Democracy in the Digital Age: Investigating Fake News,Political Polarization,and Media’s Role in Taiwan and South Africa

Abstract

The global proliferation of “fake news” has become a critical concern in media and political discourse, extending beyond the United States and Europe to regions like Africa and Asia. This study conducts an exploratory comparative analysis of Taiwan and South Africa, utilizing data from the Taiwan Institute for Governance and Communication Research (TIGCR) and Afrobarometer, to examine the impact of misinformation and disinformation on democratic processes in these emerging democracies. Our findings reveal that perceived exposure to disinformation is significantly high in both Taiwan and South Africa, and this exposure negatively impacts the two democratic societies. Specifically, we observe correlations between various factors related to disinformation exposure and how people perceive fake news as worsening polarization and affecting the role of media in democracy. The analysis focuses on three key aspects: the relationship between awareness of false information and reliance on various media platforms, the impact of misinformation dissemination on societal polarization, and the role of media in promoting democratic processes. This research provides critical insights into how misinformation affects democracies in Asia and Africa, with broader implications for global democratic health. The study also offers recommendations for future research directions to further understand and address the challenges posed by misinformation in diverse democratic contexts.

Keywords

Misinformation fake news South Africa Taiwan democratic support

Introduction

In the contemporary media landscape, social media platforms have redefined traditional communication models, transforming information dissemination from a one-directional process into a dynamic, interactive exchange. The advent of the Internet has empowered individuals to become both consumers and producers of information, accelerating the spread of news through the ubiquitous use of smartphones. This shift toward a more democratized media environment, however, has introduced significant challenges.

One major issue is the creation of “echo chambers' on social media, where individuals are primarily exposed to information that confirms their pre-existing beliefs” (Adamic and Glance, 2005; Colleoni et al., 2014; Conover et al., 2012). This selective exposure reinforces confirmation bias, limiting the diversity of viewpoints and facilitating the spread of misinformation. These environments are particularly conducive to the proliferation of fake news, as they amplify content that aligns with the dominant ideologies of the group.

The 2016 Brexit referendum highlighted the significant influence of media on political processes and public opinion, with many attributing the surprising results to the widespread dissemination of misleading information online (DiFranzo and Gloria-Garcia, 2017). This event underscored the profound impact that media can have, both in shaping political outcomes and eroding public trust in mainstream media. As confidence in established news sources declines, susceptibility to deceptive information increases, providing a larger audience and greater impact for fake news (Allcott and Gentzkow, 2017).

The issues of truth, veracity, and deception have become central topics in journalism scholarship and public debate, particularly in the context of the 2016 US presidential election and the subsequent use of the term “fake news” by Donald Trump to dismiss critical reporting (Boyd-Barrett, 2019; Waisbord, 2018). While the United States is often at the forefront of these discussions, countries worldwide, including Myanmar (Stevenson, 2018), Nigeria (Funke, 2018), and India (Udupa, 2017), are also grappling with the consequences of fake news.

Globally, the proliferation of disinformation and the weaponization of the term “fake news” by populist politicians have been identified as significant factors in diminishing trust in news media (Knight Foundation, 2018). However, this decline in media trust is part of a broader, long-term trend (Nicolaou and Giles, 2017).

In this article, we contribute to the ongoing exploration of the relationship between fake news and democracy by analyzing data from surveys conducted in 2019 and 2020 in South Africa and Taiwan, both of which have experienced politically motivated disinformation campaigns. This study represents the first comparative quantitative overview of the prevalence of fake news between Africa and Asia. While our findings are exploratory, they provide valuable insights and a foundation for future research across different regions.

Fake news poses a significant threat to both young and mature democracies by disrupting democratic processes, endangering public welfare, and eroding trust in institutions. In young democracies like South Africa and Taiwan, fake news exploits existing political and social vulnerabilities, thereby undermining the democratic foundation. Conversely, in mature democracies, fake news exacerbates political polarization and diminishes public trust in media and government.

This study seeks to explore the similarities and differences in the impact of fake news on democracies with varying cultural backgrounds and levels of democratization. By examining South Africa and Taiwan, the research aims to provide a nuanced understanding of how fake news influences democratic processes across different contexts.

We investigate several key questions: How often do individuals in South Africa and Taiwan encounter fake news? Understanding the frequency of exposure helps to gauge the pervasiveness of misinformation in each context and its potential impact on public opinion. We explore the extent to which fake news influences democratic processes and public perception in these countries, including how misinformation affects political polarization and overall evaluations of the role of media in democracy. By addressing these questions, the study contributes to the comparative analysis of fake news in diverse democratic contexts, offering insights into the global challenge of misinformation and its implications for democratic health.

Our findings aim to shed light on the critical issue of fake news, emphasizing its impact on different democratic systems and providing valuable information for policymakers, media practitioners, and scholars to develop more effective strategies to combat misinformation. By comparing the experiences of South Africa and Taiwan, we highlight both the universal and context-specific challenges posed by fake news, ultimately contributing to a deeper understanding of this complex phenomenon in the contemporary digital age.

The following section will review the literature on fake news and discuss its significance in the context of elections in Taiwan and South Africa. The third section will examine the consequences of fake news, particularly its impact on polarization and democracy. The fourth section will explain the methodology of this study, followed by a presentation of preliminary findings and a discussion. The final section will conclude with the implications of this study for future research and policy development.

Fake news: a contested term and its implications for democracies

The term “fake news” gained prominence in late 2016, yet it remains conceptually ambiguous and broadly applied. Boyd-Barrett (2019) underscores its historical connections to “propaganda,” “disinformation,” and “perception management,” encompassing hoaxes, clickbait, political disinformation campaigns, and outright propaganda (Schapals, 2018). In addition, it can include political satire and parodies (Borden and Tew, 2007). Despite these variations, contemporary definitions emphasize fake news as viral misinformation presented as legitimate news, characterized by intentionality and the potential to mislead (Tandoc et al., 2018).

Tandoc et al. (2018) categorize fake news into six types: satire, parody, fabrication, photo manipulation, propaganda, and advertising/public relations. The primary motivations behind its creation are financial—through click-driven advertising revenue—and political—serving ideological agendas. These motivations parallel traditional public relations (“spin”) and propaganda, often blending both to exploit political fears and generate financial benefits.

The broad scope of fake news complicates efforts to delineate its various forms. For instance, the 2019 Nigerian election saw “social media entrepreneurs” fabricating or amplifying disinformation based on political biases, often for financial gain or political favor (Hassan and Hitchen, 2019). Similarly, during South Africa’s 2016 election, the ruling African National Congress (ANC) was accused of orchestrating disinformation campaigns, including fake election materials, misleading radio broadcasts, and paid social media influencers (Comrie, 2019; Plaut, 2019).

Institutional bodies have critiqued the term “fake news” for its conceptual vagueness. The UK government, following expert recommendations, favors “disinformation” for deliberate falsehoods intended to mislead and “misinformation” for unintentionally shared false information (UK Government, Digital, Culture, Media and Sport Committee, 2018). Likewise, UNESCO rejects the term as an oxymoron that erodes trust in legitimate journalism (UNESCO, 2018).

Despite these concerns, some scholars, including (Chakrabarti et al., 2018) Farkas and Schou (2018), and Tandoc et al. (2018), continue using “fake news” due to its widespread public recognition. Chakrabarti et al. (2018) define it as information lacking full factual support, irrespective of format or platform. The persistent use of this contested term highlights the ongoing struggle to define and combat the spread of false information in democratic societies.

The crisis of misinformation and political manipulation in Taiwan and South Africa

The study of “fake news” has largely focused on the Global North, leaving significant gaps in understanding its impact in other regions. UNESCO (2018) recognizes dis- and misinformation as an emerging global challenge that threatens journalism and democratic stability.

