The Digital Authoritarian: Everyday behavioral patterns collected with smartphones predict authoritarianism

Authoritarianism has reemerged around the world. Here, we show that digital records of everyday behavior can be used to predict individuals’ authoritarian tendencies. More than 280 million phone events (measuring app usage, music listening, keyboard typing, calling and texting, unlocking and locking, and GPS location) were digitally recorded from the smartphones of 669 volunteers for three to six months. We then derived behavioral variables from the phone events following psychological theories of authoritarianism. These behavioral variables predicted self-reported authoritarianism levels in random forest models (ρmd = .30) significantly better than and independently of demographic characteristics. The behavioral patterns that most strongly predicted authoritarianism reflected authoritarians’ limited exposure to unknown people, places and cultures (generally) and greater Facebook use (specifically). Overall, we provide a behavioral portrait of authoritarianism in everyday life that is updated for the digital era. Our findings have implications for interventions that combat authoritarianism as democracies erode globally.