Contrasting Frames: Visual Coverage at Urban and Regional News Outlets in Australia and China

ABSTRACT

Visual news can be a broad topic and encapsulate a myriad of forms, conventions, and representational content. The present study delves deeply into the image-based side of news to explore how visual news is produced, framed, and presented at four print and digital news outlets in urban and regional parts of Australia and China. We first conducted a qualitative denotative, stylistic-semiotic, and ideological framing analysis of a total of 1,408 images, published as part of 674 articles. Next, we deepened our understanding of the image analysis results through semi-structured interviews with 14 editorial staff in the visuals departments at these four outlets. Using framing theory and the hierarchy of influences model as theoretical lenses, this study uncovers how economic, social, and political factors affect the types of visual stories that journalists in these countries cover and sheds light on how those stories are presented. Specifically, our results indicate that outlets in Australia illustrate their news more than comparable outlets in China and that the types of visuals used also differed considerably. Online templates and editor directives influenced visual news in Australia to a greater degree while comparable Chinese news outlets paid more attention to audience expectations and political considerations.

KEYWORDS:

• Comparative news analysis
• hierarchy of influences
• image analysis
• visual journalism
• Chinese and Australian news
• urban and regional news
• photojournalism
• visual news

Introduction

A common refrain in scholarship that examines visual communication phenomena is to comment on the degree to which contemporary society is inundated with images (Aiello and Parry 2019). Yet, at the same time, many studies that deal with this subject matter take images—the most basic building block of most multi-modal media expression—in a monolithic fashion rather than critically interrogating generalised assertions to see how they differ across contexts, formats, genres, and platforms (Aiello and Parry 2019). Taking this provocation to heart, this study purposively selects two countries, China and Australia, with a shared geographic affiliation in the Asia-Pacific but with markedly different systems of government and cultures, to explore how journalists make and use images used in comparable print and digital outlets in four locations (one urban and one regional in each country). Doing so allows us to underscore how various cultural, economic, and ideological differences are reflected in the visual news both countries make and whose audiences consume. It also adds to the scholarship on image-based news in China, extends scholarship on urban journalistic practice to regional areas, and focuses on routine, daily coverage rather than on coverage of special events that has previously occupied research attention. Focusing on regional in addition to urban news is important for representativeness and to help address an imbalance in the scholarly literature. It also allows for an exploration of how the organisational level in the hierarchy of influences model is affected by larger market forces and greater resources in urban areas compared to smaller markets and fewer resources in regional areas.

This study brings in journalism scholars from Australia and China to research the process and practices of visual news making, using a comparative and contextualist approach. Such two-country comparisons provide a valuable tool to “theorize relationships between contextual influences and the object of investigation” (Esser and Hanitzsch 2012, 8; Guo, Holton, and Jeong 2012). Australia and China are rich sites for such comparative analysis as they enjoy close economic ties and geographic proximity but are also drastically different in terms of population (25 million compared to more than 1.4 billion), government (democracy compared to a socialist state), and social policy in areas such as religion, the death penalty, and treatment of political prisoners and ethnic minorities (Department of Foreign Affairs and Trade 2022). These differences extend, also, to the countries’ media systems.

The article is structured as follows: it begins with a literature review of visual news framing and visual news and culture in Australia and China. It proceeds with a theoretical framework drawing on the hierarchy of influence model. We then overview the research methods we have adopted in this project and present the findings from both a visual framing analysis and an interview-based qualitative analysis to underscore the similarities and differences between how visual news is made and appears in these countries.

Literature Review

The theoretical framework for this study centres around two thematic topics: we use framing as a theoretical perspective to analyse the (stylistic-semiotic, denotative, and ideological attributes of) news images made during the study’s sampling time frame and then contexualise those results with interviews with Chinese and Australian journalists through the lens afforded by the hierarchy of influences model, which we introduce later.

Visual News Framing

Framing in the communication and journalism literature refers to the means through which people organise their subjective experiences and communicate those to others (Goffman 1974). Frames are meaning-making devices that people use to parse the issues and events they encounter (Reese, Gandy, and Grant 2001). Framing highlights some attributes of a scene, situation, or person while downplaying or outright hiding others. This can affect how the message is interpreted by those who see it (Aday 2006; Schudson 2001). The work of identifying and selectively shaping specific aspects of an issue or event is a routine part of news work (Greenwood and Thomson 2020; McQuail 2010). In such contexts, journalists and news workers evaluate different features of a story against an institutionally defined list of characteristics that make a story “newsworthy” and then configure the story from that perspective in a way that will meet perceived expectations of the audience. This was the case, for example, when U.S. journalists pivoted framing the Iraq war in 2003 from a conflict to a human-interest focused frame (Schwalbe, Silcock, and Keith 2008). Because news audiences tend to not interpret news stories as social constructions but instead generally accept them as descriptions of reality (Gamson 1992), frames are effective and powerful communication devices.

The same tendency to accept the content of the message applies to news visuals. People, at least in Western countries, generally accept photos and videos as being closer to truth than written or spoken communication because of the perceived mechanical recording ability of the camera (Messaris and Abraham 2001). News visuals, thus, frame less obtrusively than do written or spoken messages. However, scholars have paid more attention to the framing of words compared to the framing of visuals (Bock 2020). Visual frames are powerful because they carry “excess meaning” (Rodriguez and Dimitrova 2011, 50) by activating related experiences and memories more strongly than words alone and by being more memorable (Vaidya and Gabrieli 2000). As noted previously, visuals require less cognitive processing power than words to understand, and the content of images is often more readily familiar to viewers than the content of text, making images more accessible and influential to an audience. In this study, we adopt Rodriguez and Dimitrova’s (2011) levels of visual framing approach to analyse three levels of image content: the stylistic/semiotic, denotative, and the ideological. We do this by studying the modality (e.g., high-modality photographs compared to low-modality cartoons) of images published, which allows us to consider the stylistic-semiotic framing dimensions of the images; who or what they represent, which allows us to consider the denotative framing of the images; and the eliteness of who is shown, which allows us to consider the ideological framing of the images. Our visual coding and analysis approach is further detailed in the methods section.

