Article Text

Download PDFPDF

Newspaper reporting of suicides in Hong Kong, Taiwan and Guangzhou: compliance with WHO media guidelines and epidemiological comparisons
  1. King-Wa Fu1,
  2. Yuen-Ying Chan1,
  3. Paul S F Yip2
  1. 1Journalism and Media Studies Centre, The University of Hong Kong, Hong Kong, PR China
  2. 2HKJC Centre for Suicide Research and Prevention, and Department of Social Work and Social Administration, The University of Hong Kong, Hong Kong, PR China
  1. Correspondence to Dr King-Wa Fu, Journalism and Media Studies Centre, The University of Hong Kong, G24 Eliot Hall, Pokfulam Road, Hong Kong, PR China; kwfu{at}


Background Media guidelines for suicide reporting are available in many countries. However, to what extent the mass media comply with the guidelines is unknown. Few studies are available that investigate systematically whether the mass media reflect the epidemiological reality of suicide deaths in their articles.

Methods Based on the WHO media guidelines, this study investigated the characteristics of newspaper articles of suicides in three Chinese communities, namely Hong Kong, Taiwan and Guangzhou. Epidemiological comparisons were conducted to identify the age and gender differences between the suicide victims as reported in the newspapers and the official records of suicide deaths in all three places.

Results The results found that one media characteristic complied with the WHO media guidelines (ie, only about 2% of the articles were printed on the front page), but there were a number of instances of non-compliance (ie, only 4–14% provided sources for help-seeking and 27–90% printed with photos). The epidemiological comparisons revealed an over-representation of younger suicides and an under-representation of late-life suicides in the newspapers of all three places. Furthermore, female suicides were found to be under-reported in Taiwan and Guangzhou newspapers, but not in Hong Kong papers.

Conclusion Non-compliant suicide articles are prevalent in the newspapers of these three Chinese settings. The observed media misrepresentations may potentially mislead the public and the policy makers about the actual risk for suicide in some demographic groups.

  • Media misrepresentation
  • suicide
  • Hong Kong
  • Taiwan
  • Guangzhou
  • communication
  • suicide SI

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Suicide is a significant global public health problem. Every year, about 1 million people die prematurely by suicide worldwide, contributing 1.5% of the overall global burden of disease.1 The aetiology of suicidal behaviour is complex and is known to be multifaceted, including biological, psychological and sociocultural factors.2

Suicides of ethnic Chinese have accounted for more than a quarter of global suicides and demonstrate a unique epidemiological pattern and demographic profile.3 4 Previous studies noted marked discrepancies in the suicide rates within the People's Republic of China (hereafter China) between rural and urban areas, as well as between genders.5 In the other two Chinese communities, Hong Kong and Taiwan, their increases in suicide rates since 1997 were mostly attributed to the economic downturn arising from the Asian financial crisis.6 7 A comparative study reveals that the variations in epidemiological characteristics of suicides between Hong Kong, Taiwan and Beijing were ascribed to their demographic variations between geographical areas and the extent of their societal changes over decades.8

Among the contributing factors for suicidal behaviours, the media's influence has become a global agenda for suicide prevention.9–12 In previous decades, a substantial number of studies have been devoted to investigating the influence of media on suicidal behaviours, examining newspaper and television reports of actual suicides as well as film and television's fictional portrayals of suicides. By analysing 293 published studies on media influence on suicides quantitatively,13 Stack suggests media coverage is a ‘risk factor in suicide.’14 As put by Hawton,15 there is ‘overwhelming evidence,’ on the basis of his and other scholars' systematic reviews, to support the association between media depictions of suicides and subsequent increases in suicide rate,14–16 especially when the media stories report suicide deaths of celebrities.17–19 However, Hawton and Williams also argue that such media influence must be contextualised under a variety of psychosocial and situational factors that may contribute to suicidal behaviours.11 In addition, the theory of active audience, postulating that media users have abilities to reject media's effect, ignore media messages or interpret media texts in their own manner, poses a challenge to the notion of direct media effects on an individual's suicidality.20

There has been a long-lasting debate on the ways in which suicides are professionally reported in the mass media. Many journalists believe that self-discipline is sufficient to protect vulnerable readers against harmful effects, and media professionals are generally inclined to reject any form of regulation. Some assert it is only a matter of presentation style, and only those sensational media accounts deserve to be condemned.21 Some media professionals even deny accepting the negative consequence of media effects.22 This appears to reflect a wide gap between evidence-based findings and journalistic practices.

