| Abstract: |
The movie industry provides a unique setting to analyze consumer-driven gender
biases, as it allows clear identification of how the gender of leading actors,
directors and producers influences movie performance outcomes. Using a
hand-collected dataset of over 5, 000 globally produced movies from 1998 to
2008, we document a distinct non-linear relationship between female
representation in leading roles and audience ratings. Specifically, ratings
initially decline significantly as the number of female leads increases,
reaching a turning point at approximately two female leads, beyond which
ratings stabilize or slightly improve (convex pattern). This negative impact
on audience ratings is primarily driven by male viewers, whose proportional
presence diminishes as female representation grows. In contrast, professional
film awards exhibit an opposite, concave pattern, peaking significantly at two
female leads. Employing a Heckman- like selection test, we further reveal that
audience gender biases persist even after accounting for the selective
attrition of male viewers from movies featuring female leads. |