Communicating the authentic China: partnership agreements and the use of Chinese sources and voices by Brazilian media

Abstract

This paper explores China’s public diplomacy in Brazil by focusing on the use of media partnership agreements that allow Beijing to gain greater agency in shaping the country’s image in the region. We examine the partnership between China Central Television and Brazil’s BandNews TV and the co-production of Mundo China, a short news segment solely focused on stories about China. By conducting quantitative content analyses, we assess the ways in which Chinese sources and voices are used on BandNews TV to communicate the authentic China to Brazilian audiences, and we seek to identify any possible correlation with the tone with which China is presented. The results show that Chinese sources and voices are a prominent feature on BandNews TV, and their use is accompanied by a more positive tone than in the way China is depicted on other channels. Additionally, the use of Chinese correspondents signals a partnership that exceeds the co-production of Mundo China. Finally, the data reveal that Chinese media engagement with other broadcasters is also bearing fruit.

Keywords:

• Public diplomacy
• co-production
• country image
• news sources
• China
• Brazil

Introduction

Despite the perception of a lower presence of Chinese media in Latin America compared to the Asia-Pacific region and Africa (International Federation of Journalists, 2020), Beijing is increasingly strengthening its communicational engagement in this region (“CGTN yǔ Lāměi méitǐ gòng chàng shēnhuà hézuò,” 2021). For years, Chinese authorities have insisted that the country “must attach importance to public diplomacy, spread the Chinese voice well,¹ tell China’s story well,² and show the world an authentic,³ three-dimensional and comprehensive China” (Xi, 2014, as cited in Wang, 2014). In turn, this would strengthen Beijing’s ability to guide international public opinion, promote its participation in global governance, “eliminate misunderstandings in a timely manner, and maintain the dignity and image of the country” (Zhao & Zhang, 2022, para. 4). International broadcasting has been key to this process. In the past decade, Chinese media engagement has undergone significant global diversification and growth as part of its public diplomacy strategy (Marsh et al., 2023; Repnikova, 2022). This expansion has demonstrated adaptability to evolving circumstances across regions, with a notable emphasis on Africa and Latin America, generating considerable scholarly interest (Camoça & Araújo, 2021; H. Li, 2023; Marsh, 2017; Mihoubi, 2019; Morante & Wu, 2023; Wasserman, 2018). Initially, China invested heavily in its own transnational and multilingual media outlets with varying degrees of success (Thussu et al., 2018). Later, it sought to place content by buying advertising space in different media platforms around the world (Chang & Lin, 2014). Recently, it has been pursuing collaboration agreements with foreign media, including content sharing and co-productions (Morales, 2022a; Morales & Menechelli, 2023). China Central Television (CCTV) has been among the most active in this regard, either as itself—rebranded as China Global Television Network (CGTN) for its international operations in December 2016—or as part of China Media Group (CMG) since its fusion with China National Radio and China Radio International (CRI) in March 2018. This paper seeks to understand how such agreements allow Beijing to gain greater agency in the mediation of its national image by communicating the authentic China through Chinese sources and voices.

Partnership agreements are not a Chinese invention. In Latin America, where media organizations have limited resources, the reporting of international affairs often depends on international news agencies, which can be considered a classic example of the media imperialism paradigm, i.e. the perceived domination of global media markets by mainly USA-based companies (Chandler & Munday, 2020). Thus, partnership agreements emerge as a solution to minimize the dependency on hegemonic flows of information. For example, many broadcasters from the region are members of Alianza Informativa Latinoamericana (AIL), a non-profit organization focused on providing support and sharing news, allowing its members to gain greater international coverage while minimizing costs. Led by the Colombian broadcaster Caracol Televisión, this alliance has slowly been growing since 2005 and is now made up of 22 broadcasters from Latin America, the Caribbean, and Spain (AIL, n.d.). In 2017, AIL signed a content-sharing agreement with China’s video news agency CCTV+ (AIL, 2017). Additionally, CCTV signed individual agreements with some AIL members, including Bandeirantes Group (Camoça & Araújo, 2021). The agreement signed in 2019 enabled the co-production of Mundo China (MC), a China news segment broadcast on BandNews TV (Bandeirantes, 2019; Morales & Menechelli, 2023).

Based on Rawnsley’s (2015) observation that “Chinese authorities believe that public diplomacy depends on making sure Chinese sources remain the primary source of international news about China from Chinese perspectives” (p. 278), this paper argues that media partnerships constitute a public diplomacy tool that provide Beijing with increased agency in counteracting hegemonic representations of China and in reshaping its national image by making use of sources and voices that reflect Chinese perspectives. Acknowledging previous definitions (e.g. Chandler & Munday, 2020), we refer to agency as the capacity or ability to act and intervene in the process of mediation by exerting editorial power. Furthermore, mediation is conceptualized as “the intervening role of the mass media and journalists in communicating messages or representing reality to audiences” (Chandler & Munday, 2020, mediation). In this sense, through the co-production of MC, CCTV exerts editorial power in the mediation of China’s national image, which is largely constructed using data sourced from Chinese institutions (sources) and presenting the voices of Chinese citizens (voices). Given that Western countries are believed to spread false reports and “rumors” to discredit China (Suo, 2023), Chinese media are tasked to “tell China’s story well, and show the world an authentic, three-dimensional and comprehensive China” (Xi, 2014, as cited in Wang, 2014, p. 10). In this sense, a greater degree of authenticity is associated with Chinese sources and voices that reflect Chinese perspectives directly, i.e. unadulterated by news media from other countries.

