1.85 press releases on average, and female House members wrote 2.83 Tweets and 1.61 press releases on average. Although House members are engaging in both traditional and disruptive media, men and women are both creating more content on Twitter.
The analysis in Table 4 provides partial support for hypothesis 2, that engaging in the political dialogue is positively related to mentions. There is a positive and statistically significant relationship between Tweets and mentions in models 3 and 4. In model 3, moving from zero to one Tweet is expected to increase mentions by .006, when holding all other variables at their means. Interestingly, there is no relationship between press releases and mentions. This finding provides additional support for research that has documented the increasingly important role that disruptive media plays in political media and the proposition that social media may actually be on its way to replacing traditional communication techniques, such as press releases, in the future (Hayes and Lawless 2016; Peterson 2012). That being said, press releases do play an interesting role in model 4, which is discussed further below.
Effect of Gender and Participation on Iran Deal Mentions (n = 435).
Note: Values in parentheses are standard errors.
p < .05. **p < .01. ***p < .001.
Additionally, the analysis provides no support for hypothesis 3, that men receive more coverage when controls are included in the model. Neither female Republicans nor female Democrats receive fewer mentions than the reference group, male Republicans, at a statistically significant level. This suggest that men are not benefiting from news producers’ gender preferences. However, years served and being a member of the House Judiciary Committee or Foreign Affairs Committee have a statistically significant positive effect on press coverage.
Although male House members may not be benefiting from news producers’ gender preferences, many may benefit from the preference for covering strategic actors because male House members are generally more established than female House members, many of whom have been elected to the House recently and may struggle to be nominated to serve on committees and in leadership positions. As a result, news producers’ preference for mentioning strategic actors may not be as benign as it first appears, because male House members reap greater rewards because of the ongoing structural imbalance of male and female representation in office.
Finally, hypothesis 5, that men receive more press coverage in media covering stereotypical men’s issues, can also be rejected on the basis of the gender interaction terms’ limited statistical significance. There are no statistically significant differences between the mentions received by male Democrats, female Republicans, and male Republicans, the reference group. In model 3, female Democrats are actually expected to receive .53 more mentions than male Republicans in traditional media discussing the Iran deal, holding all other variables at their means. In model 4, it is surprising that female Democrats are the only group that receives an additional boost from writing press releases about the Iran deal, a stereotypical masculine issue. Similar to the positive and statistically significant interaction between male Democrats and Planned Parenthood press releases, this boost may be because journalists search for more traditional input from groups that are not at the center of the discussion. However, further research is needed to make sense of this unexpected finding.
These findings contradict the gender preferences argument, which suggests that men experience increased media coverage because of their gender. The fact that female Democrats experience a positive effect and then an additional effect for engaging in the traditional dialogue surrounding the Iran deal suggests that journalists and editors may do the opposite of favoring coverage of male House members when discussing masculine issues, an unexpected result that deserves greater investigation in future research.
Discussion and Conclusion
I pursued study to untangle the three predominant explanations for why gender inequality persists in political press coverage. I found that structural gender inequality continues to be amplified through proportionate underrepresentation in the media, and the structural inequality between male and female House members is deepened by the importance of seniority and committee membership for media coverage. Although we may think that House members can overcome structural inequality by engaging in political dialogue, by writing press releases and Tweets, this is not the case. However, Democrats’ gender-nonconforming involvement is better covered when they use press releases, even though Tweets generally affect coverage and press releases do not. Some suggest that media producers’ gender bias is the explanation, but again, the legacy of structural inequality is more consequential than journalists’ preferences. That bottom line is that if coverage is to change, these models suggest it will have to be a result of greater numbers of elected female representatives; the representation gap cannot be Tweeted away.
The first conclusion that can be drawn is about the influence of disruptive media, Twitter, versus traditional media, press releases. Although there was not a statistically significant relationship between press releases and mentions, there was a consistent positive and statistically significant relationship between Tweets and mentions. This finding aligns with previous studies that have highlighted the importance of disruptive media such as Twitter (Hayes and Lawless 2016; Peterson 2012) and provides partial support for hypothesis 2, that there would be a positive relationship between engaging in political dialogue and press coverage.
