Abstract
Studies about media self-censorship typically focus on its mechanism in traditional newsroom settings. But how media self-censorship may evolve in online journalism has remained largely unexplored. Using Hong Kong as a case, I examine the digital evolution of media self-censorship in a unique non-democratic context. Drawing on interviews with online journalists, my findings reveal that digital transformation has provided new valences for media self-censorship. With the financial hardship of legacy media in the digital age, Hong Kong online journalists are more directly exposed to external threats such as advertisement boycotts orchestrated by the state, and hence increasingly reluctant to offend external powerholders out of the fear of political and financial retaliation. Moreover, as online journalists adopt business-driven norms that favor the generation of clicks, political or policy news are further marginalized. These stories are often deemed boring, non-engaging to online audiences, and are not “sensationalizable” due to political risks, especially when compared to soft news types like crimes and lifestyles stories. Adapting to these changes, news managers are increasingly used to avoiding professional editorial debates that results are unpredictable but using “objective” web metrics as persuasive devices to discourage the production of sensitive news. Lastly, the dissemination of sensitive news is curbed in the social media gatekeeping process. These findings suggest that an authoritarian state can effectively influence online news production by controlling the capital that drives digital transformation, thereby limiting the liberating potential of the media in the digital age.
Keywords
Introduction
Since Hong Kong's reunion with China in 1997, its press freedom has slowly eroded 1 . Although China did not set up an official censorship system in Hong Kong, it has indirectly influenced Hong Kong news media through various means. A key strategy is the acquisition of media companies by businesspeople who have substantial business interests in China (Frisch et al. 2018). These changes in ownership often entail processes of re-negotiation of editorial directions and practices that lead to the cultivation of media self-censorship (Fung 2007).
Despite this, Hong Kong journalists still subscribe to the professional conception of journalism and see monitoring powerholders as part of their role (Leung and Lee 2015; So and Chan 2007). These professional beliefs may prevent journalists from self-censoring; Hong Kong news organizations, therefore, typically cultivate media self-censorship by tweaking professional notions (Lee and Chan 2009). Research suggests that pro-China media owners often appoint trustworthy journalists as top editors to control newsrooms indirectly. Questionable orders of self-censorship would then permeate through the newsroom hierarchy. These suspicious orders are usually inexplicit, coughed in professional terms, and established as unspoken rules to avoid resistance from professional-minded journalists (e.g., Au 2017; Lee 2007; Lee and Chan 2009; Lee and Lin 2006).
Political pressure, however, is not the only challenge to Hong Kong news media. With a population of 7.51 million, Hong Kong has one public broadcaster, two licensed radios, six licensed TV channels 2 , and seventeen major daily newspapers (Frisch et al. 2018). Yet, these statistics have not included many online-only media. Unlike traditional media, these online-only outlets operate with lower costs and are structurally more independent from external control. Their emergence unsurprisingly lured the lion's share of online readership from legacy media (Wang 2018), rendering digital transformation a more difficult task for them.
In Mainland China, how digital transformation may bring opportunities and challenges to journalists’ everyday work has long been discussed. Some scholars are optimistic that digitalization and online market competition may open up new possibilities for Chinese journalists to resist censorship (e.g., Hassid 2015; Hassid and Repnikova,2016; Stockmann 2013). Others argue that the Chinese government can contain challenges brought by the digital transformation (Lee et al. 2006; Wang and Guo 2021) and that pressures from online competition may even weaken journalistic autonomy (Wang and Sparks 2019). However, in the context of Hong Kong, how digital transformation may affect news media has remained underexplored.
Using Hong Kong as a case, scholars have offered valuable insights on how an authoritarian government may exercise de-facto control over news media in a legal system that purports to protect press freedom (e.g., Au 2017; Lai 2007; Lee and Chan 2009; Lee and Lin 2006). However, most of these studies situate their analyses within traditional newsroom settings. In this paper, I shift the attention to the cultivation of self-censorship in the online news production of Hong Kong legacy newspapers, where journalists distribute news via digital channels like web portals and social media. Specifically, I ask how the political economy in Hong Kong reshapes media self-censorship at a time when digital technologies are incorporated into daily news-making. Based on interviews with twenty journalists who actively participated in online news production, I aim to offer a useful update on the changing media dynamics in Hong Kong within the context of digital transformation.
In what follows, I first review existing studies of media self-censorship in non-democratic contexts, particularly China and Hong Kong, and examine political-economic and technological factors that might influence Hong Kong online journalism. Focusing on the Chinese language legacy press, I then present a case study that examines how digital transformation complicates media self-censorship in the unique authoritarian context of Hong Kong. I show that digital affordances and pressure of online competition have provoked strategic shifts in newsrooms and channeled political-economic risks to online journalists at unprecedented levels, thus creating new valences for media control. Finally, I discuss the implications of these findings for press freedom in Hong Kong.