Visual News in Australia and China

The news, and the larger digital media ecosystem, is now profoundly visual: it requires images to attract attention, inspire feeling, and provide more concrete perspectives of newsworthy people, places, and events (Thomson 2019). The rise of social media as a vehicle for news coverage and dissemination has only accelerated this shift toward a visual mode of information consumption (Hand 2017). This is felt by audience members and journalists alike. Text-based news remains plentiful, but is less memorable and persuasive (Fahmy, Bock, and Wanta 2014): news organisations and platforms are thus forced to provide not only written stories, which rely on a rational-cognitive processing model, but also visual stories, which rely on a more immediate, intuitive-emotive model, in order to capture and sustain audiences’ attention (Joffe 2008).

A 2020 survey of 307 regional journalists in Australia revealed that “the need for more visual content” was one of the biggest changes to their jobs over the past five years (Fisher et al. 2020). These journalists nominated photography as one of the most important skills for a regional reporter, prioritising it above audio, data journalism, data visualisation or graphics, web design, and coding/programming. But despite this critical role in news reporting, visual journalists have been among the hardest hit by job cuts over the past 20 years (Caple 2019). In 2020, News Corp shed 200 journalists and announced that it would be out-sourcing photography in Queensland and across the network (Meade 2020). In response to this loss of professional visual news-makers, news organisations are increasingly integrating generic stock or crowdsourced vision into their reporting, without always addressing the ethical circumstances of its production, editing, and circulation (Thomson et al. 2022). This results, in the case of generic stock imagery, in content with minimal relevance and nuance, and subsequently, in an apathetic and disengaged audience (Newman et al. 2021). It also results, in the case of crowdsourced vision, in uneven coverage where social responsibility is replaced by the most viral and commercially profitable content, and in content with questionable provenance being circulated and amplified with sometimes negative consequences (Dootson et al. 2021).

Historically, Australian literature on photojournalism has been scant (Anderson 2014), with recent contributions focusing on metropolitan contexts at the expense of the regions, and on coverage of special events, such as Australia Day or a federal election, rather than on the “largely neglected” area of everyday coverage (Caple 2019, 58).

Academic studies that have focused on visual news in China largely concern television or short video content on social media (Fearon and Rodrigues 2019; Liu and Zhang 2022; Xiang 2019). Analyses of visually oriented news publications are rare (see, however, Chase 2020, as a notable exception), as are studies that look at the images accompanying routine coverage in Chinese news outlets. Photos are “a common yet much neglected component of news” (Dan and Ren 2021, 202). This is especially true in a non-Western context where Chinese news studies “barely ever acknowledge visuals”, which leads to a “limited understanding of photojournalism in China” (Dan and Ren 2021).

One recent Chinese study examined news photos of people living with HIV in one national, three metropolitan, and three provincial newspapers (these last three being the Dahe Daily in Henan; the Nanguo Morning News in Guangxi; and the Spring City Evening News in Yunnan) between 2000 and 2015 (Dan and Ren 2021). Like recent scholarship that focused on visual coverage of national days (Caple, Huan, and Bednarek 2020), this study was also focused narrowly on a specific topic rather than examining routine coverage. It found that journalists most often showed people living with HIV through a “victim frame” and that regional news outlets were more likely to rely on the victim frame than urban news outlets, which were more likely to rely on the survivor frame.

Given the intertwined relationship between visual news and the larger visual culture of both societies, relevant aspects of this relationship will also briefly be discussed. Chinese attitudes toward visuals, in general, have been massively impacted by its socialist visual culture (Tang 2015). This socialist collectivisation and construction has historically affected which visuals are used and how. For example, it is easier to see oneself as a part of a greater whole symbolically through low-modality renderings, including cartoons and simple illustrations, than it is to see oneself through high-modality photographs, which are much more concrete and detailed. These types of visuals advance socialist visual culture and have historically been used by the state for particular purposes (Tang 2015). Low-modality visuals are also easier to express a particular point of view through, compared to high-modality visuals, which can be more polysemous (Crowe 2003). As an example, Tang points to the way visuals in high socialism China—from those seen in cinemas to those found on posters—are “persistently positive and didactical” (11), which help “construct a unified and progressive national culture.” Posters, in particular, Tang argues, “project a positive outlook and communal values; they also valorise ready accessibility and intelligibility over solipsistic depths or diffusive abstractions” (24).

In Australia, visual culture is less homogenous and reflects the more individualistic nature of Australian society (Hinkson 2009). This is bolstered by Australia’s greater access to the internet compared to China and to the diversity of viewpoints and visual styles this affords (Hinkson 2009), as well as its capitalistic orientation. It also reflects the “larrikin” nature of Australian society that embraces anti-authoritarianism, audacity, and a disregard for political or social conventions (Vine and Vine 2021). Photography offers a powerful medium for expressing and representing this individuality.