In response to the media's influence on suicidality, the WHO has suggested ‘minimising unnecessary reporting of suicides in the popular media’ as a national strategy for preventing suicides.23 The WHO media guidelines for journalists (appendix) recommend that media avoid glorifying and sensationalising suicide deaths, detailing suicide methods used and placing suicide stories prominently.24 As the WHO media guidelines further suggest, media can play key roles in educating the public about suicide and communicating various risk factors for suicide.24 Those roles aim to help the community identify at-risk individuals as early as possible, encourage people with needs to seek professional help and destigmatise suicide in the society. However, very few evaluation studies have been conducted to investigate whether the actual media presentation of suicide complies with what the media guidelines suggest.10

Moreover, many scholars have recognised that health-related news articles are frequently biased towards emphasising certain aspects or conceptions of health problems or under-representing some important issues or alternative viewpoints.25 For instance, some causes of death, including intriguing causes such as unintentional injuries or homicide, are disproportionally covered in the news media compared with their actual mortality, and on the contrary, deaths from some chronic diseases, for example heart disease, are under-reported in the media.26 By comparing news stories of intimate partner violence with epidemiological data, one research finding indicates that the cases involving homicide or use of alcohol are overly represented in the media, and consequently that tendency may skew the public perception of risk in the context of domestic violence.27 Another study finds the mass media selectively report sensational and unusual parricide (children killing a parent) cases while referencing national crime statistics.28 An analysis of cancer news coverage reveals an under-reporting of the morbidity and mortality of lung cancer and an over-representation of breast cancer cases in the local and national media.29

However, very few studies are currently available to examine systematically if there are any misrepresentations of suicide deaths in the mass media. Among these limited studies, one suggests that Hong Kong newspapers tended to report news articles about suicide victims who had suffered from relationship problems.30 A content analysis of newspaper suicide articles in Taiwan finds notable discrepancies between general and tabloid-style papers in some media characteristics, such as pictorial presentations of suicide victims and the location where the suicide took place and the placement of articles in the newspaper.31 Another study of four Chinese newspapers reveals an under-reporting of suicides occurring in the rural areas where most of the suicides in China took place.32

With these backgrounds, this study aims: (1) to investigate the compliance of newspaper reporting of suicides in three Chinese communities, namely Hong Kong, Taiwan and Guangzhou, with the recommendations made by the WHO media guidelines; (2) to compare the age–gender profiles of suicide victims as reported in the newspapers with those of suicide deaths records collected from the official death registries in the three Chinese communities under study.


Content analysis was conducted to capture the media characteristics of the suicide articles which were collected from the selected newspapers in Hong Kong, Taiwan and Guangzhou (a major city in southern China, which has a well-established media market and is geographically close to Hong Kong and Taiwan). These three Chinese communities have a similar ethnic composition, written language and cultural background but have a different pace in terms of their political systems and media markets.33 Based on a global comparative study,34 Hong Kong and Taiwan media are known as the societies with high levels of press and editorial freedom, and China is ranked as a place where their citizens enjoy least freedom of press.

Table 1 presents the names of the selected newspapers, which altogether constitute the large portion of newspaper market share in each place. The samples could be considered representative for each geographical region. For example, the five Hong Kong newspapers together accounted for about 80% of the local newspaper market share.30

Table 1

Frequencies of suicide articles in Hong Kong, Taiwan and Guangzhou newspapers from January 2006 to June 2006

The operationalisation of media characteristics of suicide articles was referred to the WHO media guidelines.24 Media guidelines are available in many different countries. But according to a review of nine media guidelines sampled worldwide, their contents are more or less the same.10 The WHO media guidelines are chosen as the reference for this study because these guidelines were developed as a global initiative for preventing suicide and are recommended by the WHO and the International Association of Suicide Prevention, which are authoritative organisations on suicide prevention.35

The measurement items include date of publication, name of the newspaper, number of Chinese characters, type of suicide (either completed or attempted), placement (either front page or not), gender of the victim, age group (under 15, 15 to 24, 25 to 39, 40 to 59 and 60 or above), suicide methods used (jumping, charcoal burning, hanging, poisoning or others), whether information about the victim's gender, age, suicide method or alleged problem encountered was found from the article headline, whether the article contained pictorial, graphical or tabular presentation, and whether the article contained preventative advice, which is defined as information about a helpline, community resource, risk indicators, warning signs, alternatives to suicide, professional opinions and psychiatric illness symptoms.35

Owing to resource constraints, the duration of analysis was limited to 6 months, from 1 January 2006 to 30 June 2006. Suicide articles published from 1 January 2006 to 30 June 2006 were retrieved from an online Chinese newspaper database, namely Wisenews, which stored most of the printed version of newspapers used in this study in digital format. The only exception was Taiwan's the Liberty Times, whose suicide articles were collected from the paper's official website. The inclusive criterion was that the news article must include at least one Chinese keyword related to suicidal behaviours, including ‘suicide,’ ‘jumping from a height,’ ‘hanging’ or ‘charcoal burning’ (suicide by carbon monoxide poisoning). Editorials, commentaries, fictional stories and media accounts about suicides occurred in other countries were excluded. Printed versions of newspapers were visually inspected by the coders to retrieve those non-keyword-searchable articles.