By conducting a quantitative content analysis, this paper explores the ways in which Chinese sources and voices are used on BandNews TV to counteract hegemonic narratives by communicating the authentic China to Brazilian audiences. The results show that, unlike other broadcasters, Chinese sources and voices are a prominent feature on BandNews TV, and their use is accompanied by a more positive tone in the way China is depicted. Additionally, the use of Chinese correspondents signals a partnership that exceeds the co-production of MC. Finally, the data reveal that Chinese media engagement with other broadcasters is also bearing fruit.

Media imperialism and global news flows

For decades, scholars have debated how international flows of communication reflect patterns of political and economic domination around the world (Sinclair et al., 1996). The concept of media imperialism became popular to describe the role of the media in facilitating and supporting forms of imperialism (Boyd-Barrett & Mirrlees, 2020). The circulation of news is one classic example, as the content available to audiences in the Global South has historically been shaped by the economic competitive advantage of developed nations in the Global North, which has allowed news outlets there to afford the high costs of media production, including international news (Sparks, 2007). Consequently, news bulletins in developing nations tend to focus disproportionally on events happening in the developed world, and international news more generally is filtered and reported on through the lens of editors based in developed nations (Sparks, 2007).

The proliferation of media organizations from the Global South is seen as evidence of the emergence of contraflows seeking to break with the traditional unilateral dominant flows originating mostly from the USA and Western Europe. However, a distinction is necessary between contraflows constituted by private enterprises that thrive on free-market capitalism and those that are ideologically counter-hegemonic by nature (Thussu, 2007), i.e. seeking to challenge the status quo in global information flows such as Al-Jazeera and others. To amplify their voice and maximize their impact, some news media have established collaborations with like-minded media outlets, e.g. Venezuelan broadcaster Telesur has co-produced programs with other channels such as Russia Today (Russia), CGTN (China), and Al-Mayadeen (Lebanon) with the aim of challenging hegemonic narratives spread by Western mainstream media (Morales, 2022b). However, if media imperialism is defined as the “process whereby the ownership, structure, distribution, or content of the media in any one country are singly or together subject to substantial pressures from the media interests of any other country or countries without proportionate reciprocation of influence” (Boyd-Barrett, 1977, p. 117), could this also apply to co-production agreements between asymmetrical partners?

News sources, country image, and China

News sources provide building blocks that assist in constructing our mental representations of the world. The proposition that news constructs reality has been the subject of heated debates among scholars precisely because journalism is supposed to be tasked with reporting on facts (Lau, 2012). Already in the 1970s, Gaye Tuchman (1975) argued that “in an era dependent upon quickly processed information, the newsperson’s construction of reality is, for many, synonymous with ‘reality’, part and parcel of both everyday life and the social memory” (p. 153). In this process, news people themselves inevitably select pieces of information to tell their stories. As such, the idea of media discourse as constitutive of social reality provides space for challengers to present competing constructions of reality, which in turn readers might interpret in ways beyond media imageries (Gamson et al., 1992). Amongst these competing constructions of reality, appearing authentic, genuine, and trustworthy is key, even if mediated authenticity per se constitutes a paradox (Enli, 2015).

The role of international news agencies in the process of shaping people’s image of the world has been widely discussed for decades (Fenby, 1985; MacGregor, 2013; Paterson, 2011; Tiffen, 1976). Recent studies provide evidence that Western voices tend to dominate the reporting on many parts of the globe. For example, the representation of Africa in the British press shows that, even if a certain proportion of African sources are used, the coverage is dominated by Western journalists (Ezeru, 2022). Similarly, news agencies also become prevalent sources. For example, the image of China on the Latin American press is primarily shaped by the use of Western sources (Ospina Estupinan, 2017). An even greater concern is how certain countries are depicted by international news media. Therefore, governments across the Global South have set up their own transnational media to counter hegemonic narratives spread directly through Western media or indirectly through the use of Western news agencies by journalists around the globe. A previous study on the use of sources by this type of counter-hegemonic international broadcaster shows that the use of non-elite sources is strategic and frequently connected to reporting of countries seen as politically unfriendly (Hernández & Madrid-Morales, 2020).

One of most active countries in its determination to counter hegemonic narratives and gain agency to shape its own image is China. Numerous studies have explored China’s external communication strategy as linked to public diplomacy activities to attain soft power goals (see Blanchard & Lu, 2012; Chang & Lin, 2014; Edney et al., 2020; Hartig, 2019; X. Li, 2011; Rawnsley, 2015; Thussu et al., 2018). While public diplomacy is often defined in academia as “an international actor’s policy based communication activities designed to understand, engage, inform, and influence foreign publics in support of national/institutional interests” (Banks, 2020, p. 64), Beijing has a very instrumental view of it and considers public diplomacy not only as a means to communicate with the world but more importantly to improve its national image and advance national interests (Hartig, 2016). In this sense, international broadcasting has long been considered one of the tools for public diplomatic purposes (Cull, 2019). In Latin America, China has been actively seeking to reshape its image first as an external actor (e.g. CGTN, CRI, and People’s Daily Online) and lately as an increasingly closer partner by signing agreements and even co-producing content with local news organizations in countries like Argentina and Brazil (Morales, 2022a; Morales & Menechelli, 2023).