The Hierarchy of Influences Model

Various micro-, meso-, and macro-level forces shape news content. The hierarchy of influences model (Shoemaker and Reese 2013) focuses systematic attention on these forces by using levels such as individual, routine, organisational, extra-media, and ideological. Journalists’ individual attributes and identities, such as their age, ethnicity, and sexuality, can influence the types of stories they cover and how they cover those stories. Publication conventions, such as story length and format, are examples of routines that affect what news is covered and how. Organisational influences include direction and oversight by editors and aspects like story structures and templates. At broader and more macro levels, extra-media and ideological influences can manifest in societal or culturally derived values, goals, and aspirations that manifest in coverage and influence what is covered and in which ways. For example, Chinese news media have historically tended to report mainly positive news in order to maintain the Chinese value of harmony (Fengjiao and Sun 2019; Shi 2014). Both Shoemaker & Reese as well as other scholars have revised and reconceputalised the hierarchy of influences model over the years. Keith (2011), for example, called for more dynamism and argued that not all five levels need be included, depending on the context and the nature of the media environment under study. This study adopts a four-layer model to consider the hierarchical impact of individuals, routines, organisations, and ideologies. We use the hierarchy of influences model as a theoretical lens to uncover how economic, social, and political factors affect the types of visual stories that journalists in these countries cover and shed light on how those stories are presented.

Media Systems in Australia and China

News media ownership is important because it can affect access to resources and infrastructure as well as the kind of stories that can be published and their quality. Examples of this include the ideological influence an owner exerts over opinion and/or editorial content (such as past research that has identified biased reporting by News Corp Australia-owned outlets related to mining and climate change) (Linnenluecke and Marrone 2021); the number of people on staff and their roles and skillsets, which can affect which news is covered and how; and access to content-sharing networks or resources that can affect how local and unique the news audiences consume is. On the topic of resources, in addition to staff availability and time pressures, another factor is the availability of legal counsel and how this can affect a media company’s risk tolerance. A large media conglomerate, for example, might have its own in-house legal team that journalists can use to reduce risk around defamation and other lawsuits while a small, independently owned outlet would likely lack this access unless it is part of a collective that shares costs. This can affect the kind of stories and topics journalists pursue and publish (Murray, Ananian-Welsh, and Greste 2019).

Australia has a relatively concentrated news media market (BBC 2023). The biggest players in this landscape are News Corp Australia, Seven West Media, and the Nine Entertainment group. Together, these three players, along with Australia Community Media, own the major urban news outlets in Australia’s capital cities as well as many smaller regional outlets. Others players in the Australian news media market are, in descending order by number of outlets owned, the Star News Group (with 19 outlets), the Today News Group (with 12 outlets), and the McPherson Media Group (with 11 news outlets). No other organisation owns more than 10 outlets (Attard, Dickson, and Jehangir 2022).

Commercialisation of the news media in China occurred in 2003 when businesses were permitted to invest up to 40 percent in news and media industries (Shi 2014). Despite this commercial influence and partial ownership, news outlets in China are under the control of the Chinese Communist Party and/or the government through bodies such as the Party’s Propaganda Department and State Council Information Offices (Chan 2019). These bodies work to ensure that the content news outlets report or amplify does not inspire people to challenge party rule (Shirk 2011). Nowadays, Chinese media can be roughly categorised as state-run and market-oriented media based on their ownership. According to a list published by the Cyberspace Administration of China in 2021, a total of 1,358 news service organisations are approved as credible news sources that others are allowed to repost. All the media on this list are state-run, including the ones at both central and local levels across multiple media types such as TV stations, news services, newspapers, and websites. It is worth noting that the market-driven media systems in most Western democracies differ from the state-driven media systems in China. As Zhang and Meng (2022, 116) point out, “while market logic and media logic intertwine in shaping the trajectory and practice in the West, politics overrides the market and media logic in China”.

Like ownership, a news outlet’s audience also exerts influence on what is covered and how. For example, past research in China has found that Chinese-language news media are more conservative than English-language news media, which tend to be more liberal and to broach more sensitive topics (Yu, Coffie, and Feng 2022). In Australia, crime and court reporting is more common in more populous, urban areas compared to less populous, regional ones (Attard, Dickson, and Jehangir 2022).

This study is interested in how visual news is made and presented in these two countries at the urban and regional levels. Our interest in visual news—operationalised as images accompanying news reporting—stems from the greater amount of attention audiences place on them and how they are more memorable and are processed faster than words (Geise and Lobinger 2017; Whitehouse, Maybery, and Durkin 2006). Despite these privileged attributes, visuals tend to be understudied and their nuances under-theorised (Bock 2020), especially in news contexts and in routine coverage of specific topics or special events.

As such, given the comparatively low proportion of research on visual news in Australia and China, especially in regional areas and regarding routine coverage rather than coverage of special events, and informed by the proposed theoretical framework, the study’s research questions are as follows:

RQ1: How is visual news produced, framed, and presented at four urban and regional outlets in China and Australia?

RQ2: What are the factors and forces that influence the making of visual news at these four outlets in China and Australia as representatives of different media systems?

Together, these two questions allow us to explore how visual news at these outlets compares in Western and non-Western contexts and how various forces, from the individual to the ideological, shape what is shown and how.

Methods

This internationally comparative mixed-methods study is undertaken by one researcher in Australia and three in China so that everyone on the research team has an appropriate familiarity with the contextual nuances of both countries under study. This aligns with Caple, Huan, and Bednarek's (2020) recommendation that those who engage in cross-cultural media analysis have a “high familiarity” with the target audience and with the meaning-making potential of the visuals under study. The research team also includes those with journalism industry experience in Western and non-Western contexts.