Misinformation, as defined by Thorson (2016), refers to seemingly credible information later debunked as false. Its consequences are profound, undermining democratic institutions, endangering public health, and distorting global discourse (West and Bergstrom, 2021). Misinformation ranges from benign rumors to deliberate political disinformation designed to manipulate public sentiment (Allcott and Gentzkow, 2017).

The digital age has amplified the spread of misinformation, often generating greater engagement than legitimate news sources (Willmore, 2016). Social media accelerates this trend by prioritizing virality over accuracy (Allcott and Gentzkow, 2017). Identifying fake news can be straightforward, but countering its spread is challenging due to psychological biases. Individuals often reject corrections that contradict their preexisting beliefs, and exposure to falsehoods can entrench misinformation (Swire and Ecker, 2018; Lewandowsky et al., 2012).

Historical cases, such as the 2016 U.S. presidential election and the Brexit referendum, illustrate how social media facilitate the breakdown of traditional gatekeeping, enabling widespread disinformation (Vicario et al., 2019). Similarly, recent elections in Taiwan and South Africa demonstrate how misinformation infiltrates political discourse. In Taiwan, fake news circulates widely via platforms such as WhatsApp, Facebook, Line, and PTT (Chang et al., 2020). Political parties exploit disinformation to manipulate voter perceptions: the Kuomintang (KMT) distorts statements from former US officials to cast doubt on the Democratic Progressive Party’s (DPP) foreign policy, while the DPP weaponizes anti-China sentiment against the KMT.

In South Africa, political misinformation is similarly weaponized. The “white monopoly capital” campaign, orchestrated by allies of former President Jacob Zuma and the PR firm Bell Pottinger, used fake Twitter accounts and bots to spread propaganda (Basson, 2018). Social media, particularly Twitter, plays a critical role in disinformation, often blurring the line between satire and serious critique (Wasserman, 2017).

Misinformation in Africa often takes the form of extreme speech, inciting violence or spreading xenophobic and misogynistic narratives via apps like WhatsApp (Jacobs and Wasserman, 2018; Muendo, 2017). Government responses vary, with some imposing restrictions on online spaces. Nigeria’s Minister for Information and Culture raised concerns about fake news during the 2019 elections (Vanguard, 2018), while governments in Cameroon, Ethiopia, and Togo have enacted Internet shutdowns to suppress online criticism (Fleischmann and Stefanello, 2019). However, digital platforms also serve as tools for countering censorship, as seen during the Kenyan government’s 2018 attempt to restrict coverage of opposition leader Raila Odinga’s alternative inauguration (Gathigi, 2018).

The relationship between misinformation, state control, and public trust is complex. In contexts where mainstream media are perceived as biased or elite-driven, audiences often turn to alternative narratives, including satire, gossip, and social media discourse, which may reinforce misinformation rather than counter it (Wasserman, 2017). Despite widespread concern over fake news, little research exists on how African audiences navigate misinformation, the prevalence of disinformation in media consumption, or its impact on political behavior. Comparative research between Africa and Asia is especially lacking.

This study seeks to bridge these gaps by examining fake news as both a technological and social phenomenon. Chakrabarti et al. (2018) identify three primary motivations for spreading fake news in Africa: the desire to be “in the know,” a sense of civic duty to share warnings, and the perception of democratized information access. These motivations are especially relevant where state-controlled media fosters public distrust (Nyamnjoh, 2005).

In South Africa, media efforts to expose corruption have faced threats of censorship and attacks on journalists, particularly during the 2019 elections (R2K, 2015; RSF, 2019a; RSF, 2019b SANEF (South African National Editors’ Forum), 2019). Taiwan, despite enjoying greater media freedom today, has faced similar challenges in its past. Socio-economic inequalities further complicate information credibility, as elite-dominated mainstream media often fail to reflect the lived experiences of ordinary citizens, making emotionally charged fake news more persuasive (Chakrabarti et al., 2018; Wasserman, 2017).

While Taiwan and South Africa differ in media landscapes and political structures, comparing these democracies offers insights into whether media freedom influences the role of fake news in shaping public perceptions and behaviors. Given the limited empirical data on dis- and misinformation in Africa, comparative studies remain rare. This research seeks to provide a quantitative overview of misinformation in Taiwan and South Africa by addressing three key questions:

Frequency of encounter: How often do individuals encounter fake news in Taiwan and South Africa?

Responsibility for prevention: Who do citizens believe is responsible for mitigating the spread of fake news?

Impact on democracy: To what extent does fake news influence democratic processes and the media’s role in governance?

By addressing these questions, this study contributes to the comparative understanding of how misinformation shapes political discourse across different cultural and institutional contexts. The findings will enhance discussions on media literacy, government accountability, and the resilience of democratic institutions in the face of disinformation.

The consequences of fake news: political polarization and media trust in Taiwan and South Africa

The spread of fake news exacerbates political polarization and erodes trust in media, shaping electoral dynamics and democratic stability. While Taiwan and South Africa have distinct political landscapes, both countries face the challenge of misinformation influencing voter behavior, public discourse, and institutional credibility.

Political polarization and electoral dynamics in Taiwan

Taiwan’s political polarization is deeply rooted in debates over national identity, particularly between maintaining de facto independence or pursuing unification with China. Unlike conventional left–right ideological divisions, partisanship in Taiwan is primarily shaped by stances on sovereignty and geopolitical alignments (Hsiao and Yu, 2020).

The Michigan School of political analysis posits that partisanship is a central determinant of political behavior (Layman et al., 2006), a perspective echoed in Taiwan, where partisan-ideological divisions drive electoral choices (Abramowitz, 2010, 2011). Hsiao (2014) found that political attitudes are strongly correlated with party affiliation, reinforcing polarization.

Before the 2014 elections, dissatisfaction with then-President Ma Ying-jeou led many Kuomintang (KMT) supporters toward centrist positions, leaving the party’s remaining base more ideologically extreme. Hsiao and Yu (2020) identified “affective polarization” in Taiwan, where hostility toward opposing parties outweighs policy-based considerations, contributing to democratic instability.

Disinformation amplifies polarization by shaping electoral strategies and voter mobilization. Campaigns increasingly focus on character attacks and misinformation rather than substantive policy debates. Taiwan’s geopolitical tensions further complicate this landscape, as disinformation campaigns—often attributed to the People’s Republic of China (PRC)—seek to shift public sentiment, while pro-independence actors exploit these threats to consolidate their base. The prevalence of misinformation in Taiwan necessitates targeted countermeasures to mitigate its impact on democratic processes.

Economic inequality and media trust in South Africa

In South Africa, political polarization is influenced by deep economic inequalities and racial disparities. Since the first democratic elections in 1994, racial identity has remained a primary determinant of voting behavior, with over 70% of Black voters supporting the ANC, while White voters largely align with opposition parties.

The erosion of media trust is another critical consequence of fake news. The Edelman Trust Barometer (2018) reported a decline in trust in South African media from 45% in 2014 to 35% in 2018. Nyamnjoh (2005) attributes this to state influence and disinformation campaigns. In response, many African media consumers turn to “small media” such as graffiti, underground cassettes, Internet forums, and viral messages on WhatsApp, which often reinforce misinformation rather than counteract it (Spitulnik, 2017).

The manipulation of media for political purposes further exacerbates distrust. In Kenya, political elites have co-opted mainstream media, weakening public confidence in journalism (Odhiambo, 1991). Similar trends in South Africa show that clientelism and political patronage shape perceptions of media independence (Mabweazara, 2018; Mbeke et al., 2010). The ANC’s influence over media narratives and its role in the “white monopoly capital” disinformation campaign exemplify the intersection of misinformation and political power (Basson, 2018).