This paper explores transnational media partnerships as a public diplomatic strategy that aids in “tell[ing] China’s story well” (Xi, 2014, as cited in Wang, 2014, p. 10) by amplifying the use of Chinese sources and voices. Mindful of previous definitions (e.g. Harcup, 2014), we limit the use of the concept of sources to refer only to institutions, be them journalistic (e.g. news agencies and media organizations) or non-journalistic (e.g. government institutions and international agencies) that are explicitly recognized on screen either verbally or visually as those issuing a statement about reality. The concept of voices is used to highlight the visual effect of the participation of individuals in the construction of a news report, from news readers and correspondents to commentators and vox populi interviewees. For example, a news report may be constructed with material from Xinhua news agency (source) and showing an expert from the Chinese Academy of Social Sciences (voice). To examine the use of sources and voices in the partnership between CCTV and BandNews TV, we formulate the following hypotheses:

H1: Chinese sources are prevalent in BandNews TV reporting of China (especially after the signing of the cooperation agreement), and the use of Chinese sources among other broadcasters in Brazil is significantly lower.

H2: Chinese voices are more salient in BandNews TV reporting of China (especially after the signing of the cooperation agreement), and the use of Chinese sources among other broadcasters in Brazil is significantly lower.

H3: China-related reports using Chinese sources and featuring Chinese voices present the country in a positive tone, which is more salient on BandNews TV compared to other broadcasters in Brazil.

Method

We conducted a content analysis following the guidelines set by Krippendorff (1980) and Neuendorf (2017). While the aim is to understand the contribution of MC to both BandNews TV and the overall news media ecosystem in the country, we selected a sample of the most popular audiovisual broadcasters in Brazil based on the Reuters Digital News Report 2022 (Newman et al., 2022). The list was filtered to ensure that it included both public (TV Brasil3) and private (SBT and Record TV) broadcasters in addition to two international news brands operating in Brazil (CNN Brasil3 and BBC News Brasil3). The landscape of broadcasting in Brazil is characterized by a concentration of ownership in the hands of four main corporations: Globo, SBT, Record, and Bandeirantes. During Jair Bolsonaro’s right-wing administration (2019–2022), SBT, Record, and Band were seen as leaning closer to and being favored by the government (Prazeres et al., 2021). Additionally, Record TV is owned by Edir Macedo, Bishop of the Universal Church of the Kingdom of God, the biggest Evangelical denomination in Brazil. TV Brasil is a state-sponsored broadcaster owned by Empresa Brasil de Comunicação (EBC). While BBC News Brasil was founded in 1938, the CNN brand only arrived in Brazil as a local franchise in March 2020. Belonging to Bandeirantes Group, Band News TV ranked fourth among the most trusted news brands in Brazil in 2022 (Newman et al., 2022).

Sampling

Sampling was conducted on content published by the selected media outlets on YouTube. This is in line with data showing that an increasing number of Brazilians are informed through online platforms, with YouTube being the top source (43%) for news among Brazilians on social media (Newman et al., 2022). With the aid of R and YouTube’s API, we extracted a list of China-related videos using the keywords “China” and “Chines,” the latter allowing to better identify both the masculine and feminine variants of the adjective “Chinese” in Portuguese. The date of the agreement between Bandeirantes and CMG, 11 September 2019, was set as the initial point of the time frame and the end date as 31 December 2021. For the comparison of BandNews TV output pre- and post-agreement, the starting point of this additional dataset was backdated to 8 August 2016, the opening of Band Jornalismo, BandNews TV’s dedicated news channel on YouTube. Since the resulting list included videos seemingly unrelated to China, we eliminated duplicates and clips without the word “Chin” in either the title or the description. While the final sample was set at 40% (n = 616), the total number of news stories coded was higher (N = 719) because each MC video clip included three stories on average. The subsamples for each of the media outlets was calculated proportionally. The sample retrieved from the account Band Jornalismo was the largest (n_BN = 271) including a subsample of stories from news bulletins before (n_BNB = 86) and after (n_BNA = 185) the agreement and those of MC (n_MC = 139). Among the subsamples from other channels (n_OC = 309), that of Record TV was the largest (n_R = 86), followed by CNN Brasil (n_CNN = 78), SBT News (n_SBT = 54), TV Brasil (n_TVB = 52), and BBC Brasil (n_BBC = 39).

