August 2019 Regional report Media Programme Sub-Sahara Africa

Fact vs. Fake

The impact of fact-checking and its measurement by African fact-checking organisations Franziska Krämer Shenilla Mohamed, Executive Director of Amnes- Foreigners are overburdening South Africa's pub- ty International South Africa, warns that Motsoal- lic hospitals. This is what Aaron Motsoaledi, edi fuels xenophobia with such statements. Dep- South Africa’s then Health Minister, claimed in uty Public Protector Kevin Malunga tweeted that November 2018 when he spoke at the Nurses the Minister of Health's statements are aimed at Summit in Johannesburg. There, he discussed the deflecting from his own challenges regarding the challenges hospital staff face in their everyday health system. His tweet triggered a discussion in working life. In an interview with radio station the comments on whether or not foreign nation- SAfm, Motsoaledi blamed the country's immigra- als are responsible for overburdening the health tion policy for overburdening the healthcare system. system. 700 babies were born in May 2017 at Charlotte Maxeke Academic Hospital in Johan- The fact-checking organisation Africa Check con- nesburg, "and 400 of them were not South Afri- tacted the hospitals and checked the figures the can at all,” Motsoaledi said, for instance. Minister of Health quoted in the interview.

Konrad-Adenauer-Stiftung e. V. Regional report July 2019 2

The result: none of the figures quoted by Mot- ing that the worldwide growth of the fact- soaledi could be substantiated. Wrong infor- checking industry is also being experienced in mation circulating in public can cause great harm, Africa, the Duke Reporter Lab currently mentions depending on the topic. nine fact-checking organisations on the conti- The issue of migration, for example, is particular- nent, five more than at the beginning of 2018. ly sensitive, and a lot of misinformation has been They include Africa Check, Pesa Check in Kenya, circulating in South Africa. "Misinformation about Tanzania and Uganda, BuhariMeter and Dubawa the actual figures of migrants living in the coun- in Nigeria, Congo Check in the Congo, ZimFact in try, blaming of migrants for most socio-economic Zimbabwe and the recently founded Namibia ills including crime, stealing jobs and business FactCheck. This year also saw the continent of opportunities, drugs and burdening of public Africa host the global fact-checking conference resources etc. has been presented to the public for the first time. Sixty African fact-checkers – 49 by politicians and the media. This does not only more than the past year – attended the Global result in grossly overstated figures, but fuels the Facts 6 Summit in Cape Town in June 2019 giving negative rhetoric presented on migrants in the Africa the second highest continental representa- public domain," explains Silindile Mlilo, research- tion after Europe. er at the African Centre of Migration and Society in Johannesburg. Much energy, effort and money from public insti- tutions and foundations is now being invested in Motsoaledi’s example shows how quickly false fact-checking. Society is well aware of the prob- information can reach the public – not only lem of fake news and its dissemination, especially spread by politicians but also by journalists. through social media. Fact-checking organisa- Trust in the traditional media has been weak- tions worldwide offer a possible solution for tack- ened: In the past, economic imbalances of media ling this problem. The question remains: How can companies and the resulting cost-cutting fact-checking organisations measure if their work measures led to poor research by journalists. is effective? Unreflected, one-sided reports or even fictitious stories aggravated the situation. Fact-checking Congo Check: The Size of organisations, however, try to pick up from here the Audience as Kate Wilkinson, Africa Check’s deputy editor-in- chief explains: "We try to determine whether Tracking one's own effectiveness means first of what people in the public domain say is true or all one thing: more effort. In addition to their not. We look at influential people, whether it is actual task – the checking of facts – the organisa- politicians, union leaders, business leaders, tions have to find ways to determine their own sometimes journalists and media.” impact. A simple possibility, which in particular the younger fact-checking organisations use, is to Africa Check’s goal is it to provide true infor- see how many people the organisations reach mation to the public. “If a government is trying to with the fact-checks: it is relatively easy for them make the best decision regarding education or to measure how many people visit their own health or infrastructure development and they website or react to contributions on Facebook or don't have the best available information or they Twitter without too much effort. are making a decision based on wrong infor- mation, they might end up making a decision that In 2018, Sammy Mupfuni co-founded Congo has negative consequences.” Check, the first independent fact-checking organ- isation in the Democratic Republic of Congo. Africa Check was founded in 2012 at the Universi- Today, 16 journalists are working to check infor- ty of Witwatersrand in Johannesburg as the first mation in Congo. And social media is indispensa- independent fact-checking organisation on the ble for the younger fact-checking organisations African continent. The non-profit organisation on the African continent, especially since fake has expanded rapidly and now has offices in news and fact-checks are often spread via Face- Kenya, Nigeria, South Africa and Senegal. Indicat- book, Twitter and WhatsApp. But social media is

