China’s Media Expansion in Zambia: Influence on Government, Commercial, Community, and Religious Media

Abstract

This study examines how news about China is presented, spread and impacts the diversity of opinions in state-owned, commercial, community, and religious media outlets in Zambia. Drawing from the 2012–2021 data, and guided by intermedia agenda-setting theory, we provide evidence of the counter-attitudinal influence of news content on China across the four media categories. While state-owned media emphasized issues of privatization, and government and foreign policies, community and religious media focused on issues of environmental degradation, poor labor conditions, and mining. Commercial media mirrored state-owned media, a trend also reflected in community and religious media outlets across time. These findings suggest that China shapes how the media in Zambia present content about China. As observed, over the years, China is positively presented in the Zambian media.

1. Introduction

The media in Zambia received a backlash after the state-owned media outlet, and Zambia National Broadcasting Corporation (ZNBC) accepted a joint venture with the Chinese media conglomerate StarTimes. Accordingly, the ZNBC accepted a venture bankrolled by a $232 million loan from the Exim Bank. This suggests that the Chinese media will control 60% of the revenue generated by the venture for the next 25 years (Mundia 2017). Media watchdogs criticized the Zambian government for allegedly granting excessive influence over the media landscape. For example, Chulu (2017) observed a shift in programming and content allocation after ZNBC accepted the joint venture. Commenting on their programming of ZNBC’s Radio One content, the Chulu (2017) averred that “although most programs had a local focus, the majority were Chinese sponsored—if the content does not have a Chinese narrative in it, it risks being rejected” (Mundia 2017, p. 22). However, there is little or no empirical evidence to support these allegations. Therefore, how the joint venture has shaped China’s coverage by the Zambian media has become a question of interest. Hypothetically, the underlying debates suggest that the joint venture influenced how the Zambian media has always reported China. While this is expected in any relationship, the questions of content allocation, what kind of information the media chooses to report, and what kind of information they ignore become relevant points of departure.

This study examines how news content about Chinese spreads and impacts the diversity of intonation/valence in Zambian news over time. Essentially, we examine (a) the presence of Chinese news content in Zambia, (b) how it spreads, and (c) how it shapes narratives. Traditionally, measuring the Chinese news agenda has required establishing the presence and transfer of Chinese news content from China to Zambia. This study takes news articles about China in the Zambian media as units of analysis. With the collected data, I established an intermedia agenda-setting by inferring influence over a period of time (2012–2021). I examined the changes in influence by examining topics and tone in news stories about China.

2. Literature and Theoretical Review

The democratization of Africa in the 1990s was accompanied by the euphoria of a neo-liberal media that opened avenues for different media categories (Mfumbusa 2010). Previously, the media in post-independent Africa was state-owned, with only a handful belonging to religious groups. As indicated by the Tanzanian founding president, Julius Nyerere, “African countries could not afford the luxury of a muckraking, adversarial press” (Musa and Domatob 2007, p. 324). Like many other African countries, Zambia benefitted from the neoliberal and mushrooming of different media outlets categorized as state-owned, commercial/private, community, and religious.

Considering the government mouthpiece, the Zambian government-owned media traces its history to the British colonial government (Gondwe 2021; Moyo and Chuma 2010). Its main role was to inform and entertain colonial elites (Mulauzi et al. 2013). The same roles continued even after independence but now served the government of the day. In this paper, ZNBC thought referring to broadcast media, is used synonymously with all state-owned media, including newspapers, magazines, and Internet-based websites (Ndawana et al. 2021).

Second, commercial and private media gained recognition during the democratic era. This media category is sustained by its profits, thus engaging in journalism practices that provide assured revenue. As Mulauzi et al. (2013) observed, “there was a general feeling among the private media practitioners that ethical reporting was dull, drab, uninteresting, and therefore, could not sell; that those who involved in it risked becoming bankrupt” (p. 2). The POST newspaper experienced a strong backlash until its indefinite suspension in 2016. The closure is attributed to the failure of the paper to settle tax obligations—a motive that most people have described as a political strategy to silence the media that criticizes the government. Some critics have argued that the only way for commercial media to survive is by aligning itself with powerful elites or the government (Ndawana et al. 2021).

