I The impact of independent fact-checking on journalism practice in South Africa A research project submitted in partial fulfilment of the requirements for the degree of Master of Arts, University of the Witwatersrand, Johannesburg. Carina van Wyk 1774698 Supervisor: Bob Wekesa 30 April 2021 II Declaration I declare that this research report is my own unaided work. It is submitted for the degree of Master of Arts in Journalism and Media Studies at the University of the Witwatersrand, Johannesburg. It has not been submitted before for any other degree or examination or at any other university. Signed: Carina van Wyk Date: 30 April 2021 III Acknowledgements First and foremost, I would like to thank my husband, Jannie Rykaart, for supporting and encouraging me all the way – from the first word to the last full stop of this research report, and everything in between. Thank you to my supervisor, Bob Wekesa, for your guidance. I would also like to extend my appreciation to all the editors who made time available in their busy schedules to be interviewed. Your insights proved invaluable. IV Contents Introduction 1 1.1. The evolution of fact-checking 1 1.2. Contextual background 5 1.3. Rationale 6 1.4. Research objectives 9 1.5. Research questions 9 1.6. Conclusion 10 Literature review 11 2.1. Introduction 11 2.2. The challenges newsrooms are facing in the digital media environment and its impact on journalism practice 11 2.3. The objectives of fact-checkers and how they assess their impact 14 2.4. The relationship between fact-checkers and the media 17 2.6. Conclusion 19 Theoretical framework 21 3.1. Introduction 21 3.2. Changes in theory in a digital media environment 21 3.3. The dynamics of theories and concepts related to journalism practice and professionalism 23 3.4. Application of the evolving agenda-setting theory 24 3.5. Examining the perceived impact of fact-checking within the “news ecosystem” 26 3.6. Conclusion 27 Methodology 28 4.1. Introduction 28 4.2. Data-collection method: semi-structured, in-depth interviews 29 4.3. Sample selection and size 31 4.4. Data-analysing method: thematic analysis 35 4.5. Declaration from author 35 4.6. Conclusion 35 Presentation and analysis of findings 37 5.1 Introduction 37 V 5.2. The impact of the growth in digital media on journalism practice and professionalism 37 5.2.1. Moving from the traditional to the online media space: the impact on South African newsrooms and news coverage 37 5.2.2. The link between changes in journalism practice due to the digital environment and the growth of independent fact-checking 38 5.3. The impact of fact-checking on newsrooms and news coverage 40 5.3.1. The impact of fact-checking on the news agenda and information ecosystem 40 5.3.2. The impact of independent fact-checking on journalists’ work processes: how and what journalists report on 42 5.3.3. Fact-checking in the fast-paced digital media environment 44 5.4. The use of independently fact-checked content by journalists and the role it plays in promoting accuracy in the media 45 5.4.1. Ways in which the media use the work of independent fact-checkers 46 5.4.2. Independent fact-checking and accuracy 48 5.5. Journalists’ perceptions of dedicated fact-checking and independent fact-checking organisations 49 5.5.1. Moving towards defining the role of independent fact-checking within the field of journalism 50 5.5.2. The media’s perceptions of independent fact-checkers 52 5.6. Conclusion 56 Discussion of findings 58 6.1. Introduction 58 6.2. To what extent is independent fact-checking considered to have an impact on South African newsrooms and news coverage, if at all? (RQ1) 58 6.3. In what ways do the South African media make use of work done by independent fact- checking organisations and to what extent does independent fact-checking promote accuracy in the media, if at all? (RQ2) 60 6.4. How do journalists perceive the role of independent fact-checkers and their relationship with the media, and do they consider independent fact-checking a distinct form of journalism that is growing due to factors related to journalism practice in the digital media environment? (RQ3) 61 6.5. Conclusion 64 Conclusion 65 Bibliography 68 Addendum A: Declaration from the author VI Addendum B: Interview question guide Addendum C: Organisations represented by interviewees 1 Introduction With the practice of independent fact-checking growing worldwide and in Africa (Graves & Cherubini, 2016; Funke, 2019; Bell, 2019; Stencel & Luther; 2020b) this research examines its perceived impact on journalism practice in South Africa. The purpose of the study is to establish whether the work done by fact-checking organisations is considered to have an impact on journalism practice (and by extension practices in newsrooms and news coverage) in South Africa and, if so, to what extent. It also explores the relationship between independent fact- checking organisations and the media, and how journalists perceive the practice of dedicated1 fact-checking as practised by independent fact-checkers. Through this research I aim to contribute to the fairly limited research in the field of independent fact-checking, especially in South Africa. The study is all the more important as research on the influence of fact-checking on journalists and journalistic discourse is particularly scarce (Graves, Nyhan & Reifler, 2015; Cheruiyot & Ferrer-Conill, 2018:964). In light of the Covid-19 pandemic, which broke out shortly after the research started, and the subsequent spread of misleading and false information related to the disease (Ali, 2020; Van der Linden, Roozenbeek & Compton, 2020; World Health Organization, 2020), establishing the possible impact of independent fact- checking on newsrooms and news coverage – and thus on the information consumed by media audiences – is arguably more relevant than ever. 1.1. The evolution of fact-checking As will be indicated in the discussions that follow, independent fact-checking is a fairly new phenomenon globally and in Africa, but fact-checking itself is not novel. It is therefore worthwhile charting its historical trajectory (see Figure 1) before returning to the South African context. In-house fact-checking has been around since at least 1923 when TIME magazine appointed a group of fact-checkers to verify the accuracy of information in articles prior to publication (Fabry, 2017). However, the rise and growth of what Graves (2013:2) calls “dedicated” fact-checking, which refers to verifying claims that are already in the public domain, can be considered a mostly 21st-century phenomenon. The independent fact-checking site Snopes started 1Since independent fact-checking organisations practice “dedicated fact-checking” as described by Graves (2013:2) the words “independent” and “dedicated” are used interchangeably in this study. 2 investigating “urban legends, hoaxes and folklore” in 1994 (Snopes, n.d.), while the first organisations dedicated to assessing “the truth of political claims” became operational in the United States of America (USA) in the early 2000s (Graves & Cherubini, 2016:6). Spinsanity, which was founded in 2001 and has since closed, is referred to by Graves (2013:125) as what “appears to have been” the USA’s “first dedicated fact-checking site”. Two years later FactCheck.org, based at the University of Pennsylvania, followed (Graves, 2013:126-127). FactCheck.org monitors the accuracy of statements by “major U.S. political players in the form of TV ads, debates, speeches, interviews and news releases” (FactCheck.org, n.d.). A blog by Channel 4 News in the United Kingdom (UK) that covered a parliamentary election in 2005 seems to have been the first regular European “source of political fact-checking” (Graves & Cherubini, 2016:6). Since then the practice of fact-checking has spread across six continents. Along with newsroom pressures brought about by the growth of the internet and social media, which is discussed in detail in this study, and a rise in the spread of mis- and disinformation (Berger, 2018), came a steady increase in fact-checking projects and/or outlets worldwide (see Figure 2). The Duke Reporters’ Lab (The Lab), a centre for journalism research at Duke University with a core focus on fact-checking (Duke Reporters’ Lab. n.d.a.), has been tracking fact-checking projects2 around the world since 2014. To be listed, a project – which can refer to a fact- checking outlet, a website or a fact-checking programme – has to regularly publish articles, videos or audio reports that “verify the accuracy of claims made by prominent public figures and institutions; debunk rumors, hoaxes and other forms of misinformation that spread online; or review the status of political promises made by candidates and political parties” (Adair & Stencel, 2016). The number of fact-checking projects on the database has been increasing steadily over the past seven years – showing especially rapid growth during the first half of 2020 “with elections, unrest and a global pandemic [Covid-19] generating a seemingly endless supply of falsehoods” (Stencel & Luther, 2020a). 2 Duke Reporters’ Lab identifies individual fact-checking “projects” or places, including outlets, websites and programmes, where the public can find reliable fact-checking reports. Examples include country- specific pages and projects produced by multinational fact-checking organisations, such as Africa Check. While the Reporters’ Lab counts each major satellite, the International Fact-Checking Network’s signatory list generally counts the overarching organisation (Adair & Stencel, 2016). 3 Figure 1: The evolution of fact-checking journalism Sources: Referenced in section 1.1. Figure 2: The growth of dedicated fact-checking globally Sources: Duke Reporters’ Lab (Adair, 2014a; Stencel & Luther, 2019, 2020a, 2020b) 4 In April 2014, a total of 44 active fact-checking projects were counted by the Lab globally, a number that more than quadrupled in just more than five years to 188 in June 2019 (Stencel, 2019). By October 2019 a total of 210 active fact-checking projects were counted in 68 countries (Stencel & Luther, 2019) and a year later it surpassed the 300 mark for the first time with a total of 304 in 84 countries across six continents (Stencel & Luther, 2020b). The first independent fact-checking organisation in Africa, Africa Check, was launched in Johannesburg, South Africa, in 2012 and has since also opened offices in Dakar in Senegal, Lagos in Nigeria and Nairobi in Kenya (Africa Check, n.d.). Although Africa’s number of fact- checking projects was still much lower than Europe’s 85, Asia’s 82, North America’s 72 and South America’s 40 (Stencel & Luther, 2020b) by October 2020, there are now 21 projects across the continent, which is five times more than the four listed by the Lab in February 2018 (Stencel & Griffin, 2018). The Lab’s latest map of fact-checkers around the globe (Duke Reporter’s Lab, n.d.b.) shows 21 fact-checking projects and/or outlets doing fact-checking across Africa. These are Africa Check in South Africa, Nigeria, Kenya and Senegal; AFP Fact Check in South Africa, Nigeria and Kenya; Dubawa in Nigeria and Ghana; Buharimeter and the International Centre for Investigative Reporting’s FactCheckHub in Nigeria; PesaCheck in Kenya, Uganda and Tanzania, ZimFact in Zimbabwe; Congo Check in the Democratic Republic of Congo; Namibia Fact Check in Namibia; GhanaFact in Ghana; and Matsda2sh (Don’t Believe), Saheeh Masr and Da Begad (Is This Real?) in Egypt. By December 2020 six of the organisations represented on the Lab’s map – Africa Check, AFP Fact Check, PesaCheck, GhanaFact, Congo Check and Dubawa – were also verified signatories of the code of principles of the International Fact- Checking Network (IFCN) (Poynter, n.d.). In my capacity as head of education and training at Africa Check, I manage Africa Facts, an informal network of African fact-checkers and media organisations that are involved in fact- checking activities (Africa Check, n.d.c.). Although Duke Reporters’ Lab and the IFCN’s official numbers are used in this research, it is valuable to note that at least five fact-checking outlets – TogoCheck (TogoCheck, n.d.) Vox Populi in Uganda (Vox Populi, n.d.), Ethiopia Check (Ethiopia Check, n.d.), Fasocheck (Fasocheck, n.d.) in Burkina Faso and MISACheck (MISACheck, n.d) in Mozambique – which joined the Africa Facts network in 2020 and 2021, were not yet featured on the Lab’s database at the time of the research. This means the official https://dabegad.com/ https://dabegad.com/ https://dabegad.com/ 5 count of fact-checking projects and organisations on the continent by the Lab will in all likelihood be higher soon. 1.2. Contextual background For the purposes of this research, I will distinguish between fact-checking as an internal editing function in newsrooms before publication and dedicated fact-checking as practised by independent organisations, specifically to verify claims in the public domain, i.e. after publication. Fact-checking has a traditional meaning in journalism that refers to internal processes followed in order to verify facts prior to publication (Graves & Amazeen, 2019:1). Internal fact-checking, which prevents incorrect information from being published, is mainly concerned with whether a reporter quotes someone correctly and not whether the claim made by the person who’s quoted is true, while dedicated fact-checkers assess whether claims are factually correct or not (Graves, 2013:3). Graves and Amazeen (2019:1) also describe it as an “evidence-based analysis” of statements made by politicians and other public figures or published by the media in order to verify their accuracy. Dedicated fact-checking services also provide internet users with guidance in terms of the credibility of online content (Brandtzaeg, Følstad & Domínguez, 2017). Some fact-checking organisations who are signatories of the IFCN, verify and flag false social media content as part of Facebook’s Third Party Fact- Checking Program, which was launched in late 2016 (Tardáguila, Funke & Benkelman, 2019) and by December 2020, was being worked on by more than 80 organisations in 60 languages (Full Fact, 2020). Some dedicated fact-checking outlets are affiliated to news organisations or based in newsrooms, especially in the USA (Stencel, 2019). Others are independent of media companies and often based at universities or within areas of civil society, usually aimed at better informing people and “promoting fact-based public discourse” (Graves & Amazeen, 2019:1). The need for dedicated fact-checking and the rapid growth within this field are widely attributed to the challenges the fast-paced digital media environment, with its quick turnaround times and short deadlines, brings (Cheruiyot & Ferrer-Conill, 2018:967; Graves & Amazeen, 2019:4). The public’s growing access to the internet and social media has escalated the challenges in the media environment with traditional journalists losing their dominance as gatekeepers in deciding what should be published as news, and information making it into the public domain irrespective of the media’s actions (Vos 2015:6). 6 These factors are all contributing to what has been referred to as “information disorder” (First Draft, n.d.), a “tidal wave of misinformation” (Amazeen, 2017b:2) and an “egregious rise in the extent of misinformation and disinformation available in the public domain” (Wekesa et al., 2017:21). According to Seaton et al. (2020:578), Covid-19 has “revealed and accelerated an information crisis” and they further argue that fact-checking “sits within a far more complex and chaotic media context” than a decade ago and “its expertise and understanding has never been so important”. The World Health Organization (WHO) (2020:2) says the pandemic has been accompanied by an “infodemic”, which it describes as an “overabundance of information, some accurate and some not, that makes it hard for people to find trustworthy sources and reliable guidance when they need it”. This makes it difficult for the public to know how to “protect themselves and others, and help mitigate the impact of a disease”, which is “exacerbated by the global scale of the emergency, and propagated by the interconnected way that information is disseminated and consumed through social media platforms and other channels” (WHO, 2020:2). One of the first studies about fact-checking in Africa, and more particularly South Africa, found that although fact-checking coverage gained traction over the five-year period from 2012 to 2017, the South African media – in contrast to North America (Stencel, 2019) and western Europe (Graves & Cherubini, 2016:8) – in general were not practising dedicated fact-checking within newsrooms. There was a steady increase, however, in the publishing of fact-checking reports written by independent fact-checkers such as Africa Check (Wekesa et al., 2017:25). This research goes a step further and explores whether independent fact-checking is considered to have an impact on journalism practice and whether it improves the accuracy of news reporting in South Africa. This is done by a thematic analysis of qualitative semi-structured interviews with editors across a broad spectrum of news outlets in the country. In order to make sense of this new form of journalism, I will use a combination of theories and concepts in the field of journalism practice and professionalism (especially for new media), and the intermedia- and network agenda-setting theories. 1.3. Rationale Having laid out the background and context to the study in terms of the evolution and key topics pertaining to fact-checking journalism, in this subsection I will explain the importance of the research. 7 As mentioned in section 1.1., dedicated fact-checking only started gaining traction in the USA in the early 2000s and was practised for the first time in South Africa and Africa in 2012, when Africa Check was launched, a mere nine years ago (Africa Check, n.d.). With this form of journalism being fairly new and rapidly growing (Stencel, 2019), the rationale for this study was not only to make a contribution to fact-checking research globally but even more so on a continent and in a country where it is – just like in the rest of the so-called global south – severely understudied when compared to North America and Europe (Van Wyk, 2016:3; Wekesa et al., 2017:21; Cheruiyot & Ferrer-Conill, 2018:972; Dias & Sippitt, 2020:605). Similar to Wekesa et al’s. study (2017), research investigating fact-checking coverage in a variety of newspapers with the highest circulations across the USA, suggested that although there’s been an increase in dedicated fact-checking or fact-checking coverage, these practices were mostly pursued by newspapers that have dedicated fact-checkers within their newsrooms (Graves et al., 2015:16). Graves et al. (2015:16,17) suggested further research across the news landscape – not only print, but also broadcast and online outlets – to validate the above findings and establish the effects of independent fact-checking on general political coverage, which is partly what this research will aim to do, but in the South African context and not on political news coverage only. The same study also suggests that the practice of dedicated fact-checking has a lot of potential to grow, and that the media is more receptive to fact-checking than what one might think when looking at their reporting (Graves et al., 2015:7), while a study about fact-checking and data- driven journalism in Southern Africa indicates that there’s potential for independent fact- checking “to slowly redefine how news is produced” (Cheruiyot & Ferrer-Conill, 2018:972). Both these studies suggest the need to investigate the impact of fact-checking on journalism practice. To further motivate the reasons for this study: despite academics referring to fact-checking as a journalism genre (Amazeen, Thorson, Muddiman & Graves, 2018:30) as well as an increase in literature examining the effect of fact-checking on the general public and the behaviour of politicians, globally research about the influence of fact-checking on journalists and journalistic discourse is scarce (Graves et al., 2015; Cheruiyot & Ferrer-Conill, 2018:964). Graves et al. (2015:16) found journalists interpret fact-checking in different ways, with some under the impression that it may refer specifically to stories that “formally research a specific claim by a 8 public figure” and others broadly understanding it as “any reporting which seems to challenge political rhetoric”. Wekesa et al.’s (2017:26) study also indicates a lack of understanding in some South African newsrooms about what fact-checking is and how it can be considered a specific journalistic genre. When taking into account the aforementioned factors – that independent fact-checking in Africa is new and fast-growing (Stencel, 2018), not well understood by journalists (Wekesa et al., 2017:26), understudied (Van Wyk, 2016:3) and could have an impact on journalistic practice and how news is produced (Cheruiyot & Ferrer-Conill, 2018:972) – obtaining and providing knowledge in this field could prove invaluable for other researchers, media practitioners and fact-checking organisations, not only in South Africa and the rest of Africa but worldwide. It also contributes towards concepts in the field of journalism practice and professionalism by establishing whether the growth in independent fact-checking and the work done by such organisations are perceived to have an impact on the way journalists approach and write their stories. This research, to some extent, builds on Wekesa et al.’s (2017) study, which showed that the publishing of fact-checking findings by the South African media is increasing. Through the analysis of qualitative in-depth, semi-structured interviews with editors from a variety of news organisations, including online, print and broadcast, the study also aims to provide an in-depth look at the relationship between independent fact-checking and journalism, how fact-checking is perceived by journalists and whether the growth of fact-checking is considered to have an impact on the media and the accuracy of their news coverage. With the flood of false information accompanying the Covid-19 pandemic, I argue that research to determine the likely impact of independent fact-checking on journalism practice – and thus the information consumed by media audiences – is more necessary than ever. Democracies rely on “well-informed and politically educated citizens” (Turčilo & Obrenović, 2020), which means information disorders threaten democratic values and good governance by denying the public access to accurate information. Studying the “actual and potential impact” of fact- checking as a way to counter misinformation could therefore lead to a better understanding of “the contribution that fact-checking can make to our democracies” (Dias and Sippitt, 2020:61). 9 Being the head of education and training at Africa Check, I am aware that my role at the fact- checking organisation could introduce the potential of bias, but I believe that the value of research done in this field – especially in South Africa – by far surpasses the possible harm of bias (see Addendum A). In summary, a shortage in intellectual literature that investigates the actual and potential impact of dedicated fact-checking on journalism globally and in South Africa, the novelty and dynamic nature of fact-checking, and the contribution fact-checking can make to democracies clearly indicate the relevance of this study. 1.4. Research objectives This research investigates whether the growth in independent fact-checking and the work of fact-checking organisations are considered to have an impact on journalism practice: what journalists report on and how accurately they report on it. Through semi-structured qualitative interviews with editors from a variety of South African media organisations, I gain insights into the relationship between fact-checking and journalism and how fact-checking is perceived by journalists. Ultimately, I aim to contribute to research in the fairly understudied and relatively new field of fact-checking. 