Social Media, Misinformation and Social Cohesion in Plural Societies: A Theoretical Account of Algorithmic Amplification, Ethnic Mobilisation and Trust in Africa

Abstract

Africa's plural societies manage ethnic, religious and regional difference through institutions, norms and everyday relationships that social media increasingly bypass, delivering content selected by engagement-based ranking and by networks of in-group trust. Research on misinformation, hate speech and political communication in Africa is growing but fragmented, and its focus on falsity obscures the harm done by divisive content that need not be false. This theoretical paper addresses that gap. Its objective is to explain how social media affect cohesion in plural societies, to specify when the outcome is polarisation and violence rather than noise, and to identify the factors and policies that moderate harm. Methodologically, the study is an integrative, theory-building review. The population was peer-reviewed journal articles published between 2023 and 2026; a purposive sample of 45 articles was retained after each was verified against its publisher record and screened for retractions and corrections. Findings were coded against provisional mechanisms, compared across countries and platforms, and synthesised abductively. The analysis yields an amplification–mobilisation–erosion model. Amplification describes what ranking systems and messaging networks select; mobilisation, how political entrepreneurs such as campaigns, paid influencers and diaspora networks exploit that selection; and erosion, what happens to inter-group and institutional trust. The paper argues that the principal harm is the systematic over-supply of identity-threat content to populations whose cohesion depends on its under-supply. Seven propositions specify the roles of engagement ranking, closed networks, high-stakes political moments and historical grievance, and of three resilience factors: cross-cutting ties and shared identity, trusted local intermediaries, and institutional credibility. Nigeria is the anchoring case. The paper recommends integrating information integrity into election security, co-regulation rather than platform bans, preference-based ranking and local-language moderation, support for trusted intermediaries and peer correction, and disclosure rules for paid political influence.