Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Partisan media bias vs. structural media bias: which one dominates in a polarized environment?
(USC Thesis Other)
Partisan media bias vs. structural media bias: which one dominates in a polarized environment?
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
PARTISAN MEDIA BIAS VS. STRUCTURAL MEDIA BIAS:
WHICH ONE DOMINATES IN A POLARIZED ENVIRONMENT?
by
Michelle Renae Cornelius
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
POLITICAL SCIENCE AND INTERNATIONAL RELATIONS
May 2024
Copyright 2024 Michelle Renae Cornelius
ii
Dedication
In loving memory of Dr. Kelly J. Madison
Professor Madison was an amazing teacher and mentor to all her students,
and I was blessed to call her a friend.
She will be forever missed by everyone who knew her.
iii
Acknowledgements
Writing a dissertation is a marathon not a sprint, and I would not have made it to the
finish line without the patience, guidance, and support of my adviser, Dr. Ann Crigler. It was
inspiring to watch her mentor other students – both undergraduate and graduate – and I wish her
nothing but the best in her well-deserved retirement.
I would also like to thank Dr. Alison Dundes Renteln for her unwavering support and
encouragement, Dr. Jeb Barnes for his support and advice, particularly with regard to helping me
re-think my original proposal and in interpreting my results, and to Dr. Tom Hollihan for his
support and feedback, particularly in the proposal stage, which helped re-focus my research.
I would like to express my deepest appreciation to my fellow graduate students for
sharing their wisdom with me, particularly Whitney Hua, Laura Brisbane, Raquel Centeno, and
Ruth Kelly.
Finally, I would like to express my gratitude to the department’s graduate advisors,
Veridiana Chavarin and Danielle DeRosa Ballard, for their words of encouragement and support.
Funding for this research was generously provided by the Ralph O. and Katherine S.
Bartling Endowed Political Science Graduate Fellowship and the POIR Summer Funding
Program.
iv
TABLE OF CONTENTS
Dedication....................................................................................................................................ii
Acknowledgements......................................................................................................................iii
List of Tables ...............................................................................................................................vi
List of Figures..............................................................................................................................vii
Abstract........................................................................................................................................viii
Introduction..................................................................................................................................1
Chapter 1: Literature Review.......................................................................................................8
Introduction......................................................................................................................8
The Pervasiveness of SV and Its Impact..........................................................................14
Why Media Coverage of SV is Important .......................................................................16
Partisan Media Bias .........................................................................................................18
Structural Media Bias ......................................................................................................21
Evaluating News Coverage of SV for Bias......................................................................24
How Liberals and Conservatives Are Different...............................................................25
Newsworthiness...............................................................................................................27
#MeToo and Newsworthiness..........................................................................................29
#MeToo and High-Risk Groups.......................................................................................31
Framing............................................................................................................................40
#MeToo and Framing ......................................................................................................44
Episodic and Thematic Framing ......................................................................................46
#MeToo and Episodic/Thematic Framing .......................................................................47
#MeToo and Solutions.....................................................................................................48
Chapter 2: Methods......................................................................................................................50
Data Selection ..................................................................................................................50
Sampling ..........................................................................................................................54
Inter-rater Reliability .......................................................................................................56
Variables of Interest.........................................................................................................56
Newsworthiness...............................................................................................................56
Framing............................................................................................................................58
Episodic and Thematic Framing ......................................................................................59
Solutions ..........................................................................................................................59
Descriptive Statistics........................................................................................................60
v
Chapter 3: Newsworthiness Analysis ..........................................................................................64
Pre-#MeToo Analysis of News Values............................................................................64
Post-#MeToo Analysis of News Values..........................................................................72
High-Risk Groups Post-#MeToo .....................................................................................82
Conclusion .......................................................................................................................86
Chapter 4: Framing Analysis.......................................................................................................89
Pre-#MeToo Analysis of Primary Frames.......................................................................89
Post-#MeToo Analysis of Primary Frames......................................................................98
Pre- and Post-#MeToo Analysis of Episodic and Thematic Framing .............................106
Pre-#MeToo Analysis of Solutions..................................................................................109
Post-#MeToo Analysis of Solutions................................................................................112
Conclusion .......................................................................................................................114
Chapter 5: Conclusion..................................................................................................................116
Implications for Democracy ............................................................................................119
Opportunities for Future Research...................................................................................121
Recommendations............................................................................................................122
References....................................................................................................................................127
Appendices...................................................................................................................................150
Appendix A: Methods......................................................................................................150
Appendix B: Codebook....................................................................................................160
Appendix C: Logit Regression Models............................................................................180
vi
List of Tables
Table 1. Summary of Data Selection and Sampling...................................................................53
Table 2. Summary of Results for News Values Pre-#MeToo in NYT and WSJ........................65
Table 3. Summary of Results for News Values Post-#MeToo in NYT and WSJ......................73
Table 4. Subcategories for the Government Action News Value in NYT and WSJ..................80
Table 5. Summary of Results for High-Risk Groups Post-#MeToo in NYT and WSJ..............83
Table 6. Summary of Results for Primary Frames Pre-#MeToo in NYT and WSJ ...................90
Table 7. Summary of Results for Primary Frames Post-#MeToo in NYT and WSJ..................99
Table 8. Logit Regression of Episodic Coverage .......................................................................108
Table 9. Comparison of Taxonomies of News Values...............................................................152
Table 10. Logit Regression of Episodic Coverage: Model 1 vs. Model 2..................................180
vii
List of Figures
Figure 1. Timeline of Major Events............................................................................................5
Figure 2. Number of Articles in the Full Sample by Month.......................................................55
Figure 3. Number of Articles in the NYT and WSJ, Pre- and Post-#MeToo .............................60
Figure 4. Frequency Bar Chart for News Values in the NYT ....................................................61
Figure 5. Frequency Bar Chart for News Values in the WSJ.....................................................61
Figure 6. Frequency Bar Chart for Primary Frames in the NYT................................................62
Figure 7. Frequency Bar Chart for Primary Frames in the WSJ.................................................63
viii
Abstract
There are two main competing theories of media bias which have been used to explain
news content: partisan media bias and structural media bias. Researchers have not found much
support for the supply-side theory of partisan media bias which posits that the personal political
views of journalists influence how they write news stories. However, economists have found
support for demand-side partisan media bias which argues that in a highly polarized competitive
environment like the one we are in now, news organizations have an economic incentive to
fragment the audience along ideological lines. Structural bias, on the other hand, results in
similar news coverage because of shared production values stemming from the professional
norms and routines of journalists. Political communication scholars are concerned that the
highly polarized political environment will influence news content as news organizations try to
provide content consistent with the beliefs of the audience. This study examined a partisan issue,
sexual violence, in two ideologically different newspapers – the New York Times and Wall Street
Journal – over a four-year period, which includes #MeToo going viral, to determine if partisan
bias or structural bias is controlling. The results show that coverage of sexual violence was
remarkably similar across both newspapers, strongly suggesting that the professional routines
and norms of journalism have remained strong in this highly polarized political environment.
1
Introduction
Journalists…become more polarized, and more polarizing, when we start
spending our time in polarizing environments. I have seen it in myself, and I have
watched it in others: When we’re going for retweets, or when our main form of
audience feedback is coming from highly partisan social media users, it subtly but
importantly warps our news judgment. It changes who we cover and what stories
we chase. And when we cover politics in a more polarized way, anticipating or
absorbing the tastes of a more polarized audience, we create a more polarized
political reality.
- Ezra Klein (2020)
In the early 19th century, many newspapers were sponsored by political parties and
explicitly partisan (Pierson and Schickler 2020). Cook (1998) traces the beginning of the
partisan press to the split between Hamiltonians and Jeffersonians which created two rival
parties: the Federalists and the Republicans. Newspapers were a critical component of party
strategy, and party organizers founded newspapers to reach voters and mobilize support for the
party. Party leaders also served as editors and gave speeches on behalf of the party. By the
1890s, the partisan press had largely vanished, primarily due to the commercialization of the
press, but also due in part to political shifts (e.g., meetings and other forms of campaigning
replaced the role party newspapers had previously served) (Cook 1998).
Towards the end of the 19th century, technological and market changes in the news
industry transformed news production in the United States from a partisan press to one which
valued independence and objectivity as guiding principles in order to reach a broader audience
(Hamilton 2004). This market-based model relies on revenue from advertisers who want to sell
products to their news consumers. The economic link between the newspaper and its audience is
what motivates structural contributors to media bias (Graber and Dunaway 2015). In order to
maximize profits, newspapers developed routines to make newsgathering cheap and efficient.
2
These routines help journalists quickly identify certain events as “newsworthy” while excluding
others, as well as provide story formulas to help them write stories quickly (Shoemaker and
Reese 2014; Bennett 2016). Since identifying what is news and how to report it can be
contested, editors and journalists compare their work to rivals and want their work to be
reinforced by others (Cook 1998). This form of news production creates coverage which looks
similar across different news outlets because they share the same production values and use the
same official sources to generate stories.
Despite similar coverage across news outlets, the notion that the media has a liberal bias
is a commonly held belief among the public and has permeated our political discourse (Hassell,
Holbein, and Miles 2020; Schiffer 2018). In tracing the origins of the “liberal” media bias
charge, historian David Greenberg (2008) found that in the first part of the 20th century media
critics Upton Sinclair and Gilbert Seldes, and politicians, including Franklin Roosevelt and Harry
Truman, accused the press of having a conservative tilt, and by the 1950s the phrase “one-party
press” was a reference to a pro-Republican press. During this same time in the 1950s, the charge
that the media had a liberal bias was made by Republican politicians such as Senator Robert Taft
and Senator and Vice President Richard Nixon, as well as William F. Buckley, founder of the
conservative magazine National Review, and the most significant attacks came from Senator Joe
McCarthy and other leaders of the anti-Communist crusade.
It was not the threat of communism, however, which cemented the notion of a liberal
press as a conservative mantra, but the civil rights movement according to Greenberg. White
Southerners who opposed integration were resentful of the critical coverage by national
journalists from the North, which they believed favored the civil rights activists and therefore
violated their professional standards of objective reporting. The idea that an elite, Northeastern
3
media was distorting the news to fit their political agenda was advanced by prominent
Southerners such as Alabama Governor George Wallace, and it became a conservative dogma
under President Nixon.
The search for a liberal bias in the news has been the subject of scholarly research for
decades (Lichter, Rothman, and Lichter 1986; Niven 1999; Watts et al. 1999; Lowry 2008;
Groseclose 2011) but most research has failed to support this theory (Schiffer 2018; D'Alessio
and Allen 2000). However, recent research into partisan bias has argued that for-profit news
organizations have an incentive in a competitive market to fragment the audience ideologically
and create content to target those audiences (Mullainathan and Shleifer 2005; Gentzkow and
Shapiro 2010).
Recent technological innovations have enabled news organizations to better know their
audience which allows them to tailor content to match audience preferences. Journalist Ezra
Klein (2020) writes that prior to the digital age, the only information news organizations had
about their audience came from network ratings, subscription renewals and letters from the
public. News organizations now have real-time digital analytics which can tell you how many
people are reading an article at any given time and how they found the article (e.g., through a
social media network or a search engine).
At its most basic level, what Klein is arguing in the epigraph above is that polarization
among the public is influencing the production of news. While all news is biased since it
involves making decisions about who and what to cover and how it should be covered
(Lippmann 1922; Shoemaker and Reese 2014; Entman 1990), partisan media bias is concerning
because previous research has found that it is a significant contributor to polarization (DiMaggio
2020) which can have implications for democratic stability (Pierson and Schickler 2020). This
4
study will examine a partisan, contested issue – sexual violence, an umbrella term used by
experts to include sexual harassment, assault, rape, incest, child sexual abuse, exploitation,
human trafficking, exposure and voyeurism (Bing and Lombardo 1997; National Sexual
Violence Resource Center 2015) – in two ideologically disparate national newspapers to
determine if the issue is covered differently, which would be evidence of partisan bias, or
similarly, which would suggest that structural biases in news production are still dominant.
The time frame of this study – October 2015 through September 2019 – provides a
unique opportunity to analyze the issue of sexual violence as it includes several high-profile
individuals and the #MeToo movement (see Figure 1 for a timeline of events). In July 2016,
former Fox News anchor Gretchen Carlson filed a sexual harassment lawsuit against the founder
and CEO of Fox News, Roger Ailes (Grynbaum and Koblin 2016). In October 2016, the
Washington Post published the infamous Access Hollywood tape in which presidential candidate
Donald Trump was caught on a hot mic bragging in vulgar terms about sexually assaulting
women (Fahrenthold 2016). Despite multiple women coming forward to say they had been
sexually harassed or assaulted by Trump (Kurtzleben 2016), he was able to win the Electoral
College and assume the presidency. The day after he was inaugurated, millions of people
attended the Women’s March, which may have been the largest single-day protests in U.S.
history at that point (Frostenson 2017). A few months later, the New York Times reported that
Fox News and Bill O’Reilly had paid millions of dollars to settle sexual harassment claims
against the anchor, and after an internal investigation which revealed further misconduct, his
show was eventually canceled (Steel and Schmidt 2017a). In October 2017, the New York Times
and the New Yorker revealed that Hollywood movie mogul Harvey Weinstein had been sexually
assaulting and raping women for decades (Kantor and Twohey 2017; Farrow 2017). Several
5
days later, the hashtag #MeToo went viral after actor Alyssa Milano asked people to reply “me
too” to her tweet if they had been sexually harassed or assaulted so that people could understand
the magnitude of the problem. In the months that followed, hundreds of high-profile men were
made to answer for past their past behavior, and many resigned or were terminated (Green 2018;
Carlsen et al. 2018).
Figure 1. Timeline of Major Events
July 6, 2016
Gretchen Carlson
files sexual
harassment
lawsuit against
Roger Ailes.
Jan. 21, 2017
Millions of people
attend the
Women’s March.
Oct. 5, 2017
NYT reveals that
Harvey Weinstein
had been sexually
assaulting and
raping women for
decades.
Oct. 7, 2016
Access Hollywood
tape of Trump
bragging about
assaulting women
is revealed.
April 1, 2017
NYT reveals Fox
News and Bill
O’Reilly paid
millions to settle
sexual harassment
lawsuits.
Oct. 15, 2017
Alyssa Milano
asks people to
reply “me too” to
her tweet if they
have been
sexually harassed
or assaulted.
Both public opinion polling and academic research has demonstrated partisan differences
among liberals and conservatives in their attitudes and opinions towards SV (Economist/YouGov
2018; van der Linden and Panagopoulos 2019; Naseralla, Baker, and Warner 2021). There are
three important reasons to study news coverage of a partisan issue. First, Shultziner and Stukalin
(2021) argue that partisan media bias is context sensitive and can be time, issue or event specific.
Several studies have found partisan media bias in news coverage of political scandals (Budak,
Goel, and Rao 2016), unemployment (Larcinese, Puglisi, and Snyder 2011), mass shootings
6
(Pelled et al. 2021), and large-scale corporate accounting scandals (Benediktsson 2010). This
study will examine both an issue (SV) and an event (#MeToo); if partisan media bias is
controlling, we should observe it when comparing news coverage in the NYT and WSJ.
Second, scholars have argued that the current media landscape is dominated by national
news and this incentivizes politicians to focus on national issues and cleavages, thereby
increasing the salience of nationally oriented political and social identities at the expense of
geographical identities, thereby furthering trends towards nationalization and polarization which
can threaten democratic stability (Pierson and Schickler 2020; Graham and Svolik 2020). Lastly,
scholars have expressed concern that the high level of polarization in the United States could
affect news content. Since news content can influence the public’s attitudes and opinions about
the problem itself and possible solutions (Iyengar 1991), if an issue is covered differently by
ideologically dissimilar newspapers, it could result in a partisan divide regarding possible causes
and solutions, or whether it is a problem policymakers should address at all.
In Chapter 1 I will discuss the importance of studying news coverage of sexual violence
and the review the theories of partisan and structural media bias, discussing how liberals and
conservatives are different and how that may result in different news coverage. I will also
discuss #MeToo and describe how this will serve as another test to see if partisan or structural
media bias is controlling. Chapter 2 will review the data and methods used to address the
research question in this study, including data selection, sampling, the variables of interest and
how they were measured. In Chapter 3 I will examine what gets covered and in Chapter 4 I will
look at how SV is covered. The results show that both the NYT and WSJ relied on similar news
values when selecting stories and told the stories in much the same way, thus providing support
for the structural media hypothesis. In the final chapter I will summarize the results of this study
7
and provide some critiques, as well as make recommendations to improve coverage of SV and
offer some suggestions for future research.
8
Chapter 1
Literature Review
The die may therefore be cast for increasingly replacing traditional journalism with
partisan media.
- Matt Grossmann, Senior Fellow at the Niskanen Institute and
Associate Professor of Political Science at Michigan State University
(2017)
This chapter lays out the framework for the research question in this study, beginning
with a brief discussion of the importance of news in opinion formation, polarization in the United
States, and the concern that could affect news content and the implications for democratic
governance. This is followed by a brief overview of partisan and structural media bias and the
research questions. I then establish sexual violence is a partisan issue and discuss the
pervasiveness of problem and why media coverage of this issue is important. This is followed
by a more detailed review of previous research concerning partisan and structural media bias,
how liberals and conservatives are different, and how media bias will be evaluated in this study.
Introduction
An independent, pluralistic media is considered essential for democracy (Levitsky and
Ziblatt 2018; United Nations n.d.). The role of the news media in democracy has been
conceptualized to inform the public, provide a public forum for debate, hold government and
other powerful organizations accountable, and voice public opinion (Graber 2003; Gans 2003;
Curran 2005; Patterson 2013). Research has shown that the news plays a role in constructing our
social reality (Neuman, Just, and Crigler 1992; Just et al. 1996; Levendusky 2022; Smith and
Searles 2013) and how the traditional media interprets events and frame the news can determine
the political consequences (Graber and Holyk 2011). The news media can raise the profile of a
9
social problem and legitimize certain voices and frames (Bekkers et al. 2011) as well as
influence the public’s attitudes and opinions about the problem and possible solutions (Iyengar
and Kinder 2010; Gilliam and Iyengar 2000; Jordan and Page 1992; Soroka 2003; Iyengar 1991).
Recently scholars have argued that the current media landscape contributes to political
polarization (Levitsky and Ziblatt 2018; Pierson and Schickler 2020; Padgett, Dunaway, and
Darr 2019) and studies have found a stronger effect for those who consume conservative media
(Garrett, Long, and Jeong 2019; DiMaggio 2020). Surveys reveal that polarization has been
growing in the United States since the 1990s (Pew Research Center 2014, 2019). A 2021 poll by
the University of Virginia Center for Politics and Project Home Fire revealed that nearly 90
percent of Biden and Trump voters surveyed agreed that people like them will not belong in
America anymore if the “other side” has its way (UVA Center for Politics 2021). They also
found that 41 percent of Biden voters and 52 percent of Trump voters agreed at least somewhat
that the situation in the U.S. was so untenable that they supported Blue/Red states seceding from
the union to form their own separate country. Historian Jon Meacham says that he sees
similarities between the U.S. today and the 1850s, just before the Civil War (Coleman 2022).
Experts have argued that this high level of polarization is dangerous for democratic stability
(Levitsky and Ziblatt 2018; Pierson and Schickler 2020; Graham and Svolik 2020) and can lead
to political violence (Kalmoe and Mason 2022; Kleinfeld 2021).
Given the high level of polarization in the country, political communication scholars have
expressed concern that this could affect news content (Iyengar and Hahn 2009). According to the
demand side theory of partisan media bias, since consumers prefer news which is consistent with
their prior beliefs, news organizations have an economic incentive in a competitive market to
fragment the audience in order to cater to this consumer demand (Mullainathan and Shleifer
10
2005). Several studies have found support for the demand side theory (Mullainathan and
Shleifer 2005; Hamilton 2004; Gentzkow and Shapiro 2010; Agirdas 2015) and research has
found that partisan news sources cover issues differently (Kenix and Jarvandi 2019; Brown and
Mourão 2022; Pelled et al. 2021; Agirdas 2015).
Other scholars, however, argue that professional journalistic norms, which Bennett
(2016) describes as “moral standards, codes of ethics, and guidelines about inserting voices and
viewpoint into stories” (142), constrain partisan media bias (Shultziner and Stukalin 2021). The
professional norms of independence and objectivity – which include fact-checking and providing
balanced coverage – are designed to avoid charges of political bias (Tuchman 1972; Bennett
2016). Adherence to these professional norms is also important for career advancement, since
most journalists build their careers by moving from one news organization to another (Cook
1998) and ideologically biased news reporting could harm their reputation and future job
prospects (Bedingfield 2012).
In their pursuit of objectivity to avoid charges of partisan bias, journalists adhere to
established norms or practices of newsworthiness. These may create a different form of bias:
structural media bias (Cook 1998). Structural media bias stems from the profit-driven imperative
of news organizations which encourages similar production values and notions of
newsworthiness (Cook 1998; Graber 2003; Graber and Dunaway 2015; Bagdikian 2004). These
values include news which favors the rich and powerful, and a reliance on stories from official
sources and news beats (Bagdikian 2004; Cook 1998; Bennett 2016). The result of this bias is
news coverage which looks similar across different news outlets (Graber 1997; Cook 1998;
Boczkowski 2010). This coverage is biased because it distorts reality since it favors the
11
perspective of official sources and powerful individuals and organizations (Shoemaker and
Reese 2014; Bagdikian 2004).
Since the news media can raise the profile of a social problem, define it, and provide
causal attributions, how an issue is covered by journalists can also influence the range of possible
solutions (Iyengar 1991; Entman 1993). Given recent research which shows that consumers of
partisan media almost never watch news from the other side, and only a minority also watch
national broadcast news (Broockman and Kalla 2023), if ideologically disparate news outlets are
covering a social problem in very different ways, it could adversely impact the ability of
policymakers to address it since democratic governance depends on shared values and the ability
to find common ground and compromise to solve problems (Graber 2003; Garrett, Long, and
Jeong 2019). However, if professional norms and routines are constraining partisan media bias,
then coverage should look similar. This leads to the first research question:
RQ1: Do ideologically dissimilar newspapers cover a partisan issue differently in a
highly polarized environment?
This study will examine the social problem of sexual violence (SV) in two ideologically
disparate newspapers, the New York Times (NYT) and the Wall Street Journal (WSJ) (Ad Fontes
Media 2023). In line with the definition of SV used by experts, this study includes articles which
discuss sexual harassment, assault, rape, incest, child sexual abuse, exploitation, human
trafficking, exposure, and voyeurism (National Sexual Violence Resource Center 2015; Bing and
Lombardo 1997).
Academic research and public opinion polling have revealed partisan differences in
attitudes and experiences of sexual harassment and assault. While both Democrats and
Republicans view sexual assault as a serious problem, 99 and 94 percent respectively
(Quinnipiac University Poll 2019), when asked about sexual harassment the differences were
12
stark: 61 percent of Democrats said it is very serious problem compared to 21 percent of
Republicans (Economist/YouGov 2018). Consistent with public opinion polling, van der Linden
and Panagopoulos (2019) found that liberals believe that sexual harassment is a significantly
bigger problem than conservatives. Another study found that Republicans and conservatives
were less likely to report severe forms of sexual harassment and assault (Jose, Fowler, and Raj
2021).
Public opinion polling has also revealed stark partisan differences surrounding SV in
electoral and political contexts. In a CBS News/New York Times poll on November 3, 2016,
when asked if they believed then-candidate Donald Trump made unwanted sexual advances,
only 20 percent of Republicans said they believed the allegations were mostly true compared
with 84 percent of Democrats. When asked about the infamous Access Hollywood tape in which
Trump bragged about sexually assaulting women, 74 percent of Republicans said it did not
change their view of him (CBS News 2016). In a Quinnipiac University poll taken in December
2017, 48 percent of Republicans stated that they would still consider voting for a political
candidate who had been accused by multiple women of sexual harassment if they agreed with
them on the issues; only 10 percent of Democrats indicated a willingness to do the same. These
partisan differences extend past elections and into Supreme Court nominations: only 10 percent
of Republicans believed Christine Blasey Ford who said that Supreme Court justice nominee
Brett Kavanaugh sexually assaulted her when they were in high school, compared to 86 percent
of Democrats (Quinnipiac University Poll 2018). When asked if Kavanaugh should be
confirmed even if he was guilty of assaulting Ford, 54 percent of Republicans said yes compared
to 12 percent of Democrats (Marist Poll/National Public Radio/PBS NewsHour 2018).
13
Partisan differences have also been found regarding attitudes towards the #MeToo
movement. In an Economist/YouGov survey taken a year after the hashtag went viral, 75
percent of Democrats said they had a very or somewhat favorable view of the movement
compared to 21 percent of Republicans. Furthermore, 66 percent of Republicans stated that the
#MeToo movement had gone too far while only 11 percent of Democrats agreed with that
statement. A survey from Pew Research five years after #MeToo found that 70 percent of
Democrats supported the movement compared to 22 percent of Republicans (Brown 2022).
1
While both Democrats and Republicans believed that offenders are more likely to be held
responsible when compared to five years ago, Republicans were nearly twice as likely than
Democrats to say that false reporting was extremely or very common (20 percent compared to 11
percent). When asked how common it was for people NOT to report sexual harassment or
assault, 58 percent of Democrats said it was extremely or very common compared to 33 percent
of Republicans.
The partisan divide in support of #MeToo presents the opportunity for a better test of
partisan media bias. Social movements have historically had a difficult time getting their
message into the mainstream news to reach a larger audience. Since journalists gather most of
their information from their media beats which represents selected public and private institutions,
Graber (2003) argued it was a myth that mass media in American democracy gives voice to
public opinion. The overreliance on official sources from these media beats has made it difficult
for social movements to broadcast their message to a larger audience via the news (DiMaggio
2020; Vliegenthart and Walgrave 2012). However, social media has allowed politicians,
businesses and the public to bypass traditional media and communicate directly with each other
1 Pew Research combined Democrats with Democratic leaners and Republicans with Republican leaners in their
report.
14
(Bennett and Pfetsch 2018; Paulussen and Harder 2014) and journalists have used social media
to source stories (Bouvier 2019; Billard 2021).
In the 45 days after actor Alyssa Milano asked Twitter users to respond to her tweet with
the words “me too” if they had been sexually harassed or assaulted, #MeToo was used over 85
million times in 85 different countries (Sayej 2017). Several victims came forward and over the
next 18 months, over 400 high-profile executives and other employees from a wide range of
industries found themselves under scrutiny for their behavior and many resigned or were
terminated (Green 2018). Since liberals are more supportive of #MeToo than conservatives
(Brown 2022), if partisan media bias is controlling the NYT would be predicted to be more
receptive than the WSJ to the messages of the #MeToo movement. However, if structural media
bias is controlling, coverage of SV will be similar in the NYT and WSJ post-#MeToo. This is
the second research question:
RQ2: Are there partisan differences in coverage of SV in the NYT and WSJ post-
#MeToo?
The Pervasiveness of SV and Its Impact
SV is a significant social problem in the U.S. In a survey conducted shortly after
#MeToo went viral in the fall of 2017, 81 percent of women and 43 percent of men reported that
they had experienced sexual harassment and/or sexual assault (Kearl 2018). Verbal sexual
harassment was reported by 77 percent of women and 34 percent of men and cyber sexual
harassment was reported by 41 percent of women and 22 percent of men. Nearly half of women
(51 percent) and 17 percent of men had experienced unwanted sexual contact and 30 percent of
women and 12 percent of men reported that someone had exposed their genitals to them. More
than 1 in 4 women (27 percent) and 1 in 14 men (7 percent) were survivors of sexual assault.
15
Previous research has documented the substantial health and economic consequences of
SV. Survivors may experience short-term physical injuries (e.g., bruising) as well as more longterm gynecological, gastrointestinal, and sexual health problems (Basile et al. 2016). SV is also
associated with risk behaviors – like smoking and excessive alcohol use – which can result in
chronic medical conditions (Basile et al. 2016). Other adverse physical health outcomes include
headaches, nausea, musculoskeletal pain, and weight loss/gain (Cortina and Berdahl 2008). The
psychological impacts include depression, anxiety, post-traumatic stress disorder, decreased selfesteem, self-blame, alcohol and prescription drug abuse, and increased risk for suicide (Basile et
al. 2016; Griffith 2019; see Cortina and Berdahl 2008 for a review).
Adverse occupational outcomes include diminished performance, work withdrawal (e.g.,
absenteeism, tardiness, work neglect), turnover, an increase in team conflicts, and a decreased
commitment to the organization, performance, and productivity (Loya 2015; Basile et al. 2016;
see also Cortina and Berdahl 2008). This can have a detrimental impact on a victim’s earning
potential and economic well-being (Loya 2015).
The CDC estimates that the lifetime economic burden from rape alone is $3.1 trillion2
(in
2014 U.S. dollars) with the government sources paying approximately $1 trillion of that burden.
Just over half of the cost, $1.6 trillion (52 percent), is the result of lost work productivity among
both the victim and the perpetrator while medical costs account for approximately $1.2 trillion
(39 percent) (Peterson et al. 2017).
It is not only the victim who is harmed by SV. Friends and family of survivors often
struggle to cope with their own intense and painful emotions. They may experience shock and
disbelief, anger, sadness, powerlessness, guilt, fear, and anxiety (Opening the Circle n.d.).
2 Amount calculated based on data which indicates more than 25 million U.S. adults are rape survivors at a cost of
$122,461 per victim.
16
Employees who witness sexual harassment in the workplace may also be harmed. Miner-Rubino
and Cortina (2007) found that employees who observed incivility and harassment towards
women had lower well-being which translated into higher organizational withdrawal for both
men and women.
The pervasiveness of SV and the magnitude of the impact it has on our society, both in
health and economic terms, makes the issue of SV worthy of news coverage as it can increase
awareness of the problem, influence the debate over possible solutions (De Benedictis, Orgad,
and Rottenberg 2019), and lend legitimacy to certain frames and voices about an issue (Bekkers
et al. 2011).
Why Media Coverage of SV is Important
The constructionist paradigm “posits that for public opinion to be understood, we must
examine the artifacts of political culture” (Johnson-Cartee 2005, 15). Since we cannot
personally experience everything, Lippmann (1922) argues that what we learn about the world
outside is brought to us by others, including news media which helps create the “pictures in our
heads.” Gamson et al. (1992) argue that we then use these pictures “to construct meaning about
political and social issues” (374). Because the world is so massive, journalists and editors have
to choose which stories to publish, thereby excluding others. As a result, Lippmann contends
that “the pictures in our heads” are not necessarily representative of the world outside and this
has an impact on public opinion formation. Research has shown that media coverage shapes not
only opinion formation, but it also influences our socio-cultural understanding of SV itself
(Terkildsen and Schnell 1997; Edwards et al. 2011; McDonald and Charlesworth 2013; Amber et
al. 2019; Cobos 2014; Franiuk et al. 2008; Pennington and Birthisel 2016; Sacks, Ackerman, and
Shlosberg 2018; Schneider and Hannem 2019; Siefkes-Andrew and Alexopoulos 2019).
17
While practitioners warn that media coverage which distorts the social reality of SV can
be damaging, news stories about SV can increase social awareness and lead to an increase in
reporting to both counseling organizations and law enforcement (Greer 2003), as well as
influence how we interpret our own experiences (Amber et al. 2019; see also Kitzinger 2004).
Research has found that victims sometimes do not label what has happened to them as
SV (Magley et al. 1999; Franiuk et al. 2008) and the news media can increase the audience’s
understanding of what it is, as well as what causes it, and how it can be prevented (Amber et al.
2019; Centers for Disease Control and Prevention n.d.). A study by Amber et al. (2019) found
that reading news stories of high-profile incidents of sexual misconduct triggered personal recall
of being sexually harassed in the past and reinterpretation of those experiences (e.g., giving the
incidents a different label or seeing them in a different light). These findings were stronger
among women and those with a more progressive political ideology, thus providing further
evidence of an ideological divide when it comes to a value-laden social issue such as sexual
harassment and assault.
Media coverage of a social problem can also raise the profile of a social issue (Mejia,
Cheyne, and Dorfman 2012) as well as influence who the audience blames for the problem and
therefore who, or what group, is responsible for solving it (Iyengar 1991). Furthermore, media
coverage can also legitimize certain voices and frames. For example, in their study of a studentled protest in the Netherlands over new education requirements, Bekkers et al. (2011) found that
the mass media strengthened the frames students produced and legitimized an extremely wellspoken student as the leader of the protest.
18
Partisan Media Bias
A 2017 Gallup poll found that 62 percent of respondents believed that the news media
favors one party over the other, a 14-percentage point increase from 2003 (Swift 2017). Out of
those who believed the news favored one party, nearly two-thirds said it favored the Democratic
party. The percentage of people who believe the news media favors one side has increased since
2017: a Pew Research poll before the 2020 presidential election found that nearly eight-in-ten
Americans (79 percent) said that news organizations tend to favor one side when reporting on
political and social issues (Walker and Gottfried 2020).
While the liberal media bias thesis, a supply-side theory of media bias which posits that
journalists and editors hold liberal attitudes which they project into news stories (Perloff 2018),
has been supported by some studies (Groseclose and Milyo 2005; Groseclose 2011; Schiffer
2006), most research has failed to support it (Lichter 2017; Schiffer 2018). The increasing
perception of media bias among the public has been attributed to the “hostile media effect”
which occurs when “partisans perceive identical news coverage of a controversial issue as biased
against their own side” (Feldman 2018, 549) and strategic communication efforts by
conservative political elites (Domke et al. 1999; Schiffer 2018).
Recent studies have attributed partisan media bias to economic incentives rather than the
political attitudes of individual journalists and editors. The demand-side theory of media bias, as
described above, attributes partisan media bias to competitive market environments which
incentivize news organizations to fragment the audience along ideological lines. Agirdas (2015)
found that there was more bias when there was a rival newspaper in the same media market;
however, when the rival closed, the remaining paper moderated its content, presumably to attract
readers from the closed newspaper. Gentzkow and Shapiro (2010) indexed media slant by
19
comparing phrases used by members of Congress in the 2005 Congressional Record to the
language used in newspapers. They combined this measure with zip-code level data and found a
strong fit between a newspaper’s slant and the ideology of potential readers in that geographic
market. Haselmayer, Wagner, and Meyer (2017) found that media are more likely to report on
messages from political parties their audience favor.
In a highly polarized environment like the one we are in now (Levitsky and Ziblatt 2018;
Mason 2018), Iyengar and Hahn (2009) argue that rational media owners can gain market share
by injecting political bias into their news coverage, a sentiment shared by Stroud (2008).
Research has shown that partisan media contributes significantly to polarization, particularly for
conservatives (DiMaggio 2020). Polarization among the public can also lead to an increase in
selective exposure to news, resulting in a less informed and more polarized electorate (Iyengar
and Hahn 2009), with different impressions about the world around them (Kull, Ramsay, and
Lewis 2003-04).
Evidence of audience influence on news content was recently part of the public discourse
as a result of the defamation lawsuit brought by Dominion Voting Systems against Fox News for
false claims made on their network about the company’s machines which were used to count
votes in the 2020 presidential election. After depositions in the case, Dominion filed a brief
which revealed the concern among Fox News executives that they were losing viewers because
they had correctly called Arizona for Joe Biden and their coverage was not aligning with thenPresident Trump’s claims that the election was stolen (Bauder 2023). Dominion alleged this fear
of losing voters played a key role the network’s decision to allow conspiracy theories to be
broadcast despite knowing they were false. Fox News eventually settled the case with Dominion
for $787.5 million (Bauder, Chase, and Mulvihill 2023).
20
The #MeToo movement provides another opportunity to test for partisan media bias since
there are partisan differences in support for #MeToo. The movement, started by Tarana Burke in
2006, gained widespread attention in October 2017 when actor Alyssa Milano asked everyone
who had been sexually harassed or assaulted to write “me too” as a status so that the public could
understand the pervasiveness of the problem (Ohlheiser 2017; Park 2017). The hashtag quickly
went viral and within 10 days it had been used in 1.7 million tweets in 85 countries and
Facebook reported that 45 percent of users had friends who had posted “me too” (Park 2017).
Historically, the gatekeeping practices of news organizations have made it difficult for
social movements to get their message out to a wider audience because journalistic routines and
beat reporting favors news stories from official sources (Cook 1998; Schudson 2002; Graber and
Dunaway 2015; Bennett 2016) and other elites whose unequal access to economic and symbolic
resources allow them to define the terms of a debate (Curran 2005). Communication scholars
have argued that the internet and social media have changed the traditional model of gatekeeping
as both elites and the general public can broadcast their messages to a large audience without the
assistance of the news media (Shoemaker 2017; Bennett and Pfetsch 2018). Social movements
have used hashtags on social media to organize protests and raise awareness about social issues
(Clark 2016; Yang 2016). The use of hashtags to bring attention to social problems is known as
hashtag activism, and research has shown that hashtag campaigns concerning violence against
women have allowed survivors to tell their own stories, challenge rape culture, support other
survivors, and influence the media narrative (Bogen et al. 2021; Clark 2016; Baker, Williams,
and Rodrigues 2020; Storer and Rodriguez 2020).
The virality of the hashtag combined with the Hollywood connection was sure to get the
attention of journalists, but if or how they responded to #MeToo in news coverage of SV remains
21
an open question. If partisan media bias is controlling, then higher support of #MeToo among
liberals should result in the NYT responding differently to the movement than the WSJ.
However, if structural bias is controlling, the NYT and WSJ should respond to #MeToo in a
similar way, either by influencing changes in coverage in the same direction or by ignoring the
movement and continuing to cover SV in the same way they had prior to October 2017.
Partisan media bias is just one form of media bias. Cook (1998) argues that the norm of
objectivity combined with media routines developed to meet the market-based demands of news
production create shared production values among news organizations which result in similar
coverage, a phenomenon identified by Hofstetter (1976) as structural media bias. Since partisan
media bias can be constrained by these professional norms and routines (Shultziner and Stukalin
2021), this study will analyze news content to see which form of bias dominates coverage of SV
in a polarized political environment.