1.5. Research questions The research is guided by the following question: To what extent is independent fact-checking considered to have an impact on journalism practice in South Africa? In order to answer the above research question, the following sub-questions are included: RQ1: To what extent is independent fact-checking considered to have an impact on South African newsrooms and news coverage, if at all? RQ2: In what ways do the South African media make use of work done by independent fact- checking organisations and to what extent does independent fact-checking promote accuracy in the media, if at all? RQ3: How do journalists perceive the role of independent fact-checkers and their relationship with the media, and do they consider independent fact-checking a distinct form of journalism that is growing due to factors related to journalism practice in the digital media environment? 10 1.6. Conclusion In section 1, I provided background on the growth of the practice of independent fact-checking worldwide, explained the rationale for undertaking the research, discussed the aims of the study and set out the research questions. The next section will cover literature applicable to the study. 11 Literature review 2.1. Introduction Literature covering the following themes are relevant to this research: the challenges newsrooms are facing in the digital media environment and its impact on journalism practice, the objectives of fact-checkers and how they assess their impact, the relationship between fact- checkers and the media, and the growth of dedicated fact-checking as a new form of journalism. 2.2. The challenges newsrooms are facing in the digital media environment and its impact on journalism practice As will be discussed in this section, a key challenge that newsrooms face globally is job losses due to the digital revolution, a drop in advertising income and the subsequent economic pressure on business models. These job losses have an impact on the quality of journalism, which is linked to the core of the study, fact-checking journalism. It is estimated that newsroom employment across the newspaper, radio, broadcast television, cable television, and digital-native3 sectors in the United States declined by almost a quarter from 2008 to 2019, amounting to about 27 000 job losses (Grieco, 2020). Not surprisingly, newspapers were hit the hardest, in 2019 employing only half the number of journalists they did 11 years before (Grieco, 2020). Came 2020 and the Covid-19 pandemic and the situation became more dire with scores of newsrooms across the country either closing, laying off or furloughing employees, or cutting salaries (Hare, 2021). In South Africa, it is estimated that the “professional journalist workforce” in the country, which is believed to have been around 10 000-strong in 2008, was “slashed in half” in the decade leading up to the publication of the State of the Newsroom 2018 report by Wits University, with many of the job cuts affecting senior, experienced journalists (Daniels, 2018:18). The School of Journalism and Media Studies at Rhodes University found that revenues of what is often referred to as the “mainstream media” are continuing to decline sharply as digital technologies and social media platforms are “attracting audiences away from local content and rapidly reducing income for news organisations” (Dugmore, 2018:2). A report by the university further 3 News sites that started as digital only as opposed to those that started as broadcast or paper-based publications: https://www.waldenu.edu/online-bachelors-programs/bs-in-communication/resource/what-is- digital-native-media https://www.waldenu.edu/online-bachelors-programs/bs-in-communication/resource/what-is-digital-native-media https://www.waldenu.edu/online-bachelors-programs/bs-in-communication/resource/what-is-digital-native-media 12 states that “newspaper and print magazine circulations, which have been dropping steadily since about 2007, were now plummeting” while the slight increase in online news consumption is not enough to counteract the declining income of legacy news operations “and some of the lost legacy audience is not showing up online” (Dugmore, 2018:3). Findlay, Bird & Smith (2018) found that South African media organisations struggle to earn the income they used to through print publications in an online space. The State of the Newsroom 2019-20 report painted yet another bleak picture of restructuring and job losses, which, according to Krüger (2020:4) illustrates the “deep crisis facing the traditional business model of journalism”. Not surprisingly, in 2019 newspaper circulation further declined, accompanied by warnings of more possible job cuts at Independent Media and Mail & Guardian, and retrenchments and salary freezes at Tiso Blackstar (Finlay, 2020:7-8). Tiso Blackstar has since sold its news assets to Lebashe Investment Group that now houses titles such as Sunday Times, TimesLIVE, Business Day, BusinessLIVE, Sowetan and SowetanLIVE under Arena Holdings (TimesLIVE, 2019). Covid-19 lockdown restrictions, first announced on 23 March 2020, and its severe economic impact, accelerated the challenges that the South African media are experiencing with declining advertising income, the closure of several print publications, further job losses and a threat of widespread retrenchments (SANEF Management Committee, 2020:3). According to an interim report on the impact of the Covid-19 pandemic on journalism published by the South African National Editors’ Forum (SANEF), “print media consumption has been devastated” and even though broadcasters and online platforms have experienced a “dramatic uptake in their production of news as citizens seek sources of trustworthy, credible information in the time of uncertainty”, difficulties faced by the media industry have intensified (Rumney, R, 2020:11). Not surprisingly, job losses have had a huge impact on journalists’ work processes and the quality of journalism globally. Due to economic pressures, news organisations have had to start sharing content and, because of the decline in staff in American, Canadian and UK newsrooms, fewer reporters now have more work to do, with changes not only in terms of covering more news events, but also in how they are covered in these “converged newsrooms” (Compton & Benedetti, 2010:494). The journalists who remain are “required to post to the Internet, collect audio and video clips, shoot digital photographs, post updates to blogs and most recently, update live to Twitter” (Compton & Benedetti, 2010:494-495). 13 South African media is no exception when it comes to changes and challenges in newsrooms. The State of the Newsroom 2017 newsroom survey found that the majority of employees at the media organisations surveyed4 had fewer than 15 years’ experience (Otter & Grant, 2017:28-47) and that staff at some media companies – who used to work exclusively for one news title – were now creating content for “content hubs”, which was then distributed among different titles, meaning the same content “is likely to be published by a number of titles in the group” (Otter & Grant, 2017:30). Economic pressure has had a domino effect, leading to a shortage of mentors to upskill young journalists, the demise of beat journalism – which used to allow journalists to specialise in a specific field – (Daniels, 2018; Jordaan, 2019) and therefore a lack of in-depth reporting (Findlay et al., 2018). However, it is not only economic pressure and job losses that prove challenging in the online space. With social media providing a platform for inaccurate information to spread widely (Dijkstra et al., 2018:1; Wardle & Derakhshan, 2018:49) and journalists using social media as a news source (Thomas, 2013; Findlay et al., 2018), they arguably run a greater risk of publishing or broadcasting information that is factually incorrect than when interviewing trusted contacts and experts in a specific field. Findlay et al. (2018) found that South African journalists can’t fulfil their gatekeeping role to the extent that is necessary as they don’t have time to adequately “fact-check and do justice to stories”. One could further argue that a reduction in data costs (Goldstuck, 2020; Khumalo & Van der Merwe, 2020; Malinga, 2020) could lead to an increase in social media usage and exacerbate the spread of false information. Findlay et al. (2018) describe the impact of social media and digital technology on South African newsrooms as follows: One of the most striking changes indicated in the shift [to digital] is the huge and unrelenting pressure on journalists. This stems not only from the fact that there are simply fewer journalists (because of the failing economic models and rampant retrenchments in the sector), but also because of the increased expectations of individual journalists to churn out multiple stories in a quicker turn-around time with more than just written copy (multimedia content, etc.). They are also now expected to keep 4 Data was obtained from 25 print titles and 16 online titles published by Independent Media, Media24 and M&G Media, and five independent publishers – the AmaBhungane Centre for Investigative Journalism, Daily Maverick, GroundUp, The Daily Vox and Health-e News. 14 tabs on an ever-increasing volume of potential content through platforms like social media ... This has critical implications for the quality of reporting being undertaken. It is against this backdrop of challenges that the aims of this study include to establish whether journalists perceive fact-checking to be a form of journalism that is necessitated by factors related to changes in journalism practice in the digital media environment. 2.3. The objectives of fact-checkers and how they assess their impact As we investigate the possible impact of independent fact-checking on journalism practice in South Africa, it is valuable to look at what fact-checkers’ objectives are. A study about fact- checking and data-driven journalism in sub-Saharan Africa, indicated that the aims of non-profit fact-checkers are a combination of, on the one hand, traditional normative journalistic goals, for example, informing the public, seeking truth and verifying the accuracy of information, and, on the other, attempts to create and establish data-driven tools that “change the ways in which news organisations operate and how users access data” (Cheruiyot & Ferrer-Conill, 2018:972). To bring us closer to understanding their aims and how they assess their impact, let’s consider how the two fact-checking organisations with offices in South Africa – Africa Check and AFP Fact Check – set out their roles and functions. Africa Check describes itself as a “non-partisan organisation that exists to promote accuracy and honesty in public debate and the media in Africa” (Africa Check, n.d.b.), its ultimate goal being to “strengthen democracy, foster engaged citizenship and improve life outcomes” (Africa Check, n.d.a.). The organisation also debunks “dangerous false statements” on social media platforms, including WhatsApp (Africa’s first independent fact-checking organisation, 2021; Africa Check, n.d.d.; What’s Crap on WhatsApp?, 2021). AFP Fact Check, the digital verification service of the Agence France- Presse (AFP) news service, which has a bureau in Johannesburg, monitors online content to “investigate and disprove false information, focusing on items which can be harmful, impactful and manipulative” (AFP Fact Check, n.d.). Both these organisations have been verifying the accuracy of Facebook content through the social media platform’s third-party fact-checking programme since 2018 (Rawlins, 2018). Although I found that fact-checking research that specifically refers to or is