Introduction Globalization of the mass media has greatly facilitated
Globalization of the mass media has greatly facilitated the flow of information around the
world. Hot issues in each country are being reported speedily and often instantly by news
organizations around the world. With around-the-clock stories and images across a wide range of
news categories, newspapers, television channels, radio stations and other online or offline news
organizations are making news more accessible to their audiences. Different news organizations,
however, may apply their own criteria in “selecting and highlighting some facets of events or
issues and making connections among them so as to promote a particular interpretation,
evaluation, and/or action” (Entman, 2004, p.5). In other words, an issue or event may be “framed”
in a particular way when it is presented to the audiences in the form of a news story.
Of particular interest in this study is the frames used by news organizations in their
coverage of events taking place in different geographic, cultural, political and economic settings.
The recent protests in Hong Kong (or the Occupy Central Movement) present a robust
opportunity for research not only because the event was broadly covered by major news media
around the world, but also because the event took place in Hong Kong, the crossroads of
democracy and authoritarianism, capitalism and communism, modernism and traditionalism.
Purpose of Study
The study has two goals. The first goal is to provide a theoretical framework for
approaching the issue of news framing. A major argument derived from the theoretical
framework is that framing is commonly applied in news reporting and it can affect or bias
audiences’ interpretation and evaluation of news events. The second goal is to describe a
quantitative content analysis designed to test research hypotheses derived from the theoretical
framework. It is hypothesized that foreign/western and domestic/Hong Kong news media applied
different frames in their coverage of the protests. Specifically, western/foreign news media would
place greater emphasis on the positive and beneficial aspects of the protests, while
domestic/Hong Kong news media would emphasize the negative and detrimental aspects. It is
further hypothesized that western/foreign news media would more likely frame the protests as a
civil disobedience and democratic movement, whereas domestic/Hong Kong news media would
highlight the anti-government and anti-social aspects of the event.
Chapter Two
Literature Review
To provide the theoretical framework of the study, a review of the framing theory and its
applications in mass media research is presented first, followed by a description of the
background of the Hong Kong protests.
Frame and Framing
In social-psychological research, a frame is a schema of interpretation, a collection of
anecdotes and stereotypes that individuals rely on to understand and respond to events (Goffman,
1974). This concept was first introduced by Goffman (1974) in his book Frame Analysis: An
Essay on the Organization of Experience which explains how frames are used to organize
experiences and guide actions. According to Goffman, people tend to understand the issues that
take place around the world on the basis of their own “primary framework” (1974, p. 21). Like a
frame that can give a picture a greater sense of structure and meaning, the primary framework
can be “seen as rendering what would otherwise be a meaningless aspect of the scene into
something that is meaningful” (p. 21). There are two types of primary frameworks: natural and
social. Natural frameworks seek to “identify occurrences that are undirected, unoriented,
unanimated and unguided”; social frameworks tend to “provide background understanding for
events that incorporate the will, aim and controlling effort of an intelligence, a live agency, the
chief one being the human being” (p. 22). Moreover, Goffman points out that while the outcomes
of events identified by the natural frameworks cannot be guided by any physical efforts, the
events of the social frameworks can generate outcomes that can be controlled by continuous
physical actions. People use these frameworks as the basis of understanding, and the frameworks
used can greatly influence their processing and interpretation of incoming information.
Frameworks are therefore essential tools for sense making and discovery.
The concept of frame was quickly adopted and extended in mass communication research
whereby the focus gradually shifted from pre-existing mental schemata (frames) to the ways and
means of creating or shaping the schemata (framing) through the process of communication.
Framing, a process-centered conception thus began to replace frame, a representation-centered
conception. In his 1993 essay titled Framing: Towards Clarification of a Fractured Paradigm,
Entman elaborates framing as a theory of media effects. “To frame is to select some aspects of a
perceived reality and make them more salient in a communicating text, in such a way as to
promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment
recommendation for the item described” (Entman, 1993, p. 52). Framing thus involves two
related processes: the “selection” process by which frames shed light on some specific topics of a
certain event during the communication process, and the “salience” process by which more
attention is paid to evaluating these selected topics than others. Entman (1993) further contends
that framing may take place at four locations during the process of communication: the
communicator, the text, the receiver, and the culture. A good example is the process of news
coverage: The communicators (reporters and editors) play essential roles in deciding what frames
can be used in news reports, regardless of whether they are making these decisions in a conscious
or an unconscious manner. The text in news reports “contains frames, which are manifested by
the presence or absence of certain keywords, stock phrases, stereotyped images, sources of
information, and sentences that provide thematically reinforcing clusters of facts or judgments”
(Entman, 1993, p. 52). The receiver’s (or the audience’s) attitude towards certain events can be
influenced or guided by the frames that are contained in the text. The culture is “the stock of
commonly invoked frames,” and it is usually regarded as “the empirically demonstrable set of
common frames exhibited in the discourse and thinking of most people in a social grouping”
(Entman, 1993, p. 53). Being contained in these four locations, framing is able to generate effects
on the structure of the news reports as well as the audiences’ attitudes and evaluations of the
news events. In Entman’s view, framing can be used as a pivotal tool during the process of
communicating, as it not only “offers a way to describe the power of a communicating text” (p.
51), but also exerts effects in forming or guiding one’s attitude toward certain issues.
In explicating the concept of framing, Scheufele (1999) drew the distinction between
frame building and frame setting. Frame building is usually concerned with how frames create
societal discourse about certain issues and how frames are being chosen by journalists in news
reporting. Frame setting, on the other hand, deals with how frames exert their influence on
audiences’ attitudes and opinions. There are three related areas of frame building: journalist
norms, political actors, and cultural contexts. Additional factors that may affect frame building
include pressures and constraints from within the newsroom; political affiliations or inclinations
of journalists, and coercion and enticement from government enterprises and political or
corporate elites. Some research, for example, has shown that the extrinsic factors (e.g.,
government and political members) can place strong influence on the framing of political issues
when reported by journalists (e.g., Tuchman, 1978).
Frame building is often followed by frame setting, which involves “the interaction
between media frames and individuals’ prior knowledge and predispositions” (de Vreese, 2005,
p. 52). A number of studies have examined audience frames as the outcome or dependent
variable, given specific kinds of media frames. The general finding is that media frames can
affect the accessibility and perceived importance of frames in individuals’ minds (e.g., Iyengar,
1991; Nelson, Clawson & Oxley, 1997). As Reese (2010) points out, frames are built upon
previous knowledge and existing cultural norms, so it is important to analyze whether the frames
used within a news story are aligned with the beliefs and mood of the target audience on a key
issue. This results in an interaction between the incoming message structure and the
psychological characteristics of the receiver. Frame resonance occurs when media and individual
frames are linked in congruency.