Community media is based on participatory communication. According to Wigston (2001), for a medium to qualify as community media, it had to aim at ‘empowering’ and reporting ‘culturally’ specific information that aligns with a particular community informed by their geographical and sociocultural (including linguistic) necessities (p. 31). Banda and Fourie (2004) summarize the criterion into four main features:

• should be fully controlled by a non-profit entity and carried on for non-profitable purposes
• should serve a particular community
• should encourage members of the community it serves, or persons associated with or promoting the interests of the such community to participate in the selection and provision of programs to be broadcast during such broadcasting
• may be funded by donations, grants, sponsorships, advertising, membership fees, or by any combination of the aforementioned

Unfortunately, these features are questioned as the most powerful elites and the government has invested in such media, thereby shaping the agenda (Banda and Fourie 2004).

The fourth is religious media, which is probably among the oldest in Zambia. The religious media—one owned by a certain religion, that is Catholic radio–is considered a powerful tool not only for evangelization but also for creating an agenda believed to influence political decisions. According to Chita and Mwale (2017), religious media is least influenced by either the government or other donations. This is because of their modes of financing and station sustainability. Like community media, most religious media are run by their members, on what one would consider a voluntary basis. The Catholic media, for example, is managed by the clergy and nuns, who are well-educated in the field but do not directly receive a salary. Other donations come from church members as well as from well-wishers outside their countries. There is little financial support from outside individuals and organizations that are not connected to the church.

Because of its independence, religious media have become a symbol of authentic journalism practice, but with particular religious propaganda. Even in the face of Chinese investment, religious media researchers assert that the Chinese news agenda has little or no influence on their agenda (Ojo 2017). However, over time, religious media are now connected with other secular media through symposia and media training offered by either BBC, USAID, or China (Gondwe 2022).

Agenda-Setting and the Flow of Information

News flows in Africa are often understood in the context of global information flow (De Beer 2010). Since the publication of the 1980s ‘Many Voices, One World’ or UNESCO’s MacBride Report, a series of research have chronicled to challenge the operation of foreign news values and information (De Beer 2010, p. 601). Most studies have examined a country’s salience in foreign news coverage by observing the dynamic interactions between news outlets in different countries. These interactions are understood within the broader framework of agenda-setting theory. Stemming from the idea that the media tells us what and how to think about certain issues (McCombs 2014), intermedia agenda-setting focuses on the interaction of different media outlets. The idea is that within intermedia interactions, some elite media act as central nodes and influence other media by transferring issues or attribute salience (Guo and Vargo 2017).

Supporting evidence has consistently found that elite media sets the agenda for smaller-scale or less-elite media. According to McCombs (2014), this happens partly because journalists are constantly endeavoring to validate their sense of the news through the lens of more established media. Since most media in Africa operate on very low budgets and monitor news content from well-funded elite media organizations, (Gondwe 2022) some scholars have argued that the quest for a balanced flow is still a far-fetched endeavor. As Thussu (2022) puts it, “While formal colonialism may have ended decades ago, vestiges of informal and intellectual colonialism persist and can also be discerned within the global news and information systems” (p. 2). This argument is consistent with Paterson (2013) observation that although studies on the free flow of information have declined, their impact on the news agendas of countries in the Global South continues to manifest. According to Paterson (2013), several other variables, including complicated language markets, impede the quest for a balanced flow of information in Africa. In addition to technological challenges, the flow of information in most of sub-Saharan Africa is consistently underscored by the presence of distinct geolinguistic media spheres (Paterson 2013, p. 81). Essentially, these distinctions account for why most African media find it challenging to share content within the continent, and across the globe, therefore, relying on Western-created content.

On the other hand, China’s media expansion within the African continent provides opportunities for the local media to utilize new technologies and amplify the flow of information. However, this does not come without scrutiny, given the control that China places on the content produced in Africa (Ojo 2017). Similarly, Umejei (2020) observed that the Chinese media operating in Africa had two kinds of gatekeepers: The “Upstairs gatekeepers” and the “Downstairs gatekeeper”. According to Umejei (p. 55), upstairs gatekeepers were high-level management journalists comprised of Chinese nationals assigned to monitor and respond to news items that talk about China. On the other hand, there are downstairs gatekeepers, a group of African journalists employed by the Chinese media to report on other issues that do not mention China. In this view, Umejei asserted that the relationship was asymmetrical in the sense that Chinese gatekeepers (upstairs) limited input from local journalists. A particular example is the 2015 visit of the pope to Kenya. African journalists working in Chinese media organizations in Africa explained that they were not allowed to cover the Pope’s visit. Most expressed displeasure that they could not cover the Pope when other international media organizations reported extensively on the visit of the Catholic pontiff to East Africa.