based on media or other academic theory is fairly limited, looking at the above-mentioned finding of Cheruiyot & Ferrer-Conill (2018:972), one could argue that the objectives of fact-checkers span across a 15 broad spectrum of journalism theory – dating back more than a century, with the normative theories marking the first journalism studies in the early 1900s, up to the latest theory applicable to the digital media environment and the changes it brought about. The normative theories of the press looked at the ideal views of the role the media should play in society and how these functions are affected by the freedom of, and restrictions on, the media (Fourie, 2001:269). In a study of three non-profit fact-checking organisations “outside the well- studied media environments of North America and Europe” (Van Wyk, 2017:2), it was found that all of them – Chequeado in Argentina, FactChecker in India and Africa Check – fulfilled at least three functions of the social responsibility theory through their work (Van Wyk, 2017:91). These functions were enlightening society, providing the public with information to facilitate and further accurate debate, and keeping governments to account (Van Wyk, 2017:91). In research that examines journalists’ motivations for doing fact-checking, Graves et al. (2016:106) describes fact-checking as “one of the most significant innovations in journalistic practice in recent years” becoming popular among journalists in the United States, with the majority of print and broadcast news organisations doing some form of fact-checking. They found that the practice of fact-checking is on the rise mainly because it “appeals to the professional values and status concerns of journalists” (Graves et al., 2016:121). In the USA, fact-checking has been embraced by leading news organisations such as the New York Times, while PolitiFact won the Pulitzer Prize for its coverage of the 2008 election, “a clear signal of the new genre’s embrace by journalistic elites” (Graves et al., 2016:107). In 2007 PolitiFact started as an election project of Florida’s largest daily newspaper, the Tampa Bay Times (then the St. Petersburg Times), rating the accuracy of certain statements made by politicians. Direct ownership of PolitiFact was transferred to the Poynter Institute for Media Studies in 2018 to allow it to function as a not-for-profit news organisation (Drobnic Holan, 2019). Until now, research has mainly been focused on the “short-term, persuasive effects” of fact- checking on people’s beliefs (Dias & Sippitt, 2020:605), which doesn’t provide a clear picture of what fact-checkers consider the impact of their work to be. According to fact-checking organisations Africa Check, Chequeado in Argentina and Full Fact in the United Kingdom, there is value in publishing fact-checking reports as they might reach influential audiences of other journalists and policy makers using fact-checkers’ work, and they raise awareness among governments, the media and pressure groups that they might be fact-checked, but they believe 16 publishing fact-checking reports is not enough (Africa Check, Chequeado & Full Fact, 2019). These organisations thus seek corrections on the record and encourage people who were found to have shared inaccurate information not to do so again. If a statement is found to be incorrect, unproven or misleading, Africa Check contacts the source of the claim, presents them with the evidence that led to the finding and requests that the source amends the claim in line with certain recommendations (Africa Check, n.d.d.). Dias & Sippitt (2020:609) state, however, that it’s not yet clear how such actions affect the “future behaviour” of politicians and journalists. Graves and Amazeen (2019:8) suggest considering the “potential impact” of fact-checking not only on individual attitudes and the behaviour of public officials but also on journalistic practice, as the effects on journalists are not clear (Graves et al., 2015:1). Graves et al. (2015:1) suggest that fact-checkers could help to provide clarity on factual disputes, making it more likely that public figures who repeat falsehoods might be challenged by the media while “the increasing use of the format across the profession may encourage other reporters to question political claims in their own coverage”. This has indeed happened in countries like the USA where media organisations like CNN and The Washington Post have journalists practising fact-checking themselves (CNN, n.d; The Washington Post, n.d.). The actual impact of fact-checking on the media could be even more difficult to determine than its impact on politicians’ behaviour (Mantzarlis, 2015). Mantzarlis (2015) suggests that the “uptake of practices that are specific to fact-checking” rather than only the coverage of fact- checking by the media, should be used as an indicator of the impact of fact-checking. Brandtzaeg et al. (2017:114) suggests that the effectiveness of fact-checking and verification depends on whether these services help users reach their goals, whether they are perceived to be necessary, whether they increase users’ job performance and enable them to “discern factual information from other online content”. A research study commissioned by Africa Check to determine the awareness of and impact of the organisation’s work in Nigeria found that Africa Check needs to improve its visibility and create more awareness around its work in order to increase its impact on public debate in the country (Amobi, 2019:33). The findings further suggested a need to “spread the word of Africa Check’s activities on new and traditional media” and indicated that none of the interviewees became aware of Africa Check through traditional media platforms (Amobi, 2019:36). 17 In this subsection, the objectives of fact-checkers were elaborated on. These objectives include promoting accuracy in public debate and the media. In the next section I will discuss the relationship between independent fact-checkers and the media. 2.4. The relationship between fact-checkers and the media In order to assess the possible impact of independent fact-checkers on journalists’ work processes and news coverage, it is worthwhile to investigate the relationship between the two parties – a relationship that could be considered quite complex. On the one hand, fact-checkers depend on traditional newsrooms to publicise their work for a wider reach, while, on the other, they often criticise work done by the media and consider it to be an institution in need of change (Graves & Cherubini, 2016:6). Graves et al. (2015:16) describe the value each of the two parties can derive from the other as follows: In some instances, fact-checkers can help to create a consensus on factual disputes, making it more likely that elites who repeat falsehoods on issues ranging from vaccine risks to birth certificates will be challenged by other journalists. Widespread coverage of fact-checkers’ conclusions also helps to increase the number of people who are exposed to corrective information, widening its audience beyond the relatively few who make the effort to visit fact-checking sites. In a qualitative evaluation of Africa Check’s work5, commissioned by the Poynter Institute, interviewees whose work had been fact-checked responded to it in different ways, with two saying Africa Check did not properly take into account the resource, skills and capacity shortages within newsrooms when fact-checking news content (Finlay, 2016:7). Some journalists also felt that the media should not be held accountable for inaccuracies in “third-party content” such as media releases and that Africa Check should take into consideration that content is created to “achieve a particular effect in audiences” (Finlay, 2016:7). Although interviewees in general didn’t agree on whether or not Africa Check had a clear impact on public conversation, two interviewees said that it did have an effect “on the way their organisations 5 It is important to note that representatives from media organisations only made up about a quarter of the sample size. There were 26 interviewees, of which seven were Africa Check staff, and the rest a mix of representatives from research organisations, the health sector, media (online, print and columnists/public commentators) and the development/civil society sector (Finlay 2016:3). 18 worked, influencing their editorial decisions and producing a sense of caution when working with statistics” (Finlay, 2016:7). According to a study by Kubheka (2017:19), the work done by Africa Check has encouraged newsrooms to take fact-checking seriously – both because of the training the organisation does in newsrooms and the possible embarrassment should an article be found to contain information that is not correct. One of her interviewees, former fact-checker Ina Skosana, who once worked for the Mail & Guardian, described how embarrassed the paper was to have a story rated as “incorrect” by Africa Check, and “quickly improved its internal processes” (Kubheka 2017:19). More research that would cast light on why fact-checking is gaining traction in some media organisations and not in others, has also been suggested (Graves et al., 2016:121). 2.5. The growth of dedicated fact-checking as a new form of journalism One of the major pursuits of this study is to examine the status of fact-checking as a professional journalistic practice and hence I look at whether it is considered a distinct form of journalism. Although using different descriptions of the practice, several academics agree that fact-checking is a journalistic genre (Graves, 2013:1; Amazeen et al., 2018:30; Juneström, 2020). Journalistic genres are described as “sustainable groups of published items with common conceptual and formal attributes” (Serdali, Ashirbekovab, Orazbekulya & Abiev, 2016:1079). In 2014, Tim Franklin, then president of the Poynter Institute journalism school in St. Petersburg, Florida, referred to fact-checking as a “new form of accountability journalism” vital to democracies worldwide (Adair, 2014b). This point of view is shared by Luengo & García-Marín (2020:405,411) in light of a “post-truth discursive context in which the importance of facts and scientific evidence is diminished to assert ideological positions”, highlighted by the spread of Covid-19-related misinformation. Graves (2013:15) describes it as a type of “annotative” journalism that “assembles news out of other stories, addresses gaps in those stories, and is designed in myriad ways to be incorporated into future stories”. The study also examines whether the practice of fact-checking came about and is growing due to factors related to journalism practice in the digital media environment. This environment, with its quick turnaround times and increased deadline pressures (Cheruiyot & Ferrer-Conill, 19 2018:967; Graves & Amazeen, 2019:4), as well as the media losing their dominance as gatekeepers (Vos, 2015:6) has led to the continuous spread of false information in the public domain (Wekesa et al., 2017:21). Amazeen (2017a:5) states that the spread of fact-checking internationally appears to be a reform effort to address the deterioration of journalism standards, while it may also “be understood as a democracy-building tool that emerges where democratic institutions are perceived to be weak or are under threat” (Amazeen, 2017b:1). Graves (2013:291) refers to fact-checkers as “a new species in the media ecosystem” and argues that the work done by “these new journalistic actors” offers insights into the “systemic changes in the news media”. South African newsrooms, in contrast to the USA, haven’t embraced dedicated fact-checking journalism similar to that done by fact-checking organisations, but are mostly republishing these organisations’ reports. Drawing from the results of a study by Wekesa et al. (2017:23) on the prevalence of fact-checking reporting, this research is pinned on the premise that the coverage of fact-checking findings by the South African media has increased since 2012. Wekesa et al.’s (2017:23) research, which was based on coverage of fact-checking organisations’ work across nine online news platforms showed an increase in the publication of the number of articles related to fact-checking from 0 in 2012, 29 in 2013 to 50 in 2016 and 68 from January to September 2017. It should be noted, however, that the vast majority of the articles (235 out of 277, or almost 85%) published from 2013 to September 2017 were “wholly based” on the reports of Africa Check (219) or other fact-checking organisations (16), mostly the American news organisation Associated Press (Wekesa et al., 2017:24,25). Only seven content items (two by Daily Maverick, two by Mail & Guardian and three by News24) were found to have relied “wholly on the news organisation’s own verification methods”, while the rest were mostly items where Africa Check was referred to “in one way or another” (Wekesa et al., 2017:25). This study will further examine how the media uses or draws from the work of independent fact-checking organisations. 