Variables

To test the hypotheses, coding included the following variables: sources (H1), voices (H2), and tone (H3). The operationalization of the first two categories was informed by previous research (Hernández & Madrid-Morales, 2020) but followed an inductive and deductive approach by qualitatively examining the sample and determining which elements were significant to test our hypotheses. A distinction was made in the category sources between journalistic (media organizations, including news agencies) and other (independent institutions, governments, Non-Governmental Organizations, etc.) sources. These were coded as Chinese (only for the mainland of the People’s Republic of China), Brazilian, Western (operationalized as stemming from the Global North, particularly North American or European), other (from other countries in the world), or any combination up to 15 possibilities. The absence thereof was also coded. The category voices was divided between journalistic and non-journalistic voices. For journalistic voices, we coded anchors (news readers or presenters), reporters/journalists (any other journalistic voices, such as reporters, correspondents, specialized journalists, or experts from the same media organization), and correspondents in China (only in the PRC mainland). The variable non-journalistic voices included experts (external to the media outlet, including scholars; researchers; and analysts from universities, think tanks, and Non-Governmental Organizations), officials (government officials or representatives from public institutions), and other interviewees (the general public, vox populi interviews, etc.). All of these were coded according to the apparent nationality of the people on the screen, i.e. Brazilian, Chinese, other (any other country), or any combination, leading to a total of seven possibilities. The absence of any of these voices was coded as 0. Finally, tone was coded as either positive, negative, or neutral based on Peng’s (2004) own operationalization of favorability, which has been adopted by previous studies (X. F. Li, 2012).

Two Portuguese-speaking coders were trained for two weeks using a comprehensive coding schedule, which was partly adjusted to solve any design issues. Table 1 shows intercoder reliability scores as calculated on a subsample of 10% using ReCal2. SPSS was used for the analysis of all data collected and related statistical calculations.

Table 1. Intercoder reliability coefficients.

Category	Variable	Krippendorff’s α
Sources	Journalistic sources	0.823
	Other sources	0.809
Journalistic voices	Anchors	1.000
	Reporters	0.970
	Correspondents in China	1.000
Non-journalistic voices	Experts	0.949
	Officials	0.792
	Other interviewees	0.720
Tone		0.836
Average		0.878

Results

To test H1, we first observed the type of sources used in China-related reporting on BandNews TV, both by MC and other stories included in news bulletins. We then compared data before and after the signing of the agreement. Finally, we contrasted this with data from the other channels. As described above, we made a distinction between two kinds of sources: journalistic sources and other sources.

Table 2 shows how MC used journalistic sources much more often (99.3%) than other types (19.4%). Additionally, MC stories were almost entirely based on Chinese-only sources (94.4%), whereas most sources for other stories in news bulletins were not disclosed or absent (83.2%). This led to the general China reporting on BandNews TV being skewed and therefore perceived as mostly shaped using Chinese sources (46.9%). Video clips from the news bulletin subsample that used Chinese-only sources were still more frequent than any other type or combination of sources, with the only exception being Brazilian-only sources (13% in non-MC coverage). This might be explained by the COVID-19 pandemic and the increased attention paid to Brazilian nationals in (or evacuated from) China or the import of Chinese vaccines, for example. When compared to the scenario prior to the signing of the cooperation agreement, Brazilian-only sources were relatively rare, both as journalistic (1.2%) or other (3.5%) sources. The increase in Brazilian-only sources seems to have skewed the findings in relation to Chinese-only sources in news bulletins, which before stood at 18.6%. The combination of Chinese and Western (7.0% before and 5.4% after) seemed to be more frequent among stories disclosing more than one type of non-journalistic source.

Table 2. Distribution of types of sources per channel (%).

Type	Origin	Band before	Band after	MC	Total bandnews*	Total other*	SBT news	Record	TV Brasil	CNN Brasil	BBC News Brasil
Journalistic	CN only	12.8	9.7	96.4	46.9	5.2	3.7	1.2	7.7	10.3	2.6
	BR only	1.2	2.7	–	1.5	1.0		1.2	1.9	1.3	–
	W only	1.2	3.8	–	2.2	7.4	3.7	1.2	21.2	9.0	5.1
	O only	–	0.5	–	0.3	–	–	–	–	–	–
	CN + BR	–	–	–	–	0.3	–	–	–	1.3	–
	CN + W	–	–	0.7	0.3	2.9	–	–	13.5	2.6	–
	CN + BR + W	–	–	–	–	1.0	–	–	–	–	7.7
	None	84.9	83.2	2.9	48.8	82.2	92.6	96.5	55.8	75.6	84.6
Other	CN only	18.6	12.4	12.9	12.7	12.0	5.6	16.3	17.3	10.3	7.7
	BR only	3.5	13.0	0.7	7.7	9.1	9.3	9.3	9.6	11.5	2.6
	W only	3.5	6.5	–	3.7	9.1	20.4	3.5	1.9	12.8	7.7
	O only	2.3	1.6	2.2	1.9	1.6	–	4.7	–	1.3	–
	CN + BR	1.2	3.2	–	1.9	2.6	–	2.3	1.9	5.1	2.6
	CN + W	7.0	5.4	2.2	4.0	6.8	3.7	10.5	11.5	3.8	2.6
	CN + O	–	1.6	1.4	1.5	1.9	1.9	1.2	3.8	2.6	–
	BR + W	–	0.5	–	0.3	–	–	–	–	–	–
	W + O	–	1.1	–	0.6	2.6	7.4	4.7	–	–	–
	CN + BR + W	–	–	–	–	0.3	–	–	–	–	2.6
	CN + W + O	1.2	1.1	–	0.6	1.0	3.7	1.2	–	–	–
	CN + BR + W + O	–	0.5	–	0.3	–	–	–	–	–	–
	None	62.8	53.0	80.6	64.8	53.1	48.1	46.5	53.8	52.6	74.4
Total (n)		86	185	139	324	309	54	86	52	78	39

* Journalistic sources: χ2 = 152.628, df = 7, p < 0.001. Significant at the 0.5 level. Other sources: χ2 = 21.239, df = 12, p = 0.047. Significant at the 0.5 level.