Self-Censorship in Authoritarian Contexts
Media self-censorship refers to editorial practices that individual journalists or news organizations employ to curry rewards and avoid punishment from external players (Lee 1998, 2019). Unlike direct censorship, self-censorship is conducted internally by news media without receiving explicit orders about what not to write (Simons and Strovsky 2006).
Media self-censorship can develop anywhere, even in consolidated democracies. So far, however, scholars have paid considerable attention on media self-censorship in societies that still suffer from various threats to press freedom. For instance, studies find that journalists self-censor to avoid risks of violent attacks in Mexico (Hughes and Márquez-Ramírez 2017) and Africa (Frère 2014) and to avoid judicial sanctions in Turkey (Yesil 2014). Sometimes, scholars even describe self-censorship as a side-effect of censorship regimes (Yesil 2014).
There are several reasons why authoritarian governments cultivate self-censorship even when official censorship is in place. For one, no written rules could pre-emptively cover all possible scenarios. Hence, enforcing all-encompassing censorship might be inefficient and costly (Gehlbach and Sonin 2014). Also, outright repression may invoke resistance and damage the government's reputation (Lee 2019). Given these concerns, authoritarian governments may instead prefer journalists to censor themselves.
In the case of China, the official censorship system is also far from all-encompassing (e.g., Gueorguiev et al. 2017; Lorentzen 2014; Roberts 2018). Studies find that the Chinese government usually avoids blatantly removing previously available information to prevent incentivizing citizens to circumvent censorship, such as getting news from foreign sources (Roberts 2018). Instead of applying absolute media control, the Chinese government often carefully calibrates the magnitude of censorship to prevent backfire and ease social tensions (Lorentzen 2014).
As a result, the cultivation of self-censorship remains an essential means of media control in China. Scholars suggest that the Chinese government keeps off-limits topics purposefully ambiguous in the sanctum of censorship because clear boundaries of permissible journalism may help journalists confirm what is “safe” and prompt more critical views (Gueorguiev et al. 2017). Because flashes of repression are not entirely predictable, there are always deep-rooted uncertainties embedded in Chinese journalists’ daily work. These uncertainties, in turn, instill fear and cultivate self-censorship (Stern and Hassid 2012). Hence, it is almost impossible to pinpoint what counts as “sensitive” by the Chinese government, as it does not simply issue a list of forbidden topics. Instead, political red lines in China are redefined constantly through strategic interactions among journalists, internal censors, and external powerholders (Repnikova 2017; Stern and Hassid 2012). Simply put, every story related to politics can be deemed sensitive in the Chinese context (Lee and Chan 2009).
Self-Censorship in Hong Kong
Although press freedom has been gradually eroded in Hong Kong, a free press tradition still characterizes its news industry. Speaking from a legal perspective, Article 27 of the Basic Law, the mini-constitution of Hong Kong, guarantees freedom of publication in the city. Therefore, there is no newspaper licensing system imposed. Moreover, Hong Kong journalists still maintain a strong sense of professionalism. In journalism schools, professionalism has long been a staple of the curriculum. Local journalists also uphold professional norms like objectivity and balanced reporting (Lee and Chan 2009). In general, studies confirm that Hong Kong journalists believe they must write professionally, play a watchdog role, and report based on facts (Leung and Lee 2015; So and Chan 2007). Therefore, explicit censorship orders may provoke resistance from professional-minded journalists.
The lack of a censorship system and the persistence of journalistic professionalism thus present unique challenges to Hong Kong news outlets. Hong Kong news organizations have to figure out how to cultivate self-censorship as the de-facto norm of newsroom operation within a de-jure system of the free press where most journalists believe in the professional conception of journalism. One key strategy is the acquisition of media ownership. By the mid-2000s, pro-China businesspeople have already acquired ownership of most Hong Kong legacy newspapers (Frisch et al. 2018). These owners could then appoint trustworthy journalists as top editors to control newsrooms indirectly. As Lee and Chan (2009) reveal, self-censorship orders would then filter down the newsroom hierarchy. News managers may strategically assign inexperienced or conservative journalists to take charge of sensitive stories, as they are typically easier to control. Usually, they disguise suspicious orders in professional terms. Supervisors may delete political criticisms that are not “objective,” impose superficial “balance” in stories covering the wrongdoings of powerholders, and put on hold or truncate sensitive stories out of excuses like limited airtime or space. Over time, reporters would have to read between the lines to learn these unspoken rules through observation.
As an example, Lee and Lin (2006) demonstrate how a reputable Hong Kong newspaper, Mingpao, positioned itself as an impartial arbitrator in the debate of democratic reform. By claiming “objectivity,” Mingpao was able to restrict the scope of their reportage to merely communicate opinions from different sides instead of providing critical views on democratic reforms. Some researchers even argue that self-censorship has already been normalized or constitutionalized without anyone being aware of it (Au 2017).