2.6. Conclusion In this section, literature pertaining to dedicated fact-checking was discussed. I covered the challenges experienced by newsrooms in the digital environment and its impact on journalism practice, the objectives of fact-checkers and how they assess their impact, fact-checking 20 coverage by the South African media, the relationship between fact-checkers and the media and the growth of dedicated fact-checking as a new form of journalism. In section 3, the theories and concepts that form the basis of the research will be examined. 21 Theoretical framework 3.1. Introduction Different to internal pre-publication fact-checking, which has been around for almost a century (Silverman, 2007:275), dedicated fact-checking is such a new phenomenon that media scholars are still grappling with its theoretical and conceptual interpretation and, according to Bannikov & Sokolova (2018), there is not a commonly agreed academic definition that “characterise[s] the role of a fact-checker within the journalism system”. The various ways in which the practice of fact-checking is described by academics, as examined earlier, confirms this. I use a combination of concepts and theories related to the field of journalism practice and professionalism, as well as the intermedia- and network agenda-setting theories as framework for the research – first, to gain a better understanding of this new type of journalism and how it came about and, second, to contribute to research about factors that influence changes in journalistic practice. The following subsections cover four aspects outlining the theoretical framework of this study: changes in media theory within a changing media environment, the dynamics of theories and concepts related to journalism practice and professionalism, the evolving agenda-setting theory, and drawing on these theories and concepts as a means of answering the research questions. 3.2. Changes in theory in a digital media environment In a changing media environment (Kammerl & Kramer, 2016; Daniller, Allen, Tallevi & Mutz, 2017), the objectives of dedicated fact-checking seem to add a new nuance to journalism and the growth of the practice provides an opportunity for new conceptual approaches. This suggests that media theories have to be enhanced if we are to understand the impact of fact- checking on journalism. Journalism has an evolving nature due to “conflicts emerging from the social context in which it operates” (Breit, 2004:81), which is evident when looking at the growth of dedicated fact-checking in response to challenges given rise to by the internet as discussed earlier. As early as 2001, South African academic Pieter Fourie (2001:252,253) suggested that the concept “network society” be used to describe the new media environment brought about by the growth in information and communication technology and the subsequent worldwide 22 “interconnection between people, groups, nations, and so on”. Five years before the advent of social media platforms such as Facebook (Barr, 2018) and Twitter (“We look back …”, 2016) in 2006, and 11 years before the practice of dedicated independent fact-checking started in South Africa in 2012, Fourie (2001:260) warned that the impact of “the Internet on public communications is increasingly also becoming an area of investigation and concern”. One could argue that the emergence and growth of social media platforms over the past 15 years could only have exacerbated Fourie’s concerns, and for the purposes of the study it would be valuable to investigate the link between not only online news, but also information shared on social media, and the growth of dedicated fact-checking. This is evident when considering the growth in the amount of fact-checking that is done on posts shared on platforms such as Facebook and WhatsApp (Facebook, 2021; What’s Crap on WhatsApp?, 2021). More recently, the term “news ecosystem” has been used in journalism studies to focus on the players that influence the news narrative and the ways in which news circulates via different technologies (Wiard, 2019:1). Wiard (2019:2) argues that “putting journalism processes in boxes prevents us from fully grasping the diversity of actors and practices that make up the news” and social networking sites are driving researchers “to address how these ecosystems influence news production”. Add to this that the use of the word “network” as well as keywords related to “ecosystems” and “landscapes” has increased significantly in journalism studies in recent years (Ahva & Steensen, 2019:48) and it becomes clear that one cannot look at media theories with only print and broadcast, and even online, news media in mind – one has to consider as many factors as possible within the news and information ecosystem. As pointed out in section 2.3, fact-checkers’ objectives relate to a broad spectrum of journalism theory, from the normative theories of the early 1900s to the latest theory applicable to the digital media space. Although some older media theories are still relevant in the media today, as indicated by Van Wyk (2017:91,92), who found that Africa Check fulfilled four functions of the social responsibility theory, new media technologies are changing journalism practice (Obalanlege, 2015:1) and digitisation forces us to reassess the theories with which we try to understand journalism (Steensen & Ahva, 2015:1). According to Ahva & Steensen (2019:39) there’s a “need to rebuild our fundamental understanding of what journalism is, owing to the many changes mostly related to digitalization that have affected the profession and its practices since the turn of the millennium”, which is also when the practice of dedicated fact-checking began, suggesting a link between digitisation and this new form of journalism. 23 3.3. The dynamics of theories and concepts related to journalism practice and professionalism Over the years, many media theories, such as the normative theories of the press, as well as the gatekeeping and agenda-setting theories, have contributed to the conceptual field of journalism practice and professionalism in different ways. The normative theories of the press looked at the ideal views of the role the media was considered to play in society and how these functions are affected by the freedom of, and restrictions on, the media (Fourie, 2001:269). Normative journalistic roles, including acting as a watchdog holding governments to account, often refer to what journalists ought to do to contribute to the “proper workings of democracy” (Hanitzsch, 2017:4). Touching on professionalism, is the observation of Graves et al. (2016:121) that fact-checking in the USA is gaining popularity among the media mainly because it “appeals to the professional values and status concerns” of journalists. Different to South Africa, dedicated fact-checking is not only practised by independent fact-checkers but also embraced by leading news organisations. Dedicating fact-checking services such as PolitiFact, FactCheck.org and the Washington Post’s “Fact Checker” column have become “established voices in national political discourse, cited frequently by news outlets” (Graves, 2013:4). PolitiFact, which started as an election project of the Tampa Bay Times (then the St. Petersburg Times) and won a Pulitzer Prize for fact-checking more than 750 political claims during the 2008 US presidential campaign (Adair, 2019), is now functioning as a not-for-profit news organisation (Drobnic Holan, 2019). The award was described as “the most unusual, and most important, Pulitzer Prize” of that year, indicating that online databases, such as that of PolitiFact, were “rapidly becoming one of the important tools of watchdog journalism in the digital age” (“PolitiFact is most …”, 2009). In order to assess the possible impact of dedicated fact-checking on newsrooms and news coverage, it’s important to consider the workflow and news production processes in newsrooms in the digital era as described in section 2.2. A study covering newsrooms in South Africa, Zimbabwe and Kenya showed a “remarkable shift in news production processes and cultures” associated with a reliance on social media for story ideas and dissemination of news on multiple platforms (Moyo, Mare & Matsilele, 2019:503). 24 Arguably one of the biggest challenges for journalism practice and professionalism, and a possible reason why the US journalists referred to above consider dedicated fact-checking a practice that represents professional journalistic values, is the loss of the media’s traditional roles as gatekeepers and agenda setters in determining what the news of the day is going to be (Brooks, 2018; Gene, 2017) due to the internet and social media platforms. Gatekeeping refers to how information goes through a series of checkpoints [journalists] before being published or broadcast as news content (Oosthuizen 2001:196). Shoemaker and Vos (2009:1) describe gatekeeping as “culling and crafting countless bits of information into the limited number of messages that reach people each day”. The new online news and social media environment, with its overabundance of information and subsequent impact on newsrooms and news coverage as discussed in section 2.2, calls for the new conception of journalistic professionalism. Ramaker, Van der Stoep & Deuze describe the professional practice of journalism as “increasingly precarious and complex” as journalists encounter multiple constraints accelerated by technological and economic pressures (2015:345). Job losses within the media sector, as discussed earlier, is one of the severe consequences of these pressures, leading to even more precariousness within the industry. The term “professionalism” implies characteristics such as “autonomy, exclusive knowledge, ethical codes of conduct, occupational ideals or culture, and altruistic features (i.e., to act in the public interest) and jointly, these values function as a framework for journalists in everyday practice” (Wiik 2019:1). Today journalism’s identity is existentially shaken (Hanitzsch, 2017:2) and journalists’ professional status is questioned as “exclusivity is broken, autonomy declines, and other actors are increasingly redefining the field. In this context, new methods and ideals arise” (Wiik 2019:1). Ramaker et al. (2015:345) call for so-called “reflective practices” – where journalists reflect on the way in which they do their work – that should fit within the “precarious professional environment of contemporary news production”. As I examine the perceived impact of independent fact-checking on journalists’ news coverage and work processes, and on accuracy in the media, I also look at whether newsroom practices are becoming more reflective or not. 3.4. Application of the evolving agenda-setting theory The agenda-setting theory is a good example of a media theory that has had to evolve with the