Research continues to show that most media outlets in Africa monitor news content from international news sources (Reta and Chekol 2000). Similar studies found evidence of the intermedia influence of larger news organizations operating within Africa, such as BBC Africa, Aljazeera, Radio France International (RFA), and Deutsche Welle (DW). Due to a lack of resources, most African media tends to have a special section in which they monitor and present news from established organizations. For example, a local news media that could not afford to join a political campaign trail would assign an individual to collect news content from BBC Africa and report it (Akinfeleye et al. 2009). Nonetheless, despite these debates, it is evident that the Chinese media has rapidly expanded its operations in Africa (James and Boukes 2017). Second, there is little or no quantifiable evidence to suggest the presence of African news content in the Chinese media. Existing studies examine how the Chinese media presents crisis issues about Africa (Zhang and Matingwina 2016) and why China is interested in Africa (Madrid-Morales 2016). To further understand the complexity of this relationship, we must explore the intermedia news agenda, not as a transnational phenomenon (i.e., Guo and Vargo 2020), nor a networked global media system (Himelboim 2010), but as an agenda happening within a particular country, and in our case, Zambia. Therefore, to advance our knowledge of influence, we investigated the degree to which each media category in Zambia covers China in the news. Therefore, we asked the following questions:

• RQ1: How do media categories in Zambia (State-owned, Commercially owned, Community owned, and Religious-owned) cover China’s engagement with Zambia?
• RQ2: Is there a reciprocal intermedia agenda-setting relationship on how (a) State-owned media (b) Commercially media (c) Community-owned media and (d) Religious-owned media cover China in Zambia?

3. Results

3.1. Overall Findings

This study examined how different media categories report news about China in Zambian media. I examined the content from the state-owned, commercial, community, and religious media. Overall, the results indicate that the government-owned media had more routine stories that presented China in a positive light. Second, commercial media seemed to mirror state-owned media, then community and religious media. However, community and religion were more inclined to criticize China’s expansion and thus had stories that presented China in a negative spotlight. The results also suggest an intertwined networked relationship across the four media categories, which helps explain the influence permeating across the media in Zambia.

Table 1. Overall sample characteristics.

ZNBC	Private Community Religious
Variable	Freq	%	Freq	%	Freq	%	Freq	%
Story Source
Routine	458	62.06%	188	52%	19	10%	16	37%
Intermedia source	196	27%	74	20%	47	25%	2	5%
Enterprise	71	10%	87	24%	116	62%	25	58%
Not indicated	3	0.4%	14	4%	5	3%	0	0%
Total	728	100	363	100	187	100	43	100
Story Topic
Political	407	56%	141	39%	13	7%	2	5%
Economic	291	40%	67	18%	89	48%	21	49%
Social	9	1.06%	39	11%	3	2%	1	2%
Scientific	3	0.4%	0	0%	0	0%	0	0%
Environment	18	2%	101	28%	63	34%	11	26%
Cultural	0	0%	13	6%	2	1%	0	0%
Human Interest	0	0%	2	1%	18	10%	8	19%
Other	0	0%	0	0	0	0	0	0%
Total	728	100	363	100	187	100	43	100
Story Type
Hard news	502	69%	119	33%	47	25%	23	53%
Brief News	145	20%	7	2%	39	21%	5	12%
Opinion	51	7%	22	6%	86	46%	7	16%
Feature	14	2%	215	59%	15	8%	8	19%
Other	16	2%	0	0%	0	0	0	0%
Total	728	100	363	100	187	100	43	100
Story Enhance
Picture	673	92%	218	60%	106	57%	9	21%
Illustration	28	4%	122	34%	4	2%	0	0%
None	27	4%	23	6%	77	41%	34	79%
Total	728	100	363	100	187	100	43	100
Story Frames
Episodic	614	91%	137	38%	70	37%	5	12%
Thematic	14	9%	226	62%	117	63%	38	88%
Total	728	100	363	100	187	100	43	100

