Not the Bots You Are Looking For: Patterns and Effects of Orchestrated Interventions in the U.S. and German Elections

Abstract

Zooming in on automated and semiautomated social actors created to influence public opinion on social media, we employ a novel analytic approach to identify patterns of inauthentic behavior across election campaigns on Twitter. Comparing two recent national election campaigns, the 2016 U.S. presidential election and the 2017 German federal election, we analyze patterns and effects of orchestrated intervention in political discourse on Twitter. Focusing on two main aspects of information flows—scale and range—we find that orchestrated interventions help amplify, but not diffuse, the candidates’ messages, mostly failing to reach new audiences in the process. This study adds an information diffusion perspective to a growing body of literature on computational propaganda, showing that although false amplification is quite effective in increasing the scale of information events, in most cases the information fails to reach new depths.