Closely associated with framing is the concept of agenda setting (McCombs & Shaw,
1972). Some scholars argue that framing should be construed as an extension of agenda setting,
as framing affirms the ability of the news media to influence the salience of topics on the public
agenda, which in turn determines what the public thinks about, although not necessarily what
they think. That is, the concentration of new media on a few issues and topics leads the public to
perceive those issues as more important than others (Lane, 2001). Other researchers maintain that
framing is different from agenda setting (e.g. Scheufele & Tewksbury, 2007). While agenda
setting emphasizes the perceived importance of issues as a function of the amount of media
coverage (or salience) the issues receive, framing is more concerned with the perceptions and
interpretations of factual information as a function of the ways the information is presented. In
other words, audiences’ reactions to issues and events are affected not only by the amount of
media coverage the issues and events receive (agenda setting), but also by the ways through
which the issues and events are presented (framing).
In their recent article titled News Framing Theory and Research, Tewksbury and
Scheufele (2009) traced the theoretical roots of framing back to psychology at the micro-level,
and sociology at the macro-level. They indicate that framing, especially news framing, serves as
an explanatory model in studying the effects of media content on audiences’ beliefs, attitudes and
behaviors. Framing effects, they argue, should be distinguished from information effects,
persuasion effects and agenda-setting effects, although all these effects can “result from exposure
to news message” (p. 22).
It seems safe to say that framing has become one of the most influential theories in mass
communication research. “Today, virtually every volume of the major journals features at least
one paper on media frames and framing effects” (Scheufele & Iyengar, 2012). In the next section,
we review some of the representative studies of framing in the news.
Framing the News
News framing research focuses on “the process by which people develop a particular
conceptualization of an issue or reorient their thinking about an issue” (Chong & Druckman,
2007, p.2), with the assumption that the way that an issue is framed in news reports can influence
the audiences’ understanding and interpretation of the issue (Scheufele & Tewksbury, 2007).
Much research on news framing examined the ways politicians, issue advocates and other
stakeholders manipulate reporters and editors to convey their preferred meanings through news
media (D’Angelo & Kuypers, 2010). It has also been shown that critical news events can alter
the parameters of a policy debate, and political elites often use such events to influence public
policy and mobilize public opinion in support of their objectives (e.g., Birkland, 1997; Cobb &
Elder, 1983; Kingdon, 1984). Examples of these critical events are military actions, hijackings,
assassinations and natural disasters. These events may also introduce new or additional frames
which are inconsistent with or unrelated to previous frames and thus have the capacity to reshape
public opinion on issues (Schnell & Callaghan, 2004).
When critical events took place, the news media become “massive search engines”
looking for dramatic ways to frame the issues for audience consumption (Wolsfeld, 2003, p.
229). Lawrence (2010) further suggests that, in the context of foreign policy and national
security, mainstream news media typically show less independence in framing critical events;
instead, they tend to rely more upon high-level government officials to frame the news. Reporters
and editors may see it as irresponsible to introduce perspectives that lie outside of what Hallin
(1986,
p. 116-117) described as the “sphere of legitimate controversy.” This trait manifests itself
prominently in a reluctance to challenge high officials during wartime (e.g., Aday, Livingston &
Hebert, 2005). This can also be traced to the idea that raising questions that challenge the official
position on foreign affairs can call into question the patriotism of journalists (Lawrence, 2010).
However, Entman (2004) challenges the position that the news media should “privilege some
definitions at the expense of others” (Kinder & Sanders, 1966, p. 163) by focusing on the frames
and interpretations of the elites and officials in power. Instead, “the media should provide enough
information independent of the executive branch that citizens can construct their own
counterframes of issues and events” (Entman, 2004, p. 17).
News framing has also been examined in the context of political communication. Jacoby
(2000), for example, shows how framing is used as a propaganda tool in political campaigns
through which advocates are able to “mobilize voters behind their policies by encouraging them
to think about those policies along particular lines” (p.751). Arbour (2014) makes similar
arguments that partisan framing exists in political campaigns to “activate the images and values
citizens already possess” (p.608). McCombs and Reynolds (2008) also state that the government
has been using the news media as an effective tool to exert control over the public opinion polls
during elections, as voters tend to “use the media to help them sort through important issues
before they vote” (p.1).
Several studies have shown that framing is often used in news coverage of such events as
political or social movements. Different effects on public opinions on the movements often
follow different news frames used. For instance, studies of valence framing have shown that
factually identical news stories may result in opposite evaluations, depending on whether the
stories are framed positively or negatively (e.g., de Vreese & Boomgaarden, 2003; Iyengar,
1991). A valenced news frame implicitly or explicitly highlights either the positive (gain,
advantageous or favorable) aspects or the negative (loss, disadvantageous, unfavorable) aspects
of factually identical news events, resulting in frame-cued evaluation and assessment among the
audience.
De Vreese, C. & Boomgaarden (2003), for example, investigated how a key summit of
the Council of the European Union was framed in the news and how the frames affected public
support for the movement toward European integration. Through a content analysis of newspaper
and television stories in three European countries, they observed that the summit was framed in
terms of political-institutional consequences and the overall valence of the news as
disadvantageous toward the EU and EU integration. A subsequent experiment further confirmed
that participants exposed to the advantageous “opportunity-framed” news stories showed higher
levels of support for EU than participants exposed to disadvantageous “risk-framed” news
stories.
Johnson, Maio and Smith-McLallen (2005) put forward the hypothesis that loss-framed
messages are generally more persuasive than gain-framed messages. The hypothesis has received
some support from studies showing that negative information is more persuasive than positive
information (e.g., Rozin & Royzman, 2010) and that people are more inclined to take a risk in
order to reduce losses than increase gains (e.g., Kuhberger, Schulte-Mechlenbeck & Perner,
1999). A meta-analysis of 165 loss-gain framing studies by O’Keefe and Jensen (2006), however,
found no significant difference in persuasive effects between the two message frames.
More recently, Fu, Zhou, Zhang, Chan and Burkhart (2012) examined news coverage of
natural disasters in China, and found that the Chinese government used the news media as a
“watchdog” tool to watch over the citizens, especially during the time when disasters broke out.
Of the news articles that covered natural disasters, few of them framed the stories by
“commenting on the government negatively” or being “openly critical to the government” (p.81).
Through news framing, the news media were able to successfully curb the negative attitudes
toward the government.
The research reviewed above suggests that framing is rampant in the production of media
content, including the reporting of factually identical news events. In the next section, we will
review the background of the 2014 Hong Kong protests, the focal news event examined in this
study.
Background of the Hong Kong Protests
The Hong Kong protests were a historical event that took place in Hong Kong between
September 28 to December 11, 2014. The protests occurred at the busiest sections of Hong Kong,
the most preeminent financial center in Asia, and presented what many commentators considered
the biggest challenge to the Chinese government since the 1989 Tiananmen protests (e.g.,
Tharror, 2014).