rise of digital media and in its new forms can be applied to studies on the impact of fact- 25 checking journalism, which can also be considered a response to the challenges the online environment has brought along as discussed in section 2.5. Historically, the agenda-setting theory was about shaping public opinion through the topics presented to media audiences and the way in which they were presented (Fourie 2001:304). This played a role in how significant the public perceived certain issues to be (McQuail 2005:512) as media audiences tend to assess the importance of a topic based on the emphasis placed on it by the media (McCombs, 2011:1). However, while the mainstream media used to hold what could be considered a monopoly on setting the news agenda (Boynton & Richardson 2016:1918), journalists’ decisions about what to report on are now often guided by content shared on social media platforms (Kushin, 2010:147; Moyo et al., 2019:503). Another factor contributing to journalism not being as dominant in setting the agenda and leading public debate as it used to be, is that the media often cover what their audiences seem to prefer based on online audience analysis (Moyo et al., 2019:502-503). Today, the agenda-setting theory includes several facets, such as the network agenda-setting (NAS) model (McCombs, Shaw & Weaver, 2014:782) and what used to be known as the theory of news diffusion, which is now called intermedia agenda-setting (McCombs, Shaw & Weaver, 2006:61). These theories, which look at the impact of the “networked media agenda” on the “networked public agenda” (McCombs, Shaw & Weaver, 2014:782), could prove valuable in assessing the relationship between independent fact-checkers and the South African media, and whether fact-checkers are perceived to have an impact on news coverage and hence the “public agenda”. An analysis of the online media landscape in the USA from 2014 to 2016 suggested that fact- checkers didn’t play a role “in predicting the agenda of news media overall” (Vargo et al. 2017). Luengo & García‐Marín (2020:410) state that fact-checking usually doesn’t set the agenda as it is done in reaction to what others have said. However, when taking into account how the agenda-setting theory has expanded beyond the impact of the mass media on the public agenda (McCombs & Shaw, 2006:59) and that it is now also used to investigate the role of social science, communication and journalism subfields in shaping the news agenda (McCombs & Shaw, 2006:60), I argue that it would be valuable exploring whether fact-checking plays such a role in the South African context. 26 Intermedia agenda setting provides a framework for analysing “how the media discuss issues that have already made the agenda” (Ahva & Steensen, 2019:45), which is essentially what fact-checking journalism does as it investigates the accuracy of information that is already in the public domain (How we fact-check [Africa Check], 2021). According to Graves (2013:9), intermedia links and effects should be understood “as the question of how different news organizations, genres, and media technologies or platforms interrelate and interact at different moments in media history,” referring to the example of a network news producer who would go through the day’s newspapers in the morning to see which news events are covered (Graves 2013:10). In this study, I use the intermedia agenda-setting theory as a basis to investigate the intermedia links between the media and fact-checking organisations by assessing their level of interaction and how the work of journalists and that of fact-checkers interrelate. The NAS model adds to traditional agenda-setting theory by measuring not only how an issue is covered by a news organisation at a certain stage, but also how often certain issues are reported on simultaneously during certain times and what the relationships between these different news stories are (Vargo, C.J., Guo, L. & Amazeen, A., 2018:2030). Research done by Vargo et al. (2018:2030) found that the “network relationships among different news items and messages can be transferred between varied stakeholder agendas: from media to public, from different interest groups to media, as well as from media to media”. Graves et al. (2015:17) use the example of cable news networks that use research done by fact-checkers to produce video segments in which political claims are debunked as a way in which fact-checking spreads in “today’s news ecosystem”. Similarly, in this research the NAS model is used as a basis to establish whether the South African media also use research done by dedicated fact-checkers in their reporting. 3.5. Examining the perceived impact of fact-checking within the “news ecosystem” As mentioned in section 3.2., the growth of independent fact-checking within the evolving field of journalism provides an opportunity for developing new conceptual approaches. In order to answer the research questions, one has to consider several factors within the news ecosystem. I therefore aim to interpret the perceived impact of fact-checking within the so-called “network society” or “news ecosystem” by drawing on concepts and theories related to journalistic practice and professionalism, and the intermedia- and network agenda-setting theories. 27 The agenda-setting theory has evolved in line with journalism practice in a digital media environment to now include the concepts of intermedia- and network agenda setting, which are used to examine the impact of the “networked media agenda” on the “networked public agenda” (McCombs, Shaw & Weaver, 2014:782). These theories are used to not only assess the relationship between independent fact-checkers and the South African media but also how the work of journalists and that of fact-checkers interrelate and therefore in which ways the media use fact-checking work in their coverage. These results will then, once again, touch on changes in journalism practice and professionalism. 3.6. Conclusion In this section, I covered the evolving media theories that form the basis of the research. These include several theories and concepts pertaining to journalism practice and professionalism as well as the evolving agenda-setting theory within an expanding media network or ecosystem. In section 4, I will explain how I will incorporate the theoretical framework into my research methodology. 28 Methodology 4.1. Introduction This study seeks to establish the perceived impact of independent fact-checking on journalism practice in South Africa by finding answers to the following questions: RQ1: To what extent is independent fact-checking considered to have an impact on South African newsrooms and news coverage, if at all? RQ2: In what ways do the South African media make use of work done by independent fact- checking organisations and to what extent does independent fact-checking promote accuracy in the media, if at all? RQ3: How do journalists perceive the role of independent fact-checkers and their relationship with the media, and do they consider independent fact-checking a distinct form of journalism that is growing due to factors related to journalism practice in the digital media environment? In this section, I start moving toward the empirical dimension of the study by explaining how I will collect and subsequently analyse data. Taking into account that I am using concepts related to journalism practice and professionalism, including variants of the agenda-setting theory, as reference for answering the research questions and contributing to fact-checking research, I found a qualitative approach to be more suited than a quantitative one. The theoretical subfield of journalistic professionalism and practice is more inclined towards qualitative than quantitative methods, as illustrated in studies that share similarities with this research, such as the use of semi-structured interviews by Borges-Rey (2016) to look at data journalism practice in professional newsrooms and a similar approach followed by Kurambayev & Issenov (2020) to examine obstacles in journalism professionalism for aspiring journalists. In fact, journalism practice and qualitative research methods are closely linked in that both have a strong focus on observation and in-depth interviewing when gathering information, sceptical interpretation of this information, and providing perspective when explaining findings (Iorio, 2011:7). 29 In probing intermedia- and network agenda-setting related to the circulation of news items between independent fact-checkers and selected media, my objective was not to quantitatively demonstrate the level of uptake but to indicate the extent to which there is an uptake. Therefore, although some researchers who explore the intermedia- and network agenda-setting theories use a quantitative approach (Vargo & Guo, 2016; Nygaard, 2020), I chose a qualitative approach because of the incorporation of journalistic professionalism and practice as mentioned above. Qualitative methods have also successfully been used in exploring evolving agenda-setting theories (Rogstad, 2016; Geiß, 2019), while several studies in the field of fact-checking approach the topic from a qualitative rather than quantitative approach, too (Cheruiyot & Ferrer- Conill, 2018; Van Wyk, 2017). While quantitative research is defined as being “primarily based on measurement rather than on description”, qualitative research is mostly based on description instead of measurement (Bertrand & Hughes, 2005:260). It is important to note that one of the limitations of the study is that the impact of independent fact-checking on journalism practice is not measurable. A descriptive interpretation of the information I gathered through in-depth qualitative interviews is thus used to assess the perceived impact thereof. Furthermore, qualitative research’s best value lies in “achieving in-depth understanding of social reality in a specific context” (Lindlof and Taylor, 2011:109), which is vital in gaining insights into the day-to-day news coverage of journalists, the conditions they work under and whether independent fact-checking plays a role in the way they work – all of which tie in with the exploration of factors related to journalism practice and professionalism. 4.2. Data-collection method: semi-structured, in-depth interviews Qualitative research methods for collecting data include observations, interviews, written reports, elicitation techniques and mapping, which can be further divided into subcategories (Given, 2008:2). Observation is a way of gathering data by observing and spending time with people within their environments and produce written accounts of their activities (Wästerfors, 2018), elicitation is used to establish how groups of people organise the ways in which they understand the world, and mapping is a continuous activity for researchers who investigate changes and continuities within specific settings over time (Given, 2008). Since research of journalistic professionalism in the wake of the rise of fact-checking journalism is a work in progress, and little is known about the extent of its actual or possible impact on journalism 30 practice within the newsroom environment, I deemed neither observation nor elicitation or mapping to be suitable methods for research on the topic. In order to explore the perceived impact of fact-checking in detail and discover patterns of similarity and difference based on journalists’ individual experience, beliefs and behaviour (Given, 2008:5), I opted for in-depth interviews with editors from a variety of platforms and media organisations across the South African media landscape. Semi-structured interviews, which lie between strictly structured interviews with a set of questions requiring specific answers and an open-ended interview that takes the form of an in-depth discussion (Blaxter, Hughes & Tight, 2010:193), were used. A researcher who conducts semi-structured interviews has more freedom than one opting for a structured one as the interviewer does not have to strictly follow a detailed interview guide (Kajornboon, 2004:6). For example, the guide used for semi-structured interviews may consist of “carefully worded questions or a list of topics to be covered” based on the research question/s, and interviewers can decide if they want to follow the questions “to the letter” and in a specific order, or move between topics based on the interviewee’s responses (Ayres, 2012:811). Corbetta (2003:271) states that, “Within each topic, the interviewer is free to conduct the conversation as he thinks fit, to ask the questions he deems appropriate in the words he considers best, to give explanations and ask for clarification if the answer is not clear, to prompt the respondent to elucidate further if necessary, and to establish his own style of conversation.” The interviewer decides how detailed the interview guide should be – whether a checklist of topics to cover, or a list of questions supplying the interviewer with guidelines – giving both the researcher and the interview ample freedom in terms of the discussion, while ensuring all the necessary themes are covered and information collected (Corbetta, 2003:271). The interviews used in this study comprised broad, in-depth questions (Addendum B) encouraging detailed answers based on topics related to the research questions in order to gain in-depth insights into current journalism practice in newsrooms, the interconnectedness of different news platforms and how that ties in with the evolving agenda-setting theory, and where and how independent fact-checking comes into play in this environment. 