Visual Analysis

The first method is a qualitative visual analysis of 1,408 images published in 2022 over a consistent, one-week period in four print and digital news outlets (two in each country). The one-week period selected was intended to be “typical” and, thus, did not contain any holidays in it nor any major country- or region-wide events, such as elections, festivals, or natural disasters, that might have skewed the comparative nature of the sample. The team identified four comparable (in terms of relative size and geographic positioning) news outlets in each country. This included one urban outlet in a capital city and one regional news outlet in a regional centre in each country. The online websites for these outlets were visited at the end of each day to manually download all the images that the outlet had published that day, along with the headline for the accompanying article, the image’s caption (if one was present), and any provided image attribution information. The downloaded images reflected articles spanning a wide range of topics, including crime and courts, culture, politics, economics, sports. We adopted a census-based sampling approach so all images published during that week that were attached to news articles were included in the data pool for analysis.

Regarding the image analysis, we first explored the denotative visual framing of the images (Rodriguez and Dimitrova 2011) to identify and code for the people, places, or objects present. Next, we explored the stylistic-semiotic visual framing of the images to account for their modality (as high, medium, or low). Lastly, we explored the ideological visual framing of the images to identify which types of people (specifically “elite” versus “non-elite”) were featured in the media spotlight in each of the geographic locations under study. Doing so allows us to better understand the distribution of power in the society and the degree to which concerns and issues of the ordinary person or of elite social actors tend to dominate everyday news coverage. In addition to this, we also coded for the (a) number of images per article, (b) the image source (internal: staff, internal: freelance, internal: file/archive, external: social media embed, external: directly contributed/handout, external: wire, other, or unknown), and (c) whether the image was accompanied by a caption. Each category reached an inter-coder reliability rating of.85 or higher using Krippendorff’s alpha, indicating high agreement (Hayes and Krippendorff 2007).

Interviews with Journalists

The study’s second method is interviews with staff—four Australian and 10 Chinese—employed at the same four outlets that produced the images analysed in the study’s first phase. This included six photo editors or equivalent, four photojournalists, two reporters who also had to make images for the stories they wrote, one designer, and one with a dual role as photojournalist and editor. Space considerations don’t allow for a reporting of individual demographic characteristics, but aggregate totals are provided below. Fourteen individuals (eight men and six women) from the four outlets participated in the interviews. Their average age was 40.25 years and they had an average of 16.5 years of professional journalism experience.

The interview protocol was developed, in part, through the results of the visual analysis and, in part, through a review of the relevant scholarly literature. Specifically, questions were included that allowed the staff at each outlet to contexualise their and their colleauges’ practices related to the variables under study (e.g., the image type; image sourcing practices; preference for focusing on elite versus non-elite actors, etc.) and the concepts from the scholarly literature, such as how the influences of individual attributes, routines, organisational attributes, and ideological influences, affected how the staff did their work. Interviews took between 1 and 2 hours and were conducted in late 2022 and early 2023 via Zoom in the participant’s preferred language (English or Mandarin). The research team manually transcribed (and translated, as needed) the interview audio and then imported it into NVivo for coding and analysis using the constant comparative method (Glaser 1965).

Findings

In order to explore the production and presentation of visual, image-based news across these four outlets in these two countries, we first used the raw journalistic output in terms of number of articles and number of images per article as a proxy for the relative importance that each outlet placed upon different storytelling modes (visual and written). We acknowledge that there are potential crossovers between these two modes (for example, in the case of infographics, which blend words and images) so conducted and report later a more granular analysis that takes into account image modality to fully assess its type and the degree to which the outlets favoured verbal over visual storytelling approaches.

“Visualness” of Australian and Chinese News

Over the same consistent week, which didn’t include any holidays, natural disasters, elections, or other major news events that might skew the findings, all four outlets together produced 674 articles. The range was from 76 (at the Australian regional news outlet) to 247 (at the Australian urban news outlet) and the average was 168.5 articles. The two Chinese outlets produced slightly more articles (n = 351) compared to the two Australian outlets (n = 323). Altogether, all four outlets published 1,408 images accompanying articles during the same timeframe. The range was from 80 (at the regional Chinese outlet) to 617 (at the urban Australian outlet) and the average was 352 images. On an outlet-by-outlet basis, the average number of images per article was, in ascending order, .51 at the regional Chinese outlet, 1.75 at the urban Chinese outlet, 2.49 at the urban Australian news outlet, and 4.88 at the regional Australian news outlet. These numbers, as well as numerical findings from the sections that follow, can be found in Table 1.

Table 1. Numerical findings summary.

	Australian urban outlet	Australian regional outlet	Chinese urban outlet	Chinese regional outlet
Number of articles	247	76	194	157
Number of images	617	371	340	80
Average number of images per article	2.49	4.88	1.75	0.51
Number of articles without images	2	2	67	103
Source of image
Internal (staff)	30.86	34.22	24.71	1.25
Internal (freelance)	10.83	6.08	0.00	0.00
Internal (file/archive)	4.87	4.94	0.00	2.5
External (via social media)	1.62	8.74	3.82	0.00
External (contributed/provided/supplied)	4.33	14.06	60.00	7.5
External (wire)	28.88	11.4	7.06	0.00
Other	6.49	0.07	0.00	1.25
Unknown	12.09	19.77	4.41	87.5
Modality (stylistic/semiotic framing level)
High modality (photograph)	94.21	95.81	85.59	71.25
High modality (screenshot)	1.08	0.38	0.88	1.25
Low modality (infographic)	1.62	0.76	10.29	1.25
Low modality (Illustration/rendering)	2.35	3.04	2.35	22.5
Low modality (cartoon)	0	0	0.59	3.75
Other	0.072	0	0.29	0.00
Denotative framing level
People	77.73	77.92	39.71	42.5
Place: built environment	8.3	11.41	23.82	16.25
Place: natural environment	1.32	1.57	3.24	8.75
Place: mixed built and natural environment	0.00	0.00	5.29	0
Objects	7.54	5.51	15.88	26.25
Animals	1.88	3.14	0.59	1.25
Other	2.83	0.39	11.47	5
Ideological framing level
Elite domestic	30.75	22.83	11.76	17.5
Elite international	9.62	3.14	5.59	7.5
Mix of elite (domestic and international)	5.09	1.57	3.24	0.00
Non-elite domestic	26.22	49.6	13.24	11.25
Non-elite international	4.33	0	2.65	1.25
Mix of non-elite (domestic and international)	0.03	0.39	0.88	0.00
Mix of elite and non-elite	1.69	0.39	2.35	5

