Due to precautions NYU is taking to prevent the spread of coronavirus, this event, and all other events for the remainder of the semester, have been cancelled. Please check back soon for information regarding rescheduling!
In this talk, Dr. Sobolev will explore the behavior and impact of several hundred “trolls” — paid supporters of Vladimir Putin’s regime in Russia who were allegedly employed in late 2014 and early 2015 to leave pro-government comments on the popular social media platform LiveJournal. First, he will devise a classification method of the possible objectives that would motivate governments to employ Internet trolls, the strategies trolls use to achieve these objectives, and these strategies’ observable implications. Second, combining text analysis with modern approaches in causal inference, he will develop a method to measure the natural evolution of online discussions so as to estimate the causal effect of troll interventions. Using a modified regression discontinuity approach and a set of partially testable assumptions about the timing of such interventions, he will discover that Russian troll activity was more successful in diverting online discussions away from politically charged topics than in promoting a pro-government agenda. Moreover, while trolls succeeded in diverting discussions away from purely political topics, their interference apparently had no effect when the topic under discussion was the national economy. Those social media users who were discussing poor economic growth, unemployment, or price inflation seemed not to be responsive to troll interventions.
Anton Sobolev is a postdoctoral associate in the Leitner Program on Political Economy at Yale University. His research studies classic questions of comparative politics using text analysis, machine learning, and causal inference. His recent projects focus on mass protest and political control in autocracies. More broadly, he studies communication technology’s effects on society and political behavior. Anton’s dissertation explores the information tools employed by modern non-democratic leaders to maintain political control and their citizens’ probable responses to the strategies these tools make possible. It focuses primarily on government hiring of agents to impersonate ordinary citizens and engage online and offline with members of the political opposition. His work has been published in World Politics, European Journal of Political Economy, Post-Soviet Affairs, Europe-Asia Studies, and Problems of Post-Communism. Prior to his appointment at Yale, Anton received his Ph.D. in Political Science and M.Sc. in Statistics from University of California, Los Angeles.
This event is part of the NYU Department of Politics Comparative Politics Speaker Series.