Linking Taiwan and South Africa: the role of media and political trust

Despite differing socio-political contexts, Taiwan and South Africa face common challenges: growing distrust in media, misinformation shaping electoral outcomes, and political polarization driven by identity-based divisions. While Taiwan’s polarization is rooted in national identity debates, South Africa’s is influenced by economic inequality and historical injustices. In both cases, misinformation exploits societal divisions, deepening distrust in institutions and fueling partisan conflicts.

Media responses to fake news also reflect broader structural challenges. In South Africa, state influence over media fosters public skepticism, whereas in Taiwan, the proliferation of disinformation campaigns—particularly from foreign actors—undermines confidence in political institutions. Journalists in both countries struggle to navigate this complex media environment, as reaffirmations of professionalism and objectivity are often insufficient to counter digital misinformation (Carlson, 2018; Waisbord, 2018).

The transition to the “post-truth” era, despite its conceptual limitations, signifies a broader crisis of trust in journalism and factual information (Vos and Thomas, 2018). Political campaigns’ disregard for factual accuracy, combined with systematic attacks on journalists, weakens democratic resilience and the ability to hold leaders accountable.

Given the discussion above, this study therefore aims to examine the consequences of fake news by addressing two key questions:

Perceptions of media and democracy: How do citizens in Taiwan and South Africa view the role of media in sustaining democracy?

Polarization and political trust: How does perceived polarization influence trust in political institutions in both countries?

By investigating these questions, this research offers comparative insights into how misinformation affects democratic governance across diverse cultural and political landscapes. Findings from this study will contribute to the broader discourse on media trust, democratic resilience, and strategies for combating misinformation in emerging and consolidated democracies alike.

Theoretical underpinnings of political communication in the context of fake news

Diane Mutz (2006) identifies two primary mechanisms through which media shapes political communication. The first pertains to content exposure: individuals are inundated with media that reinforces their ideological leanings, potentially deepening existing political preferences. The second involves shifts in the communicative environment. While media liberalization has increased access to diverse information sources, individuals tend to selectively engage with outlets that align with their preconceptions, thereby reinforcing polarization.

The Theory of Selective Exposure suggests that consistent consumption of ideologically congruent information hardens political attitudes. Over time, individuals become adept at filtering out dissenting viewpoints, preferring media that reaffirms their beliefs (Klapper, 1960; Mutz, 2006). Similarly, the Theory of Biased Assimilation posits that individuals interpret new information through the lens of their existing beliefs. Rather than objectively evaluating counterarguments, they are more likely to question the credibility of information that challenges their views (Lord et al., 1979).

The interplay between selective exposure and political polarization remains a subject of scholarly debate. While some argue that selective exposure fuels polarization, others contend that pre-existing polarization drives individuals to seek out confirmatory information (Mutz, 2006). This reciprocal relationship complicates causal inferences regarding media influence on political divisions.

Selective exposure fosters echo chambers, where individuals insulate themselves within homogeneous media environments, a phenomenon intensified by algorithmic filtering on digital platforms that curates content based on user preferences (Pariser, 2011). Conversely, existing polarization may heighten selective exposure, as entrenched societal divisions lead individuals to actively seek information that affirms their ideological stance, reinforcing the cycle of polarization (Stroud, 2010).

In Taiwan, these dynamics are particularly relevant given the island’s polarized political landscape, where national identity and cross-strait relations dominate discourse. Politicians and partisan groups exploit these communication patterns by disseminating information—or misinformation—strategically tailored to resonate with specific voter segments. This targeted messaging deepens societal divides and underscores the urgency of addressing fake news and political manipulation through media literacy initiatives and regulatory frameworks. A nuanced understanding of these mechanisms is essential for developing strategies to mitigate the adverse effects of disinformation and foster a more informed electorate.

Hypotheses on the interplay between fake news and democratic concerns

The theoretical framework suggests that misinformation, particularly in the form of fake news, exacerbates societal polarization and deepens ideological conflicts. The increasing prevalence of misinformation necessitates a critical examination of its impact on political polarization and broader democratic stability.

Fake news, characterized as deliberately misleading information masquerading as legitimate news, reinforces pre-existing biases and sharpens ideological divisions. Social media further amplifies these effects by fostering echo chambers, where individuals primarily encounter information that aligns with their beliefs while being shielded from dissenting viewpoints (Pennycook et al., 2020; Vicario et al., 2019). This dynamic contributes to a more polarized public sphere, entrenched group identities, and a diminished capacity for constructive political discourse.

This study examines two dependent variables: perceived political polarization and perceived media influence on democracy. The hypotheses explore how fake news affects these critical dimensions of democratic governance in Taiwan and South Africa.

Hypotheses on fake news and political polarization

H1-1: The prevalence of fake news correlates with increased negative sentiment toward political opponents, intensifying affective polarization.

H1-2: Individuals are likely to attribute the production of fake news to political parties, further exacerbating partisan divisions.

H1-3: Higher levels of education correlate with stronger perceptions of political polarization stemming from fake news exposure.

H1-4: Partisanship shapes perceptions of fake news and its effects, with supporters of the ruling and opposition parties holding divergent views on its impact.

Hypotheses on fake news and media’s role in democracy

H2-1: Fake news contributes to increased skepticism toward the media’s role in democracy.

H2-2: Greater news consumption correlates with more positive attitudes toward the media’s role in democracy.

H2-3: Political affiliation significantly influences perceptions of fake news and democracy, with ruling and opposition party supporters interpreting its implications differently.

These hypotheses aim to capture the complex relationship between misinformation and political polarization, particularly in the distinct political landscapes of Taiwan and South Africa. Given the different partisan and ideological cleavages in these countries, it is crucial to assess how fake news interacts with these divides to shape public attitudes and democratic processes. This comparative study will provide empirical insights into how misinformation erodes media trust and exacerbates political polarization, ultimately challenging democratic resilience.

Method: data and measurement section

This study utilizes survey data from the Taiwan Institute for Governance and Communication Research (TIGCR) and Afrobarometer, two extensive databases that provide insights into communication patterns in Taiwan and South Africa. Both datasets span multiple survey waves, allowing for robust cross-sectional and time-series analyses.

The primary objective is to examine how individuals perceive fake news and its impact on democratic discourse. The 2020 TIGCR dataset and 2019 Afrobarometer (8th wave) are particularly relevant, as both include dedicated questions on fake news. Therefore, our analysis primarily relies on cross-sectional data from these years.

Case selection: Taiwan and South Africa as comparative democracies

Taiwan and South Africa serve as representative cases of young democracies in Asia and Africa due to their shared commitment to democratic values and media freedom. While not fully reflective of their respective continents, their media landscapes influence neighboring countries, making them valuable exploratory cases.

Both countries exhibit Western-aligned media ecosystems, facilitating the use of parallel survey instruments and ensuring high comparability. Moreover, Taiwan and South Africa were among the few young democracies where robust survey data could be reliably collected (Lau et al, 2018), enhancing the feasibility of this comparative study.

Challenges and considerations

A key limitation is disparities in Internet penetration, particularly in South Africa, where surveys may over-represent respondents with higher education and socio-economic status. However, this is a common issue in both developing and developed countries, as survey response rates often skew toward specific demographics (Pew Research, 2018).

Despite these challenges, the current availability of high-quality data from TIGCR and Afrobarometer allows for meaningful comparative analysis. Leveraging these datasets, this study provides insights into the role of fake news in shaping democratic discourse across distinct cultural and political contexts.

Data collection and survey implementation

To ensure data quality and country-level reliability, national partners oversaw the survey implementation in each country. Afrobarometer managed data collection in South Africa, while National Chengchi University oversaw the TIGCR survey in Taiwan.

Interviewer training was rigorous in both countries, typically involving multi-day workshops to ensure familiarity with the questionnaire in national and local languages, adherence to sampling protocols, and proficiency in fieldwork methodology.