But professional notions, like “objectivity,” are also strategic devices that journalists can use to defend themselves against external pressure (Tuchman 1972). In Hong Kong, strategic-minded journalists have used “objective” polling figures to indirectly express citizens’ distrust of the government instead of giving direct criticism (Lee and Lin 2006). Legacy newspapers also “outsource” their opinions to talk radio by selectively incorporating comments from the shows to reflect defiant voices without adopting a reporter-driven approach (Lee 2012). Moreover, if reporters can support their stories with solid professional arguments, they would become difficult to censor (Lee and Chan 2009).
Overall, there is a strong scholarly assumption that professionalism, as a strategic device, is central to the dynamics of media freedom in Hong Kong. Using professional creeds as the boilerplate language of self-censorship, professionalism creeds and self-censorship can hence coexist peacefully in the newsroom.
Self-Censorship in Digital Contexts
While much has been written about how the political economy in Hong Kong shapes practices of self-censorship in legacy media, very few studies have considered influences from digital transformation. Some studies about self-censorship in Hong Kong were conducted when Hong Kong legacy media had yet to readily incorporate digital technologies, like social media, into their trade (e.g., Lai 2007; Lee and Chan 2009; Lee and Lin 2006). On the other hand, some studies have sought to spotlight specific forms of news, such as radio reporting (Lee 2012) and television news (Au 2017), rather than news produced online and distributed through digital channels. Therefore, it remains unclear how digital transformation influences the cultivation of self-censorship in Hong Kong online news production.
By “digital” or “online” journalism, here I refer to the modes of news production and distribution made possible by computer programming that transcend the temporal and spatial constraints of traditional media, in which news production and distribution become spontaneous, individualized, and networked. As Peters (2016) suggests, computer programming enables such experiences by inventing digital means for people to communicate instantly at all times. News organizations can now distribute their products via digital channels, like websites and social platforms, that transcend the physical limits of print newspapers. Moreover, traces of digital communication on these channels are countable and indexable. Delivering, consuming, or responding to news could now be coded into quantifiable metrics that index the external reality and are retrievable with a click.
These digital affordances affect many aspects of journalism. Using digital performance indexes, news managers could now track the performance of individual stories and monitor reporters more efficiently (Bunce 2019). Web metrics and audience feedbacks are increasingly influential in editors’ selection of topics, headlines, and story assets (Ferrucci 2020; Lee and Tandoc 2017). As online readership grows, the shrinkage of print readership has also weakened the independence of journalists from advertisers in some newsrooms (Goyanes and Rodriguez-Castro 2019). Thus, there are reasons to believe that media self-censorship, just like other journalistic practices, would also be influenced by this technological shift.
In 2020, about 85% of Hong Kong citizens regularly use online sources for news, including social media, with 58% of social media users reading news on Facebook and only 31% of the population still reading print news. 3 In 2019, digital ad spending in Hong Kong had amounted to $1.15 billion. 4 However, tech giants like Facebook and Google are allegedly the biggest beneficiaries. They claimed more than 60 percent of the total digital ad revenues in the Asia Pacific, 5 resulting in a lack of sustainable digital business models for legacy newspapers in Asia. As Hong Kong legacy newspapers restrict political criticism in their reportage, many citizens have turned to alternative digital-only news channels for alternative views. These outlets, such as The Stand News, InMediaHK, and Initium Media, typically operate on a smaller scale and are thus structurally less susceptible to external control, so they could cater to a particular gratification niche of users by providing critical views absent from the mainstream media (Luqiu 2017; Wang 2018).
So far, how these changes have influenced the professional autonomy of Hong Kong journalists has received little scholarly attention. However, in the broader Chinese context, how digital technological shifts have impacted the development of the professional autonomy of journalists has long been a debated topic. Optimists argue that digital transformation has challenged media control in China because journalists may heed to more critical forms of journalism to gratify demands from an increasingly diversified public and compete for online readership (Hassid 2015; Hassid and Repnikova 2016; Stockmann 2013). On the other hand, others counter that the government did not loosen its political control of Chinese newspapers during their marketization and digital transformation (Lee, et al. 2006; Wang and Guo 2021). Some even content that fierce online competition might precipitate financial problems for legacy newspapers, thus weakening their professional autonomy (Wang and Sparks 2019). What remains unclear is how we may apply these insights to the case of Hong Kong, where its journalism is also heavily pressured by the politics of authoritarian China and fierce competition of online readership.
Using Hong Kong as a case, this study thus provides a helpful case analysis about the influences of digital transformation on journalistic autonomy in a unique authoritarian context. Seeking to explore how Hong Kong legacy newspapers maintain a delicate balance between self-censorship and professionalism amid rapid technological transition, I ask:
RQ: Would digital transformation challenge or open up new possibilities of political control in the newsrooms of Hong Kong legacy newspapers?
Method
Selecting Outlets
I collected my data in Hong Kong from June 2018 to December 2018 through in-depth interviews with legacy newspaper journalists who actively participate in online news-making. I applied a few criteria to pick the sampling outlets.