Also known as the Occupy Central Movement or the Umbrella Revolution, the event was
triggered by the decision made by the Standing Committee of the National People’s Congress
(SCNP) of the PRC on issues relating to the selection of the Chief Executive of Hong Kong by
universal suffrage and on the method for forming the Legislative Council of Hong Kong in 2016.
Officially announced on August 31, the decision in effect ruled out a fully democratic election for
Hong Kong leaders because only candidates approved by Beijing would be allowed to run.
To protest against the decision, thousands of students, teachers and scholars in Hong
Kong gathered around the Hong Kong governmental headquarters building on the evening of
September 20. They then occupied several major city intersections to show their dissent against
Beijing’s decision. Drawing attention from around the world, the protests dragged on for more
than two months.
The protests brought about interruptions to Hong Kong’s transportation, domestic
commerce as well as international trade. As more and more students in Hong Kong became
involved in the protests, schools in the central and western areas in Hong Kong were suspended.
The protests soon became a wedge issue that exacerbated social, economic and political divisions
that had long existed in Hong Kong. Case in point: People formed into pro- and anti-protesters
(or “yellow” and “red”) groups during the protests, resulting in more chaos and conflict in the
society. The pro-protesters group insisted on the basic principles of democratic government and
demanded less intervention from the Chinese government; however, the anti-protesters group
argued that the protests should be curbed by the government because they were bringing Hong
Kong’s economy to a halt. The student-initiated protests soon evolved into furious debates
intermingled with occasional physical confrontations between people with opposing views, all of
which were covered extensively by major news media around the world.
To better understand what triggered the Hong Kong protests, we turn to the political and
economic factors involved.
Political Factors. Consisted of three main territories (Hong Kong Island, Kowloon Peninsula
and the New Territories), Hong Kong was colonized by the United Kingdom for 156 years (1841
to 1997). Although they were never granted full rights and British citizenship, people in Hong
Kong did regard democracy and the protection of human rights as issues of great importance.
Consequently, in accordance with the “one country, two systems” policy, the Seventh National
People’s Congress (NPC) of the People’s Republic of China adopted the Hong Kong Basic Law
(constitution) in 1990 which went into effect in 1997, the year Hong Kong was handed over to
the PRC. Of particular relevance to political rights are Article 45 and 68 of the Basic Law:
(Article 45)
• The Chief Executive of the Hong Kong Special Administrative Region shall be
selected by election or through consultations held locally and be appointed by the
Central People's Government.
• The method for selecting the Chief Executive shall be specified in the light of the
actual situation in the Hong Kong Special Administrative Region and in accordance
with the principle of gradual and orderly progress. The ultimate aim is the selection
of the Chief Executive by universal suffrage upon nomination by a broadly
representative nominating committee in accordance with democratic procedures.
(Article 68)
• The Legislative Council of the Hong Kong Special Administrative Region shall be
constituted by election.
• The method for forming the Legislative Council shall be specified in the light of the
actual situation in the Hong Kong Special Administrative Region and in accordance
with the principle of gradual and orderly progress. The ultimate aim is the election
of all the members of the Legislative Council by universal suffrage.
The implementation of “universal suffrage” soon became one of the most controversial
elements of the Basic Law. Since 1997, as Kan (2013, p. 73) indicates, Hong Kong has been
“standing at a critical juncture in the struggle for universal suffrage toward realizing full
democracy.” While Hong Kong citizens have repeatedly demanded that the Chinese government
realize its promise, the Chinese authorities have successfully delayed the implementation of
universal suffrage over the years: In 2004, the Standing Committee of the NPC (NPCSC) ruled
out the possibility of universal suffrage in 2007 and 2008; in 2007, the Committee ruled out
universal suffrage being introduced by 2012.
In September 2014 the NPCSC announced its decision on the selection of Hong Kong’s
Chief Executive in 2017. The decision stated that the selection of the Chief Executive might be
implemented by the method of universal suffrage. However, it also stated that candidates for the
Chief Executive position would be nominated by “a broadly representative nominating
committee” formed “in accordance with the number of members, composition and formation
method of the Election Committee for the Fourth Chief Executive.” Public election would be
held after the Chief Executives nominees have been selected by the Nominating Committee,
subject to the approval of the NPCSC. According to Bush (2014), Director of Center for
Northeast Asian Policy Studies of the Brookings Institution and a Senior Fellow of Foreign
Policy, the NPCSC decision insures that “candidate selection will be in the hands mainly of
people who are sympathetic to Beijing. In addition, over half of the members of the nominating
committee would have to approve each candidate, which means that no pan-democrat could get
nominated if Beijing disapproved of him or her.” Martin (2011), a specialist in Asian affairs at the
Congressional Research Service, also stated that “It is generally thought that the transformation of
the Election Committee into a nominating committee will provide the Chinese government with
enough insurance that the elected Chief Executive will be approved by the NPCSC, thereby
avoiding an embarrassing and unpleasant situation” (p. 13).
Public protests immediately broke out after the NPCSC announced its decision. Initially
led by college students, the protests were soon joined by citizens and leaders from all sectors of
society. It should be noted that pro-democracy movements have had a long history in Hong Kong
(Early, 2014). For example, demonstrations have been held on June 4 every year in Hong Kong to
honor the memory of the 1989 Tiananmen Square Protests.
The NRCSC decision on universal suffrage might have triggered the 2014 Hong Kong
protests, but beneath the surface the protests might be an expression of the humiliation and anger
of more than 150 years of colonial occupation, the disappointment and distrust in the Chinese
government, and the deep hunger of Hong Kong people for greater political freedom and
democracy.
Economic Factors. It is generally known that Hong Kong has been experiencing an economic
decline in recent years. On the macro level, Hong Kong is slowly losing its competitiveness,
according to the International Institute for Management Development (Steinbock, 2013). In terms
of economic performance, Hong Kong fell in 2013 from fourth to eighth place in world ranking.
Furthermore, Hong Kong has performing increasingly worse in government efficiency, business
efficiency and infrastructure, according to the same report.
In her recent article in Foreign Policy, Lu (2014) notes that Hong Kong has been a city
that “is losing its edge as a global financial and commercial center” and its economy “will be
overshadowed by China’s major cities in 2022.” She also points out that the economy of Hong
Kong has grown at a slower speed compared to that of Beijing and Shanghai – Hong Kong’s
GDP annual rate has been hovering around two percent in recent years, while the growth rates of
major regional districts in Beijing and Shanghai have reached more than seven percent per year.
Yglesias (2014) makes an interesting but no less significant observation about a particular
problem in Hong Kong – the sales of luxury goods has kept declining in recent years – and
attributed the problem to “a Chinese government crackdown on corruption and conspicuous
consumption.”
As one of Asia’s Four Little Dragons, Hong Kong has always been marked by a great
sense of pride in its economic development which, almost miraculously, transformed the tiny port
city into one of the world’s largest financial centers. Under a capitalist system, Hong Kong was
able to economically outperform every major city in mainland China until recent years. Today,
however, major Chinese cities like Shanghai, Beijing, Guangzhou and Shenzhen have emerged as
comparable financial and employment hubs that are attracting more investments from overseas.