Survey samples were proportionally distributed across regions, states, provinces, and urban/rural areas, ensuring that every adult citizen had an equal chance of selection. Interviews, conducted in the respondent’s language of choice using electronic tablets, lasted approximately 1 hour, with all responses kept strictly confidential.

TIGCR Sample: 4507 completed questionnaires (Taiwan)

Afrobarometer Sample: 1600 completed surveys (South Africa, 8th wave, post-2019 election)

Margin of Error: ±2–3 percentage points at a 95% confidence level

The TIGCR survey is a multi-year panel study, initiated in 2019, with participants voluntarily continuing their involvement. This continuity contributed to a larger-than-average sample size. In contrast, the Afrobarometer survey was cross-sectional, conducted immediately following the 2019 South African election.

No significant differences were found in key variables across the two datasets, ensuring comparability. Table 1 provides a summary of descriptive sample statistics after weighting, further demonstrating the robustness and reliability of the data collected from Taiwan and South Africa.

Table 1.

Summary descriptive sample statistics after post-stratiﬁcation.

Countries	Taiwan	South Africa
Gender	Female 48.6%	Female 50.06%
Age	M = 33.37; SD = 11.59	M = 38.37; SD = 14.98
Age group
18–29	14.2%	23.7%
30–39	15.9%	23.7%
40–49	18.3%	17.95%
50–59	18.7%	11.51%
Over 60	33.0%	11.32%
Education	M = 3.4; SD = 1.45	M = 2.81; SD = 1.23
Elementary or less	15.63%	15.30%
Junior high	10.8%	23.49%
High school diploma	26.99%	32.42%
Come college	11.29%	16.69%
University and above	35.29%	10.20%
Media use
TV	M = 3.01; SD = 0.91	M = 3.51; SD = 0.90
Newspaper	M = 3.05; SD = 0.89	M = 2.16; SD = 1.13
Radio	M = 2.42; SD = 0.97	M = 3.33; SD = 1.01
Online	M = 2.64; SD = 0.98	M = 2.76; SD = 1.33
Media norms
Freedom	M = 2.98; SD = 1.10	M = 3.60; SD = 1.36
Consumption	M = 2.35; SD = 2.55	M = 3.41; SD = 1.78

Note: TIGCR and Afrobarometer Survey.

Research design

This study examines the impact of fake news on democratic stability through two primary statistical analyses. First, we conducted mean comparisons between Taiwan and South Africa to identify cross-national differences. Second, following Bryan and Jenkins (2016)—who recommend using separate linear regression models instead of multilevel modeling when working with a small number of countries—we fitted three regression models independently for each country.

The analysis focuses on two key dependent variables: perceived political polarization and the perceived negative impact of media on democracy. These variables serve as the foundation for testing our hypotheses on how fake news shapes political dynamics in Taiwan and South Africa. In addition, we assess the role of media consumption patterns in influencing polarization, incorporating measures of media usage from TIGCR and Afrobarometer datasets to examine how different media sources shape public perceptions.

To ensure a comprehensive analysis, we include several demographic control variables, such as age, income, and education level, which account for variations in media engagement, fake news susceptibility, and political attitudes. These controls help isolate the effects of misinformation while allowing for a more robust interpretation of findings within distinct political contexts.

By integrating mean comparisons, regression models, and demographic controls, this research provides a comparative perspective on how fake news influences democratic processes and media trust. The study contributes to a broader understanding of misinformation’s role in polarization and public confidence in media, offering insights into how emerging democracies navigate the challenges posed by digital disinformation.

Analysis

This study employs ordinary least squares (OLS) regression to examine the impact of fake news on political polarization and democratic perceptions during the 2019 South African general election and the 2020 Taiwanese presidential election. These two electoral events were selected to ensure data relevance and coherence, with non-relevant years and missing responses excluded from analysis. OLS regression, a widely used method in econometrics and social sciences, estimates relationships between dependent and independent variables by minimizing the sum of squared residuals, providing interpretable coefficients and statistical significance measures (Greene, 2001).

The study focuses on two dependent variables: (1) perceived political polarization, measured by respondents’ agreement with the statement, “Fake news worsens polarization in the country,” and (2) the perceived impact of media on democracy, assessed through the statement, “Despite concerns about fake news, individual media exposure is good for democracy.” Respondents from Taiwan and South Africa, sampled via TIGCR and Afrobarometer surveys, rated their agreement on a 4-point Likert-type scale ranging from “strongly disagree” to “strongly agree,” with higher values indicating stronger agreement.

The primary independent variable is fake news exposure via media, measured by responses to the question: “How often do you receive information through media?” on a 5-point scale from “rarely” to “always.” Additional models introduce explanatory variables such as support for media freedom, assessed on a 5-point scale measuring respondents’ endorsement of press freedom. Media source selection captures the frequency of receiving political news from TV, newspapers, radio, and online platforms on a 4-point scale (“never” to “often”). The perceived responsibility for fake news is measured by respondents’ attribution of blame to the ruling party, opposition party, or media, on a 4-point scale ranging from “not responsible” to “fully responsible.”

To control for potential confounding effects, the study includes demographic controls: education (categorized from limited education to higher degrees), gender (binary: 1 = male, 0 = female), age (grouped into five categories), and income (measured on an appropriate scale). These controls ensure a comprehensive analysis of the relationship between media exposure, fake news perception, and political polarization in Taiwan and South Africa. By integrating these variables, this research aims to provide comparative insights into how misinformation shapes democratic resilience across distinct political contexts.

The following equations specify the models used in this analysis:

For fake new and polarization:

Base model:

\begin{matrix} Polarization= = β 0 + β 1 * Exposure to Fake News + β 2 * Education \\ + β 3 * Age + β 4 * Gender + β 5 * income + ϵ \end{matrix}

Media freedom model:

\begin{matrix} Polarization = β 0 + β 1 * Exposure to Fake News + β 2 * Media Freedom \\ + β 3 * Education + β 4 * Age + β 5 * Gender + β 6 * Income + ϵ \end{matrix}

Media source model:

\begin{matrix} Polarization = β 0 + β 1 * Exposure to Fake News + β 2 * MediaFreedom MediaFreedom \\ + β 3 * News Sources - TV + β 4 * News Sources - Newspaper \\ + β 5 * News Sources - Radio + β 6 * News Sources - Online \\ + β 7 * Education + β 8 * Age + β 9 * Gender + β 10 * Income + ϵ \end{matrix}

Comprehensive partisan model:

\begin{matrix} Polarization = β 0 + β 1 * Exposure to Fake News + β 2 * Media Freedom \\ + β 3 * News Sources - TV + β 4 * News Sources - Newspaper \\ + β 5 * News Sources - Radio + β 6 * News Sources - Online \\ + β 7 * Ruling Party Support + β 8 * Fake News by Ruling Party \\ + β 9 * Fake News by Opposition + β 10 * Fake News by Media \\ + β 11 * Education + β 12 * Age + β 13 * Gender + β 14 * Income + ϵ \end{matrix}

For the role of media and democracy:

Base model:

\begin{matrix} Media Good For Democracy = β 0 + β 1 * Exposure to Fake News + β 2 * Education \\ + β 3 * Age + β 4 * Gender + β 5 * income + ϵ \end{matrix}

Media Freedom Model:

\begin{matrix} Media Good For Democracy = β 0 + β 1 * Exposure to Fake News + β 2 * Media Freedom + β 3 * \\ Education + β 4 * Age + β 5 * Gender + β 6 * Income + ϵ \end{matrix}

Media source model:

\begin{matrix} Media Good For Democracy = β 0 + β 1 * Exposure to Fake News + β 2 * Media Freedom \\ + β 3 * News Sources - TV + β 4 * News Sources - Newspaper \\ + β 5 * News Sources - Radio + β 6 * News Sources - Online \\ + β 7 * Education + β 8 * Age + β 9 * Gender + β 10 * Income + ϵ \end{matrix}

Comprehensive partisan model:

\begin{matrix} Media Good For Democracy = β 0 + β 1 * Exposure to Fake News + β 2 * Media Freedom + β 3 * \\ News Sources - TV + β 4 * News Sources - Newspaper \\ + β 5 * News Sources - Radio + β 6 * News Sources \\ - Online + β 7 * Ruling Party Support + β 8 * Fake News by \\ Ruling Party + β 9 * Fake News by Opposition \\ + β 10 * Fake News by Media + β 11 * Education \\ + β 12 * Age + β 13 * Gender + β 14 * Income + ϵ \end{matrix}

These models enable a comprehensive investigation of the role of fake news in shaping political polarization and attitudes toward democracy. By integrating demographic, media consumption, and partisan factors, this research provides a comparative framework for understanding how misinformation operates within distinct political and cultural contexts in Taiwan and South Africa.