Note: This and the following tables to be interpreted as follows: the first column on the left presents data from BandNews TV before the signing of the coproduction agreement and can be contrasted to the next column showing the situation thereafter. The two middle columns contain aggregated data that compare the output of BandNews TV (not including the dataset before the agreement) and that of all other sampled broadcasters combined, which are then presented individually on the right side of the table.

When contrasting the data from BandNews TV with that of other broadcasters, it is evident that on average, journalistic sources were mostly absent or at least not disclosed (82.2%), and less than half of non-journalistic sources were present or disclosed (46.9%). On average, Western-only sources were more frequently used (7.4%) than Chinese (5.2%) or a combination of the two (2.9%). These figures seem to be skewed by the data from TV Brasil. While most channels used both types of sources at similar rates, TV Brasil preferred Western-only sources (21.2%) or a combination with Chinese sources (13.5%) rather than Chinese only (7.7%). A closer inspection of annotations by the coders shows that TV Brasil frequently used footage from international news agencies such as Reuters and occasionally from CCTV. A unique case is that of BBC News Brasil, where among journalistic sources disclosed, the combination Chinese, Brazilian, and Western was more frequently found (7.7%) than either Western (5.1%) or Chinese (2.6%) only. This finding is further explored in the discussion section.

Regarding other types of sources used by broadcasters (excluding BandNews TV), the use of Chinese-only sources tended to be slightly higher (12%) than Brazilian (9.1%) or Western (9.1%) only, followed by the combination Chinese and Western (6.8%). Stark differences were found between single broadcasters. Stories based on Western-only sources were more frequently found on SBT News (20.4%) followed by CNN (12.8%). In the case of CNN, stories with Brazilian-only sources (11.5%) were even higher than Chinese ones (10.3%). Unlike journalistic sources, non-journalistic sources on TV Brasil tended to be more frequently Chinese only (17.3%) or a combination of Chinese and Western (11.5%).

To test H2, we examined whose voices were present in China-related reporting on BandNews TV, both in terms of MC and other stories included in news bulletins. Then, we compared data before and after the signing of the co-production agreement, and finally, we contrasted these with data from the other channels included in the sample. As described above, a distinction was made between journalistic voices and non-journalistic voices.

Table 3 shows the use of journalistic voices. Most stories were anchored by Brazilians both on regular news bulletins (68.6%) and even on MC (67%). However, MC was also frequently anchored by Chinese presenters (15.8%). Reporters and journalists were also overwhelmingly Brazilian both on regular news bulletins (52.5%) and MC (49.6%). Similar to anchors, Chinese reporters and journalists also featured frequently on MC (30.9%). On other channels, Chinese journalistic voices were completely absent.

Table 3. Distribution of journalistic voices per channel (%).

Voice	Origin	Band before	Band after	MC	Total bandnews*	Total other*	SBT news	Jornal da record	TV Brasil	CNN Brasil	BBC News Brasil
Anchor	None	46.5	29.7	16.5	24.1	11.3	1.9	1.2	1.9	6.4	69.2
	Brazilian	53.5	68.6	67.6	68.2	88.3	98.1	98.8	98.1	92.3	30.8
	Chinese	–	1.6	15.8	7.7	–	–	–	–	–	–
	Other	–	–	–	–	0.3	–	–	–	1.3	–
Reporter	None	67.4	45.4	18.7	34.0	49.2	16.7	40.7	65.4	44.9	100
	Brazilian	30.2	53.5	49.6	51.9	50.8	83.3	59.3	34.6	55.1	–
	Chinese	2.3	1.1	30.9	13.9	–	–	–	–	–	–
	BR + CN	–	–	0.7	0.3	–	–	–	–	–	–
Correspondent in CN	None	93.0	89.7	90.6	90.1	100	100	100	100	100	100
	Brazilian	3.5	–	0.7	0.3	–	–	–	–	–	–
	Chinese	3.5	10.3	8.6	9.6	–	–	–	–	–	–
Total (n)		86	185	139	324	309	54	86	52	78	39

* Anchor: χ2 = 47.508, df = 3, p < 0.001. Significant at the 0.5 level. Reporter: χ2 = 52.779, df = 3, p < 0.001. Significant at the 0.5 level. Correspondent in CN: χ2 = 32.143, df = 2, p < 0.001. Significant at the 0.5 level.

Table 4. Distribution of non-journalistic voices per channel (%).