First, as the purpose of this study is to explore how digital technologies and their corresponding impacts on the economics of newspapers may reshape self-censorship practices, I purposefully excluded non-commercial press. As profit-making is not a crucial task for non-commercial newspapers, they likely view the competition of readership starkly differently from other commercial outlets. Thus, I exclude non-commercial press, such as the state-owned and subsidized Ta Kung Pao and Wen Wei Po.
Second, as there is a scholarly assumption that Hong Kong legacy media cultivates self-censorship mainly through the tweaking of professionalism, I deliberately excluded newspapers that report with overt partisan intent, such as Apple Daily, Oriental Daily, and SingTao Daily. For example, Apple Daily is a pro-democracy advocacy newspaper known for its anti-government stance. It almost always adopts critical approaches against the Chinese and Hong Kong governments, and even printed slogans like ‘Go to the street’ on headlines to mobilize anti-government protestors. For newspapers with strong political penchants like Apple Daily, ideological principles may overwhelm professional notions like “objectivity” in their editorial decision-making. They would also be more or less likely to succumb to government pressures, rendering their self-censorship dynamics dissimilar from other non-partisan outlets. Therefore, this study deliberately excludes partisan newspapers.
Finally, this study deliberately excluded English newspapers. Hong Kong English newspapers typically target elite readers and foreign audiences (Guo 2010; Lin and Jackson 2020). As a result, journalists working for English newspapers are “more diverse and intend to provide a balanced picture to serve a growing global readership” (Lin and Jackson 2020: 9). Therefore, I excluded English newspapers to ensure that this study conceptually focuses on a group of most similar outlets.
Eventually, I concentrated on a group of non-partisan Chinese commercial newspapers that adopt a relatively centric stance, including Mingpao, Hong Kong Economic Times (HKET), and HK01. Of them, Mingpao is a typical professional non-partisan newspaper (e.g., Fung 2007; Lee and Lin 2006). Scholars also regard HKET as a professional legacy newspaper with a “centrist political stance” (Huang and Song 2018: 25). Similarly, scholars describe HK01 as having a politically “even-handed” stance (Nip 2019: 402).
Selecting Informants
Since this study focuses on online news production, I purposefully selected informants responsible for either producing texts, images, videos, or any multi-media content dedicated to online publishing or distributing news via digital channels such as social networking platforms. There is a clear division of labor between online and print news production in all three sampled outlets. In other words, these outlets solely assigned their online news production and distributing tasks to specific teams. Therefore, I sampled informants from these specialized teams.
Mingpao designates its online news production tasks exclusively to one specialized team: the instant news team. These tasks involve the production of instant news, posting and promoting print stories online, and managing the outlet's flagship Facebook page, Mingpao Instant News, which was established in 2013 when Mingpao decided to engage in the competition on social media. Mingpao transferred most members of this team internally. Although editorial materials are freely exchanged between this team and print colleagues, their collaboration is voluntary and sporadic.
The case of HKET is similar. As a financial daily, HKET mainly produces financial news dedicated to the niche market of investors. In 2014, the outlet joined the competition of the online mass market by establishing TOPick, a general digital news channel that produces all types of news with the potential to become “top picks” on the Internet but does not participate in the print news-making of HKET. The TOPick team is also responsible for managing HKET's most-followed Facebook page, TOPick News. All founding members of TOPick are from HKET. Although TOPick regularly borrows HKET's stories and other editorial materials, their collaboration remains sporadic.
HK01 is a digital-leaning news outlet established in 2016 with online and offline presences (publishing a weekly broadsheet newspaper). HK01 structures its newsroom very similarly to large-scale legacy outlets. It adopts a traditional beat system and hires more than 200 staff reporters. HK01 recruited most of its staff from other legacy newspapers, including Mingpao and HKET (Huang and Song 2018: 26 and 91). HK01 assigns print and online news tasks exclusively to different teams like the other two outlets. Similarly, these online and print journalists may share editorial materials but rarely collaborate.
From the perspective of newsroom print-digital integration, the models of Mingpao, HKET, and HK01 all fall into the category of “co-ordination of isolated platforms” (García Avilés et al. 2009), in that their online and offline news production remain coordinated but separated. Therefore, online journalists of Mingpao's instant news team, TOPick, and HK01 all work in a solely digital news production environment. Since most of them are transferred or recruited from newspapers, these online journalists typically carry professional mindsets developed from their print experience. Studying these print-turned-online journalists, thus, creates an excellent opportunity to observe how traditional journalists experienced digital transformation.
In all, I interviewed twenty journalists: seven from Mingpao, six from HKET, and the remaining seven from HK01. In terms of their positions, three were top editors, nine were mid-level editors, and eight were frontline reporters. I interviewed a top editor who supervises the whole outlet for each newspaper to incorporate perspectives from journalists working at different levels. All informants are Hong Kong journalists with print experience, except two recent graduates from journalism schools. I began my recruitment through personal networks. Then, the informants recommended others in a snowball manner (See Appendix A for further information).
