Title: Exposing the Obscured Influence of State-Controlled Media: A Causal Estimation of Influence Between Media Outlets Via Quotation Propagation

Abstract: This study quantifies influence between media outlets by applying a novel methodology that uses causal effect estimation on networks and transformer language models. We demonstrate the obscured influence of state-controlled outlets over other outlets, regardless of orientation, by analyzing a large dataset of quotations from over 100 thousand articles published by the most prominent European and Russian traditional media outlets, appearing between May 2018 and October 2019. The analysis maps out the network structure of influence with news wire services serving as prominent bridges that connect outlets in different geo-political spheres. Overall, this approach demonstrates capabilities to identify and quantify the channels of influence in intermedia agenda setting over specific topics.