Voice	Origin	Band before	Band after	MC	Total bandnews*	Total other*	SBT news	Jornal da Record	TV Brasil	CNN Brasil	BBC News Brasil
Experts	None	86.0	83.8	82.7	83.3	86.1	70.4	94.2	86.5	83.3	94.9
	Brazilian	11.6	15.1	1.4	9.3	12.6	24.1	5.8	13.5	16.7	2.6
	Chinese	1.2	–	13.7	5.9	–	–	–	–	–	–
	Other	1.2	0.5	1.4	0.9	1.0	3.7	–	–	–	2.6
	BR + CN	–	0.5	–	0.3	–	–	–	–	–	–
	CN + other	–	–	0.7	0.3	0.3	1.9	–	–	–	–
Officials	None	89.5	76.2	87.1	80.9	75.1	57.4	81.4	67.3	75.6	94.9
	Brazilian	8.1	18.9	–	10.8	15.2	25.9	10.5	26.9	12.8	–
	Chinese	–	–	12.2	5.2	1.3	–	1.2	1.9	2.6	–
	Other	2.3	1.6	0.7	1.2	6.1	14.8	3.5	1.9	7.7	2.6
	BR + CN	–	1.6	–	0.9	1.6	–	2.3	1.9	1.3	2.6
	BR + other	–	1.6	–	0.9	0.3	–	1.2	–	–	–
	CN + other	–	–	–	–	0.3	1.9	–	–	–	–
Other interviewees	None	89.5	89.2	79.9	85.2	87.4	61.1	93.0	84.6	97.4	94.9
	Brazilian	4.7	4.9	2.2	3.7	8.7	25.9	7.0	9.6	2.6	–
	Chinese	1.2	3.2	14.4	8.0	1.9	5.6	–	3.8	–	2.6
	Other	–	1.1	1.4	1.2	1.6	7.4	–	1.9	–	–
	BR + CN	4.7	1.1	0.7	0.9	–	–	–	–	–	–
	CN + other	–	–	1.4	0.6	–	–	–	–	–	–
	BR + CN + other	–	0.5	–	0.3	0.3	–	–	–	–	2.6
Total (n)		185	86	139	324	309	54	86	52	78	39

* Experts: χ2 = 20.860, df = 5, p < 0.001. Significant at the 0.5 level. Officials: χ2 = 23.566, df = 6, p < 0.001. Significant at the 0.5 level. Other interviewees: χ2 = 23.104, df = 6, p < 0.001. Significant at the 0.5 level.

Table 5. Mean tone score per type of source and voice.

Type	Subtype	Origin	Band before	Band after	MC	Total band news	Total other	SBT news	Jornal da record	TV Brasil	CNN Brasil	BBC News Brasil
Sources	Journalistic	None	1.92	2.12	2	2.11	1.85	1.64	1.99	2	1.75	1.85
		CN	2.27	2.67	2.8	2.78	1.93	1	3	2.27	1.55	2.25
		Non-CN	1	1.92	–	1.92	1.92	2	2	2.17	1.5	2
	Other	None	1.94	2.27	2.79	2.54	1.95	1.58	2.1	2.21	1.83	1.97
		CN	1.83	2.07	2.74	2.29	1.71	1.25	1.81	1.89	1.53	1.83
		Non-CN	2.25	2	2.75	2.07	1.83	1.85	2.05	2.17	1.55	1.5
Journalistic voices	Anchors	None	1.92	2.11	3	2.37	1.94	2	3	2	1.8	1.93
		CN	–	3	2.91	2.92	–	–	–	–	–	–
		Non-CN	1.96	2.16	2.69	2.38	1.85	1.62	1.99	2.1	1.68	1.83
	Reporters	None	1.9	2.3	2.92	2.45	1.97	1.44	2.29	2.03	1.83	1.9
		CN	3	3	3	3	–	–	–	–	–	–
		Non-CN	1.96	2.02	2.58	2.25	1.75	1.67	1.8	2.22	1.58	–
	Correspondent in CN	None	1.86	2.07	2.75	2.36	1.86	1.63	2	2.1	1.69	1.9
		CN	3	2.95	3	2.97	–	–	–	–	–	–
		Non-CN	3	–	3	3	–	–	–	–	–	–
Non-journalistic voices	Experts	None	1.88	2.21	2.74	2.44	1.87	1.74	2.02	2.09	1.6	1.89
		CN	2	3	2.95	2.95	1	1	–	–	–	–
		Non-CN	2.36	1.83	3	1.97	1.81	1.4	1.6	2.14	2.15	2
	Officials	None	1.94	2.09	2.74	2.39	1.87	1.52	2.06	1.94	1.75	1.92
		CN	–	2.67	3	2.95	2	2	1.33	3	2	2
		Non-CN	2	2.37	3	2.38	1.82	1.77	1.85	2.33	1.44	1
	Interviewees	None	1.86	2.14	2.72	2.37	1.88	1.61	2.04	2.11	1.68	1.89
		CN	2.8	2.78	3	2.94	2.14	2	–	2.5	–	2
		Non-CN	2.5	1.91	3	2.25	1.66	1.61	1.5	1.83	2	–
Mean tone score per channel			1.94* ^†	2.16*	2.78	2.42^†	1.86	1.63	2	2.1	1.69	1.63

* M_BA = 1.94, SD = 0.845 and M_AA = 2.16, SD = 0.775; t(269) = 2.06, p = 0.040.

^† M_BA = 1.94, SD = 0.845 and M_AA+MC = 2.42, SD = 0.736; t(408) = 5.216, p < 0.001.