Interview and Coding
Since I expected my informants to talk about sensitive internal information, I promised them anonymity. Each interview took 2–3 h. The interviews were semi-structured, adopted an open-ended strategy to encourage unguided introspection, and were audio-taped and transcribed. I applied a two-cycle coding process for the analysis (Miles et al. 2014). After going through the transcripts, the first-cycle codes were identified, then grouped according to their themes in the second-cycle coding. I constantly compared and recoded new and existing codes. Finally, I translated the Cantonese interview excerpts into English.
Findings
Pressures from the Business Side
With the shrinkage of print readership in the digital age, researchers have observed that the commercial pressure to look for viable online business models are increasingly pressuring journalists and has incentivized them to change their reporting priorities (Bunce 2019; Goyanes and Rodriguez-Castro 2019). Similarly, the testimonials of my informants point to the reality that commercial consideration is now fully integrated into their daily trade. A consensus from the informants is that print journalists traditionally were reluctant and rarely required to attend business meetings. But business meetings about online news production are now ordinary. Moreover, as they revealed, online journalists have become increasingly approving of attending these business meetings. When asked why, they frequently referred to the financial crisis they faced: while the traditional print business is shrinking, competition from alternative online media is fierce. Two informants commented: I don't want to be the traditionalist—being very defensive, and would reject all suggestions from the advertising department…Without them, we suck (a mid-level editor).
We have only limited resources….We need to know what content could help our business; I no longer ask why we need to meet with the business team (a web news team head).
However, as online journalists begin to legitimate business considerations, they are inevitably exposed to unprecedented political-economic risks, as pressures channeled from the business side make them wary of the financial consequences of their works. They do not only worry about offending individual advertisers, such as private companies; now, business-minded journalists worry most about the state, as the state in the Chinese context could disproportionally influence the advertising market.
A telling example is a boycott suffered by Next Media, the publisher of the major pro-democracy newspaper Apple Daily. It had reportedly suffered huge advertising losses due to a boycott orchestra by the Chinese government,
6
despite its massive success in the online competition.
7
When my informants explained their financial concerns, they repeatedly mentioned the experience of Next Media. Thus, its story has become a strong argument against the production of contentious news online, no matter how many clicks these stories may generate. A senior web editor further illustrated: Apple Daily captures the largest online market share in Hong Kong…If they still suffer tremendous pressure from advertising boycott, we would naturally infer that any website keen on political news, especially one that reports with a pro-democracy stance, will share the same fate.
Indeed, in day-to-day exchanges between online journalists and sales representatives, online journalists may now receive direct reminders about the consequences of advertising boycotts. As a mid-level editor recalled, they were once explicitly told that critical views against the government may backfire: They told us candidly why we get advertisements. Because we offer great content but also not producing too much political news… In the business world, [companies like] the Bank of China, China Construction Bank are all owned by Beijing. If you publish stories full of criticisms, who dare to place ads?
Therefore, many informants deem attracting online readership by providing critical views as financially unwise for large-scale commercial media because they could hardly convert those clicks into advertising income. A top news manager described that the story of Apple Daily is a cautionary tale: “They are too radical, the advertisers simply do not accept that.”
Here, what motivates online journalists to self-censor is not strict orders from the business departments. Instead, the pressure to seek viable online business models convinced them to prioritize business concerns voluntarily. Indeed, reminders from their business colleagues fall short of being coercive, as almost all informants insisted that they only follow suggestions when they see fit since online journalists do not report to their business colleagues.
Of course, it may seem too simplistic, or even naïve, to believe journalistic autonomy and the consideration not to offend powerholders could coexist without struggles. The question “how to ensure autonomy when you need to please advertisers” turns out to be a very difficult one. The most common response was “it depends.” An editor elaborated, “Are we pressured? It depends on you…The business department only gives you suggestions.” Nonetheless, it remains difficult for my informants to pinpoint precisely how to balance autonomy and business/political consideration. Some simply said they need to “survive in the competition” to justify potential compromises. On the whole, they did not give any explanations better than vague statements like “we will strive for balance overall” or “decisions are made on a case-by-case basis.”
Avoiding Sensitive News
Digital journalism studies usually suggest that sensational reporting approaches are now more liberally employed in online news production to draw more clicks. Online media now treats “hard” news topics like government affairs and political controversies just as sensationally as traditionally “soft” news categories, such as crime or lifestyle news. (Garcia-Perdomo et al. 2018; Kilgo et al. 2018). However, this is not the case for my informants.
In general, Hong Kong legacy news newspapers now encourage online journalists to report more sensationally. However, such sensational treatments in online political news, such as spotlighting controversies in headlines, sharp criticisms, or writing satirically, are still easy targets for reprimands. To illustrate what may happen if journalists cross the line, an interviewee used the editorial intervention regarding a “sensational” social media headline as an example. When their team published a story about the re-election of Vladimir Putin for his fourth term on March 18, 2018, they wrote a line “King Putin Forever? (普帝永續?)” as the promotional rhetoric of a Facebook post that provided a link to the news. Their supervisor quickly scolded them that the statement was too opinionated and unprofessional. Later, they replaced the original expression with an emotionally detached news excerpt.
