Social Media and Mis/disinformation in Electoral Context in the DRC: Research Report

"Social media platforms are used to intentionally spread disinformation, exploit social divisions, and reach a wide audience."

This report, published by Internews, shares the results of research that sought to understand the disinformation ecosystem during the 2023 electoral campaign period in the Democratic Republic of Congo (DRC) - in particular, the dynamics of consuming and sharing false information on WhatsApp groups. Through the quantitative and qualitative study of activities in 155 WhatsApp groups, the objective of the research was to explore not just the nature of the messages but also the mechanisms through which those messages are identified, perceived, and relayed by users. The study forms part of the Media Sector Development Activity (MSDA) implemented by Internews in the DRC (see Related Summaries, below).

As explained in the report, the study focused on WhatsApp, as 54% of Congolese people surveyed by the market research company, Target, declared that people use the internet there primarily to connect to WhatsApp, making the platform a central tool to share information. The study is based on a sample of 520 messages identified in 155 WhatsApp groups of journalists and community leaders in Kinshasa, North Kivu, and Haut Katanga during a 15-day period in the December 2023 electoral campaign. The study also included a short questionnaire that was administered to all users of messages of misinformation/disinformation and hate speech identified during the monitoring of the WhatsApp groups, as well as semi-structured focus group discussions with participants of the groups to supplement and triangulate the information collected during the monitoring of WhatsApp groups.

The report starts with an analysis of the overall landscape of mis/disinformation in the DRC, focusing on the actors who produce false, misleading, or harmful content, as well as the political and security context in which these practices take place. The factors contributing to information pollution in the DRC include, for example, weak media institutions, limited access to education, polarisation and political tensions, increased use of social media, and external influence by foreign stakeholders. The three main groups described in the report that are behind information pollution in the DRC are: private political entrepreneurs, digital militias, and private digital companies, such as the Archimedes Group. The report also looks at the practices of information pollution and how information influence is used to conquer public opinion in the electoral context and in military operations in the province of North Kivu. Examples include the creation of fake online identities and media infiltration, which involves the infiltration of agents or propaganda officers within the mainstream media to alter or steer media coverage of events, thus favouring a certain point of view. 

The next section in the report offers an exploration of the findings related to the nature and characteristics of mis/disinformation messages, as well as hate speech disseminated in the WhatsApp groups. In brief, the study, reveals the following: 
 

• Overall, and across the monitoring period, the vast majority of messages documented (88.5%) were mis/disinformation, with a smaller proportion of hate speech.
• Urban centres are hotspots for the diffusion of both mis/disinformation and hate speech.
• Some users are encouraged by politicians or political entrepreneurs to share false information.
• The spread of mis/disinformation around the electoral campaign increased slightly in the weeks prior to the elections.
• Identity politics strongly shape the type of mis/disinformation disseminated on WhatsApp groups.
• In North Kivu, fear and insecurity created by conflict in the East is used to shape the electoral debate and spread mis/ disinformation about presidential candidates.

The study also looked at users' motivations to share mis/disinformation and hate speech on WhatsApp groups and the strategies they use to identify mis/disinformation, verify information, and correct other online users. In brief, the findings show the following:
 

• Information shared is equally obtained from official and informal sources.
• A large proportion of the messages disseminated by users on WhatsApp groups are shared from other WhatsApp groups (65.5%) and social media, including Facebook (8.5%), X, TikTok, and YouTube.
• Users are somewhat selective in joining and leaving groups that share false or harmful information.
• While users understand the value of fact-checking information before sharing it, they do not systematically do so. 
   
• In a context of mistrust of official sources of information, users deliberately share unverified information to fact-check it.
• Users who share false or harmful information are also motivated by getting reactions rather than informing.
• Users are proactive in correcting mis/ disinformation but are less willing to confront authors of hate speech.  
   

The final chapter in the report provides an analysis of the format, language, and types of content that WhatsApp group users trust more, shedding light on information preferences. In summary, the findings show that: 
 

• Users are less inclined to verify information when it is shared by people or sources they trust.
• While the source of the information matters more than its format, users tend to trust videos more than text or image-based information, which they find are easier to manipulate.
• Most mis/disinformation shared by users is in French, as it is widely understood and used across the medias.

In conclusion, the report states that "While this research sheds light on WhatsApp group members' practices relating to information sharing, it also acknowledges its limitations. The study was conducted in a non-statistically representative sample and was not able to clearly ascertain the precise extent to which users deliberately shared false information. Further research would be required to explore the motivations driving the creation and sharing of false and harmful information online, as well as the extent to which users rely on it to shape their electoral preferences and behaviors."

Click here for the 60-page French version of this report in PDF format.