Structural Media Bias
Structural media bias stems from the profit-driven incentive of news organizations and
the professional journalistic norms and routines created to address the needs of the market
model. Prior to the 1890s, daily newspapers were explicitly aligned with a political party, but
technological and market changes created a financial incentive to establish editorial
independence from political parties and journalists adopted objectivity as a guiding principle
(Hamilton 2004; see also Cook 1998). Hamilton (2004) highlights two economic changes which
led to the rapid decline of the partisan press:
The development of expensive high-speed presses made it possible for a newspaper to
serve many more readers within a given city. To reach more readers, and therefore
spread the high fixed costs across many consumers, newspapers stopped talking about
politics in an explicitly partisan manner. Independent papers could draw readers from
across the political spectrum. At the same time, advertising became an important way for
companies with nationally and locally distributed brands to raise awareness about their
22
products. Papers with larger audiences attracted more attention from advertisers, another
incentive to increase readership. As a result, papers began to drop overt political bias and
proclaim their independence in covering news of government and politics. (3)
In order to maximize profits, news organizations have to fill the daily “news hole” with
content as cheaply and efficiently as possible. Everyday news organizations are faced with an
infinite and unpredictable number of events which can become news, so they have developed
routines in order to help editors and journalists identify what events to cover and how to cover
them (Shoemaker and Reese 2014; Bennett 2016; Tuchman 1973). These routines not only help
manage the flow of information, but they also socialize reporters to view some events as
newsworthy while excluding others, and in the process they actively construct reality
(Shoemaker and Reese 2014).
In order to increase efficiency, news organizations assign reporters to “beats” where
newsworthy events occur on a daily basis (Bennett 2016; Ryfe 2021). These beats generally
involve government institutions such as local and state legislatures, police, courts, Congress, and
the presidency, but also can include issues (e.g., technology) and regions (e.g., city or county)
(Bennett 2016; Ryfe 2021). Ryfe (2021) contends that the beat system increases efficiency in
two important ways. First, it allows journalists to quickly obtain information which can be
turned into a news story. Second, instead of deploying large numbers of reporters across a city
in the hopes of finding newsworthy events (which would be cost prohibitive), it allows editors to
place a small number of reporters at key institutions where news is guaranteed to happen.
Partially in response to these news beats, public officials began holding scheduled events
specifically designed to be newsworthy. The beat system makes journalists dependent on the
sources they cover for news (Bennett 2016), and it privileges news coverage of events which
23
occur on these beats over other important events which occur outside of it (Graber and Dunaway
2015).
Cook (1998) argues that uncertainty “over what news is, who makes it, and how to
produce it…leads to convergence of news content and newsmaking processes among individual
journalists and across news organizations” (78). Editors and journalists compare their work to
that of rivals and want their account of events to be reinforced by others; if it is not, their
competence is called into question. In his classic book The Boys on the Bus about the 1972
presidential campaign, Timothy Crouse (1973) describes how a reporter’s eyewitness account of
Hubert Humphrey being nearly reduced to tears at a campaign event was called into question by
his managing editor because stories from other news outlets did not mention it. The reporter
began to doubt himself and was relieved when he found another news story which corroborated
his account.
Since for-profit news organizations share notions of newsworthiness and rely on the same
sources for information, news coverage looks similar across news outlets (Bennett 2016; Cook
1998; Graber 2003; see Ryfe 2021 for an analysis of for-profit and nonprofit news
organizations). Bennett (2016) argues that the story formulas which result from these norms and
routines work because standardized news is safe and any departures from the norm must be
defended. This also creates predictable news content which is appealing to advertisers who are
concerned about their reputation and want to know in advance what type of news content their
ads dollars are supporting.
While both demand-side partisan media bias and structural media bias are motivated by
the for-profit business model, if partisan media bias is controlling then news coverage of a
24
partisan issue should be different when comparing ideologically dissimilar newspapers.
However, if structural bias is controlling, then news coverage should look similar.
Evaluating News Coverage of SV for Bias
In order to investigate whether ideologically disparate newspapers cover a partisan issue
differently in a highly polarized environment, this study will examine the newsworthiness factors
of published news stories on SV and how it was framed for the audience. In order to identify the
ideology of the audience for the NYT and WSJ, this study uses a proxy: trust in the news source.
A poll by Pew Research found that the NYT was the most trusted newspaper among Democrats
and the WSJ was the most trusted newspaper among Republicans (Jurkowitz et al. 2020).
Business reporting on the two newspapers provides further evidence that the readership for the
NYT is mostly liberal and the WSJ is mostly conservative. The Columbia Journalism Review
reported that an internal WSJ poll revealed that 85 percent of readers said they are economic
conservatives and 69 percent said they had conservative social values (Piore 2021). Vox reported
that subscribers of the NYT are “older, richer, whiter, and more liberal than the rest of America”
(Kafka 2022).
Since #MeToo will be an additional test for partisan media bias, this study will be divided
into two time frames: before and after #MeToo. The pre-#MeToo time frame will look for
partisan or structural media bias by comparing news coverage in the NYT to the WSJ. The post-
#MeToo will make a different comparison. Since the theory is that more support for the
movement among liberals will influence coverage in the NYT but not the WSJ if partisan media
bias is controlling, then post-#MeToo coverage needs to be compared to the newspaper’s pre-
#MeToo coverage to detect a change; if post-#MeToo coverage in the NYT is compared to post-
#MeToo coverage in the WSJ, then a significant increase in the NYT and an insignificant
25
increase in the WSJ could produce an insignificant result, thereby masking the influence of
#MeToo.
Since demand-side partisan media bias is driven by the interests of the audience, the
differences between liberals and conservatives would influence how ideologically disparate
newspapers choose which stories to tell (news values) and how to tell those stories (framing),
particularly in a highly polarized environment. However, if structural bias is controlling, then
the professional norms and routines should result in similar coverage between the newspapers.
How Liberals and Conservatives Are Different
Jost, Federico, and Napier (2009) argue that ideology is not simply a way to organize the
world or make heuristic judgements, it is also used as “a device for explaining and even
rationalizing the way things are or, alternatively, how things should be different than they are”
(326). In their review of the literature on political ideology, Jost, Federico, and Napier (2009)
wrote that when compared to conservatives, liberals exhibit greater egalitarianism and openness
to change, and they are more tolerant of uncertainty. Conservatives are more likely than liberals
to avoid uncertainty and ambiguity, have a higher need for order, structure, and discipline, and to
express hostility toward outgroups, particularly those who are stigmatized or low status. There is
also evidence “that conservatism serves system-justifying ends,” meaning that it defends and
justifies the status quo and “views current social arrangements as fair, legitimate, and desirable”
(Jost, Federico, and Napier 2009, 309 and 321).
Moral foundations theory has also been used to explain ideological differences observed
on a variety of issues, including sexual harassment and attitudes towards rape (van der Linden
and Panagopoulos 2019; Barnett and Hilz 2018). This theory contends that there are five basic
foundations which influence moral decision-making: harm/care, fairness/reciprocity,
26
ingroup/loyalty, authority/respect, and purity/sanctity (Haidt and Graham 2007). Research has
shown that liberals perceive harm and fairness as significantly more relevant to moral judgment
than conservatives who view ingroup, authority and purity as much more important (Haidt and
Graham 2007).
In a recent study, van der Linden and Panagopoulos (2019) used moral foundations
theory to explain the results of their survey on sexual harassment. They found that liberals were
more concerned about sexual harassment in general and conservatives were more likely to
punish a liberal (out-group) offender than a conservative (in-group) offender. They argue that
liberals were more concerned with sexual harassment because of the emphasis they place on
harm, fairness, and justice and conservatives were less likely to condemn an in-group offender
because they place a greater value on in-group loyalty.
Barnett and Hilz (2018) investigated the relationship between moral foundations, political
ideology, and acceptance of rape myths. They found that the positive association between
conservative political ideology and rape myth acceptance was mediated by their higher
endorsement of binding foundations which emphasizes social structure and hierarchy.
They contend that liberals have lower levels of rape myth acceptance because they place a
greater value on individualizing foundations which focuses on individual liberty and freedom
from harm; this makes them less likely to engage in victim blaming (a function of rape myths)
and instead ascribe blame to the perpetrator for violating their victim’s rights.
If the NYT and WSJ are tailoring their news content to the ideology of their audience,
then the differences between liberal and conservatives should influence which stories are told,
since news values help journalists decide what stories the audience would find interesting, and
how they are told, thus providing evidence of partisan media bias.
27
Newsworthiness
Gatekeeping is the term used to describe how journalists decide which voices and
messages make it into the news (Bennett 2016) and these decisions are guided by journalistic
values and routines (Shoemaker 2017; Shoemaker and Reese 2014). Which stories make it
through these gates is determined by news values, which Shoemaker and Reese (2014) describe
as an audience-oriented routine since they are used by journalists to “predict what an audience
will find appealing and important” (170). These news values are taught in journalism school and
can also be learned on the job (Shoemaker and Reese 2014; Bennett 2016). Research on
newsworthiness is important because “it goes to the heart of what is included, what is excluded
and why” (O'Neill and Harcup 2009, 162).
Several research studies have developed taxonomies of news values which can apply to
any type of news story (Graber and Dunaway 2015; Harcup and O'Neill 2017; see O'Neill and
Harcup 2009 for a review). Allern (2002) argues that in order to attract an audience, there
should be something in the news story which encourages identification with the consumer such
as people with name recognition, familiar events, or something which is close to them either
geographically or culturally. Other commonly identified news values include timeliness,
conflict, and events which affect a large number of people or have a strong impact on the
audience (Harcup and O'Neill 2017; Graber and Dunaway 2015).
This first series of hypotheses on newsworthiness focuses on the pre-#MeToo time frame.
Since this study compares competing forms of media bias, competing hypotheses are used for
most of the variables. If partisan media bias is controlling, we would expect to see differences in
the news values used by the NYT and WSJ to select stories about SV. Since conservatives are
interested in preserving the status quo and have a need for order and structure, the legal news
28
value, which involves the legal system, should be higher in the WSJ than the NYT (H1a). If
structural media bias is controlling, there will be no differences in the use of the legal news value
by the NYT and WSJ (H1b).
Since conservatives want to avoid uncertainty and they have a greater fear of threat and
loss than liberals, if partisan media bias is controlling then the WSJ would be higher in the
immediacy news value as conservatives would want to know about things happening right now
(H1c). If structural media bias is controlling, there should be no differences (H1d).
Conservatives want to avoid uncertainty and maintain the status quo. Since conflict can
signal to readers that something is uncertain or the status quo may be threatened, the WSJ will
use the conflict news value more than the NYT if partisan media bias is controlling (H1e). If
structural media bias is controlling, there will be no differences between the NYT and WSJ
(H1f).
Since liberals place more value on harm, fairness and justice, there would be more stories
which involve large numbers of people (magnitude) in the NYT when compared to the WSJ if
partisan media bias is controlling (H1g). However, there will be no differences if structural
media bias is controlling (H1h).
There are no differences expected for the other news values coded in this study: rich and
famous, familiarity, cover-up and government action (H1i). There is no reason to believe that
ideological differences would influence stories about rich and famous people in the pre-#MeToo
time frame. While conservatives have a higher need to avoid uncertainty and ambiguity – which
suggests more stories with the familiarity news value in the WSJ – there is no reason to think that
liberals would not be equally interested in knowing about familiar people, organizations, and
institutions involved in wrongdoing since they place an emphasis on harm, fairness, and justice.
29
The cover-up news value is expected to be found frequently in relation to stories about the
Catholic Church. Since that is a major scandal which has been in the news for decades, there is
no reason to expect that ideological differences in the audience would result in significantly
different coverage for this variable. As for the government action news value, both
conservatives and liberals are interested in what the government is doing but for different
reasons. For conservatives, they want the status quo to be maintained and also advocate for
smaller government. Liberals, on the other hand, are more interested in social change which can
come from government action and support government social programs.
#MeToo and Newsworthiness
Previous research has found that a large proportion of articles written about SV involve
court cases (Greer 2003; Hersch and Moran 2013) and in interviews with editors they said that a
story about sexual assault or rape typically is not published unless the offender has been charged
or indicted (Tebben 2018). This research suggests that stories about SV were largely being
generated from the beats of reporters (i.e., crime, courts) and the sources journalists had
established by working on those beats. The widespread sharing of stories via social media
during #MeToo may have encouraged more victims to report what happened to them through
various means such as reporting directly to the company itself, or by sharing their stories with
journalists directly. If this occurred, then the source of the story would have changed from the
legal system to the victim or the company (who would announce an internal investigation,
suspension, or termination), thereby reducing the number of stories with the legal news value and
increasing the conflict news value. If the ideology of the audience was influencing news content,
then support among liberals for #MeToo and the emphasis on harm and fairness among liberals
would result in an increase in the conflict news value and less reliance on the legal news value in
30
the post-#MeToo time frame in the NYT but not the WSJ (H2a). However, if structural bias is
controlling, we would see changes in the same direction – suggesting an influence of #MeToo –
or no change in these news values – suggesting #MeToo had no influence (H2b).
Since support for #MeToo was higher among liberals, this would signal to the NYT that
their audience was interested in this issue thereby providing an incentive to publish more articles
about it. #MeToo resulted in a large amount of coverage, and to keep the reader’s attention they
may have relied more heavily on the familiarity news value than they did in their pre-#MeToo
coverage (H2c). Alternatively, if structural bias is controlling, there will be no change or a
change in the same direction in both the NYT and WSJ (H2d).
The #MeToo hashtag went viral after a Hollywood actor asked people to reply “me too”
to her tweet if they had been sexually assaulted or harassed. Since liberals are supportive of
#MeToo, this signals to the NYT that they are interested in more stories about this issue and
more victims who are associated with Hollywood may have been encouraged to come forward
revealing their experience with rich and powerful men in the industry. If partisan media bias was
controlling, there would be more coverage using the rich and famous news value in the NYT
post-#MeToo and no change in the WSJ (H2e). However, if structural bias is controlling, there
will be an increase or no change at all (H2f).
Since there were several revelations in the wake of #MeToo resulting in a high volume of
news articles, immediacy may have increased. Given the greater support of #MeToo among
liberals, it is expected that this news value will increase in the NYT but not the WSJ if partisan
media bias is controlling (H2g). However, if structural media bias is controlling, there will be an
increase in both newspapers or no change at all (H2h).
31
In the pre-#MeToo time frame, since it is already expected that the NYT will use the
magnitude news value more than the WSJ because liberals emphasize harm and fairness, it is not
expected that support for #MeToo among liberals will lead to a significant increase in this news
value. Therefore, there is no change expected in this news value for either the NYT or the WSJ
(H2i).
The large amount of media attention given to this issue after #MeToo went viral suggests
a response from elected officials is expected, which would result in an increase in this news
value for both the NYT and the WSJ since both liberals and conservatives have an interest in
government actions, although for different reasons (H2j).
There are no differences expected in the cover-up news value in the post-#MeToo time
frame (H2k). As discussed above, it is expected that cover-up will mainly apply to the Catholic
Church which has been a major scandal for decades and there is no reason think that support of
#MeToo by liberals will significantly influence this variable in the NYT or WSJ.
#MeToo and High-Risk Groups
In their 2019 Impact Report, the “Me too.” organization said they were “about ending
sexual violence for everyone, everywhere, while prioritizing the needs of our most vulnerable,
marginalized communities” (8), so one of their goals was to make the experiences of these
groups newsworthy so the response includes these groups. As Tarana Burke, the founder of
#MeToo, wrote in the Washington Post (quoted in Onwuachi-Willig 2018-2019, 105):
What history has shown us time and again is that if marginalized voices – those of people
of color, queer people, disabled people, poor people – aren’t centered in our movements
then they tend to become no more than a footnote. I often say that sexual violence knows
no race, class or gender, but the response to it does. “Me too.” is a response to the
spectrum of gender-based sexual violence that comes directly from survivors – all
survivors. We can’t afford a racialized, gendered or classist response. Ending sexual
violence will require every voice from every corner of the world and it will require those
32
whose voices are most often heard to find ways to amplify those voices that often go
unheard.
Since liberals were more supportive of the #MeToo movement, if partisan media bias is
controlling, then the NYT would be more receptive to this message and increase their coverage
of these groups, especially considering that conservatives generally express hostility towards
outgroups, particularly those who are stigmatized or low status. However, if structural media
bias is controlling, then the media will continue to rely on elites for stories and coverage of these
groups will not change. Furthermore, if the news does cover these groups, it could lend
legitimacy to their experiences and help break down the barriers they currently face to receiving
the help they need to recover.
In order to understand why it is necessary to include stories about high-risk groups, it is
important to discuss the unique risk factors and barriers to receiving services faced by racialized
minority groups, the LGBTQ+ community, low-paid workers, immigrants, and people with
disabilities.
Racialized minority groups. Survivors of SV from racialized minority groups face many
barriers when they seek help. According to the Oregon Coalition Against Domestic & Sexual
Violence (n.d.), these barriers include:
• Cultural and/or religious beliefs that restrain the survivor from leaving the abusive
relationship or involving outsiders.
• Strong loyalty binds to race, culture and family.
• Distrust of law enforcement, criminal justice system, and social services.
• Lack of service providers that look like the survivor or share common experiences.
• Lack of culturally and linguistically appropriate services.
• Lack of trust based on history of racism and classism in the United States.
• Fear that their experience will reflect on or confirm the stereotypes placed on their
ethnicity.
• Assumptions of providers based on ethnicity.
• Attitudes and stereotypes about the prevalence of domestic violence and sexual assault in
communities of color.
• Legal status in the U.S. of the survivor and/or the batterer.
33
• Oppression, including revictimization is intensified at the intersections of race, gender,
gender identity, sexual orientation, ability, legal status, age and socioeconomic status.
Stereotypes about racial groups – particularly the promiscuity of Black and Hispanic
women and the submissiveness of Native American and Asian American women – serve to
devalue them and condone SV committed against them (Ohio Alliance to End Sexual Violence
n.d.). Historically, SV has been used as a tool of oppression against both Black women and
Black men (Pennsylvania Coalition Against Rape 2017) as well as to keep them from gaining
legal or civil rights (Ohio Alliance to End Sexual Violence n.d.).
The “strong Black woman” stereotype, which posits Black women as strong and resilient,
can harm black women and act as yet another barrier to getting the help they need to heal from
the trauma of SV. As Zounlome et al. (2019) explain:
Black women who are survivors of sexual violence…may not be believed by others or
may not want to admit to themselves they experienced an assault due to the perception
that Black women should be strong enough to “fight off” perpetrators. In addition, Black
women may not be given space or time (literally or metaphorically) to heal after sexual
violence, due to their own or others’ perceptions that they ought to “get over it.” These
oppressive societal systems in place seem particularly harmful to Black women. (892)
Furthermore, since women from racially minoritized groups are more likely to be raped
by someone from within their own race, the fear of “being ostracized as betraying their own
communities and/or contributing to racism practiced by white individuals and systems”
discourages them from reporting and receiving services (Ohio Alliance to End Sexual Violence
n.d.).
LGBTQ+ community. Survivors who are LGBTQ+ also face unique barriers to
reporting and receiving services. Filing a police report or seeking medical help may require
them to disclose their sexual orientation or gender identity and the fear of being asked and not
being able to say they are from these communities prevents them from coming forward (National
34
Sexual Violence Resource Center n.d.-a). They may also fear discrimination from police or
service providers, and some are denied services because of their sexual orientation or gender
identity (National Sexual Violence Resource Center n.d.-a).
The fear of reinforcing stereotypes or betraying their community can also serve as a
barrier to receiving services. Since stereotypes of LGBTQ+ people are that they are “bad” or
prey on children, they fear disclosing a sexual assault by a member of this community will
reinforce those stereotypes (National Sexual Violence Resource Center n.d.-a). They may also
feel like they are betraying their community by naming another LGBTQ+ person as the offender
(National Sexual Violence Resource Center n.d.-a).
Some LGBTQ+ people experience multiple layers of discrimination and oppression
because they are also immigrants, members of a racially marginalized group, or have a disability;
these intersecting identities can further isolate victims and prevent them from reporting and
seeking help (National Sexual Violence Resource Center n.d.-a).
Low-income women. Women who are low-wage workers are at increased risk for SV
for several reasons. One risk factor is workplace power disparities which is especially
problematic for tipped workers (Lewis 2018). For example, restaurant workers who rely on tips
often feel pressured to stay quiet in the face of harassment by customers so they do not
jeopardize their tips (Lewis 2018). In areas where the tipped minimum wage is only $2.13/hour,
tipped workers rely almost exclusively on their tips to earn a living since their wages go entirely
towards taxes and they do not receive paychecks (Lewis 2018; Fitzgerald 2019). The reliance on
tips also makes workers vulnerable to abuse from supervisors who have the power to assign them
to shifts which generate more or less tips (Lewis 2018).
35
Physical isolation also puts low-wage workers at higher risk. Hotel workers clean rooms
alone, janitors often clean empty shopping malls or commercial office buildings alone at night,
and farmworkers work in remote agricultural fields – conditions which make it less likely that
that anyone can help them in the event of an attack (Fitzgerald 2019; Lewis 2018).
Finally, women in low-wage jobs often face intersectional discrimination based on their
race, national origin, gender identity, and immigration status (Lewis 2018). For example,
immigrant workers may be unaware of their rights, face language barriers, and fear deportation
(Lewis 2018; more on immigrants below).
Their status as low-wage workers also place them at risk with regard to housing because
they are poor and lack financial resources to move. Women are threatened with eviction if they
resist a landlord’s advances, and some landlords try to control their female tenant by telling her
other men could not come around or live in the tenant’s apartment, a violation which could result
in eviction (Fitzgerald 2019; Reed, Collinsworth, and Fitzgerald 2005). Landlords have also
used their key to access the apartment when the victim was home alone which was often a
prelude to sexual assault (Fitzgerald 2019; Reed, Collinsworth, and Fitzgerald 2005).
Since survivors of SV often need to take time off from work to deal with mental health
symptoms, attend medical or legal appointments, or simply out of fear of seeing the perpetrator
at work, low-wage workers are disadvantaged because they do not have paid time off so if they
do not show up, they do not get paid (Loya 2015). If they have to take time off without pay, it
places an additional strain on their financial situation which puts them at risk of becoming
homeless (Loya 2015). The risk of job loss is also higher among low-wage workers who are
either unable to take time off or are terminated for taking time off (Loya 2015).
36
Immigrant communities. Immigrants are at higher risk for SV for several reasons.
Many immigrants are often afraid to report to law enforcement officials out of fear they will be
deported (National Sexual Violence Resource Center n.d.-b). Victims who are in the U.S. on a
work or student visa risk losing their legal status if they quit their job, leave school, or take an
extended leave of absence (National Sexual Violence Resource Center n.d.-b). Since work visas
are not transferable from one employer to another, this makes victims even more dependent on
their employer and less likely to complain (Fitzgerald 2019).
In the domestic sphere, violence of all kinds can be exacerbated in marriages where a
non-citizen’s legal status depends on his or her marriage to the abuser (National Sexual Violence
Resource Center n.d.-b). In the workplace, perpetrators “view farmworkers and other
undocumented individuals as ‘perfect victims’ because they are isolated due to language and
cultural barriers, generally do not know their rights, and may be vulnerable because they lack
legal status” and “they use these fears and conditions to exert power and control over their
victims” (National Sexual Violence Resource Center n.d.-b).
In 2012, Human Rights Watch released a report, “Cultivating Fear: The Vulnerability of
Immigrant Farmworkers in the US to Sexual Violence and Sexual Harassment,” about immigrant
farmworkers and SV. Their investigation found that in most cases perpetrators were foreman,
supervisors, farm labor contractors, company owners, and others who had the power to hire and
fire workers and provide other benefits such as breaks and better hours. Farmworkers are often
also dependent on their employers for housing and transportation, making them more vulnerable
to abuse. Furthermore, they found that violence and harassment was rarely a singular occurrence
and single perpetrators were known to have abused multiple victims over a period of months or
years.
37
Human Rights Watch (2012) also found that young women and girls, new immigrants,
single women, indigenous women, and women who are LGBTQ+ are even more vulnerable to
SV. Young women and girls more vulnerable as they are less able to defend themselves from
SV because they don’t understand what is going on or how to deal with it. Perpetrators may also
separate parents from their children by having them work in different fields so they can more
easily abuse young girls. The report also said that girls and young women who work with their
families have to contend with the risk that if they reject advances, not only do they risk losing
their own jobs, but the perpetrator could retaliate by firing the victim’s family members as well.
The investigation found that some perpetrators target new immigrants because they
“don’t know many things” (36). The report describes one incident where the older women at the
packing house came to the defense of the new immigrant and all were fired. Single women, or
women not working alongside their husbands, are often singled out for abuse and harassment.
Indigenous people are at increased risk because they do not speak Spanish or English and since
many experienced discrimination in their home countries, they already distrust governments and
authorities. Lastly, the investigation found that people who identify as LGBTQ+, or are even
perceived to be LGBTQ+, are more vulnerable to SV and other forms of harassment.
Those living with disabilities. People with disabilities are at higher risk for sexual assault
than the general population and, like other groups discussed above, face unique barriers to
receiving services (National Sexual Violence Resource Center n.d.-c). To begin with, people
with disabilities “may be limited in their ability to communicate or may be socially isolated due
to their disability or how the disability is perceived by others” and this can be further
compounded by a reluctance of people to believe those with actual or perceived cognitive
38
disabilities (National Sexual Violence Resource Center n.d.-c). The National Sexual Violence
Resource Center describes circumstances which serve as a barrier to reporting:
• A victim may be dependent on the abuser. People with disabilities who rely on
caregivers for some level of routine care are often taught to comply with the
requests or orders of their caregivers. While not inherently bad, this learned
compliance can be readily exploited by an offender. Losing a caregiver’s support
would likely be a consequence of reporting, and the fear of losing a caregiver with
few prospects of a quick replacement is a real barrier to reporting abuse.
• Individuals with disabilities who live in an institutional setting are exposed to a
large number of staff, with frequent staff turnover. This increased exposure can
lead to more risk, and the victim may not understand how to make a report of
abuse.
• Community and legal resources can be difficult to access. This is by no means
limited to victims with disabilities. However, this problem is particularly acute
for people with disabilities who live in a group setting or who rely on an in-home
attendant or caregiver. The fear of having no place to live if one reports sexual
abuse is real and justified. In addition, a person who relies on family, friends, or
another caregiver for personal care and to manage finances, negotiate the maze of
public benefits, or travel to work, play, or appointments may be far less willing to
report sexual assault and risk losing necessary supports.
The importance of discussing the needs of these high-risk groups. When people from
these high-risk groups are not part of the public debate about SV, then their unique experiences
are not part of the discussion of solutions and prevention. A recent report by the National
Women’s Law Center criticized the media discourse around workplace harassment after #MeToo
and the effect it had on reforms:
The media conversation around workplace harassment in the wake of #MeToo going
viral was initially driven by a focus on the experience of women in Hollywood and other
wealthy white women victimized by sexual assault and harassment. The stories of
women of color and low-paid workers who are most likely to experience harassment and
the least likely to have the resources to address it were frequently left out. This critical
omission is reflected not only in the failure to move laws that would be most impactful
for women of color, immigrant women, and low-paid workers…but in reforms that apply
narrowly to sexual harassment only. (Johnson, Ijoma, and Kim 2022, 9; emphasis in the
original)
39
The report argues that “[l]ike sexual harassment, workplace discrimination and
harassment based on race, disability, color, religion, age, or national origin all undermine
workers’ equality, safety, and dignity – and these forms of harassment and discrimination often
intersect in working people’s actual experiences” (Johnson, Ijoma, and Kim 2022, 9). For
example, a Black woman may experience racial slurs in addition to sexual harassment (Johnson,
Ijoma, and Kim 2022). Onwuachi-Willig (2018-2019) argues that the harassment comedian
Leslie Jones in response to her being cast in an all-female remake of Ghostbusters and the
harassment sports journalist Jemele Hill experienced after tweeting that Trump was a white
supremacist, were both racialized and gendered.
While experts have criticized the news coverage of #MeToo for not highlighting the
stories of people who are particularly vulnerable to SV (Fitzgerald 2019; Tyson 2019), to my
knowledge there has not been a large, systematic analysis done to determine the extent to which
these groups were – or were not – covered. In this study I will first code whether victims from
these groups were mentioned in a story and then I will examine those articles further to
determine whether the article discussed why they were at higher risk because of their
membership in a group.
It is important to first analyze the pre-#MeToo coverage to establish whether differences
already exist between the two newspapers. Since liberals are more concerned with
egalitarianism, justice and fairness and conservatives are more likely to express hostility towards
outgroups, particularly people from low status or stigmatized groups, if partisan media bias is
controlling there would be more coverage of them in the NYT than the WSJ (H3a). If structural
media bias is controlling, there should be no differences (H3b).
40
In the post-#MeToo time frame, since support of #MeToo is higher among liberals, if
partisan media bias was controlling we would expect to see increased coverage of high-risk
groups in the NYT but not the WSJ in the post-#MeToo time frame (H3c). If structural bias
were controlling and #MeToo influenced coverage, we would expect to see an increase in
coverage of these groups in both the NYT and WSJ (H3d). However, if #MeToo had no
influence, there would be no change in the coverage of these groups in both the NYT and the
WSJ (H3e).
Framing
Frames are a way for journalists (and the public) to simplify and process the information
they receive and help them make sense of facts (Neuman, Just, and Crigler 1992; Shoemaker and
Reese 2014). Neuman, Just, and Crigler (1992) define frames as “conceptual tools which media
and individuals rely on to convey, interpret, and evaluate information” (60). Gamson and
Modigliani (1987) define a media frame as “a central organizing idea” which provides meaning
to an issue or event (143). Robert Entman (1993) states that framing involves the selection of
certain aspects of a perceived reality which is made salient in a way as to “promote a particular
problem definition, causal interpretation, moral evaluation, and/or treatment recommendation
for the item described” (52; emphasis in the original). Entman further notes that frames can be
defined by what they exclude as well as what they include.
Frames can be conceptualized as either equivalency or emphasis frames (Chong and
Druckman 2007). Equivalency frames present logically equivalent information in different ways
(e.g., 90 percent employment vs. 10 percent unemployment) and this framing alerts an
individual’s perception of the issue (Chong and Druckman 2007). Emphasis frames direct a
person’s attention to qualitatively different considerations about an issue (e.g., free speech or
41
public safety) which are closer to journalistic news coverage (Chong and Druckman 2007; de
Vreese and Lecheler 2012). This study will examine emphasis frames which journalists used to
convey stories about SV.
Frames can be categorized as either generic or issue-specific. Semetko and Valkenburg
(2000) built on the work by Neuman, Just, and Crigler (1992) and Iyengar (1991) and identified
five generic news frames which could be used across different topics: 1) the conflict frame which
emphasizes conflict between individuals, groups or institutions; 2) the responsibility frame which
presents an issue or problem in a way that attributes responsibility for solving it to either the
government or individual/groups; 3) the economic consequences frame highlights the economic
impact a problem, event, or issue will have on individuals, groups, institutions, a region, or a
country; 4) the human interest frame presents a problem, event, or issue with an emotional angle;
and 5) the morality frame places a problem, event, or issue in the context of religious tenets or
morality.
Issue-specific frames are frames which only apply to certain issues or events (de Vreese
and Lecheler 2012), and several studies have examined the frames used in news coverage of SV
(Bing and Lombardo 1997; McDonald and Charlesworth 2013; Pennington and Birthisel 2016;
Schneider and Hannem 2019; Aroustamian 2020; Cuklanz 2020). In their study on news
coverage of sexual harassment, Bing and Lombardo (1997) argued that different frames created
different expectations about what behaviors constituted harassment and suggested different
courses of action in response to it. For example, the judicial frame, which focused on legal
actions, suggested reactive strategies because action is only taken after-the-fact and prevention
would only result from the deterrent effect of sanctions. The victim frame, on the other hand,
42
focuses on the experience of the victim with an emphasis on prevention which includes
understanding the behaviors and conditions which lead to harassment.
Recent research has used case studies to examine the framing of news coverage about
SV. Schneider and Hannem (2019) analyzed the 2016 U.S. presidential campaign between
Donald Trump and Hillary Clinton and found that SV was framed as a political problem where
harm to victims was minimized or ignored, and allegations of misconduct were “dismissed
outright as ‘partisan attacks’ or as politically motivated lies” (756). In her analysis of news
coverage involving Charlie Rose and Matt Lauer, Cuklanz (2020) argued that coverage which
focused on the costs to the organization was problematic because these stories did not discuss the
abuse of power involved or how power operated in the description of abuse which victims said
occurred.
This study will use a combination of general and issue-specific frames to analyze news
coverage to better understand how journalists framed stories about SV. The five primary frames
analyzed in this study are legal, conflict, business, victim, and political (these frames will be
discussed in more detail in Chapter 2). Since framing involves how information is conveyed to
an audience, journalists may frame stories which is consistent with the ideology of the audience.
Only a few studies have examined partisan differences in framing and the results have been
inconsistent: McCluskey (2017) found that audience ideology had little influence on the framing
of school shootings but Haynes, Merolla, and Ramakrishnan (2016) found partisan differences in
the framing of immigration policy.
Since court cases have been identified as newsworthy in previous research on SV (Greer
2003; Hersch and Moran 2013; Tebben 2018), it is expected that a lot of coverage will use the
legal frame in the pre-#MeToo time frame. As with the legal news value, since conservatives are
43
interested in preserving the status quo and have a need for order and structure, the legal frame
should be higher in the WSJ than in the NYT if partisan media bias is controlling (H4a); if
structural media bias is controlling, no changes are expected (H4b).
In previous studies, the conflict frame has been used to code stories about court cases
because of their adversarial nature. Since so much of the coverage is expected to involve the
courts, I wanted to separate those out from other examples of conflict like a survivor coming
forward and revealing what happened to her, either to a journalist or to a current or former
employer. This was important because the barriers to entering the legal system are much higher
than disclosing to a journalist or employer. For example, in criminal cases the police have to
believe the victim and refer the case to the prosecutor who will only file charges if he or she
believes they have strong enough evidence to convict beyond a reasonable doubt. In civil cases,
a lawyer will not even take a case unless they believe the payout will be significant enough for it
to be worth their time (Johnston 2017). This could result in differences between which victims
and offenders are discussed in connection with court proceedings and those who are not.
Since conflict can signal a threat to the status quo which conservatives are interested in
maintaining, the conflict as a frame is expected to be higher in the WSJ than in the NYT if
partisan media bias is controlling (H4c); if structural media bias is controlling then there will not
be any differences (H4d).
Since liberals emphasize harm, fairness, and justice, it is expected that the NYT will use
the victim frame, which focuses on victims and their experiences, more than the WSJ, if partisan
media bias is controlling (H4e). Furthermore, a study by Aalberg and Beyer (2015) found that
conservatives do not think the human-interest frame (which is the equivalent of the victim frame
in this study) is helpful in understanding the issue of irregular immigration, suggesting that this
44
frame will be used more by the NYT. If structural media bias is controlling, then there will be
no differences (H4f).
A subset of the victim frame includes the discussion of harm to victims. This was coded
separately because scholars have criticized news coverage of SV for not discussing it (Schneider
and Hannem 2019; Pennington and Birthisel 2016; Cuklanz 2020; O'Hara 2012; Wolfe and
Brown 2015) and without a discussion of the impact SV has on victims, it is difficult to
galvanize support among the public to solve the problem. Since news coverage which discusses
harm to victims would appeal to liberal readers because of their interest in equality, social
change, and their emphasis on harm and fairness, if partisan bias is controlling the NYT will
discuss harm to victims more than the WSJ (H4g). However, if structural bias is controlling,
there should be no differences between the NYT and WSJ in the discussion of harm to victims
(H4h)
Since conservatives are “sensitive to losses and emphasize financial security” (Thevenot
2018, par. 2) and the WSJ covers “the news through the lens of business” (Piore 2021, par. 7), it
is expected that the business frame will be used more by the WSJ than the NYT if partisan media
bias is controlling (H4i). If structural media bias is controlling then, there will be no differences
in the use of the business frame (H4j)
I do not expect differences in the political frame, which includes polling and strategy,
because the “game frame” is so ubiquitous in news coverage (Cappella and Jamieson 1997;
Patterson 1993, 2016) that it is unlikely to be influenced by the ideology of the audience (H4k).
#MeToo and Framing
Just as with the news values, if journalists were able to use sources other than the legal
system to generate stories because so many victims were coming forward, then the proportion of
45
articles using the legal frame should go down and the number of articles using the victim and
conflict frame should increase. Since a goal of the #MeToo movement is to highlight victims,
and liberals support #MeToo more than conservatives, if partisan bias is controlling then the
legal frame should decrease and the victim and conflict frame – which includes survivors coming
forward – will increase in the NYT but not the WSJ (H5a). If structural bias is controlling and
#MeToo has an influence, then those changes would occur in both the NYT and WSJ (H5b). If
#MeToo has no effect, then there will be no change in either newspaper (H5c).
Part of the mission of the #MeToo movement is to help survivors heal from the trauma of
SV, so discussing the harm to victims is important so other survivors know that they are not
alone in their experience. It can also help convey the seriousness of the problem in order to
create public support for resources and programs to end SV. Since liberals support #MeToo
more than conservatives, if partisan bias is controlling then there should be an increase in the
discussion of harm to victims in the NYT but not the WSJ post-#MeToo (H5d). If structural bias
is controlling and #MeToo influences coverage, we should see these changes in both the NYT
and WSJ (H5e). If #MeToo has no effect, there will be no changes in the discussion of harm to
victims in either newspaper (H5f).
Since so much of the #MeToo coverage involved the workplace, it is expected that the
business frame will increase post-#MeToo but not significantly because it is not a frame which
appeals to liberals who are concerned with harm and fairness and the WSJ already frames much
of its news coverage through the lens of business (H5g).
There were several politicians who were accused of sexual misconduct after #MeToo
went viral so it is expected that the use of this frame will increase in both the NYT and WSJ as
46
the parties, political observers, and politicians themselves debate the implications and try to
control the narrative (H5h).
Episodic and Thematic Framing
Framing not only affects support for policy actions, but it can also affect the perception of
a problem’s severity (Wiest, Raymond, and Clawson 2015) as well as who – or what – is to
blame for a problem (Iyengar 1991). Attribution of blame can be influenced by episodic or
thematic framing: episodic framing focuses attention on an individual or event without
acknowledging the broader social or political context in which it is occurring while thematic
framing connects issues to a broader social, political and economic context (Iyengar 1991).
People who consume episodic news stories are more likely to blame the individuals in the story,
thereby making them responsible for resolving it. On the other hand, people who consume
thematic stories are more likely to think that societal or institutional responses are needed
(Iyengar 1991). This type of framing can also lead to expectations about possible solutions,
whether they be individual solutions or policy recommendations by government or other
organizations.
Previous research on SV has found that the majority of coverage is episodic (Mejia,
Cheyne, and Dorfman 2012; De Benedictis, Orgad, and Rottenberg 2019; Greer 2003). This
type of reporting has been criticized for focusing attention on individuals and events rather than
the larger structural and systemic problem of gender inequality (De Benedictis, Orgad, and
Rottenberg 2019; McDonald and Charlesworth 2013). For example, news coverage of sexual
harassment which focuses on court cases has been highlighted as problematic because it situates
the issue in the private sphere (as a conflict to be resolved between two parties) which precludes
47
even a debate about potential institutional or structural responses (McDonald and Charlesworth
2013).
