Suchergebnisse

Online political astroturfing—hidden information campaigns in which a political actor mimics genuine citizen behavior by incentivizing agents to spread information online—has become prevalent on social media. Such inauthentic information campaigns threaten to undermine the Internet's promise to more equitable participation in public debates. We argue that the logic of social behavior within the campaign bureaucracy and principal–agent problems lead to detectable activity patterns among the campaign's social media accounts. Our analysis uses a network-based methodology to identify such coordination patterns in all campaigns contained in the largest publicly available database on astroturfing published by Twitter. On average, 74% of the involved accounts in each campaign engaged in a simple form of coordination that we call co-tweeting and co-retweeting. Comparing the astroturfing accounts to various systematically constructed comparison samples, we show that the same behavior is negligible among the accounts of regular users that the campaigns try to mimic. As its main substantive contribution, the paper demonstrates that online political astroturfing consistently leaves similar traces of coordination, even across diverse political and country contexts and different time periods. The presented methodology is a reliable first step for detecting astroturfing campaigns.

Online political astroturfing-hidden information campaigns in which a political actor mimics genuine citizen behavior by incentivizing agents to spread information online-has become prevalent on social media. Such inauthentic information campaigns threaten to undermine the Internet's promise to more equitable participation in public debates. We argue that the logic of social behavior within the campaign bureaucracy and principal-agent problems lead to detectable activity patterns among the campaign's social media accounts. Our analysis uses a network-based methodology to identify such coordination patterns in all campaigns contained in the largest publicly available database on astroturfing published by Twitter. On average, 74% of the involved accounts in each campaign engaged in a simple form of coordination that we call co-tweeting and co-retweeting. Comparing the astroturfing accounts to various systematically constructed comparison samples, we show that the same behavior is negligible among the accounts of regular users that the campaigns try to mimic. As its main substantive contribution, the paper demonstrates that online political astroturfing consistently leaves similar traces of coordination, even across diverse political and country contexts and different time periods. The presented methodology is a reliable first step for detecting astroturfing campaigns.

Online political astroturfing—hidden information campaigns in which a political actor mimics genuine citizen behavior by incentivizing agents to spread information online—has become prevalent on social media. Such inauthentic information campaigns threaten to undermine the Internet's promise to more equitable participation in public debates. We argue that the logic of social behavior within the campaign bureaucracy and principal–agent problems lead to detectable activity patterns among the campaign's social media accounts. Our analysis uses a network-based methodology to identify such coordination patterns in all campaigns contained in the largest publicly available database on astroturfing published by Twitter. On average, 74% of the involved accounts in each campaign engaged in a simple form of coordination that we call co-tweeting and co-retweeting. Comparing the astroturfing accounts to various systematically constructed comparison samples, we show that the same behavior is negligible among the accounts of regular users that the campaigns try to mimic. As its main substantive contribution, the paper demonstrates that online political astroturfing consistently leaves similar traces of coordination, even across diverse political and country contexts and different time periods. The presented methodology is a reliable first step for detecting astroturfing campaigns.