On Statistical Discrimination as a Failure of Social Learning: A Multi-Armed Bandit Approach
Published in Management Science, 2024
This paper investigates statistical discrimination through the framework of multi-armed bandit theory. We show that statistical discrimination can emerge as a rational response to uncertainty in social learning environments, providing theoretical insights into the mechanisms underlying discriminatory behavior in algorithmic and human decision-making contexts.
Recommended citation: Komiyama, J., & Noda, S. (2024). “On Statistical Discrimination as a Failure of Social Learning: A Multi-Armed Bandit Approach.” Management Science. To appear.
Recommended citation: Komiyama, J., & Noda, S. (2024). "On Statistical Discrimination as a Failure of Social Learning: A Multi-Armed Bandit Approach." Management Science. To appear.
Download Paper