The heavy reliance on episodic reporting can be partly explained by economic and
organizational constraints. Episodic stories are easier to write under time constraints and cost
less to produce than thematic stories which require background research and data collection
(Iyengar 1991). Individual stories, exemplars, and human-interest stories are known to attract a
reader’s attention (Gans 1979; Kim, Carvalho, and Davis 2010; Aarøe 2022). Exemplars are
also perceived as more newsworthy and some news values – such as immediacy and conflict –
are best portrayed through specific acts and individuals (Greer 2003; Skidmore 1995; Cavender
and Mulcahy 1998).
Since episodic coverage is so widespread and common, it is not expected that the
ideology of the audience will influence this variable, therefore there are no differences expected
between the NYT and WSJ in the pre-#MeToo time frame (H6).
#MeToo and Episodic/Thematic Framing
In the aftermath of #MeToo going viral, it seemed like every day another powerful and
influential individual was facing allegations of sexual misconduct. This would suggest an
increase in episodic coverage, but since it is already anticipated that episodic coverage will
dominate in the pre-#MeToo time frame, there is no reason to expect that the #MeToo movement
will affect this variable in a significant way because of the ceiling effect, so no changes are
expected in the NYT or WSJ in the post-#MeToo time frame (H7).
Solutions
As discussed above, research has shown that media coverage influences our sociocultural understanding of SV and possible solutions. My interest in exploring solutions found in
48
news coverage follows the view of De Benedictis, Orgad, and Rottenberg (2019) that journalists
are not responsible for devising solutions, but rather they should inform the public about
potential solutions.
In their study of elder abuse, Mastin et al. (2007) identified individual-level solutions and
societal-level solutions. Individual-level solutions were those that people could implement on
their own such as how to select a financial representative, being aware of financial scams,
identifying signs of abuse, and reporting abuse. Societal-level solutions were actions which
could be taken by the government such as regulation, legislation, or intervention.
Previous research has found differences in the types of solutions to social problems
emphasized in liberal and conservative media. Kenix and Jarvandi (2019) examined the framing
of refugees and found that liberal newspapers emphasized governmental responsibility for the
refugee crisis while conservative newspapers emphasized individual responsibility. Kim,
Carvalho, and Davis’ (2010) study on poverty similarly found that conservative newspapers were
significantly less likely to mention societal causes and solutions. These findings correspond to
liberal and conservative political platforms (Kenix and Jarvandi 2019; Kim, Carvalho, and Davis
2010) and suggest that the WSJ will emphasize individual-level solutions (e.g., policy changes at
an organization or encouraging others to speak out) while the NYT will emphasize societal-level
solutions (e.g., government legislation) if partisan media bias is controlling (H8a). However,
since it is expected that both newspapers will publish more episodically framed articles, this
suggests that both newspapers will emphasize individual-level solutions (H8b).
#MeToo and Solutions
The goal of Milano’s tweet and the massive response was to bring awareness to the
pervasiveness of the problem with the hope that something would be done to end SV, either
49
through individual-level solutions (e.g., policy changes in workplaces) or societal-level solutions
(e.g., government legislation). Since the #MeToo movement has more support among liberals
than conservatives, the discussion of solutions should increase in the NYT but not the WSJ if
partisan media bias is controlling (H9a). If structural bias is controlling and #MeToo influenced
coverage, then there would be an increase in both the NYT and WSJ (H9b); if MeToo had no
effect then there would be no change in the discussion of solutions in either newspaper (H9c).
In the next chapter, I will discuss the methodology used to evaluate news coverage of SV
for media bias, including the data selection process, the sampling procedure used, and the
variables of interest coded for analysis. The chapter concludes with some descriptive statistics of
the sample, including the frequency of the news values and primary frames found in the NYT
and WSJ in the pre-and post-#MeToo time frames.
50
Chapter 2
Methods
This study will use a quantitative content analysis to examine whether ideologically
disparate newspapers covered SV differently in a highly polarized political environment. A
quantitative content analysis allows a systematic and replicable examination of news content and
provides a way to describe the communication as well as the ability to analyze relationships
using statistical methods (Riffe et al. 2019). I will also use qualitative examples from my dataset
to help illustrate the findings from my quantitative analysis.
Data Selection
Since a key theoretical argument in this study is that professional journalistic norms and
routines produce structural media bias as well as constrain partisan media bias, the data used in
this study is limited to news articles; editorials, opinion, and commentary are excluded since the
authors of these articles are not constrained by those same norms and routines. Furthermore,
there is a wall of separation between news and opinion which both the NYT and WSJ have
publicly reiterated (The Wall Street Journal n.d.; Lehman 2015).
News articles were downloaded from the Factiva database using the search terms
“sexual* harass*” or “sexual misconduct” or “sexual* assault*” or “rape*” or “raping”. Since
#MeToo serves as an inflection point in this study, the time frame begins two years before the
hashtag went viral in October 2017 and ends two years after: October 1, 2015 – September 30,
2019.
The ability to analyze the issue of SV from different ideological news sources is an
integral part of my research question. Previous research has found The New York Times (NYT)
to be liberal and The Wall Street Journal (WSJ) to be conservative (Habel 2012), and other
51
studies have studied ideological differences in the media using this categorization (Hmielowski,
Hutchens, and Beam 2020; Kenix and Jarvandi 2019). The search was limited to these two
national newspapers based in New York City. Given that more people prefer to get their news
online versus print (Mitchell 2018), articles were downloaded from the online versions of these
two news sources. Using the online version of news articles had the additional benefit of
eliminating space constraints as a possible explanation for any differences found. Identical
duplicates were removed as well as republished news, recurring pricing and market data,
personal announcements, obituaries, calendars, captions, letters, weather news, food items,
routine traffic reports, and sports and recreation stories.
For the NYT, columns, reviews, and commentaries/opinions (which includes editorials)
were also removed.3
This resulted in a total of 6,453 articles. During the downloading process, I
manually de-selected 767 news briefings, corrections, transcripts, and reader comments.4
The
total number of articles downloaded for the NYT was 5,686.
For the WSJ, commentaries/opinions, news digests, reviews, and editorials were
removed.5
This resulted in 2,569 articles. During the downloading process, I manually deselected 208 articles because they were news briefings (including What’s News, The 10-Point,
3 The same filters were not available for both news outlets; while the WSJ had a “news digest” filter, the NYT did
not and that is why “news digest” was not used for the NYT articles.
4 The same categories were not removed from both the NYT and WSJ because some categories were relevant to a
particular outlet while others were not. For example, the NYT did not have any “Photos of the Day” articles and the
WSJ did not have any reader comments. In addition, these articles were located by using the “find” function
(Control + F) and searching for key terms on the results page. This was imperfect but anything which was missed
during this process was corrected when I screened the downloaded articles to remove irrelevant ones.
5 The types of articles removed are not the same for each news source because Factiva appears to be inconsistent
with how they categorize articles. For example, when I realized that I did not have any articles from the WSJ for
2019, I discovered that news articles had been categorized as “columns” – this was not a problem with the NYT
articles. Columnists such as Peggy Noonan were removed when I excluded “commentaries/opinions.” Furthermore,
excluding commentaries/opinions removed some WSJ editorials but not all, so it was necessary to filter out
editorials as well.
52
and The Score), photo articles, or corrections. The total number of articles downloaded by from
the WSJ was 2,361.
In reviewing the articles, I realized that the term “sexual abuse” was frequently used to
discuss this social problem, so I conducted a second search using the search terms “sex* abuse*”
or “sexually abusing” or “molest*”. This search in the NYT (with columns, reviews, and
commentaries/opinions removed) returned 1,794 news articles. After manually deselecting
briefings, corrections, transcripts, reader comments, and duplicates from the first search (1,145
articles), 649 articles were downloaded. This brings the total of articles NYT articles
downloaded to 6,335.
The second search in the WSJ returned 622 news articles after commentaries/opinions,
news digests, and reviews were removed.6
After manually deselecting the same categories and
duplicates from the first search (287 articles), 335 articles were downloaded bringing the total to
2,696.
After all the articles were downloaded, the dataset needed to be cleaned because some
articles which should have been removed during the filtering process (like personal
announcements) were still present. Additionally, there were some news briefings/roundups and
transcripts which were missed when I manually de-selected articles prior to download.
Furthermore, some articles were primarily about another topic and the issue was simply
mentioned. Since the criteria of newsworthiness and framing of news stories on SV are key
variables in this study, articles which were not primarily about this issue needed to be removed.
Irrelevant articles were defined as ones which did not reference the topic of SV in the headline or
6 There were no editorials in the second search to be removed.
53
the first five paragraphs.7
This process removed 2,836 articles from the NYT and 1,057 from the
WSJ (a total of 43 percent of all articles downloaded). A combined total of 5,138 articles
remained in the dataset for analysis (see Table 1 for a complete breakdown).
Table 1. Summary of Data Selection and Sampling
New York
Times
Wall Street
Journal
Results for keyword search 1:
sexual* harass* or sexual misconduct or
sexual* assault* or rape* or raping 6,453 2,569
Results for keywords search 2:
sex* abuse* or sexually abusing or molest* 1,794 622
De-selected prior to download (search 1 and 2) 1,912 495
Articles downloaded in search 1 and 2 6,335 2,696
Irrelevant articles removed 2,836 1057
Total articles in dataset 3,499 1,639
Sample (10 percent) 350 163
This high number of irrelevant articles warrants some further investigation. A small
number of articles were removed because they were a news digest, an entertainment review or
recommendation, reader comments, or a key word was used metaphorically, such as Donald
Trump denouncing the Trans-Pacific partnership as “a rape of our country” (Swanson 2018).
The vast majority were removed because the topic did not appear in the headline or the first five
7 References to the topic of SV included acts of sexual harassment or assault (e.g., inappropriate touching, groping,
rape) as well as well as sex trafficking, sexual exploitation, child pornography, revenge porn, consent, and sex
offenders. Articles which discussed fictionalized or dramatized versions of sexual harassment or assault (e.g., the
movie Spotlight) were removed, but articles which discussed documentary films (e.g., Surviving R. Kelly, Finding
Neverland) remained in the sample. Review of the first five paragraphs was chosen because the inverted pyramid
style of reporting places the most important facts earlier in the story (Bennett 2016).
54
paragraphs. Some examples of these articles included using SV to add historical context (a story
about LGBTQ+ people and religion mentions that Catholics have debated the role of
homosexuality in sexual abuse scandal; an African American prisoner on death row and a brief
discussion regarding how Black men have been executed for nonlethal crimes such as robbery
and sexual assault in the past; a midterm election story about Senator Dianne Feinstein which
contains criticism of how she handled the letter from Christine Blasey Ford). Other articles are
about agencies or organizations who provide services (like adoption agencies or mental health
units on college campuses) and some of their clients are survivors of SV. Almost all articles
included links to other news stories in the form of reading recommendations, and some of the
titles of those recommendations included key words used to search for articles.
Sampling
I used systematic sampling because it allows for a more even selection of news articles
over time and by source. As seen in Figure 2, which displays the number of articles published in
the NYT and WSJ over the four-year period being analyzed, there was a large spike in coverage
after #MeToo went viral in October 2017, which suggests the media was responding to public
opinion. There was another increase in coverage around September and October of 2018 which
corresponds to the Supreme Court confirmation hearings for Brett Kavanaugh and the 2018
midterm elections. I was concerned that a straight random sample would result in a
disproportionate number of articles from these two spikes in coverage and much fewer articles in
the pre-#MeToo time period. I was also concerned that the higher number of articles published
by the NYT (see Table 1) would result in a higher proportion of articles being included from that
newspaper. Since a highly uneven proportion of articles from the NYT and the post-#MeToo
time period would make comparisons between newspapers and across time more challenging, I
55
opted for systematic sampling. In order to sample 10 percent of each newspaper, I first ordered
the articles by news outlet and then by date and then I coded every 10th article. I used a random
number generator to select a number between 1-10 to determine which article would be the first
one to be coded.8
One limitation of this method of sampling is that because there were so many
more articles published in the NYT than in the WSJ, there was more time between articles in the
WSJ than the NYT. For example, there was only ten days between the first and second article
coded in the NYT, but there were 47 days between the first and second article coded in the WSJ.
Figure 2. Number of Articles in the Full Sample by Month
If an article was discovered to be irrelevant during coding, it was coded as irrelevant and
replaced with either the article above or below it, whichever was the closest to the date of the
irrelevant article. If both articles were from the same date as the article being replaced, or the
8 Random.org to select a random number.
56
difference was the same (e.g., the article above is the day before and the article below is the day
after), then the article below the irrelevant article was selected for coding.
Intra-rater Reliability
Intra-rater reliability was checked after every 100th relevant article which was coded. I
used a random number generator to select three numbers between 1-75 to select the article to
recode; I excluded the last 25 articles so that I did not re-code an article I recently coded. I used
this method for the first 200 articles; after that, all articles had the possibility to be selected for
recoding with the exception of the most recent 25 articles. So, after coding 300 articles, I
randomly selected three numbers between 1-275 to recode; after coding 400 articles, I randomly
selected three numbers between 1-375 to recode, and after coding 500 articles, I randomly
selected three numbers between 1-475 to recode. A total 15 articles were re-coded and the
average of the intra-reliability scores was 0.9 (or 0.898933).
Variables of Interest
In order to examine whether partisan or structural media bias is dominant in news
coverage of SV in the NYT and WSJ, this study examines which stories were deemed
newsworthy for publication and how those stories were framed.
Newsworthiness
The news values used for this study were developed from four taxonomies: two which
apply broadly to all issues in the news (Harcup and O'Neill 2017; Graber and Dunaway 2015)
and two used in studies specifically analyzing news coverage of SV (Greer 2003; Cheit, Shavit,
and Reiss-Davis 2010). After pilot testing, a list of 10 news values were coded but I am not
reporting the results for two of them, speculation and NGO (see Appendix A for more
information). The eight news values analyzed for this study are as follows:
57
• Magnitude (multiple offenders or victims, or large numbers of people involved)
• Legal system (civil or criminal proceedings, law enforcement)
• Immediacy (just occurred or was revealed, includes announcements; offender atlarge)
• Rich and/or famous (e.g., wealthy people, entertainers, professional athletes, person
has name recognition with the public)9
• Familiarity – the person, organization, or institution is familiar to the audience10
• Conflict (war, arguments, controversies, e.g., a victim, most often a woman,
identifying a person she says violated her)
• Cover up (element of cover up to the story)
• Government action (e.g., statements, press conference, announcement, hearings,
legislation, reports, can be agencies or officials, including elected politicians)11
Since a single news story can contain multiple news values (Cheit, Shavit, and Reiss-Davis
2010), this variable was coded as “check all that apply.”
High-risk groups. The inclusion of high-risk groups was part of the analysis of
newsworthiness. High-risk groups were identified by previous research and advocacy
groups: racialized minority groups (Basile et al. 2016), LGBTQ+ (Kearl 2018; Raghu and
Suriani 2017; Basile et al. 2016), immigrants (Raghu and Suriani 2017; Fitzgerald 2019;
Human Rights Watch 2012), those who are poor (Kearl 2018; Basile et al. 2016), and
those with a medical condition (Khalifeh et al. 2015; Kearl 2018; Van Deinse et al.
2019). A binary variable (yes/no) was created for each group and an article was coded as
“yes” if a victim was identified as belonging to one or more of these groups and “no” if
not.
9 Rich and famous people could also be coded as familiar. Since not all wealthy people are familiar to the public, it
was important to keep this as a separate news value.
10 Organizations can include companies (e.g., CBS, Uber) or sports teams and within that category includes
familiarity with their products (e.g., a movie by a production company); institutions can include government bodies
(e.g., state legislature, Congress) or agencies (e.g., Border Patrol).
11
Government action can be foreign or domestic and includes law enforcement and the courts.
58
Framing
This study used a combination of issue-specific and generic frames to identify the
primary frames used by journalists in their stories on SV (see Appendix A for details on
pilot coding and how the frames were chosen). The five frames used were as follows:
1. Legal frame – criminal and civil cases, government investigations/hearings,
internal investigations, and legislation
2. Conflict frame – SV which occurs it the context of a conflict (e.g., war) or a
controversy (e.g., a survivor, most often a woman, identifying a person who
violated her)
3. Victim frame – article focuses on victims and their experience with SV
4. Business frame – discusses SV as it relates to the impact it has on business (e.g.,
stock prices, revenue, value of the company, successor or interim replacement
being named, implications of CEO resigning or being terminated)
5. Political frame – discusses the political implications of allegations, (e.g., impact on a
campaign, how allegations influenced voters), political strategy for dealing with
allegations including how political parties are handling allegations differently, offender
evading political consequences (e.g., Trump), politician resigning or not seeking reelection
While a single news story can contain multiple frames, Bing and Lombardo (1997)
found in their analysis of news coverage on sexual harassment that articles rarely
contained more than one frame. In order to simplify the analysis, only one primary frame
for each news article was coded.
Harm to victims. Harm to victims is part of the victim frame and was coded because
scholars have criticized news coverage for not discussing it and acknowledging the harm to
victims is important to the #MeToo movement. What constituted harm to victims was developed
from previous research (Kearl 2018; Basile et al. 2016) and from reading articles during pilot
coding. The categories included physical harm, psychological harm (e.g., depression, anxiety,
PTSD, eating disorders), suicide or attempted suicide, substance abuse, retaliation or fear of
retaliation, economic or professional harm (e.g., job loss, having to look for another job),
stopping a hobby or activity, changing routine, and school-related loss (e.g., grades suffered,
59
changing schools, changing major or advisor, withdrawing from a course, dropping out of school
or a training program), had to move, or was murdered. This variable was coded as “yes” if it was
discussed and “no” if it was not.
Episodic and Thematic Framing
The categories for episodic and thematic framing were adapted from Barnes and Hevron
(2018) who created five categories based on Iyengar (1991). An episodic article was one which
presented the issue by discussing a specific incident, individual or event. A primarily episodic
articles mostly focused on victim(s) or offender(s) but provided additional context which
illustrated something broader than the experience of the personalized example. A mixed article
discussed SV at an organization or as a problem in an industry. In both thematic and primarily
thematic articles, the issue was presented by using contextual, historical, or statistical
information; the difference was that specific examples were used in primarily thematic articles
but not in thematic ones.12
Solutions
Solutions are a subset of the episodic and thematic framing category. Solutions were
coded as present or not present (yes/no) and information about the solution was recorded and
later coded as individual and societal-level solutions based on the categories created by Mastin et
al. (2007) and adapted for the issue of SV. Individual-level solutions are those which are
specific to an individual or organization and include policy changes at a company or
organization, public pressure campaigns targeted at specific companies, hiring a firm for
independent investigation and advice, industry-specific recommendations, encouraging others to
speak out. Societal-level solutions are those which are broad and applied to large segments of
12 For more complete definitions, please see the codebook in Appendix B.
60
the population. Examples include government legislation, government installation of
surveillance cameras in public places, new ratings system to combat gender stereotypes, and
groups to combat SV like Male Champions of Change in Australia.
The discussion of results will be divided into two chapters. I will first review the
findings related to newsworthiness, both pre- and post-#MeToo, and then in the following
chapter I will review the pre- and post-#MeToo results for the framing variables, including
episodic/thematic and solutions.
Descriptive Statistics
The sample for this study is divided into two time frames: two years before #MeToo and
two years after. In both the NYT and WSJ, there was a nearly 150 percent increase in the
number of stories published about SV in the post-#MeToo time frame. In the NYT, there were
101 articles pre-#MeToo and 249 articles post-#MeToo; in the WSJ there were 47 articles pre-
#MeToo and 116 articles post-#MeToo (see Figure 3).
Figure 3. Number of Articles in the NYT and WSJ, Pre- and Post-#MeToo
61
Figures 4 and 5 below represent the frequencies of the news values found in the NYT and
WSJ, and the top three news values are the same for each newspaper in both times periods. In
Figure 4. Frequency Bar Chart for News Values in the NYT
Figure 5. Frequency Bar Chart for News Values in the WSJ
62
the pre-#MeToo time frame for both the NYT and WSJ, immediacy was the most frequent news
value, followed by magnitude and legal. In the post-#MeToo time frame, immediacy remained
the most frequent news value in both newspapers but they differed slightly after that: in the NYT
magnitude was the second most frequent news value followed by familiarity, and in the WSJ
familiarity was second followed by magnitude.
Figures 6 and 7 represent the frequency of primary frames in the NYT and WSJ,
respectively. The legal frame, which includes criminal and civil cases, is by far the most
frequently used frame to tell the story of SV in both the NYT and WSJ. The second most
frequently used frame is conflict, but that appears to be driven by an increase in the post-
#MeToo time frame.
Figure 6. Frequency Bar Chart for Primary Frames in the NYT
63
Figure 7. Frequency Bar Chart for Primary Frames in the WSJ
In the next two chapters I will discuss the results for the analysis of news values and
framing which provides strong support for the structural media bias hypothesis: both the NYT
and WSJ used similar criteria to select which stories to publish and similar frames to discuss the
issue of SV.
64
Chapter 3
Newsworthiness Analysis
This study is comparing two main forms of media bias to determine which one is
controlling in the coverage of a partisan issue in ideological disparate newspapers in a politically
polarized environment. If partisan media bias is controlling, we would expect to see different
coverage of SV in the NYT and WSJ; however, if structural media bias is dominant, then
coverage in both newspapers would be similar. This chapter will review the newsworthiness
results of stories in the NYT and WSJ. The first test of media bias will compare the NYT to the
WSJ in the pre-#MeToo time frame. The second test of media bias will examine news coverage
post-#MeToo, comparing each newspaper to its own pre-#MeToo coverage to determine if the
movement influenced coverage; if partisan media bias is controlling, we would expect changes in
one newspaper but not the other.
There was a total of 513 articles coded in this study, 350 in the NYT and 163 in the WSJ.
When broken down further, in the pre-#MeToo time frame there were 101 articles in the NYT
and 47 in the WSJ; in the post-#MeToo time period, there were 249 articles in the NYT and 116
in the WSJ. The unit of analysis was the article, and the variables were dichotomous (yes/no).
Statistical analysis was performed on each variable in which there were five or more
observations. A proportion Z-test was used to test variables in which there were more than 10
observations in every cell ("Two Proportion Z-Test" n.d.) and a Fisher’s Exact test was used for
variables which had fewer than 10 observations in every cell ("Fisher’s Exact Test" n.d.).
Pre-#MeToo Analysis of News Values
Most of the hypotheses in this section were presented as competing hypotheses: one
which predicted partisan media bias (i.e., differences in coverage between the NYT and WSJ)
65
and the other which predicted structural media bias (i.e., no differences in coverage between the
NYT and WSJ). The results below show that in the pre-#MeToo time frame, the NYT and WSJ
relied on similar news values, thus providing support for the structural media hypothesis (see
Table 2 for a summary of the results).
Table 2. Summary of Results for News Values Pre-#MeToo in NYT and WSJ
News Value NYT WSJ P-value
Legal 62% 62% p = 0.9373
Immediacy 83% 96% p = 0.03323*
Conflict 10% 11% p = 1
Magnitude 73% 66% p = 0.3618
Rich/Famous 27% 28% p = 0.9059
Familiarity 50% 60% p = 0.3027
Cover-up 6% 4% p = 1
Government Action 47% 47% p = 0.9752
*p < .05
Notes: October 2015 – September 2017; n = 101 articles in the NYT and 47 articles in the WSJ
Legal news value. The first hypothesis for the legal news value predicted that if partisan
media bias were controlling, the WSJ would use this news value more frequently than the NYT
since conservatives are interested in preserving the status quo and have a greater need for order
and structure (H1a); if structural bias is controlling there would be no differences (H1b). The
results revealed no differences between the NYT and WSJ pre-#MeToo, thus confirming H1b.
A total of 62% of the articles in the pre-#MeToo time frame were coded as using the
legal news value. This finding is consistent with previous research which found that a large
66
proportion of articles about SV were told in the context of a court case, either civil or criminal
(Greer 2003; Hersch and Moran 2013). The majority of articles in this category were reports on
well-known individuals (e.g., Bill Cosby, Julian Assange) and well-known organizations or
institutions (law enforcement, Columbia University, Fox News, Catholic Church, and elite
private schools). Just over a third of the coverage with this news value were about four people
and organizations: Bill Cosby, Julian Assange, the Catholic Church, and Fox News.
Articles coded for the legal news value focused on various stages of court proceedings or
investigations by law enforcement. For example, Cardinal George Pell was under investigation
by the authorities in Australia for sexual abuse, Julian Assange was questioned by Swedish
authorities regarding rape allegations in Sweden, the trial date was set for Bill Cosby and the
judge ruled on various motions by the prosecution and defense, a police officer in Oklahoma was
convicted of rape, and former Speaker of the House Dennis Hastert being sentenced to 15
months in prison. The civil cases included those filed by Gretchen Carlson and Andrea Tantaros
against Fox News for sexual harassment and retaliation, a victim filing defamation lawsuit
against Bill Cosby for calling her a liar after she came forward, and civil lawsuits against the
Catholic Church for child sexual abuse.
Immediacy news value. Since conservatives want to avoid uncertainty and have a
greater fear of threat and loss than liberals, H1c predicted that if partisan media bias is
controlling, the immediacy news value would be present more in the WSJ because they would
have a higher need to know about things happening right now; H1d predicted no differences.
The results showed that WSJ the immediacy news value was present in 96 percent of their
articles which was significantly higher than the NYT (83 percent), z = 2.129225, p = 0.03323.
H1c, which supports the partisan media bias hypothesis, is confirmed.
67
The immediacy news value was the most common news value coded for both the NYT
and the WSJ. These included something happening in a criminal case such as a babysitter
pleading not guilty to charges of murder and aggravated sexual abuse of a 16-month-old boy, a
transgender woman reporting that she was sexually assaulted at the Stonewall Inn, an officer at
Riker’s being indicted on rape charges. Other articles discussed an offender apologizing for their
behavior, advertisers who canceled advertising spots on The O’Reilly Factor after the NYT
revealed O’Reilly and Fox News had settled several sexual harassment lawsuits for millions of
dollars, and events in the 2016 presidential campaign (e.g., Access Hollywood tape, polling).
In terms of the difference between the two newspapers, the NYT did more stories which
discussed the aftermath of an event which were not initiated by the actions of a person or
organization (such as a press conference or a ruling in a case), thus providing more contextual
stories about SV than the WSJ. For example, after a “star” professor at Yale was cleared of
sexual harassment against an undergraduate student, the university was criticized for their history
of not taking the issue seriously (Remnick 2016). In another article, the NYT used the case of a
man from Long Island who was arrested on charges of sexually abusing his adopted children and
endangering the welfare of two foster children to discuss the risks of out-of-state foster care as
some of the children had come from the State of Washington (Stewart and Goldstein 2016).
These types of stories align well with the emphasis liberals place on harm, fairness, and justice,
suggesting that the NYT invested in these stories to appeal to their readers even though they
were not something which “just happened.” Even though this coverage aligns with the ideology
of the NYT readership, the difference may be explained by organizational priorities: the WSJ
simply does not want to pay their reporters to write stories about the aftermath of an event and
focuses their attention on events which just occurred.
68
Conflict news value. Regarding the conflict news value, the partisan media bias
hypothesis, predicted that this news value would be more prominent in the WSJ since conflict
can signal a threat to the current order and conservatives have a higher need to avoid uncertainty
and maintain the status quo (H1e). The structural bias hypothesis predicted that there would be
no differences between the WSJ and NYT (H1f). The results show no differences between the
two newspapers in the pre-#MeToo time frame, thus confirming H1f.
There were only 15 articles coded as conflict (if the legal system had been included, this
would have been much higher). Donald Trump and the 2016 presidential campaign accounted
for 40 percent of the articles in this category. This included the release of the Access Hollywood
tape, Trump’s initial apology video, reports of more women coming forward, and the effect it
was having on Trump’s presidential campaign. Other articles were about the atrocities in South
Sudan which included women of all ages being gang-raped, and the debate over the termination
of a software engineer at Google who criticized the company’s diversity efforts.
Magnitude news value. Since liberals place more of an emphasis on harm, fairness, and
justice, H1g predicted that there would be more articles in the NYT which feature the magnitude
news value, which indicates something which involves or affects large numbers of people, if
partisan media bias is controlling. The structural media bias hypothesis, H1h, predicted no
differences between the NYT and WSJ. The results showed no differences between the two
newspapers, thus confirming H1h.
Magnitude was the second most frequent news value coded in the pre-#MeToo time
frame (71 percent), and just over a third of the articles were about Bill Cosby, Donald Trump,
Fox News, and the Catholic Church. A vast majority of the stories coded as magnitude featured
multiple victims, a subcategory which was coded in 61 percent of the pre-#MeToo sample,
69
strongly suggesting that the stories with only one known victim are less newsworthy. This
finding is consistent with guidelines for journalists published by the Poynter Institute (2017)
after #MeToo in which they advised that journalists had a greater obligation to investigate stories
which involve many victims, which can also be taken to mean they also have a greater obligation
to report stories involving multiple victims.
Bill Cosby was accused of drugging and sexually assaulting at least 50 women and after
the Access Hollywood tape was released, multiple women came forward stating that Donald
Trump had sexually assaulted them. News articles about the Catholic Church and Fox News
featured multiple victims and offenders. The world-wide scandal involving the Catholic Church
involves numerous victims, most of whom were children at the time, who said that priests
sexually abused them. The coverage of SV at Fox News included multiple victims (e.g.,
Gretchen Carlson, Andrea Tantaros, and the multiple women who settled lawsuits with Bill
O’Reilly) as well as multiple offenders (Roger Ailes, Bill O’Reilly, and Eric Bolling).
There were several other instances of magnitude discussed in both newspapers. One
article in the NYT reported on the results of an investigation by the Boston Globe’s Spotlight
team which revealed that over 200 students from at least 67 private schools in New England had
accused teachers, staff, or administrators of sexually abusing or harassing them. An article in the
WSJ discussed the results of a survey in which 50% of women working in the advertising
industry reported they experienced sexual harassment, and another was an announcement by the
Alameda County District Attorney that seven police officers were charged with a sexual act
involving a minor.
Rich/famous, familiarity, cover-up, and government action news values. There were
no differences expected for the remaining four news values: rich/famous, familiarity, cover-up,
70
and government action (H1i). The ideological differences between liberal and conservatives do
not suggest that one group would be more interested than the other in people who are rich and
famous. The familiarity news value would appeal to conservatives because they have a higher
need to avoid uncertainty and ambiguity. However, this news value may also appeal to liberals
in this context because they emphasize harm, fairness, and justice and they would want to know
about wrongdoing by people or organizations which are familiar to them. There were no
differences predicted in the cover-up news value because it was expected to be associated with
the Catholic Church and the story has been in the news for decades. With regard to the
government action news value, there were no differences expected but for different reasons:
conservatives are interested in what the government does because they want to maintain the
status quo and advocate for “smaller” government while liberals are interested in social change
which can come from government policy and programs. There were no differences between the
NYT and WSJ in the pre-#MeToo time frame for these four news values which confirm H1i,
thus providing support for the structural media bias hypothesis.
Approximately a quarter of the articles featured rich and famous people. Furthermore,
out of the 40 articles in this category, 70 percent of them were about just four people: Bill Cosby,
Julian Assange, Donald Trump, and Bill O’Reilly. The articles about Bill Cosby dealt with the
civil and criminal cases he had pending against him. The stories about Julian Assange concerned
the rape allegations he was facing in Sweden, an investigation he claimed was being backed by
the United States as a way to extradite him for publishing classified documents. The articles
about Donald Trump were related to the 2016 presidential campaign, and the articles about Bill
O’Reilly were about the settlements for sexual harassment he and Fox News had paid several
women. Among the other rich and famous people discussed in the news were singer Kesha’s
71
dispute with her music producer, who she accused of physical, emotional, and sexual abuse
(most of the allegations were withdrawn or rejected by the court and the producer in turn sued
her for defamation and breach of contract), a lawsuit against singer Taylor Swift by a former
radio host (she said he groped her and he lost his job so he sued her for damages and she
countersued for assault and battery).
Regarding the familiarity news value, just over 60 percent of the articles were about five
people or organizations: Bill Cosby, Donald Trump, Julian Assange, Fox News, and the Catholic
Church. Other articles discussed SV at well-known universities (e.g., U.C. Berkeley, Harvard,
Columbia, Yale), sexual harassment at Uber which became public after a former engineer posted
about her experience at the company and it went viral, and a $1.65 million judgement Rolling
Stone magazine had to pay to a fraternity after a story about gang rape had to be retracted when a
report from the Columbia Journalism School revealed that they failed to take basic journalistic
steps to verify the woman’s account.
While there were only eight articles coded with the cover-up news value in the pre-
#MeToo time frame, as expected over half of articles involved stories about the Catholic Church.
Other articles included a lawsuit by a Nebraska state trooper alleging the agency covered up and
failed to investigate a male doctor assigned to perform pre-employment exams and another
article was a story about widespread abuse of underage students at elite private schools. The
government action category was intended to capture any actions by governments (e.g.,
congressional investigations, law enforcement, press conferences, legislation) since that is a beat
for reporters and the official sources generate a substantial amount of news. More than half of
this category was dominated by criminal cases (e.g., investigations, arrests, charges, sentencing).
Just under 20 percent of articles involved some other government action such as the mayor of
72
Seattle resigning after several men said he sexually abused them when they were underage, the
head of the Royal Canadian Mounted Police apologizing to women for years of harassment and
abuse as part of two class-action settlements, and Education Secretary announcing a change in
policy regarding campus sexual assault. Only five articles were about changes to the laws or
proposed legislation.
On balance, the results provide strong support for structural media bias in the pre-
#MeToo time frame as only one of the eight news values analyzed – immediacy – was
significantly different between the NYT and WSJ. This suggests that the norms and routines of
journalism had a stronger influence on news content than the ideology of the audience.
Furthermore, a substantial amount of coverage was about just a few well-known individuals and
organizations: Bill Cosby, Donald Trump, Julian Assange, Bill O’Reilly, Fox News, and the
Catholic Church. In the next section, I will review the results for the post-#MeToo time frame to
see if the #MeToo movement was able to influence news coverage of SV.
Post-#MeToo Analysis of News Values
Since liberals were more supportive of #MeToo than conservatives, if the NYT and WSJ
were catering their content to their audience the NYT would be more receptive to the messages
of the movement. The main comparison in this section will be the newspaper’s post-#MeToo
coverage compared to its own pre-#MeToo coverage. Most of the hypotheses in this section are
presented as competing hypotheses. If partisan media bias is controlling, then we would expect
to see changes post-#MeToo in the NYT but not the WSJ. If structural media bias is controlling,
there would either be no change or the change would be in the same direction (increase or
decrease) for both newspapers. The results strongly suggest that the NYT and the WSJ
73
responded to #MeToo in a similar way, thus providing support for the structural media
hypothesis (see Table 3 for a summary of the results).
Table 3. Summary of Results for News Values Post-#MeToo in NYT and WSJ
News Value NYT
Pre
NYT
Post
P-value
NYT Pre vs. Post
WSJ
Pre
WSJ
Post
P-value
WSJ Pre vs. Post
Legal 62% 39% p = 6.742e-05** 62% 35% p = 0.002073**
Immediacy 83% 88% p = 0.1933 96% 91% p = 0.3509
Conflict 10% 37% p = 4.529e-07** 11% 26% p = 0.03575*
Magnitude 73% 73% p = 0.9733 66% 67% p = 0.8746
Rich/Famous 27% 35% p = 0.1377 28% 44% p = 0.05347
Familiarity 50% 71% p = 0.0001831** 60% 76% p = 0.03757*
Cover-up 6% 6% p = 1 4% 4% n/a
Government
Action 47% 38% p = 0.129 47% 53% p = 0.4423
*p < .05, **p < .01
n/a = not enough observations for analysis
Notes: The pre-#MeToo time frame is October 2015 – September 2017 and the post=MeToo Time frame is October
2017 – September 2019. Sample: n = 101 articles NYT pre-#MeToo, 249 articles NYT post-#MeToo, 47 articles
WSJ pre-#MeToo, 116 articles WSJ post-#MeToo.
Legal and conflict news value. The first set of hypotheses predicted that with so many
victims coming forward, it would have been easier for journalists to generate stories from outside
the legal system (i.e., court cases). If this happened, then there would be a decrease in the legal
news value and an increase in the conflict news value. If partisan media bias is controlling, then
there would be changes in the NYT but not the WSJ (H2a) but if structural media bias is
controlling there would be no changes or changes in the same direction for both news values
(H2b). The results reveal a significant decrease in the legal news value for the NYT post-
74
#MeToo (39 percent) when compared to its pre-#MeToo coverage (62 percent), z = 3.985223, p
= 6.742e-05. There was also a significant decrease in the WSJ post-#MeToo (35 percent) when
compared to its pre-#MeToo coverage (62 percent), z = 3.079513, p = 0.002073. The results for
the conflict news values show a significant increase for the NYT post-#MeToo (37 percent)
when compared to its pre-#MeToo coverage (10 percent), z = 5.045295, p = p = 4.529e-07. In
the WSJ, the presence of this news value increased 15-percentage points (from 11 percent to 26
percent) post-#MeToo (p = 0.03575, Fisher’s exact test). The decrease in the legal news value
and increase in the conflict news value in both the NYT and WSJ confirm the structural media
bias hypothesis, H2b.
Although the legal news value decreased, as with the pre-#MeToo time frame, the vast
majority of articles involved criminal and civil cases and many of them involved well-known
individuals, organizations, or institutions. A little more than a third of the articles coded for the
legal news value was about just five people or organizations: Harvey Weinstein, Bill Cosby,
Jeffrey Epstein, Julian Assange, or the Catholic Church. While many of the criminal cases
included well known individuals or organizations (Michigan State, R. Kelly, Kevin Spacey),
several did not. In these cases, there was another element which seemed to prompt coverage
such as multiple offenders or victims, like articles about a serial killer or the 12 Israelis held in
Cyprus for the rape of a British woman. Other cases involved tragic facts such as the case of a
man charged with raping a teenager who killed her before she could testify against him at his
trial. Other articles covered people who were in positions of authority, such as the police officers
accused of SV, a worker at a migrant detention center convicted of sexually abusing underage
boys, and a humanitarian aid worker charged with child rape in Nepal. These findings are
consistent with the guidelines for journalists published by the Poynter Institute (2017) which
75
instructs journalists to publish articles on cases involving those who have access or influence
over children or other vulnerable people.