Eight coders were recruited to code the retrieved suicide articles. If more than one victim was reported in the article, the one who had more salient media coverage was chosen to be coded. A pilot study of a sample of 80 suicide articles showed that the κ coefficient of each measurement item exceeded 0.9, and their mean was 0.94, indicating good inter-rater reliability.

χ2 statistics or a Fisher exact test was used to detect the statistical differences between the categories of nominal or ordinal variables. The mean differences in continuous variables between groups were analysed using a univariate analysis of variance (ANOVA).

Official suicide data on Hong Kong and Taiwan in 2006 were collected from the Hong Kong Coroner's Court and the Department of Health of Taiwan respectively. Since suicide data in Guangzhou are not publicly available, suicide data on China (the most recent figures were from 2000) collected from the WHO Statistical Information System were used instead, with an assumption that the deviations in suicide patterns between Guangzhou and the whole of China, if any, could have not affected the study outcomes.


A total of 2279 suicide articles were retrieved. These articles consisted of 226 (10%) from the Guangzhou newspapers, 1086 (48%) from the Hong Kong newspapers and 967 (42%) from the Taiwan newspapers. The mean number of suicide articles per day per paper in Guangzhou, Hong Kong and Taiwan was 0.21, 1.21 and 1.34, respectively. The frequencies of suicide articles for each paper are shown in table 1. Their media characteristics are presented in table 2.

Table 2

Characteristics of suicide articles (%)

Testing compliance with media guidelines

As shown in table 2, a majority of the suicide articles in the Guangzhou and Taiwan newspapers were the reports of completed suicide deaths, contrary to the Hong Kong papers in which the articles of attempted suicide were more frequently found. Moreover, 1.9% and 0.9% of the suicide articles were printed on the front page of the Hong Kong and the Taiwan newspapers respectively, and yet the Guangzhou newspapers did not run any front-page suicide articles during the study period.

The Guangzhou newspapers carried relatively lengthy articles, over 50% of which had more than 600 Chinese characters, but in contrast, only 20% of the suicide articles in the Hong Kong and the Taiwan newspapers contained more than 600 Chinese characters. Suicide articles in the Hong Kong newspapers were less wordy than the articles in other places, and more than 43% of them had fewer than 200 Chinese characters.

Approximately 90% of the suicide articles in the Hong Kong newspapers were accompanied by pictorial presentations, but only 34% and 28% were found to have accompanying images in the Guangzhou and Taiwan newspapers respectively. The Taiwan and Hong Kong newspapers used tables and graphical descriptions more often than did the Guangzhou newspapers. Moreover, the Hong Kong newspapers had a lower percentage, only 4%, of suicide articles providing preventative advice, such as helpline numbers or information on referral services, than did the newspapers of Guangzhou and Taiwan, 13.7% and 12.7% respectively.

Analysing the headline contents as shown in table 2, headlines of suicide articles in the Hong Kong newspapers were found to carry the victims' demographic identities, gender, age, suicide methods used and alleged problem encountered more frequently than the article headlines of the Guangzhou and the Taiwan newspapers. For example, over 91% of headlines in the Hong Kong newspapers reported the suicide methods used, compared with 76% and 70% of those in the Guangzhou and the Taiwan papers respectively. Furthermore, 53% of headlines in the Hong Kong newspapers mentioned the alleged problem encountered by the suicide victims, whereas only 38% and 33% of those in the Guangzhou and the Taiwan papers, respectively, reported this piece of information.

Epidemiological comparisons

The age–gender profiles of the completed suicide victims (excluding attempted suicides) portrayed by the newspaper articles in Guangzhou, Hong Kong and Taiwan were compared with the demographic statistics of suicide deaths recorded by the official sources. These comparisons are made as shown in tables 3, and 4.

Table 3

Epidemiological comparisons of gender and age profiles between completed suicide victims reported in newspapers (excluding attempted suicides) and suicide death statistics from official sources, N (%)

Table 4

Epidemiological comparisons of age–gender profiles between completed suicide victims reported in newspapers (excluding attempted suicides) and suicide death statistics from official sources, N (%)

Table 3 illustrates that the suicide articles in the newspapers of Guangzhou and Taiwan under-reported the female suicides (p<0.05), that is a lower percentage of the reported female suicides articles than that of the official female suicide deaths. However, no such disparity in genders was observed in the Hong Kong newspapers.