Note: Percentages were estimated to the nearest whole number.

presents an overall summary of our sample characteristics as they relate to how the media in Zambia reported China:

The database on which I base my analysis indicates changes in how China is presented in the news. Essentially, there is a shift from negative to positive across time series in all media categories. Whether these changes are the result of China’s positive influence or the influence exerted by power is yet to be substantiated. However, the decline in negative news content about China and the lack of criticism about some events involving China in Zambia (i.e., people protesting poor working conditions) attest to soft power influences.

3.2. Topic and Tone of News about China in Zambia

Generally, the results indicate that the issues were categorized into two different types: Those found in state-owned and commercial media and those found in community and religious media. News stories about China from state-owned and commercial media were positive, whereas those in the two other media categories were negative. For example, while positive news would have headlines like, “President Lungu lauds China over investments in Africa”, “Zambia launches China-funded project to improve access to water,” “Zambia-China Partnership Yields $5bn Infrastructure Projects”, “Govt Unveils USD 2.5 Billion Development Package from China”, etc., negative news content would be represented by headlines like, ” “Hakainde Is A Useful Idiot of The West, He Is Fighting China for Their Interests”, “Stop being crybabies over Chinese, compete with them ‚Kitwe DC”, “Say no to China: Anger mounts in Zambia over Beijing presence”, etc.

Essentially, the findings suggest considerable degrees of similarity in issues and topics between state-owned and commercially owned media, as indicated by the Pearson correlation coefficient r > 0.5. Similarities (r > 0.5) were also observed between the community and the religious media. In each case, several related issues were prominent and appeared to influence the discussion. While the ZNBC and commercial media depicted government policies, investment, foreign policies, and mining as the most important issues, community and religious media focused on issues about how Chinese investment policies affected the environment, agriculture, and poor labor/working conditions for individuals working for Chinese companies. Additionally, community and religious media appeared to cover more diverse topics than ZNBC and commercially owned media that focused much on politically inclined issues. For example, stories of social justice–that is, environmental degradation, poverty, education, and human rights–were more evident in community and religious media than in state-owned and commercial media.

3.3. Intermedia Agenda-Setting Effects

RQ2 examined the reciprocity of intermedia relationships across the four media categories. Similar to the findings in RQ1, the results indicate divided interaction levels between state-owned and commercial, versus community and religious. However, I also found moderate, statistically significant levels of interactions between commercial and community versus state-owned and community. Granger causality was also identified in these interactions, placing religious media that might not have clear connections with state-owned and commercial media at the receiving end.

Table 2. Granger test causality on salient issues.

China News Agenda	Local Media Agenda
Issue	GC Results	Issue	GC Results
Government policies	334.31	Governance	71.34
Privatization	387.67	Privatization	60.23
Education	21.45	Education	238.72
Health	23.18	Health	156.91
Politics	289.14	Politics	66.32
Tourism	88.41	Tourism	78.12
Labor Laws	9.14	Labor Laws	236.12
Mining	55.43	Mining	162.4
Foreign policy	247.34	Foreign policy	52.51
Environment	6.14	Environment	77.91
Media & Culture	3.11	Media & Culture	68.33
Security	167.32	Security & Crime	23.14
Religion	0.23	Religion	256.26
Human Interest	0.37	Human Interest	182
Agriculture	73.2	Agriculture	166

Note: lag = 1-F(17,183); p < 0.01.

presents a summary of the Granger test causality.

To further illustrate these differences, I performed cross-lagged correlations across four media categories in Zambia. The results indicate that at Time lag 1, ZNBC influenced the news agenda of commercial/private media (r = 0.897), community media (r = 0.834), and religious media (r = 0.547) as indicated in Figure 1 below.

The Rozelle-Campbell baseline results indicate that ZNBC “caused” the other media categories to focus on the importance of the Chinese news agenda (rc1e2 = 0.47 > rc1e2 = 0.02), where r stands for the correlation coefficient, c stands for cause, e stands for effect, and 1 and 2 stand for successive time periods. At Time Lag 2, the joint effects of the ZNBC news agenda increased in commercial, private, community, and religious media than in Time Lag 1. This joint effect is greater than the influence of each of the three media news agendas on each other. In summary, Topics that seemed to criticize China, i.e., poor labor conditions, lack of environmental conservation, health, and human interest, received less attention in state and commercially owned media.