The increase in the conflict news value suggests that victims coming forward increased
this news value, and these results suggest that #MeToo played a role in encouraging victims to
report what happened to them. Many of the articles in this category seemed to be about the
aftermath of new revelations (i.e., resignations, honors previously awarded to offenders being
withdrawn). This could be because the initial revelation of wrongdoing is only news for a day
and the aftermath is an ongoing story which includes drama and conflict, something that is
appealing to journalists. For example, an article in the WSJ detailed the fallout for Harvey
Weinstein 10 days after the story first broke. In that time, the Academy of Motion Picture Arts
and Sciences – the group who hands out the Oscars – had voted to expel him, the British
Academy of Film and Television Arts had suspended him, the Producers Guild of America was
debating whether to dismiss him, new criminal investigations had opened in New York and
London, and French President Emmanuel Macron had taken steps to remove him from the
Légion d'Honneur.
Other stories were about the impact of #MeToo more broadly, such as workers at
McDonalds staging protests in several cities to get the company to address sexual harassment
and Time’s Up – an organization formed in the wake of #MeToo – throwing their support behind
#MuteRKelly, a grassroots social media campaign which called for companies to cut their ties
with musician R. Kelly who had been repeatedly accused of sexual assault and abuse. To a
lesser extent, the conflict category also included SV in conflict zones, mainly the civil war in
South Sudan and the genocide against the Rohingya in Myanmar.
76
Familiarity news value. The support of #MeToo among liberals would have signaled to
the NYT that their audience was interested in this subject and incentivized them to publish more
stories about it. Since there were so many stories which could have been told as a result of
victims coming forward and limited resources with which to tell those stories, the NYT may
have become more reliant than the WSJ on the familiarity news value if partisan media bias is
controlling (H2c). If structural media bias is controlling, there would be no change or change in
the same direction (H2d). The results show an increase in the presence of the familiarity news
value in both the NYT and WSJ. There was a 21-percentage point increase in the NYT post-
#MeToo (71 percent) when compared to pre-#MeToo (50 percent), z = 3.741256, p = 0.0001831.
Similarly, there was a 15-percentage point increase in the WSJ post-#MeToo (76 percent) when
compared to pre-#MeToo (60 percent), z = 2.079471, p = 0.03757. These results confirm H2d
and the change in the same direction for both newspapers provide evidence for structural media
bias.
The familiarity news value was the fourth most frequently used in the pre-#MeToo but
was second only to immediacy in the post-#MeToo time frame. The significant increase in both
the NYT and the WSJ suggests that given the volume of possible stories about SV after #MeToo
went viral both newspapers used familiarity to narrow down which stories they would pursue
with their limited resources. They also focused a significant amount of attention on a small
number of people/organizations. Nearly 40 percent of the articles coded with the familiarity
news value were about seven people or organizations: Harvey Weinstein, Donald Trump, Brett
Kavanaugh, Jeffrey Epstein, Steve Wynn, and Bill Cosby. There were several other people and
organizations in the entertainment industry including Kevin Spacey, Russell Simmons, Morgan
Freeman, Les Moonves of CBS, and Amazon Studios. Articles about SV in news and publishing
77
included Charlie Rose, Matt Lauer, Fox News, The New Republic and the Boston Globe. Wellknown universities included USC, Michigan State (because of Larry Nassar), Ohio State, and
Columbia. The well-known companies and organizations discussed include Nike, Under
Armour, Bank of America, Uber, the Boy Scouts, United Nations, and Oxfam. Government
entities were also considered familiar and included state legislatures, the military, and Congress.
Rich/famous news value. Since the tweet which led to the hashtag going viral was sent
by a Hollywood actor, and the story which prompted it was about a Hollywood movie mogul
who had sexually abused women for decades with impunity, this could have led to more people
who had experienced sexual harassment in the entertainment industry to come forward. This
may have resulted in more stories involving wealthy and powerful men in the industry, thereby
increasing the number of articles with the rich/famous news value. Since liberals were more
supportive of #MeToo than conservatives, if partisan media bias is controlling, there would be
more articles with this news value in the NYT than the WSJ (H2e). If structural media bias is
dominant, there would be either no change or an increase in both newspapers (H2f). There was
no change in the presence of rich/famous news value in either the NYT or the WSJ, thereby
confirming H2f. News reports on just six people – Harvey Weinstein, Donald Trump, Brett
Kavanaugh, Jeffrey Epstein, and Steve Wynn – accounted for nearly 50 percent of the articles
coded in this category. These results suggest that not only does the media focus on famous and
powerful people, but they also write a substantial number of stories about a small subset of them.
Immediacy news value. For the immediacy news value, H2g predicted that if partisan
media bias is controlling, the presence of this news value would increase post-#MeToo in the
NYT because the movement had more support among liberals and the onslaught of revelations
would have meant that more stories were being published as victims came forward. If structural
78
bias is controlling, H2h predicted an increase or no change. The results show that there was no
change post-#MeToo in either the NYT or the WSJ, thereby confirming H2h. The articles in his
category included the announcement of investigations (e.g., the police investigate a report of rape
against Russell Simmons, CBS hires a law firm to investigate reports about Les Moonves), the
fallout from new revelations (e.g., professor at Columbia accused of sexual harassment resigns,
Rep. Ruben Kihuen will not seek re-election, Steve Wynn resigns from his company),
congressional hearings (e.g., Brett Kavanaugh), as well as government reports and court cases
(e.g., arrests, hearings, sentencing, appeals). This was the most frequently used news value in
both the pre- and post-#MeToo time frame – it was coded in 87 percent of articles in the NYT
and 92 percent of articles in the WSJ – which suggests there was a ceiling effect for this news
value. It also reflects Patterson’s (2013) observation that journalism is a “shortsighted
discipline”, and journalists are “trained to see the world changing in important ways from one
day to the next” (82). The hyper focus on what is going on right now means that journalists,
regardless of the partisan leanings of their newspapers, place an outsized value on immediacy
when selecting stories, leaving little opportunity for follow-ups which provide more context.
Magnitude news value. Since the magnitude news value was expected to be higher in
the NYT in the pre-#MeToo time period because liberals emphasize harm and fairness, the
higher degree of support among liberals for #MeToo was not expected to significantly influence
this news value in the post-#MeToo time frame, and no change was expected in the WSJ (H2i).
There was no change in either newspaper post-#MeToo, thus confirming H2i and providing
support for structural media bias. As with the pre-#MeToo time frame, a substantial volume of
the coverage focused on a small number of individuals and groups: nearly 40 percent of the
coverage was about the Catholic Church, Harvey Weinstein, Bill Cosby, Donald Trump, Brett
79
Kavanaugh, Jeffrey Epstein, Steve Wynn, and Fox News. Furthermore, nearly all of the articles
in this category included multiple victims (93 percent). Other articles in this category included
Dr. Larry Nassar who sexually abused hundreds of girls and young women at Michigan State,
Dr. George Tyndall at USC, SV in the modeling industry, the Boy Scouts, sexual assault in the
military, sexual misconduct by employees Oxfam, and SV in conflict zones such as South Sudan
and Myanmar. This finding is consistent with the guidelines for journalists published by the
Poynter Institute (2017) which encouraged reporting stories which involved multiple victims.
Government action news value. Given the large amount of coverage generated after
#MeToo went viral, it is reasonable to expect a government response which would increase the
government action news value. However, since both liberals and conservatives are attentive to
government action, albeit for different reasons, there was expected to be an increase in both
newspapers (H2j). There were no significant changes post-#MeToo in either the NYT or the
WSJ, so H2j is not confirmed. This provides further support for the structural media bias
hypothesis.
However, I noticed that when comparing the pre- and post-#MeToo proportions in each
newspaper there was an insignificant decrease in the NYT and an insignificant increase in the
WSJ, so I ran a z-test comparing the NYT post-#MeToo to the WSJ post-#MeToo. The results
show that this news value was used significantly more in the WSJ (53 percent) than the NYT (38
percent), z = 2.822694, p = 0.004762. To investigate this difference further, I first organized all
the articles into three subcategories: criminal cases, other government, and legislation. As Table
4 shows on the next page, there was a similar decrease in the proportion of articles in both the
criminal and legislation subcategories, but there was a much larger increase in the other
government subcategory in the WSJ so I focused my attention there. For each of the 34 WSJ
80
articles in the other government subcategory, I searched for a similar story in the NYT on or
around the same date to identify which stories were reported in the WSJ but not the NYT. I
found 13 stories which were published by the WSJ but not the NYT and five (38 percent) were
related to stories that the WSJ was involved in breaking (Steve Wynn of Wynn Resorts, the
hiring of a campaign aide accused of sexual assault by the Murphy Administration in New
Jersey, and a doctor at the Indian Health Services who had been abusing children for decades), so
it appears to be an organizational decision to publish more articles about these particular
individuals and organizations. The other stories did not seem to have a theme tying them
together (e.g., the Chinese government clamping down on a rideshare company, a sexual
harassment probe in Japan which was weakening the prime minister, a statement by the
Secretary of State on Myanmar, and Trump defending a speech mocking Christin Blasey Ford in
an interview on "60 Minutes").
Table 4. Subcategories for the Government Action News Value in NYT and WSJ
Gov’t Action
Subcategories
NYT
Pre
NYT
Post
Difference
NYT Pre vs. Post
WSJ
Pre
WSJ
Post
Difference
WSJ Pre vs. Post
Criminal 31% 18% -13% 30% 16% -14%
Other Gov’t 10% 14% +4% 6% 29% -23%
Legislation 2% 1% -1% 6% 4% -2%
Notes: The pre-#MeToo time frame is October 2015 – September 2017 and the post=MeToo Time frame is October
2017 – September 2019. Sample: n = 101 articles NYT pre-#MeToo, 249 articles NYT post-#MeToo, 47 articles
WSJ pre-#MeToo, 116 articles WSJ post-#MeToo.
Cover-up news value. For the final news value, H2k predicted that there would not be
any differences post-#MeToo in the cover-up news value since it was expected to be highly
associated with the Catholic Church. The results showed no differences in either the NYT or the
81
WSJ, thus confirming H2k which supports the structural media bias hypothesis. As expected, 70
percent of articles coded in this category were about the Catholic Church. Other examples
include the cover-up of sexual abuse by Southern Baptist Convention, how Bill Shine helped
Roger Ailes cover up sexual harassment at Fox News, and the failure of USC to make their
report about gynecologist Dr. George Tyndall public and report the findings of their investigation
to the California Medical Board.
The overall results in the post-#MeToo time frame suggest that the NYT and WSJ
responded to #MeToo in very similar ways, supporting the structural media hypothesis that
journalistic norms and routines remain strong in a highly polarized political environment. In
both the NYT and the WSJ, the decrease in the use of the legal news value and increase in the
conflict news value suggests that the source of the news stories changed. Whereas in the pre-
#MeToo time frame a victim needed to either be in a position to report the incident to the police
or possess the knowledge, resources, and strong facts to engage the services of a lawyer to file a
lawsuit, it appears #MeToo encouraged many victims to come forward and report incidents to
their current or former employer as many articles were about the aftermath of reports being filed
(e.g., investigations, suspensions, resignations, terminations). However, the stories which were
told were largely about people or organizations with which people were familiar, such as people
associated with the entertainment industry (e.g., Harvey Weinstein, Bill Cosby, Kevin Spacey, R.
Kelly, Les Moonves of CBS), well-known individuals and companies in the news media (e.g.,
Matt Lauer, Charlie Rose, Fox News), politicians and people associated with other branches of
government (e.g., Donald Trump, Roy Moore, Al Franken, Brett Kavanaugh), well-known
corporations (e.g., Wynn Resorts, Nike, Uber), and the Catholic Church. These results suggest
82
that the focus of news is heavily skewed towards the actions of officials and other elites which is
consistent with previous research (Entman 1989; Schudson 2002; Graber 2003; Bennett 2016).
In the next section, I will discuss the results of the coverage of high-risk groups. It was
important to the #MeToo movement that people from these groups have their stories highlighted,
particularly so that any potential responses could address the unique challenges they face.
High-Risk Groups Post-#MeToo
One of the goals of the #MeToo movement was to highlight the experiences of groups
who are at high risk – racialized minority groups, LGBTQ+, immigrants, those who are poor,
and people with medical conditions – and make their stories newsworthy. In order to assess the
influence – or lack thereof – of #MeToo, we need to first establish whether there were any
differences pre-#MeToo.
Since liberals are more concerned with justice and fairness and conservatives are more
likely to express hostility towards outgroups, it was expected that there would be more coverage
of these groups in the NYT if partisan media bias is controlling (H3a); if structural media bias is
controlling there would be no differences (H3b). There were no differences between the NYT
and WSJ, although the results for race (non-white victims) was almost significant at p = 0.05584
(Fisher’s exact test). These results confirm H3b and the presence of structural rather than
partisan media bias.
Since #MeToo prioritizes these groups and liberals support the movement, if partisan
media bias is controlling there would be more coverage of them in the NYT than the WSJ (H3c).
If structural media bias is controlling and #MeToo was able to influence coverage of these
groups, there would be an increase in both the NYT and WSJ (H3d). However, if structural
media bias is controlling and #MeToo had no impact, then coverage would not change in either
83
newspaper (H3e). The overall number of articles in these categories is quite small (see Table 5)
and the results reveal that when comparing a newspaper’s pre-#MeToo coverage to its own post-
#MeToo coverage, there are no changes in coverage of victims from any of these high-risk
groups, thus confirming the structural media bias hypothesis H3e.
Table 5. Summary of Results for High-Risk Groups Post-#MeToo in NYT and WSJ
Group
NYT
Pre
Freq.
NYT
Post
Freq.
P-value
NYT Pre vs.
Post
WSJ
Pre
Freq.
WSJ
Post
Freq.
P-value
WSJ Pre vs.
Post
Racialized
Minority 8 23 p = 0.8364 0 2 n/a
LGBTQ+ 4 4 n/a 1 1 n/a
Immigrants 4 9 p = 0.7646 0 1 n/a
Poor 1 6 p = 0.6778 1 0 n/a
Medical Condition 1 1 n/a 0 0 n/a
*p < .05, **p < .01
n/a = not enough observations for analysis
Notes: The pre-#MeToo time frame is October 2015 – September 2017 and the post=MeToo Time frame is October
2017 – September 2019. Sample: n = 101 articles NYT pre-#MeToo, 249 articles NYT post-#MeToo, 47 articles
WSJ pre-#MeToo, 116 articles WSJ post-#MeToo.
In looking at all the articles in the dataset which discussed high-risk groups,
approximately half of the articles, which only accounts for 20 articles, discussed why these
factors put them at higher risk of SV.
Racialized minority groups. One of the stories in the NYT provided more detail than
most about how the race of the victim was factor in their victimization. The article was about a
police officer in Oklahoma convicted of raping several black women while he was on duty and
they cited a report from the African American Policy Forum which found that black women were
84
“disproportionally sexually assaulted because they were often assumed to be promiscuous and to
be less likely to report the crime” (Fenwick and Schwarz 2015).
Other articles in the NYT discussed how black women felt a responsibility to protect
Black men from negative stereotypes (Dickerson and Saul 2017; Jacobs and Robertson 2019).
One Black female college student who reported sexual harassment by a “distinguished professor”
said that she “had to choose being a woman over being black” (Jacobs and Robertson 2019). In
foreign news articles, SV was discussed in terms of genocide and conflicts between ethnic or
religious majorities and minorities (e.g., Rohingya in Myanmar, civil war in South Sudan).
LGBTQ+. There were very few articles about LGBTQ+ victims and only one article
discussed why identifying as LGBTQ+ could put them at higher risk. In a story about the killing
of four gay men in London, the sister of one of the victims believed that her brother’s sexuality
played a role in the police overlooking important details which led them to initially label his
death as a suicide rather than a murder (Bennhold 2016). In a WSJ article about the results of a
survey of students in the University of Texas system, the very last line of the article reads: “The
most vulnerable group was LGBTQ+, of whom 43% of undergraduate and 41% of graduate
students reported being harassed” (Belkin 2017). There was no further discussion as to why
LGBTQ+ students were experiencing harassment at much higher rates than other students.
Poverty. While poverty was discussed on both a societal and individual level, few
discussed how poverty affected their level of risk. The articles which discussed poverty on a
societal level were in other countries and discussed how the governments in those countries were
highly reliant on humanitarian aid. Since many of these aid organizations operated with limited
government oversight, one expert said they could be used as cover for human trafficking
(Schultz and Bhandari 2018). In an article about the misconduct of Oxfam workers in Haiti, the
85
NYT quoted Mark Schuller, an associate professor of anthropology who wrote a book on the
negative consequences of humanitarian aid in the country after the 2010 earthquake. Professor
Schuller said that the “reward structures of the aid industry reinforce inequality, imperialism and
patriarchy” but unfortunately the article did not elaborate further (Porter 2018).
There were few stories on individual level poverty. One story in the NYT involved
victims who were part of a welfare-to-work program for low-income New Yorkers (Neuman
2019). One victim said that she was fearful of losing her job and she was a single mother who
needed to provide for her son, so she complied with the supervisor’s demands for sex acts.
These cases took place within the Department of Parks and Recreation, presumably a job which
involves physical isolation which puts them at higher risk, yet the article did not address this
structural risk factor or press New York City officials about how they could better protect their
employees.
Immigrants. In cases of undocumented immigrants, fear of deportation was discussed as
a barrier to reporting, particularly after Trump was inaugurated in 2017 (Medina 2017).
However, deportation was just one threat used to silence victims: one smuggler threatened to sell
her 3-year-old daughter if she reported him (Fernandez 2019). One article discussed how
incarceration and legal status made a woman more vulnerable. In the case, two supervisors and a
guard were charged with raping nine women at federal prison in Brooklyn and prosecutors noted
that one of the defendants appeared to have chosen his victim because he knew she was going to
deported at the conclusion of her sentence in five months, she spoke minimal English and she
had virtually no visitors (Goldstein 2017).
In cases of documented immigrants at universities, none of the articles discussed how
international students could be at higher risk of losing their student visa if they took time off
86
from school. However, only one article discussed how having a work visa tied to the employer
is problematic. The case involved a model from Europe who was assaulted by the executive
director of a modeling agency who had sponsored her visa. When she left the agency because of
the assault, she also lost her visa and had to leave the country (Friedman 2018).
Medical conditions. There were only two articles which discussed victims with
disabilities, both in the NYT. In one article about a man accused of sexually abusing his adopted
and foster children, many of whom were special needs, the police admitted that the allegations
made by two of the boys were believed because, unlike other boys who had reported abuse in the
past, they did not have severe disabilities or emotional problems (Stewart and Goldstein 2016).
The other article was about victims who attended a Catholic school for the deaf in Verona, Italy.
One victim, who was sexually abused by a priest at the school in the 1950s, said that he
remembered calling for help and then realizing “We were deaf, no one could hear us” (Povoledo
2019).
These results provide further evidence of structural media bias which favors the activities
and interests of those who are in positions of power and influence.
Conclusion
Overall, the analysis of newsworthiness shows strong support for structural media bias
and little evidence of partisan media bias despite the current polarized political environment.
The NYT and WSJ used similar criteria to choose which stories to publish about SV and a
substantial amount of coverage involved well-known individuals or organizations such as
Harvey Weinstein, Bill Cosby, Donald Trump, Brett Kavanaugh, Jeffrey Epstein, Fox News, and
the Catholic Church. With the possible exception of Jeffrey Epstein, journalists already had a
pre-existing relationship with these entities (e.g., corporate public relations departments,
87
spokespersons, press secretaries, members of the U.S. Senate) – in essence, they were part of a
reporter’s “beat” in business, entertainment, or politics. These findings are consistent with
previous research that news coverage looks similar and focuses on the activities of elites
(Bennett 2016; Cook 1998; Graber 2003; Bagdikian 2004).
The results also suggest that the #MeToo movement encouraged more victims to come
forward, thus changing the impetus for a news story from the court system (as seen in the
decrease of legal news value post-#MeToo) to victims revealing their stories on social media, to
journalists directly, or announcements from organizations about high-level employees (e.g.,
investigations, suspensions, resignations, terminations). The increase in the familiarity news
value post-#MeToo suggests that both the NYT and WSJ used familiarity to decide which stories
to pursue in the wake of so many victims coming forward.
Despite the effort of #MeToo to make high-risk groups and their experiences
newsworthy, there was not a significant increase in the coverage of these groups. Overall, only
10 percent of articles in the post-#MeToo sample (37 articles) featured victims from these
groups, which was similar to the proportion of articles in this category pre-#MeToo (11 percent).
These findings are consistent with previous research which has found that non-white victims of
crime are less newsworthy than white victims (Weiss and Chermak 1998; Gilchrist 2010; White,
Stuart, and Morrissey 2021). This makes victims from high-risk groups invisible which could
have implications for how law enforcement and counselors interacts with victims from these
groups (e.g., finding them less credible), which causes further harm to survivors and prevents
them not only from receiving justice, it also adversely impacts their ability to heal from their
trauma, something that could negatively affect their interpersonal relationships with family and
friends as well as their employment.
88
These findings are problematic in that the coverage does not illuminate the racial and
structural conditions which make certain groups more vulnerable. For example, without
discussing how racial harassment and sexual harassment are often intertwined, policies of
prevention on an organizational and societal-level are unlikely to be addressed (Johnson, Ijoma,
and Kim 2022). Without public pressure to highlight the unique concerns of these groups, it is
unlikely that governmental resources will be properly allocated to address them.
89
Chapter 4
Framing Analysis
This chapter will examine how the stories of SV were told in NYT and WSJ. If partisan
media bias is controlling then there will be differences between the two newspapers but if
structural bias is controlling, then coverage will be similar. As with the previous chapter, the
first set of tests will examine coverage in the pre-#MeToo time frame between the NYT and
WSJ. The second test of media bias in the post-#MeToo time frame will compare each
newspaper’s coverage to its own pre-#MeToo coverage. If partisan media bias is controlling and
#MeToo influenced coverage, then there will be changes in one newspaper but not the other but
if structural bias is controlling there will be change in the same direction or no change at all.
Similar to the analysis of newsworthiness, the unit of analysis was the article and the
variables were dichotomous (yes/no). Statistical analysis was performed on each variable in
which there were five or more observations. A proportion Z-test was used to test variables in
which there were more than 10 observations in every cell ("Two Proportion Z-Test" n.d.) and a
Fisher’s Exact test was used for variables which had fewer than 10 observations in every cell
("Fisher’s Exact Test" n.d.).
Pre-#MeToo Analysis of Primary Frames
As with the chapter on newsworthiness, most of the hypotheses in this section are
presented as competing hypotheses: one which predicts partisan media bias and the other which
predicts structural media bias. Overall, the results show that there were no differences between
the NYT and WSJ for four out of the five primary frames analyzed, thus providing support for
the structural media bias hypotheses. The only significant difference between the two
90
newspapers was for the business frame, which only constituted 11 articles in the pre-#MeToo
time frame (see Table 6 for a summary).
Legal frame. The first set of hypotheses were about the legal frame. Since
conservatives are interested in preserving the status quo and have more of a need for order and
structure, if partisan media bias is controlling we would expect that the legal frame to be used
more in the WSJ (H4a). However, if structural media bias is controlling, there would be no
differences between the two newspapers (H4b). The results show no differences between the
NYT and the WSJ in the pre-#MeToo time frame, confirming H4b.
Table 6. Summary of Results for Primary Frames Pre-#MeToo in NYT and WSJ
Frame NYT WSJ P-value
Legal 71% 68% p = 0.6915
Conflict 9% 6% p = 0.7529
Victim 6% 0% p = 0.1772
Business 3% 17% p = 0.004683**
Political 10% 9% p = 1
Harm to victims
(Subcategory: victim frame) 34% 13% p = 0.009259**
*p < .05; **p < .01
Notes: October 2015 – September 2017; n = 101 articles in the NYT and 47 articles in the WSJ
The legal frame discussed SV in the context of the law and included criminal and civil
cases, legislation, government investigations and hearings, as well as internal investigations
which were often conducted by law firms. The legal frame was the primary frame in 70 percent
of articles in the pre-#MeToo time period, and the majority of these articles concerned criminal
or civil cases (75 percent). The remainder of articles in this category were about internal
91
investigations and legislation (there were no congressional investigations or hearings).
Furthermore, a third of the articles using this frame were about just four people and
organizations: Bill Cosby, Julian Assange, Fox News, and the Catholic Church.
Articles which covered SV in the context of civil and criminal cases present the issue as a
dispute to be resolved between two parties and include arguments over the facts of the case (who
is telling the truth?), the validity of the complaint (as can be seen in motions to dismiss),
admissibility of evidence, and standards of proof. For example, in an article about the criminal
trial for Bill Cosby, the charges against him are discussed as well as the motion practice (e.g.,
filing motions to dismiss or exclude evidence) by both the prosecution and defense attorneys.
The prosecution filed a motion to allow the testimony of 13 other women who said were drugged
and sexually assaulted by Cosby as evidence of “prior bad acts” and Cosby’s defense attorneys
sought to suppress testimony he provided in civil proceedings and the recording of a phone call
between him and the victim’s mother (Hurdle and Bowley 2016).
Conflict frame. The conflict frame focuses on scandal or controversy (e.g., a victim,
most often a woman, coming forward and revealing what happened to her; the reaction or
aftermath of revelations) and acts of SV which occurred in conflict zones. If partisan media bias
is controlling, it was expected that the conflict frame would be used more in the WSJ since
conflict can signal a threat to the status quo which conservatives are interested in maintaining
(H4c). If structural bias is dominant, there would be no differences between the NYT and WSJ
in the pre-#MeToo time frame (H4d). The results show that there were no differences between
the two newspapers, thus confirming H4d.
There were only 12 articles (8 percent) which used the conflict frame in the pre-#MeToo
time period (this frame would have been higher if court cases were included as they are in other
92
studies). This frame included articles about a social club at Harvard refusing to admit women,
universities facing pressure to resist the Trump administration’s efforts to relax policies put in
place by the Obama administration to address campus sexual assault, and civil war in South
Sudan. There was also a story about the controversial decision by Google to fire a software
engineer for violating company policy after he wrote a memo criticizing the company’s diversity
efforts (in the memo he claimed biological differences explain some of the gap between male
and female tech workers). These stories also present SV as a dispute between parties (social club
vs. Harvard, universities vs. Trump administration, warring parties in South Sudan, Google vs.
fired software engineer), and suggest that there could be winners and losers.
Victim frame. Articles which used the victim frame highlight the experience of victims
and the impact SV has had on their lives. It was predicted that if partisan media bias was
controlling, the NYT would use the victim frame more than the WSJ since liberals emphasize
harm, fairness, and justice (H4e); if structural media bias was controlling, there would be no
differences between the two newspapers in the pre-#MeToo time frame (H4f). There were no
significant differences between the NYT and the WSJ in the use of the victim frame, confirming
the structural media bias hypothesis, H4f.
However, the number of articles in this category are small making it difficult to observe
statistically significant differences. The victim frame was only used in six articles in the pre-
#MeToo time frame and all of them came from the NYT. In a story about a teenage survivor of
sexual assault by another student at an elite prep school in New Hampshire, she said “I want
everyone to know that I am not afraid or ashamed anymore, and I never should have been”
(Hauser 2016). She also said how upset and disgusted she was that the jury did not believe he
did it knowingly, and although she was determined to return to school, some male friends
93
refused to speak to her, boys made jokes about consent and treated the issue of SV as a joke, and
she eventually had to leave the school. In a story about repeat offenders on college campuses,
the first paragraph reads:
She was 18, majoring in pre-medicine and settling in for her first year at Kansas State
University, her dream school. Barely six weeks later, Crystal Stroup’s college career was
suddenly and violently derailed. (Saul 2017)
The article goes on to describe how she was living in off-campus housing and they had a
small get together at their apartment where she had too much to drink. Her friends asked a
fellow student who lived downstairs, Jared Gihring, to look after her while they went out for
food. She woke up the next morning “disoriented and in pain” with large bruises in the shape of
hand prints emerging on her upper arm and thigh. It turns out that two years earlier, another
student had reported that Gihring raped her at a fraternity house but Kansas State took the
position that it was not responsible for investigating reports of rape at a fraternity house because
they were off campus and no action was taken. The article ends by discussing the immense
stress she has been under, partly because it was impossible to avoid Gihring on campus or at her
apartment complex. The stress negatively affected her grades, and she was forced to drop out of
school. The last sentence of the article reads:
She now works in a nursing home kitchen, not exactly the health care career she had
hoped for. “I don’t know how to pick up the pieces and start over again,” she said.
Articles such as these provided information which help readers understand the long-term impact
SV has on survivors and the challenges they face in their recovery.
Harm to victims is a subcategory of the victim frame and was coded in this study because
experts have criticized news coverage of SV for not discussing it. Since discussing harm to
victims would appeal to liberal readers because of their interest in social change and equality, as
well as their emphasis on harm and fairness, if partisan media bias was controlling then the NYT
94
would discuss harm to victims more than the WSJ (H4g). If structural media bias is controlling,
then there would not be any differences between the two newspapers (H4h). The results show
that the NYT discussed harm to victims significantly more than the WSJ (p = 0.009259, Fisher’s
exact test), thus confirming the partisan media bias hypothesis H4g.
In addition to the NYT discussing harm to victims more frequently than the WSJ in the
pre-#MeToo time period, the NYT also discussed more types of harm to victims. For example,
the WSJ did not discuss the fear of retaliation or physical harm in any of their news stories. The
NYT also discussed psychological harm in a broader context than the WSJ. While the WSJ only
discussed the suicide of two men abused by Catholic priests in one article, the NYT documented
the suffering of victims more broadly, including depression, anxiety, post-traumatic stress
disorder, and difficulty with trust and intimacy in relationships. However, the discussion of
harm to victims was usually brief and vague and was not like the harm described above in the
article on campus sexual assault. For example, in an article which discusses how Yazidi women
and girls, some as young as 8, were kidnapped by ISIS and made to join “a systematic network
of sex slaves,” aid workers said “that depression and suicide attempts are increasing” (Chan and
Sengupta 2016).
Both newspapers discussed retaliation which included defamation, demotions,
terminations, contracts not being renewed, and job offers being rescinded. Gretchen Carlson, a
former Fox News anchor, stated in her lawsuit that Roger Ailes fired her after she rebuffed his
advances. In another case at Fox News, former Fox News host Andrea Tantaros stated that she
was demoted after she rebuffed advances from Ailes. A former contestant on The Apprentice
came forward during the 2016 presidential campaign to say Donald Trump groped and kissed her
without her consent and she filed a lawsuit against him for defamation after he said she was lying
95
and was just seeking attention at his expense. In a case at Yale, a postgraduate fellowship was
revoked after the student reported the professor who ran the program to the university for sexual
harassment.
Business frame. The business frame discussed SV in terms of impact on the business:
stock prices, revenue, value of the company, successor or interim replacement being named,
implications of CEO resigning or being terminated. It was expected that if partisan media bias is
controlling the business frame would be used more by the WSJ since they report news through a
business lens and conservatives are sensitive to loss and emphasize financial security (H4i). If
structural media bias was controlling, there would not be any differences (H4j). The results
confirm H4i: the WSJ used the business frame significantly more than the NYT in the pre-
#MeToo time frame (p = 0.004683, Fisher’s exact test).
Even though there were partisan differences found with this frame, the overall number of
articles using the business frame is small – three articles in the NYT and eight in the WSJ – and
accounts for only seven percent of articles in the pre-#MeToo time period. All three articles in
the NYT were about Bill O’Reilly, a case which was made public by a NYT investigation (Steel
and Schmidt 2017b). Two of the stories were about sponsors pulling their advertising (there was
a public pressure campaign on companies to pull their advertising from his show after the story
came out in the NYT). The other story was about whether Fox News would terminate O’Reilly,
who was described as “ratings draw and a revenue generator,” and how the problems at Fox
News could jeopardize the parent company’s attempt to acquire the British satellite company
Sky because they needed to convince British regulators that they were “fit to acquire the 61
percent in Sky that they do not own” (Steel, Schmidt, and Rutenberg 2017).
96
The WSJ published three articles about sexual harassment (and racism) in the advertising
industry which is an issue that had been in the news because of several high-profile incidents.
One article discussed pressure from powerful corporate clients to diversify their creative team
and detailed the changes at several advertising agencies, including appointing women into top
creative positions (Bruell 2016). Other articles discussed the resignations of top-level executives
as a result of sexual harassment allegations, and one article discussed the well-known corporate
brands pulling their advertising from The O’Reilly Factor (Bruell 2017). Another article was
about the quarterly earnings report from 21st Century Fox which detailed $10 million in claims it
had paid out related to Fox News Chief Executive Roger Ailes (Flint 2017). The stories which
used this frame involved the aftermath of SV becoming public (e.g., the impact on the
organization, dismissals, and replacement of key organizational leaders).
The articles which use the business frame focus more on the impact the allegations of SV
have on the company and what was lost or could be lost in both in terms of money and
employees who have been integral to a company’s success (e.g., Roger Ailes and Bill O’Reilly).
Cuklanz (2020) has criticized the use of these frames in discussion of SV because in focusing on
what the company and offenders lost, readers are less likely to be sympathetic with the
experiences of victims.
Political frame. The political frame presents SV as a political problem. It includes the
implications of the allegations, opinion polling, political strategy for dealing with allegations,
offenders evading political consequences, and politicians resigning or not seeking re-election.
Since the “game frame” is so ubiquitous in news coverage, there were no differences expected
for the political frame based on the ideology of the audience (H4k). The results confirm this
97
hypothesis: there were no differences between the NYT and the WSJ in the use of the political
frame in the pre-#MeToo time frame.
There were only 14 articles coded with the political frame and all but one of the domestic
articles was about Donald Trump; the other was about the resignation of the mayor of Seattle
after several men came forward and said he sexually abused them decades ago when they were
underage. The articles about Donald Trump concerned the fallout from the release of the Access
Hollywood tape, including Republicans who were distancing themselves from him and the
recommended strategy from Chris Christie and Reince Preibus that Trump attack Hillary Clinton
for defending her husband’s behavior with women to rally Republicans, which he did by inviting
three of the women who said Bill Clinton sexually assaulted or raped them to the second
presidential debate (Bender and Hook 2016a; Haberman 2016a, b). Another article further
described Trump’s strategy of denying all allegations against him, calling the women liars and
claiming that they were all part of a vast conspiracy by the news media and Hillary Clinton’s
campaign to derail his candidacy (Burns and Corasaniti 2016). Other articles discussed how the
allegations against Trump affected polling in the race and how his path to victory was
diminishing (Hook 2016; Bender and Hook 2016b).
These findings are consistent with the coverage of politics as a game, especially in an
election context (Bennett 2016; Cappella and Jamieson 1997; Patterson 1993, 2016) and is
similar to the legal frame which presents the issue as one with winners and losers. This frame
also reflects one of the elements of Schneider and Hannem’s (2019) concept of the politicization
of sexual misconduct in which SV is framed as a political issue and statements by victims are
dismissed as politically motivated lies and partisan attacks by sources reporters use to tell the
story.
98
The balance of the evidence in the pre-#MeToo time frame provides support for
structural bias. The NYT and WSJ covered SV in a similar way in 92 percent of the articles in
the pre-#MeToo time frame; the only significant difference was in the use of the business frame
which only accounted for 11 articles. Most of the coverage presented SV in the context of a
dispute between parties with winners and losers, as opposed to a pervasive social problem in
which gender inequality was playing a major role. In the next section, I will review the results
for the post-#MeToo time frame to see if the #MeToo movement was able to influence how the
story of SV was framed.
Post-#MeToo Analysis of Primary Frames
Since liberals were more supportive of the #MeToo movement than conservatives, if the
NYT was catering their content to their audience the NYT would be expected to be more
receptive to the messages of the movement. As with the section on news values, the main
comparison in this section will be the framing of the newspaper’s post-#MeToo coverage
compared to its own pre-#MeToo coverage. Most of the hypotheses in this section are presented
as competing hypotheses. If partisan media bias is controlling, then we would expect to see
changes post-#MeToo in the NYT but not the WSJ. If structural media bias is controlling, there
would either no change or the change would be in the same direction (increase or decrease) for
both newspapers. The results suggest that structural media bias was still controlling in the post-
#MeToo time frame, and the movement had little effect on how the NYT and WSJ told the
stories of SV. The results are summarized in Table 7.
99
Table 7. Summary of Results for Primary Frames Post-#MeToo in NYT and WSJ
Frame NYT
Pre
NYT
Post
P-value
NYT Pre vs. Post
WSJ
Pre
WSJ
Post
P-value
WSJ Pre vs. Post
Legal 71% 52% p = 0.0008391** 68% 57% p = 0.1863
Conflict 9% 29% p = 3.841e-05** 6% 21% p = 0.03459*
Victim 6% 5% p = 0.7971 0% 2% n/a
Business 3% 6% p = 0.296 17% 10% p = 0.2922
Political 10% 8% p = 0.4851 9% 9% p = 1
Harm to victims
(Subcategory:
victim frame)
34% 34% p = 0.9898 13% 16% p = 0.6386
*p < .05, **p < .01
n/a = not enough observations for analysis
Notes: The pre-#MeToo time frame is October 2015 – September 2017 and the post=MeToo Time frame is October
2017 – September 2019. Sample: n = 101 articles NYT pre-#MeToo, 249 articles NYT post-#MeToo, 47 articles
WSJ pre-#MeToo, 116 articles WSJ post-#MeToo.
Legal, conflict and victim frames. As #MeToo encouraged more victims to come
forward, this could have changed motivation of the news story from the court system to victims
themselves or press releases from organizations announcing investigations, suspensions,
resignations, or terminations. It was expected that if partisan media bias is controlling, this
would result in a decrease of the legal frame, and an increase in victim and conflict frames in the
NYT post-#MeToo but not the WSJ (H5a). If structural media bias is controlling, there would
either be change in the same direction (H5b) or no change (H5c). The results partially confirm
each of these hypotheses. The NYT decreased the use of the legal frame post-#MeToo (52
percent) when compared to their pre-#MeToo coverage (71 percent), z = 3.339611, p =
0.0008391, but there was no change in the WSJ (partial support for H5a). There was no change
in the use of the victim frame post-#MeToo in either the NYT or the WSJ (partial support for
100
H5c). The conflict frame increased in both newspapers post-#MeToo. There was a 20-
percentage point increase in the conflict frame in the NYT when compared to their pre-#MeToo
coverage (p = 3.841e-05, Fisher’s exact test). In the WSJ, there was a 15-percentage point
increase when compared to their pre-#MeToo coverage (p = 0.03459, Fisher’s exact test). This
finding partially supports H5b. The decrease in the legal frame in the NYT but not the WSJ
appears to be the result of the NYT increasing the use of the conflict frame five percentage
points more than the WSJ. On balance, these results provide more evidence for structural bias.