The motivation behind the intervention, as the informant second-guessed, is the fear of offending the Chinese government. On March 11, 2018, a week before the Russian election, China also removed its presidential term limits. Indeed, China has already announced the intent to make such a constitutional amendment in late February. At that time, the public broadcasting service of Hong Kong, Radio Television Hong Kong (RTHK), published a post on Facebook with a hashtag “King Xi Forever? (習帝永續?)” to question the constitutional amendment. After being heavily criticized by the pro-Beijing camp, RTHK took down the post, halted the renewal of all Facebook pages, and reformed its policy to prevent future incidents. The informant opined that the fear of similar political retaliation triggered the intervention, “I think, they don't want others to associate the post to Xi.”
Sometimes, warnings are pre-emptive. Supervisors may remind online journalists to treat sensitive stories “professionally” in advance if the timing of an event is known. Two examples an informant used were the sentencing of rioters who participated in democratic protests and a court verdict that involved companies owned by the states. The informant said, “[Supervisors would say,] hey, this is a serious topic, don't play tricks.”
Consistent with Lee and Chan’s (2009) argument that propagators of self-censorship often issue suspicious commands in ambiguous terms, it remains not entirely clear what counts as “playing tricks” and what “tricks” are unprofessional. Sometimes, writing critically and analytically is not easily distinguishable from writing sensationally. Therefore, the safest strategy for online journalists to avoid conflicts and reprimands is to return to the “traditional”—write in plain language when reporting sensitive political news. A social media editor elaborated: If the story is sensitive, I won't dare test the limit. But if it's a car accident, if it is just about the details of the wedding of Prince Harry, I would feel safe to evoke emotions from social media users.
Reprimands from supervisors can be highly selective. Bosses of online journalists usually condone sensationalized treatments of non-sensitive news topics, such as crime or lifestyle news, but they routinely regard sensationalism in political news reporting as unprofessional. This double-standard strategy has inevitably motivated online journalists to evaluate political and non-political stories differently. My informants repeatedly mentioned animal rights, thrilling crime, and inspirational stories as ideal story types for online use because they could report these stories sensationally to draw clicks without getting reprimands. In contrast, because political stories are often plainly written, online journalists generally perceive them as boring, having no hope of going viral online, and not worth producing.
Nonetheless, it is hard to imagine how online journalists could avoid writing sensitive political news altogether. As many informants admitted, complete avoidance of sensitive issues may damage their professional reputations. Despite being marginalized, political stories have never ceased in their online news repertoires. Still, propagators of self-censorship need additional strategies to restrict the reporting of sensitive issues. One such strategy is to “soften” those sensitive issues by shifting attention to interesting trivialities. The example that best demonstrates this was in 2017 when President Xi Jinping visited Hong Kong for the first time since becoming the president of China to join the 20th anniversary of the handover. Instead of focusing on substantive issues, like Xi's speech on national security controversies, the team of an informant turned its attention to write about trivialities: After Xi landed, someone carried a luggage case with him. Some colleagues found that the case is bullet-proof. Then we noticed that his bodyguards all wore sunglasses—is it because they don't want to be recognized? We visited Xinhua (the official news agency of China) to find the answers. Turns out it's because the bodyguards want to hide sightlines…In the end, we didn't write much about what Xi did.
Another case is a feature interview about the retirement of the vice-chancellor of a prominent university. During his incumbency, controversies about freedom of expression on campus were common. For instance, this university was among several institutions that insisted on taking down banners bearing “Hong Kong independence” on campus. But as an informant recalled, their news team deemed these substantive issues “not engaging” and deemphasized them as side stories. Instead, they prioritized interesting trivialities in repertoires—how the vice-chancellor exercised to lose weight so that he could fit in the new Korean-style slim-fit suits that he just bought.
Overall, these findings suggest that the tweaking of professional notions remains a vital tactic of media self-censorship in Hong Kong online news production. Using professionalism as an excuse, news managers have been restricting the application of critical tones and sensational reporting tactics in online political reportage. However, the emerging culture favoring clicks generation among online journalists has seemingly cultivated a new self-censorship mechanism. Because political stories are usually blandly written, online journalists often perceive them as boring, non-engaging, and without any viral potential on social media. Therefore, when online journalists prioritize interesting trivialities and eschew political news, they are not only motivated by the fear of offending external powerholders. They also want to avoid spending time and resources on products that do not fit with the social media culture and are unlikely to go viral.