Note: Mean values range from 1 (negative) to 3 (positive).

As our aim was to understand China-related reporting, correspondents were only coded if they were seen reporting from China. Here we saw a stark difference before and after the signing of the agreement. While Chinese and Brazilian correspondents were both seen in 3.5% of videos prior to the agreement, Chinese correspondents were seen reporting from China in 10.3% of the stories after the agreement. Only on one occasion (0.7%) was a Brazilian correspondent present on MC. When compared to other channels, it is evident that only BandNews TV featured correspondents in China. However, the identity of the correspondents is a subject that deserves close inspection and is covered in the discussion section below.

Regarding non-journalistic voices, Table 4 shows how there was a stark contrast between MC and regular bulletins on BandNews TV. When present, experts (13.7%), officials (12.2%), and interviewees (14.4%) tended to be Chinese on MC, while on regular news bulletins, they tended to be Brazilian (15.1%, 18.9%, and 4.9% respectively). Other voices and combinations were absent in most cases. A similar pattern was seen prior to the co-production agreement. As explained above, the apparent higher presence of Brazilian officials might be partly explained by the government’s response to dealing with the COVID-19 pandemic and other relevant stories. In comparison, Chinese voices were almost completely absent on other channels.

To test H3, we examined the tone with which China was reported on in relation to the use of Chinese sources and voices. To simplify this process, each subcategory was recoded through SPSS in three groups: none (remained the same, i.e. no sources or voices were identified as belonging to any other type), Chinese (including any combination with other types of sources), and non-Chinese (excluding any combination with Chinese). Table 5 shows how the results were clear. Overwhelmingly, the presence of Chinese sources and voices seemed to correlate with a more positive tone in the way China was reported on. There were a few exceptions where this correlation was not identified. In the case of BandNews TV, the use of Chinese non-journalistic sources did not seem to be in correlation with a more positive tone. Stories where sources were not disclosed seemed to report on China with a more positive tone, albeit with an almost negligeable difference compared to those that either used other Chinese or other sources. Another exception seemed to be the used of correspondents, where the median for non-Chinese correspondents was the highest. However, this is somewhat misleading because, out of the whole sample across all channels, there was only one videoclip featuring a non-Chinese correspondent.

Chinese sources and voices were almost completely absent across other broadcasters. In the few cases where these were used, there seemed to be no apparent correlation to positive reporting. There was no instance in which the mean for the use of Chinese sources or voices among other broadcasters was higher than for BandNews TV. On average, almost all indicators were below neutral (2.00). For example, the category officials (2.00) only accounted for 3.2% of the subsample (10 stories), and the category interviewees (2.14) for 2.3% (6 stories). In other cases, the differences between categories were nonsignificant, for example between Chinese (1.93) and non-Chinese (1.92) journalistic sources. The only interesting case seems to be TV Brasil, where Chinese journalistic sources were used in up to 21.2% of the stories sampled, and the mean score (2.27) was higher than any other. However, the difference between those that used non-Chinese sources (2.17) and unidentified sources (2.00) was nonsignificant. In the case of non-journalistic sources used by TV Brasil, the mean tone score for Chinese sources (34.6%) was lower (1.89) than any other and rather negative (< 2.00). Other seemingly positive indicators among Chinese voices on TV Brasil, such as officials (3.00) and interviewees (2.50), were negligible because they only represented two videoclips for each category.

Discussion and conclusion

This study explores how Chinese sources and voices are used on BandNews TV to communicate China to Brazilian audiences and the correlation with the tone with which China is presented. The results showed that the partnership agreement has provided CCTV with a platform for Chinese sources and voices to play a protagonistic role in shaping the way China is portrayed on BandNews TV. While this is mainly articulated through MC, Chinese sources and voices also feature on BandNews TV news bulletins at a higher rate than on other channels. The data from BandNews TV indicate that reports with Chinese sources and voices present China with an overwhelmingly positive tone, which responds to the ambitions of the Chinese government to “tell China’s story well” (Xi, 2014, as cited in Wang, 2014, p. 10), whereby “well” means “positively.”

The presence of correspondents in China constituted an innovative feature in the mediation of China’s image by BandNews TV. However, closer inspection showed that these were reporters working for CMG, who were introduced as either a “reporter in China” or “our correspondent in Beijing,” and while their names appeared on screen, no affiliation was mentioned. A simple online search revealed that they worked for CCTV or CRI’s Portuguese-language service. In fact, some appeared reporting on Yanzheng Road, near the headquarters of CRI in Beijing’s Shijingshan district. On one occasion, a reporter was seen standing outside the People’s Procuratorate of the Beijing Municipality, located just opposite CRI’s main building. Regarding the only Brazilian correspondent, this was in fact a reporter from BandNews TV that was sent to Beijing to cover Bolsonaro’s visit to China in 2019. The non-disclosure of the affiliation of Chinese reporters might lead audiences to believe they work for BandNews TV. This issue echoes accounts from other parts of the world, where certain broadcasters with content sharing agreements do not always disclose Chinese sources (International Federation of Journalists, 2020). However, we cannot categorically confirm nor deny that this is the case with BandNews TV.