The majority of articles in the legal frame category (61 percent) were civil and criminal
cases. The articles which discussed civil cases were mostly about the initial filing of the lawsuit
and details of the complaint. The criminal cases discussed various stages of the case, from
investigation to charges, trial, verdict, and appeal. The legal frame can also present the issue of
SV in terms of winners and losers, similar to the way journalists cover politics. The NYT
published an article called “Is the Case Against Harvey Weinstein in Jeopardy?” in which they
report that the case “appears to be fraying” because a detective failed to turn over important
evidence to prosecutors, the judge dismissed part of the indictment and evidence emerged which
undermines the credibility of one of the victims (Ransom 2018). In addition, the District
Attorney’s office thought that once they filed charges several more women would come forward
but that had not happened. The article also cites former prosecutors who describe having this
information coming out after the indictment as “a horror” which hurt the investigation and could
discourage other witnesses from coming forward. In this way, the prosecution is portrayed as the
“loser” and Weinstein as the “winner.”
The legal frame also included government hearings and investigations, and the most
prominent one during the post-#MeToo period was the Supreme Court confirmation hearings of
101
Brett Kavanaugh, which was another opportunity to portray the issue of SV as a dispute between
parties, in this case Republicans and Democrats. After Dr. Ford testified, one article detailed
what would happen next procedurally, which senators were undecided, and what would happen
next if Kavanaugh was not confirmed.
Journalists still relied heavily on conflict and drama to tell the story, which is a wellknown formula for writing stories (Bennett 2016). Several of the articles in the conflict frame
category focused on the loss to the offender. For example, two universities announced they had
decided to strip Charlie Rose of journalism honors after decades of sexual harassment, including
groping women, exposing himself to them and making inappropriate phone calls. Hollywood
director Ridley Scott announced he was going to replace actor Kevin Spacey in a film that had
been completed and was weeks away from being released.
Other articles focus on controversy, such as an article about the Catholic Church in which
people in Chile were upset and angry at Pope Francis who, when speaking to reporters, defended
a bishop they said protected a pedophile priest. Similar to the legal frame, these stories frame
SV as a conflict between individuals or groups to be resolved in the private sphere, instead of as
a symptom of gender inequality.
Since one of the goals of the #MeToo movement was to support victims, discussing the
harm to victims SV has on their lives is important. If partisan media bias is controlling, we
would expect to see an increase in the discussion of harm to victims in the NYT but not the WSJ
(H5d). If structural media bias is controlling, then there would be an increase in both
newspapers if #MeToo influenced coverage (H5e) or no change in either newspaper if #MeToo
did not influence coverage (H5f). The results show no change post-#MeToo in the NYT or the
WSJ which confirms H5f and a lack of partisan bias.
102
While #MeToo did not influence the proportion of articles which discussed harm to
victims in either newspaper, the WSJ did discuss more types of harm to victims than it did in the
pre-#MeToo time frame including fear of retaliation, having to quit their job, seeking a new
assignment, and both physical and psychological harm. In both the NYT and the WSJ, the most
common type of harm to victims was psychological harm, retaliation, and fear of retaliation.
Psychological harm included depression, anxiety, PTSD, eating disorders, humiliation, guilt,
shame, thoughts of suicide and trouble with relationships. Retaliation included losing their job,
social ostracism, and harm to future career prospects. Both Harvey Weinstein and Les Moonves
(CBS) were accused of using their powerful positions in the entertainment industry to harm the
careers of women who rejected them. For example, Ashley Judd filed a defamation lawsuit
against Weinstein arguing that the lies he spread about her prevented her from getting work. In
another case, a jury found that Columbia University retaliated against an assistant professor who
reported sexual harassment by revoking her paid leave, speeding up her tenure process and
eventually denying her tenure (Robertson 2018).
Business frame. It was expected that the business frame would increase post-#MeToo
but not significantly because it is not something which appeals to liberals and the WSJ already
frames much of its coverage through the lens of business (H5g). There were no changes in either
the NYT or the WSJ post-#MeToo, thereby confirming H5g that #MeToo did not influence the
use of the business frame. It is important to note that the total number of articles this category is
small: 15 articles in the NYT and 12 in the WSJ, which represents just seven percent of articles
post-#MeToo. Similar to articles in the pre-#MeToo time frame, the coverage mostly discussed
the effect the public revelations had on the organization (e.g., resignations, terminations,
announcement of replacements). While the NYT discussed a wide variety of organizations (Fox
103
News, PBS/Charlie Rose, the Today show on NBC, WNYC public radio, Wynn Resorts, New
York Philharmonic, City Ballet, and two well-known restaurants), 75 percent of the coverage in
the WSJ which used this frame focused on just three people and organizations: Harvey Weinstein
and The Weinstein Company, Steve Wynn of Wynn Resorts, and Les Moonves of CBS. These
results show that despite the expectation that the business frame would not appeal to liberals, the
NYT reported on a wider variety of organizations in which SV was occurring which could help
the audience understand the pervasiveness of the problem.
Political frame. Since several politicians were accused of sexual misconduct after
#MeToo went viral, it was expected that the use of the political frame would increase in both the
NYT and the WSJ. There were no changes post-#MeToo in either the NYT or the WSJ, so H5h
was not confirmed, meaning that both newspapers continued to discuss politicians in the same
way they did before #MeToo. There were 30 articles which used the political frame in the post-
#MeToo time period and just over half of them were about three people: Donald Trump, Roy
Moore, and Brett Kavanaugh.
These articles included resignations from politicians such as Senator Al Franken and
Representative Trent Franks, a major Democratic donor who was reconsidering supporting
Democrats who urged Senator Franken to resign, and the Republican Governors Association
who returned donations from casino magnate Steve Wynn. In December 2017, three women
who first came forward in 2016 re-stated what Trump did to them and he attacked them as liars
and claimed Democrats were resurrecting allegations from 2016 because they could not prove
his campaign worked with Russia to influence the 2016 election. In an article in the NYT, this
incident was used to discuss how the issue of SV could help Democrats in the 2018 midterm
elections (Tackett 2017).
104
In a special election to fill the U.S. Senate seat of Jeff Sessions who was nominated to be
the new Attorney General, Republican candidate Roy Moore was accused of making sexual
advances towards four teenage girls four decades ago when he was 30 years old (Bromwich
2017). Initially Moore called it a political attack by the Democratic party, but as fellow
Republicans like Senator Mitch McConnell urged him to drop out of the race, Moore then
claimed it was “an effort by Mitch McConnell and his cronies to steal this election from the
people of Alabama” (Jamerson and Hook 2017).
The dismissal of allegations as politically motivated lies is one of the elements of
Schneider and Hannem’s (2019) concept of the politicization of sexual misconduct. Statements
by sources which claimed that allegations were a politically motivated attack were published in
both the NYT and WSJ in the post-#MeToo time frame, suggesting that at least part of their
findings extend beyond the 2016 presidential election. The dismissal of allegations as politically
motivated lies was mostly associated with Republican party (e.g., Trump, Moore, and
Kavanaugh) as only one Democratic congressman, John Conyers, suggested it was partisan
(Stolberg and Alcindor 2017).
The Supreme Court confirmation hearings for Brett Kavanaugh, which were just weeks
before the 2018 midterm elections, was also framed as a political problem. One article in the
NYT reported on a public opinion poll taken in 10 House battleground districts to gauge support
for Kavanaugh’s nomination (Cohn 2018). An article in the WSJ discussed the political strategy
and messaging ahead of the hearing in which Dr. Christine Blasey Ford would testify, and the
concern over the optics of having men question Dr. Ford since the GOP did not have any women
on the Judiciary Committee (Peterson, Andrews, and Nicholas 2018). Another article in the WSJ
discussed how the Kavanaugh hearings were energizing the GOP base and suggested it could
105
influence whether undecided senators would vote to confirm him to the U.S. Supreme Court
(Hughes 2018).
Framing SV as a political problem is harmful to victims for several reasons. First, as the
examples above illustrate, it reduces their traumatic experience to a means to an end for an
individual or group (e.g., win an election, get someone confirmed as a justice on the Supreme
Court). There were several examples in the coverage of journalists publishing unfounded claims
of an allegation being a part of a partisan attack, but the claim was not presented to the reader as
being unsupported by evidence. In a story about senate candidate Roy Moore and the statements
by four women that he made sexual advances towards them when he was in his thirties and they
were teenagers, Moore said that it was “completely false” and a “desperate political attack by the
National Democrat Party and the Washington Post” (Bromwich 2017). In the same article, a
former chairman of the Mobile County Republican Party echoed Moore’s assertion that the
Democratic Party was behind the Washington Post story about Moore, stating that most
Republicans in Alabama expect “these shenanigans being pulled by the Democrats as standard
operating procedure” (Bromwich 2017). In an article which described how the battle over
confirming Brett Kavanaugh to the Supreme Court may be “reinvigorating the Republican base,”
the story simply said that Kavanaugh denied the charges and repeated his unsubstantiated claim
that they were the result of Democratic anger over the outcome of the 2016 elections (Hughes
2018); there were no further details provided, not even a denial from Democrats.
Furthermore, if SV is presented as a political weapon by sources in the coverage to
influence political events, then victims – at least implicitly – are being portrayed as liars who use
SV as a tool to achieve a political goal. Outside the political context, this can lead to accusations
that victims have ulterior motives to say they are victims of SV, such as to “get back” at the man
106
or to win a lawsuit against a wealthy individual. The idea that victims lie about SV is a wellknown rape myth with has far reaching implications for how victims deal with the trauma,
including not coming forward at the time out of fear they will not be believed, as well as how
they are perceived by actors within the legal system, including law enforcement and juries (see
Edwards et al. 2011 and Franiuk et al. 2008 for a discussion of rape myth acceptance and its
implications).
The results show that on balance, structural media bias was controlling in the post-
#MeToo time frame and the #MeToo movement had little effect on how the stories of SV were
told. The conflict and legal frames, which are story formulas journalists are very familiar with,
represented nearly 80 percent of the coverage. The issue of SV was again presented as dispute
between parties, a contest with winners and losers, as opposed to a pervasive social problem
which disproportionately affects women and helps perpetuate inequality. In the next section I
will examine episodic and thematic framing as well as solutions.
Pre- and Post-#MeToo Analysis of Episodic and Thematic Framing
Since episodic coverage is so common, the ideology of the audience was not expected to
influence this variable and no differences were expected between the NYT and WSJ in the pre-
#MeToo time frame (H6). The results confirm H6: there were no differences between the NYT
and WSJ with regard to episodic and thematic framing. The coverage was episodic or primarily
episodic in 75 percent of articles in the NYT and 74 percent of articles in the WSJ.
#MeToo was not expected to influence the amount of episodic coverage, so H7 predicted
no differences in the post-#MeToo time frame. The results show that there were no changes in
episodic or thematic coverage when comparing the NYT’s pre-#MeToo coverage to its post-
#MeToo coverage, and there were no changes when comparing the WSJ’s pre-#MeToo coverage
107
to its post-#MeToo coverage, thus confirming H7. Similar to the results in the pre-#MeToo time
frame, episodic framing dominated the coverage: 76 percent in the NYT and 77 percent in the
WSJ.
Scholars have argued that some news values are best portrayed through specific acts and
individuals which results in episodic coverage (Greer 2003; Skidmore 1995; Cavender and
Mulcahy 1998). If episodic coverage is the result of journalistic routines, it could help explain
why episodic coverage is so common across both newspapers. In order to test this argument, I
ran two logistic regressions using episodic coverage as the dependent variable. In the first model
I included all eight news values as predictors and three control variables (pre-post-#MeToo,
newspaper, and article word count), and in the second model I included everything in model 1
but added the five frames. Model 2 had better explanatory value and the results, as well as the
predicted probabilities, are displayed in Table 9 (see Appendix C for a comparison of the two
models). The results show that the conflict, legal, and rich and famous news values and the
business and political frames increased the likelihood of episodic coverage.
The articles coded with the conflict and legal news values in the NYT and WSJ often
involved disputes between individuals or parties. In both newspapers, those coded as
rich/famous involved individuals who are rich and famous, and the vast majority of articles were
about well-known offenders, a type of storytelling which lent itself well to episodic reporting.
However, the absence of magnitude in both the NYT and WSJ was more likely to predict
episodic coverage. A review of the articles coded as mixed, primarily thematic or thematic (i.e.,
not episodic) show that a vast majority of them were also coded with magnitude news value, and
these articles discussed SV at organizations like the Catholic Church or in an industry and did
not focus on a single offender or victim.
108
Table 8. Logit Regression of Episodic Coverage
Variables
Coefficient
(S.E.)
Predicted Probabilities
Frames
Legal 1.961
(1.284)
--
Conflict 0.604
(1.372)
--
Victim 1.924
(1.372)
--
Political 3.582*
(1.409)
Present: 79%
Not Present: 10%
Business 2.796*
(1.355)
Present: 64%
Not present: 10%
News Values
Conflict 1.070*
(0.537)
Present: 24%
Not Present: 10%
Cover-up -0.680
(0.456)
--
Familiarity -0.032
(0.285)
--
Gov't Action 0.368
(0.276)
--
Immediacy 0.627
(0.372)
--
Legal 1.517***
(0.318)
Present: 33%
Not Present: 10%
Magnitude -0.612*
(0.290)
Present: 5%
Not Present: 10%
Rich/Famous 1.637***
(0.329)
Present: 35%
Not Present: 10%
Post-#MeToo 0.464
(0.289)
--
Newspaper -0.388
(0.276)
--
Word Count -0.0004
(0.0003)
--
Constant -1.867
(1.363)
--
*p<0.05; **p<0.01; ***p<0.001
Percent correctly predicted: 79.5%; (n = 513)
Predicted probabilities are while holding other variables at their mean.
109
The results show that the framing of an article can also predict episodic coverage. When
holding the other variables at their mean, the presence of the political frame increased the
likelihood of episodic coverage from 10 percent to 79 percent; for the business frame, the
likelihood of episodic coverage increased from 10 percent to 64 percent. Articles coded with the
political frame were almost always about a particular individual, such as Donald Trump, Roy
Moore, Brett Kavanaugh, and members of Congress. The articles coded with the business frame
were often about the revelation and aftermath that a prominent individual within the organization
– such as Bill O’Reilly and Roger Ailes at Fox News, Harvey Weinstein, Steve Wynn of Wynn
Resorts, and Les Moonves at CBS – engaged in SV.
The findings in both the pre- and post-#MeToo time frames are consistent with previous
research which found that the majority of coverage of SV is episodic (Mejia, Cheyne, and
Dorfman 2012; De Benedictis, Orgad, and Rottenberg 2019; Greer 2003). Iyengar (1991) has
argued that episodic coverage is more common because it is easier to write under time
constraints and since thematic reporting requires background research and data collection, it is
more costly to produce. Not only is episodic reporting easier and more cost effective, it also
reflects how journalists are trained to report on differences from one day to the next (Patterson
2013). The findings presented here also suggest that the news values used to select stories and
the framing used to tell stories contributes to episodic reporting.
Pre-#MeToo Analysis of Solutions
Based on previous research, H8a predicted that individual solutions would be emphasized
in the WSJ while societal-level solutions would be emphasized in the NYT if partisan media bias
was controlling. However, both newspapers were expected to publish more episodically framed
articles suggesting they would emphasize individual-level solutions (H8b). The results reveal
110
that there were no differences in the presence of individual-level or societal-level solutions
between the NYT and WSJ in the pre-#MeToo time period, so the structural media bias
hypothesis, H8b is confirmed and the partisan media bias hypothesis H8a is rejected.
These findings are inconsistent with previous research which found that conservative
newspapers were more likely to emphasize individual responsibility and less likely to mention
societal solutions for immigration and poverty (Kenix and Jarvandi 2019; Kim, Carvalho, and
Davis 2010). One possible explanation is that these two social problems are viewed differently
than SV because there is long-standing government infrastructure in place to address
immigration (e.g., Border Patrol, Homeland Security) and poverty (e.g., federal food and housing
assistance programs, tax credits), whereas the federal initiatives to address SV are not as wellknown or recently established. For example, the Office on Violence Against Women (within the
Department of Justice) was established in 1995 as part of the Violence Against Women Act
(VAWA) and it has to be to reauthorized every five years ("The History of the Office on
Violence Against Women" n.d.; Sullivan and Vazquez 2022). The legislation was recently
renewed in 2022 after it lapsed because of a partisan disagreement over whether a provision
which bans spouses convicted of domestic violence from owning firearms should be extended to
dating partners and stalkers, a change which was opposed by Republicans and the National Rifle
Association (Sullivan and Vazquez 2022). The provision, referred to as the “boyfriend
loophole,” had to be dropped in order to get the Act reauthorized. The White House Gender
Policy Council was recently established in 2021 ("Executive Order" 2021), and in May 2023 the
Biden-Harris Administration released the U.S. National Plan to End Gender-Based Violence:
Strategies for Action ("U.S. National Plan to End Gender-Based Violence" 2023). It could be
111
that the government spending on these programs has not risen to the level which would attract
vocal opposition from conservatives, so it is not reflected in the coverage.
Solutions were only discussed in 31 articles (21 percent) in the pre-#MeToo time period,
and when they were discussed, it was very brief. The most common type of solutions discussed
in both newspapers were individual-level solutions such as new policies or procedures at an
organization dealing with SV and changes to organizational culture. One article in the WSJ
about the major ad firm WPP discussed new programs they were implementing to close the
gender gap, including a networking and mentoring program for the company’s most senior
women (Tadena 2016). One article which devoted more than a sentence to changes which were
being made in the ad industry was published by the WSJ (Bruell 2016):
Despite the quick progress, a harder task awaits, say many observers and people
in the industry. If Madison Avenue wants to leave the "Mad Men" culture in the past,
they say, it must change the composition of ad agencies' workforces and root out
discriminatory thinking at all levels of the business.
"I don't think hiring is enough," says Susan Credle, global chief creative officer at
Interpublic Group's FCB. "We need to create cultures where underrepresented people can
find their voice. We need to continue to educate people on unconscious bias and what it
feels like to be in the minority."
Agencies such as Omnicom's DDB and BBDO are slotting in "unconscious bias"
training—seminars meant to instruct employees on how prejudices can creep into their
decision making without their knowing it. Internal recruiting targets for women and
minorities are becoming more common.
DDB also hired gender-equity certification firm Edge to audit and analyze gender
equality across the business, including pay, promotions and hiring practices.
The most frequent societal-level solution discussed was legislation, and within the U.S.
context the focus was mostly on extending the statute of limitations in criminal and civil case so
that survivors of child sexual abuse could file a police report or pursue civil litigation. This
solution was mainly discussed in the context of the Catholic Church and the thousands of victims
of clergy sex abuse.
112
Post-#MeToo Analysis of Solutions
The #MeToo movement wanted to bring awareness to the pervasiveness of the problem
of SV so that it would be addressed. Since liberals supported #MeToo more than conservatives,
if partisan media bias was controlling the discussion of solutions would increase in the NYT
post-#MeToo but not the WSJ (H9a). If structural media bias was controlling and #MeToo had
an influence, there would be an increase in both newspapers (H9b); if #MeToo had no influence
there would be no change in either newspaper (H9c). The results show no change in the
discussion of solutions post-#MeToo in both the NYT and WSJ, thus confirming H9c. This goal
of the #MeToo movement to expand awareness of possible solutions did not seem to be
achieved. This finding is consistent with previous research on #MeToo coverage in the U.K.
press. De Benedictis, Orgad, and Rottenberg (2019) argued that by failing to inform the public
about potential solutions, “the press can be understood to have helped defuse any potential that
#MeToo might contain as a mobilizing social force” (734).
Further analysis was performed on the individual and societal-level variables to
determine if there was an increase in one and a decrease in the other which would not have been
revealed by just analyzing the discussion of solutions. There were no significant changes in the
discussion of individual or societal-level solutions in either newspaper.
Solutions were discussed in 67 articles (18 percent) in the post-#MeToo time period and
similar to the pre-#MeToo time frame, the majority of solutions were individual-level ones such
as new organizational policies and procedures and changes to organizational culture and the
discussion was brief and vague. For example, in a story about sexual harassment at Nike, the
NYT wrote, “As women – and men – continue to come forward with complaints, Nike has begun
a comprehensive review of its human resources operations, making management training
113
mandatory and revising many of its internal reporting procedures” (Creswell, Draper, and
Abrams 2018). At Fox News, a workplace culture panel was created and there were a number of
networking and mentoring initiatives implemented to advance women at the network (Steel and
Grynbaum 2018).
As with the pre-#MeToo time period, legislation was the most common societal-level
solution discussed post-#MeToo. While extending or temporarily lifting the statute of
limitations on child sexual abuse was discussed, the post-#MeToo period included legislation
about sexual harassment specifically. For example, in New York City, the city council passed 11
bills to address sexual harassment in the workplace. The Stop Sexual Harassment in NYC Act
requires all private employers with more than 15 employees to conduct annual training for all
their employees, including interns and executives. Other bills extended protections to all
employees, regardless of the company’s size and another extends the statute of limitations for
filing a sexual harassment complaint from one year to three years.
Other societal-level solutions included a group called Male Champions for Change in
Australia which addresses cultural attitudes and social norms by redefining men’s roles as a way
to combat gender inequality. A lawmaker in Virginia said that more women needed to be
elected, noting how the increase in the number of women elected to the state House in in 2017
had “drastically reshaped the culture of the chamber” (Martin and Wines 2019). These solutions
are consistent with those of experts who recommend promoting social norms which protect
against violence, including modeling positive masculinity, and ending workplace sex segregation
and inequality (Basile et al. 2016; Schultz 2018-2019).
The overall results in this section show that episodic coverage is still pervasive and there
are no partisan differences between the two newspapers, either before or after #MeToo. In terms
114
of the solutions presented, there were no differences found between the two newspapers, and
individual-level solutions were the most common ones presented, which is consistent with the
majority of coverage being episodic and blame and responsibility being attributed to individual
people or organizations (Iyengar 1991).
Conclusion
The overall results for the analysis of framing shows strong support for structural media
bias, suggesting that the journalistic norms and routines have remained strong in a polarized
environment, and the exogenous shock of the #MeToo movement did not influence how
journalists told the story of SV. Most of the coverage was framed as a dispute between parties
with winners and losers rather than a social problem with gender inequality as the main cause.
Most of the articles which used the legal frame were about civil and criminal cases.
(Bing and Lombardo 1997) argue that discussing SV in this way is problematic because civil
litigation and criminal actions are reactive strategies, with law enforcement and attorneys only
intervening after the bad act has been committed and a legally defined standard of proof has been
met. Furthermore, prevention is only achieved through the deterrent effect of sanctions, whether
that is monetary or serving time in prison. Whether this is an effective strategy is debatable. As
legal scholar Vicki Schultz (2018-2019) argues, most men who harass know their behavior
bothers their victims or are indifferent to their feelings because “they are harassing them in order
to reinforce their own status, power, and social identity” (59). The cases of Harvey Weinstein
and Bill O’Reilly seem instructive here as the threat of monetary sanctions had absolutely no
deterrent effect; both men had paid substantial sums of money to multiple women and it did
nothing to deter their behavior.
115
The vast majority of coverage was episodic which is consistent with previous research on
SV (Mejia, Cheyne, and Dorfman 2012; De Benedictis, Orgad, and Rottenberg 2019; Greer
2003). Most of the articles did not discuss solutions but when they did, most of them were
individual-level solutions which aligns well with the episodic nature of the news coverage.
Since societal-level solutions were not presented as often, readers may perceive them to be less
significant and not worthy of time and other resources.
In the next chapter I will summarize my findings, discuss opportunities for future
research, provide some recommendations, and offer some concluding thoughts on my findings
and the media’s role in democracy.
116
Chapter 5
Conclusion
The role of the news media in democracy is to inform the public, provide a forum for
public debate, hold government and other powerful organizations accountable, and voice public
opinion (Graber 2003; Gans 2003; Curran 2005; Patterson 2013). Recently scholars have been
concerned that the high level of political polarization in the U.S. has affected news content, and
this could further exacerbate polarization and threaten democratic stability. The purpose of this
study was to examine news coverage of a partisan issue in ideologically disparate newspapers to
see if it was covered differently, which would be evidence of partisan media bias, or similarly,
which would be evidence of structural media bias. Since partisan media bias can be context
sensitive (Shultziner and Stukalin 2021), this study examined both an issue (SV) and an event
(#MeToo) which allowed me to test for partisan media bias in two different ways: I first
examined how SV was covered in the pre-#MeToo time frame, then again using the exogenous
shock of #MeToo in the post-#MeToo time frame.
On balance, the results provide much more evidence that structural media bias still
dominates news production and suggests that the norms and routines of journalism have thus far
been able to resist being influenced by the current polarized political environment. As the results
indicate, the NYT and WSJ relied on similar news values when deciding which stories to publish
about SV in both time frames. Even though the frequency in which the newspapers used certain
news values increased or decreased after #MeToo, the fact that both newspapers moved in the
same direction is evidence of strong journalistic norms and routines at work. In the post-
#MeToo time frame, there was a decrease in the legal news value and an increase in the conflict
and familiarity news values in both the NYT and WSJ. This finding demonstrates that both
117
newspapers were less reliant on the courts for stories about SV in the post-#MeToo time frame
and were able to generate stories from other sources, many of which were part of a news
reporter’s beat (e.g., politics, business, and entertainment).
Furthermore, both newspapers emphasized the viewpoints of powerful individuals and
organizations in both time frames, and a substantial amount of coverage was about a few wellknown individuals and organizations such as Harvey Weinstein, Donald Trump, Bill Cosby,
Brett Kavanaugh, Jeffrey Epstein, Fox News, and the Catholic Church. While the elites receive
a lot of attention, groups who are at highest risk received very little coverage in the NYT and
WSJ, and despite #MeToo’s efforts to highlight their experiences, coverage of those at high risk
did not increase in the post-#MeToo time frame.
In terms of how SV was covered, the NYT and WSJ heavily emphasized conflict and
drama which are well-known story formulas for journalists. The legal, conflict, and political
frames accounted for over 80 percent of the articles in the NYT and WSJ in both time periods.
The legal frame, which primarily consisted of court cases, accounted for the majority of coverage
in the pre- and post-#MeToo time periods, but it decreased significantly in the NYT post-
#MeToo (there was an insignificant decrease in the WSJ but it was trending in the hypothesized
direction). The conflict frame, which involved scandal and controversy, increased significantly
for both newspapers post-#MeToo. Several articles which used the conflict frame focused on
new revelations and the aftermath, including what the offender lost (e.g., the loss of professional
honors) and controversy (e.g., people upset with Pope Francis for comments he made when
speaking to reporters). The decrease in the legal frame accompanied by the increase in the
conflict frame suggests that the number of victims coming forward shifted the coverage away
from the courts which offered the public a different perspective on the issue.
118
The business and political frames, which were used by journalists less often, presented
SV as a problem with economic or political implications. In the economic frame, SV is
discussed in terms of how it can affect the bottom line of a corporation, in terms of the value of
the company (stock prices, revenue), and/or the potential impact of a high-level executive
leaving, or a replacement being named. In the political frame, SV can be presented as a problem
to by managed by a politician and his party (e.g., dismissing allegations as a partisan attack), or it
can be presented as an electoral advantage for the opposing party (e.g., how SV could help
Democrats in the 2018 midterm elections). Furthermore, these frames present SV as an issue
which can cost an individual or organization financially or politically, not as an ethical or moral
issue of fairness and justice for victims.
The victim frame, which focused on victims and their experiences with SV, accounted for
just four percent of the articles in the NYT and WSJ in the pre-#MeToo time frame and even
though #MeToo wanted to highlight victims and the impact SV has on their lives, the proportion
of articles using the victim frame did not change post-#MeToo (it remained at four percent).
This study also found that most of the coverage was episodic in nature during both time
periods, a type of framing that presents SV as an individual aberration which encourages
individual-level solutions, as opposed to a broader systemic and cultural problem which needs
societal-level solutions. Solutions were only discussed in 19 percent of articles, or to put it
differently, solutions were not mentioned in 81 percent of articles on SV. Many of the solutions
were individual-level solutions (e.g., new policies or procedures at an organization) which were
only briefly mentioned.
119
Implications for Democracy
This study found that the issue of SV was covered similarly in two ideologically different
national newspapers, suggesting that the journalistic norms and routines which produce structural
bias remain strong, even when presented with an exogenous shock like #MeToo. The fact that
#MeToo was associated with an increase in the amount of coverage but not the way the SV is
covered is further evidence of how much professional routines and norms dictate coverage.
Structural media bias, however, comes with its own challenges because in deciding what
to include and what to exclude, the news produced using these norms and routines resulted in
distorted coverage of SV. This has implications for democracy since the public needs good
information in order to make informed opinions.
The news media’s over-reliance on elites and official sources meant that they were the
allowed to control the terms of the debate and this made it difficult for the messages of #MeToo
to break through the gates and become part of the coverage in the NYT and WSJ. #MeToo
wanted to highlight the experiences of victims, particularly those from high-risk groups, yet the
victim frame was absent from over 90 percent of the articles in the sample. Victims are an
important voice in the story of SV, and the public rarely got to hear directly from them about
how SV has affected their lives, including the impact on their interpersonal relationships and
how it may have interfered with their economic opportunities. Their experiences are important
because they can help the public understand the gravity of the problem and create a sense of
urgency for those in power to do something about it.
The fact that people from high-risk groups were rarely covered is problematic because
poverty and various forms of oppression (racism, sexism, ableism) are risk factors for SV and
experts recommend that anti-poverty and anti-oppression efforts be included in sexual violence
120
prevention programs (National Sexual Violence Resource Center 2019; Greco and Dawgert
2007). Since these risk factors barely made it into the news at all, it was difficult for the public
to learn why they are at higher risk. Without this discussion included as a part of the public
discourse, solutions which could address their unique needs are unlikely to enter the public
debate.
The large amount of episodic coverage also directed the focus to the individual behavior
of the offender, as opposed to the structural conditions which enable the behavior, implicitly
suggesting that once that individual has been removed, the problem has been solved. If the
problem is solved simply by removing the offender, then there is no reason to have a debate
about preventative solutions, nor is there any reason to allocate resources on a societal-level to
stopping SV.
The frequent use of criminal and civil cases to tell the story of SV framed it as a dispute
between parties to be resolved in the courts and this type of coverage can preclude a debate about
broader potential institutional or structural responses (McDonald and Charlesworth 2013),
particularly the cultural norms or attitudes which reinforce gender inequality and create the
conditions for SV.
Furthermore, the legal and political framing presented the issue of SV as one with
winners and losers as opposed to a pervasive social problem which disproportionately affects
women and helps perpetuate inequality. This “game frame” (Patterson 1993, 2016) can also lead
to distrust and cynicism (Cappella and Jamieson 1997) and further polarization (Young 2017),
which is harmful to democracy.
Despite these shortcomings, the media did devote significantly more coverage in the
post-#MeToo time frame to SV. This alerted the public to a major social problem and powerful
121
people and organizations were held accountable. Recently there was public concern that
#MeToo was “dead” after Harvey Weinstein’s rape conviction was overturned by New York’s
highest court. However, #MeToo founder Tarana Burke said that the movement is bigger than
any single court case and as proof she said, “Ten years ago we could not get a man like Harvey
Weinstein into a courtroom” and argued the movement was responsible for the cultural shift
which allowed that to happen (Noveck 2024).
Opportunities for Future Research
There are many avenues of future research which can build on the findings in this study.
First, interviews with editors and journalists could help explain the “why” behind the results
found in this content analysis. For example, editors and journalists can explain what criteria they
use when deciding which stories to publish about SV and whether #MeToo had any impact on
their story selection process. They can also explain why the majority of the coverage of SV is
centered around civil and criminal cases, why victims and their experiences were not emphasized
in the coverage (particularly after #MeToo), and why solutions were not frequently discussed.
Second, future research can extend the analysis beyond the two well-resourced
newspapers studied here to examine regional newspapers and cable television. In their study of
partisan media bias, Gentzkow and Shapiro (2010) excluded the NYT and WSJ from their
analysis because they did not have a defined geographic market, and this suggests that regional
newspapers could be different from national newspapers. Since many people get their news
from cable television ("Cable News Fact Sheet" 2023), future research could compare the news
coverage of SV at Fox News and MSNBC.
Third, experts have expressed concern about a backlash to #MeToo and other gender (and
racial) advocacy efforts in the form of abortion bans and restrictions, prohibitions on workplace
122
antidiscrimination training, and defamation lawsuits against survivors of sexual violence
(Johnson, Ijoma, and Kim 2022). Future research could extend the time frame of this study to
look for evidence of backlash and broaden the analysis done here to examine the issue of gender
equality in news coverage.
Lastly, future research could build on this study by examining the coverage of SV outside
the U.S. This study was guided by previous research about news values and framing developed
using news coverage in Western democracies. While some journalistic values and methods are
universally agreed upon, the practice of journalism can be influenced by local language and
culture (Duran 2005). For example, ideologically driven journalism is part of the culture at
British newspapers, even among prestigious news outlets, whereas in the U.S. a blatant partisan
slant could damage the reputation and revenue potential of a prestigious newspaper (Bedingfield
2012). A study on #MeToo coverage in the U.K. press found that the focus was on victims in the
entertainment and fashion industry and excluded workers from sectors that are racialized and
classed such as those in hospitality and domestic service (De Benedictis, Orgad, and Rottenberg
2019), findings which are consistent with the results of this study. Since SV is a problem
throughout the world, including in conflict zones where it is used as a weapon of war, it would
be interesting to see how the issue is covered in other countries where news values and framing
could be quite different.
Recommendations
Political communication experts have argued that the news media does not have the
knowledge, power, resources or structure to perform the democracy-sustaining function expected
of it (Graber 2003; Cook 1998), leading Patterson (1993) to describe the news media as a
“miscast institution.” I agree with this assessment as well as Graber’s (2003) argument that the
123
media does not have a mandate to teach the public. However, Graber also argues that the
modern press provides Americans with sufficient knowledge to cope with their duties as
monitorial citizens, a term created by Schudson which means the public is informed and alert
enough to identify danger to their own well-being as well as the public good (Graber 2003).
There is currently a debate over whether this still holds true as many are concerned that people
are not paying attention to the threat to democracy posed by Donald Trump and the current
Republican Party. Former Republican Congresswoman Liz Cheney has warned that the U.S. is
“sleepwalking into dictatorship” (CBS News 2023).
The current threat to democracy has prompted experts to remind journalists of the critical
role they play in preserving democracy (Sullivan 2022; Maruf 2021) and journalists are
reflecting on how they cover the news (Darcy 2023). This provides journalists with a good
opportunity to consider how changing the way they report the news can help better inform the
public on important issues. One change they can make is to press sources for evidence when
they make baseless claims. For example, in the story where Roy Moore claimed that the
allegations of sexual misconduct against him were a “desperate partisan attack” by the
Washington Post and the Democratic party, there was nothing in the article which suggested that
the journalists pressed Moore or his campaign for evidence of this claim. When a politician
makes a baseless claim like this, journalists need to press him or her for evidence and if it is not
provided, state in the article that they were asked and they either did not respond or refused to
provide it. News organization could also keep a list of these types of unfounded statements by
politicians in order to contextualize future articles by saying something along the lines of “the
campaign [or his office] refused to provide evidence for this claim, just as it did when asked
about X and Y.”
124
While social media has presented challenges to news production both in terms of
economic survival and disinformation (Natividad 2021; Marston 2021), it can also provide an
opportunity for new and old media to work together to provide information to the public
(Chadwick 2017). Since the same journalistic norms do not apply to publishing on social media,
there is more information available to consume but it is not necessarily reliable, so people using
social media to disseminate information need the mainstream media to provide them with
legitimacy (Chadwick 2017). Experts and advocacy organizations can use social media to
challenge misperceptions about SV, highlight the unique needs of people from high-risk groups,
and offer solutions. If they are able to get the attention of mainstream media, their message will
be broadcast to a larger audience which could lead to changes in how the public perceives SV as
a social problem and greater support for policies to end it, both on a micro-level (e.g., at an
organization) and a macro level (e.g., changes in socio-cultural attitudes).
Finally, since structural media bias stems from the market-driven incentives of news
production, the nonprofit model of news could complement reporting on SV from the for-profit
sector. Because nonprofit news organizations generate revenue from donations, gifts,
memberships, and grants from foundations, Ryfe (2021) argues that they “are no longer in the
business of selling news to audiences, or selling audiences to advertisers” and this affects how
they produce the news (70). For example, periodicity plays a less prominent role in nonprofit
news production and many of the nonprofit news sites Ryfe studied did not publish on a regular
basis. Since journalistic impact is more important than profit, exclusivity is not a priority for
nonprofit news organizations and allows more room for collaboration. The government or other
foundations could issue grants to nonprofit news organizations to report on the challenges highrisk groups face in preventing SV and getting the help they need to recover from the trauma.
125
Stories from nonprofit news organizations could be picked up by the mainstream press and
disseminated more widely, much like the stories from ProPublica which revealed the
undisclosed gifts received by Supreme Court Justices Clarence Thomas and Samuel Alito.
One example from a nonprofit news organization which had an impact was the 2015 PBS
Frontline documentary Rape on the Night Shift that revealed the widespread sexual harassment
and rape of janitors who worked the night shift in Los Angeles office buildings ("Janitors Fast to
end Rape" n.d.). United Service Workers West showed the film and provided a safe space for
janitors to share their experiences and they decided that as part of their 2016 contract campaign
they would fight for legislation to protect all janitors from SV. This led to the passage of
AB1978, which “requires that all janitorial contractors register with the State and that all janitors
and supervisors get annual training to prevent sexual harassment and sexual violence” ("Janitors
Fast to end Rape" n.d.).
***
This study was undertaken to see if news coverage of a partisan issue was portrayed
differently by ideologically disparate newspapers in a highly polarized political environment.
The results suggest that the institutional norms which produce structural media bias work
similarly across newspapers with different partisan leanings, and while social movements like
#MeToo can get the attention of the media, those same norms continued to dictate how the issue
was covered. Since experts are highlighting the important role of journalism in preventing
democratic backsliding, it is encouraging to find that the professional journalistic norms and
routines are holding up in a polarized environment.
However, while structural media bias appears to have constrained the possibility of
partisan media bias, it still comes with advantages and drawbacks. The advantages are that the
126
media was able to draw the public’s attention to the problem of SV and hundreds of powerful
men were held accountable for their behavior. However, the media’s hyperfocus on elites and
official sources meant that other important voices were excluded, particularly that of victims.