31 The wording of the questions was adjusted where necessary based on the flow of the interview as well as the type of media platform interviewees work on as online platforms often publish fact-checking reports as is (Wekesa et al., 2017), print publications might publish an entire report or make reference to findings by a fact-checking organisation, and radio and TV broadcasters would interview independent fact-checkers (Mwiti, 2020). Where the interviewer deemed it necessary, follow-up questions were asked as they are valuable to clarify an answer, get an explanation for or examples of a response, or to double- check whether an answer is correctly understood (Du Plooy, 2002:144). Follow-up questions can also “elicit underlying ideas, feelings, sentiments and suggestions” the researcher might not have been aware of (Du Plooy, 2002:143). With the exception of the first interview, all of them were conducted online due to Covid-19- related lockdown regulations and precautions, as even the least strict level 1 regulations stated that one should travel to “perform and acquire services only where such services cannot be provided from the safety of one’s home” (Department of Health, South Africa, 2020). Since interviewees are in an environment they know and are familiar with during an online interview, they might feel more at ease sharing their opinions, while other advantages include that there’s a reduction in time as well as costs related to the process (Bolderston, 2012:72-73). Disadvantages include possible technical issues such as unclear sound and a lack of visual quality (Bolderston, 2012:73), which I addressed during the interviews by asking interviewees to repeat answers that weren’t clear. All interviews were recorded and transcribed6 before the thematic analysis was conducted. 4.3. Sample selection and size In the case of random sampling, all members of a population have an equal chance of being selected, but when doing qualitative research one’s goal is not often to interpret findings based on information from an entire population (Lindlof & Taylor, 2011:109), which is why one should sample deliberately and not at random (Moser & Korstjens, 2016). Although qualitative researchers might be interested in how often something happens, most of them choose to do “purposeful” or “purposive sampling”, making “informed judgements about what to observe or whom to interview” (Lindlof & Taylor, 2011:109,110). Although one of the disadvantages of 6 Transcripts available on request. https://www.tandfonline.com/doi/full/10.1080/13814788.2017.1375091 32 purposive sampling is an “inability to generalize research findings”, it is effective when “only limited numbers of people can serve as primary data sources due to the nature of research design” (Dudovskiy, 2019) as in the case of this research. After doing a study about the uptake of fact-checking reporting in the newsrooms of US newspapers, similar to the one by Wekesa et al. (2017) done based on South African news websites, Graves et al. (2015:16,17) suggested further research across the media landscape – including print, online and broadcast outlets – to establish the effects of independent fact- checking on news coverage. While I included interviews with journalists across all three of these types of media platforms, I also bore in mind Bertrand and Hughes’ suggestion to select “a population relevant to the question being asked, small enough to be practically manageable and large enough to provide the sort of data you need” (2005:64). I initially proposed interviews with a variety of about 15 journalists, including reporters, subeditors, news editors, editors and producers from various platforms. As I started contacting media organisations and journalists with interview requests, however, I found that in the majority of cases, editors were the most suitable interviewees for purposes of the research as they were more likely able to provide a broader overview of their editorial teams’ news-gathering practices as well as the content on their platforms. Based on this finding, as well as my knowledge in the fact-checking field, the final sample was a purposive selection of a variety of news, online, managing and associate editors. The limited time and willingness of some editors to take part in interviews due to their added workload in covering Covid-19 and adapting their work processes to keep staff safe, had an impact on the size and composition of the final research sample as well as the time period over which the interviews took place, as editors weren’t as readily available for interviews as initially expected. I realise the size of the research population makes it difficult to generalise the findings across newsrooms in South Africa. The study, however, doesn’t look at the impact of independent fact-checking on journalism practice in a conclusive manner by aiming to measure it, but in a manner that provides qualitative interpretive insights into the perceived impact thereof. I am, however, confident that the final sample of 13 editors was as representative of the country’s media landscape (see Figure 3) as possible and large enough for data-collection purposes, while still practically manageable for the purposes of the research in line with Bertrand and Hughes’ recommendations (2005:64). When I became aware of the challenges of 33 getting interviews completely in line with my initial planned sample in my research proposal, I moved my main focus to interviewing at least one editor from a wide variety of well-known media organisations in the country. This approach worked well since it’s become common over the past few years for media houses to have the same staff creating content for several titles and platforms within the same company in what is often referred to as content or subbing “hubs” (Otter & Grant 2017:30; Daniels 2018:19). This meant that one interviewee often had insights into the content and work processes of more than one publication, news website, or TV or radio station. However, in cases where it became clear that the digital departments of certain organisations might use the work done by independent fact-checking organisations differently to their traditional platforms, online editors were contacted for clarification to ensure that the results reflect the reality within newsrooms. The findings can thus serve as a solid basis for future research in the field of fact- checking journalism. Figure 3: A helicopter view of the South African media landscape Source: Rumney (2020:5) The approach followed ensured a varied sample, especially in terms of media ownership, while the different media formats (print, online, radio and TV) were all still represented. The interviews 34 led to valuable insights and a comprehensive overview of journalism practice and whether independent fact-checking is considered to have an impact within different media environments and organisations – all with fairly large audiences. The organisations represented by the interviewees include privately, state-owned and funded media; a variety of talk and music radio stations; free-to-air TV news as well as 24-hour subscription TV news channels; a daily and weekly newspaper; and several of the most popular online news platforms in terms of unique South African browsers. When taking into account the South African media landscape as depicted in Figure 3, the research sample (Addendum C) not only represented editors from the public and commercial sectors, but also from a variety of different platforms within these sectors. In the public sector, interviewees included editors at both SABC Radio News and SABC TV News. In the commercial space, the so-called “Big 4” traditionally print, but now also online, media companies are all represented: Media24 (News24), Arena Holdings (Sunday Times, Sunday Times Daily, TimesLIVE, Sowetan and SowetanLIVE), Caxton (The Citizen) and Independent Media (a weekend title). In terms of commercial TV stations, editors from free-to-air eTV News as well as the paid-for 24-hour channel, Newzroom Afrika, that is broadcast on DStv, were interviewed, while in terms of radio an editor from Eyewitness News (EWN), which covers news for two talk and two music stations owned by Primedia, was interviewed. All interviewees have between 10 and 32 years of experience working in the media industry. I specifically included the independent investigative newspaper Mail & Guardian and its online platform (mg.co.za) as well as investigative news platform Daily Maverick, which “is gaining much social prestige and respect among the news-consuming elite” (Ndlovu, 2015:127), since Wekesa et al.’s study (2017:23,24) indicated that these two online platforms published more content related to fact-checking from 2012 to 2017 than seven other online platforms covered in their research. For purposes of the research, I distinguish between dedicated fact-checking reporting as referring to the assessment of the accuracy of statements in the public domain (Graves, 2013:2; Africa Check, 2021) and internal fact-checking quality-control measures within newsrooms such as sub-editing and the verification of claims before publication (Graves & Amazeen, 2019:1). This was explained to each interviewee to ensure clarity of what is being referred to in the questions. 35 4.4. Data-analysing method: thematic analysis With thematic analysis, researchers look for “recognizable recurring topics, ideas, or patterns (themes) occurring within the data”, a method often chosen when “investigating a phenomenon for which little prior understanding exists” (Hawkins, 2018:2), which made it ideal for researching the perceived impact of independent fact-checking on journalism practice in South Africa, a topic very little research has been done on as far as I could establish. Since thematic analysis is known for its flexibility, it’s ideal to uncover “issues, problems, similarities and differences” in the field of communication (Hawkins, 2018:2), but I also kept in mind that in qualitative content analysis, such as thematic analysis, text is “open to subjective interpretation, reflects multiple meanings, and is context dependent” (Julien, 2012:122). Keeping the literature review and theoretical framework in mind, I followed a thematic analysis approach to analyse the information obtained during the in-depth interviews. With the ultimate aim of answering the research questions, I first organised the interview transcripts by question and then switched to a thematic strategy where I used colour codes to mark and organise themes based on recurring statements that started emerging (Wallace Knowledge Center, Workbook E). 