The number-of-images-per-story metric reveals starkly different views on the importance that each outlet places on visual storytelling in image form. For both the Australian outlets, almost every article was accompanied by at least one image. Indeed, only two articles (out of 76 and 247, respectively, for the regional and urban outlet) lacked any images whatsoever. In contrast, for the Chinese outlets, a completely text-driven approach was much more common. For the urban Chinese outlet, slightly more than one-third (34.5 percent) of its articles lacked images while for the regional Chinese outlet, this proportion grew to nearly two-thirds (65.6 percent). Interviews with staff at each of these outlets provides context about these differences.

Staff at the Australian outlets acknowledged that attitudes toward visual news can change across different cultural contexts but also reflected on how the use of online templates affected editorial decision-making regarding visuals. In the words of one of the staff photographers at the urban Australian outlet:

Part of it has also come from our online template, which, essentially, it has to have an

image . . . and that's mainly driven by I guess a managerial and production decision years ago … because it's expensive to change those things, especially at the bigger organisations. (Urban Australian outlet staffer, 2023)

Staff at the Australian outlets perceived that Australia is somewhat unique in the western world regarding visual news and that, unlike other countries that had cut back on visual news workers (Thomson 2018), the situation in Australia was more stable. This staffer said:

I think Australian media has, by and large, has bucked a lot of global trends. I think newspapers, certainly in the UK, were divesting out of staff photographers many years ago. Where Australian press held onto to staff photographers. And still even to an extent today, they still have. (Urban Australian outlet staffer, 2023)

The Australian staffers attributed this higher importance placed on visual news to advocacy from editors and to the engagement potential that images had, especially for outlets that had paywalls, as the visual was often key in attracting attention and encouraging subscriptions. Staff at the regional outlet in China said the historical reluctance there to create visual news was due to economic as well as logocentric reasons.

Newspapers in this province used to be dense and full of text. Their belief was that words carried the newspaper’s depth, so writing skills were taken seriously, even for career advancement … During our early careers in journalism, in 2003 and 2004, pictures were rarely used due to space constraints, and because the text content itself required more space. It was also more difficult to manage pictures in the past. (Regional Chinese outlet staffer, 2023)

The equivalent staffers at the urban Chinese news outlets noted that the decision to include visuals was driven by how concrete the story was and that more abstract stories were less likely to include visuals. Political news was less likely to have visuals accompanying it than sports or tourism news. This was due, in part, to the sensitivity of the topic. “Very often it is still necessary for the municipal propaganda department and the municipal committee to determine the text and images before we can finally publish,” one regional Chinese staffer said (2023). “There are more restrictions in this area.” The urban Chinese staffers also noted that they were less likely to include images when re-publishing wire content compared to when the article was written in-house. Overall, the shift in Chinese journalists’ practice from historically using no or few visuals to the current practice of visuals being used for roughly half of the articles was due, Chinese staff said, to an increasingly saturated media market where legacy news has to compete with social media.

Interestingly, though Australian news was more “visual” overall than Chinese news, the Australian visual news staff didn’t see this as uniformly positive. They said they felt their time and talent was sometimes wasted on stories that weren’t very visual or didn’t necessarily need visuals and resented having to direct their energies where their text-based colleagues mandated.

Image Sourcing

In terms of image sourcing, it was sometimes difficult to determine the provenance of an image when it lacked explicit attribution. This lack of attribution for images was most prevalent at the regional Chinese outlet where 87.5 percent of the images it published were shown without any attribution information. This trend wasn’t consistent across the Chinese outlets in the sample. The other outlet, in an urban area, provided attribution information in all but 4.41 percent of cases. Between the Australian outlets, they provided attribution in 87.91 percent of cases (for images published by the urban outlet) and 80.23 percent of cases (for the images published by the regional outlet).

All five regional Chinese staffers interviewed mentioned frequent staffing turnovers and inadequate training as reasons for the high proportion of unattributed images being used in regional areas of China. One regional Chinese staffer also added that labelling the source of visuals acquired through stock photography databases wasn’t required, nor was it uniform practice to label the source of images when they are contributed by the sources interviewed in the article.

For images that did include attribution, the primary source of imagery for the Australian news outlets was internal staff or freelancers who contributed more than half of their outlets’ attributed images. In contrast, the Chinese news outlets both relied much more heavily on external sources for visuals through contributed imagery. Contributed images by external sources—such as through wire services, social media accounts, or directly from the sources quoted in news articles—were the most popular source of images for both Chinese outlets, accounting for 62.76 percent of all attributed images published by the urban outlet (and 60 percent of all attributed images published by the regional outlet).

Australian staff perceived there to be a much clearer demarcation between roles compared to Chinese staff and that this demarcation helped to preserve specialised roles. “If they had the writers taking photos on their phones, then that would put staff photographers, in particular, and even freelancers out of work. So it's been important to keep some sense of demarcation in place about who should be doing what,” he said. This was quite different to how regional Chinese journalists saw their roles and the opportunity for specialisation. One regional Chinese journalist in the sample said,

There’s no clear line between text journalists and visual journalists, and our professional identity has blurred. All of us are now general graphic and text journalists. In the past, I’ve been a text-based journalist, but now I need to collect images during interviews and take photos based on how I envision the news to be presented.