This type of coverage can distort the public’s perception of the issue itself by not demonstrating
the devastating impact it has on victims, particularly those from high-risk groups, as well as who
commits SV and why. Furthermore, the emphasis on episodic reporting presents SV as an
individual aberration, a type of framing which suggests that removing the individual from the
workplace or society (e.g., prison) will solve the problem. Episodic reporting of this nature
generally does not lend itself well to discussions of socio-cultural and structural solutions to end
SV, making these solutions less likely to become part of the public debate about how to solve the
problem.
However, this is not a reason to lose hope in our media institutions. The fact that
journalists are willing to reflect on how changes to news production will better serve the public is
a testament to their dedication to preserving democracy. New technologies like social media and
the nonprofit model of news production can work alongside the mainstream news media in
creating a news product which better informs the public about important issues like SV.
127
References
Aalberg, Toril, and Audun Beyer. 2015. "Human Interest Framing of Irregular Immigration: An
Empirical Study of Public Preferences for Personalized News Stories in the United
States, France, and Norway." The American Behavioral Scientist 59 (7):858-875. doi:
10.1177/0002764215573258.
Aarøe, Lene. 2022. "The Foundations of Journalists’ Use of Exemplars in News Reporting."
American Political Science Association, Montréal, Québec, Canada September 17, 2022.
Ad Fontes Media. 2023. "The Media Bias Chart Version 11.0." Ad Fontes Media, accessed
December 16, 2023. https://adfontesmedia.com/gallery/.
Agirdas, Cagdas. 2015. "What Drives Media Bias? New Evidence From Recent Newspaper
Closures." Journal of Media Economics 28 (3):123-141. doi:
10.1080/08997764.2015.1063499.
Allern, Sigurd. 2002. "Journalistic and Commercial News Values: News Organizations as
Patrons of an Institution and Market Actors." Nordicom Review 23 (1-2):137-152. doi:
10.1515/nor-2017-0327.
Amber, Brittney, Tuyen K. Dinh, Arielle N. Lewis, Leidy D. Trujillo, and Margaret S. Stockdale.
2019. "High-Profile Sexual Misconduct Media Triggers Sex Harassment Recall and
Reinterpretation." Equality, Diversity and Inclusion: An International Journal 39 (1):68-
86. doi: 10.1108/EDI-12-2018-0222.
Aroustamian, Camille. 2020. "Time’s Up: Recognising Sexual Violence as a Public Policy Issue:
A Qualitative Content Analysis of Sexual Violence Cases and the Media." Aggression
and Violent Behavior 50:101341. doi: 10.1016/j.avb.2019.101341.
Bagdikian, Ben H. 2004. The New Media Monopoly. Boston: Beacon Press.
Baker, Andrea, Katrina Williams, and Usha M. Rodrigues. 2020. "#metoo 2.0 to #meNOmore:
Analysing Western Reporting About Sexual Violence in the Music Industry."
Journalism Practice 14 (2):191-207. doi: 10.1080/17512786.2019.1674683.
Barnes, Jeb, and Parker Hevron. 2018. "Framed? Judicialization and the Risk of Negative
Episodic Media Coverage." Law & Social Inquiry 43 (3):1059-1091. doi:
10.1111/lsi.12346.
Barnett, Michael D., and Emily N. Hilz. 2018. "The Psychology of the Politics of Rape: Political
Ideology, Moral Foundations, and Attitudes Toward Rape." Violence Against Women 24
(5):545-564. doi: 10.1177/1077801217708887.
Basile, Kathleen C., Sarah DeGue, Kathryn Jones, Kimberly Freire, Jenny Dills, Sharon G.
Smith, and Jerris L. Raiford. 2016. "STOP SV: A Technical Package to Prevent Sexual
Violence." National Center for Injury Prevention and Control, Centers for Disease
128
Control and Prevention, accessed July 24, 2020.
https://www.cdc.gov/violenceprevention/pdf/sv-prevention-technical-package.pdf.
Bauder, David. 2023. "Dominion Voting Case Exposes Post-election Fear at Fox News." Last
Modified February 23, 2023, accessed December 16, 2023.
https://apnews.com/article/politics-television-donald-trump-business1a4337a89c8abd952a814c60fa269b3c.
Bauder, David, Randall Chase, and Geoff Mulvihill. 2023. "Fox, Dominion Reach $787M
Settlement over Election Claims." Associated Press, Last Modified April 18, 2023,
accessed December 16, 2023. https://apnews.com/article/fox-news-dominion-lawsuittrial-trump-2020-0ac71f75acfacc52ea80b3e747fb0afe.
Bedingfield, Sid. 2012. "Journal's Health Care Plan Coverage Free of Murdock's [sic]
Conservative Bias." Newspaper Research Journal 33 (2):68-83. doi:
10.1177/073953291203300206.
Bekkers, Victor, Henri Beunders, Arthur Edwards, and Rebecca Moody. 2011. "New Media,
Micromobilization, and Political Agenda Setting: Crossover Effects in Political
Mobilization and Media Usage." The Information Society 27 (4):209-219. doi:
10.1080/01972243.2011.583812.
Belkin, Douglas. 2017. "University of Texas Survey Finds High Rate of Reported Rapes at
Flagship School; Few Students Reported Assaults, Underscoring the Challenges
Universities Face in Prevention." The Wall Street Journal March 24, 2017. Accessed
October 20, 2020 in Factiva.
Bender, Michael C., and Janet Hook. 2016a. "Donald Trump’s Lewd Comments about Women
Spark Uproar; Republican Candidate Apologizes for 2005 Recording, but Party Leaders
and Evangelicals Are Severely Critical." The Wall Street Journal, October 8, 2016.
Accessed October 20, 2020 in Factiva.
Bender, Michael C., and Janet Hook. 2016b. "Trump’s Path to Victory Diminishes; Donald
Trump Denies Sexual Misconduct Claims Amid Fresh Polls That His Presidential
Campaign Is Losing Ground." The Wall Street Journal, October 13, 2016. Accessed
October 20, 2020 in Factiva.
Benediktsson, Mike Owen. 2010. "The Deviant Organization and the Bad Apple CEO: Ideology
and Accountability in Media Coverage of Corporate Scandals." Social Forces 88
(5):2189-2216. doi: 10.1353/sof.2010.0032.
Bennett, W. Lance. 2016. News: The Politics of Illusion. 10th ed. Chicago, IL: University of
Chicago Press.
Bennett, W. Lance, and Barbara Pfetsch. 2018. "Rethinking Political Communication in a Time
of Disrupted Public Spheres." Journal of Communication 68 (2):243-253. doi:
10.1093/joc/jqx017.
129
Bennhold, Katrin. 2016. "Killings of 4 Young Gay Men in London Fuel Fears Over Date-Rape
Drug." The New York Times, November 25, 2016. Accessed October 17, 2020 in Factiva.
Billard, Thomas J. 2021. "Movement–Media Relations in the Hybrid Media System: A Case
Study from the U.S. Transgender Rights Movement." The International Journal of
Press/Politics 26 (2):341-361. doi: 10.1177/1940161220968525.
Bing, Janet M., and Lucien X. Lombardo. 1997. "Talking Past Each other about Sexual
Harassment: An Exploration of Frames for Understanding." Discourse & Society 8
(3):293-311. doi: 10.1177/0957926597008003002.
Boczkowski, Pablo J. 2010. News at Work: Imitation in an Age of Information Abundance.
Chicago: University of Chicago Press.
Bogen, Katherine W., Kaitlyn K. Bleiweiss, Nykia R. Leach, and Lindsay M. Orchowski. 2021.
"MeToo: Disclosure and Response to Sexual Victimization on Twitter." Journal of
Interpersonal Violence 36 (17-18):8257-8288. doi: 10.1177/0886260519851211.
Bouvier, Gwen. 2019. "How Journalists Source Trending Social Media Feeds: A Critical
Discourse Perspective on Twitter." Journalism Studies 20 (2):212-231. doi:
10.1080/1461670X.2017.1365618.
Bromwich, Jonah Engel. 2017. "Alabama Republicans Defend Roy Moore: ‘It Was 40 Years
Ago’." The New York Times, November 10, 2017. Accessed October 18, 2020 in Factiva.
Broockman, David, and Joshua Kalla. 2023. Selective Exposure and Partisan Echo Chambers in
Partisan Television Consumption: Evidence from Linked Viewership, Administrative,
and Survey Data. OSF Preprints July 28. Accessed December 15, 2023.
doi:10.31219/osf.io/b54sx.
Brown, Anna. 2022. "More Than Twice as Many Americans Support Than Oppose the #MeToo
Movement." Pew Research Center, Last Modified September 29, 2022, accessed October
6, 2023. https://www.pewresearch.org/social-trends/2022/09/29/more-than-twice-asmany-americans-support-than-oppose-the-metoo-movement/.
Brown, Danielle K., and Rachel R. Mourão. 2022. "No Reckoning for the Right: How Political
Ideology, Protest Tolerance and News Consumption Affect Support Black Lives Matter
Protests." Political Communication 39 (6):737-754. doi:
10.1080/10584609.2022.2121346.
Bruell, Alexandra. 2016. "Ad Industry Tries to Leave ‘Mad Men’ Mentality Behind; Agencies
Try to Combat Racism and Sexism with New Initiatives and Executive Appointments."
The Wall Street Journal, September 27, 2016. Accessed October 20, 2020 in Factiva.
Bruell, Alexandra. 2017. "BMW, Glaxo and Allstate Join Wave of Advertisers Withdrawing
from Bill O’Reilly’s Show; Brands Flee Controversy after Revelations That Fox News
Star and 21st Century Fox Paid $13 Million to Settle Five Harassment Claims." The Wall
Street Journal, April 4, 2017. Accessed October 20, 2020 in Factiva.
130
Budak, Ceren, Sharad Goel, and Justin M. Rao. 2016. "Fair and Balanced? Quantifying Media
Bias through Crowdsourced Content Analysis." Public Opinion Quarterly 80 (1):250-
271. doi: 10.1093/poq/nfw007.
Burns, Alexander, and Nick Corasaniti. 2016. "Donald Trump Assails His Accusers as Liars, and
Unattractive." The New York Times, October 14, 2016. Accessed October 16 ,2020 in
Factiva.
"Cable News Fact Sheet." 2023. Pew Research Center, Last Modified September 14, 2023,
accessed April 27, 2024. https://www.pewresearch.org/journalism/fact-sheet/cable-news/.
Cappella, Joseph N., and Kathleen Hall Jamieson. 1997. Spiral of Cynicism: The Press and the
Public Good. New York: Oxford University Press.
Carlsen, Audrey, Maya Salam, Claire Cain Miller, Denise Lu, Ash Ngu, Jugal K. Patel, and Zach
Wichter. 2018. "#MeToo Brought Down 201 Powerful Men. Nearly Half of Their
Replacements Are Women.". New York Times, Last Modified October 29, 2018,
accessed November 19, 2023.
https://www.nytimes.com/interactive/2018/10/23/us/metoo-replacements.html.
Cavender, Gray, and Aogan Mulcahy. 1998. "Trial by Fire: Media Constructions of Corporate
Deviance." Justice Quarterly 15 (4):697-717. doi: 10.1080/07418829800093951.
CBS News. 2016. "Do Donald Trump's Comments on the Recently Released Video Make You
Think Better of Donald Trump, Worse of Donald Trump, or Don't They Change Your
View of Donald Trump?". CBS News, accessed December 16, 2023 in Polling the
Nations.
https://ptn.infobase.com/articles/UG9sbFF1ZXN0aW9uOjYyODEwOA==?aid=98123.
CBS News. 2023. "Liz Cheney tells 'CBS News Sunday Morning' that the U.S. is 'sleepwalking
into a dictatorship'." CBSNews.com, Last Modified December 1, 2023, accessed April 7,
2024. https://www.cbsnews.com/news/liz-cheney-tells-cbs-news-sunday-morning-thatthe-u-s-is-sleepwalking-into-a-dictatorship/.
CBS News/New York Times. 2016. "From What You've Heard or Read, Do You Think the
Allegations That Donald Trump Made Unwanted Sexual Advances against Women Are
Mostly True or Mostly Not True?". CBS News/New York Times, accessed December 16,
2023 in Polling the Nations.
https://ptn.infobase.com/articles/UG9sbFF1ZXN0aW9uOjYyODE3Mg==?aid=98123.
Centers for Disease Control and Prevention. n.d. "Reporting on Sexual Violence." National
Sexual Violence Resource Center, Centers for Disease Control and Prevention, accessed
December 16, 2023. https://www.cdc.gov/violenceprevention/pdf/sv/sv-mediaguide508.pdf.
Chadwick, Andrew. 2017. The Hybrid Media System: Politics and Power. Second ed. New
York, NY: Oxford University Press.
131
Chan, Sewell, and Somini Sengupta. 2016. "Nadia Murad, Yazidi Woman Who Survived ISIS
Captivity, Wins Human Rights Prize." The New York Times, October 16, 2016. Accessed
October 16, 2020 in Factiva.
Cheit, Ross E., Yael Shavit, and Zachary Reiss-Davis. 2010. "Magazine Coverage of Child
Sexual Abuse, 1992-2004." Journal of Child Sexual Abuse 19 (1):99-117. doi:
10.1080/10538710903485575.
Chong, Dennis, and James N. Druckman. 2007. "Framing Theory." Annual Review of Political
Science 10 (1):103-126. doi: 10.1146/annurev.polisci.10.072805.103054.
Clark, Rosemary. 2016. "'Hope in a Hashtag': The Discursive Activism of #WhyIStayed."
Feminist Media Studies 16 (5):788-804. doi: 10.1080/14680777.2016.1138235.
Cobos, April. 2014. "'Rape Culture' Language and the News Media: Contested versus Noncontested Cases." Essachess 7 (2(14)):37-52.
Cohn, Nate. 2018. "Voters in Battleground Districts Are Divided on Kavanaugh." The New York
Times, September 25, 2018. Accessed October 19, 2020 in Factiva.
Coleman, Peter T. 2022. "Half the U.S. Believes Another Civil War Is Likely. Here Are the 5
Steps We Must Take to Avoid That." Last Modified January 6, 2022, accessed December
15, 2023. https://time.com/6133380/us-avert-civil-war/.
Cook, Timothy E. 1998. Governing with the News: The News Media as a Political Institution.
Chicago: University of Chicago Press.
Cortina, Lilia M., and Jennifer L. Berdahl. 2008. "Sexual Harassment in Organizations: A
Decade of Research in Review." In The SAGE Handbook of Organizational Behavior,
edited by Julian Barling and Cary L. Cooper, 469-497. London: SAGE Publications Ltd.
Creswell, Julie, Kevin Draper, and Rachel Abrams. 2018. "At Nike, Revolt Led by Women
Leads to Exodus of Male Executives." The New York Times, April 28, 2018. Accessed
October 18, 2020 in Factiva.
Crouse, Timothy. 1973. The Boys on the Bus. 1st ed. New York: Random House.
Cuklanz, Lisa. 2020. "Problematic News Framing of #MeToo." The Communication Review 23
(4):251-272. doi: 10.1080/10714421.2020.1829302.
Curran, James. 2005. "What Democracy Requires of the Media." In The Press, edited by Geneva
Overholser and Kathleen H. Jamieson, 120-140. Cary: Oxford University Press.
D'Alessio, D., and M. Allen. 2000. "Media Bias in Presidential Elections: A Meta-Analysis."
Journal of Communication 50 (4):133-156. doi: 10.1111/j.1460-2466.2000.tb02866.x.
Darcy, Oliver. 2023. "News Organizations Are Struggling to Cover Media Figures and
Politicians Peddling Delusions." CNN, Last Modified June 20, 2023, accessed April 28,
132
2024. https://www.cnn.com/2023/06/20/media/news-organizations-media-figurespoliticians-reliable-sources/index.html.
De Benedictis, Sara, Shani Orgad, and Catherine Rottenberg. 2019. "MeToo, Popular Feminism
and the News: A Content Analysis of UK Newspaper Coverage." European Journal of
Cultural Studies 22 (5-6):718-738. doi: 10.1177/1367549419856831.
de Vreese, Claes H., and Sophie Lecheler. 2012. "News Framing Research: An Overview and
New Developments." In The SAGE Handbook of Political Communication, edited by
Holli A. Semetko and Margaret Scammell, 292-306. SAGE Publications Inc.
Dickerson, Caitlin, and Stephanie Saul. 2017. "Two Colleges Bound by History Are Roiled by
the #MeToo Moment." The New York Times, December 2, 2017. Accessed October 18,
2020 in Factiva.
DiMaggio, Anthony. 2020. "Slanting the News: Media Bias and Its Effects." In News on the
Right: Studying Conservative News Cultures, edited by Anthony M. Nadler and A.J.
Bauer, 190-212. New York, NY: Oxford University Press.
Domke, D., M. D. Watts, D. V. Shah, and D. P. Fan. 1999. "The Politics of Conservative Elites
and the ‘Liberal Media’ Argument." Journal of Communication 49 (4):35-58. doi:
10.1111/j.1460-2466.1999.tb02816.x.
Duran, Ragip. 2005. "Trainers Can Remain Foreign to Local Journalists." Nieman Reports, Last
Modified June 15, 2005, accessed April 7, 2024.
https://niemanreports.org/articles/trainers-can-remain-foreign-to-local-journalists/.
Economist/YouGov. 2018. "How Serious of a Problem Do You Think Sexual Harassment Is in
the United States?". Economist/YouGov, accessed December 23, 2023 in Polling the
Nations.
https://ptn.infobase.com/articles/UG9sbFF1ZXN0aW9uOjcxNDk4MQ==?aid=98123.
Edwards, Katie M., Jessica A. Turchik, Christina M. Dardis, Nicole Reynolds, and Christine A.
Gidycz. 2011. "Rape Myths: History, Individual and Institutional-Level Presence, and
Implications for Change." Sex Roles 65 (11-12):761-773. doi: 10.1007/s11199-011-
9943-2.
Entman, R. M. 1993. "Framing: Toward Clarification of a Fractured Paradigm." Journal of
Communication 43 (4):51-58. doi: 10.1111/j.1460-2466.1993.tb01304.x.
Entman, Robert M. 1990. Democracy without Citizens: Media and the Decay of American
Politics. New York, NY: Oxford University Press.
"Executive Order on Establishment of the White House Gender Policy Council." 2021. White
House, Last Modified March 8, 2021, accessed December 17, 2023.
https://www.whitehouse.gov/briefing-room/presidential-actions/2021/03/08/executiveorder-on-establishment-of-the-white-house-gender-policy-council/.
133
Fahrenthold, David A. 2016. "Trump Recorded Having Extremely Lewd Conversation about
Women in 2005." The Washington Post, Last Modified October 8, 2016, accessed
November 11, 2023. https://www.washingtonpost.com/politics/trump-recorded-havingextremely-lewd-conversation-about-women-in-2005/2016/10/07/3b9ce776-8cb4-11e6-
bf8a-3d26847eeed4_story.html.
Farrow, Ronan. 2017. "From Aggressive Overtures to Sexual Assault: Harvey Weinstein's
Accusers Tell Their Stories." Conde Nast, Last Modified October 10, 2017, accessed
October 23, 2018. https://www.newyorker.com/news/news-desk/from-aggressiveovertures-to-sexual-assault-harvey-weinsteins-accusers-tell-their-stories.
Feldman, Lauren. 2018. "The Hostile Media Effect." In The Oxford Handbook of Political
Communication, edited by Kate Kenski and Kathleen Hall Jamieson, 549–564. Oxford
University Press.
Fenwick, Ben, and Alan Schwarz. 2015. "In Rape Case of Oklahoma Officer, Victims Hope
Conviction Will Aid Cause." The New York Times, December 12, 2015. Accessed
October 16, 2020 in Factiva.
Fernandez, Manny. 2019. "‘You Have to Pay With Your Body’: The Hidden Nightmare of
Sexual Violence on the Border." The New York Times, March 3, 2019. Accessed October
19, 2020 in Factiva.
"Fisher’s Exact Test." n.d. StatsTest.com, accessed May 28, 2023.
https://www.statstest.com/fischers-exact-test/.
Fitzgerald, Louise. 2019. "Unseen: The Sexual Harassment of Low-Income Women in America."
Equality, Diversity and Inclusion: An International Journal 39 (1):5-16. doi:
10.1108/EDI-08-2019-0232.
Flint, Joe. 2017. "Fox News Parent Had $10 Million in Harassment Settlement Costs in Quarter;
Costs Related to Sexual Harassment Claims against Former Fox News Chief Executive
Roger Ailes." The Wall Street Journal, May 10, 2017. Accessed October 20, 2020 in
Factiva.
Franiuk, Renae, Jennifer L. Seefelt, Sandy L. Cepress, and Joseph A. Vandello. 2008.
"Prevalence and Effects of Rape Myths in Print Journalism: The Kobe Bryant Case."
Violence Against Women 14 (3):287-309. doi: 10.1177/1077801207313971.
Friedman, Vanessa. 2018. "Modeling in the #TimesUp Era." The New York Times, September 6,
2018. Accessed October 18, 2020 in Factiva.
Frostenson. 2017. "The Women's Marches May Have Been the Largest Demonstration in US
History." Vox, Last Modified January 31, 2017, accessed December 16, 2023.
https://www.vox.com/2017/1/22/14350808/womens-marches-largest-demonstration-ushistory-map.
134
Gamson, William A., and Andre Modigliani. 1987. "The Changing Culture of Affirmative
Action." In Research in Political Sociology, edited by Richard Braumgart, 137-177. JAI
Press Inc.
Gamson, William, David Croteau, William Hoynes, and Theodore Sasson. 1992. "Media Images
and the Social Construction of Reality." Annual Review of Sociology 18:373-393.
Gans, Herbert J. 1979. Deciding What’s News: A Study of CBS Evening News, NBC Nightly
News, Newsweek, and Time. 1st ed. New York: Random House.
Gans, Herbert J. 2003. Democracy and the News. Oxford: Oxford University Press.
Garrett, R. Kelly, Jacob A. Long, and Min Seon Jeong. 2019. "From Partisan Media to
Misperception: Affective Polarization as Mediator." Journal of Communication 69
(5):490-512. doi: 10.1093/joc/jqz028.
Gentzkow, Matthew, and Jesse M. Shapiro. 2010. "What Drives Media Slant? Evidence From
U.S. Daily Newspapers." Econometrica 78 (1):35-71. doi: 10.3982/ECTA7195.
Gilchrist, Kristen. 2010. ""Newsworthy" Victims?: Exploring Differences in Canadian Local
Press Coverage of Missing/Murdered Aboriginal and White Women." Feminist Media
Studies 10 (4):373-390. doi: 10.1080/14680777.2010.514110.
Gilliam, Franklin D., and Shanto Iyengar. 2000. "Prime Suspects: The Influence of Local
Television News on the Viewing Public." American Journal of Political Science 44
(3):560-573.
Goldstein, Joseph. 2017. "Prison Supervisors Who Educated Subordinates About Rape Are
Charged With Sexual Assault." The New York Times, May 25, 2017. Accessed October
17, 2020 in Factiva.
Graber, Doris. 2003. "The Media and Democracy: Beyond Myths and Stereotypes." Annual
Review of Political Science 6:139-160. doi: 10.1146/annurev.polisci.6.121901.085707.
Graber, Doris A. 1997. Mass Media and American Politics. 5th ed. Washington, D.C: CQ Press.
Graber, Doris A., and Johanna Dunaway. 2015. Mass Media and American Politics. 9th ed.
Thousand Oaks, California: CQ Press, an imprint of SAGE Publications.
Graber, Doris A., and Gregory G. Holyk. 2011. "The News Industry." In The Oxford Handbook
of American Public Opinion and the Media, edited by George C. Edwards, Lawrence R.
Jacobs and Robert Y. Shapiro. Oxford: Oxford University Press.
Graham, Matthew H., and Milan W. Svolik. 2020. "Democracy in America? Partisanship,
Polarization, and the Robustness of Support for Democracy in the United States." The
American Political Science Review 114 (2):392-409. doi: 10.1017/S0003055420000052.
135
Greco, Donna, and Sarah Dawgert. 2007. "Poverty and Sexual Violence: Building Prevention
and Intervention Responses." Pennyslvania Coalition Against Rape, accessed August 24,
2022. https://pcar.org/resource/poverty-and-sexual-violence-building-prevention-andintervention-responses.
Green, Jeff. 2018. "#MeToo Has Implicated 414 High-Profile Executives and Employees in 18
Months." Last Modified June 25, 2018, accessed December 5, 2023.
https://time.com/5321130/414-executives-metoo/.
Greenberg, David. 2008. "The Idea of 'The Liberal Media' and Its Roots in the Civil Rights
Movement." The Sixties 1 (2):167-186. doi: 10.1080/17541320802457111.
Greer, Chris. 2003. Sex Crime and the Media: Sex Offending and the Press in a Divided Society.
Cullompton: Willan.
Griffith, James. 2019. "The Sexual Harassment–Suicide Connection in the U.S. Military:
Contextual Effects of Hostile Work Environment and Trusted Unit Leaders." Suicide and
Life‐Threatening Behavior 49 (1):41-53. doi: 10.1111/sltb.12401.
Groseclose, Tim, and Jeffrey Milyo. 2005. "A Measure of Media Bias." The Quarterly Journal
of Economics 120 (4):1191-1237. doi: 10.1162/003355305775097542.
Groseclose, Timothy. 2011. Left Turn: How Liberal Media Bias Distorts the American Mind. 1st
ed. New York: St. Martin's Press.
Grossman, Matt. 2017. "Media Bias (Real and Perceived) and the Rise of Partisan Media."
Niskanen Center, Last Modified November 6, 2017, accessed November 11, 2023.
https://www.niskanencenter.org/media-bias-real-perceived-rise-partisan-media/.
Grynbaum, Michael M., and John Koblin. 2016. "Gretchen Carlson of Fox News Files
Harassment Suit Against Roger Ailes." The New York Times, July 6, 2016. Accessed
October 16, 2020 in Factiva.
Habel, Philip D. 2012. "Following the Opinion Leaders? The Dynamics of Influence Among
Media Opinion, the Public, and Politicians." Political Communication 29 (3):257-277.
doi: 10.1080/10584609.2012.694986.
Haberman, Maggie. 2016a. "Inside Trump Tower, an Increasingly Upset and Alone Donald
Trump." The New York Times, October 9, 2016. Accessed October 16, 2020 in Factiva.
Haberman, Maggie. 2016b. "Women’s Accusations Drag Donald Trump’s Behavior Back to
Center Stage." The New York Times, October 13, 2016. Accessed October 16, 2020 in
Factiva.
Haidt, Jonathan, and Jesse Graham. 2007. "When Morality Opposes Justice: Conservatives Have
Moral Intuitions That Liberals May Not Recognize." Social Justice Research 20 (1):98-
116. doi: 10.1007/s11211-007-0034-z.
136
Hamilton, James. 2004. All the News That’s Fit to Sell: How the Market Transforms Information
into News. Princeton, N.J: Princeton University Press.
Harcup, Tony, and Deirdre O'Neill. 2017. "What Is News?: News Values Revisited (Again)."
Journalism Studies 18 (12):1470-1488. doi: 10.1080/1461670X.2016.1150193.
Haselmayer, Martin, Markus Wagner, and Thomas M. Meyer. 2017. "Partisan Bias in Message
Selection: Media Gatekeeping of Party Press Releases." Political Communication 34
(3):367-384. doi: 10.1080/10584609.2016.1265619.
Hassell, Hans J. G., John B. Holbein, and Matthew R. Miles. 2020. "There Is No Liberal Media
Bias in Which News Stories Political Journalists Choose to Cover." Science Advances 6
(14):eaay9344. doi: 10.1126/sciadv.aay9344.
Hauser, Christine. 2016. "Sexual Assault Victim From Elite New Hampshire School Speaks
Out." The New York Times, August 30, 2016. Accessed October 16, 2020 in Factiva.
Haynes, Chris, Jennifer Lee Merolla, and S. Karthick Ramakrishnan. 2016. Framing
Immigrants: News Coverage, Public Opinion, and Policy. New York: Russell Sage
Foundation.
Hersch, Joni, and Beverly Moran. 2013. "He Said, She Said, Let's Hear What The Data Say:
Sexual Harassment in the Media, Courts, EEOC, and Social Science." The Kentucky Law
Journal 101 (4):753-788.
"The History of the Office on Violence Against Women." n.d. Department of Justice, accessed
November 27, 2023. https://www.justice.gov/ovw/history.
Hmielowski, Jay D., Myiah J. Hutchens, and Michael A. Beam. 2020. "Asymmetry of Partisan
Media Effects?: Examining the Reinforcing Process of Conservative and Liberal Media
with Political Beliefs." Political Communication 37 (6):852-868. doi:
10.1080/10584609.2020.1763525.
Hofstetter, C. Richard. 1976. Bias in the News: Network Television Coverage of the 1972
Election Campaign. Columbus: Ohio State University Press.
Hook, Janet. 2016. "Hillary Clinton's Lead Jumped After Lewd Trump Video, Before Debate, in
WSJ Poll; Democratic Candidate's Advantage over Donald Trump Grows to 11
Percentage Points from 6 Points in September." The Wall Street Journal, October 10,
2016. Accessed October 20, 2020 in Factiva.
Hughes, Siobhan. 2018. "Kavanaugh Battle Revs Up GOP Base; Republicans Make Enthusiasm
Gains on Democrats, Who Have Been Energized by Their Opposition to Trump's
Policies, Tweets." The Wall Street Journal, October 4, 2018. Accessed October 21, 2020
in Factiva.
Human Rights Watch. 2012. "Cultivating Fear: The Vulnerability of Immigrant Farmworkers in
the US to Sexual Violence and Sexual Harassment." Human Rights Watch, Last
137
Modified May 15, 2012, accessed October 11, 2022.
https://www.hrw.org/report/2012/05/15/cultivating-fear/vulnerability-immigrantfarmworkers-us-sexual-violence-and.
Hurdle, Jon, and Graham Bowley. 2016. "Bill Cosby to Stand Trial in Sexual Assault Case in
June." The New York Times, September 6, 2016. Accessed October 16, 2020 in Factiva.
Iyengar, Shanto. 1991. Is Anyone Responsible?: How Television Frames Political Issues.
Chicago: University of Chicago Press.
Iyengar, Shanto, and Kyu S. Hahn. 2009. "Red Media, Blue Media: Evidence of Ideological
Selectivity in Media Use." Journal of Communication 59 (1):19-39. doi: 10.1111/j.1460-
2466.2008.01402.x.
Iyengar, Shanto, and Donald R. Kinder. 2010. News That Matters: Television and American
Opinion. Updated ed. Chicago: University of Chicago Press.
Jacobs, Julia, and Campbell Robertson. 2019. "L. Douglas Wilder, Ex-Governor of Virginia,
Accused of Sexual Harassment by College Student." The New York Times, March 28,
2019. Accessed October 19, 2020 in Factiva.
Jamerson, Joshua, and Janet Hook. 2017. "Democrat Doug Jones, Roy Moore's Rival, Gains
Traction in Alabama Senate Race; Former U.S. Attorney Focuses on Issues, Steers Clear
of Sexual Allegations Swirling Around Republican." The Wall Street Journal, November
16, 2017. Accessed October 20, 2020 in Factiva.
"Janitors Fast to End Rape on the Night Shift." n.d. SEIU: United Service Workers West,
accessed December 5, 2023. https://www.seiu-usww.org/janitors-fast-to-end-rape/.
Johnson-Cartee, Karen S. 2005. News Narratives and News Framing: Constructing Political
Reality. Lanham, MD: Rowman & Littlefield Publishers.
Johnson, Andrea, Samone Ijoma, and Da Hae Kim. 2022. "#MeToo Five Years Later: Progress
and Pitfalls in State Workplace Anti-Harassment Laws ". National Women's Law Center,
Last Modified October 2022, accessed October 15, 2022. https://nwlc.org/wpcontent/uploads/2022/10/final_2022_nwlcMeToo_Report.pdf.
Johnston, Katie. 2017. "For Low-Wage Workers, Many Obstacles to Reporting Sexual
Harassment." Boston Globe, November 26, 2017. Accessed December 11, 2023 in
ProQuest Central.
Jordan, Donald L., and Benjamin I. Page. 1992. "Shaping Foreign Policy Opinions: The Role of
TV News." The Journal of Conflict Resolution 36 (2):227-241. doi:
10.1177/0022002792036002001.
Jose, Rupa, James H. Fowler, and Anita Raj. 2021. "Political Differences in American Reports of
Sexual Harassment and Assault." Journal of Interpersonal Violence 36 (15-16):7695-
7721. doi: 10.1177/0886260519835003.
138
Jost, John T., Christopher M. Federico, and Jaime L. Napier. 2009. "Political Ideology: Its
Structure, Functions, and Elective Affinities." Annual Review of Psychology 60 (1):307-
337. doi: 10.1146/annurev.psych.60.110707.163600.
Jurkowitz, Mark, Amy Mitchell, Elisa Shearer, and Mason Walker. 2020. "Democrats Report
Much Higher Levels of Trust in a Number of News Sources than Republicans." Pew
Research Center, Last Modified January 24, 2020, accessed September 22, 2023.
https://www.pewresearch.org/journalism/2020/01/24/democrats-report-much-higherlevels-of-trust-in-a-number-of-news-sources-than-republicans/.
Just, Marion R., Ann N. Crigler, Dean E. Alger, Timothy E. Cook, Montague Kern, and Darrell
M. West. 1996. Crosstalk: Citizens, Candidates, and the Media in a Presidential
Campaign. Chicago: The University of Chicago Press.
Kafka, Peter. 2022. "The New York Times’s Old White Democrats Problem." Vox, Last
Modified April 6, 2022, accessed October 16, 2023.
https://www.vox.com/recode/23011969/new-york-times-subscriber-athletic-age-peterkafka-media-column.
Kalmoe, Nathan P., and Lilliana Mason. 2022. Radical American Partisanship: Mapping Violent
Hostility, Its Causes, and the Consequences for Democracy. First ed. Chicago: University
of Chicago Press.
Kantor, Jodi, and Megan Twohey. 2017. "Harvey Weinstein Paid Off Sexual Harassment
Accusers for Decades." New York Times, October 5, 2017. Accessed October 17, 2020 in
Factiva.
Kearl, Holly. 2018. "The Facts Behind the #MeToo Movement: A National Study on Sexual
Harassment and Assault." Stop Street Harassment, Raliance, & U.C. San Diego Center on
Gender Equity and Health, Last Modified February 2018, accessed July 22, 2020.
http://www.stopstreetharassment.org/wp-content/uploads/2018/01/Full-Report-2018-
National-Study-on-Sexual-Harassment-and-Assault.pdf.
Kenix, Linda Jean, and Reza Jarvandi. 2019. "The Role of Ideology in the International
Mainstream News Media Framing of Refugees: A Comparison between Conservative and
Liberal Newspapers in United States, United Kingdom and Australia." Journal of
Applied Journalism & Media Studies 8 (3):349-365. doi: 10.1386/ajms_00006_1.
Khalifeh, H., P. Moran, R. Borschmann, K. Dean, C. Hart, J. Hogg, D. Osborn, S. Johnson, and
L. M. Howard. 2015. "Domestic and Sexual Violence against Patients with Severe
Mental Illness." Psychological Medicine 45 (4):875-886. doi:
10.1017/S0033291714001962.
Kim, Sei-Hill, John P. Carvalho, and Andrew C. Davis. 2010. "Talking about Poverty: News
Framing of Who is Responsible for Causing and Fixing the Problem." Journalism &
Mass Communication Quarterly 87 (3-4):563-581. doi: 10.1177/107769901008700308.
139
Kitzinger, Jenny. 2004. "Media Coverage of Sexual Violence Against Women and Children." In
Women and Media: International Perspectives, edited by Karen Ross and Carolyn M.
Byerly, 13-38. Oxford, UK: Blackwell Publishing.
Klein, Ezra. 2020. "Why the Media Is So Polarized — And How It Polarizes Us." Last Modified
January 28, 2020, accessed September 22, 2023.
https://www.vox.com/2020/1/28/21077888/why-were-polarized-media-book-ezra-news.
Kleinfeld, Rachel. 2021. "The Rise of Political Violence in the United States." Journal of
Democracy 32 (4):160-176.
Kull, Steven, Clay Ramsay, and Evan Lewis. 2003-04. "Misperceptions, the Media, and the Iraq
War." Political Science Quarterly 118 (4):569-598. doi: 10.1002/j.1538-
165X.2003.tb00406.x.
Kurtzleben, Danielle. 2016. "Here's The List Of Women Who Accused Donald Trump Of Sexual
Misconduct." National Public Radio, Last Modified October 20, 2023, accessed
December 16, 2023. https://www.npr.org/2016/10/13/497799354/a-list-of-donaldtrumps-accusers-of-inappropriate-sexual-conduct.
Larcinese, Valentino, Riccardo Puglisi, and James M. Snyder. 2011. "Partisan Bias in Economic
News: Evidence on the Agenda-Setting Behavior of U.S. Newspapers." Journal of
Public Economics 95 (9):1178-1189. doi: 10.1016/j.jpubeco.2011.04.006.
Lehman, Susan. 2015. "News and Opinion Strictly Separate at The Times." The New York
Times Company, Last Modified September 9, 2015, accessed April 5, 2024.
https://www.nytimes.com/2015/09/09/insider/news-and-opinion-strictly-separate-at-thetimes.html.
Levendusky, Matthew S. 2022. "How Does Local TV News Change Viewers' Attitudes? The
Case of Sinclair Broadcasting." Political Communication 39 (1):23-38. doi:
10.1080/10584609.2021.1901807.
Levitsky, Steven, and Daniel Ziblatt. 2018. How Democracies Die. First ed. New York: Crown.
Lewis, A. Elaine. 2018. "Who Is at Highest Risk of Sexual Harassment?". American Civil
Liberties Union, Last Modified January 18, 2018, accessed December 16, 2023.
https://www.aclu.org/news/womens-rights/who-highest-risk-sexual-harassment.
Lichter, Robert S. 2017. "Theories of Media Bias." In The Oxford Handbook of Political
Communication, edited by Kate Kenski and Kathleen Hall Jamieson, 403-416. New
York, NY: Oxford University Press.
Lichter, S. Robert, Stanley Rothman, and Linda S. Lichter. 1986. The Media Elite: America's
New Powerbrokers. Bethesda, MD: Adler & Adler.
Lippmann, Walter. 1922. Public Opinion. New York: Harcourt, Brace.