As Guy (2014) warns, “Economic and financial forces have been set in motion that will
dramatically, but quietly, shift the operations and focus of international banks and asset managers
to Beijing or Shanghai. And it will occur far faster than most people could have anticipated.”
The most direct consequence of Hong Kong’s declining economy is the widening wealth
gap and a lack of future opportunities for younger people (Pau, 2015). Property prices are
persistently too high and wages too low (the minimum wage is 30 HK dollars or $3.87 an hour)
for many Hong Kong residents to make a living. Although the overall unemployment rate in
Hong Kong has remained steady at about 3.3 percent, the unemployment rate for residents aged
15 to 24 is at a much higher rate of 9.4 percent. About 1.3 million people, or 19.6 percent, of the
population were deemed to be living below the poverty line in 2012, according to the Hong Kong
Poverty Situation Report. Figures from the Hong Kong Council of Social Service showed the
poverty rate for the 15- to 24-year-olds to have increased to 17.7 percent, which means nearly
200,000 youngsters are officially living in poverty (Kao, 2013). In Hong Kong, the first
generation whose future living standard may be worse than that of their parents is coming of age
(Pau, 2015).
Today many Hong Kong residents fear Hong Kong will become economically
marginalized. The economic fear is compounded by the political fear of the Chinese authoritarian
government. The increasingly economic and political uncertainty finally reached the breaking
point on September 28, 2014 when downtown Hong Kong turned into a battlefield of tear gas
and seething crowds after the police attempted to crack down a student-led democracy protest,
inciting public fury that brought tens of thousands of people onto the streets of a city long known
as a stable financial center.
For more background information, please refer to Appendix A: Timeline of Hong Kong's
pro-democracy movement and Appendix B: Hong Kong Profile.
Chapter Three
Research Hypotheses
The purpose of this study is to examine the frames used in news coverage of the Hong
Kong protests in 2014. To that end, news stories on the protests from three different newspapers
– South China Morning Post, The New York Times, and The Guardian – were selected and
analyzed.
With a history of 112 years and a daily circulation of 104,000, South China Morning Post
is the number one English-language newspaper published in Hong Kong. The newspaper enjoys
a steady reading circle and has long been one of the most influential newspapers in the region. It
should be noted that both English and Chinese are official languages in Hong Kong and children
start learning English and Chinese at the age of three.
First published in 1851, The New York Times is perhaps the world’s most respected and
influential newspaper that enjoys more than two million total print and digital circulation. The
paper has won 141 Pulitzers Prizes, more than any other newspaper in America (Rainey and
Garrison, 2012).
Founded in 1821, The Guardian is a British newspaper with a daily circulation of
189,000. The paper’s readership is generally on the center-left of the British political spectrum.
In 2014, the paper was named newspaper of the year by the British Press Awards for its reporting
on government surveillance.
The three selected newspapers are likely to follow similar daily routines of journalism,
with reporters covering established beats and relying on government sources for regular
information to reinforce prevailing norms and state-society relationships (Tuchman, 1978). Thus,
all three papers practice routine news coverage which “legitimates the existing political order by
disseminating bureaucratic idealizations of the world and by filtering out troublesome
perceptions of events” (Fishman, 1980).
It also seems likely that the three newspapers might have different journalistic values and
geopolitical characteristics that are reflected in their framing of factually identical news events
like the Hong Kong protests. For example, while upholding objectivity as a journalist norm,
American news media tend to cater to the existing elite consensus on a given issue (Powlick &
Katz, 1998). British journalists, on the other hand, find greater appeal in notions of truth,
independence and “fair play” than objectivity (Maras, 2013). Both countries (U.S. and U.K.) are
fully democratic and known around the world for their respect and tolerance of free speech. In
contrast, Chinese news media generally serve as a servant of the party as well as the state. The
practice of strict news censorship dating back to the Mao era still persists, although market-
oriented reforms since the 1970s have created a “partially liberalized authoritarian media system”
(Chan & Qiu, 2002; Daniel, 1999).
As a result, when it comes to the Hong Kong protests, the news reports from South China
Morning Post, under the pressures of the Chinese government, are more likely to stress the anti-
social and negative consequences of the protests (e.g., transportation chaos caused by the protest
and its economic costs to the city), while The New York Times and The Guardian are more likely
to highlight the pro-democracy and positive aspects of the protests (e.g., protesters’ demand for
political freedom and resistance against the rule by an authoritarian government).
Therefore, the first research hypothesis is:
H1: In reporting the Hong Kong protests, the South China Morning Post would be more likely to
focus on the negative effects or costs of the protests than The New York Times and The Guardian.
For the Hong Kong protesters, they had also been likely to be framed as playing negative
roles (e.g. the protesters’ attacks on the policemen and the inconveniences that were brought
forward to the Hong Kong citizens) by news reports from the South China Morning Post when
compared to news articles from The New York Times and The Guardian. Therefore, the second
hypothesis comes forward:
H2: In reporting the Hong Kong protests, protesters are more likely to be framed in a negative
way by the South China Morning Post compared to news reports from The New York Times and
The Guardian.
It came as no surprise that government officials in Hong Kong and in Beijing uniformly
denounced the protests as illegal, and Chinese state media and officials claimed repeatedly that
the West had played an instigating role in the protests. The camaraderie between Hong Kong and
Chinese officials could place strong influence, if not pressure, on local journalists in their
reporting of the protests. The reasoning leads to the third hypothesis:
H3: The news articles in South China Morning Post would be more likely than those in The New
York Times and the Guardians to frame the Hong Kong government and its officials as playing
positive roles during the protests.
It is well known that the Chinese government has been playing a strict scrutinizing role
over the news media. As a servant for the Chinese central government, few Chinese news
agencies tend to cover the censorship issue in their news articles, as any report containing the
issue is scrutinized and restricted from publishing. Not surprisingly, the 2014 Hong Kong
protests were severely censored in mainland Chinese news media. On the surface, there seems to
be no censorship in Hong Kong – freedom of the press in Hong Kong is theoretically guaranteed
by a bill of rights. As a “special administrative region” of China, however, the Hong Kong
government faces mounting pressure from China to have more control over its press. The
pressure results in indirect or self-imposed news censorship in Hong Kong, a region that had
once enjoyed unfettered Western-style press freedom under British rule. It would therefore be
more likely for British and American news media to cover censorship-related issues in stories of
the protests than Hong Kong news media. The fourth hypothesis is:
H4: In news coverage of the Hong Kong protests, news censorship would be less likely reported
in South China Morning Post than in The New York Times and The Guardian.