The urban Chinese journalists, in contrast, said their words-based reporter colleagues would use their smartphones to make images on-site and that news photographers were increasingly being required to make videos, too, despite the presence of a separate video department, which existed for more complex projects, such as documentaries. In the words of one of the urban Chinese news outlet’s editors: “The roles of photographers are transforming. They are required to make videos in addition to photography and to do some simple video editing themselves.”

Image Type and Modality (Stylistic-semiotic Framing)

As foreshadowed earlier, a potential crossover exists between some types of images (in the case of infographics, for example, which blend words and images) so we also conducted a more granular analysis that takes into account the image modality to fully assess its type and the degree to which the outlets favoured verbal over visual storytelling approaches. By evaluating this at a granular level, this helps bring clarity to what can be a very broad category of “visual news” and helps to illuminate industry practice relating to which image-based visuals are being produced and privileged. It also allows us to focus explicit attention on the stylistic-semiotic framing level within Rodriguez and Dimitrova’s (2011) model.

The most dominant image type across the sample was photographs, which are a type of high-modality visual (i.e., a visual with a high degree of realism [Bell 2001]) and which accounted for a majority of images at each outlet. Both Australian outlets used photographs in similar proportions (in 94.21 percent of cases at the urban outlet and 95.81 percent of cases at regional outlet). The Chinese outlets used photographs less often, in 71.25 percent of cases at the regional outlet and in 85.59 percent of cases at the urban outlet. Screenshots, another high-modality type of image, were used sparingly across all outlets. They were used, on average, about .75 percent of the time in Australian articles and around 1 percent of the time in Chinese articles.

Infographics, a low-modality image type, were also used infrequently in the Australian news outlets (in 1.62 percent of cases for the Australian urban outlet and in .76 percent of cases for the Australian regional outlet). The Chinese regional news outlet also used infographics at similar rates (in 1.25 percent of cases) but the Chinese urban outlet used them most of all (in 10.29 percent of cases). Speaking to this difference, one of the staffers from the urban Chinese news outlet said that their outlet took a more expansive view toward news and tried to use the mode that best suited the story. In their words: “We don’t restrict ourselves in any of the modes. Photos, videos, and infographics … all can be used alone or simultaneously according to the needs of the content of the articles” (2023).

Illustrations and renderings, which are also low-modality image types, were used heavily by the regional Chinese outlet (in 22.5 percent of cases) but were used less frequently at the other three outlets (in an average of 2.58 percent of cases) (See Figure 1). Cartoons, a low-modality image type, weren’t used at all in the Australian articles published over this time period but were used in .59 percent of cases in the Chinese urban outlet’s articles and in 3.75 percent of cases in the Chinese regional outlet’s articles.

Figure 1. To accompany a story about travel, the regional Chinese outlet used an illustration, left, which it did in nearly a quarter of its stories with images, in contrast to the regional Australian outlet, which used a photograph for a similar type of story and did so in more than 95 percent of its articles with images.

The primary reason for the relatively high proportion of low-modality images used by the regional Chinese outlet was due to cost concerns. In the words of a news visuals department supervisor at the regional Chinese outlet:

Our outlet prefers to use some non-realistic visuals these days, probably because all its talented visual staff have left. Nowadays, newspapers can't keep any talent, and people with superior visual skills have started their own studios and companies. This has made it difficult for newspapers to find talented visual staff, as the cost of hiring new staff is quite high. As a result, newspapers have had to resort to using non-realistic visuals, as they are cheaper and easier to produce. (regional Chinese staffer, 2023)

A second reason for the use of low-modality images in Chinese news articles was due to sensitivity concerns. In the words of a woman editor:

“I think that when it comes to news about social controversies and political issues, it is wise not to use real visuals [i.e., photorealistic ones] as much as possible. It’s inevitable that people will comment on real or live images.”

For this editor, audience engagement and commentary about such stories was not desired and so using different types of visuals allowed her outlet to prevent this.

Image Content and Power Relations (Denotative and Ideological Framing)

The final aspect of the image analysis was who or what was depicted in the images published over this one-week timeframe across the four outlets in the sample, and allows us to focus explicit attention on the denotative and ideological framing layers, respectively, of Rodriguez and Dimitrova’s (2011) model. Overall, regarding the denotative level of visual framing, the Australian news outlets were both more people-focused with their images compared to the Chinese news outlets. The Australian news outlets both published images of people in 77.73 (urban) and 77.92 (regional) percent of cases. In contrast, both the Chinese outlets published images of people in less than half of all cases (in 42.5 percent of cases for the regional outlet and in 39.71 percent of cases in the urban outlet).

The Chinese outlets focused on both place (32.35 percent for the urban outlet and 25 percent for the regional outlet) and objects (15.88 for the urban outlet and 26.25 for the regional outlet) more heavily than did the Australian outlets. The Australian outlets focused on place in 9.62 percent of cases for the urban outlet and in 12.98 percent of cases for the regional outlet. The urban and regional outlets also focused on objects in 7.54 and 5.51 percent of cases, respectively.

In addition to identifying the image content by focus on people, place, or objects, we also identified the type of people depicted by their status (as elite domestic, elite international, non-elite domestic, non-elite international, or a mix) to explore power structures in each country and to shed insight on the ideological visual framing level of analysis. In all cases, each outlet focused more on elite domestic actors compared to elite international figures. The regional outlets in Australia and China focused on elites in roughly equal proportions (in 27.54 and 25 percent of cases, respectively) while the Australian urban outlet featured a much higher proportion of elite people (in 45.46 percent of cases) compared to the urban Chinese outlet (which only featured elites in 20.59 percent of cases, the lowest among the sample).

