Dynamic Network Influence: The Art of Strategic Messaging
Prof. Wei Li
Associate Professor of Economics
University of British Columbia
Strategic influencers send costly messages over time to persuade agents in a network.
Each influencer maximizes her total discounted payoff, which decreases in the agents’
opinion deviations from her agenda. Agents update opinions by taking weighted
averages of neighbors’ opinions and influencers’ messages. In a single influencer
benchmark, early messages are more extreme to hasten agenda adoption, followed
by moderate messages to align opinions with her agenda. The influencer is worse
off in networks where the weights agents attach to their own opinions are farther
apart from the weights agents attach to their neighbors’ opinions. With multiple
influencers, if they have the same impact on agents, consensus emerges in any network
as the average agenda of the influencers. If they have different impacts, in symmetric
networks, consensus still emerges but it favors the agenda of the stronger influencer. In
asymmetric networks, influencers with different impacts often target different subgroups
with differing intensity, generating perpetual disagreement and polarization.