Since the transfer of sovereignty over Hong Kong in 1997, government restrictions on
access to information have persisted and violent attacks against journalists and media executives
have increased significantly (Freedom House, 2014). Out of 180 countries, Hong Kong’s World
Press Freedom Index slid from the 18th place in 2002 to the 61st place in 2014. Furthermore, many
of Hong Kong's media outlets are owned by local tycoons who have significant business ties in
the mainland, so nearly all of them have adopted self-censorship at some level and have mostly
maintained a conservative editorial line in their news coverage. Citing Article 23 of the
Basic Law which prohibits crimes against national security and sedition, the Hong Kong
government declared it illegal for the news media to be seditious by disclosing “national secrets.”
The ambiguous definition of national secrets led to concerns that the law might be used as a tool
for silencing dissidents’ voices, as has happened in mainland China. Thus the fifth hypothesis:
H5: In news coverage of the Hong Kong protests, politically sensitive topics such as human
rights, democracy, civil disobedience, references to the Tiananmen Square Movement and
criticism of Chinese communism would be less likely reported in South China Morning Post than
in The New York Times and he Guardian.
Chapter Four
Methodology
A quantitative content analysis of news stories on the Hong Kong protests was used to
test the research hypotheses. Harwood and Garry (2003) describe content analysis as “a
technique that enables analysis of ‘open-ended’ data to be structured for purpose of diagnosis” (p.
479). The method allows for the systematic and objective analysis of messages in all types of
documents, speeches, media releases, scripts, and interviews (Wrench, Thomas-Maddox,
Richmond, & McCroskey, 2008). It has also become a popular research method in analyzing
media content, including news stories (Neundorf, 2002). Considering the proposed research
hypotheses, content analysis is an appropriate method of research for this study as it can
demonstrate how different newspapers construct stories about a historical event.
Data Sources
News reports of the Hong Kong protests examined in this study were collected from three
newspapers: South China Morning Post, The New York Times, and The Guardian. A total of 249
news stories of the protests published during September 29 and December 18 in 2014 were
selected from the online news archive of each newspaper (see Table 1). The selection was
conducted through a search process involving the use of such keywords as “Hong Kong
protests”,
“occupy central”, “Hong Kong students”, “Hong Kong government”, “Hong Kong police”,
“Hong Kong public opinion”, “democratic movements in Hong Kong”, “news censorship”,
“Hong Kong economy”, “Hong Kong politics”, “Chinese government”. Of the 249 articles
selected, 96 (38.6%) were selected from The New York Times, 77 (30.9%) from The Guardian
and 76 (30.5%) from South China Morning Post.
Table 1
Distribution of Newspaper Stories
Frequency
%
Valid %
Cumulative %
New York Times
96
38.6
38.6
38.6
The Guardian
77
30.9
30.9
69.5
South China Morning Post
76
30.5
30.5
100.0
Total
249
100.0
100.0
Coding Scheme
Each news article was used as the primary unit of analysis. The coding categories ranged
from basic structural elements (e.g., time, dateline, byline and story type) to more complex
coding of news frames. To test the research hypotheses presented earlier, the 249 news articles
were content analyzed in terms of five different news frames listed below:
1. Valence Frames of the Protests
Consistent with prior studies on valance framing, four coding alternatives were developed:
positive, neutral, negative and unclear frame. The news articles that contained positive or
beneficial aspects of the protests (e.g., protests as part of a larger democratic movement,
exemplification of civil-disobedience, progress toward greater political freedom, nonviolence or
peaceful resistance) were coded as positively framed stories, while articles presenting negative or
detrimental aspects of the protests or showing passive attitudes toward the protests (e.g.,
interrupting social order and commercial practice, fostering social and economic instability,
perpetuating of social and political divisiveness, upsetting the Chinese communist regime) were
coded as negative frames. News articles that presented both positive and negative aspects of the
protests at the same time were coded as having a neutral frame. News articles that presented
neither positive nor negative aspects of the protests were coded as unclear.
2. Valence Frames of the Protesters
Four coding alternatives were used to analyze the news articles’ frames of the protesters:
positive, neutral, negative, and unclear. The news articles that contain positive or supportive
descriptions of the protesters (e.g., freedom or human rights fighters, leaders of democracy,
intelligent or responsible citizens) were coded as positively framed stories, while articles
presenting negative or detrimental aspects of the protesters (e.g., troublemakers, separatists,
anarchists, opportunists, naïve and irrational citizens, Western sympathizers) were coded as using
a negative frame. News articles that presented both positive and negative aspects of the protesters
at the same time were coded as a neutral frame. News articles that presented neither positive nor
negative aspects of the protests were coded as unclear.
3. Valence Frames of the HK Government
The news article’s frames of the Hong Kong government (including its officials) were
categorized into four alternatives: positive, neutral, negative, and unclear. News articles that
expressed support, justification or defense for the government and its actions and reactions
during the protests (e.g., rationalizing the necessity of police violence, defending the use of delay
tactics in negotiations with protesters) were coded as positive. News articles that expressed
criticism, disapproval or dissent against the Hong Kong government (e.g., questioning the
excessive use of police force, the lack of sincerity in conflict resolution or expertise in crisis
management) were coded as negative. News articles that expressed both supportive and critical
views of the Hong Kong government were coded as neutral, and articles that did not address the
Hong Kong government were coded as unclear.
4. Framing of News Censorship
Four coding alternatives were applied when coding the news articles’ framing of news
censorship: positive, neutral, negative and unclear. News articles that regarded news censorship
in Hong Kong and China as necessary or justifiable were coded as positive. News articles that
contained both criticism of news censorship or deemed it unjustifiable were coded as negative.
Articles that contained both positive and negative elements of censorship were coded as neutral,
and those that did not address censorship at all were coded as unclear.
5. Mentioning of Politically Sensitive Issues
There are a number of politically sensitive issues in China. Coding of this content
category was largely based on the presence or absence of certain “blacklisted” keywords,
including “Tiananmen Square protests”, “political dissidents”, “human rights”, “authoritarian
government”, “dictatorship”, “anti-democracy”, “anti-communist”, “media censorship”, “Internet
content blocking”, “freedom of speech”. News articles that included the use of any of the
keywords were coded as Yes, while articles that contained none of the keywords were coded as
No.
Appendix A provides examples of news content coded into the categories defined above.
Chapter Five
Results
Intercoder Reliability
In content analysis, intercoder reliability refers to the extent to which independent coders
agree on the coding of the content of interest with an application of the same coding scheme. The
following procedure was taken to assess intercoder reliability of the present study. First, the
author (coder 1) coded the 249 Hong Kong protests news articles that appeared in South China
Post, The Guardian and New York Times according to the coding scheme presented earlier. Using
the same coding scheme, two graduate students (coder 2 and coder 3) in mass communications at
USF were then invited to code 46 stories randomly selected from the pool of 249. The three
coders are citizens of China, Taiwan and the U.S. respectively. Table 2 presents two sets of
intercoder reliability measures: Percentages of coding agreement between coders and
Krippendorff's alphas. According to Neuendorf (2002, p. 145), percentage agreements “of .90 or
greater are nearly always acceptable, .80 or greater is acceptable in most situations, and .70 may
be appropriate in some exploratory studies for some indices.” All but one percentages of
agreement in this study were greater than .80, indicating acceptable agreement among the three
coders.