The Australian urban journalists in the sample uniformly bemoaned the focus on elite sources over non-elite ones but attributed this to editorial directions from colleagues or to the 24/7 nature of the news cycle, which privileged reactive rather than proactive coverage. One journalist working for the urban Australian outlet put it like this:

I've always believed we do too many business jobs. They're the sorts of jobs, that as an editor, I would remove from coverage. I can't see the benefit of photographing CEOs all the time. It doesn't provide any information to people about what's going on. I would sooner have my focus on increasing the representation of ordinary people and less CEOs and politicians. (Urban Australian staffer, 2023)

The visuals editor for the regional Australian news outlet, which used photos of non-elites in a similar proportion to the regional Chinese news outlet, revealed a different strategy, one that privileged getting “fresh faces in the paper and online” as a way of increasing subscriptions to its paywalled content. In her words:

Elite people have already seen themselves, they're not going to necessarily always read a story about themselves because they're used to that. And more often than not, especially if it tends to be a politician, they're actually subscribed to the paper anyway, so you're not engaging I guess a new audience. If you're looking for . . . normal people, or non-elite, you are more often than not targeting more of an audience that hasn't been in the paper or been online. (Regional Australian staffer, 2023)

Chinese news staff provided equivalent perspectives on their proportionately lower use of elite figures in their visual coverage, which they said stemmed from concerns about legal or political risk. In the words of an editor at the regional Chinese outlet:

We generally make scant use of portraits, especially of elites. There is a requirement in the domestic media that photos of national leaders in general cannot be used by any other media, except for national media such as Xinhua News Agency and People’s Daily. Unless it is a particularly significant political figure conveying a particularly meaningful speech or spirit, then it is possible to display their portrait, such as Xi Jinping. Otherwise, photos of the conference venue or of the delegates are used instead. (Regional Chinese staffer, 2023)

In addition to featuring other types of elites, such as performers or sports stars, rather than political elites, this editor said his outlet tended to use more images of landscapes, architecture, and food visuals because “these types of visuals don’t have much legal risk.” Staffers at the urban Chinese outlet said they wanted to be more representative of ordinary people and this affected the proportion of elite to non-elite people they focused on. This is exemplified by the statement: “I never care about elites. I think pictures or videos about ordinary people are most important.”

The semi-structured interview protocol did not include any questions about media freedom; however, staff from the regional outlets (but not the urban ones) organically raised the topic, saying that political factors constrained how visual news was produced and presented in China. In the words of one Chinese photo editor, “There’s more freedom in the media abroad. They’re not afraid to talk about or film anything … I don’t think the domestic news is news since there’s a propaganda department above it” (Regional Chinese staffer, 2023). The staff interviewed said this was especially true for news about the government, which was heavily regulated, especially in regional areas. In the words of another regional Chinese editor:

The political factor dominates. In Chinese politics, the party controls everything, and that of course includes the press. That's a very important part of it, and politics always comes first. The media is a recorder of history, and what is published leaves a trail, and if it goes wrong, there are perhaps a lot of political implications. So that's why we say our media are very careful about what they publish. The government decides what gets published and what doesn't, so any deviation from the party line can have serious consequences. In this environment, media outlets are very aware of the consequences of making a mistake, which leads to high self-censorship.

The Australian staffers didn’t mention press freedom concerns but, instead, lamented the effects of the 24-hour news cycle and the pressure to constantly produce and produce quickly. This sometimes led to “churning” rather than doing quality work. They also lamented access issues and how the democratisation of photography through smartphones led to a perceived lack of respect for their craft.

Overall, this analysis of image-based news in Australia and China, contextualised with interviews from staff in both countries, has led to a richer exploration of how visual-heavy the news is in these countries, how these visuals are sourced, which image types and modalities they draw on, and which content they feature. The results inform our understanding of how the hierarchy of influences model operates in Western and non-Western contexts, allows a more granular understanding of the ways the stylistic-semiotic, denotative, and ideological attributes of visual framing operate in each context, and identifies the social, political, and economic forces that influence which visuals are used and how. All these aspects will be discussed more fully in the section that follows.

Discussion and Conclusion

Given the paucity of research on news images in China and the focus on coverage of special events rather than on everyday news in Australia, this study sought to help fill this gap and deeply explore the image side of visual news in a comparative fashion. Using visual framing and the hierarchy of influences model as theoretical lenses, this study also sought to uncover how economic, social, and political factors affect the types of visual stories that journalists in urban and regional outlets in Australia and China cover and shed light on how those stories are presented.

Based on this sample, Chinese news outlets seem to regard visual news as a complement or as an add-on to text-based news rather than a primary form of news content in its own right. This is exemplified by the one-third to two-thirds of Chinese articles that lacked accompanying images, which can be explained, in part, by organisational influences, in the form of online templates that affected which content was presented. These image-driven templates deployed in Australian newsrooms ensured that almost every article included visuals, even though Australian visual journalists sometimes bemoaned this fact, as it meant that they had to sometimes generate images for what they perceived as stories that didn’t lend themselves to visual coverage rather than spend their time more freely on the stories the visuals departments thought mattered most. These influences affect the types of visuals that are created and published in each context and lend insight into the stylistic-semiotic dimension of visual framing.

Economic, social, and political forces also contributed to the proportion of visual to non-visual news at each of these four outlets. In Australia, the capitalistic imperative sees visuals as a way to cultivate audience engagement and drive audience subscriptions. (“A good visual is going to grab people to click through, and that's what I guess we're really aiming for these days, in the day of subscriptions,” in the words of one Australian visual journalist.) In contrast, Chinese editors said their audiences—especially the older ones—expressed a preference for more “logical” text compared to more “emotional” visual news. The additional time required to vet images for sensitive topics like politics also meant that Chinese news editors were likely to forgo visuals for these types of stories.