4. Methods

To capture a holistic picture of the role of China and the interplay of intermedia agenda-setting in Zambian media, we used both automated and non-automated (manual) tools to identify issues in the news content stored online. Through a two-constructed week model, we collected all Zambian news articles that mentioned China from 1 January 2012 to 31 December 2021. This means that for each month, two weeks were constructed, thus providing a total of 24 days per year and 216 days to collect from (24 days × 9 years). Amazon’s Alexa and the Media Institute of Southern Africa (MISA-Zambia) were used to catalog the top news sites in Zambia. The two apps helped identify legitimate news sources based on daily time spent on the site, percentage of daily page views per visitor, percentage of traffic from search, and a number of total sites linked to that source. A total of 127 news sources were identified, each of which owned an online website from which stories were collected. The 124 news sources were then categorized into state-owned, commercial, community, and religious media.

Second, we manually went through the websites and social media platforms of each cataloged media and selected content falling under the two-constructed week dates. Third, we used Noovell.com Global News to collect online stories. We also collected stories from print copies of newspapers. Subsequently, with two trained coders, we manually inspected and collected stories that fit the criteria. If the stories collected correlated with those collected earlier, we only picked one Based on a 10% sample of the stories collected by separately trained coders, the Krippendorf’s alpha (α) the intercoder reliability tests were recorded at ranges 0.87 and 0.93. This resulted in a total of 1812 articles, with 1022 articles coming from ZNBC, 538 from commercial media, 137 from community media, 72 from community media, and 43 from religious media.

Measures

This study employed the latent Dirichlet allocation model (LDA) to automatically identify salient topics from each news story. First, I cleaned the data by examining grammatical issues from the text and then trained our model to generate “topics”. This step provided us with a larger picture from which a list of 23 salient topics was identified.

Second, to find specific topics, I employed supervised machine learning (SML) and the Support Vector Machine (SVM) to help identify and predict each topic separately. Essentially, I used a word-frequency text feature to transform the text into a vector of numbers. With only a corpus of 1812, most of the processes were manually aided. Therefore, our analysis detected 15 topics labeled as government policies and China, China and investment, China and Culture, China and health, China and the environment, China and security/surveillance, China and Agriculture, China and tourism, China and stories of human interest, China and labor/working conditions, China and politics, and China and foreign policies (i.e., Zambia’s engagement with the US, China and mining, China and education, and China and religion). These labels were created by the researcher based on the words generated by the model.

I collected the top ten associated words for each topic with the highest probability estimate and created labels. For example, the labels government policies and China were created from associated words like (government, debt, refinancing, loans, Sino-Africa, opposition, diplomatic, policy, development, and economy). On the other hand, China and labor labels were created from associated terminology such as protest, remuneration, poor sanitation, dangerous conditions, no boots, delayed salary, injury, no compensation, harsh boss, and racism). In addition, the levels of importance for each topic were calculated by observing the frequency and degree of centrality. Degree centrality refers to the number of connections an attribute/word has with other attributes in associated networks. To avoid reporting multiple topics under the same theme, I mapped each topic in the corpus to a multinomial topic distribution that encoded how much each of the |K| topics was entailed within the article, arriving at the 15 topics, as indicated above.

The results were largely determined through inter-annotator agreement and intent classification of the topics. The findings indicate a high intent classifier probability (∼0.80) of topics in China, given a set of posts P = {Pi | 1 ≤ i ≤ n | C(Pi) = Topic}. The data also revealed that if the taxonomy of a post was unknown, then each algorithm had a probability of approximately (0.20) to classify it as an intent post. In this sense, the topic classification of posts and filtering of non-topic-based (or noisy) posts improved the performance of the proposed intent classification.