Konrad-Adenauer-Stiftung e. V. Regional report July 2019 3

also important in order to receive reactions from readers and thus to understand how effective the Instead of only including the number of people work is. reached by the fact-checks, more aspects can be taken into account when measuring the impact. "We measure the impact of our online work by Rather, for the Dubawa’, the effect is also defined the size of the audience on our social networks by the reaction of other media and politicians, mainly and then from the number of visits to our but also journalists and editorial offices, who are website. We are currently relying much more on supported by Dubawa in fact-checking, says the number of our subscribers on our social me- Oputa. "For us, we know we are contributing to dia channels, namely Facebook and Twitter, and the society through audience reaction, that is, also on the level of commitment that our publica- when people delete or edit false information tions generate. This is for example the number of online based on a fact-check we did; when our comments and sharing," says Mupfuni. Congo fact-checks are cited or reposted by mainstream Check also offers seminars for young journalists, media or even government officials; and when bloggers and Internet users in which participants the journalists or newsrooms we train continue learn to fact-check information online. The team to produce fact-checks without our direct sup- also considers the number of participants in its port." evaluation. Africa Check: Setting goals for His conclusion: Congo Check's fact-checks have successful fact-checking an impact as the team receives messages and supportive comments for the work the fact- For newly founded and smaller fact-checking checkers do. However, Mupfuni is not yet satis- organizations, measuring impact is a problem fied: "Indeed, our work has an impact in the since they lack resources such as money and community but it is still too low compared to time. As the first independent and largest fact- what we want. Among the biggest challenges is checking organisation on the African continent, the fact that we are not yet reaching a wide target Africa Check has existed for seven years now – on social networks and this affects the number of but even for them the question of how its work’s visits to our website. And this is mainly due to the impact can be measured systematically has not fact that we don't have the money to sponsor our yet been clarified. Facebook page continuously." In addition to that, better tools need to be implemented to measure Nicola Theunissen is Impact Manager at Africa the impact of Congo Check, as Mupfuni points Check since November 2017. Her work consists out. of answering exactly this question. "I have worked as a communication consultant in the Dubawa: Fact-checks must reach monitoring and evaluation field for five years and politicians and media attended conferences in many countries but I have not attended a single session that sets out The younger fact-checking organisations in par- how to rigorously measure the impact of the ticular are only gradually beginning to measure media. And even at journalism and fact-checking their impact more systematically. Ebele Oputa, conferences, speakers agree that it's not an exact works as deputy editor-in-chief for the Nigerian science," says Theunissen. fact-checking organisation Dubawa. As a small team with few resources, it was more important Africa Check’s team considers fact-checking as to first check the facts and publish the fact- effective when it makes a noticeable difference. checks. In the future, Dubawa’s team would also "We consider our long-term impact if we have like to focus on monitoring and evaluation – improved the quality of debate and if we have which, however, involves a lot of work. "To meas- improved the decision-making capacity of policy- ure the impact on the side of the public, we makers and the public. And that will eventually would have to undergo surveys and intensive help to strengthen democracy and life out- social media monitoring," says Oputa.