140
Lowry, Dennis T. 2008. "Network TV News Framing of Good vs. Bad Economic News under
Democrat and Republican Presidents: A Lexical Analysis of Political Bias." Journalism
& Mass Communication Quarterly 85 (3):483-498. doi: 10.1177/107769900808500301.
Loya, Rebecca M. 2015. "Rape as an Economic Crime: The Impact of Sexual Violence on
Survivors’ Employment and Economic Well-Being." Journal of Interpersonal Violence
30 (16):2793-2813. doi: 10.1177/0886260514554291.
Magley, Vicki J., Charles L. Hulin, Louise F. Fitzgerald, and Mary DeNardo. 1999. "Outcomes
of Self-Labeling Sexual Harassment." Journal of Applied Psychology 84 (3):390-402.
doi: 10.1037/0021-9010.84.3.390.
Marist Poll/National Public Radio/PBS NewsHour. 2018. "If the Charge of Sexual Assault
during a Party in High School by Christine Blasey Ford against Brett Kavanaugh Is True,
Do You Think Brett Kavanaugh: Should Be Confirmed to the U.S. Supreme Court or
Should Not Be Confirmed to the U.S. Supreme Court?". Marist Poll/National Public
Radio/PBS NewsHour, accessed December 23, 2023 in Polling the Nations.
https://ptn.infobase.com/articles/UG9sbFF1ZXN0aW9uOjcxNTkxNw==?aid=98123.
Marston, Celeste Katz. 2021. "Challenges and Possible Solutions to Winning Back Trust in
Journalism Post-Trump." Nieman Foundation for Journalism at Harvard, Last Modified
November 12, 2020, accessed April 22, 2024.
https://niemanreports.org/articles/challenges-possible-solutions-winning-back-trust-injournalism-post-trump/.
Martin, Jonathan, and Michael Wines. 2019. "Unmistakable Divide Among Women in Virginia
Over Accusations Against Fairfax." The New York Times, February 13, 2019. Accessed
October 19, 2020 in Factiva.
Maruf, Ramishah. 2021. "News Outlets Should Be Openly ‘Pro-democracy,’ Journalism
Professor Says." CNN, Last Modified December 12, 2021, accessed April 7, 2024.
Mason, Lilliana. 2018. Uncivil Agreement: How Politics Became Our Identity. Chicago, IL: The
University of Chicago Press.
Mastin, Teresa, Jounghwa Choi, Gia Elise Barboza, and Lori Post. 2007. "Newspapers' Framing
of Elder Abuse: It's Not a Family Affair." Journalism & Mass Communication Quarterly
84 (4):777-794. doi: 10.1177/107769900708400408.
McCluskey, Michael R. 2017. News Framing of School Shootings: Journalism and American
Social Problems. Lanham: Lexington Books.
McDonald, Paula, and Sara Charlesworth. 2013. "Framing Sexual Harassment through Media
Representations." Women's Studies International Forum 37. doi:
10.1016/j.wsif.2012.11.003.
Medina, Jennifer. 2017. "Too Scared to Report Sexual Abuse. The Fear: Deportation." The New
York Times, April 30, 2017. Accessed October 20, 2017 in Factiva.
141
Mejia, Pamela, Andrew Cheyne, and Lori Dorfman. 2012. "News Coverage of Child Sexual
Abuse and Prevention, 2007-2009." Journal of Child Sexual Abuse: Preventing
Childhood Sexual Abuse: A Range of Strategies 21 (4):470-487. doi:
10.1080/10538712.2012.692465.
metoo. 2019. "metoo Impact Report 2019." metoo., accessed December 18, 2023.
https://metoomvmt.org/wp-content/uploads/2020/01/2019-12-
09_MeToo_ImpactReport_VIEW_4.pdf.
Miner-Rubino, Kathi, and Lilia M. Cortina. 2007. "Beyond Targets: Consequences of Vicarious
Exposure to Misogyny at Work." Journal of Applied Psychology 92 (5):1254-1269. doi:
10.1037/0021-9010.92.5.1254.
Mitchell, Amy. 2018. "Americans Still Prefer Watching to Reading the News – and Mostly Still
Through Television." Pew Research Center, Last Modified December 3, 2018, accessed
December 17, 2023. https://www.pewresearch.org/journalism/2018/12/03/americansstill-prefer-watching-to-reading-the-news-and-mostly-still-through-television/.
Mullainathan, Sendhil, and Andrei Shleifer. 2005. "The Market for News." The American
Economic Review 95 (4):1031-1053. doi: 10.1257/0002828054825619.
Naseralla, Eyad J., S. Glenn Baker, and Ruth H. Warner. 2021. "The Influence of Political
Partisanship on Perceptions of Sexual Assault." Analyses of Social Issues and Public
Policy 21 (1):210-236. doi: 10.1111/asap.12273.
National Sexual Violence Resource Center. 2015. "What is Sexual Violence." National Sexual
Violence Resource Center, accessed March 17, 2024.
https://www.nsvrc.org/sites/default/files/publications_nsvrc_media-packet_1.pdf.
National Sexual Violence Resource Center. 2019. "Risk & Protective Factors." National Sexual
Violence Resource Center, Last Modified April 2019, accessed August 23, 2022.
https://www.nsvrc.org/risk-and-protective-factors.
National Sexual Violence Resource Center. n.d.-a. "Gender and Sexual Identity." National
Sexual Violence Resource Center, accessed December 16, 2023.
https://www.nsvrc.org/sarts/toolkit/6-8.
National Sexual Violence Resource Center. n.d.-b. "Immigrant Victims of Sexual Assault."
National Sexual Violence Resource Center, accessed December 16, 2023.
https://www.nsvrc.org/sarts/toolkit/6-12.
National Sexual Violence Resource Center. n.d.-c. "Victims Who Have Disabilities." National
Sexual Violence Resource Center, accessed December 23, 2023.
https://www.nsvrc.org/sarts/toolkit/6-6.
Natividad, Ivan. 2021. "How Did Trump Change American Journalism?". UC Regents, Last
Modified January 27, 2021, accessed April 22, 2024.
https://news.berkeley.edu/2021/01/27/how-did-trump-change-american-journalism.
142
Neuman, W. Russell, Marion R. Just, and Ann N. Crigler. 1992. Common Knowledge: News and
the Construction of Political Meaning. Chicago, IL: University of Chicago Press.
Neuman, William. 2019. "They Were Accused of Sexual Harassment, but the City Hid It From
Future Employers." The New York Times, March 19, 2019. Accessed October 19, 2020 in
Factiva.
Niven, David. 1999. "Partisan Bias in the Media? A New Test." Social Science Quarterly 80
(4):847-857.
Noveck, Jocelyn. 2024. "#MeToo Advocates Vow the Reckoning Will Continue after
Weinstein’s Conviction Is Overturned." Associated Press, Last Modified April 26, 2024,
accessed April 27, 2024. https://apnews.com/article/metoo-weinstein-tarana-burke-anitahill-8272c7113b53714c51cb54111cd01bee.
O'Hara, Shannon. 2012. "Monsters, Playboys, Virgins and Whores: Rape Myths in the News
Media's Coverage of Sexual Violence." Language and Literature: Journal of the Poetics
and Linguistics Association 21 (3):247-259. doi: 10.1177/0963947012444217.
O'Neill, Deirdre, and Tony Harcup. 2009. "News Values and Selectivity." In The Handbook of
Journalism Studies, edited by Karin Wahl-Jorgensen and Thomas Hanitzsch, 161-174.
New York: Routledge.
Ohio Alliance to End Sexual Violence. n.d. "Sexual Violence & Women of Color: A Fact
Sheet." Ohio Alliance to End Sexual Violence, accessed August 25, 2022.
https://oaesv.org/wp-content/uploads/2021/04/oaesv-sexual-violence-women-ofcolor.pdf.
Ohlheiser, Abby. 2017. "The Woman behind ‘Me Too’ Knew the Power of the Phrase When She
Created It - 10 Years Ago." The Washington Post, October 19, 2017. Accessed October
13, 2023 in Global Newsstream, ProQuest Central.
Onwuachi-Willig, Angela. 2018-2019. "What about #UsToo?: The Invisibility of Race in the
#MeToo Movement." The Yale Law Journal Forum 128:105-120.
Opening the Circle. n.d. "For Families & Friends of Survivors." Opening the Circle, accessed
December 16, 2023. http://www.openingthecircle.ca/defining-abuse/for-families-friendsof-survivors.
Padgett, Jeremy, Johanna L. Dunaway, and Joshua P. Darr. 2019. "As Seen on TV? How
Gatekeeping Makes the U.S. House Seem More Extreme." Journal of Communication 69
(6):696-719. doi: 10.1093/joc/jqz039.
Park, Andrea. 2017. "#MeToo Reaches 85 Countries with 1.7M Tweets." CBS Interactive, Last
Modified October 24, 2017, accessed May 10, 2018.
https://www.cbsnews.com/news/metoo-reaches-85-countries-with-1-7-million-tweets/.
Patterson, Thomas E. 1993. Out of Order. 1st ed. New York: A. Knopf.
143
Patterson, Thomas E. 2013. Informing the News: The Need for Knowledge-Based Journalism.
New York: Vintage Books.
Patterson, Thomas E. 2016. "News Coverage of the 2016 Presidential Primaries: Horse Race
Reporting Has Consequences." Shorenstein Center on Media, Politics and Public Policy,
Last Modified July 11, 2016, accessed December 5, 2023.
https://shorensteincenter.org/news-coverage-2016-presidential-primaries/.
Paulussen, Steve, and Raymond A. Harder. 2014. "Social Media References in Newspapers:
Facebook, Twitter and YouTube as Sources in Newspaper Journalism." Journalism
Practice 8 (5):542-551. doi: 10.1080/17512786.2014.894327.
Pelled, Ayellet, Josephine Lukito, Jordan Foley, Yini Zhang, Zhongkai Sun, Jon C. W.
Pevehouse, and Dhavan V. Shah. 2021. "Death Across the News Spectrum: A Time
Series Analysis of Partisan Coverage Following Mass Shootings in the United States
Between 2012 and 2014." International Journal of Communication 15:2116-2135.
Pennington, Rosemary, and Jessica Birthisel. 2016. "When New Media Make News: Framing
Technology and Sexual Assault in the Steubenville Rape Case." New Media & Society
18 (11):2435-2451. doi: 10.1177/1461444815612407.
Pennsylvania Coalition Against Rape. 2017. "Racism and Sexual Violence: What's the
Connection?". Pennsylvania Coalition Against Rape, accessed December 17, 2023.
https://pcar.org/sites/default/files/resourcepdfs/tab_2017_racismsexual_violence_connections-508d.pdf.
Peterson, Cora, Sarah DeGue, Curtis Florence, and Colby N. Lokey. 2017. "Lifetime Economic
Burden of Rape Among U.S. Adults." American Journal of Preventive Medicine 52
(6):691-701. doi: 10.1016/j.amepre.2016.11.014.
Peterson, Kristina, Natalie Andrews, and Peter Nicholas. 2018. "Senators Get Their Message Out
Ahead of Brett Kavanaugh Hearing; Assault Allegation against Supreme Court Nominee
Spurs Competing Narratives by GOP, Democrats." The Wall Street Journal, September
18, 2018. Accessed October 21, 2020 in Factiva.
Pew Research Center. 2014. "Political Polarization in the American Public." Pew Research
Center, Last Modified June 12, 2014, accessed March 21, 2018. http://www.peoplepress.org/2014/06/12/section-1-growing-ideological-consistency/.
Pew Research Center. 2019. "In a Politically Polarized Era, Sharp Divides in Both Partisan
Coalitions." Pew Research Center, Last Modified December 17, 2019, accessed
December 16, 2023. https://www.pewresearch.org/politics/2019/12/17/in-a-politicallypolarized-era-sharp-divides-in-both-partisan-coalitions/.
Pierson, Paul, and Eric Schickler. 2020. "Madison's Constitution Under Stress: A Developmental
Analysis of Political Polarization." Annual Review of Political Science 23 (1):37-58. doi:
10.1146/annurev-polisci-050718-033629.
144
Piore, Adam. 2021. "Breaking Right: The Wall Street Journal's Stubborn Conservatism."
Columbia Journalism Review, Last Modified September 14, 2021, accessed November
11, 2023.
https://www.cjr.org/special_report/breaking_right_wall_street_journal_stubborn_murdoc
h.php.
Porter, Catherine. 2018. "Haiti Suspends Oxfam Great Britain After Sex Scandal." The New York
Times, February 22, 2018. Accessed October 18, 2020 in Factiva.
Povoledo, Elisabetta. 2019. "The Vatican Is Talking About Clerical Abuse, but Italy Isn’t. Here’s
Why." The New York Times, February 22, 2019. Accessed October 19, 2020 in Factiva.
Poynter Staff. 2017. "Which Sexual Harassment and Assault Stories Should You Cover? Here
Are Some Guidelines." Poynter Institute, Last Modified November 29, 2017, accessed
May 31, 2020. https://www.poynter.org/ethics-trust/2017/which-sexual-harassment-andassault-stories-should-you-cover-here-are-some-guidelines/.
Quinnipiac University Poll. 2017. "If a Political Candidate Has Been Accused of Sexual
Harassment by Multiple People, Would You Still Consider Voting for Them If You
Agreed with Them on the Issues, or Would You Definitely Not Vote for Them?".
Quinnipiac University Poll, accessed December 16, 2023 in Polling the Nations.
https://ptn.infobase.com/articles/UG9sbFF1ZXN0aW9uOjcwMzAwMw==?aid=98123.
Quinnipiac University Poll. 2018. "As You May Know, Dr. Christine Blasey Ford Has Accused
Judge Brett Kavanaugh of Sexually Assaulting Her in High School. Judge Kavanaugh
Has Denied This Accusation. If You Had to Choose, Which Do You Tend to Believe
Most: The Accusation of Sexual Assault Made by Dr. Christine Blasey Ford or the
Denial Made by Judge Brett Kavanaugh?". Quinnipiac University Poll, accessed
December 16, 2023 in Polling the Nations.
https://ptn.infobase.com/articles/UG9sbFF1ZXN0aW9uOjcxNjA3MQ==?aid=98123.
Quinnipiac University Poll. 2019. "In General, Do You Think That Sexual Assault Is a Serious
Problem, or Not?". Quinnipiac University Poll, accessed December 16, 2023 in Polling
the Nations.
https://ptn.infobase.com/articles/UG9sbFF1ZXN0aW9uOjc0MjA1NA==?aid=98123.
Raghu, Maya, and Joanna Suriani. 2017. "#MeTooWhatNext: Strengthening Workplace Sexual
Harassment Protections and Accountability." National Women's Law Center, Last
Modified December 21, 2017, accessed June 27, 2018.
https://nwlc.org/resource/metoowhatnext-strengthening-workplace-sexual-harassmentprotections-and-accountability/.
Ransom, Jan. 2018. "Is the Case Against Harvey Weinstein in Jeopardy?" The New York Times,
October 29, 2018. Accessed October 19, 2020 in Factiva.
Reed, Maggie E., Linda L. Collinsworth, and Louise F. Fitzgerald. 2005. "There’s No Place Like
Home: Sexual Harassment of Low Income Women in Housing." Psychology, Public
Policy, and Law 11 (3):439-462. doi: 10.1037/1076-8971.11.3.439.
145
Remnick, Noah. 2016. "After a Professor Is Cleared of Sexual Harassment, Critics Fear ‘Cultural
Silence’ at Yale." The New York Times, July 8, 2016. Accessed October 16, 2020 in
Factiva.
Riffe, Daniel, Stephen Lacy, Brendan Watson, and Frederick Fico. 2019. Analyzing Media
Messages: Using Quantitative Content Analysis in Research. Fourth ed. New York:
Routledge.
Robertson, Aaron. 2018. "Former Assistant Professor Wins Harassment Suit Against Columbia."
The New York Times, July 26, 2018. Accessed October 18, 2020 in Factiva.
Ryfe, David. 2021. "The Economics of News and the Practice of News Production." Journalism
Studies 22 (1):60-76. doi: 10.1080/1461670X.2020.1854619.
Sacks, Meghan, Alissa R. Ackerman, and Amy Shlosberg. 2018. "Rape Myths in the Media: A
Content Analysis of Local Newspaper Reporting in the United States." Deviant Behavior
39 (9):1237-1246. doi: 10.1080/01639625.2017.1410608.
Saul, Stephanie. 2017. "When Campus Rapists Are Repeat Offenders." The New York Times,
January 24, 2017. Accessed October 17, 2020 in Factiva.
Sayej, Nadja. 2017. "Alyssa Milano on the #MeToo Movement: 'We're Not Going to Stand for It
Any More'." The Guardian, Last Modified December 1, 2017, accessed December 16,
2023. https://www.theguardian.com/culture/2017/dec/01/alyssa-milano-mee-too-sexualharassment-abuse.
Schiffer, Adam J. 2006. "Assessing Partisan Bias in Political News: The Case(s) of Local Senate
Election Coverage." Political Communication 23 (1):23-39. doi:
10.1080/10584600500476981.
Schiffer, Adam J. 2018. Evaluating Media Bias. Lanham, Maryland: Rowman & Littlefield.
Schneider, Christopher J., and Stacey Hannem. 2019. "Politicization of Sexual Misconduct as
Symbolic Annihilation: An Analysis of News Media Coverage of the 2016 'Rape
Election'." Sexuality & Culture 23 (3):737-759. doi: 10.1007/s12119-019-09587-6.
Schudson, Michael. 2002. "The News Media as Political Institutions." Annual Review of
Political Science 5:249-269. doi: 10.1146/annurev.polisci.5.111201.115816.
Schultz, Kai, and Rajneesh Bhandari. 2018. "Noted Humanitarian Charged With Child Rape in
Nepal, Stunning a Village." The New York Times, May 11, 2018. Accessed October 18,
2020 in Factiva.
Schultz, Vicki. 2018-2019. "Reconceptualizing Sexual Harassment, Again." The Yale Law
Journal Forum 128:22-66.
146
Semetko, H. A., and P. M. Valkenburg. 2000. "Framing European Politics: A Content Analysis
of Press and Television News." Journal of Communication 50 (2):93-109. doi:
10.1111/j.1460-2466.2000.tb02843.x.
Shoemaker, Pamela. 2017. "The Gatekeeping of Political Messages." In The Oxford Handbook
of Political Communication, edited by Kate Kenski and Kathleen Hall Jamieson. Oxford
University Press.
Shoemaker, Pamela J., and Stephen D. Reese. 2014. Mediating the Message in the 21st Century:
A Media Sociology Perspective. Third ed. New York: Routledge.
Shultziner, Doron, and Yelena Stukalin. 2021. "Distorting the News? The Mechanisms of
Partisan Media Bias and Its Effects on News Production." Political Behavior 43 (1):201-
222. doi: 10.1007/s11109-019-09551-y.
Siefkes-Andrew, Ashlie J., and Cassandra Alexopoulos. 2019. "Framing Blame in Sexual
Assault: An Analysis of Attribution in News Stories About Sexual Assault on College
Campuses." Violence Against Women 25 (6):743-762. doi: 10.1177/1077801218801111.
Skidmore, Paula. 1995. "Telling Tales: Media Power, Ideology and the Reporting of Child
Sexual Abuse in Britain." In Crime and the Media: The Post-Modern Spectacle, edited by
David Kidd-Hewitt and Richard Osborne, 78-106. London: Pluto Press.
Smith, Glen, and Kathleen Searles. 2013. "Fair and Balanced News or a Difference of Opinion?
Why Opinion Shows Matter for Media Effects." Political Research Quarterly 66
(3):671-684. doi: 10.1177/1065912912465922.
Soroka, Stuart N. 2003. "Media, Public Opinion, and Foreign Policy." The International Journal
of Press/Politics 8 (1):27-48. doi: 10.1177/1081180X02238783.
Steel, Emily, and Michael M. Grynbaum. 2018. "Suzanne Scott Named First Female Chief
Executive of Fox News." The New York Times, May 17, 2018. Accessed October 18,
2020 in Factiva.
Steel, Emily, and Michael S. Schmidt. 2017a. "Bill O’Reilly Is Forced Out at Fox News." New
York Times, April 19, 2017. Accessed October 17, 2020 in Factiva.
Steel, Emily, and Michael S. Schmidt. 2017b. "Bill O’Reilly Thrives at Fox News, Even as
Harassment Settlements Add Up." The New York Times, April 1, 2017. Accessed October
17, 2020 in Factiva.
Steel, Emily, Michael S. Schmidt, and Jim Rutenberg. 2017. "Fox’s Leaders Are Starting to Sour
on Bill O’Reilly." The New York Times, April 18, 2017. Accessed October 17, 2020 in
Factiva.
Stewart, Nikita, and Joseph Goldstein. 2016. "Long Island Abuse Case Reveals Risks of Out-ofState Foster Care." The New York Times, September 13, 2016. Accessed October 16,
2020 in Factiva.
147
Stolberg, Sheryl Gay, and Yamiche Alcindor. 2017. "Congress Returns to Intense Pressure to
End Secrecy Over Sex Harassment." The New York Times, November 26, 2017. Accessed
October 18, 2020 in Factiva.
Storer, Heather L., and Maria Rodriguez. 2020. "Mapping a Movement: Social Media, Feminist
Hashtags, and Movement Building in the Digital Age." Journal of Community Practice
28 (2):160-176. doi: 10.1080/10705422.2020.1757541.
Stroud, Natalie Jomini. 2008. "Media Use and Political Predispositions: Revisiting the Concept
of Selective Exposure." Political Behavior 30 (3):341-366. doi: 10.1007/s11109-007-
9050-9.
Sullivan, Kate, and Maegan Vazquez. 2022. "Biden Marks Reauthorization of Violence Against
Women Act." CNN, Last Modified March 16, 2022, accessed November 27, 2023.
https://www.cnn.com/2022/03/16/politics/biden-violence-against-women-act-whitehouse/index.html.
Sullivan, Margaret. 2022. "If American Democracy Is Going to Survive, the Media Must Make
This Crucial Shift." The Washington Post, Last Modified January 3, 2022, accessed April
4, 2024. https://www.washingtonpost.com/media/2022/01/03/media-democracy-jan6-
atlantic-npr/.
Swanson, Ana. 2018. "Trump Proposes Rejoining Trans-Pacific Partnership." New York Times,
April 12, 2018. Accessed October 17, 2020 in Factiva.
Swift, Art. 2017. "Six in 10 in U.S. See Partisan Bias in News Media." Gallup, Last Modified
April 5, 2017, accessed December 16, 2023. https://news.gallup.com/poll/207794/sixpartisan-bias-news-media.aspx.
Tackett, Michael. 2017. "Trump’s Combative Denials Again Draw Him Into the Sexual
Harassment Debate." The New York Times, December 12, 2017. Accessed October 18,
2020 in Factiva.
Tadena, Nathalie. 2016. "Martin Sorrell Acknowledges Sexism Is a Problem in the Ad Industry;
WPP CEO Addressed the Advertising Community for the First Time since the Head of
His JWT Agency Resigned." The Wall Street Journal, March 23, 2016. Accessed
October 20, 2020 in Factiva.
Tebben, Susan. 2018. "Finding A Voice: Newspaper Editors and the Effect of Sexual Assault
and Rape News." Media Report to Women 46 (4):6-22.
Terkildsen, Nayda, and Frauke Schnell. 1997. "How Media Frames Move Public Opinion: An
Analysis of the Women's Movement." Political Research Quarterly 50 (4):879-900. doi:
10.1177/106591299705000408.
The Wall Street Journal. n.d. "News. Opinion.". Dow Jones & Company, Inc., accessed April 5,
2024. https://newsliteracy.wsj.com/news-opinion/.
148
Thevenot, Maya. 2018. "Republican vs. Democrat CEOs: Who's Better for Shareholders?".
Florida Atlantic University College of Business, Last Modified July 3, 2018, accessed
November 12, 2023. https://business.fau.edu/faculty-research/stone-initiative/fau-stonepress-room/republicans-vs-democrat-ceos.php.
Tuchman, Gaye. 1972. "Objectivity as Strategic Ritual: An Examination of Newsmen's Notions
of Objectivity." The American Journal of Sociology 77 (4):660-679. doi:
10.1086/225193.
Tuchman, Gaye. 1973. "Making News by Doing Work: Routinizing the Unexpected." The
American Journal of Sociology 79 (1):110-131. doi: 10.1086/225510.
"Two Proportion Z-Test." n.d. StatsTest.com, accessed May 28, 2023.
https://www.statstest.com/two-proportion-z-test/.
Tyson, Vanessa. 2019. "Understanding the Personal Impact of Sexual Violence and Assault."
Journal of Women, Politics & Policy: Me Too Political Science 40 (1):174-183. doi:
10.1080/1554477X.2019.1565456.
"U.S. National Plan to End Gender-Based Violence: Strategies for Action." 2023. White House,
Last Modified May 2023, accessed November 27, 2023.
https://www.whitehouse.gov/wp-content/uploads/2023/05/National-Plan-to-EndGBV.pdf.
United Nations. n.d. "World Press Freedom Day: Background." United Nations, accessed
December 16, 2023. https://www.un.org/en/observances/press-freedom-day/background.
UVA Center for Politics. 2021. "New Initiative Explores Deep, Persistent Divides Between
Biden and Trump Voters." UVA Center for Politics, Last Modified September 30, 2021,
accessed December 16, 2023. https://centerforpolitics.org/crystalball/articles/newinitiative-explores-deep-persistent-divides-between-biden-and-trump-voters/.
Van Deinse, Tonya B., Rebecca J. Macy, Gary S. Cuddeback, and Amanda J. Allman. 2019.
"Intimate Partner Violence and Sexual Assault among Women with Serious Mental
Illness: A Review of Prevalence and Risk Factors." Journal of Social Work 19 (6):789-
828. doi: 10.1177/1468017318766425.
van der Linden, Sander, and Costas Panagopoulos. 2019. "The O'Reilly Factor: An Ideological
Bias in Judgments about Sexual Harassment." Personality and Individual Differences
139:198-201. doi: 10.1016/j.paid.2018.11.022.
Vliegenthart, Rens, and Stefaan Walgrave. 2012. "The Interdependency of Mass Media and
Social Movements." In The SAGE Handbook of Political Communication, edited by Holli
A. Semetko and Margaret Scammell, 387-398. London: SAGE Publications Ltd.
Walker, Mason, and Jeffrey Gottfried. 2020. "Americans Blame Unfair News Coverage on
Media Outlets, Not the Journalists Who Work for Them." Pew Research Center, Last
Modified October 28, 2020, accessed December 23, 2023.
149
https://www.pewresearch.org/short-reads/2020/10/28/americans-blame-unfair-newscoverage-on-media-outlets-not-the-journalists-who-work-for-them/.
Watts, Mark D., David Domke, Dhavan V. Shah, and David P. Fan. 1999. "Elite Cues and Media
Bias in Presidential Campaigns: Explaining Public Perceptions of a Liberal Press."
Communication Research 26 (2):144-175. doi: 10.1177/009365099026002003.
Weiss, Alexander, and Steven M. Chermak. 1998. "The News Value of African-American
Victims: An Examination of the Media’s Presentation of Homicide." Journal of Crime &
Justice 21 (2):71-88. doi: 10.1080/0735648X.1998.9721601.
White, Kailey, Forrest Stuart, and Shannon L. Morrissey. 2021. "Whose Lives Matter? Race,
Space, and the Devaluation of Homicide Victims in Minority Communities." Sociology
of Race and Ethnicity 7 (3):333-349. doi: 10.1177/2332649220948184.
Wiest, Sara L., Leigh Raymond, and Rosalee A. Clawson. 2015. "Framing, Partisan
Predispositions, and Public Opinion on Climate Change." Global Environmental Change
31:187-198. doi: 10.1016/j.gloenvcha.2014.12.006.
Wolfe, Lauren, and Stacia L. Brown. 2015. "Writing Rape: How U.S. Media Cover Campus
Rape and Sexual Assault." Women's Media Center, Last Modified December 15, 2015,
accessed August 15, 2020. https://womensmediacenter.com/reports/writing-rape-how-us-media-cover-campus-rape-and-sexual-assault.
Yang, Goubin. 2016. "Narrative Agency in Hashtag Activism: The Case of #BlackLivesMatter."
Media and Communication 4 (4):13-17. doi: 10.17645/mac.v4i4.692.
Young, Dannagal G. 2017. "Stop Covering Politics as a Game." Nieman Lab, Last Modified
December 2017, accessed December 4, 2023. https://www.niemanlab.org/2017/12/stopcovering-politics-as-a-game/.
Zounlome, Nelson O. O., Y. Joel Wong, Elyssa M. Klann, Jessica L. David, and Nat J. Stephens.
2019. "‘No One . . . Saves Black Girls’: Black University Women’s Understanding of
Sexual Violence." The Counseling Psychologist 47 (6):873-908. doi:
10.1177/0011000019893654.
150
Appendix A
Methods
Pilot Coding
Pilot coding was performed on every 25th article to select the categories to be coded for
newsworthiness and framing in this study. There was a total of 206 articles reviewed but some
were coded as irrelevant. In total, there were 92 articles coded in the NYT and 47 in the WSJ for
a total of 139.
News Values
There are several taxonomies of news values in the literature, and I compared four of
them to develop the list used for pilot coding. The taxonomies from Harcup and O’Neill (2017)
and Graber and Dunaway (2015) apply broadly to all issues in the news and the taxonomies used
by Greer (2003) and Cheit, Stavit, and Reiss-Davis (2010) were used in studies specifically
analyzing news coverage of SV. Below is a detailed list of each taxonomy, followed by a
comparison table (see Table 8), and the results of pilot coding which narrowed down the news
values used in this study.
Harcup and O’Neill (2017) identified 15 news values and argue that any potential news
story has to include at least one of the following (1482, italics in the original):
1. Exclusivity: Stories generated by, or available first to, the news
organization as a result of interviews, letters, investigations, surveys, polls,
and so on.
2. Bad news: Stories with particularly negative overtones such as death,
injury, defeat and loss (of a job, for example).
3. Conflict: Stories concerning conflict such as controversies, arguments,
splits, strikes, fights, insurrections and warfare.
4. Surprise: Stories that have an element of surprise, contrast and/or the
unusual about them.
5. Audio-visuals: Stories that have arresting photographs, video, audio and/or
which can be illustrated with infographics.
6. Shareability: Stories that are thought likely to generate sharing and
comments via Facebook, Twitter and other forms of social media.
151
7. Entertainment: Soft stories concerning sex, showbusiness, sport, lighter
human interest, animals, or offering opportunities for humorous treatment,
witty headlines or lists.
8. Drama: Stories concerning an unfolding drama such as escapes, accidents,
searches, sieges, rescues, battles or court cases.
9. Follow-up: Stories about subjects already in the news.
10. The power elite: Stories concerning powerful individuals, organizations,
institutions or corporations.
11. Relevance: Stories about groups or nations perceived to be influential
with, or culturally or historically familiar to, the audience.
12. Magnitude: Stories perceived as sufficiently significant in the large
numbers of people involved or in potential impact, or involving a degree
of extreme behavior or extreme occurrence.
13. Celebrity: Stories concerning people who are already famous.
14. Good news: Stories with particularly positive overtones such as recoveries,
breakthroughs, cures, wins and celebrations.
15. News organization’s agenda: Stories that set or fit the news organization’s
own agenda, whether ideological, commercial or as part of a specific
campaign.
Graber and Dunaway (2015) identify five criteria of newsworthiness:
1. Strong impact – the story has a strong impact on the audience
2. Violence, conflict, disaster, or scandal – examples include, wars, murders, natural
disasters and sex scandals
3. Familiarity – the person or situation is familiar to the audience
4. Proximity – people are most interested in what happens near them, so they pay
closer attention to local news than to national or international news
5. Timely and novel – something which just occurred and is not a part of people’s
everyday lives
Greer (2003) examined news coverage of sex crimes in Northern Ireland and identified 11
criteria for newsworthiness:
1. Speculation – the story can make the reader think “this could happen to me or
someone I care about”
2. Immediacy – crime just occurred or was revealed; warning that the sex offender
could strike again at any time
3. Personalization – the person or people involved are notable individuals of high
status
4. Level of public interest – person or people involved are elected officials or in a
position of authority over children
152
5. Individual pathology – the pathology of the offender marks them as distinct from
the rest of society
6. Seriousness – seriousness of the offense; harm done to the victim(s)
7. Dramatization – action and violence
8. Titillation –revealing the forbidden/voyeurism
9. Spatial proximity – closeness to the reader in geographical terms
10. Cultural proximity – closeness to the reader in cultural terms; the extent to which
an event “resonates within a news reader’s existing framework of values,
interests, beliefs and concerns” (48)
11. Novelty – rare, extraordinary event; angle to a story which sets it apart from the
norm
Table 9. Comparison of Taxonomies of News Values
Harcup and O’Neill
(2017)
Graber and
Dunaway (2015)
Greer (2003) Cheit, Stavit, and
Reiss-Davis (2010)
Magnitude Strong impact Multiple parties
Conflict
Drama
Violence, conflict,
disaster, or scandal
Dramatization
Seriousness
Extra violence
The power elite
Relevance
Celebrity
Familiarity Cultural proximity
Personalization
Level of public
interest
Upstanding accused
Celebrity status
Surprise Timely and novel Novelty
Immediacy
Titillation
Bizarre facts
Proximity Spatial proximity
No corollary with
other taxonomies:
Exclusivity
Bad news
Audio-visuals
Shareability
Entertainment
Follow-up
Good news
News organization’s
agenda
No corollary with
other taxonomies:
Speculation
Individual pathology
No corollary with
other taxonomies:
Cover up
153
In their study of child sex abuse, Cheit, Stavit, and Reiss-Davis (2010) examined six news values
(105, italics mine):
1. Upstanding accused – The accused is an upstanding member of his or her
community (e.g., a politician, teacher, or coach)
2. Extra violence – More violence than normal in child sexual abuse stories is
present
3. Bizarre facts – There are especially strange or memorable facts in the story (e.g.,
Satan worship)
4. Multiple parties – There is more than one victim, accused, or both
5. Cover-up – There is an element of a cover-up to the story
6. Celebrity status – The victim, the accused, or both are celebrities for something
other than the case (e.g., Michael Jackson)
Based on this table, I developed categories to capture at least one of the news values in
the first 5 rows of this table. I also came up with two categories of my own: legal system and
government action. Legal system is similar to drama and conflict (which usually involves court
cases), but since so much of SV involves civil and criminal cases (Greer 2003; Hersch and
Moran 2013), I wanted to use a category which would allow me to separate out how much the
legal system is driving coverage. Government action is similar to Harcup and O’Neill’s power
elite, but since journalists rely so much on official government sources from their beats to
generate news content (Bennett 2016), I used this term to be able to demonstrate how much news
is driven by actions taken by the government. Instead of celebrity, I used the term rich/famous to
capture a broader array of individuals.
There were several categories of news values from the comparison table that I did not use
based on the results from pilot coding. One was novelty/surprise/bizarre facts. This was
described in previous literature as an unusual or rare event (Greer 2003, Harcup and O’Neill
2017), something which is not part of people’s everyday lives (Graber and Dunaway 2015), and
strange or memorable facts (Cheit, Stavit, and Reiss-Davis 2010). During pilot coding I realized
that this variable was difficult to operationalize based on these definitions. For example, what
154
constitutes surprising or memorable facts in the context of SV? Unfortunately, as #MeToo made
quite clear, SV is not an unusual or rare event. Even if people were surprised by the Harvey
Weinstein story, once hundreds of powerful men were identified as offenders, at what point was
it no longer surprising? For these reasons I decided to exclude this news value. Furthermore, as
Greer (2003) argues, novelty has the most limited lifespan of any of the news values. As a
journalist explained in an interview with Greer, rape dominated the headlines in the 1970s, but
child sex abuse displaced it in the 1980s and 1990s, so much so that the rape has to be quite
unusual to get a headline in the current news environment.
Spatial proximity is the geographical closeness to the reader or news outlet. During pilot
coding, the location could only be determined in 58 percent of the articles examined. Upon
further reflection of spatial proximity as a news value, since so many #MeToo stories involved
the news and entertainment industry which are concentrated in New York City and Los Angeles,
I determined that coding location of where it occurred (e.g., a film festival in Cannes) would not
provide any useful insight into why the story was selected for publication. Furthermore, both the
NYT and WSJ are national newspapers with a national readership which would make drawing
conclusions from spatial proximity data problematic.
Exclusivity was described by Harcup and O’Neill (2017) as news articles “generated by,
or available first to, the news organization as a result of interviews, letters, investigations,
surveys, polls, and so on” (1482). This variable was difficult to operationalize for coding and
replicability purposes because it was not always clear that the newspaper was the first to publish
the story (many #MeToo stories were about well-known individuals and were covered by several
news outlets) so this news value was not included in the analysis.
155
Bad news was defined by Harcup and O’Neill (2017) as news articles “with particularly
negative overtones such as death, injury, defeat and loss of a job, for example)” (1482). While
this is a good variable for categorizing news in general, all the news articles in this study involve
illegal behavior so all (or virtually) all stories would have been coded as bad news. For this
reason, it was excluded from the final codebook.
Audio-visuals are news stories which “have arresting photographs, video, audio and/or
which can be illustrated with infographics” (Harcup and O’Neill 2017, 1482). Since the vast
majority of articles downloaded only contained text, this news value could not be coded.
Shareability was described by Harcup and O’Neill (2017) as news articles which make
the reader laugh or cry and would generate sharing and comments on social media platforms like
Facebook and Twitter. While some stories are obviously funny or tragic, this would be difficult
to operationalize in the context of SV because there is no way to reliably predict the emotional
response of a reader. Furthermore, since the articles downloaded from the database did not
include how many times a news article was shared or liked, it was not possible to gather this data
objectively.
Entertainment is described as soft news “concerning sex, showbusiness, sport, lighter
human interest, animals, or offering opportunities for humorous treatment, witty headlines or
lists” and good news is defined as stories with “positive overtones such as recoveries,
breakthroughs, cures, wins and celebrations” (Harcup and O’Neill 2017, 1482). Since all these
news articles concern SV, both of these news values would not be relevant and therefore were
excluded.
Follow-up is defined by Harcup and O’Neill (2017) as news articles “about subjects
already in the news” (1482). This news value would be difficult to code in my study because the
156
news articles are all about the same subject. Furthermore, since the news articles were randomly
sampled, it was often difficult to know if the story was a follow up to something previously
published.
Harcup and O’Neill (2017) described a news organization’s agenda as stories
which “set or fit the news organization’s own agenda, whether ideological, commercial or
as part of a specific campaign” (1482). None of the stories during pilot coding fit this
description and it was excluded.
