日本語

Graduate School of Economics and Management, Tohoku University

This study investigates the outcome bias by exploiting natural experimental situations, where outcomes are likely to be exogenously determined, from Major League Baseball. Specifically, I extract hits near the wall where a slight distance affects the outcome and test the difference in evaluation between random successes and failures. The estimation results provide evidence of the persistent outcome bias and its heterogeneity based on player race and manager experience. Managers are more likely to continually allow players who happen to be successful to play in subsequent games. Additionally, they tend to over-evaluate the random success of white players compared to racial minorities, although their decisions become less biased with increased experience. The racial disparities in evaluation may further complicate the problems posed by the outcome bias; however, the improvement in decisionmaking with experience suggests the potential solution.