Nigeria’s General Elections: The Online Information & Media Landscape

Fake news is not a new concept in Nigeria, but the exponential growth in the use and availability of mobile Internet, as well as the sheer amount of information accessible in the age of digital media, has made the task of filtering out false information far more difficult. To gain a deeper understanding of the problem, the Centre for Democracy and Development have conducted a study of the online information landscape in order to paint a clearer picture of how information flows, and understand the precise nature of disinformation (information shared with the intent to mislead), misinformation (information not necessarily shared with intent), hateful or dangerous speech, propaganda and other harmful forms of content in Nigeria.

 

Our analysis focused primarily on Facebook and Twitter, with some data collection and fact checking through WhatsApp. The initial results of the analysis revealed a number of interesting issues:

 

  • The Facebook data, collected from December 28, 2018 until January 27, 2019, reflected a broader, international perspective of the electoral process, and CDD’s analysis divided the content into three principle categories: politics, news and culture oriented clusters.
  • CDD collected tweets and related metadata for accounts of major politicians, parties, media and political organizations as well as hashtags related to the campaign, candidates and the general political process.
  • The Twitter dataset contained over 1 million tweets collected from December 31 2018 to January 30, 2019, revealed a dense and interconnected landscape with an above-average prevalence of automated accounts – roughly 19.5% of collected accounts show signs of automation, which suggests a high level of bot activity around the hashtags, accounts and subjects collected.
  • One observable trend showed that accounts with bot like tendencies promoted messages concerning Biafra and calls to boycott the elections.
  • False news sites and Facebook groups are sharing false or questionable content across platforms, in some cases to hundreds of thousands of users.
  • Examples of content found include false stories about the supposed cloning of President Buhari, a false invasion of Nigeria and Ghana by North Korea, as well as satirical items that users are often sharing as true stories.

 

INITIAL CONCLUSIONS

 

It is important to note that these are preliminary findings, with time and further analysis, CDD will continue to build a fuller picture and release more comprehensive analysis, and are continuing to collect data on all platforms. In terms of day-to-day communication, WhatsApp is by far the most common information sharing platform today in Nigeria according to a report by WeAreSocial, a social analytics firm, and this penetration is growing. The nature of WhatsApp’s closed, end to end encrypted platform means that it is difficult to capture data from WhatsApp, and researchers can only collect information, messages and content manually for analysis. However, this manually collected data from WhatsApp combined with the automated collections from Facebook and Twitter will help to build a better picture of the online information landscape in Nigeria for the election and beyond.

 

This study is conducted in collaboration with the National Democratic Institute.