Konrad-Adenauer-Stiftung e. V. Regional report July 2019 4

comes," says Theunissen. However, these aspects ing the significance of Africa Check. At the begin- are not easy to measure. ning of 2019, for example, a major South African party asked the Africa Check team to review its Nicola Theunissen and her colleague Unathi Beku manifesto before it was published. "Politicians measure the influence of fact-checking on differ- are aware that South African voters are now ent levels: on a micro-level, i.e. on the level of the aware of misinformation and interested in an individual reader, where Theunissen and Beku accurate debate," says Unathi Beku, Africa take the amount and kinds of reactions to stories Check’s Impact Researcher. The mere fact that on social media into consideration. At a meso Africa Check exists ensures that politicians worry level, it is about the influence on a larger com- about whether the information they provide is munity or organisation. The Nigeria Centre for correct which shows the impact of fact-checking. Disease Control (NCDC), for example, referred to the work of Africa Check in 2018: "Chimezie The problem of the target group Anueyiagu, Communications Director of the Nige- ria Centre for Disease Control, said a health Neither small nor large fact-checking organisa- workshop Africa Check held in Abuja, fills a crucial tions are currently able to answer one aspect gap in the health communication space and satisfactorily: How do the fact-checkers reach builds the NCDC's capacity to achieve its mission people who do not want to follow the organisa- of protecting Nigerians and ensuring the integrity tions on social media or are not even able to do of health information," Theunissen said. so? How do the fact-checkers reach people who do not actively inform themselves about their At the macro level, the impact of Africa Check's work – or even share false information? work can be considered as a social change on an institutional or political level. As an example, In Nigeria, many people have no access at all to Theunissen mentions a speech by South Africa's the fact-checks that are shared online. This is a current Minister of Transport, Fikile Mbalula, in big challenge for Dubawa’s team. "We have been which he referred positively to Africa Check's working on partnering with media organizations research on the ANC's manifesto before the 2019 that have access to offline platforms like radio or South African elections. "We considered this as print media," says Ebele Oputa. However, such quite impactful because it's an indicator that we cooperations are difficult to establish and main- improved the quality of public debate when a tain in Nigeria. politician references us in public.” Africa Check tries to reach more people via com- Various tools, such as programs to record results, munity radio and by translating the fact-checks help Theunissen and Beku to be more systematic into different local languages. It is not known in their analysis – but that is it. "There are a whether the fact-checks reach people who share bunch of tools, but I don't think there are a lot of false information, Nicola Theunissen says. "You them that demonstrated an actual worth for the actually don't know whether your fact-checks are fact-checking organisations," Theunissen says. reaching the right persons. What we know is, and According to her, the African fact-checking com- this is one of our challenges, that we are reaching munity should work together on a universal tool educated people. The majority of our audience to understand their impact: for readers, donors, has a post-graduate level, and this is an interna- and not least also for its own editorial staff and tional trend by the way," says Theunissen. These the fact-checking community in order to perceive are not necessarily the people who disseminate its own successes. or believe false information.

However, the most important questions remain: Social media offers an opportunity to tackle this Is democracy strengthened by fact-checking? problem: some fact-checking organisations like Does the fact-checking of Africa Check have any Dubawa or Africa Check are now cooperating impact? Theunissen agrees, but also speaks of with Facebook. If false information is identified as "anecdotal evidence", i.e. individual stories show- such by the fact-checkers, Facebook users are

Konrad-Adenauer-Stiftung e. V. Regional report July 2019 5

less likely to see it through a Facebook algorithm. standing fact-checking organisations like Africa Users who have shared false information will also Check. There is a lack of relevant methods and receive a notification. "In that way this is an excel- instruments for this purpose. For younger organ- lent way for me to reach the people who spread isations such as Congo Check, which are still in the misinformation," Theunissen says. the process of being set up, it is even more diffi- cult because they need money, time and skills to At the same time, Facebook collects data on the do so. The ambition of the fact-checkers to extent to which this cooperation and fact- measure their impact as accurately as possible checking generally works by measuring how of- has, however, been awakened. At conferences ten users actually delete information that they like Global Fact 6, they discuss new possibilities have shared, and which were identified as false and tools and exchange their ideas. by fact-checkers for instance. However, the com- pany does not pass this data on to the fact- Their motivations are noble – to identify and checking organisations although it is precisely correct misinformation. And although it might this kind of information that the fact-checking still be a challenge to measure their positive in- organisations would need in order to measure fluence in African society, their work may per- the influence of the fact-checks much more sys- haps hold the key to re-establishing credibility in tematically and thus close this gap. media.

In the current era of fake news and deliberate misinformation fact-checking has emerged as a new discipline in journalism. Still, the impact of fact-checking is currently not possible to under- stand systematically, even for large, long-

Konrad-Adenauer-Stiftung e. V.

Franziska Krämer Research Assistant Media Programme Sub-Sahara Africa https://www.kas.de/web/medien-afrika

[email protected]

The text of this publication is published under a Creative Commons license: “Creative Commons Attribution- Share Alike 4.0 international” (CC BY-SA 4.0), https://creativecommons.org/licenses/by-sa/4.0/legalcode www.kas.de