DP11724 Do women respond less to performance pay? Building evidence from multiple experiments
|Author(s):||Oriana Bandiera, Greg Fischer, Andrea Prat, Erina Ytsma|
|Publication Date:||December 2016|
|Keyword(s):||econometrics, Gender, meta-analysis, wage differentials|
|JEL(s):||C11, J16, J31|
|Programme Areas:||Labour Economics, Industrial Organization|
|Link to this Page:||www.cepr.org/active/publications/discussion_papers/dp.php?dpno=11724|
Performance pay increases productivity but also earnings inequality. Can it widen the gender gap because women are less responsive? We provide answers by aggregating evidence from existing experiments on performance pay that have both male and female subjects, regardless of whether they test for gender differences. We develop a Bayesian hierarchical model (BHM) that allows us to estimate both the average effect and the heterogeneity across studies. We find that the gender response difference is close to zero and heterogeneity across studies is small. We also find that the average effect of performance pay is positive, increasing output by 0.28 standard deviations. The data are thus strongly supportive of agency theory for men and women alike.