Krippendorff’s alpha (α) is generally considered the most reliable measure of intercoder
reliability. The general form of the measure is: α = 1 – Do/De, where Do is observed disagreement
and De is expected disagreement based on an interpretation of chance. According to
Krippendorff (2011, p. 1), “when observers agree perfectly, observed disagreement Do=0 and
α=1, which indicates perfect reliability. When observers agree as if chance had produced the
results, Do=De and α=0, which indicates the absence of reliability.” To interpret alpha,
Krippendorff (2004, p. 241) suggests that “It is customary to require α ≥ .800. Where tentative
conclusions are still acceptable, α≥ .667 is the lowest conceivable limit.” Table 2 shows that all
but one of the five coding categories generated alphas greater than or equal to .80, with an overall
alpha of .85 among the three coders.
Table 2
Intercoder Reliability
Coding Category
%
Agreement
among 3
Coders
%
Agreement
between
Coder 1 &
3
%
Agreement
between
Coder 1 &
2
%
Agreement
between
Coder 2 &
3
Krippendorff's
Alpha
Protests
85.9
91.1
80%
86.7
.80
Protesters
82.6
86.9
78.3
82.6
.75
HK government/police
93.5
91.3
95.7
93.5
.90
Censorship
91.3
91.3
91.3
91.3
.82
Politically sensitive issues
89.9
86.9
93.5
89.1
.80
Overall
88.7
89.6
87.8
88.7
.85
Hypotheses Testing
All hypotheses in this study were tested with the following two-step procedure. First,
Pearson’s Chi-square (X2) test was used to test against the null hypothesis that the frequency of
content distribution is the same across all three newspapers. If the Chi-square test results in a p-
value smaller than .05, then there is strong evidence to state that there is a difference between at
least two of the newspapers. Second, following a significant Chi-square test, a pairwise z-test of
difference in proportions was performed to determine which two newspapers were different in a
specific content category. All pairwise z-tests were performed with Bonferroni corrections to
hold the overall type I error level at 5% or less.
H1. Valence Frames of the Protests
The first hypothesis states that the South China Morning Post would be more (less) likely
to focus on the negative (positive) effects or costs of the protests than The New York Times and
The Guardian. Chi-square test results (Table 3) indicated that such hypothesized difference did
exist among the three newspapers (X2=73.12, df=6, p<.001). Pairwise z-tests provided additional
support for the hypothesis. As Table 4 and Figure 1 show, The New York Times (n=38, 39.6%)
and The Guardian (n=27, 35.1%) published significantly more positively framed stories about
the protests than South China Morning Post (n=8, 10.5%), whereas South China Morning Post
(n=44, 57.9%) published more negatively framed stories of the protests than both The New York
Times (n=13, 13.5%) and The Guardian (n=7, 9.1%). Together, the results provided strong
support for Hypothesis 1.
Table 3
Chi-Square Test: Valence Frame of the Protests
Value
df
Sig. (2-sided)
Pearson Chi-Square
73.133a
6
.000
Likelihood Ratio
72.436
6
.000
Linear-by-Linear Association
19.339
1
.000
N of Valid Cases
249
a. 0 cells (0.0%) have expected count less than 5. The minimum expected
count is 6.41.
Table 4
Crosstab: Valence Frame of the Protests
Newspaper
Total
New York
Times
The Guardian
South China
Morning Post
Valence frame of the
protests
Positive Count
38a
27a
8b
73
% within
Newspaper
39.6%
35.1%
10.5%
29.3%
Neutral Count
33a
41b
17a
91
% within
Newspaper
34.4%
53.2%
22.4%
36.5%
Negative Count
13a
7a
44b
64
% within
Newspaper
13.5%
9.1%
57.9%
25.7%
Unclear
Count
12a
2a
7a
21
% within
Newspaper
12.5%
2.6%
9.2%
8.4%
Total
Count
96
77
76
249
% within
Newspaper
100.0%
100.0%
100.0%
100.0%
Each subscript letter denotes a subset of Newspaper categories whose column proportions do not differ
significantly from each other at the .05 level.
Figure 1. Valence Frame of the Protests
H2. Valance Frames of the Protesters
Hypothesis 2 postulates that Hong Kong protesters were more likely to be framed in a
negative way by the South China Morning Post when compared to news reports from The New
York Times and The Guardian. Chi-square test (Table 5) showed the differences among the three
newspapers attained statistical significance (X2=60.24, df=6, p<.001). Pair-wise t-tests (Table 6)
further revealed that South China Morning Post framed the protesters negatively in more news
stories (n=17, 22.4%) than the New York times (n=8, 8.3%) and The Guardian (n=1, 1.3%). On
the other hand, the protested were framed positively more often in The New York Times (n=49,
51.0%) and The Guardian (n=56, 72.2%) than in South China Morning Post (n=12, 15.8%).
Hypothesis 2 was thus supported.
Table 5
Chi-Square Test: Valence Frame of the Protesters
Value
df
Sig. (2-sided)
Pearson Chi-Square
60.244a
6
.000
Likelihood Ratio
65.280
6
.000
Linear-by-Linear Association
19.889
1
.000
N of Valid Cases
249
a. 0 cells (0.0%) have expected count less than 5. The minimum expected
count is 7.94.
Table 6
Crosstab: Valence Frame of the Protesters
Newspaper
Total
New York
Times
The
Guardian
South China
Morning Post
Valence frame of the
protesters
Positive
Count
49a
56b
12c
117
% within
Newspaper
51.0%
72.7%
15.8%
47.0%
Neutral
Count
25a
15a
23a
63
% within
Newspaper
26.0%
19.5%
30.3%
25.3%
Negative
Count
8a
1a
17b
26
% within
Newspaper
8.3%
1.3%
22.4%
10.4%
Unclear
Count
14a
5a
24b
43
% within
Newspaper
14.6%
6.5%
31.6%
17.3%
Total
Count
96
77
76
249
% within
Newspaper
100.0%
100.0%
100.0%
100.0%
Each subscript letter denotes a subset of Newspaper categories whose column proportions do not differ
significantly from each other at the .05 level.
Figure 2. Valence Frame of the Protesters
H3. Valence Frames of the HK government
The third hypothesis states that the news articles in South China Morning Post would
more likely than those in The New York Times and The Guardian to frame the Hong Kong
government and its officials as playing positive roles during the protests. Results presented in
Table 7 provided initial evidence that differences existed among the three newspapers (X2=41.68,
df=6, p<.001).