Web Metrics and Social Media Gatekeeping
When asked about their professional integrity as journalists, many editors defended their decisions to avoid or soften political news as bias-free. The most common response is that these editorial judgments are data-driven and made solely based on web metrics. This implies another strategy to limit the agency of journalists: weaponizing digital data or web metrics as tools of manipulation. As a social media editor revealed, traditional-minded journalists may occasionally use professional notions like “newsworthiness” to challenge his decisions, such as bargaining limited Facebook promotion slots for blandly written political stories that are unlikely to go viral. However, he usually refuse to join such a debate. Instead, he prefers to use web metrics as “objective criteria” to justify himself: Someone may think they are more experienced in making editorial decisions…you know, traditionalists love to talk about experience? So, when you ask why I’m not promoting your stories, I’d refer to the figures. Hey, last time your story did not reach the standard, it fell below the average engagement rate. So, sorry, I can't help you. I have a target to meet.
In a similar vein, a top news manager who oversees the whole outlet admitted that they may now refer to web metrics as an alternative justification of editorial decisions. This is especially the case if similar stories previously did not go viral. “No need for coercion, I just give them figures and persuade them with the business reality.”
Conventionally, rules of self-censorship are known to be ambiguous and consensual (Lee and Chan 2009). However, these findings suggest another possibility. Unlike fluid criteria such as “newsworthiness,” web traffic data provides solid and quantifiable criteria for editorial judgments. As web metrics are seemingly objective, news censors could claim bias-free decisions by referring to figures. By doing so, news managers could avoid debates carried out in ambiguous professional terms where results are unpredictable. They can now use web metrics as an alternative source of authority in editorial debates.
Of course, it is still difficult to snub well-crafted sensitive news. As noted in the literature (Au 2017; Lee and Chan 2009), a conventional resistance strategy for non-conforming is to offer solid arguments and evidence to support a story. Because Hong Kong journalists generally share similar professional backgrounds, it is difficult for anyone to reject highly important and professionally well-prepared stories. As there is no shortage of space online, news managers also cannot use reasons like “not enough space/airtime” to kill stories. When reporters insist on publishing on a sensitive topic that was professionally written, they will probably get their way.
Yet, propagators of self-censorship develop tactics correspondingly. If news managers think a story may draw political backlash but is too solid to be rejected, they might give it a “cold-shoulder treatment” in terms of online promotion. Simply put, they will publish these stories without giving them any app-push notification and social media promotion, or they may only promote them on social media pages that have fewer followers. A news manager candidly elaborated: We use various means to make them less viral… otherwise, the media owner would feel threatened. So, the best scenario is that the story reaches fewer people; to an extent the media owner does not even know it is published. Then, everyone is happy, and balance is achieved.
By limiting their visibility and reach online, legacy news media can bury politically sensitive stories quietly. Because these stories are eventually published, supervisors of online journalists need not give explicit censorship orders that harm their reputation or get themselves entangled in debates that they may lose. If non-conforming journalists still argue against their decisions, news managers may then refer them to the web metrics to reject their pleas.
Discussion
Previous studies about media self-censorship in Hong Kong have mainly situated their analyses in the setting of traditional newsrooms (e.g., Au 2017; Lai 2007; Lee and Chan 2009; Lee and Lin 2006). But now, most news outlets have incorporated digital technologies into their trade. Focusing on online news production in legacy newspapers, this study explores how media self-censorship in Hong Kong has evolved in the digital age. It provides a useful update on the changing media dynamics in Hong Kong within the context of digital transformation and its growing authoritarianism. The findings show that media self-censorship in Hong Kong is transitioning toward a hybrid culture where new digital and old traditional practices coexist.
As the evidence suggested, the tweaking of professionalism is still an essential means of self-censorship in online news production of Hong Kong media. By reminding online journalists to “not play tricks” and write “professionally,” legacy newspapers can inhibit political criticism in their online reportage and ensure online journalists write political stories plainly and without critical views. These findings extend the assumption that media self-censorship in Hong Kong is exercised by manipulating professional notions to online settings (Au 2017; Lee and Chan 2009).
But digital transformation also envisages new possibilities for media self-censorship. Two key factors are the availability of web metrics and the business pressure to generate more clicks. As my finding suggests, the struggles of Hong Kong legacy newspapers to survive in the digital competition have incentivized online journalists to legitimate business-driven norms. Whether a story would go viral online is now an important concern for online journalists. Therefore, the standards of “good journalism” in online news production have been gradually changing—from emphasizing news that is important to news that would go viral. Therefore, when online journalists perceive these blandly written political stories having no hope for online virality, avoiding or “softening” these stories by diverting attention to interesting trivialities has become increasingly justifiable.
Hence, these online journalists do not just eschew sensitive political news out of the fear of offending external powerholders but also the financial concern to avoid spending resources on non-profitable products. For them, the avoidance of sensitive news is both a tactic to prevent administrative punishments and an adaptive financial strategy to cope with the changing economics of newspapers in the digital age. In addition to tweaking professional principles, Hong Kong legacy newspapers now also employ non-professional notions to justify self-censorship practices in their newsrooms.
These findings have important implications for Hong Kong journalism. First, most sensitive news topics are related to the wrongdoings of powerholders. If news media deprioritize these stories or avoid reporting them through a critical lens, then the watchdog role of Hong Kong media will be further weakened. Second, if Hong Kong legacy newspapers constantly divert online readers’ attention away from substantive issues, they might no longer provide citizens with sufficiently reliable information to make sound political decisions.