Results and Discussion: Media, Polarization, and Democracy in South Africa’s 2019 Election

This section presents the findings on the relationship between fake news, political polarization, and perceptions of democracy during the 2019 South African general election. The analysis addresses two key questions: (1) Does exposure to fake news increase political polarization? (2) How does fake news influence perceptions of media’s role in democracy? To explore these dynamics, four regression models were employed: the Base Model, Media Freedom Model, Media Source Model, and Partisan Model.

Political polarization and media influence

The analysis highlights multiple factors shaping political polarization, with results progressively refined across the four models. The Base Model, which includes media exposure, education, age, gender, and income, reveals that age negatively correlates with polarization (β = −0.0575, p < 0.05), suggesting that older individuals are less likely to be polarized. Conversely, higher income levels are associated with increased polarization (β = 0.0428, p < 0.10).

Incorporating perceptions of media freedom in the Media Freedom Model demonstrates a strong positive effect on polarization (β = 0.119, p < 0.01), indicating that individuals who perceive greater media freedom tend to exhibit stronger polarization. The negative effect of age remains significant (β = −0.0596, p < 0.05), suggesting a consistent pattern across models.

The Media Source Model, which examines the role of specific news platforms, reveals differentiated effects by media type. TV news exposure significantly increases polarization (β = 0.0945, p < 0.05), while radio news consumption significantly reduces it (β = −0.0767, p < 0.05). Meanwhile, the impact of perceived media freedom remains robust (β = 0.120, p < 0.01). These findings underscore the varying influence of different media sources, suggesting that television contributes to polarization, while radio may serve as a moderating force.

The Partisan Model further integrates political alignment and exposure to fake news. Supporting the ruling party correlates with lower polarization (β = −0.134, p < 0.10), while fake news propagated by the media significantly increases polarization (β = 0.115, p < 0.01). Notably, media exposure becomes a significant positive predictor of polarization (β = 0.633), indicating a more complex relationship when partisan dynamics and misinformation are accounted for. These results reinforce the argument that media influence, particularly through fake news and partisan narratives, plays a critical role in shaping polarization.

Implications for political polarization

The findings, summarized in Table 2, confirm that fake news exacerbates political polarization, particularly when disseminated through certain media channels. Perceived media freedom consistently correlates with greater polarization, suggesting that while media liberalization enhances information access, it may also deepen ideological divides. Furthermore, the distinct effects of TV and radio news consumption highlight how different media platforms contribute to polarization in varied ways. Finally, partisan alignment and exposure to fake news further amplify societal divisions, underscoring the complex interplay between media consumption, misinformation, and political attitudes in South Africa’s democratic landscape.

Table 2.

Rival models of fake news and polarization in the 2019 South Africa election.

Dependent variable polarization
	Base Model	Media Freedom	Media Source	Partisan
Media Expo	0.0174	0.0206	–0.194	.633
Education	0.0346	0.0268	0.0282	0.0168
Age	–0.0575**	–0.0596**	–0.0575**	–0.0503*
Gender	–0.0383	–0.0277	–0.0250	–0.0430
Income Media freedom	0.0428*	0.0421* 0.119***	0.0396* 0.120***	0.0335 0.119***
News—TV			0.0945**	0.109***
News—Newspaper			–0.0112	–0.0183
News—Radio			–0.0767**	–0.0793**
News—Online			0.0493	–0.0155
Ruling party support				–0.134*
Fake news by gov.				–0.0599
Fake news by opp				0.0306
Fake news by media				0.115***
Constant	3.957***	3.522***	3.462***	3.312***
N	1256	1242	1239	1149
Pseudo R²	.0130	.0348	.0437	.0602
Chi-square	3.295	7.420	5.609	5.188
AIC	3715.87	3642.97	3629.02	3378.89
BIC	3746.69	3678.85	3685.39	3454.59

***

p ⩽ 0.001, **p ⩽ 0.01; *p ⩽ 0.05, one-tailed tests.

AIC—Akaike Information Criterion; Smaller values indicate better model performance. Composite model includes all predictors for other models plus tactical voting.

Source: 2019 Afrobarometer surveys.

Perceptions of democracy

The relationship between media and perceptions of democracy during the 2019 South African election reveals several key factors. The analysis as shown in Table 3 incorporates progressively more variables across different models to provide a comprehensive understanding of how media exposure influences views on democracy.

Table 3.

Rival models of fake news and democracy in the 2019 South Africa election.

Dependent variable media good for democracy
	Base Model	Media Freedom	Media Source	Partisan
Media Expo	–0.0206	–0.0201	–0.0912	–0.0938
Education	–0.0980***	–0.101***	–0.0969***	–0.0539**
Age	–0.00865	–0.00632	0.000106	0.00553
Gender	0.119**	0.120**	0.132***	0.148***
IncomeMedia freedom	0.0771***	0.0761*** 0.0125	0.0751*** 0.0132	0.0651*** 0.0318
News—TV			0.00798	0.0112
News—Newspaper			–0.0797***	–0.0690***
News—Radio			0.00631	–0.0102
News—Online			0.117	0.125
Ruling party support				0.273***
Fake News By Gov.				–0.0875***
Fake News By Opp				–0.0400
Fake News By Media				0.115***
Constant	2.932***	2.889***	2.888***	3.020***
N	1532	1512	1508	1387
Pseudo R²	.0278	.0283	.0354	.0661
Chi-square	8.715	7.313	5.501	6.938
AIC	4271.73	4216.16	4205.72	3813.70
BIC	4303.74	4253.41	4264.22	3892.22

***

p ⩽ 0.001, **p ⩽ 0.01; *p ⩽ 0.05, one-tailed tests.

AIC—Akaike Information Criterion; Smaller values indicate better model performance.

Source: 2019 Afrobarometer surveys.

The Base Model includes fundamental predictors such as media exposure, education, age, gender, and income. Education has a significant negative effect on the perception that media is good for democracy (beta = −0.0980), indicating that more educated individuals are more critical of media’s democratic role. Conversely, gender and income show positive and significant effects, suggesting that women and higher-income individuals are more likely to view media positively in terms of its contribution to democracy.

The Media Freedom Model reveals that adding media freedom does not significantly alter the findings from the Base Model. Media freedom itself is not a significant predictor (beta = 0.0125), while the significance and direction of other variables, such as education, gender, and income, remain consistent.

The Media Source Model incorporates specific media sources such as TV, newspapers, radio, and online news. Results indicate that newspaper exposure significantly decreases the perception that media is good for democracy. Other sources, including TV, radio, and online news, do not show significant effects. The negative impact of education and the positive impacts of gender and income persist, reinforcing the patterns observed in the earlier models.