4.5. Declaration from author Being head of education and training at Africa Check, I was aware that my role at the fact- checking organisation could introduce the potential of bias (see Addendum A). I believe that the value of this research, however, by far surpassed the possible harm of bias. My insider knowledge of the fact-checking industry proved invaluable as I could easily gain access to information related to Africa Check and other fact-checking organisations, which added depth to the research. Prior to each interview, I informed the interviewee that I work for Africa Check but that the research was conducted in my personal capacity as a master’s degree student and researcher with an interest in fact-checking. I encouraged all interviewees to be frank about Africa Check and its work throughout the interview process. 4.6. Conclusion Section 4 covered and motivated the research methodology used in the study: the data- collection and analysing methods, the sample size and selection process. Data was largely collected through semi-structured, in-depth interviews with a variety of editors across the South 36 African media landscape. This was followed by a thematic content analysis of the interviews. In section 5, the research findings will be presented in line with the research questions and recurring themes that were identified. 37 Presentation and analysis of findings 5.1 Introduction In this section the information gathered by means of the in-depth qualitative interviews is presented and analysed in four categories, namely: the impact of the growth in digital media on journalism practice and professionalism; the impact of fact-checking on newsrooms and news coverage; the use of independently fact-checked content by journalists and the role it plays in promoting accuracy in the media; and journalists’ perceptions of dedicated fact-checking and independent fact-checking organisations. In each of the categories, information is presented and analysed in line with both the interview questions and themes that emerged during the course of the interviews. 5.2. The impact of the growth in digital media on journalism practice and professionalism 5.2.1. Moving from the traditional to the online media space: the impact on South African newsrooms and news coverage All interviewees indicated that the growth in digital media has had a substantial impact on newsrooms and news coverage, with more than half attributing changes in news coverage to social media specifically. It emerged that most journalists are now either covering news for both traditional and online platforms, or print and broadcast journalists work closely together with digital teams. To summarise the interviewees’ overall description of newsrooms at this point in time: fewer journalists, who are often junior, have to do more work at a faster pace than before, leading to a lack of depth in news coverage. The interviewees attributed this situation in newsrooms to several factors. First, the media have to keep up with a constant flow of news not only on news websites but also on social media platforms, accelerating the pace that journalists have to work at. There are no more deadlines, literally. It's a deadline, like, every minute, you've got to make sure you're on top of things, you're getting things out. (Interviewee 2, News24) Social media has given the people their power back … Members of the public are actually our chief editors, because when you wake up in the morning they have already 38 set the agenda for you in terms of what they think you should be discussing on your platforms or what they think they want to hear today or see on TV. (Interviewee 13, SABC TV News) Second, according to the interviewees, a drop in circulation and advertising income has led to the retrenchment of mostly experienced staff members, and thus smaller and so-called “juniorised” newsrooms. Third, due to the time pressures that these small newsrooms are under, there’s little opportunity for journalists to spesialise or build a network of contacts in a specific field. The basics of journalism is reliance on trusted sources of information. Nobody's going to trust you if they don't know you; if they don't know what you have done over time ... and that actually is one of the greatest, you know, tragedies that we're looking at … We expect very, very junior people to do some of the most complex of stories. (Interviewee 8, Sowetan and SowetanLIVE) These findings confirm the changes in news production processes and the challenges newsrooms are facing in the digital media environment discussed in the literature review. 5.2.2. The link between changes in journalism practice due to the digital environment and the growth of independent fact-checking The vast majority of interviewees thought there is a link between journalism practice in the fast- paced, high-pressure digital media environment they experience in their daily work and the growth of independent fact-checking, but even more so between the rise of social media and the need for fact-checking. Speaking to the normative journalistic goals of seeking truth and informing the public, discussed in section 2.3, they believe fact-checkers have more time than journalists for fact-checking and verifying information in order to seek the truth, something that is especially challenging due to the plethora of content that’s available online. I think there is a lot of nonsense that's around that needs to be debunked, and your online and traditional newsrooms don't have the capacity to do so … and independent fact-checking organisations play a vital role in doing that. (Interviewee 4, Daily Maverick) 39 Sadly, sometimes that journalistic ideal of truth, the North Star if you will, is sort of obscured by this firehose of information and we're constantly playing catch-up, trying to be on top of the latest information. The consequence there is that we [journalists] are not always able to properly fact-check stuff. (Interviewee 9, Newzroom Afrika) Several interviewees touched on the role fact-checkers play in providing accurate information and improving the quality of journalism. … If you think about lack of resources in the newsroom, lack of specialisation, lack of depth of skills and, you know, knowledge and loss of institutional memory, it would have been very difficult without some [fact-checking] organisations being able to play in that space and actually close the gap. (Interviewee 8, Sowetan and SowetanLIVE) I think that sort of horrible time of journalism [when traditional media just started providing news on online platforms] where we [the media in general] were just chasing what audiences think is relevant and not doing as much quality journalism alarmed enough people that you ended up with independent fact-checking organisations and non-profit journalism centres and all of those things, which are saying we actually do need quality journalism that is fact-checked. (Interviewee 6, Mail & Guardian) Almost half of the interviewees also attributed the growth of independent fact-checking to the spread of false information via social media and other online sources, while at least three believe that the amount of false information spreading on social media plays a far bigger role in creating a need for fact-checking than changes in the media industry and information on news websites. I think the 24-hour broadcast cycle – Monday to Sunday – the websites, the social media have definitely led to the establishment and the creation and the rise of independent fact-checkers. But probably more due to social media where everyone can tweet what they like, say what they like, you've got the bots, you've got anonymous people ... (Interviewee 7, SABC Radio News) 40 5.3. The impact of fact-checking on newsrooms and news coverage 5.3.1. The impact of fact-checking on the news agenda and information ecosystem With independent fact-checking practised for the first time in the country less than a decade ago and all the challenges in newsrooms brought about by digitalisation, it’s not surprising that interviewees’ thoughts on whether the work of fact-checking organisations have an impact on the news agenda in South Africa in any way varied substantially. Only four of the 13 editors thought independent fact-checking has an impact on the news agenda. Three of the four interviewees who did think so, indicated that this impact is fairly limited. I think there is an impact, because when fact-checking has been done often it forms part of news bulletins. EWN, I think, does them quite often. Whenever Africa Check does debunk something, there's normally a news piece. It's more the fast-news organisations that pick up on those trends, whereas for us it influences the depth of the story rather than a story in itself. (Interviewee 4, Daily Maverick) I do think that that [independent fact-checking] has an impact on how things are put across eventually … sometimes, like I say, something’s been fact-checked and it's changed the whole … way that the story was initially put out. (Interviewee 2, News24) The majority of editors were unsure about whether fact-checking has an impact on the news agenda, while three interviewees believe it doesn’t have an impact on the news agenda in South Africa. These answers could be considered a reflection of findings in an analysis of the online media landscape in the USA, which indicated that fact-checking typically doesn’t set the agenda as it is done in reaction to statements that have already been made (Luengo & García‐ Marín (2020:410). However, some interviewees touched on the impact fact-checking could have by changing the coverage of an item that is already on the news agenda or the direction in which a story is heading. This corresponds with the intermedia agenda-setting theory that looks at “how the media discuss issues that have already made the agenda” (Ahva & Steensen, 2019:45). 41 If the news was about XYZ said ABC and suddenly a fact-checking organisation can come along and say, ‘Well, hold on, actually, this is not true’, then that affects the news agenda and that gives journalists a different avenue to follow that they may not necessarily have had time to research themselves. So certainly, I think it can have a huge effect … (Interviewee 9, Newzroom Afrika) Two editors referred to the role that fact-checkers play in the information ecosystem, in adding to the conversation and keeping the narrative within the so-called “information ecosystem” accurate. This confirms my earlier argument that one cannot look at media theories with only print and broadcast, and even online, news media in mind – one has to consider as many factors as possible within the ecosystem of information. The kind of stuff they [fact-checkers] check, you can say, ‘Okay, just take a step backwards ... just because a thousand people say it's so doesn't mean it's so, here's something credible that says it isn't, so actually put that into the mix’. I'm not sure if it's the news agenda so much but it's definitely a really important part of giving citizens information that's useful. (Interviewee 11, Sunday Times, Sunday Times Daily and TimesLIVE) Throughout the interviews several references were made to the bigger impact that interviewees thought fact-checking has in the USA when compared to South Africa – mostly based on their observations while following and covering that country’s presidential election in 2020. “I think it's in huge public interest ... that you have somebody who can actually point to inaccuracies and falsehoods that are being peddled. I've seen it in the US and I think it works fantastic. The likes of New York Times do that. I mean, Trump