With respect to the discrepancy in age, the newspaper articles in all three places tended to under-represent the suicides of late-life adults who were aged above 60 years (p<0.05), and to overemphasise the youth suicides who were aged below 25 years (p<0.05).

On closer examination of the discrepancy in specific age–gender groups between the reported suicide articles and the official records of suicide deaths as shown in table 4, such a discrepancy was marginally statistically significant among Hong Kong males (p=0.03) and was substantial among Hong Kong females and both genders in Guangzhou and Taiwan (p<0.05).


Based on the criteria set by the WHO media guidelines, this study first examined the ways in which the suicides were characterised by the Guangzhou, Hong Kong and Taiwan newspapers, and tested their compliance with the guidelines. Epidemiological comparisons were then made to identify the differences between the age–gender profiles of the suicide victims as reported in media accounts and the demographic statistics obtained from the official suicide death records.

Compliance and non-compliance

Based on the findings, one characteristic of the suicide articles in the Guangzhou, Hong Kong, and Taiwan newspapers was found to be compliant with the WHO media guidelines, but many of the others were not. First, a low percentage of the suicide articles were prominently printed on the newspaper front page. Among the three places, the Hong Kong newspapers demonstrated the most frequent front-page suicide articles, amounting to 1.9% of the total. At face value, less than 2% of the total suicide articles featured on the front page were not deemed high. But while considering suicide is a relatively rare event and referencing to typical front page news topics with public interest and societal importance such as politics or election, one suicide article out of every 50 newspaper's front pages could not be considered a small number. Admittedly, it is hard to interpret this number quantitatively without comparing with the proportions of front-page coverage of other newsworthy topics. That comparison exceeds the scope of this study but could be a question for the future; second, less textual and more visual presentations were observed. Suicide articles had on average about 400 Chinese characters per article; 58% were accompanied by pictorial presentations or illustrations, compared with 14% of the suicide articles in Australian newspapers36 and 8% in Swiss newspapers.37 In comparison, the Hong Kong newspapers had the highest percentage of suicide articles printed with photos, which was close to 90%; third, the headline of suicide articles, which is the most eye-catching and visually arousing position of a news story, was found to carry many important attention-seeking features, including the suicide method used and the alleged problem encountered by the victim, which could heighten an individual's suicidal thoughts.20 A majority, about 81%, of the article headlines reported the suicide method used, and about 43% reported the alleged problem encountered by the victims, even though the death verdict was typically not yet issued and certificated by coroner or medical professional at the time of newspaper publication; fourth, the Hong Kong newspapers had substantially fewer articles on preventative advice than did the Guangzhou and Taiwan papers. In an American study conducted with a similar methodology, preventative advice for suicide was offered in about 13–15% of all suicide articles,38 which is very close to the percentages found in the Guangzhou and Taiwan papers but far higher than that in the Hong Kong papers.

One must be aware of the vagueness of some media guidelines. For example, it is suggested that ‘reporting specific details of the method used’ and ‘giving simplistic reasons’ be avoided, but what exactly constitutes ‘details’ and ‘simplistic’ remains unclear. Without a clearer specification, media professionals may find the guidelines difficult to follow.

Moreover, we call for more future studies on comparisons of suicide articles within and between nations as well as thorough analyses of suicide articles in a different medium, such as film39 and the internet.40 Future studies using content analysis may include more variables, for example principal motives for suicide, and making use of the newer revision of the WHO media guidelines.24

Epidemiological comparisons

Scholars have recognised the disparity between mediated reality and the physical world, as well as the media's effect on an individual's perception of social reality.41 42 For example, cultivation theory suggests that a person's view towards a social issue can be shaped by media exposure and thus can deviate from social reality.43 44 Empirical studies find that media misrepresentations could skew the public perception of risk of social problems or health issues.27–29

The findings of this study indicate that the newspaper articles of all three Chinese settings tended to underplay the late-life adult suicides, aged 60 or over, but to over-represent youth suicides younger than 25. These media misrepresentations with respect to age are consistent with a similar analysis in Israel.45 On the contrary, Pirkis and colleagues46 report an over-representation of elderly suicides, aged above 64, and an under-reporting of youth suicides, who were aged under 25, in the Australian newspapers. Pirkis and colleagues46 account for their results by stating that the Australian media may be more sensitive to suicide contagions among the youths who seem to be particularly vulnerable to copycat suicides.