No statistically significant results were found when looking at the Tweet interaction terms. It is interesting that there was a not a gendered return on Tweeting, because scholars have proposed that there would be gender differences although there is debate as to whether social media amplifies or diminishes existing inequalities (Evans and Hayes Clark 2016; Vergeer 2015). The fact that male and female House members receive the similar results from their Twitter output is encouraging.
Additionally, press releases should not be abandoned just yet. The fact that male Democrats received a positive boost from writing press releases about Planned Parenthood and female Democrats received a positive boost from writing press releases about the Iran deal cannot be disregarded. This boost may be because journalists search for more traditional input from groups that are not at the center of the discussion. However, further research is needed to make sense of this unexpected finding.
The second conclusion that can be drawn is that gendered issue expertise does not seem to influence press coverage in the ways that previous research might predict (Deason et al. 2015; Dolan and Lynch 2014). Female House members did not receive more mentions by discussing Planned Parenthood, and male House members did not receive more mentions by discussing the Iran deal. Instead, male Democrats received fewer mentions in news discussing Planned Parenthood and female Democrats received more mentions in news discussing the Iran deal than male Republicans. Given that gender preferences related to issue expertise do not influence the number of mentions male and female House members receive, we can reject both hypotheses related to gendered issue expertise argument (hypotheses 4 and 5).
The third conclusion that can be drawn is that there does not appear to be support for hypothesis 3, that there is a media-level bias for male politicians (Carlin and Winfrey 2009; Dunaway et al. 2013; Luhiste and Banducci 2016). Instead, it appears as though female politicians’ underrepresentation is most likely due to their underrepresentation in the House, which provides support for hypothesis 1 (Baitinger 2015; Shor et al. 2015). The fact that male Republicans and Democrats did not consistently receive more mentions than their female counterparts when controls were added into the models suggests that there are not media-level gender preferences influencing mentions. Instead, the null coefficients that consistently accompany the gender and party variables suggest that women are underrepresented in the media because they are underrepresented in the House of Representatives. Although the absence of strong gender effects may quell some concerns about gender bias in the amount of press coverage politicians receive, ongoing structural barriers that prevent women from entering public office are troubling, especially because the effects appear to be mirrored in the news.
Limitations of this study highlight future avenues for research. The sample in this study is the U.S. House of Representatives. Future research may investigate the press coverage received by members of other governmental bodies in the United States and abroad, to analyze how additional individual- and country-level variables such as leadership positions and gender attitudes may influence gender differences in press coverage. Additionally, mentions related to two topics, Planned Parenthood and the Iran deal, were collected and analyzed. Future studies may collect data on a wide variety of topics that may be classified as “masculine” and “feminine,” and researchers may also develop and apply coding schemes that document not only occurrences of press coverage but the context in which these occurrences take place, which may illuminate nuanced differences in male and female politicians’ press coverage. Finally, mentions of House members were collected from six specific news sources. Although these sources were strategically selected to represent liberal, centrist, and conservative media, additional sources may provide greater insights into variation in press coverage across news sources.
Gender scholars interested in many arenas of social, economic, and political life have been perplexed by the persistence of gender inequality. This work contributes to this avenue of research by testing three explanations for why gender inequality persists in political press coverage. Although previous research has found evidence of gendered differences in communication and news producers’ gendered preferences, this study uncovers how representation in public office is the key contributor to why gender inequality persists in political news coverage by analyzing coverage of House members in news about Planned Parenthood and the Iran deal. In addition to writing Tweets that are consistently positively related to press coverage, female House members will be more present in the media when more women are elected to serve in the House of Representatives. Although the lack of support for gender preferences–related theories may be seen as a positive for those fighting gender inequalities in the political arena, addressing the structural barriers women face when running for and serving in public office will be an onerous process. However, increasing women’s representation in public office will not only increase women’s visibility in press coverage; it is likely to have wide-ranging impacts on politics and society more broadly.
Footnotes
Acknowledgements
I would like to thank Jennifer Earl for advising me on this project, Natasha Burrell and Levi Ybarra for their research assistance, Terrence Hill and Simone Rambotti for their guidance and support, and Robin Stryker and Corey Abramson for their feedback.