The Employer Bias: A Lab-based Study on Evaluating Employees’ Self-Assessments

Abstract: In many organisational decisions, managers must rely on limited employee information, such as self-evaluations or observed performance. While ample evidence documents biased self-representations, less is known about how such signals are judged by others. This study examines how managers interpret employee performance and self-confidence. I conduct a laboratory experiment with 192 employers evaluating 576 employees, focusing on whether perceptions differ by manager gender and whether confidence biases are viewed as gender-typical. Results show substantial gender differences: female employers consistently perceive employees as about 0.2–0.3 standard deviations less confident than male employers. Providing explicit information about confidence inaccuracies activates gender stereotypes, with underconfidence strongly associated with female employees and overconfidence with male employees. These findings highlight social biases in interpreting performance and self-confidence, with implications for gender dynamics and managerial decisions in organisations that rely on employee self-evaluations.

Keywords: Perception bias, gender stereotypes, employers, hiring and promotion decisions, lab experiment