Influence Competition in Networks: a scientific talk by Sudipta Sarangi

Date: 04 June 2026, 10:00-11:00 

Location: Corvinus University of Budapest, Building E, Room E.1.118.2 

Language of the event: English 

Abstract: 

We study strategic influence competition between two actors: an influencer, who can shape interaction patterns and directly influence agents, and a counter-influencer, who can also influence agents. Both actions are costly, and outcomes are determined by a voting process in which agents’ final opinions reflect both direct influence and peer effects. 

We characterise equilibrium network structures and influence strategies as functions of relative influence technologies, costs, and voting rules. When the influencer has superior technology, equilibrium networks tend to be empty or highly centralised, amplifying influence through peer reinforcement while reducing persuasion costs. When the counter-influencer is stronger, networks become more symmetric and decentralised, limiting targeted influence. 

Repeated interactions before voting significantly reduce the influencer’s resource requirements, as opinion dynamics allow initially influenced agents to persuade others over time. Relaxing unanimity to majority voting further lowers costs and reduces the need for dense network structures. Finally, we show that these results are robust to limited unplanned interactions but break down when such interactions become sufficiently frequent.