This paper asks whether consumption of politically charged movies produced in democracies affects anti-government demonstrations in autocratic countries. Using a newly compiled dataset of political protest films from 2000–2018, the authors examine whether widely watched imported protest movies are associated with changes in protest levels and test mechanisms such as imitation, coordination, and diffusion of democratic values.
Novel cross-country movie dataset Instrumental-variable estimation Placebo and robustness checks
Data and methods
The authors build a novel dataset of political protest movies and their viewership between 2000 and 2018 and link these measures to protest activity in autocratic countries. The abstract reports the use of instrumental-variables estimation and a battery of placebo tests; specific instruments, control variables, sample composition and exact outcome measures are described in the full paper rather than the abstract.
Main findings
According to the abstract, widely watched imported protest movies are positively associated with anti-government demonstrations in autocracies. The authors report that this association survives instrumental-variable estimation and several placebo checks. The abstract also notes null results: little-watched protest films and widely watched family-romance movies do not affect protests, and widely watched imported protest movies do not predict prior protests in autocracies or later protests in democracies.
- Widely watched imported protest movies: reported positive association with anti-government demonstrations.
- Little-watched protest films and widely watched family-romance films: reported no effect.
- No evidence reported that the films predict protests before exposure in autocracies or subsequent protests in democracies.