The Partisan Model incorporates variables related to partisan alignment and exposure to fake news. Supporting the ruling party is associated with a significantly more positive view of media’s role in democracy. Exposure to fake news by the government significantly decreases this perception, while fake news by the media has a significant positive effect, Gender and income continue to be significant predictors, though the negative effect of education is somewhat reduced.

The analysis of the 2019 South African election underscores the significant influence of fake news on both political polarization and perceptions of democracy. As presented in Graph 1 and 2, Media freedom consistently correlates with higher polarization levels, and specific media sources like TV and radio have distinct impacts. Furthermore, partisan alignment and exposure to fake news play critical roles in shaping public perceptions. Education, gender, and income emerge as significant predictors of how individuals perceive the role of media in democracy. More educated individuals tend to be more critical, while women and higher-income individuals view media more favorably. These insights provide a nuanced understanding of the complex dynamics between media, political polarization, and democratic perceptions in South Africa.

Graph 1.

Fake news and polarization in South Africa 2019 Election.

Graph 2.

Fake news and democracy in South Africa 2019 Election.

Empirical findings: media and political dynamics in Taiwan’s 2020 election

Media and political polarization

Applying the same four models used to analyze South Africa’s 2019 election, we investigated the factors influencing political polarization during Taiwan’s 2020 election. As shown in Table 4, these models—Base Model, Media Freedom Model, Media Source Model, and Partisan Model—progressively incorporate additional variables to provide a comprehensive understanding of media influence on polarization.

Table 4.

Rival models of fake news and polarization in the 2020 Taiwan election.

Dependent variable polarization
	Base Model	Media Freedom	Media Source	Partisan
Media Expo	0.0897	0.00897	0.00645	0.00445
Education	0.0652***	0.0642***	0.0582***	0.0385**
Age	0.0231*	0.0221*	0.0222	0.0165
Gender	0.0427	0.0411	0.0545	0.0269
Income	0.0029	0.00391	0.000742	0.00289
Media freedom		0.000668	0.00311	0.00741
News—TV			–0.00252	0.00361
News—Newspaper			0.0413*	0.0328
News—Radio			–0.0168	–0.0122
News—Online			0.0464**	0.0239
Ruling party support				0.0997***
Fake News By Gov.				0.176***
Fake News By Opp				0.0778*
Fake News By Media				0.111***
Constant	0.232**	0.233**	0.0986	0.0765
N	879	870	814	782
Pseudo R²	.028	.028	0.039	0.113
Chi-square	5.069	4.153	3.247	6.952
AIC	3715.87	3642.97	1097.34	986.54
BIC	3746.69	3678.85	1149.06	1056.47

***

p ⩽ 0.001, **p ⩽ 0.01; *p ⩽ 0.05, one-tailed tests.

AIC—Akaike Information Criterion; Smaller values indicate better model performance. Composite Model includes all predictors for other models plus tactical voting.

Source: 2020 TIGCR surveys.

Our analysis reveals several key insights. The Base Model shows that fundamental predictors such as media exposure, education, age, gender, and income influence political polarization. Education has a significant positive effect on polarization, suggesting higher educational attainment is associated with increased polarization. Age also shows a significant positive effect, indicating older individuals are more polarized. Media exposure, gender, and income do not show significant effects in this model.

Incorporating media freedom in the Media Freedom Model does not significantly alter the findings. Media freedom itself is not a significant predictor of polarization, while the positive effects of education and age remain consistent. Gender and income continue to be non-significant.

The Media Source Model reveals that specific media sources influence polarization. Online news exposure significantly increases polarization, and newspaper exposure also has a significant positive effect. Education remains a strong positive predictor. Other media sources and demographic variables do not show significant effects.

The Partisan Model highlights the significant role of partisan alignment and exposure to fake news. Supporting the ruling party is associated with a more positive view of media’s role in democracy. Conversely, exposure to fake news by the government significantly increases polarization, while fake news by the opposition, and by the media also contribute to increased polarization. The significance of education is somewhat reduced, suggesting a complex interplay of factors.

Media and perceptions of democracy

Our analysis also examines how media influences perceptions of democracy. Again, we apply the same four models and present the results in Table 5—Base Model, Media Freedom Model, Media Source Model, and Partisan Model.

Table 5.

Rival models of fake news and democracy in the 2020 Taiwan election.

Dependent variable media good for democracy
	Base Model	Media Freedom	Media Source	Partisan
Media Expo	0.0180	0.0173	0.0141	0.00627
Education	0.0141	0.0125	0.0177	0.0305
Age	–0.0411	–0.0410	–0.0478*	–0.0255
Gender	–0.0191	–0.0255	–0.0599	–0.0234
Income	0.00405	0.00527	–0.00448	–0.00945
Media Freedom		–0.00571	–0.016	–0.0295
News—TV			0.0218	0.00941
News—Newspaper			0.0136	0.00630
News—Radio			–0.0203	–0.00725
News—Online			0.0786**	0.0636*
Ruling party support				0.316***
Fake News By Gov.				–0.245**
Fake News By Opp				0.168**
Fake News By Media				–0.221**
Constant	3.498***	3.526***	3.388***	3.383***
N	880	871	817	780
Pseudo R²	.0072	.0073	.0168	.1062
Chi-square	1.276	1.062	1.375	6.493
AIC	2333.55	2311.14	2169.15	2015.87
BIC	2362.23	2344.53	2220.91	2085.76

***

p ⩽ 0.001, **p ⩽ 0.01; *p ⩽ 0.05, one-s tests.

AIC—Akaike Information Criterion; Smaller values indicate better model performance.

Source: 2020 TIGCR surveys.

In the Base Model , fundamental predictors such as media exposure, education, age, gender, and income do not significantly influence the perception that media is good for democracy. The coefficients for media exposure, education, age, gender, and income are all non-significant.

Incorporating media freedom into the Media Freedom Model does not significantly change the results. Media freedom itself is not a significant predictor. Other variables, such as media exposure, education, age, gender, and income, remain non-significant, indicating that perceived media freedom does not have a direct effect on democratic perceptions.

The Media Source Model shows that specific media sources shape perceptions of democracy. Online news exposure significantly increases the perception that media is good for democracy while other media sources, including TV, newspapers, and radio, do not show significant effects. The negative impact of age suggests that younger individuals may perceive media more positively in promoting democracy.

The Partisan Model highlights the significant role of partisan alignment and exposure to fake news. Supporting the ruling party is associated with a significantly more positive perception of media’s role in democracy. Conversely, exposure to fake news by the government and by the media significantly decreases this perception, whereas exposure to fake news by the opposition increases it. Online news continues to show a significant positive effect.

Our findings from the analysis of Taiwan’s 2020 election confirm our hypotheses that fake news significantly influences both political polarization and perceptions of how media may improve democracy. Partisan alignment emerges as the strongest factor, consistent with expectations.

In terms of political polarization, as shown in higher educational attainment and age are associated with increased polarization. Specific media sources, such as online news and newspapers, contribute to this polarization. Partisan alignment and exposure to fake news further amplify these effects, illustrating the nuanced and multifaceted nature of media’s impact on political attitudes in Taiwan.

Graph 3.

Fake news and polarization in the Taiwan 2020 election.

Graph 4.

Fake news and democracy in Taiwan 2020 election.

Regarding perceptions of democracy, political party affiliation, and positions significantly shape individuals’ views on media’s role in promoting democracy. Support for the ruling party positively influences these perceptions, while exposure to fake news by the government and media negatively impacts them. Conversely, fake news from the opposition has a positive effect on perceptions of democracy.

These insights from the Taiwan TIGCR data highlight the complex dynamics between media, political polarization, and democratic perceptions, reinforcing the significant influence of political context and media consumption patterns on shaping public attitudes.