We make the supposition that the reasons for the observed discrepancies in the Hong Kong, Taiwan and Guangzhou newspapers would be because the Chinese media tend to consider youth suicides as more newsworthy (relatively rare event) but deem late-life adult suicides less newsworthy (relatively common). In addition, youth suicides could be reported in a sensational manner in the media to attract readers. Unlike their counterparts in Australia, media professionals in Chinese settings seem to have a lower awareness of the negative impact of suicide contagion among youths.

The under-representation of late-life adult suicides in the media can have mixed meanings. On the one hand, the suicide rate of older adults is relatively higher than that of other age groups in all three places.8 Media under-reporting may therefore misinform the public and policy makers about the risk and the severity of older suicides. As a result of this misinformation, government's policy priority may be misguided. Even worse, early identification of geriatric depression or suicidal risk among older individuals may prove less efficacious because the at-risk individual's relatives, friends or other gatekeepers have underestimated the actual risk of older suicides. But from another viewpoint, senior people have relatively less media exposure to suicide articles about victims who are in their same age group, and consequently become more immune to imitative suicide. It is of significant and practical value to explore the effective ways of reporting suicide in the media for fulfilling both objectives: minimising risk of imitative suicides and promoting public awareness of suicide problem.

Moreover, female suicides in Taiwan and Guangzhou were under-reported in the media. Hong Kong media was distinctively found to have no such disparity. Comparatively, no such under-representation of female suicides is found in Australian and Israeli newspapers.45 46 We speculate that the misrepresentation of female suicides is attributable to the traditional Chinese thinking that it is prone to stereotyping women as relatively more emotional or hysterical, as well as suicidal, than men, prejudicing female suicides to be seen as less unusual, and thereby less newsworthy in relatively traditional Chinese settings such as Taiwan and Guangzhou. However, this does not seem to be the case for Hong Kong where people in this international city generally adopt more global values, such as gender equality; further evidence is needed to support this speculation.


Some shortcomings of the comparative approach are worth noting. The measurements might be incomparable between different communities, and this could threaten the validity of the study; however, the selected variables were mostly manifest items that could maximise the measurements' reliability and validity. The study period was in the first 6 months of 2006, and the seasonal fluctuation of suicide articles was assumed to be minimal. The suicide articles were retrieved electronically by keyword searches, and the accuracy of the searches depended largely on the keyword strings used and the capability of the third-party newspaper archive system. The suicide data in China were slightly outdated, and the national suicide victim profile might deviate from that of the Guangzhou suicides, but the dataset used in this study is currently the best available data source.


Non-compliant suicide articles are found to be prevalent in the Hong Kong, Taiwan and Guangzhou newspapers. This phenomenon may be associated with the market competition as well as the unique sociocultural background in Chinese communities. It may also partly contribute, though further study is required, to the higher-than-global-average suicide rate in Chinese settings.

As observed in this study, the media misrepresentations of younger aged suicides and late-life adult suicides could lead to a biased social attitude towards over- or underestimation of the severity of suicide problems with respect to specific age groups. Such media distortion is not conducive to establishing a better-informed society and at worst could mislead policy makers in prioritising policy direction. Understanding these media misrepresentations can contribute to policy formulation for suicide prevention. Journalists must be alert to these misrepresentations in order to minimise reporting biases.

What is already known on this subject

Mass media can play a key role in educating the public about suicide and may help to prevent suicide, but health-related news articles are frequently biased towards emphasising certain aspects or targets.

What this study adds

Non-compliant suicide articles with the WHO media guidelines are prevalent in Hong Kong, Taiwan and Guangzhou newspapers. Some demographic groups are misrepresented in these media accounts. Such media misrepresentation may mislead the public and the policy makers about the actual risk for suicide of these groups.


We wish to acknowledge the contributions made by D Chan, V Chan, G Ip, V Ko, J Lam, J Lee, G Wong, J Yip, KH Law and SP Lin, for their assistance in data collection and coding of newspaper articles.


What to do

  • Work closely with health authorities in presenting the facts

  • Refer to suicide as a completed suicide, not a successful one

  • Present only relevant data, on the inside pages

  • Highlight alternatives to suicide

  • Provide information on helplines and community resources

  • Publicise risk indicators and warning signs

What not to do

  • Don't publish photographs or suicide notes

  • Don't report specific details of the method used

  • Don't give simplistic reasons

  • Don't glorify or sensationalise suicide

  • Don't use religious or cultural stereotypes

  • Don't apportion blame

Source: WHO. Preventing Suicide: A Resource for Media Professionals: Department of Mental Health, Social Change and Mental Health, WHO; 2000.



  • Funding This work was supported by the Hong Kong Jockey Club Charities Trust.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.