Greer (2003) defined individual pathology as something that marks the offender as
distinct from the rest of society (e.g., describing the offender as a monster). The pilot coding
revealed that this category would not yield any useful results, so it was excluded.
The speculation category from Greer (2003), which was defined as a story which makes a
reader think it could happen to them or someone they care about, was so infrequent (only two
articles had this news value), that it was excluded from the final analysis.
Lastly, during pilot coding I noticed that several news stories seemed to be motivated by
a report from a non-governmental organization such as the United Nations or Human Rights
Watch. However, upon further reflection, as a category for a news value it did not make sense
because it would not be generalizable beyond the present study. Since these stories were also
captured by the familiarity news value, I decided to eliminate the NGO category from the study.
Framing
The issue-specific frames come from Bing and Lombardo (1997) and Schneider and Hannem
(2019), and the generic frames are adapted from Semetko and Valkenburg (2000). Bing and
Lombardo (1997) analyzed news coverage of sexual harassment and identified four issuespecific frames:
157
1. Judicial – comparing behavior to predetermined legal policies or statutes; used legal
terms to discuss issues like guilt, evidence, and credibility of witnesses.
2. Victim – the focus is on the experience of harassment with a primary emphasis on the
harm to the victim.13
3. Initiator – focuses on behavior of perpetrator as opposed to harm to victim; responsibility
may be shifted to the target (“she misunderstood”) and justifications may be provided
(“boys will be boys”). Often admit behavior occurred but redefine it as a
misunderstanding or misperception.
4. Social science – articles in this frame tend to discuss the issue in a general and abstract
way and avoid the specifics of an individual case.
Since Bing and Lombardo only focused on sexual harassment, I adapted the judicial frame to
include criminal and civil cases, government investigations, internal investigations, and
legislation and renamed it “legal.”
The pilot coding of these four frames revealed the initiator and social science frames
were rarely used as a primary frame and were therefore dropped from the codebook.14
Since the
legal frame accounted for half the articles, the category was disaggregated into the following
subcategories: criminal, civil, civil and criminal, unclear civil / criminal, government
hearing/investigation, internal investigation, civil and internal, and legislative. This would allow
the categories to be analyzed separately or they could be added back together in the legal frame.
In their analysis of news coverage of the sexual misconduct allegations against Donald
Trump during the 2016 U.S. presidential campaign, Schneider and Hannem (2019) found that the
issue was “framed in partisan terms and as, foremost, a political issue” which they referred to as
the politicization of sexual misconduct (737). During pilot coding, there were articles which
framed the issue of SV as a political problem, so I created a political frame. News articles which
used the political frame discussed the political implications of allegations, (e.g., impact on a
13 The victim frame is similar to the human-interest frame from Semetko and Valkenburg (2000) and the human
impact frame from Neuman, Just, and Crigler (1992).
14 There were only two articles which met the initiator criteria and one article which met the social science criteria.
158
campaign, how allegations influenced voters), the political strategy for dealing with allegations
including how political parties were handling allegations differently, the offender evading
political consequences (e.g., Trump), and politicians resigning or not seeking re-election.
Of the five frames identified by Semetko and Valkenburg (2000), two were adapted for
use in this study: conflict and economic consequences. While conflict typically includes court
cases, as with the news values, since so much of SV involves litigation I wanted to have a
separate category to represent these articles. In this study, the conflict frame is defined as one
which frame focused on scandal or controversy (e.g., a victim, most often a woman, coming
forward and revealing what happened to her, or the reaction or aftermath of revelations) and acts
of SV which occurred in conflict zones. Examples of articles from the pilot sample include
criticism over Pope Francis’ handling of child sex abuse, an actor not presenting an award at the
Oscars because of allegations against him, and criticism of a political party in New Zealand for
their lack of response to assault allegations at a youth summer camp.
Semetko and Valkenburg (2000) defined the economic consequences frame as one which
highlighted the economic impact of a problem. During pilot coding, there were several articles
which hinted at economic consequences, such as the CEO resigning or being terminated, or a
successor or interim replacement being named, but did not explicitly discuss it so I created the
business frame. This frame captured how allegations of SV were affecting the business
economically (e.g., stock prices, value of the company) as well as other important events which
can impact a company’s future (e.g., resignations, terminations, successors being named).
As for the other three frames by Semetko and Valkenburg (2000) not used in this study,
the human-interest frame, which presents a problem with an emotional angle, is similar to the
victim frame from Bing and Lombardo (1997). The responsibility frame, which attributes
159
responsibility for solving a problem to either government or individuals/groups, is captured
through episodic and thematic coding as well as solutions. The morality frame, which places a
problem, event, or issue in the context of religious tenets or morality, was not relevant for this
issue.
160
Appendix B
Codebook
Q1
Article Number [enter number]
Q2
Newspaper (check one):
□ New York Times [1]
□ Wall Street Journal [2]
Q3
Choose which answer best describes this article:
□ Relevant [1]
□ Not relevant (e.g., not a news article, “rape” used metaphorically, not written by a
reporter (i.e., columnists or expert); text entry available) [2]
□ Mention SV but no details (appears in one paragraph; text entry available) [3]
□ Other (SV discussed in less than half the article; text entry available) [4]
If “Not relevant,” “Mention SV,” or “Other,” was selected, no further coding was performed on
the article.
Q4
Does the story primarily take place inside the U.S. or a foreign country?15 (check one)
□ Domestic [1]
□ Foreign [2]
□ Unclear [3]
Q5
Is the article written by at least one female journalist? (At least one name in the byline is
female.) (Check one):
□ Yes [1]
□ No [2]
□ Unknown author [3]
□ The Associated Press [4]
□ Unclear (with text entry) [5]
15 Note: A Senate hearing about SV and UN Peacekeeping troops would be considered domestic because the action
is occurring within the U.S.
161
Q6
News values (check all that apply):
□ Magnitude (multiple offenders or victims, or large numbers of people involved) [1]
□ Legal system (civil or criminal proceedings, law enforcement) [2]
□ Immediacy (just occurred or was revealed, includes announcements; offender at-large)
[3]
□ Rich and/or famous (e.g., wealthy people, entertainers, professional athletes, person has
name recognition with the public)16 [4]
□ Familiarity – the person, organization, or institution is familiar to the audience17 [5]
□ Conflict (war, arguments, controversies) [6]
□ Cover up (element of cover up to the story) [7]
□ Speculation (random act of violence by a stranger, can happen to anyone) [8]
□ Government action (e.g., statements, press conference, announcement, hearings,
legislation, reports, can be agencies or officials, including elected politicians)18 [9]
□ NGOs (officials or agencies; examples include announcements or reports from groups
like the U.N., Amnesty International, or Human Rights Watch) [10]
□ Other (with text entry) [11]
Q7
Did the article have multiple victims of SV?19 (Check one)
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
Q8
Did the article have multiple offenders SV?20 (Check one)
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
16 Rich and famous people could also be coded as familiar.
17 Organizations can include companies (e.g., CBS, Uber) or sports teams and within that category includes
familiarity with their products (e.g., a movie by a production company); institutions can include government bodies
(e.g., state legislature, Congress) or agencies (e.g., Border Patrol).
18
Government action can be foreign or domestic and includes law enforcement and the courts.
19 Multiple victims mean more than one victim of the same offender or at the same organization (e.g., university,
Catholic Church). This is for the primary focus of the story; other examples used for context are not coded
(otherwise every article which referred to Weinstein or Cosby could be coded as multiple victims).
20 Multiple offenders – more than one offender involved in an incident (e.g., gang rape), or at same organization
(university, Catholic Church).
162
Q9 – Episodic / thematic21
Type Value Description
Episodic 1 Presents the issue by discussing a specific incident or individual, or an eventoriented report. (Example: U.N. report about South Sudan.)
An episodic article focuses on the experience of a particular victim, offender, or
organization.
It will not include any context about the systemic nature of the issue by citing
additional examples, nor will it provide any statistical context.
Primarily
Episodic
2 Mostly focused on the victim or victims of a single offender or an incident with
multiple offenders, but provides some context (a few sentences or a few short
paragraphs) that illustrates something broader than the experience of the
personalized example.
The context can include citing additional examples (e.g., references to Weinstein
or Cosby), gender inequality, or some statistical data.
Mixed 3 This category includes articles about the issue at an organization (e.g., Catholic
Church, a school) or in an industry and a single offender or victim is not the
primary focus.
Primarily
Thematic
4 Mostly presents the issue by using contextual, historical, or statistical information,
but uses specific examples. Includes articles which discuss the issue as a weapon
of war (e.g., mass rapes in conflict zones).22
Thematic 5 Presents an issue by using contextual, historical, or statistical information.
Does not use any specific examples (victim or offender) to discuss the issue.
21 Definitions were adapted from Barnes and Hevron (2018)
22 Mass rapes were coded as primarily thematic because it is not being presented as an individual aberration.
163
Q10
Primary frame of the article [check one]:
Frame Value Description
Criminal 1 Initial report, investigation by law enforcement (includes reference to
the possibility of an investigation), court (charges, plea agreement,
trial, sentencing), appeals, release from prison (parole or completed
sentence), statements by district attorney defending decision not to
bring charges. Can include international cases and cases in U.S.
military courts.
Civil 2 Civil lawsuit (including the intention to file a lawsuit), trial,
disposition of case (dismissal of case, settlement, verdict, damages);
concerns over liability; mediation, arbitration, and victim
compensation funds. Also includes DOJ civil rights investigations.
Criminal and
civil
3 There are elements of the criminal and civil categories in about equal
parts.
Unclear if
civil or
criminal
4 It is unclear if it is a civil or criminal matter. (Example: DOJ sending
preserve orders to the Catholic Church.)
Government
investigation /
hearing
5 Investigation by a state or federal agency that is not law enforcement
(e.g., Department of Education, EEOC; DOJ and state attorneys
general are excluded – they can conduct civil and criminal
investigations); hearings by a government body (e.g., nomination to
U.S. Supreme Court).
Internal 6 Organization conducts its own internal investigation into allegations
(includes announcement that there will be an investigation and report
issued after investigation is complete). Discussion of procedures for
conducting investigations. Actions taken as result of an investigation
(person disciplined or terminated, new policies implemented).
Criticism of the way an organization handled complaints.
Announcement of new policy. The organization can be a company,
non-profit, or government agency. Survey results or report from an
organization.
Civil and
internal
7 There are elements of the civil and internal investigation categories
in about equal parts.
Legislative 8 Stories about proposed, recommended, or current legislation, both
criminal or civil; includes criticism of current or proposed laws.
Examples include a change to the statute of limitations, closing a
loophole in current law, and enacting a new sentencing law.
164
Victim 9 Focus of article is on victims and their experience with SV.
Business 10 Relates to the business impact: stock prices, revenue, value of the
company, successor or interim replacement being named,
implications of CEO resigning or being terminated.
Political 11 Includes political implications of allegations, (e.g., impact on a
campaign, how allegations influenced voters), political strategy for
dealing with allegations including how political parties are handling
allegations differently, offender evading political consequences (e.g.,
Trump), politician resigning or not seeking re-election.
Conflict 12 Scandal, controversy (e.g., a victim, most often a woman, coming
forward and revealing what happened to her; the reaction or
aftermath of revelations) and acts of SV which occurred in conflict
zones
Public health 13 Prevention and education
Other 14 (with text entry)
Q11
Is SV discussed as a human rights issue? (Can include discussion gender inequality and impacts
on women as a group.) [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
Q12
Tone towards victim(s) [check one]:
□ Positive [1]
□ Mostly positive [2]
□ Negative [3]
□ Mostly negative [4]
□ Mixed (equal positive and negative) [5]
□ Neutral [6]
□ N/A (victim(s) not discussed) [7]
Examples of positive tone towards victims:
• Sympathetic portrayal of victim
• Specific description of behavior which harmed victim
• Harm to victim (short and/or long-term)
• Abuse of power (offender had power over victim)
165
• Victim describes his or her own experience (includes not only their experience with SV,
but also with the organization (e.g., workplace, school), and the criminal legal system.
• Support for victims (march, rally, or demonstration to support victims, a policy solution,
or raise awareness).
Examples of negative tone towards victims:
• Victim blaming
o An example includes drinking, but only if presented as if she was at least partly to
blame for what happened to her.
o Offender said it was a misunderstanding or behavior was misinterpreted (puts
responsibility back on her)
• She’s too sensitive (e.g., if you can’t stand the heat get out of the board room)
• Behavior made women uncomfortable (can tie into sexual stereotypes about women
being overly sensitive)
• Offender saying allegations are false or denying nonconsensual sex (implication is that
the victim is lying)
• Victim has an ulterior motive for coming forward; the victim is motivated by political or
financial reasons, she is trying to destroy his career
• Questioning the character and/or credibility of the victims (usually done by the offender
or defense attorney and may be tied to claim of false allegations)
Q13
Tone towards offender(s) [check one]:
□ Positive [1]
□ Mostly positive [2]
□ Negative [3]
□ Mostly negative [4]
□ Mixed (equal positive and negative) [5]
□ Neutral [6]
□ N/A (offender(s) not discussed) [7]
Examples of positive tone towards offenders:
• Favorable or sympathetic descriptions/portrayal of offender
o Examples include the offender being described as someone who transformed the
industry, Weinstein being described as someone who “has seen his company,
career, and reputation implode” (article 4325), describing Roman Polanski as a
Holocaust survivor (article 250), or a company who justifies their decision to hire
an offender by saying he learned his lesson and he assured them he would behave
in a professional manner (article 2725).
• How important offender is to company (e.g., Roger Ailes and Fox News)
• Defend the offender (e.g., fans of Michael Jackson at premiere of documentary)
• Offender (or his attorney/spokesperson) says he is hurt by allegations
166
• Offender apology if he takes responsibility and it is not a non-apology (an example of a
non-apology is “I apologize if I hurt someone”) and it does not add any “buts” to the
apology such as “but I never harmed a woman’s career.”
• Justifications provided (boys will be boys, locker room talk – these suggest the behavior
is normal).
Examples of negative tone towards offenders:
• Accused by one or more people of SV
• Offender presented as a bully, charged or convicted of a crime, irresponsible (offender
who also tolerated hostile work environment)
• Offender is responsible for harm to victims
NOTE: In episodic articles, only the primary victim(s) and offender(s) are coded for tone; in
mixed and thematic articles, all are coded.23
The N/A option in the tone variable was included for articles which do not refer to a specific
case, do not mention victim(s) or the victim is dead, and/or offender(s) not yet identified.
Q14
Is the ideology or political party of the offender mentioned? (Only coded for domestic articles,
N/A if not domestic; if a former or current POTUS is the offender, ideology or party does not
need to be explicitly mentioned to be coded because it is so well known.) [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A [4]
If answer to Q14 is “Yes”, answer Q15, otherwise skip to Q16.
Q15
Party or ideology of offender(s) [check one]:
□ GOP / conservative [1]
□ Democrat / liberal [2]
□ Both [3]
□ Other (with optional text entry) [4]
23 Sometimes more than one offender in an episodic article was discussed equally so all were coded.
167
Q16
Does the offender (or his representative) deny nonconsensual sex or touching? (Only applies to
articles in which a sexual contact took place; if not, code as N/A.) [check one]:
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A [4]
Q17
Does the offender apologize? [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A [4]
Q18
Does article include statements that the process is unfair to the offender or the offender is being
denied due process? (Includes references to being framed, a witch hunt.) [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A [4]
Q19
Is victim praised for coming forward and/or portrayed as courageous? [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A [4]
Q20
Does the article include example(s) of an abuse of power, due to either the offender’s position or
age? [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A [4]
168
Q21
Is the victim criticized for a delay in reporting? [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A [4]
Q22
Is victim accused of having an ulterior motive? (Examples of ulterior motive could be to gain
leverage in a dispute, get money in a settlement, harm someone’s reputation or hurt a politician’s
chances for election/re-election.) [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A [4]
Q23
Does this article discuss sexual harassment? [check one]
□ Yes [1]
□ No [2]
If yes, answer question 24; if no, skip to question 26.
Q24
Are nonsexual forms of harassment discussed? [check one]
□ Yes [1]
□ No [2]
If yes, answer question 25, if no, skip to question 26.
Q25
Nonsexual forms of harassment discussed [check all that apply]24:
□ Hostile behavior, personal ridicule, intimidation, and inappropriate comments or jokes
(includes yelling, bullying, being berated, insults, inappropriate comments about
someone’s body or appearance, gender-based or sexual orientation-based slurs) [1]
□ Patronizing treatment (examples include being asked to perform “gendered roles” like
cleaning up after a meeting, “women are all…” “all gay people are…”), saying “calm
down, dear” to a female lawmaker. [2]
□ Social ostracism, exclusion, and marginalization (e.g., gossiping about someone’s
personal relationships or sex life, not invited on work outings, excluded from meetings)
[3]
□ Denial of information and work sabotage (e.g., failing to share crucial job information)
[4]
24 Categories for nonsexual harassment came from Schultz 2018-2019, Equality Rights Advocates.
169
□ Display of offensive materials (non-sexual in nature) [5]
□ Leering / stalking [6]
□ Blocking someone’s movement [7]
□ Other (with text entry) [8]
Q26
Was specific language used to describe the offensive behavior? [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A (no specific incident described) [4]
Examples of specific language:25
• Use of the word “force,” “forced,” or “forcibly” along with action (e.g., forcibly kissed,
forced oral sex) or other language which makes it clear it was against their will (coerced,
against her will)
• “Pushed his fingers inside her vagina”
• Rape – in active voice which makes it clear the offender committed act. “Occupy
Cleveland Protester Alleges She Was Raped” becomes “Occupy Cleveland Protester
Reports a Man Raped Her” – the first headline makes the offender invisible. (Example
from Chicago Taskforce Media Toolkit)
• “[he] had unzipped the young man’s pants and rubbed his penis for about three minutes”
• Grabbed her crotch, breast, buttocks
• Unwanted workplace hugging
• “Admitted to masturbating in front of colleagues”
• Stared at her breasts
• Exposed himself
• Pinned her against the wall
• Repeatedly showed law clerk pornography and asked if it turned her on
Examples of vague language:26
• Groped/groping (not specific enough – better to describe exactly what offender did like
grabbed her crotch)
• Sexual harassment
• Sexual abuse
• Sexual misconduct
• Sex trafficking
• Sexual exploitation
• Unwanted touching, inappropriate touching
25 Examples came from pilot coding.
26 Examples came from The Crime Report article, Maine Coalition Against Sexual Assault, Chicago Taskforce, and
pilot coding.
170
• Inappropriate behavior, sexual misbehavior, dehumanizing behavior, sexual improprieties
• Unwanted sexual advances
• Verbal abuse, lewd comments/remarks, inappropriate comments
• Stalking
• Engaging in
• Sexual activity/assault/molest
• Fondle
• Performed oral sex
• Oral sex, anal sex, anal intercourse
Q27
Was harm to victims discussed? [check one]
□ Yes [1]
□ No [2]
□ N/A (victims not discussed) [3]
If yes answer question 28; if no skip to question 29.
Q28
Harm to victims [check all that apply]:27
□ Retaliation (harm career prospects/adversely affect career, loss of advancement
opportunities/deny promotion, loss of job, demotion, defamation, financial loss or fear of
financial loss) [1]
□ Fear of retaliation or threatened with retaliation or physical harm [2]
□ Quit a job [3]
□ Sought new job assignment or began looking for a new job [4]
□ Avoided working there (e.g., because of offender’s reputation) [5]
□ Physical injuries (short and long-term) [6]
□ Psychological (self-blame, depression, anxiety, PTSD, eating disorders, etc.) [7]28
□ Suicide / attempted suicide [8]
□ Substance abuse (drugs, alcohol, prescription) [9]
□ Stopped a hobby or activity, changed route or routine [10]
□ Changed schools, changed advisor, change major/research focus, grades suffered,
dropped out of school, dropped a course, did not complete training (job or volunteer
work) [11]
□ Moved from a dorm, apartment, house or other form of residence [12]
□ Murdered [13]
□ Other (with text entry) [14]
27 The harm to victim categories were developed from Kearl 2018, Basile et al. 2016, and pilot coding.
28 The psychological category includes references to trauma and conditions like depression or PTSD, but not generic
references to harm.
171
Q29
Was harm to friends, family members, witnesses and/or other employees discussed in the
article?29 [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
□ N/A (no specific incident discussed) [4]
Q30
Are solutions discussed in the article? [check one]
□ Yes [1]
□ No [2]
□ Unclear (with optional text entry) [3]
Possible solutions included creating or implementing new policies and/or procedures, education
or training, legislation (e.g., changes to the statute of limitations), hiring and promoting more
women, men as allies, men positive masculinity, programs or policies to empower women and
girls.
Q31
Sources [check all that apply and number of sources in each category]:30
□ Victim(s) [1]
□ Family, friends, and colleagues of victim(s) [2]
□ Offender(s) [3]
□ Family, friends, and colleagues of offender(s) [4]
□ Family, friends, and colleagues of victim(s) and offender(s) [5]
□ Witness(es) [6]
□ Police (or other criminal investigative agency like the FBI, can include reports, could be
foreign or domestic) [7]
□ Prosecutor (can include indictments or motions, could be foreign or domestic) [8]
□ Victim’s attorney or spokesperson (can include legal documents) [9]
□ Offender’s attorney or spokesperson (can include legal documents) [10]
□ Other attorneys (e.g., those hired to perform internal investigation; includes reports but
excludes lawyers used as experts for comment) [11]
□ Court (includes judges, court rulings, references to court documents and records, jury
verdicts, grand jury reports) [12]
29 Harm in this context includes those used for victims – like fear of retaliation – but uses a broader definition as it
can refer to groups of people. For example, harm can refer to anger over how the company has handled the issue.
The purpose of this question is to demonstrate that the offender’s behavior harms more than the victim.
30 Sources did not need to be directly quoted (i.e., police said, court filings revealed, the victim told police). Sources
were only coded under one category, so if a source qualified under two categories – Bill Cosby is both an offender
and an entertainer – then the category most relevant to the news story would be coded (in this case Cosby would be
coded as an offender).
172
□ Politicians (or their spokespeople, campaign; past or present; domestic) – DEMOCRAT
[13]
□ Politicians (or their spokespeople, campaign; past or present; domestic) – REPUBLICAN
[14]
□ Politicians (or their spokespeople, campaign; past or present; domestic) – NO PARTY
IDENTIFIED [15]
□ Political Party (domestic, DNC, RNC, includes spokespeople and party officials, local,
state, or national level) [16]
□ Foreign governments (politicians, officials, government agencies, military, reports;
excludes police and prosecutors) [17]
□ For-Profit Organization (e.g., CEO, Board of Directors, company spokesperson,
organization’s attorney; can include reports) [18]
□ Nonprofit/NGO (e.g., advocacy groups, U.N. – includes people and reports) [19]
□ Right-wing groups (e.g., nationalist or white supremacy groups; not a political party) [20]
□ Religious organization (includes leaders, spokespeople, reports) [21]
□ Academic / expert (includes studies, think tanks and consultants, but not those affiliated
with an NGO or other nonprofit like Human Rights Watch) [22]
□ Academic institution (includes administrators and spokespeople, K-12, college and
universities) [23]
□ Government (domestic only– includes people, reports, public records, legislative bodies
and committees, and transition teams (e.g., EEOC, Department of Education, House
Ethics Committee; does not include police or prosecutors or former staff) [24]
□ News media (includes journalists, editors, and other news outlets) [25]
□ Entertainers and professional athletes [26]
□ Public [27]
□ Unnamed sources [28]
□ Opinion polls [29]
□ Other (numeric entry) [30]
□ Other (text entry) [31]
New categories created from “other” category:
□ Current and/or former employees (includes executives)
□ Other entertainment – behind the scenes, creative professionals and representatives such
as agents and publicists.
□ School-related (e.g., students, fraternity, alumni; excludes academic institutions)
□ Activist
□ Well-known political figures (e.g., Michelle Obama, Steve Bannon; not politicians)
Notes about sources:
• Victims and offenders are involved in SV – not another crime.
• Politician category is domestic only (listed by party affiliation).
• Trump was coded as an offender (the party/ideology of offender is coded elsewhere), but
his campaign, advisers, and spokespeople would be recorded under “politicians –
REPUBLICAN.”
173
• President was coded by party even if it is not explicitly mentioned.
• Advocacy groups are those who advocate for children, women, SV, or civil / human
rights
• To be coded as an expert, the person must be an expert related to an issue in the article
(e.g., a math professor discussing sexual harassment taking place in his or her department
is not an expert in this context – this person would be coded as a colleague; if the article
involves immigration, a professor who studies immigration would be an expert).
• Unnamed sources include “people familiar with…”, critics, supporters, observers, and
analysts which cannot be connected to a category. For example, Vatican officials would
be coded under “Religious organization” and a spokesperson from CBS would be coded
under “For-Profit Organization.”
• In situations where the article says something like “according to people familiar with the
situation,” numerically it was counted as two.
• If the article said a certain number of victims have accused someone of SV (or a certain
number of women reported), that number was used for the numeric count. For example,
in article 3899 it said “Nearly 50 women have come forward saying they were
assaulted…” so this was coded as 49 victims. Other articles said, “dozens of women”
and this was coded as 24 victims.
Q32
Were experts on SV used as sources? (Includes academics, advocacy groups, government
agencies which deal specifically with this issue, and legal experts (e.g., issues with the law
related to SV)).
□ Yes [1]
□ No [2]
Q33
Source gender (only applies to individuals, not groups like “police”, “prosecutors”, “officials” or
“people”).31 [check all that apply and number of sources in each category]
□ Female sources [1]
□ Male sources [2]
□ Unclear [3]
□ N/A (no individuals identified) [4]
Q34
Is a victim or victims discussed in article?
□ Yes [1]
□ No [2]
If yes, answer questions 35 – 43; if no, skip to question 44.
31 Needed an indication of the source’s gender through the use of a pronoun or Mr./Ms. (unless person is well
known; e.g., Chris Wallace is a male anchor on Fox News).
174
Q35
Victim Characteristics [check all that apply]32:
□ High level leaders in for-profit organizations [1]
□ High level leaders in nonprofit organizations [2]
□ Politicians: elected or appointed, past or present, and political candidates (domestic only)
[3]
□ Foreign government officials (past or present) [4]
□ Other workplace (employees, colleagues, lower-level managers and supervisors, interns,
and volunteers) [5]
□ Military – domestic [6]
□ Military – foreign (includes U.N. peacekeepers) [7]
□ News media (hosts and reporters; should exclude producers who would be high-level
leaders) [8]
□ Entertainers (actors, musicians) [9]
□ Athletes [10]
□ Rich and/or famous [11]
□ Legal field, law enforcement & corrections (includes police and other government
agencies like the DOJ and Border Patrol, judges, attorneys, and corrections officers;
private prisons / detention centers) [12]
□ Health care workers (includes mental health professionals, social workers) [13]
□ Teachers / Professors [14]
□ School administrators (e.g., principals) [15]
□ Coaches [16]
□ Religious leaders [17]
□ Student(s) [18]
□ Low wage service workers (e.g., restaurant workers, farmworkers, hospitality, janitorial)
[19]
□ Friends and family member(s) (includes spouse, girlfriend/boyfriend, child of
girlfriend/boyfriend) [20]
□ Customers, clients, members of the public [21]
□ Stranger(s) [22]
□ Member of church [23]
□ Unclear / not specified [24]
□ Other (with text entry) [25]
□ None of the above [26]
Added categories to victim characteristics based on “other” category:
□ Civilian in conflict zones (did not create similar category in offender characteristics)
□ Patients
32 Includes intended victims (e.g., offender was caught before he/she could commit the crime because of an
undercover investigation).
175
Q36
Victim age (at time of incident, or when it started in cases of abuse) [check all that apply]:
□ Young (under age 18, includes descriptions of victim as boy/girl, child/children) [1]
□ Old (over 65 or described as elderly) [2]
□ Between 18-64 [3]
□ Not specified or unclear [4]
Q37
Victim race [check all that apply]33
□ Black [1]
□ Latino/a [2]
□ Asian [3]
□ White [4]
□ Other (with text entry) [5]
□ Not specified or unclear [6]
Q38
Victim gender [check all that apply]
□ Female [1]
□ Male [2]
□ Both [3]
□ Not specified or unclear (used if the victim or victims are described as a child, children,
or teenager and pronouns are not used) [4]
Q39
Victim: LGBTQ+ (victim must be identified as LGBTQ+ (i.e., this is not inferred from gender of
the offender and victim, so the sexual abuse of a boy by a priest would not be counted so as not
to conflate pedophilia with homosexuality)).
□ Yes [1]
□ No [2]
Q40
Victim: immigrant (victim must be identified as an immigrant to be coded in this category)
□ Yes [1]
□ No [2]
If yes, answer question 41; if no, skip to question 42.
33 If the race of the victim is well known (e.g., Gretchen Carlson), then the race can be coded without being stated.
176
Q41
Victim: immigration status [check all that apply]:
□ Documented [1]
□ Undocumented [2]
□ Status unclear [3]
Q42
Victim: poor (victim must be identified as poor to be coded in this category):
□ Yes [1]
□ No [2]
Q43
Victim: medical condition / disability (victim must be identified as having a medical condition
(which includes mental illness) or disability (physical or intellectual) to be coded in this
category):
□ Yes [1]
□ No [2]
Q44
Is an offender or offenders discussed in this article?
□ Yes [1]
□ No [2]
If yes, answer questions 45 – 53; if no, skip to question 54.
Q45
Offender Characteristics [check all that apply]:
□ High level leaders in for-profit organizations [1]
□ High level leaders in nonprofit organizations [2]
□ Politicians: elected or appointed, past or present, and political candidates (domestic only)
[3]
□ Foreign government officials (past or present) [4]
□ Other workplace (employees, colleagues, lower-level managers and supervisors, interns,
and volunteers) [5]
□ Military – domestic [6]
□ Military – foreign (includes U.N. peacekeepers) [7]
□ News media (hosts and reporters; should exclude producers who would be high-level
leaders) [8]
□ Entertainers (actors, musicians) [9]
□ Athletes [10]
□ Rich and/or famous [11]
□ Legal field, law enforcement & corrections (includes police and other government
agencies like the DOJ and Border Patrol, judges, attorneys, and corrections officers;
private prisons / detention centers) [12]
177
□ Health care workers (includes mental health professionals, social workers) [13]
□ Teachers / Professors [14]
□ School administrators (e.g., principals) [15]
□ Coaches [16]
□ Religious leaders [17]
□ Student(s) [18]
□ Low wage service workers (e.g., restaurant workers, farmworkers, hospitality, janitorial)
[19]
□ Friends and family member(s) (includes spouse, girlfriend/boyfriend, child of
girlfriend/boyfriend) [20]
□ Customers, clients, members of the public [21]
□ Stranger(s) [22]
□ Member of church [23]
□ Unclear / not specified [24]
□ Other (with text entry) [25]
□ None of the above [26]
Added categories to offender characteristics based on “other” category:
□ Militia (did not create similar category in victim characteristics)
□ People with a duty to care not listed above (did not create similar category in victim
characteristics)
□ Patients
Q46
Offender age (at time of incident, or when it started in cases of abuse) [check all that apply]:
□ Young (under age 18, includes descriptions of offender as boy/girl, and child/children)
[1]
□ Old (over 65 or described as elderly) [2]
□ Between 18-64 [3]
□ Not specified or unclear [4]
Q47
Offender race [check all that apply]34
□ Black [1]
□ Latino/a [2]
□ Asian [3]
□ White [4]
□ Other (with text entry) [5]
□ Not specified or unclear [6]
34 If the race of the offender is well known (e.g., Harvey Weinstein, Bill Cosby, Kevin Spacey), then the race was
coded without being stated explicitly in the article.
178
Q48
Offender gender [check all that apply]
□ Female [1]
□ Male [2]
□ Both [3]
□ Not specified or unclear (used if the offender or offenders are described as a child,
children, or teenager and pronouns are not used) [4]
Q49
Offender: LGBTQ+ (offender must be identified as LGBTQ+ (i.e., this is not inferred from
gender of the offender and victim, so the sexual abuse of a boy by a priest would not be counted
so as not to conflate pedophilia with homosexuality)).
□ Yes [1]
□ No [2]
Q50
Offender: immigrant (offender must be identified as an immigrant to be coded in this category)
□ Yes [1]
□ No [2]
If yes, answer question 51; if no, skip to question 52.
Q51
Offender: immigration status [check all that apply]:
□ Documented [1]
□ Undocumented [2]
□ Status unclear [3]
Q52
Offender: poor (offender must be identified as poor to be coded in this category):
□ Yes [1]
□ No [2]
Q53
Offender: medical condition / disability (offender must be identified as having a medical
condition (which includes mental illness) or disability (physical or intellectual) to be coded in
this category):
□ Yes [1]
□ No [2]
Notes on Victim and Offender Characteristics
• The same person can be coded in multiple categories (e.g., Harvey Weinstein would be a
“high level leader in a for-profit organization” as well as “rich and/or famous”).
179
• The law enforcement category is specific to democratic countries (the police do not hold
a position of trust in an authoritarian regime).
• I did not code victims or offenders who are simply mentioned in the story for added
context. For example, in a story about Kevin Spacey which mentions Harvey Weinstein, I
did not code Weinstein or his victims. This rule applies to stories which are about a
specific incident, offender, or victim and someone else is mentioned; all examples in
mixed or thematic stories were coded (there isn’t a primary victim or offender).35
Q54
Are other victims or offenders mentioned for context?
□ Yes [1]
□ No [2]
Q55
Who – or which organization – was the article primarily about? (Only applies to episodic and
primarily episodic articles.)
□ Catholic Church [1]
□ Bill Cosby [2]
□ Harvey Weinstein [3]
□ Fox News (Ailes, O’Reilly) [4]
□ Steve Wynn [5]
□ Les Moonves [6]
□ Donald Trump [7]
□ Congress (House or Senate) [8]
□ Brett Kavanaugh [9]
□ Roy Moore [10]
□ Jeffrey Epstein [11]
□ Julian Assange [12]
□ R. Kelly [13]
□ Other [14]
□ None of the above [15]
Q56
Text entry for notes / comments about the article.
35 I had a separate question to code whether there were other victims or offenders mentioned for context in case I
needed to go back later and code them.
180
Appendix C
Logit Regression Models
Table 10. Logit Regression of Episodic Coverage: Model 1 vs. Model 2
Variables
Model 1
Coefficient
(S.E.)
Model 2
Coefficient
(S.E.)
Legal Frame -- 1.961
(1.284)
Conflict Frame -- 0.604
(1.372)
Victim Frame -- 1.924
(1.372)
Political Frame -- 3.582*
(1.409)
Business Frame -- 2.796*
(1.355)
Conflict News Value -0.088
(0.268)
1.070*
(0.537)
Cover-up News Value -0.896*
(0.443)
-0.680
(0.456)
Familiarity News Value -0.004
(0.278)
-0.032
(0.285)
Gov't Action News Value 0.329
(0.251)
0.368
(0.276)
Immediacy News Value 0.689
(0.355)
0.627
(0.372)
Legal News Value 1.373***
(0.295)
1.517***
(0.318)
Magnitude News Value -0.565*
(0.278)
-0.612*
(0.290)
Rich/Famous News Value 1.549***
(0.314)
1.637***
(0.329)
Post-#MeToo 0.431
(0.275)
0.464
(0.289)
Newspaper -0.315
(0.264)
-0.388
(0.276)
Word Count -0.0004
(0.0003)
-0.0004
(0.0003)
Constant 0.241
(0.565)
-1.867
(1.363)
*p<0.05; **p<0.01; ***p<0.001
Percent correctly predicted in model 1: 77.2%; (n = 513)
Percent correctly predicted in model 2: 79.5%; (n = 513)
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Feeling polarized: emotional partisan polarization and its reinforcement through campaign advertising
PDF
Three text-based approaches to the evolution of political values and attitudes
PDF
The politics of transformation: mass media and the Northern Ireland peace process
PDF
Global governance of sport in a digital age: the political economy of sport integrity regulation
PDF
Environmental policy reforms in a global and regional world: mimetic, diffusive, and coercive policy adoptions in France and Korea
PDF
The fabled fourth estate: media freedom, democracy and human rights
PDF
Policing accountability: an empirical investigation of state-sponsored police reform in Riverside, California
PDF
Abolitionists in the university: student activism, institutional change, and the role of alternative media
PDF
Rebels with a cause: Youth, social movements, and media
PDF
Political strength, not skirt length: female political figures, media bias, and how public relations can help
PDF
The new news: vision, structure, and the digital myth in online journalism
PDF
The impact of 9/11 on the judicial treatment of Middle Eastern asylum applicants in the U.S. courts
PDF
In the eyes of L.A.'s Black baby boomers, Mayor Karen Bass makes the grade (barely): weighing in on the first Black woman's first 16 months as the city's chief executive
PDF
Measuing and mitigating exposure bias in online social networks
PDF
Navigating political polarization: a group case study of community engagement in the adoption of intersectional ethnic studies in a southern California K-12 school district
PDF
Intermedia agenda setting in an era of fragmentation: applications of network science in the study of mass communication
PDF
Carnival in the birdcage: a study of the film and media industries' reforms in China after 1978 and its soft power
PDF
Changing political attitudes and behavior in a diverse America: incorporating individual and contextual determinants
PDF
A theory of status-quo terrorism: democracies in conflict and their proclivity to outsource repression
PDF
How misinformation exploits moral values and framing: insights from social media platforms and behavioral experiments
Asset Metadata
Creator
Cornelius, Michelle Renae
(author)
Core Title
Partisan media bias vs. structural media bias: which one dominates in a polarized environment?
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Political Science and International Relations
Degree Conferral Date
2024-05
Publication Date
05/17/2024
Defense Date
05/03/2024
Publisher
Los Angeles, California
(original),
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
#MeToo,framing,news values,newsworthiness,OAI-PMH Harvest,partisan media bias,structural media bias
Format
theses
(aat)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Crigler, Ann (
committee chair
), Barnes, Jeb (
committee member
), Dundes Renteln, Alison (
committee member
), Hollihan, Thomas A. (
committee member
)
Creator Email
artshapeslife@gmail.com,mrcornel@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC113939977
Unique identifier
UC113939977
Identifier
etd-CorneliusM-12935.pdf (filename)
Legacy Identifier
etd-CorneliusM-12935
Document Type
Dissertation
Format
theses (aat)
Rights
Cornelius, Michelle Renae
Internet Media Type
application/pdf
Type
texts
Source
20240517-usctheses-batch-1153
(batch),
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
#MeToo
framing
news values
newsworthiness
partisan media bias
structural media bias