This study also speaks to journalism studies in the broader Chinese context. Many critics argue that digital technologies may provide new channels for Chinese journalists to evade political control, and online market competition may motivate them to broaden their range of journalism (Hassid 2015; Hassid and Repnikova 2016; Stockmann 2013). The experience of Hong Kong suggests otherwise. Because the Chinese government can disproportionally influence the advertising market in Hong Kong, it can consequently exert a strong influence on news organizations struggling to survive in the fierce competition of online readership. My findings thus echo the notion that fierce online competition may weaken the financial independence of legacy newspapers, as well as the professional autonomy of their journalists (Wang and Guo 2021; Wang and Sparks 2019).
Beyond the Chinese context, this study contributes to the literature of newsroom social control in authoritarian contexts. The findings expand the traditional scholarly understanding that rules of self-censorship are usually consensual and unwritten. Online journalists may now debate their digital editorial decisions in more concrete terms by using web metrics as “objective criteria” to judge the worthiness of a story. Researchers of newsroom social control often argue that professional norms such as “objectivity” render journalists’ behaviors more predictable and easier to control (e.g., Soloski 1989; Herman and Chomsky 2010; Hearns-Branaman 2014). But no matter how predictable it is, professionalism is still riddled with grey areas. In contrast, web metrics are quantifiable and seemingly unbiased, making them irrefutable as performance indicators (Bunce 2019). By manipulating the interpretation of these figures, the management can now monitor and discipline their journalists more efficiently.
Certainly, these findings are confined to a specific context that concerns media self-censorship in online news production of Hong Kong Chinese commercial newspapers. Future studies could compare non-partisan and partisan newspapers, Chinese and English press, or trace signs of self-censorship in other types of Hong Kong media such as broadcasting organizations. They could also explore the transferability of my findings to other authoritarian regimes, especially where the governments can exert disproportional influence on the capital market.
It is also worth bringing to the fore the newly imposed Hong Kong national security law (NSL). The NSL was imposed unilaterally by China on July 1, 2020, in response to the massive Anti-Extradition Law Amendment Bill Movement in 2019. The NSL stipulates that crimes of secession, subversion, terrorism, and collusion with foreign forces are punishable by a maximum sentence of life imprisonment. 8 Critics generally deem the NSL as a turning point of the city's freedom. Drawing on insights this study provided, it might be possible to extrapolate how the NSL would further usher changes in media self-censorship in Hong Kong.
Most fundamentally, the NSL might rip out some grey areas involving sensitive news topics. As my findings suggest, propagators of self-censorship now tend to leverage “objective” criteria than fluid professional notions to support their editorial decisions. Following similar logic, they would also likely take political redlines stipulated by the NSL as the boundaries of acceptable journalism. As the NSL officially criminalizes certain political deviances, the advocacy of separatism or regime changes is clearly off the table. When dealing with “forbidden” topics, news managers may simply avoid any debates of newsworthiness and readership but refer to the potential judiciary sanctions to snub these topics.
However, many crimes in NSL are vaguely defined. For instance, Article 29(5) of the NSL stipulates provoking hatred against the government as illegal 9 , but what counts as “provoking hatred” is not entirely clear with its vague terminology 10 . Therefore, the strategic interactions between Hong Kong journalists and internal censors will continue, in which the observed patterns of media self-censorship, i.e., self-censorship by professional and non-professional business notions, would likely stay relevant in the foreseeable future.
Conclusion
This study provides empirical evidence to contend that the digitalization of legacy news media are not automatically emancipating in nondemocratic regimes. Whether they have beneficial consequences heavily depends on the political economy of media. Unlike Western countries, where news outlets compete in a relatively free market, Hong Kong legacy media operates in an environment where commercial advertising is more dependent on political favors from the state. When print readership shrinks and newspapers struggle to find online business models, they thus become more vulnerable to external control. Although digital technologies present dialogical channels for the public to influence journalists, they do not necessarily motivate news media to broaden their range of journalism. As pro-China businesspeople have acquired many Hong Kong news organizations, they typically appoint trustworthy editors to manage the newsroom. Then, politically conservative editors can calibrate and manipulate their interpretation of web metrics to snub sensitive news.
In sum, external powerholders can effectively influence online news production in Hong Kong commercial newspapers by controlling the capital and advertising market that drive digital transformation, thereby limiting the media's liberating potential. Consequently, all these render the relations between media control from the state, digital technologies, and market competition more collaborative than contesting in the case of Hong Kong.
Footnotes
Appendix A
See Table A1.
Acknowledgments
The author would like to express his gratitude to the editors and three anonymous reviewers for their constructive comments on the manuscript. The author also thank Omar DumDum, Professor Sue Robinson, Professor Zhongdang Pan, and Zening Duan for their support and insightful comments on earlier versions of the manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