Pairwise z-tests further confirmed that The New York Times (n=42, 43.8%) and The
Guardian (n=47, 61.0%) reported more stories that framed the Hong Kong government in
negative terms than South China Morning Post (n=15, 19.7%), whereas South China Morning
Post (n=14, 18.4%) framed the Hong Kong government more often in positive terms than The
New York Times (n=2, 2.1%) and The Guardian (n=2, 2.6%). Hypothesis 3 was thus supported.
Table 7
Chi-Square Test: Framing of HK Government
Value
df
Sig. (2-sided)
Pearson Chi-Square
41.676a
6
.000
Likelihood Ratio
41.704
6
.000
Linear-by-Linear Association
4.264
1
.039
N of Valid Cases
249
a. 0 cells (0.0%) have expected count less than 5. The minimum expected
count is 5.49.
Table 8
Crosstab: Framing of HK Government
Newspaper
Total
New York
Times
The
Guardian
South China
Morning Post
Framing of HK
government
Positive
Count
2a
2a
14b
18
% within
Newspaper
2.1%
2.6%
18.4%
7.2%
Neutral
Count
15a
12a
13a
40
% within
Newspaper
15.6%
15.6%
17.1%
16.1%
Negative
Count
42a
47a
15b
104
% within
Newspaper
43.8%
61.0%
19.7%
41.8%
Unclear
Count
37a
16b
34a
87
% within
Newspaper
38.5%
20.8%
44.7%
34.9%
Total
Count
96
77
76
249
% within
Newspaper
100.0%
100.0%
100.0%
100.0%
Each subscript letter denotes a subset of Newspaper categories whose column proportions do not differ
significantly from each other at the .05 level.
Figure 3. Framing of HK Government
H4. Framing of news censorship
The fourth hypothesis stipulates that news censorship in Hong Kong and China would be
less likely reported in South China Morning Post than in The New York Times and The Guardian
in their coverage of the Hong Kong protests. The hypothesis was supported by the overall Chi-
square test (X2=24.76, df=6, p<.001) and pairwise t-tests results showed in Table 9 and 10
respectively. Specifically, critical statements and sentiments about news censorship in Hong
Kong or China were more often presented in stories from The New York Times (n=35, 36.5%) and
The Guardian (n=28, 36.4%) than from South China Morning Post (n=7, 9.2%).
Interestingly, none of the three newspapers reported any supportive views of news censorship in
their stories. Hypothesis 4 was therefore supported.
Table 9
Chi-Square Test: Framing of News Censorship
Value
df
Sig. (2-sided)
Pearson Chi-Square
24.761a
6
.000
Likelihood Ratio
29.039
6
.000
Linear-by-Linear Association
17.250
1
.000
N of Valid Cases
249
a. 6 cells (50.0%) have expected count less than 5. The minimum expected count
is .61.
Table 10
Crosstab: Framing of News Censorship
Newspaper
Total
New York
Times
The
Guardian
South China
Morning Post
Framing of News
censorship
Positive
Count
0a
0a
0a
2
% within
Newspaper
0.0%
0.0%
0.0%
0.8%
Neutral
Count
2a
0a
1a
3
% within
Newspaper
2.1%
0.0%
1.3%
1.2%
Negative
Count
35a
28a
7b
70
% within
Newspaper
36.5%
36.4%
9.2%
28.1%
Unclear
Count
57a
49a
68b
174
% within
Newspaper
59.4%
63.6%
89.5%
69.9%
Total
Count
96
77
76
249
% within
Newspaper
100.0%
100.0%
100.0%
100.0%
Each subscript letter denotes a subset of Newspaper categories whose column proportions do not differ
significantly from each other at the .05 level.
Figure 4. Framing of News Censorship
H5. Mentioning of Politically Sensitive Issues
According to the fifth hypothesis, politically sensitive topics such as human rights,
democracy, civil disobedience, references to the Tiananmen Square Movement and criticism of
Chinese communism would be less likely reported in South China Morning Post than in The New
York Times and The Guardian. The Chi-square test results shown in Table 11 indicated that the
hypothesized difference among the three newspapers reached statistical significance (X2=53.10,
df=2, p<.001). Pairwise z-tests results presented in Table 12 lent more support to the hypothesis:
Significantly more stories in The New York Times (n=59, 61.5%) and The Guardian (n=54,
70.1%) mentioned the politically sensitive issues than South China Morning Post (n=12,
15.8%).
Table 11
Chi-Square Test: Politically Sensitive Issues Mentioning
Value
df
Sig. (2-sided)
Pearson Chi-Square
53.098a
2
.000
Likelihood Ratio
56.986
2
.000
Linear-by-Linear Association
31.939
1
.000
N of Valid Cases
249
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count
is 37.85.
Table 12
Crosstab: Politically Sensitive Issues Mentioning
Newspaper
Total
New York
Times
The
Guardian
South China
Morning
Post
Politically sensitive
issues mentioning
Yes
Count
59a
54a
12b
125
% within
Newspaper
61.5%
70.1%
15.8%
50.2%
No
Count
37a
23a
64b
124
% within
Newspaper
38.5%
29.9%
84.2%
49.8%
Total
Count
96
77
76
249
% within
Newspaper
100.0%
100.0%
100.0%
100.0%
Each subscript letter denotes a subset of Newspaper categories whose column proportions do not differ
significantly from each other at the .05 level.
Figure 5. Mentioning of Politically Sensitive Issues
Chapter Six
Discussion
Results of the content analysis clearly supported the general hypothesis that different
frames were used in the coverage of the 2014 Hong Kong protests. Specifically, few news
articles from the South China Morning Post mentioned such politically sensitive issues as human
rights or democracy (15.8%), and most of the stories in the same paper tended to frame the
protests as economically or politically detrimental to Hong Kong (57.9%). In contrast, more
news reports from The New York Times and The Guardian related the protests to the political
sensitive issues (61.5% from The New York Times and 70.1% from The Guardian), and both
papers showed a higher tendency to frame the protests as beneficial to the Hong Kong society at
large (39.6% from The New York Times and 35.1% from The Guardian). Additionally, more news
articles from the South China Morning Post would frame the protesters negatively (22.4%) when
compared to news reports from The New York Times (8.3%) and The Guardian (1.3%). While
The New York Times (43.8%) and The Guardian (61%) tended to portray the Hong Kong
government negatively, the South China Morning Post presented more stories (18.4%) portraying
the Hong Kong government as playing a positive role during the protests.
Taken as a whole, results of the present study provided additional evidence for the notion
that foreign/western and domestic/Hong Kong news media applied different frames in their
coverage of the 2014 protests. Specifically, western/foreign news media would place greater
emphasis on the positive and beneficial aspects of the protests, while domestic/Hong Kong news
media would emphasize the negative and detrimental aspects. Furthermore, western/foreign news
media would more likely frame the protests as a civil disobedience and democratic movement,
whereas domestic/Hong Kong news media would highlight the anti-government and anti-social
aspects of the event.
