The 4th Workshop of Gender and Economics was a great opportunity to share the results of the experimental study on the role of averaged and individual information of workers performance on the decision on remuneration setting. Magda presented updated analysis and received great comments and suggestions.

The concept of statistical discrimination, originally introduced by Phelps (1972) and Arrow (1973), suggests that employers use group-level statistics as a proxy for individual productivity while missing the individual information on particular worker. This study examines whether providing individual performance information can mitigate such biases and lead to fairer wage distributions. The research employed a controlled laboratory experiment with university students acting as managers responsible for allocating wages.

We find that the decisions are made mostly considering fair distribution of remunerations (in the context of actual contribution of workers in the team joint result), however there is a difference between female and male-typed tasks (solve equation versus emotion recognition tasks). We find that the decisions are more likely to be driven by statistical information in female-typed task, and the wage setting decision when the information on individual performance is revealed is additionally correlated with the gender norms of the manger, and there is a premium for woman if her score was higher than their male team member.