For a range of reasons, Australian and Chinese news outlets also had a different relationship to originality as it concerned visual news. Recalling that Australian outlets sourced roughly half of their visuals internally in comparison to some Chinese outlets, which sourced as much as 70 percent of their visuals externally (through sources such as social media, directly from audience members, or through wire service imagery), news editors said their audiences were more concerned with the publishing outlet than with the specific source of a visual within an article. In the words of one of the Chinese editors from the regional outlet: “The audience generally doesn’t pay too much attention to what content is packed in and by whom … What audiences value is the platform, the outlet, not the specific person providing the content.” Another Chinese editor agreed, saying, “For the average reader, they don't really care who took the picture and where they got it from. They are more interested in whether the image is what I expect to see and whether it satisfies their curiosity.”

This audience perception seems to influence how the Chinese news templates are designed and the routines that affect how they are used. This is exemplified by the high proportion (87.5 percent) at the regional Chinese outlet of images that lacked attribution. For this outlet and for the regional audience it served, the publisher’s identity was more important than was the identity of the individual person that contributed various elements to the story. Chinese editors also posited that these images might have come from stock libraries, where attribution wasn’t required, or were due to “a high turnover of media personnel, poor training and novices [who] are unfamiliar with the norms of newspaper journalism.”

In terms of image type and modality, which allows us to direct attention to the stylistic-semiotic dimension of Rodriguez and Dimitrova’s (2011) visual framing model, while photographs were the most common type of image across the entire sample, the trend for Chinese outlets to use other image types more frequently than the Australian outlets has to do with cost concerns and normative values of taste and ethics that are societally and ideologically driven, such as the appropriateness of showing realistic images of violence and tragedy.

Overall, influences at the individual, routine, organisational, and ideological levels allow us to observe similarities and differences between the production of visual news in Western and non-Western contexts and how these extend to the content that is represented. This contributes to our understanding of what “visual news” means in more granular and concrete terms and the factors that shape this form’s creation and framing, selection, editing, and presentation.

From a review of the conversations with staff at each of these four outlets in urban and regional parts of Australia and China, it seems that ideological influences were most important in shaping news content within China and that these ideological influences—from government policies and from societal conventions—seemed to trickle down and influence organisational aspects, routines, and individual attitudes about what was acceptable or desirable in relation to visual news coverage. Conversely, Australian journalists placed the most importance on the organisational level—specifically the influence of editors and online templates—and how these affected routines and were sometimes in conflict with individual staffers’ attitudes and perceptions.

This study contributes to the development of the hierarchy of influences model by not only probing at influence through journalists’ self-reported insights but also pairing these insights with empirical assessments of their journalistic output, as well. Other conceptual–theoretical approaches to journalism studies scholarship can overstate the role of individual agency in the process of media production and presentation (Reese 2019). This work helps illustrate how influential, in this context, the organisational and ideological levels can be as they relate to the production and presentation of visual news in Australia and China, respectively. The hierarchy of influences model evolved out of the American context and, as such, reflects the traditionally separate relationship between the news media and government that exists there. Examining it in China where the relationship between the news media and government is much more intertwined, is an important contribution as is this study’s focus on the organisational level of the hierarchy, which the model’s developers note is also a concern that is most often left to other disciplines, such as economics, rather than one undertaken by journalism studies scholars (Reese 2019).

Theoretically, the paper makes a contribution by comparing a socialist market economy with a capitalist one, and does so while also acknowledging the influences of two quite-different political and media systems in the two countries. In doing so, the paper brings into sharp relief the influences of economic models, political systems, and how these interface with media systems in the two countries studied and on journalistic routines, attitudes, and outputs. This paper also makes a theoretical contribution by systematically focusing attention on the three levels (stylistic-semiotic, denotative, and ideological) of Rodriguez and Dimitrova’s (2011) framing model and how each of these manifests in urban and regional contexts, respectively, in Australia and China.

Lastly, given the high staff turnover reported in the regional newsrooms under study and the difficulty this poses for journalistic training and professional socialisation, this study contributes to scholarship that argues the individual and routine levels of the hierarchy are less influential in an age of precarity than the organisational and ideological levels.

Limitations and Opportunities for Future Directions

This paper responds to the “limited understanding of photojournalism in China” that Dan and Ren (2021, 202) identified. It also forays into the “largely neglected” area of everyday visual news coverage (Caple 2019, 58) in Australia and expands the exploration from only camera-created photos to other types of images, such as screenshots and illustrations, to provide a nuanced overview of how visuals are used in everyday news contexts at these outlets. However, given the size of the two countries under study, the findings from the Chinese outlets might not be as representative as those from the Australian sample, which represents a much smaller and more concentrated media market. Additionally, the overall sample size (n = 14) is limited by the number of visual journalism staff at each outlet and by those who agreed to participate in this study.

This study provides a number of directions for future scholarly exploration. Considering the relatively high number of articles without images, it would seem worthwhile to explore how Chinese news audiences regard the lower proportion of visuals in their news coverage. It would also seem worthwhile to explore the degree to which Australian news consumers share Australian journalists’ concerns with having highly visual news regardless of its quality. Finally, given the relatively high proportions of low-modality images used in the Chinese news outlets’ coverage, future research could address how Chinese news audiences perceive and regard low-modality visuals compared to high-modality ones, especially with the advent of generative AI and how audiences perceive AI-generated images compare to camera-produced ones.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by University of Nottingham Ningbo China.