5. Discussion

5.1. Who Influences Who?

Essentially, the news did not appear to flow in a unidirectional manner, suggesting that network relationships were intertwined. At face value, this would suggest reciprocity of influence across all media categories. These findings imply that there are two main aspects. First, media categories in Zambia have equal influence, which is theoretically good. Second, they assume that even religious media that claim independence of thought (Chita and Mwale 2017) are equally influenced by the same media that they critique for social injustices. However, if we follow the agenda of each media category, we might not appreciate its implications. For example, having state-owned and commercial media influences community and religious media might not be the best option. Neither would religious media influence the agendas of state-owned media. Consistent with Chita and Mwale (2017) and Banda and Fourie (2004), this undermines the main essence of the difference.

However, further, compounding suggests that ZNBC has a lead influence on particular issues that inform most news agendas in Zambia. In particular, we observe a strong influence on issues pertaining to government policies, investment, foreign policies, politics, and security. On the other hand, there is a strong influence on issues pertaining to social justice, such as environmental degradation, poor working conditions, and mining from the community and religious media. This counteraction suggests that the ZNBC and commercial media ignore such topics. Specifically, topics that presented China with positivity formed structures in which influence was highly centralized in state-owned and commercial media.

Issues that criticized China (i.e., poor working conditions, poor salaries, and environmental pollution) received less coverage and were mostly found in the community and religious media. The ZNBC seemed to respond to such topics only when negatively pressed. For example, stories about environmental degradation and poor Chinese labor conditions received strong criticism and protests in other media outlets but were absent from ZNBC. In instances where they were to respond, the reporting seemed to favor China, except in the case where a Chinese national shot a 16-year-old boy in 2018. Such issues have very low centrality in the network. Additionally, I found pervasive evidence of small-world characteristics which suggested that changes in the framing of a story/issue were being disseminated rapidly within the networks. The spread of such information was easily transferred from the ZNBC and commercial media (which had some direct links with Chinese funding) to community and religious media with little or no connection to China. Further findings suggest that over time, issues about China in the community and religious media are slowly mirroring the ZNBC, and the tone is positively changing.

Among several other implications that emerge from these findings include the fact that the media as a whole, in Zambia, is slowly migrating from reporting stories of social justice that are issue-based stories to stories that mimic the state-owned media—a government mouthpiece. While this trend should be a norm, the state of affairs in Africa calls for media that speaks for the majority poor—a characteristic underlying the definition of community and religious media.

5.2. Limitations and Future Research

This study is limited in several aspects, including its theoretical and methodological approaches. Theoretically, the study attempted to follow the traditional approaches of agenda-setting, but the operationalization in Zambian settings did not seem to provide convincing evidence of the nature of the Zambia-China relationship as it relates to the media. First, most studies have indicated how complex it is to measure agenda-setting influence in an African setting. Essentially, the quantitative nature and firm operationalization procedures that underscore what would be referred to as agenda-setting are sometimes not in conformity with the realities of Africa. That is why scholars such as Kwansah-Aidoo (2001) proposed a qualitative approach to agenda-setting. Second, the collection of valid and reliable data is complex. Therefore, I link different theoretical approaches to understanding intermedia agenda-setting and its influence in an environment where reliable data are difficult to access. Consequently, several approaches have been employed to collect and analyze data. This was done for validity and reliability.

Among other limitations is the lack of reliable data. In our quest for data and through the process of data collection, we received several setbacks. First, what would be considered reliable data was almost non-existent. Despite the unsubstantiated claims, the Zambian government has not provided official statements to counteract the allegations nor are there any official documents available that explain the Zambia-China media relationship. We attempted to reach out to the responsible offices for official documentation. The responses were snowballing until the last person warned us not to pursue the quest for our benefit. In other words, we were advised to stop asking questions about documents that officially describe the Zambia-China media relationship. Therefore, our online data, though coming from questionable news sources, presented a pattern worth investigating.

There are many questions I wish to answer in the future, including the nature of influence, by studying the impact of the magnitude and duration of ZNBC’s lead-lag relationship. In recent years, ZNBC’s audience news consumption has decreased and many people have migrated to online sources. In addition, most people consume their media from broadcast satellites, such as DSTV, but also China’s TopStar Communications Company Limited, a terrestrial pay-TV service. There are unquantifiable arguments that even when China owns 60% of the joint venture, content is still created by the Zambians. Future research exploring this topic, particularly through quantitative and network analyses, is needed to establish these relationships.