While a quantitative approach allows for greater understanding of the extent of the use of Chinese sources and voices by BandNews TV, greater contextualization is needed to understand the implications of the phenomenon of partnership agreements more broadly. Amongst the other broadcasters sampled, Western sources appeared more frequently, and Chinese voices were almost completely absent. TV Brasil stands out for its high use of Western-only journalistic sources (21.2%), Chinese only non-journalistic sources (17.3), and a combination of Chinese and Western sources in both categories (13.5% and 11.5%, respectively). The reason behind this might be explained by TV Brasil being part of EBC, which in 2015 signed an agreement with CCTV for content exchange and program co-production (Brandão, 2015). In 2019, EBC signed a new agreement with CMG, which expanded the cooperation to include “programme exchange, content sharing, joint production, cooperative broadcasting, staff training and technology exchange” (EBC, 2019, para. 1). Overall, the mean tone with which China was reported on TV Brasil was slightly positive (2.1), higher than the other channels (1.86 on average) with the exception of BandNews TV (2.42 on average).

While the findings are compelling, this type of collaboration is not unique. Bandeirantes Group has signed agreements with Japan’s NHK (Perfil News, 2005), the USA’s The New York Times (NYT) (Ribeiro, 2016), Russia’s BRICS media (TV BRICS, 2020), the UAE’s news agency WAM (Bandeirantes Group, 2021), and Germany’s DW (Lucas, 2022). As well as having joined AIL (see above), Bandeirantes Group is the only Brazilian broadcasting member of the European News Exchange (Wells, 2020) and reportedly the only provider of Brazilian news for Reuters (Lucas, 2022). While this paper has only explored the partnership with China’s CMG, it is unclear to what extent all these other agreements may have also influenced how other parts of the world are presented on the group’s many channels. For example, the agreement with NYT led to “Conexão com The New York Times,” a news program broadcast by BandNews TV and presented by a Brazilian reporter but featuring content created by NYT.

Collaboration, co-production, and content-sharing agreements are presented as beneficial to both parties. However, both sides face challenges. On the one hand, this type of agreement allows BandNews to improve its coverage of international news. However, the level of reciprocity is unclear, i.e. while the use of Chinese sources and voices benefits China’s interests in gaining control over how its own image is shaped, it is unknown whether BandNews TV content features on Chinese television. The degree of asymmetry in this partnership might become cause for concern by creating new patterns of media imperialism, even if scholars suggest that China is far from becoming a cultural imperialist (Sparks, 2020). On the other hand, even if the televised image of China seems to have become more positive on BandNews TV, the perception of the involvement of the Chinese government and its public diplomatic interests as the ultimate power behind CMG might be at odds with further conveying a sense of authenticity to what viewers see on their screens. Here, further research adopting a qualitative approach might provide insights into how authenticity can be communicated beyond the inclusion of Chinese voices.

This study’s contributions are twofold. Empirically, it provides evidence that China’s shift in its approach from external actor to increasingly closer partner is helping it to gain agency to reshape the country’s image and counter hegemonic narratives. Theoretically, this expands our understanding of the utility of transnational media partnerships to advance public diplomacy goals by revealing their impact on news production practices, e.g. in relation to sources and voices. Indeed, this highlights the discursive role sources and voices play in attaching a sense of authenticity to the mediation process. Additionally, this study argues that even if partnerships seem to be beneficial to both parties, media imperialism concerns might not be totally unsubstantiated speculation.

This research naturally presents limitations. First, Globo News, the most popular broadcaster in Brazil, does not publish news content on YouTube; therefore, it was excluded from the sample. Second, the dataset was limited to China’s televised image as published on social media. Third, the findings are hardly representative of overall journalism in Brazil. Finally, this study only focuses on one country. We invite future research to adopt a comparative perspective by examining countries such as Argentina, Chile, and Venezuela, where CMG has partnered up with Télam, Radio Cooperativa, and Telesur, respectively.

Disclosure statement

No potential competing interests are reported by the authors.

Additional information

Notes on contributors

Pablo Sebastian Morales

Pablo Sebastian Morales is a UKRI postdoctoral fellow in the Department of Media and Communications at LSE. After obtaining a master’s degree at Zhejiang University, he worked for China Radio International and People’s Daily Online in Beijing. He holds a PhD degree from the University of Westminster, UK. His research examines global media flows and journalistic cultures in the Global South.

Paulo Menechelli

Paulo Menechelli holds a PhD in international relations at the Institute of International Relations, University of Brasilia (UnB). He is the secretary-general of the Centre for Global Studies (UnB), researcher of the Media Studies and International Relations Lab (LEMRI) at the Federal University of Rio de Janeiro (UFRJ), editor of Mundorama Magazine, and co-founder of the think tank Observa China.

Notes

1 The interpretation of word the word “well” (好 hǎo) in the original version (“讲好中国故事”, jiǎng hǎo Zhōngguó gùshì) can be ambiguous. The prevalent interpretation amongst translators seems to favor the view that this is an adverb qualitatively indicating how the action is performed (i.e. well).

2 The word “真实” (zhēnshí) can be translated as “true,” “real,” or “authentic,” depending on the context.

3 The Portuguese-language spelling of the country’s name (Brasil) has been kept in all names of broadcasters or institutions.