It would be tempting to summarily attribute the differences in framing observed in this
study to the differences in political systems under which the three newspapers operate (Siebert,
Peterson, & Schramm, 1956). For example, the more positive and supportive frames used by the
New York Times and the Guardian may be attributed to their libertarian traditions whereby the
press serves as an “extralegal check” on government, exposing corruption and free from
excessive central governmental control. Additionally, the media served in the role of educator,
providing the general populace with information and discussion on matters of public interest. The
more negative frames shown in South China Morning Post, on the other hand, reflect the
authoritarian tradition whereby the state is the highest expression of group organization and that
an individual should be dependent on the state to achieve his/her logical means and ends. When
applied to explaining press-state relations, the authoritarian tradition holds that the press should
serve as the mechanism to “support and advance governmental policies and ideals” (Seibert et
al., 1956, p. 18).
A closer look at the results of the present study, however, suggests that such a
dichotomized (libertarian vs. authoritarian) system account might be incomplete, especially
when used to explain the framing of the protests by the South China Morning Post: Of the 76
articles from the South China Morning Post, 7 (10.5%) did frame the protests positively, 12 (15.8%)
framed the protesters positively, 15 (19.7%) framed the Hong Kong government negatively, 7
(9.2%) framed government-imposed censorship negatively and, perhaps more tellingly, 12 (15.8%)
of the articles did refer to politically sensitive issues.
A more complete and valid explanation of the results should therefore take into account
the political system as well as journalistic values. In their comparative study of Hong Kong,
Japan, People’s Republic of China, and U.S. news media, Ostini and Fung (2002) provided an
alternative and more timely view by classifying national press systems as democratic-conservative,
democratic-liberal, authoritarian-conservative, or authoritarian-liberal.
“Democratic-conservative media systems are those in which the political
system is democratic but the professional values of the majority of journalists
are conservative—that is, the professional system(s) in which they operate
emphasize support of societal status quo. Conversely, in a democratic-liberal
system, dissent and free speech are values supported by both the political
system and the individual journalists within that system. Authoritarian-
conservative systems officially control press content and professional values
within media organizations support such constraints. Authoritarian-liberal
systems are those in which official policies suppress dissent, but individuals
within media organizations support social reform and display such support in
their practice of journalism” (p. 48).
Based on empirical data, Ostini and Fung (2002) conclude that countries may “share
similar structural factors in which individual journalists operate under different levels of
professional autonomy” (p. 55). Japan and the U.S., for instance, are structurally similar in that
they both operate within a democratic-liberal system. The authoritarian-conservative press
system in China, however, is different from the authoritarian-liberal press system in Hong Kong.
As stated earlier, Hong Kong is a semi-autonomous district in China and it is governed
under the principle of “one country, two systems.” Such structural constraints, however, cannot
prevent Hong Kong citizens from expressing their liberal and critical thoughts, largely as a result
of over 150 years of British rule. It should further be noted that, under Article 27 of the Hong
Kong Basic Law, Hong Kong residents enjoy freedom of speech, press, and publication, and
these rights are generally upheld by the territory’s independent courts. And ironically, Article 35
of the current Chinese Constitution also stipulates that "Citizens of the PRC have freedom of
speech, publication, assembly, association, procession and demonstration," despite the
government’s severe restrictions on freedom of speech and expression.
The authoritarian-liberal press system in Hong Kong thus creates a conflict between the
political structure and journalistic values, and the conflict was clearly manifested in the frames
used by South China Morning Post in its reporting the 2014 Occupy Central protests. On the face
of it, there seems to be no censorship in Hong Kong – freedom of the press in Hong Kong is
theoretically guaranteed by law. As a “special administrative region” of China, however, the
Hong Kong government faces mounting pressure from China to have more control over its press.
The pressure results in indirect or self-imposed news censorship in Hong Kong, a region that had
once enjoyed unfettered Western-style press freedom under British rule.
The co-existence of authoritarian and liberal views of the press in Hong Kong, and the
resultant self-imposed news censorship thus provide the best available explanation of the mixed
use of positive and negative frames found in the present study. The explanation is corroborated
by a recent report form Reporters Without Borders (2014). According to the report, many Hong
Kong news organizations have been imposing news reporting restrictions since the 1997
takeover. Even though there had been scant news reports in Hong Kong newspapers on political
sensitive issues like human rights, these reports would generally show favorable attitudes toward
the government and condemn the instigators who openly violate the state policy.
It is worth noting that self-censorship in Hong Kong has gradually spilled from news to
other speech-related industries. For example, bookstores in Hong Kong are also exercising self-
censorship by refusing to sell books or magazines that contain unfavorable comments toward the
Hong Kong or Chinese government. A recent article (Sala, 2015) in the Guardian, for instance,
reports that several bookshops in Hong Kong refused to sell books that reveal government
scandals in exchange for a more favorable property rent.
An open question is when and to what extent Hong Kong is able to maintain its
authoritarian-liberal press system before succumbing to the authoritarian-conservative press
system in mainland China under which all news stories on the Hong Kong protests, online or
offline, before or after 2014, were strictly prohibited – with the only exception of the official
Xinhua agency that conveniently and conclusively framed the protests as “illegal.”
Chapter Seven
Conclusions
In this thesis, a content analysis was done on the newspaper coverage of the Hong Kong
protests, one of the most newsworthy events in 2014. Results of the analysis proved that different
news frames did exist in news reports from different newspapers that embodied different
professional values and structural constraints. The most noteworthy finding is the co-existence of
authoritarian and liberal views of the press in Hong Kong, as reflected in the types and valence of
the frames used in reporting the protests.
Like all empirical research, the present study has its limitations. Although the study tried
to capture the essence of the protests by investigating news articles published during September
29 and December 18, 2014, the protests continued afterwards. Indeed, the movement the protests
embodied shall continue for years, if not decades. Case in point: CNN (2015) recently reported
that the Chinese government’s decision on the Hong Kong electoral reform, the main trigger of
the protests, was rejected by the Legislative Council on June 18, 2015. Likewise, China
NewsAsia (2015) reported on the silent rally outside Hong Kong government headquarters on
September 28, 2015, to mark the first anniversary of the start of the Occupy Central pro-
democracy protests. Other stories convey the same message: Hong Kong citizens will never stop
calling for greater democracy. More research is therefore needed to cover the movement after the
2014 protests to document the continuing struggle between the government and the press, as well
as the undying hope for more freedom.
Future research should also build on the present study by examining additional media
sources (e.g., weblogs and social media) and their frames of the movement. Quantitative research
methods (e.g., surveys) should be employed to provide additional insight pertaining to the
changes in public mood or opinion as the result of framing. In doing so, a comprehensive picture
of frame building (how frames are being chosen by journalists in news reporting) and frame
setting (how frames exert their influence on audiences’ attitudes and opinions) may eventually
emerge.
