Mapping Ethiopia's Divided Society: Mathematical Model Explains Political Polarization

Polarization in Ethiopia, driven by ethnic and political divides, has been exacerbated by social media and out-group animosity, impacting social cohesion. This study aimed to model and analyzes the dynamics of polarization and choice-making in social networks, focusing on in-group and out-group influences, and proposes interventions to reduce affective polarization. A computational model simulated 100 agents in a social network (Erdős-Rényi graph, 10% connectivity) over 50 time steps, with choices (e.g., masking/supporting) influenced by in-group approval (0.7) and out-group opposition (0.5). Qualitative data included semi-structured interviews with 30 participants from Oromia, Amhara, Tigray, and Addis Ababa, and focus group discussions (FGDs) with 32 participants (four groups: youth, women, academics, civil society). Statistical analysis (chi-square tests) and thematic analysis validated findings. The model showed polarization; with Party A’s masking/support proportion dropping to 0.1 and Party B’s rising to 0.9, driven by group dynamics. Interviews confirmed in-group loyalty (70%, χ²(1, N=30) = 5.14, p < .05) and out-group opposition (60%) as key drivers, while FGDs highlighted social media’s role (75% youth) and supported dialogue (81.25% across groups, χ²(3, N=32) = 8.12, p < .05). Structural factors like ethnic federalism were noted as limitations. In-group and out-group dynamics significantly drive polarization in Ethiopia, but interventions like dialogue can mitigate divides. Implement inter-ethnic dialogue and social media regulation to reduce polarization.