Conclusion

This comparative study on fake news and democracy in Taiwan and South Africa provides significant insights into the complex relationship between media, partisan alignment, and public perceptions in differing democratic contexts. Our analysis reveals that the partisan model holds the strongest explanatory power, as evidenced by its lowest AIC and BIC scores among all the models tested. This underscores the critical role of partisanship in shaping how voters perceive fake news and the media’s influence on democracy.

Education level emerged as a significant variable in both Taiwan and South Africa, albeit in different ways. In South Africa, higher levels of education are associated with a lower trust in the media’s positive impact on democracy, indicating a strong negative correlation between education level and the belief that media benefits democracy. This suggests that more educated South Africans are more critical of the media’s role in supporting democratic processes. Conversely, in Taiwan, voters with higher levels of education are more concerned about the impact of fake news on societal polarization rather than the overall role of media in democracy. Taiwanese citizens tend to believe that while media can play a positive role in promoting democracy, the proliferation of polarized news stories contributes significantly to societal divisions.

These differences may reflect the developmental statuses of the two democracies. South Africans primarily rely on traditional media sources such as radio and newspapers for political information, which may contribute to a more skeptical view of the media’s role in democracy among the educated. In contrast, Taiwanese voters are more inclined to use online sources, reflecting a different media landscape where the concern is more about the content and its polarizing effects rather than the medium itself.

Our findings also highlight the crucial role of partisanship in shaping voters’ views on fake news and media. In both countries, supporters of the ruling party are more likely to believe that the media is beneficial for democracy, aligning their satisfaction with the news to their partisan biases. This suggests that pre-existing partisan biases significantly influence public opinion, rather than the media alone generating biased news. The ruling party supporters’ positive view of the media contrasts with the opposition supporters’ more critical stance, further illustrating how partisanship shapes media perceptions.

The study’s results indicate that people may have pre-existing biases due to their partisan alignment, which are reinforced by their media consumption patterns. This finding challenges the notion that media alone is responsible for generating biased news, highlighting the interplay between media consumption and partisan biases.

While our research provides substantial and intriguing results, it is important to interpret these findings with caution. The exploratory nature of the study, based on two different datasets from countries with similar but not identical surveys, requires careful consideration of the contextual differences. Despite these limitations, this study offers valuable insights and a starting point for future research to delve deeper into the intricate relationships between fake news, media, polarization, and democracy across various cultural and economic landscapes.

Future research should explore how these dynamics play out in different democratic settings, considering the evolving media landscapes and the role of education and partisanship in shaping public perceptions. Understanding these relationships can help develop strategies to mitigate the adverse effects of fake news on democratic processes and promote a more informed and engaged citizenry. Our findings underscore the importance of considering the broader socio-political context when examining the impact of media on democracy, highlighting the need for nuanced and context-specific approaches in future studies.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Dennis Lu Chung Weng

Kira Alberts

Author biographies

Dennis Lu-Chung Weng is an Associate Professor of Political Science at Sam Houston State University and founding Director of the Asia Pacific Peace Research Institute. He is coauthor of The Myth of War in the Taiwan Strait: Elite Perspectives from Beijing, Taipei, and Washington amid the Yizhou Dilemma, and has published widely in peer-reviewed journals and edited volumes, including Electoral Studies, Democratization, and the Journal of Asian and African Studies. In addition to his academic work, he regularly provides political commentary and serves as a policy advisor on U.S.–China–Taiwan relations.

Kira Alberts is a political scientist and Research Coordinator at the Centre for Research on Democracy (CREDO) at Stellenbosch University. Her work examines the intersection of democracy and digital communication, with a focus on how online disinformation shapes political behaviour and democratic resilience in African contexts.

References

Abramowitz

(2010) Transformation and polarization: The 2008 presidential election and the new American electorate. Electoral Studies 29(4): 594–603.

Adamic

Glance

(2005) The political blogosphere and the 2004 US election: divided they blog. In Proceedings of the 3rd international workshop on Link discovery. pp. 36–43.

Allcott

Gentzkow

(2017) Social media and fake news in the 2016 election. Journal of Economic Perspectives 31(2): 211–236.

Basson

(2018) The end of white monopoly capital. News 24, 2 November. Available at: https://www.news24.com/Columnists/AdriaanBasson/the-end-of-wmc-20181101 (accessed 3 February 2019).

Borden

Tew

(2007) The role of journalist and the performance of journalism: Ethical lessons from “fake” news (Seriously). Journal of Mass Media Ethics 22(4): 300–314. DOI: 10.1080/08900520701583586

Boyd-Barrett

(2019) Fake news and “RussiaGate” discourses: Propaganda in the post-truth era. Journalism 20(1): 87–91. DOI: 10.1177/1464884918806735

Bryan

Jenkins

(2016) Multilevel modelling of country effects: A cautionary tale. European Sociological Review, 32(1): 3–22.

Carlson

(2018) The information politics of journalism in a post-truth age. Journalism Studies 19(13): 1879–1888. DOI: 10.1080/1461670X.2018.1494513

Chakrabarti

Rooney

Kweon

(2018) Verification, duty, credibility: Fake news and ordinary citizens in Kenya and Nigeria. London: BBC News. Available at: http://downloads.bbc.co.uk/mediacentre/bbc-fake-news-research-paper-nigeria-kenya.pdf (accessed 29 January 2019).

10.

Chang

HCH

Haider

Ferrara

(2021) Digital civic participation and misinformation during the 2020 Taiwanese presidential election. Media and Communication 9(1): 144–157.

11.

Colleoni

Rozza

Arvidsson

(2014) Echo chamber or public sphere? Predicting political orientation and measuring political homophily in Twitter using big data. Journal of Communication 64(2): 317–332.

12.

Comrie

(2019) ANC gambles on Twitter influencers. Daily Maverick, 1 March. Available at: https://www.dailymaverick.co.za/article/2019-03-01-anc-gambles-on-twitter-influencers/ (accessed 5 March 2019).

13.

Conover

Ratkiewicz

Francisco

, et al. (2012) Political polarization on Twitter. Proceedings of the International AAAI Conference on Web and Social Media 5(1): 89–96.

14.

DiFranzo

Gloria-Garcia

(2017) Filter bubbles and fake news. ACM Crossroads 23(3): 32–35.

15.

Edelman (2018) Edelman trust barometer. New York: Edelman. Available at: https://www.edelman.com/sites/g/files/aatuss191/files/201810/Edelman_Trust_Barometer_Employee_Experience_2018_0.pdf (accessed 23 January 2019).

16.

Farkas

Schou

(2018) Fake news as a floating signifier: Hegemony, antagonism and the politics of falsehood. Javnost—the Public 25(3): 298–314. DOI: 10.1080/13183222.2018.1463047

17.

Fleischmann

Stefanello

(2019) African governments use internet shutdowns to silence opposition more and more. Euronews, 26 January. Available at: https://www.euronews.com/2019/01/26/african-governments-use-internet-shutdowns-to-silence-opposition-more-and-more-what-can-pe (accessed 10 February 2019).

18.

Funke

(2018) Nigeria is the next battleground for election misinformation. Poynter, 30 November. Available at: https://www.poynter.org/fact-checking/2018/nigeria-is-the-next-battleground-for-election-misinformation/ (accessed 10 February 2019).

19.

Gathigi

(2018) How closure of TV stations has narrowed democratic space in Kenya. The Conversation. Available at: http://theconversation.com/how-closure-of-tv-stations-has-narrowed-democratic-space-in-kenya-91164 (accessed 10 February 2019).

20.

Greene

(2001) Estimating econometric models with fixed effects. Department of Economics, Stern School of Business, New York University.

21.

Hassan

Hitchen

(2019) Nigeria’s “propaganda secretaries.” Mail & Guardian, 18 April. Available at: https://mg.co.za/article/2019-04-18-00-nigerias-propaganda-secretaries/ (accessed 18 April 2019).

