Skin Tone Penalties: Bottom-up Discrimination in Football

This paper investigates colorism, racial discrimination based on skin color, in men’s football. Firstly, using machine learning algorithms, we extract players’ skin tones from online headshots to examine their impact on fan-based ratings and valuations. We find evidence of a skin tone penalty, where darker-skinned players face lower fan-driven market values and ratings. Secondly, using algorithm-based ratings and employing a Difference-in-Discontinuities design with geolocated penalty kicks data, we show that lighter-skinned players enjoy a premium higher by 1.25 standard deviations than their darker-skinned peers, conditional on scoring a penalty. Additionally, we find evidence that non-native players with dark skin face a double penalty. Leveraging the COVID-19 pandemic as a natural experiment, we highlight the role of fans’ stadium attendance in algorithm-based results. The findings underscore direct skin tone discrimination in football and highlight fans’ role in perpetuating algorithmic bias.