22.

Hsiao

(2014) Analyzing Political Polarization Among the Taiwanese Public Using the Party Affect Thermometer. Journal of Electoral Studies in Chinese 21(2): 1–42.

23.

Hsiao

ECH

(2020) Polarization perception and support for democracy: The case of Taiwan. Journal of Asian and African Studies 55(8): 1143–1162.

24.

Jacobs

Wasserman

(2018) Siqalo showed how social media is reshaping protest narratives. News 24, 11 May. Available at: https://www.news24.com/Analysis/siqalo-showed-how-social-media-is-reshaping-protest-narratives-20180511 (accessed 23 October 2018).

25.

Klapper

(1960) The Effects of Mass Communication. The Free.

26.

Knight Foundation (2018) Indicators of news media trust. Miami: Knight Foundation. Available at: https://knightfoundation.org/reports/indicators-of-news-media-trust (accessed 1 January 2019).

27.

Lau

Lombaard

Baker

, et al (2018) How representative are SMS surveys in Africa? Experimental evidence from four countries. International Journal of Public Opinion Research. DOI: 10.1093/ijpor/edy008

28.

Lord

Ross

Lepper

(1979) Biased assimilation and attitude polarization: The effects of prior theories on subsequently considered evidence. Journal of Personality and Social Psychology 37(11): 2098.

29.

Layman

Carsey

Horowitz

(2006) Party polarization in American politics: Characteristics, causes, and consequences. Annual Review of Political Science 9(1): 83–110.

30.

Lewandowsky

Ecker

Seifert

, et al. (2012) Misinformation and its correction: Continued influence and successful debiasing. Psychological Science in the Public Interest 13(3): 106–131.

31.

Mabweazara

(2018) “Reinvigorating ‘age-old questions’: African journalism cultures and the fallacy of global normative homogeneity.” In: Mabweazara

(ed.) Newsmaking cultures in Africa. London: Palgrave MacMillan, pp. 1–27.

32.

Mbeke

Ugangu

Okello-Orale

(2010) The media we want: The Kenya media vulnerabilities study. Nairobi: Friedrich Ebert Stiftung.

33.

Muendo

(2017) Kenya targets WhatsApp administrators in its fight against hate speech. The Conversation, 31 August. Available at: https://theconversation.com/kenya-targets-whatsapp-administrators-in-its-fight-against-hate-speech-82767 (accessed 23 October 2018).

34.

Mutz

(2006) Hearing the Other Side: Deliberative Versus Participatory Democracy. Cambridge University Press.

35.

Nicolaou

Giles

(2017) Public trust in media at all-time low, research shows. Financial Times, 15 January. Available at: https://www.ft.com/content/fa332f58-d9bf-11e6-944b-e7eb37a6aa8e (accessed 2 February 2019).

36.

Nyamnjoh

(2005) Africa’s Media: Democracy and the Politics of Belonging. London: Zed Publishers.

37.

Odhiambo

(1991) Development journalism in Africa: Capitulation of the fourth estate? Africa Media Review 5(2): 17–29.

38.

Pariser

(2011) The Filter Bubble: What the Internet is Hiding From You. Penguin UK.

39.

Pennycook

McPhetres

Zhang

, et al. (2020) Fighting COVID-19 misinformation on social media: Experimental evidence for a scalable accuracy-nudge intervention. Psychological Science 31(7): 770–780.

40.

Pew Research (2018) Public opinion research in Sub-Saharan Africa. Pew Global. Available at: http://www.pewglobal.org/2018/01/03/public-opinion-research-in-sub-saharan-africa/ (accessed 23 October 2018).

41.

Plaut

(2019) South Africa gets help tracking down social media predators ahead of poll. The Conversation, 28 February. Available at: https://theconversation.com/south-africa-gets-help-tracking-down-social-media-predators-ahead-of-poll-112622 (accessed 5 March 2019).

42.

R2K (2015) R2K condemns the return of the secrecy bill! Available at: https://www.r2k.org.za/2015/05/12/secrecy-bill-returns/ (accessed 18 April 2019).

43.

RSF (2019a) Nigeria. Available at: https://rsf.org/en/nigeria (accessed 18 April 2019).

44.

RSF (2019b) Kenya. Available at: https://rsf.org/en/kenya (accessed 18 April 2019).

45.

SANEF (South African National Editors’ Forum) (2019) SANEF shocked and outraged by the actions of EFF leader and its supporters against veteran journalist Karima Brown.’ Available at: https://sanef.org.za/sanef-outraged-and-shocked-by-the-actions-of-eff-leader-and-its-supporters-against-veteran-journalist-karima-brown/ (accessed 18 April 2019).

46.

Schapals

(2018) Fake news: Australian and British journalists’ role perceptions in an era of “alternative facts.” Journalism: Theory Practice & Criticism. DOI: 10.1093/ijpor/edy008

47.

Spitulnik

(2017): Alternative small media and communicative spaces. In: Leslie

Ogundimu

Hyden

(eds) Media and democracy in Africa Oxon: Routledge, pp. 177–206.

48.

Stevenson

(2018) Facebook admits it was used to incite violence in Myanmar. The New York Times, 7 November. Available at: https://www.nytimes.com/2018/11/06/technology/myanmar-facebook.html (accessed 3 February 2019).

49.

Stroud

(2010) Polarization and partisan selective exposure. Journal of Communication 60(3): 556–576.

50.

Swire

Ecker

(2018) ELEVEN misinformation and its correction: cognitive mechanisms and recommendations for mass communication. In Misinformation and mass audiences University of Texas Press, pp. 195–211. University of Texas Press.

51.

Tandoc

Jr Lim

Ling

(2018) Defining “fake news..” Digital Journalism 6(2): 137–153. DOI: 10.1080/21670811.2017.1360143

52.

Thorson

(2016) Belief echoes: The persistent effects of corrected misinformation. Political Communication 33(3): 460–480.

53.

Udupa

(2017): Viral video Mobile media, riot and religious politics. In: Udupa

McDowell

(eds) Media as politics in South Asia. Abingdon: Routledge, pp. 190–205.

54.

UK Government, Digital, Culture, Media and Sport Committee (2018) Disinformation and “fake news”: Interim report. Available at: https://publications.parliament.uk/pa/cm201719/cmselect/cmcumeds/1630/163002.htm (accessed 23 October 2018).

55.

UNESCO (2018) Journalism, fake news & disinformation. Paris: UNESCO.

56.

Vanguard (2018) Fake news, biggest threat to 2019 polls—Lai Mohammed. Vanguard, 13 December. Available at: https://www.vanguardngr.com/2018/12/fake-news-biggest-threat-to-2019-polls-lai-mohammed/ (accessed 11 February 2019).

57.

Vicario

Quattrociocchi

Scala

, et al. (2019) Polarization and fake news: Early warning of potential misinformation targets. ACM Transactions on the Web (TWEB) 13(2): 1–22.

58.

Vos

Thomas

(2018) The discursive construction of journalistic authority in a post-truth age. Journalism Studies 19(13): 2001–2010. DOI: 10.1080/1461670X.2018.1492879

59.

Waisbord

(2018) Truth is what happens to news. Journalism Studies 19(13): 1866–1878. DOI: 10.1080/1461670X.2018.1492881

60.

Wasserman

(2017) Fake news from Africa: Panics, politics and paradigms. Journalism: Theory Practice & Criticism (Published Online Ahead of Print).

61.

West

Bergstrom

(2021) Misinformation in and about science. Proceedings of the National Academy of Sciences 118(15): e1912444117.

62.

Willmore

(2016) This analysis shows how viral fake election news stories outperformed real news on Facebook. BuzzFeed.